The Arabidopsis PsbO2 protein regulates dephosphorylation and turnover of the photosystem II reaction centre D1 protein

The Arabidopsis PsbO2 protein regulates dephosphorylationand turnover of the photosystem II reaction centre D1 protein

Bjorn Lundin1, Maria Hansson1, Benoıt Schoefs2, Alexander V. Vener1 and Cornelia Spetea1,*

1Division of Cell Biology, Linkoping University, SE-581 85 Linkoping, Sweden, and2Dynamique Vacuolaire et Reponses aux Stress de l’Environnement, UMR Plante-Microbe-Environnement CNRS 5184/INRA

1088/UB, Universite de Bourgogne, BP 47870, 21078 Dijon cedex, France

Received 3 August 2006; revised 4 October 2006; accepted 10 October 2006.*For correspondence (fax þ46 13 224314; e-mail [email protected]).

Summary

The extrinsic photosystem II (PSII) protein of 33 kDa (PsbO), which stabilizes the water-oxidizing complex, is

represented in Arabidopsis thaliana (Arabidopsis) by two isoforms. Two T-DNA insertion mutant lines deficient

in either the PsbO1 or the PsbO2 protein were retarded in growth in comparison with the wild type, while

differing from each other phenotypically. Both PsbO proteins were able to support the oxygen evolution

activity of PSII, although PsbO2 was less efficient than PsbO1 under photoinhibitory conditions. Prolonged

high light stress led to reduced growth and fitness of the mutant lacking PsbO2 as compared with the wild type

and the mutant lacking PsbO1. During a short period of treatment of detached leaves or isolated thylakoids at

high light levels, inactivation of PSII electron transport in the PsbO2-deficient mutant was slowed down, and

the subsequent degradation of the D1 protein was totally inhibited. The steady-state levels of in vivo

phosphorylation of the PSII reaction centre proteins D1 and D2 were specifically reduced in the mutant

containing only PsbO2, in comparison with the mutant containing only PsbO1 or with wild-type plants.

Phosphorylation of PSII proteins in vitro proceeded similarly in thylakoid membranes from both mutants and

wild-type plants. However, dephosphorylation of the D1 protein occurred much faster in the thylakoids

containing only PsbO2. We conclude that the function of PsbO1 in Arabidopsis is mostly in support of PSII

activity, whereas the interaction of PsbO2 with PSII regulates the turnover of the D1 protein, increasing its

accessibility to the phosphatases and proteases involved in its degradation.

Keywords: Arabidopsis, photosystem II, PsbO, oxygen evolution, D1 protein degradation, high light stress.

Introduction

The PsbO protein is a 33 kDa membrane-extrinsic subunit of

the water-oxidizing photosystem II (PSII) complex, exposed

at the lumenal side of the thylakoid membrane. Its primary

sequence is moderately conserved in all known oxygenic

photosynthetic organisms. The first three-dimensional

structure of the PsbO protein from the cyanobacterium

Thermosynechococcus elongatus was elucidated by Zouni

et al. (2001). Ferreira et al. (2004) assigned the primary

sequence of the cyanobacterial PsbO bound to PSII, which

allowed the construction of homologous models for the

PsbO protein in plant PSII (De Las Rivas and Barber, 2004).

The most conserved regions of interaction of PsbO with

intrinsic and extrinsic components of the PSII complex

were identified, and were suggested to be involved in sta-

bilizing the water-oxidizing complex as well as in regulatory

processes (De Las Rivas and Barber, 2004). Despite its ubi-

quitous presence in oxygenic photosynthetic organisms, the

PsbO protein is not obligatory for the process of water oxi-

dation. A mutant of the cyanobacterium Synechocystis

PCC6803 with a deleted psbO gene was able to grow pho-

toautotrophically, and exhibit almost normal PSII function

(Burnap and Sherman, 1991). By contrast, the green alga

Chlamydomonas reinhardtii mutant lacking PsbO could not

grow photoautrophically, and showed high instability of PSII

(Mayfield et al., 1987). Thus, the functional role of the PsbO

protein in PSII may differ among the oxygenic photosyn-

thetic organisms.

Sequencing of the Arabidopsis thaliana genome (AGI,

2000) revealed the existence of two psbO genes, At5g66570

and At3g50820, encoding for the PsbO1 and PsbO2 proteins,

528 ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd

The Plant Journal (2007) 49, 528–539 doi: 10.1111/j.1365-313X.2006.02976.x

respectively, which differ by only 10 residues in their mature

form. Proteomic analyses of the thylakoid lumen from

Arabidopsis confirmed the existence of the two PsbO

proteins with slightly different electrophoretic mobilities

(pI and molecular weight) (Kieselbach et al., 2000; Peltier

et al., 2002; Schubert et al., 2002). Both genes are expressed

in wild-type Arabidopsis plants, while the PsbO1 isoform is

four- to fivefold more abundant than PsbO2 (Gomez et al.,

2003; Murakami et al., 2002). An Arabidopsis mutant with an

impaired psbO1 gene showed retarded growth, although the

level of the PsbO2 protein was significantly increased

(Murakami et al., 2002, 2005). More recently, an RNA inter-

ference approach was used for simultaneous suppression of

the expression of both psbO genes, confirming the import-

ance of PsbOs for stability of PSII and photoautotrophy in

Arabidopsis (Yi et al., 2005). No mutants lacking only the

PsbO2 isoform have so far been analyzed, although under-

standing the function of this isoform can be crucial for the

regulation of photosynthesis. The low levels of PsbO2 in the

wild-type plants may indicate a regulatory rather than a

structural role for this isoform in PSII function. Indeed, in the

present work we report an involvement of PsbO2 in the

regulation of light-induced turnover of the PSII reaction

centre D1 protein.

The PSII complex is the main target for light-induced

inactivation and damage. It undergoes a continuous repair

cycle, which includes degradation of the damaged D1

protein and its substitution with the newly synthesized

polypeptide (for a recent review see Aro et al., 2005). The

light-induced degradation of the D1 protein in higher plants

is a multistep energy-dependent proteolysis (Spetea et al.,

1999), which most likely involves DegP2 and FtsH proteases

(Haußuhl et al., 2001; Lindahl et al., 2000). The PSII repair

cycle is also regulated by reversible phosphorylation of

several PSII core subunits (for recent reviews see Aro et al.,

2005; Vener, 2006). In particular, dephosphorylation of the D1

protein is a prerequisite for its degradation in vivo (Pursihe-

imo et al., 1998; Rintamaki et al., 1996). Nevertheless, a

recent report suggests that phosphorylation of PSII may not

be essential for its repair (Bonardi et al., 2005). Phosphory-

lation of PSII polypeptides is regulated by the redox state of

the thylakoid membrane (Vener et al., 1998) as well as by the

endogenous circadian rhythm (Booij-James et al., 2002), and

requires at least two thylakoid-associated protein kinases,

TAK1 (Snyders and Kohorn, 2001) and STN8 (Vainonen et al.,

2005). The thylakoid membrane also contains a PSII-specific

protein phosphatase, regulated by the thylakoid lumenal

protein TLP40 in spinach (Vener et al., 1999). This phospha-

tase is responsible for the rapid dephosphorylation of PSII

subunits, including D1, at elevated temperatures (Rokka

et al., 2000). The gene encoding for the phosphatase cata-

lyzing dephosphorylation of D1 is still unknown.

A role for PsbO in regulating D1 dephosphorylation and/or

degradation in plants has not been previously demonstra-

ted. Nevertheless, a ‘chaperone’ function for PsbO in

preventing D1 aggregation has been suggested from in

vitro experiments in spinach (Yamamoto et al., 1998). More

recently, it has been shown that spinach PsbO binds

radioactive GTP (Spetea et al., 2004), with possible rele-

vance for the GTP-dependent degradation and turnover of

the D1 protein under high- and low-light conditions (Spetea

et al., 1999, 2000).

In this study we investigate the separate functions of

PsbO1 and PsbO2 in Arabidopsis using two T-DNA insertion

mutant lines (Alonso et al., 2003), named psbo1 and psbo2.

