Báo cáo khoa học: Functional dissection of two Arabidopsis PsbO proteins PsbO1 and PsbO2 doc

Functional dissection of two Arabidopsis PsbO proteins
PsbO1 and PsbO2
Reiko Murakami
1
, Kentaro Ifuku
1
, Atsushi Takabayashi
1
, Toshiharu Shikanai
2
, Tsuyoshi Endo
1
and Fumihiko Sato
1
1 Division of Integrated Life Sciences, Graduate School of Biostudies, Kyoto University, Kyoto, Japan
2 Graduate School of Agriculture, Kyushu University, Higashiku, Fukuoka, Japan
In oxygenic photosynthesis, the multisubunit protein
complex photosystem II (PSII) uses light energy to
oxidize water and to form molecular oxygen [1–3].
Water oxidation occurs at a catalytic site of PSII that
contains four manganese atoms, and PSII contains sev-
eral extrinsic subunits that play important roles in sta-
bilizing the active manganese site. One of these, the
nuclear-encoded PsbO protein in PSII, has a molecular
mass of approximately 26 kDa, although it has been
called extrinsic