The psbo1 mutant had lower levels of PSII proteins and

activity, and it was more sensitive than wild-type (wt) plants

to short high-light stress. Despite higher levels of PSII

proteins in the psbo2 mutant, reduced plant growth and

fitness were observed under prolonged high-light stress.

Furthermore, we found differences between the two mutants

in terms of in vivo steady-state levels of phosphorylation of

the D1 protein, the ability for in vitro dephosphorylation and

degradation of D1, pointing to a specific role for PsbO2 in

regulatory events associated with D1 turnover.

Results

Characterization of homozygous psbo1 and psbo2 Arabid-

opsis mutants

Two Arabidopsis mutant lines with T-DNA insertions in each

of the two psbO genes were obtained from the SALK Insti-

tute (Alonso et al., 2003). In this study, we denote the mutant

lacking PsbO1 as psbo1 (SALK 093396) and the mutant

lacking PsbO2 as psbo2 (SALK 024720) (Figure 1a). In the

homozygous mutants, the presence of a T-DNA insertion in

each of the psbO genes was confirmed by PCR of genomic

DNA with two sets of primers: a gene-specific and a T-DNA-

specific primer (Figure 1b, lanes 1 and 3), and with only

gene-specific primers (Figure 1b, lanes 2 and 4). The pres-

ence or absence of psbO transcripts was examined by RT-

PCR analysis of mRNA from the mutants and wt plants. The

18S rRNA, used as a positive control, was detected in both

mutants and wt plants (Figure 1c, lane 1). As expected, the

psbO1 transcript was present in the psbo1 homozygous

mutant (Figure 1c, lane 2), since the T-DNA insertion was

located in the 5¢-untranslated region (5¢-UTR), which usually

contains the translation signal. The psbO2 transcript was

present in the wt plants and in the psbo1 mutant, but not in

the psbo2 homozygous mutant (Figure 1c, lane 3).

The absence of the corresponding PsbO protein in each of

the mutants was confirmed by Western blotting of thylakoid

membranes with an antibody produced against an N-

terminal peptide, which is identical in the two isoforms

(Figure 1d). Both PsbO1 and PsbO2 could be detected in

thylakoids isolated from wt plants, in a ratio of about 4:1. The

expression level of the PsbO2 protein was increased three-

Role of Arabidopsis PsbO2 protein 529

ª 2007 The AuthorsJournal compilation ª 2007 Blackwell Publishing Ltd, The Plant Journal, (2007), 49, 528–539

fold in psbo1, reaching 75% of the total amount of PsbO in

wt, in line with previous observations (Murakami et al., 2002,

2005). The expression level of PsbO1 in psbo2 was 125% of

the total PsbO in wt plants (Figure 1d).

The phenotypes of both mutants differed from each other

and from that of wt plants (Figure 2). The psbo1 mutant

grew significantly slower than the wt, as was reported earlier

(Murakami et al., 2002), with the leaves pale green and

parallel to the surface. In contrast, the psbo2 mutant grew

faster than psbo1 but slower than the wt, exhibiting different

leaf characteristics: dark green, more elongated and with

bent edges. The distinct phenotypes of the two mutants

were observed for several generations of homozygous

plants, and were therefore attributed to the absence of the

corresponding PsbO isoform.

Levels of other thylakoid proteins and PSII activity

Western blot analyses of isolated thylakoid membranes

(Figure 3a) indicated an overall reduction of PSII proteins

(CP43, D1, D2, Lhcb2, PsbP) in the psbo1 mutant to about

75% of the wt levels, and an increase to about 138% in the

psbo2 mutant. These values correspond to the determined

associated levels of PsbO (Figure 1d), indicating that the

total amount of PsbO per PSII complex is the same in wt and

mutants. Silver stained SDS gels containing light-harvesting

complex II (LHCII)–PSII supercomplexes isolated from mu-

tants revealed similar levels of PSII proteins as in wt prepa-

rations (data not shown).

Unlike the PSII proteins, the amount of the PSI reaction

centre heterodimer PsaA/B (Figure 3a) increased to about

110% in the thylakoids isolated from the psbo1 mutant, and

decreased to about 91% in the psbo2 mutant, i.e. not

significantly different from the wt (100%). The b-subunit of

CF1 (ATP synthase) was present in nearly equal amounts in

both mutants and wt plants (Figure 3a). Using electron

paramagnetic resonance (EPR) measurements, a ratio of 0.7

between PSI and PSII has recently been determined for

Arabidopsis wt plants (Suorsa et al., 2006). Based on this

ratio and the relative levels of PSI and PSII determined by

Western blotting in this work, corresponding ratios of 1.0

and 0.4 were calculated for psbo1 and psbo2, respectively.

The ratio of chlorophyll (Chl) a to Chl b determined in

thylakoids and LHCII–PSII supercomplexes was not signifi-

cantly different in wt and mutants (Table 1). Nevertheless,

the clearly distinct thylakoid contents of leaves in the

mutants as compared to the wt may explain the pale and

dark green color of the psbo1 and psbo2 mutant, respect-

ively (Figure 2). The different contents of LHCII–PSII super-

complex in leaves (Table 1) support the results of the

Western blot (Figure 3a). The fact that the psbo1 and psbo2

mutants, respectively, contain less and more PSII than the

wt, in amounts regulated by the level of total PsbO,

emphasizes the importance of PsbOs for the stability of PSII

(a)

(b) (c)

(d)

Figure 1. Screening for homozygous Arabidopsis psbo1 and psbo2 mutants.

(a) Localization of the T-DNA insertion in the psbO1 and psbO2 genes, and of

the primers used for PCR analysis (arrows numbered 1 to 5) and for RT-PCR

analysis (arrows numbered 6 to 9).

(b) Agarose gel stained with ethidium bromide of PCR products obtained from

genomic DNA with the following sets of primers (numbered as in a): 5 and 2

(lane 1), 1 and 2 (lane 2), 5 and 4 (lane 3), 3 and 4 (lane 4). As control, wild-type

(wt) genomic DNA was analyzed with the same sets of primers. The migration

of the DNA size markers is shown.

(c) Agarose gel stained with ethidium bromide of RT-PCR products obtained

with the following sets of cDNA-specific primers (numbered as in a): 18S

rRNA-specific primers (lane 1), 6 and 7 (lane 2), 8 and 9 (lane 3).

(d) Western blot with anti-PsbO antibody of thylakoids isolated from wild-type

(wt), psbo1 and psbo2 mutants. The positions of the PsbO1 and PsbO2

proteins are indicated.

wt psbo 2psbo 1

Figure 2. Phenotype of wild-type (wt) plants and psbo mutants, grown

hydroponically under optimal light conditions (120 lmol photons m)2 sec)1).

530 Bjorn Lundin et al.


and photoautotrophy in Arabidopsis, in line with the report

by Yi et al. (2005). The morphological change in leaves and

the higher PSII level in the psbo2 mutant are intriguing, and

must be the result of a mechanism of adaptation to the

absence of PsbO2.

To analyze the photosynthetic performance of the psbo

mutants, we measured the oxygen evolution activities of

isolated thylakoid membranes and LHCII–PSII supercom-

plexes as well as the maximum quantum yield for PSII

photochemistry of intact leaves. As shown in Figure 3(b), the

oxygen evolution activities determined in thylakoids isola-

ted from the psbo1 and psbo2 mutants were lower (70%) and

higher (145%), respectively, compared with the wt values,

due to the different PSII content in wt and mutants. The

corresponding activities in isolated LHCII–PSII supercom-

plexes were not significantly different in wt and mutants

(Figure 3b). Our data indicate that PsbO2 can substitute for

PsbO1 in stabilizing the water-splitting complex, and that its

absence does not affect the activity of PSII supported by the

PsbO1 protein.

The maximum quantum yield for PSII photochemistry (Fv/

Fm; see Experimental procedures) was reduced in psbo1 to

75% of the wt values (Table 1), mainly due to the high F0

value (data not shown), in line with previous observations

(Murakami et al., 2002, 2005). The psbo2 mutant presented

similar fluorescence characteristics as the wt plants, and

thus the same Fv/Fm was obtained (Table 1). To assess and

compare the light saturation of photosynthetic performance

in the mutants and wt plants, we applied a rapid light curve

approach in light-adapted detached leaves. This analysis

revealed that all plants reached saturation of the electron

transport rate (ETR) in the range 440–600 lmol pho-

tons m)2 sec)1 (Figure 3c). The psbo2 mutant showed high-

er rates than the wt plants. Nevertheless, the light curve

response for psbo1 showed much lower saturation levels of

ETR, explained by the higher value of F0 (data not shown).

Phenotypic response to high-light treatment

Wild-type plants and the two psbo mutants were grown at

optimal light intensity (GL, 120 lmol photons m)2 sec)1)

(a)

(b)

(c)

Figure 3. Analysis of PSII protein composition and PSII activity in wild-type

(wt) plants and psbo mutants.

(a) Representative Western blots of isolated thylakoids with the anti-CF1 b-

subunit of ATP synthase, PsaA/B, CP43, D1, D2, Lhcb2, PsbP and PsbQ

antibodies. The loaded amount of Chl (lg) is indicated above each lane.

(b) Oxygen evolution activities in isolated thylakoids and LHCII–PSII super-

complexes, measured as described in Experimental procedures. The standard

deviations were calculated for data obtained from seven to ten independent

preparations.

(c) Plot of relative ETR as a function of quantum flux density of the

photosynthetically active radiation (PAR). The ETR was measured via the

fluorescence parameter (as described in Experimental procedures) in light-

adapted detached leaves. The data shown represent an average of seven to

ten independent measurements.

Table 1 Chlorophyll (Chl) a/b ratio, leaf content and maximumquantum yield (Fv/Fm) in wild-type (wt) plants and psbo mutants.Chlorophyll a/b ratios and leaf contents were determined forthylakoids and LHCII–PSII supercomplexes isolated as described inExperimental procedures. Chlorophyll fluorescence was deter-mined in leaves detached from 16-h dark-adapted plants. Thevalues are means of three to five independent experiments � SD

wt psbo1 psbo2

Chl a/bThylakoids 3.2 � 0.15 2.9 � 0.15 3.2 � 0.09LHCII–PSII 3.0 � 0.08 3.0 � 0.01 2.9 � 0.06lg Chl g)1 leafThylakoids 470 � 6 320 � 7 670 � 10LHCII–PSII 32 � 19 22 � 9 83 � 11Fv/Fm

Leaf 0.8 � 0.01 0.6 � 0.01 0.8 � 0.02



until reaching full development (Figure 2a). The plants were

transferred to high-light conditions (HL, 1000 lmol pho-

tons m)2 sec)1), while maintaining the 8-h light/16-h dark

cycles. To assess and compare the photosynthetic activities

in wt and mutants, the quantum yield of PSII photochemistry

was determined by measuring Chl fluorescence during

7 days after 16-h dark and after 8-h HL periods. The pro-

portion of open PSII centers (qP) was calculated and plotted

in Figure 4(a). After the first 16-h dark period (day 0), similar

initial qP values (0.85) were determined in the wt and psbo2,

but lower (0.7) in the psbo1 mutant. The wt and both mu-

tants exhibited a drop in qP during the first day in HL,

although to different levels; psbo1 showing the lowest va-

lue, i.e. retaining 50% active PSII centres, followed by psbo2

(60%) and wt (70%). In the following 2–7 days of HL treat-

ment, a gradual recovery of PSII activity took place, finally

reaching similar levels in the wt and mutants (Figure 4a).

This recovery was transient, at least in the case of the psbo2

mutant, since the deleterious effects of HL on its fitness

became obvious after 15–28 days of HL treatment. First, the

leaf weight was compared during 15 days of growth under

GL and HL conditions (Figure 4b). No significant difference

was observed in the growth of the wt plants under these

conditions. The same observation is valid for the psbo1

mutant. The average leaf weight of the psbo2 mutant was

drastically reduced after 15 days of HL, even when com-

pared with psbo1. After 28 days of HL, the psbo2 mutant had

completely withered leaves (data not shown). The two

unexpected observations of this phenotypic experiment

could be summarized as follows: (i) the fitness of the psbo1

mutant was only slightly lower than that of wt, and (ii) the

most severe effects were on the growth of the psbo2 mutant.

They point to a critical role for the PsbO2 protein in the PSII

D1 protein repair cycle, where PsbO1 could not function

efficiently during prolonged HL stress.

Light-induced D1 protein degradation is impaired in the

psbo2 mutant

To elucidate the role of PsbO2 in turnover of D1, detached

intact leaves, pre-treated or not with a chloroplast protein

synthesis inhibitor (lincomycin), were subjected to illu-

mination for up to 3 h with the same HL intensity, followed

by measurements of the Fv/Fm parameter and D1 protein

level. The ratios, expressed in per cent, between the values

obtained for leaves treated and not treated with lincomycin,

are plotted in Figure 5(a) as a function of time. The in-

creased susceptibility to photoinhibition of the wt and

psbo1 mutant is demonstrated by both faster inactivation

and D1 degradation in the presence of lincomycin, as

compared with its absence. On the other hand, the time

courses for PSII inactivation and D1 degradation in psbo2

were not affected by the addition of lincomycin, indicating

the lack of de novo D1 synthesis, and thus the lack of D1

degradation in this mutant. An impairment of the D1 deg-

radation step of the turnover cycle in the psbo2 mutant

could explain its considerably reduced fitness after pro-

longed HL stress (Figure 4b).

To confirm these results, and to gain more information on

the initial steps leading to D1 degradation, we isolated

thylakoid membranes, illuminated them with HL for up to

120 min, then measured PSII activity and D1 protein degra-

dation. As shown in Figure 5(b), the time courses for loss of

oxygen evolution were very similar in psbo1 and in the wt.

Instead, the PSII inactivation in psbo2 showed a lag phase of

15 min, and reached similar levels as in psbo1 and the wt

after 60 min of HL treatment. Measurements of the amount

of remaining D1 protein in thylakoids under similar condi-

tions indicated faster kinetics of D1 degradation in the psbo1

mutant, and a complete block of this process in the psbo2

mutant. A possible explanation for the latter observation is

the apparent delay in PSII inactivation due to the higher

relative PSII content in this mutant (Figure 3a and Table 1).

In addition, the presence of PsbO2 may be essential for the

initiation of D1 degradation during HL stress. This possibility

is further supported by the fact that high levels of PsbO2 in

psbo1 (Figure 1d) led to a more efficient degradation of D1

Figure 4. Fitness and phenotype of wild-type (wt) plants and psbo mutants,

grown under high-light conditions (HL ¼ 1000 lmol photons m)2 sec)1).

(a) The proportion of open PSII centres (qP) was calculated as described in

Experimental procedures, and plotted versus the time of HL treatment (n ¼ 5–

7).

(b) Comparison of leaf weight for plants grown under optimal light

(GL ¼ 120 lmol photons m)2 sec)1) and HL for 15 days (n ¼ 15–40).



than in wt, despite similar kinetics for the inactivation of PSII

(Figure 5b).

In vivo phosphorylation levels of PSII proteins

Reversible phosphorylation of PSII proteins regulates the

functional stability of the PSII complex in higher plants (Aro

et al., 2005), and dephosphorylation of damaged D1 protein

is a prerequisite for its degradation (Rintamaki et al., 1996).

To test whether PsbO1 and/or PsbO2 have a role in the early

stages of D1 turnover, we first analyzed the in vivo levels of

PSII protein phosphorylation in the wt and mutants. Thyl-

akoid membranes were isolated in the presence of a phos-

phatase inhibitor (NaF), and analyzed by Western blotting

with anti-phosphothreonine antibodies. We used two dif-

ferent commercial antibodies, which have previously been

shown to be efficient in the detection of phosphorylated

thylakoid proteins, but have different immunoreactivities

towards specific phosphorylated subunits of PSII (Aro et al.,

2004; Vainonen et al., 2005). The representative picture

shown in Figure 6(a) demonstrates that both antibodies,

from Zymed (San Francisco, CA, USA) (left panel) and from

New England BioLabs (Ipswich, MA, USA) (right panel), re-

vealed the same pattern of in vivo phosphorylation of PSII

core proteins in the wt and mutants. Phosphorylated LHCII

was detected by only one antibody (from New England

BioLabs), and the level of its phosphorylation was not sig-

nificantly affected in any of the psbo mutants, as compared

with the wt. The in vivo levels of phosphorylation of D1 and

D2 in psbo2 were similar to those in the wt. However, both

antibodies detected significantly lower levels (61%) of

endogenous phosphorylation of D1 and D2 proteins in the

psbo1 mutant as compared to the wt (Figure 6a).

For further estimation of changes in levels of phosphory-

lation of D1 and D2 proteins we used mass spectrometry

coupled with differential stable isotopic labeling and affinity

chromatography (Ficarro et al., 2003; He et al., 2004; Vai-

nonen et al., 2005). Isolated thylakoid membranes from the

wt and psbo1 were resuspended to equal Chl concentration,

and surface-exposed parts of the membrane proteins,

including phosphorylated N-termini of D1 and D2, were

(a)

(b)

Figure 6. Phosphorylation of thylakoid membrane proteins in leaves of wild-

type (wt) plants and psbo mutants.

(a) Representative Western blot of endogenous protein phosphorylation in

thylakoids isolated from 16-h dark-adapted plants with two different anti-

phosphothreonine antibodies (left panel, antibody from Zymed; right panel,

antibody from New England BioLabs). The gels were loaded with 0.25 lg Chl

per lane. The positions of detected phosphorylated proteins are indicated.

(b) Relative quantification of D1 and D2 protein endogenous phosphorylation

in the wt plants and psbo1 mutant by mass spectrometry analysis of esterified

phosphopeptides differentially labeled with stable isotopes. The mass

spectrum shows two doublets of the signals corresponding to phosphoryl-

ated peptides from the D1 and D2 proteins. The signals at m/z 852.4 and 858.4

correspond to the singly charged ions of light-isotope labeled (wt) and heavy-

isotope labeled (psbo1) peptide from D1, respectively. The signals at m/z

738.4 and 741.4 correspond to the singly charged ions of light-isotope labeled

(wt) and heavy-isotope labeled (psbo1) peptide from D2, respectively.

(a)

(b)

Figure 5. Kinetics for high-light-induced inactivation of PSII electron trans-

port and D1 protein degradation in wild type (wt) and psbo mutants.

(a) The relative levels of the Fv/Fm parameter and of the D1 protein were

determined in detached leaves treated with high light (1000 lmol pho-

tons m)2 sec)1) in the presence (þ) or absence ()) of lincomycin (LN) for the

indicated periods of time. The plotted values represent the ratios between

them (þ/)), indicating the effect of inhibition of chloroplast protein synthesis.

Insert: representative Western blots with anti-D1 antibody of thylakoids

isolated from the HL-treated leaves (0.2 lg Chl per lane).

(b) The relative levels of remaining oxygen evolution activity and of the D1

protein were determined in isolated thylakoid membranes, and plotted as a

function of duration of the high light treatment. The standard deviations were

calculated for data obtained from three to five independent experiments.

Insert: representative Western blot with anti-D1 antibody.



released by proteolytic shaving using trypsin (Vener et al.,

2001). The released peptides were separated from the

residual membranes by centrifugation. Complete removal

of the phosphothreonine-containing domains from the

membrane proteins by trypsin treatment was confirmed by

Western blot analysis with anti-phosphothreonine antibod-

ies (data not shown). Then we performed a parallel differ-

ential stable isotopic labeling of the peptides released from

psbo1 and wt thylakoids. The tryptic peptides from the wt

thylakoids were esterified with d0-methanol (modification of

each carboxyl group gives a peptide mass increment of 14

Da), while the peptides from the mutant thylakoids were

esterified with d3-methanol (modification of each carboxyl

group gives a peptide mass increment of 17 Da). Peptides

from the wt thylakoids labeled with light isotopes were

mixed 1:1 with peptides from the psbo1 mutant thylakoids

labeled with heavy isotopes, the phosphorylated peptides

were enriched from this mixture by immobilized metal

affinity chromatography, and were analyzed by electrospray

ionization mass spectrometry. This allowed simultaneous

measurements of intensities for light- and heavy-isotope

labeled phosphopeptide pairs.

Figure 6(b) shows part of the mass spectrum with the

doublet signals corresponding to singly protonated ions of

N-terminal phosphopeptide Ac-tIALGK of the D2 protein

(light form 738.4 from the wt and heavy form 741.4 from the

psbo1 mutant) and to N-terminal phosphopeptide Ac-tAILER

of the D1 protein (light form 852.4 from the wt and heavy

form 858.4 from the psbo1 mutant). The obtained spectrum

revealed reduced levels of both D1 and D2 phosphorylation

in the psbo1 mutant, compared with the wt. Similar results

were obtained in the experiments with the reversed stable

isotope labeling of the peptides from wt and psbo1 mutant

thylakoids (data not shown). The amounts of phosphorylat-

ed CP43 and PsbH in psbo1 were also slightly reduced as

compared with the wt level (data not shown), most likely due

to an overall reduction in the PSII proteins in this mutant

(Figure 3a and Table 1). The mechanism behind the more

significant reduction in the phosphorylation of D1 and D2

proteins in psbo1 as compared with other PSII core proteins

was further investigated.

In vitro dephosphorylation of the D1 protein is accelerated in

psbo1 and impaired in psbo2

The endogenous level of protein phosphorylation is the re-

sult of phosphorylation–dephosphorylation reactions. In

order to investigate which reaction(s) are affected in the

psbo mutants, we carried out specific in vitro assays. First,

we analyzed and compared the phosphorylation–dep-

hosphorylation reactions of PSII proteins (LHCII and PSII

core) in the wt and mutant thylakoids using Western blotting

with anti-phosphothreonine antibody from Zymed (Fig-

ure 7a). The plot in Figure 7(b) shows the relative levels of

phosphorylated D1 protein determined in the experimental

conditions described below. The thylakoids were isolated

from 16-h dark-adapted plants in the presence of NaF, the

phosphatase inhibitor (Figure 7a, lane 1), washed from NaF,

(a)

(b)

Figure 7. In vitro protein phosphorylation induced by light or by reducing agents in darkness in thylakoids isolated from 16-h dark-adapted wild-type (wt) plants and

psbo mutants.

(a) Representative Western blots with anti-phosphothreonine antibody from Zymed are shown. Thylakoid membranes were isolated in the presence of NaF (lane 1).

The membranes were washed from NaF and incubated in the dark for 1.5 h at 22�C, to allow dephosphorylation of the phosphoproteins by endogenous membrane

protein phosphatase (lane 2). The thylakoid proteins were then phosphorylated in vitro with ATP in the light (lane 3) or in darkness in the presence of ferredoxin and

NADPH (lane 4), activating thylakoid protein kinases. The gels were loaded with 0.25 lg Chl per lane. The positions of the detected thylakoid phosphoproteins are

indicated.

(b) Relative levels of phosphorylated D1 protein were determined by quantification of Western blots as in (a).



and incubated in darkness for 1.5 h to allow dephosphory-

lation of PSII proteins (Figure 7a, lane 2). The D1 and D2

proteins were efficiently dephosphorylated in the wt and

psbo1 mutant, but no significant dephosphorylation of these

proteins was detected in psbo2. As a next step, we per-

formed in vitro phosphorylation of thylakoid proteins with

ATP, activating endogenous thylakoid kinases by two dis-

tinct mechanisms: (i) upon exposure of thylakoid mem-

branes to light (Figure 7a, lane 3), and (ii) in darkness in the

presence of the reducing agents NADPH and ferredoxin

(Figure 7a, lane 4). Phosphorylation of PSII proteins oc-

curred identically in thylakoids isolated from the wt and

mutants, indicating that neither protein kinases nor the re-

dox components of the thylakoid membrane were affected

by the T-DNA insertion. The unchanged phosphorylation

level of the PSII core proteins during the 1.5-h dark incuba-

tion of psbo2 thylakoids, implies a role for PsbO2 in regu-

lating protein dephosphorylation.

To make a detailed analysis of protein dephosphorylation

kinetics in vitro, we isolated thylakoid membranes, phos-

phorylated thylakoid proteins with ATP in the light, and

assayed their time-dependent dephosphorylation at 22�C.

The results, shown in Figure 8(a) (upper panel) and quanti-

fied in Figure 8(b), confirm the fast and specific D1 dep-

hosphorylation in thylakoids from psbo1, as compared with

those from psbo2 and wt plants. The LHCII was similarly

dephosphorylated in the wt and mutants. It was previously

shown that the PSII-specific protein phosphatase is heat

activated (Rokka et al., 2000). Thus, we also performed the

dephosphorylation experiments with isolated thylakoids at

44�C. As shown in Figure 8(a) (lower panel), D1 as well as

LHCII dephosphorylation advanced rapidly and similarly in

thylakoids from both mutants and wt plants, indicating that

the heat-activated PSII-specific protein phosphatase is per-

fectly functional in all three cases. The rapid disappearance

of the phosphorylated D1 protein was not due to a loss in the

amount of D1 protein, as revealed by control Western blots

with anti-D1 antibodies (data not shown). In conclusion, we

suggest that fast in vitro D1 dephosphorylation at 22�C may

be a consequence of specific PsbO2 interaction with the D1

protein at the lumenal side of the PSII complex, which make

the stromal N-terminus of D1 more accessible to the protein

phosphatase and also increases the susceptibility of the D1

protein to degradation.

Discussion

Arabidopsis has two different PsbO proteins with highly

similar amino acid sequences. Using the approaches of re-

verse genetics, biochemistry and physiology we addressed

the question of functional differences between PsbO1 and

PsbO2 proteins, and examined the possible specific roles of

these isoforms in Arabidopsis. We found that PsbO1 and

PsbO2 are both active in photosynthesis, and PsbO2 can

substitute for PsbO1 in PSII activity, although with lower

efficiency under high-light stress. On the other hand, we

showed that PsbO2 has additional regulatory roles in the

dephosphorylation and degradation of the D1 protein, pro-

cesses in which PsbO1 is not able to substitute for PsbO2.

We propose that both copies have a structural role in sta-

bilizing mainly the lumenal side of D1 and CP43. Addition-

ally, PsbO2 might have a function in modulation of the

turnover of D1 protein. We also demonstrated that the total

amount of PsbO regulates the PSII content and activity in the

wt and mutants, most likely by controlling the stability of the

complex, in line with previous observations (Yi et al., 2005).

We found that thylakoid membranes isolated from the

psbo mutants performed differently in terms of kinetics for

photoinactivation and D1 protein degradation under HL

stress conditions (Figure 5b). Remarkably, degradation of

the D1 protein was completely blocked in the psbo2 mutant.

This observation is important, pointing to a critical role for

PsbO2 in the PSII repair cycle, and may explain why the

fitness of the psbo2 mutant was the most affected during

growth under HL conditions (Figure 4b).

Protein phosphorylation preserves the structural integrity

of PSII complexes in the grana thylakoids before migration

to the stroma regions for dephosphorylation and degrada-

tion (Aro et al., 2005; Vener, 2006). We revealed that the

psbo1 and psbo2 mutants had different in vivo phosphory-

(a)

(b)

Figure 8. In vitro dephosphorylation of PSII proteins in thylakoids isolated

from wild-type (wt) plants and psbo mutants.

The thylakoid proteins were first phosphorylated in vitro with ATP during

30 min of illumination. Then thylakoids were transferred to the dark and

incubated at 22�C or 44�C to allow dephosphorylation for the indicated

periods of time. The thylakoid proteins were analyzed by Western blotting

with anti-phosphothreonine antibody from New England BioLabs.

(a) Representative Western blots for D1 and LHCII proteins are shown (0.25 lg

Chl per lane).

(b) Relative phosphorylation levels of the D1 protein (left panel) and of LHCII

(right panel) at 22�C were determined by quantification of Western blots as in

(a) and represented as a function of time.



lation levels of the D1 protein, partly due to different

amounts of protein and partly due to different rates of

dephosphorylation, as demonstrated by Western blotting

with anti-D1 and anti-phosphothreonine antibodies, as well

as by mass spectrometry analyses (Figures 3a, 6 and 8b).

Thylakoid membranes isolated from mutants and wt plants

exhibited similar efficiencies in phosphorylation of PSII

proteins using two activation systems (light or reducing

agents in darkness) (Figure 7). Furthermore, the efficiency of

the heat-shock-activated PSII-specific protein phosphatase

(Rokka et al., 2000) in dephosphorylation of D1 at 44�C was

similar in the mutants and wt plants. However, D1 dep-

hosphorylation at 22�C occurred at a much higher rate in

psbo1 thylakoids than in wt or psbo2 thylakoids (Figure 8).

Thus, we suggest that regulation of D1 dephosphorylation

and degradation by PsbO2 proceeds at the substrate level. In

other words, the interaction between PsbO2 and the D1

protein makes the phosphorylated N-terminus of D1 more

accessible to the stroma-exposed protein phosphatase, and

the protein itself becomes more susceptible to degradation

by specific proteases.

Under normal conditions, either PsbO1 or PsbO2 can

associate with the PSII complex, as revealed by protein

analysis of LHCII–PSII supercomplexes (data not shown).

The interaction between PsbO and the D1 protein (as well as

D2 and CP43) was revealed in the detailed three-dimensional

structure of cyanobacterial PSII (Ferreira et al., 2004), and it

is assumed that the plant PsbO makes the same interactions

(De Las Rivas and Barber, 2004). More specifically, the

interaction of the PsbO with PSII proteins occurs via a loop

linking b-strands 1 and 2, as well as via the extended head

domain between b-strands 5 and 6, both highly conserved

regions in PsbO from prokaryotic and eukaryotic organisms

(De Las Rivas and Barber, 2004). There are several amino

acids that are not identical between mature PsbO1 and

PsbO2 in the b1–b2 and b5–b6 loops, particularly E72/D,

N102/K, D140/E and T143/S. These are conservative substi-

tutions, which probably do not contribute to the differences

in interactions of PsbO1 and PsbO2 with the PSII core

proteins, more specifically D1. Murakami et al. (2005)

explained the poor performance of PsbO2 in supporting

PSII activity by a change of three amino acids in the C-

terminal part of the protein as compared to PsbO1. In the

same work, similar binding affinities of the two PsbOs for

urea-washed PSII were demonstrated in reconstitution

experiments. The differences between PsbO1 and PsbO2

that could explain the selective PsbO2-mediated dephosph-

orylation and degradation of D1, and the molecular mech-

anism of these processes are currently unknown.

It has previously been reported that D1 protein is degra-

ded in a GTP-dependent manner in plants (Spetea et al.,

1999). Spinach PsbO binds GTP, a process which is inhibited

in the presence of 3-(3¢,4¢-dichlorphenyl)-1,1-dimethylurea

(DCMU) (Spetea et al., 2004). This is an inhibitor of PSII

electron transport, able to bind and prevent light-induced

conformational changes in the D1 protein structure, thus

preventing its degradation. These findings indicate that

there is, indeed, a crosstalk between D1 and PsbO in

regulation of D1 turnover, mediated by GTP. The possibility

that the Arabidopsis PsbO isoforms may differ from each

other in the ability to bind and/or hydrolyze GTP with impact

on D1 turnover requires further investigation. Initial experi-

ments indicate a higher GTP hydrolytic activity for PsbO2 as

compared with PsbO1.

The existence of psbO paralogous genes is an example of

gene duplication that occurred 28–48 million years ago in

more than 60% of the Arabidopsis genome (Ermolaeva

et al., 2003). If a function of one of the duplicated genes is

not critical for survival then it will most likely be lost due to

evolutionary pressure. The psbO2 gene has not been

eliminated, which indicates its involvement in processes

important for plant growth, probably regulation of D1

turnover. Notably, the psbO2 gene is greatly different from

psbO1 in the promoter region (data not shown), and has an

upregulated expression under (HL) stress conditions and

programmed cell death, as indicated by Affymetrix and

ATH1 GeneChip� arrays databases (Zimmermann et al.,

2004). It seems that due to a shortage of PSII and a rapid

PsbO2-mediated dephosphorylation of the reaction centre

D1 protein, the psbo1 mutant undergoes a sustained and

highly energy demanding D1 repair process, explaining the

retarded plant growth. On the other hand, the psbo2 mutant,

although having an excess of PSII, can undergo photoinac-

tivation with no subsequent D1 degradation, explaining the

poorest fitness under prolonged HL stress.

Only one PsbO protein was characterized in spinach,

although proteomic analyses of thylakoid lumen revealed

several protein spots corresponding to PsbO (Schubert

et al., 2002). Four PsbO protein spots were also found in

pea (Peltier et al., 2000). Multigene families encoding for the

PsbO protein have been reported in the past in pea (Wales

et al., 1989), tomato (Gorlach et al., 1993) and tobacco

(Palomares et al., 1993). Genomic sequencing of other plant

species, such as rice and wheat, also revealed multiple psbO

genes, coding for highly similar PsbO proteins (data not

shown). Thus, the involvement of Arabidopsis PsbO2 in

acceleration of dephosphorylation and high-light-induced

degradation of PSII reaction centre D1 protein may explain

specific characteristics of D1 protein turnover in higher

plants (Spetea et al., 1999), and could be common for all

plant species.

Experimental procedures

Plant material

Arabidopsis (A. thaliana cv. Columbia) wt plants and T-DNAinsertion mutants from the SALK Institute (SALK 093396 for psbO1



gene and SALK 024720 for psbO2 gene) were grown hydroponi-cally at 120 lmol photons m)2 sec)1 and 22�C with 8-h light/16-hdark cycles (Noren et al., 2004). The plants were screened by PCRfor homozygosity of the T-DNA insertion with primers specific forpsbO1 (5¢-AAAAATAACAGCAAAGATGCCAAGTTCA-3¢ forwardand 5¢-GGAGACAAAAACAAACAAACAACGGCTA-3¢ reverse),psbO2 5¢-TGGATGTGAGGACGGCTTTTCA-3¢ forward and 5¢-TGTGTTTGCTCTTTATTCTCTTTTGCTCTG-3¢ reverse), and T-DNAleft border (LB) 5¢-GCGTGGACCGCTTGCAACT-3¢. The PCR cycleconditions (35 cycles) used were denaturation at 94�C for 30 sec,annealing at 63�C for 45 sec and extension at 72�C for 3 min.Seeds were collected from the homozygous mutant plants and,after several generations, plants were cultivated in a larger scalefor analysis.

Extraction of RNA and RT-PCR analysis

Total RNA of frozen leaf tissues was extracted with RNeasy PlantMini Kit (Qiagen, Hilden, Germany). The corresponding single-stranded cDNAs were obtained by using oligo (dT)15 primer and aReverse Transcription System (Promega, Madison, WI, USA). For-ward and reverse primers for the 18S rRNA were 5¢-CTGCCAG-TAGTCATATGCTTGTC-3¢ and 5¢-CCCGCCACCTATTAAGATCA-3¢.Primers specific for psbO1 cDNA were 5¢-GCCTCTCTCCAATCC-ACCGCTACATT-3¢ forward and 3¢-TGAAAAGGAATGGCACACGTT-C-ACC-5¢ reverse. Primers specific for psbO2 cDNA were5¢-GACCTCAAAGACTTCGCTGGAAAAT-3¢ forward and 3¢-TGGATAGCTCT-TCCTCGTCTCCTCT-5¢ reverse.

Isolation of thylakoids and LHCII–PSII supercomplexes

Thylakoid membranes from the wt plants and psbo mutants wereisolated as previously described (Noren et al., 1999). To assess thein vivo phosphorylation level of PSII proteins, thylakoids were iso-lated from 16-h dark-adapted plants in the presence of 10 mM NaF(general inhibitor for protein phosphatases). For some experiments,LHCII–PSII supercomplexes were prepared according to Eshaghiet al. (1999). Chlorophyll was extracted in 80% (v/v) acetone, and itsconcentration determined according to Porra et al. (1989).

In vitro and in vivo high-light treatments

Isolated thylakoid membranes (0.1 mg Chl ml)1) were illuminatedfor up to 2 h in the presence of 0.2 mM GTP with 1000 lmol pho-tons m)2 sec)1 (high light, HL) by using a Schott KL 2500 LCD lamp(Wiesbaden, Germany) at 22�C. At the indicated periods of time,samples were withdrawn for the determination of remaining PSIIactivity (oxygen evolution) and amount of D1 protein (Westernblotting).

In experiments aimed to study the effect of inhibition ofchloroplast protein synthesis, detached leaves were immersed indistilled water containing or not 2 mM lincomycin (Sigma, St.Louis, MD, USA) in darkness for 16 h. The leaves together withthe media were then transferred to Petri dishes and exposed toHL for up to 3 h at 22�C. Chlorophyll fluorescence was measuredin control and lincomycin-treated leaves after 5 min of darkadaptation. For quantification of D1 protein levels, the leaveswere frozen in liquid nitrogen and stored at )80�C until isolationof the thylakoid membranes.

Where indicated, fully developed plants were grown under HL forup to 28 days while maintaining the light–dark cycle. Chlorophyllfluorescence was determined in leaves harvested after 0 to 7 days ofHL treatment.

Measurements of photosynthetic parameters

Oxygen evolution was measured with a Clark-type electrode(Hansatech, Pentney, King’s Lynn, UK) under saturated visible light(3000 lmol photons m)2 sec)1) at 22�C by using 0.5 mM phenyl-p-benzoquinone as the electron acceptor from PSII. Isolated thylakoidmembranes and LHCII–PSII supercomplexes were suspended in50 mM 2-(N-morpholino)ethanesulfonic acid (MES)–NaOH (pH 6.0),1.0 M glycine betaine, 5 mM MgCl2, 10 mM NaCl and 5 mM CaCl2 at aChl concentration of 10 and 5 lg ml)1, respectively.

Chlorophyll fluorescence was measured using a pulsed-ampli-tude fluorimeter model PAM-210 (Walz, Effeltrich, Germany) inleaves detached from 16-h dark-adapted plants (Schreiber et al.,1986). The dark-adapted state constant fluorescent yield, F0, wasrecorded by using weak measuring light (modulation frequency of32 Hz). Then a 1-sec pulse of saturating visible light (3500 lmolphotons m)2 sec)1, modulation frequency of 8 kHz) was appliedfor measurement of the maximum fluorescence yield (Fm). Themaximum quantum yield of PSII photochemistry (Fv/Fm) wascalculated using the equation Fv/Fm ¼ (Fm ) F0)/Fm. A rapidresponse curve of photosynthesis versus irradiance was meas-ured in detached leaves adapted for 5 min to the PAM-210 lightsource. We performed 12 measurements of the quantum yield ofPSII photochemistry (Y ¢) using 1-sec saturating pulses appliedafter every 20 sec of illumination with photosynthetically activeradiation (PAR) at intensities of 0 to 1250 lmol photons m)2 sec)1,increased stepwise (standardized automatic recording developedby Walz). The relative ETR was calculated by the equationETR ¼ 0.84 · R · PAR · Y ¢ (Genti et al., 1989). It was assumedthat 84% of the incident quanta were absorbed (factor 0.84), andthat the fraction of the absorbed quanta distributed to PSII (factorR) was 0.6 for wt (Suorsa et al., 2006). Accordingly, based on therelative PSII and PSI levels in the mutants, the calculated R valueswere 0.5 and 0.7 for psbo1 and psbo2, respectively. Whereindicated, the proportion of open PSII centers (qP) was deter-mined in detached leaves as the ratio between quantum yield ofPSII at the indicated time of HL treatment (Y ¢), and the maximumquantum yield (Fv/Fm) (Genti et al., 1989).

In vitro assay for phosphorylation and dephosphorylation

of PSII proteins

Isolated thylakoid membranes (0.4 mg Chl ml)1) were phosphoryl-ated with 0.4 mM ATP in the light (120 lmol photons m)2 sec)1) orin darkness in the presence of 1 mM NADPH and 10 lM ferredoxinfor 30 min at 22�C. Thylakoids phosphorylated in the light werefurther incubated in darkness for 1 h at 22�C or 44�C, to allow dep-hosphorylation. The proteins were separated by gel electrophoresisand analyzed by Western blotting using rabbit anti-phosphothreo-nine antibodies (Aro et al., 2004).

Quantitative analysis of PSII protein phosphorylation by

mass spectrometry

Proteolytic cleavage of the surface-exposed peptides from thethylakoid membranes isolated from mutant and wt plants, prepar-ation of the peptide methyl esters using d0-methyl alcohol or d3-methyl d-alcohol (Aldrich, Milwaukee, WI, USA), and isolation ofmethyl-esterified phosphopeptides by immobilized metal affinitychromatography were performed as described earlier (Vainonenet al., 2005). The isolated methyl-esterified phosphopeptides wereanalyzed by electrospray ionization mass spectrometry using ahybrid mass spectrometer API QSTAR Pulsar i (Applied Biosystems,



Foster City, CA, USA) equipped with a nano-electrospray ion source(MDS Protana, Odense, Denmark).

Protein analysis

Thylakoid proteins were separated by gel electrophoresis in 14%(w/v) acrylamide and 6 M urea (Spetea et al., 1999), and visualizedby silver staining. For Western blotting, the proteins were trans-ferred to poly(vinylidene difluoride) membranes (PVDF-Plus0.45 lm; Micron Separations, Westborough, MA, USA), reactedwith specific antibodies followed by an enhanced chemilumines-cence (ECL-Plus) detection system (Amersham Biosciences, Chal-font St Giles, UK), and analyzed by chemiluminescence imaging(LAS-1000; Fuji, Japan). For quantification of thylakoid proteins, allgels were loaded on Chl basis, in amounts allowing the immuno-staining response to be in the linear range. A rabbit polyclonalantibody against the N-terminal EGAPKRLTYDEIQS peptide, whichis identical in Arabidopsis PsbO1 and PsbO2 proteins (Swiss-ProtP23321 and Q9S841, respectively), was produced by Innovagen(Lund, Sweden). Where indicated, rabbit anti-phosphothreonineantibodies obtained from Zymed or New England BioLabs wereused. Other antibodies were raised in rabbit against CP43, D1, PsbPand PsbQ proteins from spinach. Antibodies against the CF1 b-subunit of ATP-synthase and Lhcb2 protein were purchased fromAgrisera (Umea, Sweden). The antibodies against D2 from Syn-echocystis and PsaA/B from spinach were kindly provided by Pro-fessor Eva-Mari Aro (Turku University).

Acknowledgements

We thank the Salk Institute Genomic Analysis Laboratory for pro-viding the sequence-indexed Arabidopsis T-DNA insertion lines andProfessor Eva-Mari Aro (Turku University) for antibodies against theD2 and PsaA/B proteins as well as for helpful suggestions con-cerning lincomycin experiments. This work was supported bygrants from the Swedish Research Council (C.S., A.V.V.), theGraduate Research School in Genomics and Bioinformatics (C.S.),the Swedish Research Council for Environment, Agriculture andSpace Planning (A.V.V.), Nordiskt Kontaktorgan for Jordbruksfors-king (A.V.V.), the Hagberg Foundation (C.S.) and the Helge Ax:sonJohnson Foundation (B.L.).

References

Alonso, J.M., Stepanova, A.N., Leisse, T.J. et al. (2003) Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science,301, 653–657.

AGI (The Arabidopsis Genome Initiative) (2000) Analysis of thegenome sequence of the flowering plant Arabidopsis thaliana.Nature, 408, 796–815.

Aro, E.M., Rokka, A. and Vener, A.V. (2004) Determination ofphosphoproteins in higher plant thylakoids. Meth. Mol. Biol. 274,271–286.

Aro, E.M., Suorsa, M., Rokka, A., Allahverdiyeva, Y., Paakkarinen,

V., Saleem, A., Battchikova, N. and Rintamaki, E. (2005) Dynamicsof photosystem II: a proteomic approach to thylakoid proteincomplexes. J. Exp. Bot. 56, 347–356.

Bonardi, V., Pesaresi, P., Becker, T., Schleiff, E., Wagner, R.,

Pfannschmidt, T., Jahns, P. and Leister, D. (2005) Photosystem IIcore phosphorylation and photosynthetic acclimation require twodifferent protein kinases. Nature, 437, 1179–1182.

Booij-James, I.S., Swegle, W.M., Edelman, M. and Mattoo, A.K.

(2002) Phosphorylation of the D1 photosystem II reaction centerprotein is controlled by an endogenous circadian rhythm. PlantPhysiol. 130, 2069–2075.

Burnap, R.L. and Sherman, L.A. (1991) Deletion mutagenesis inSynechocystis sp. PCC6803 indicates that the Mn-stabilizingprotein of photosystem II is not essential for O2 evolution. Bio-chemistry, 30, 440–446.

De Las Rivas, J. and Barber, J. (2004) Analysis of the structureof the PsbO protein and its implications. Photosynth. Res. 81,329–343.

Ermolaeva, M.D., Wu, M., Eisen, J.A. and Salzberg, S.L. (2003) Theage of the Arabidopsis thaliana genome duplication. Plant Mol.Biol. 51, 859–866.

Eshaghi, S., Andersson, B. and Barber, J. (1999) Isolation of a highlyactive PSII-LHCII supercomplex from thylakoid membranes by adirect method. FEBS Lett. 446, 23–26.

Ferreira, K.N., Iverson, T.M., Maghlaoui, K., Barber, J. and Iwata, S.

(2004) Architecture of the photosynthetic oxygen-evolving centre.Science, 303, 1831–1838.

Ficarro, S., Chertihin, O., Westbrook, V.A. et al. (2003) Phospho-proteome analysis of capacitated human sperm. Evidence oftyrosine phosphorylation of a kinase-anchoring protein 3 andvalosin-containing protein/p97 during capacitation. J. Biol. Chem.278, 11579–11589.

Genti, B.E., Briantais, J.M. and Baker, N.R. (1989) The relationshipbetween the quantum yield of photosynthetic electron transportand quenching of chlorophyll fluorescence. Biochim. Biophys.Acta. 990, 87–92.

Gomez, S.M., Bil, K.Y., Aguilera, R., Nishio, J.N., Faull, K.F. and

Whitelegge, J.P. (2003) Transit peptide cleavage sites of integralthylakoid membrane proteins. Mol. Cell. Proteomics, 2, 1068–1085.

Gorlach, J., Schmid, J. and Amrhein, N. (1993) The 33 kDa protein ofthe oxygen-evolving complex: a multi-gene family in tomato.Plant Cell Physiol. 34, 497–501.

Haußuhl, K., Andersson, B. and Adamska, I. (2001) A chloroplastDegP2 protease performs the primary cleavage of the photo-damaged D1 protein in plant photosystem II. EMBO J. 20, 713–722.

He, T., Alving, K., Field, B., Norton, J., Joseloff, E.G., Patterson, S.D.

and Domon, B. (2004) Quantitation of phosphopeptides usingaffinity chromatography and stable isotope labeling. J. Am. Soc.Mass Spectrom. 15, 363–373.

Kieselbach, T., Bystedt, M., Hynds, P., Robinson, C. and Schroder,

W.P. (2000) A peroxidase homologue and novel plastocyaninlocated by proteomics to the Arabidopsis chloroplast thylakoidlumen. FEBS Lett. 480, 271–276.

Lindahl, M., Spetea, C., Hundal, T., Oppenheim, A.B., Adam, Z. and

Andersson, B. (2000) The thylakoid FtsH protease plays a role inthe light-induced turnover of the photosystem II D1 protein. PlantCell, 12, 419–431.

Mayfield, S.P., Bennoun, P. and Rochaix, J.D. (1987) Expression ofthe nuclear encoded OEE1 protein is required for oxygen evolu-tion and stability of photosystem II particles in Chlamydomonasreinhardtii. EMBO J. 6, 313–318.

Murakami, R., Ifuku, K., Takabayashi, A., Shikanai, T., Endo, T. and

Sato, F. (2002) Characterization of an Arabidopsis thaliana mutantwith impaired psbO, one of two genes encoding extrinsic 33-kDaproteins in photosystem II. FEBS Lett. 523, 138–142.

Murakami, R., Ifuku, K., Takabayashi, A., Shikanai, T., Endo, T. and

Sato, F. (2005) Functional dissection of two Arabidopsis PsbOproteins: PsbO1 and PsbO2. FEBS J. 272, 2165–2175.



Noren, H., Svensson, P. and Andersson, B. (1999) Auxiliary photo-synthetic functions of Arabidopsis thaliana-studies in vitro and invivo. Biosci. Rep. 19, 499–509.

Noren, H., Svensson, P. and Andersson, B. (2004) A convenient andversatile hydroponic cultivation system for Arabidopsis thaliana.Physiol. Plant. 121, 343–348.

Palomares, R., Herrmann, R.G. and Oelmuller, R. (1993) Post-tran-scriptional and post-translational regulatory steps are crucial incontrolling the appearance and stability of thylakoid polypeptidesduring the transition of etiolated tobacco seedlings to white light.Eur. J. Biochem. 217, 345–352.

Peltier, J.B., Friso, G., Kalume, D.E., Roepstorff, P., Nillson, F.,

Adamska, I. and van Wijk, K.J. (2000) Proteomics of the chloro-plast: systematic identification and targeting analysis of lumenaland peripheral thylakoid proteins. Plant Cell, 12, 319–341.

Peltier, J.B., Emanuelsson, O., Kalume, D.E. et al. (2002) Centralfunctions of the lumenal and peripheral thylakoid proteome ofArabidopsis determined by experimentation and genome-wideprediction. Plant Cell, 14, 211–236.

Porra, R.J., Thompson, W.A. and Kriedemann, P.E. (1989) Deter-mination of accurate extinction coefficients and simultaneous-equations for assaying chlorophyll a and chlorophyll b extractedwith 4 different solvents – verification of the concentration ofchlorophyll standards by atomic absorption spectroscopy. Bio-chim. Biophys. Acta. 975, 384–394.

Pursiheimo, S., Rintamaki, E., Baena-Gonzalez, E. and Aro, E.M.

(1998) Thylakoid protein phosphorylation in evolutionally diver-gent species with oxygenic photosynthesis. FEBS Lett. 423, 178–182.

Rintamaki, E., Kettunen, K. and Aro, E.M. (1996) Differential D1dephosphorylation in functional and photodamaged photosys-tem II centers. Dephosphorylation is a prerequisite for degrada-tion of damaged D1. J. Biol. Chem. 271, 14870–14875.

Rokka, A., Aro, E.M., Herrmann, R.G., Andersson, B. and Vener, A.V.

(2000) Dephosphorylation of photosystem II reaction center pro-teins in plant photosynthetic membranes as an immediate re-sponse to abrupt elevation of temperature. Plant Physiol. 123,1525–1536.

Schreiber, U., Schliwa, U. and Bilger, W. (1986) Continuousrecording of photochemical and non-photochemical chlorophyllfluorescence quenching with a new type of modulation fluo-rometer. Photosynth. Res. 10, 51–62.

Schubert, M., Petersson, U.A., Haas, B.J., Funk, C., Schroder, W.P.

and Kieselbach, T. (2002) Proteome map of the chloroplast lumenof Arabidopsis thaliana. J. Biol. Chem. 277, 8354–8365.

Snyders, S. and Kohorn, B.D. (2001) Disruption of thylakoid-asso-ciated kinase 1 leads to alteration of light harvesting in Arabid-opsis. J. Biol. Chem. 276, 32169–32176.

Spetea, C., Hundal, T., Lohmann, F. and Andersson, B. (1999) GTPbound to chloroplast thylakoid membranes is required for light-induced, multienzyme degradation of the photosystem II D1protein. Proc. Natl. Acad. Sci. USA, 96, 6547–6552.

Spetea, C., Keren, N., Hundal, T., Doan, J.M., Ohad, I. and Anders-

son, B. (2000) GTP enhances the degradation of the photosystemII D1 protein irrespective of its conformational heterogeneity atthe Q(B) site. J. Biol. Chem. 275, 7205–7211.

Spetea, C., Hundal, T., Lundin, B., Heddad, M., Adamska, I. and

Andersson, B. (2004) Multiple evidence for nucleotide metabo-lism in the chloroplast thylakoid lumen. Proc. Natl. Acad. Sci.USA, 101, 1409–1414.

Suorsa, M., Sirpio, S., Allahverdiyeva, Y., Paakkarinen, V.,

Mamedov, F., Styring, S. and Aro, E.M. (2006) PsbR, a missing linkin the assembly of the oxygen-evolving complex of plant photo-system II. J. Biol. Chem. 281, 145–150.

Vainonen, J.P., Hansson, M. and Vener, A.V. (2005) STN8protein kinase in Arabidopsis thaliana is specific in phos-phorylation of photosystem II core proteins. J. Biol. Chem. 280,33679–33686.

Vener, A.V. (2006) Phosphorylation of thylakoid proteins. In Pho-toprotection, Photoinhibition, Gene Regulation and Environment(Demmig-Adams, B., Adams, W. & Matoo, A.K., eds). The Neth-erlands: Springer, pp. 107–126.

Vener, A.V., Ohad, I. and Andersson, B. (1998) Protein phosphory-lation and redox sensing in chloroplast thylakoids. Curr. Opin.Plant Biol. 1, 217–223.

Vener, A.V., Rokka, A., Fulgosi, H., Andersson, B. and Herrmann,

R.G. (1999) A cyclophilin-regulated PP2A-like protein phospha-tase in thylakoid membranes of plant chloroplasts. Biochemistry,38, 14955–14965.

Vener, A.V., Harms, A., Sussman, M.R. and Vierstra, R.D. (2001)Mass spectrometric resolution of reversible protein phosphory-lation in photosynthetic membranes of Arabidopsis thaliana. J.Biol. Chem. 276, 6959–6966.

Wales, R., Newman, J.B., Pappin, D. and Gray, C.J. (1989) Theextrinsic 33 kDa polypeptide of the oxygen-evolving complexof photosystem II is a putative calcium-binding protein and isencoded by a multi-gene family in pea. Plant Mol. Biol. 12, 439–451.

Yamamoto, Y., Ishikawa, Y., Nakatani, E., Yamada, M., Zhang, H.

and Wydrzynski, T. (1998) Role of an extrinsic 33 kilodalton pro-tein of photosystem II in the turnover of the reaction center-binding protein D1 during photoinhibition. Biochemistry, 37,1565–1574.

Yi, X., McChargue, M., Laborde, S., Frankel, L.K. and Bricker, T.M.

(2005) The manganese-stabilizing protein is required for photo-system II assembly/stability and photoautotrophy in higherplants. J. Biol. Chem. 280, 16170–16174.

Zimmermann, P., Hirsch-Hoffmann, M., Hennig, L. and Gruissem,

W. (2004) GENEVESTIGATOR. Arabidopsis microarray databaseand analysis toolbox. Plant Physiol. 136, 2621–2632.

Zouni, A., Witt, H.T., Kern, J., Fromme, P., Krauss, N., Saenger, W.

and Orth, P. (2001) Crystal structure of photosystem II fromSynechococcus elongatus at 3.8 A resolution. Nature, 409, 739–743.



Documents

The Arabidopsis PsbO2 protein regulates dephosphorylation and turnover of the photosystem II reaction centre D1 protein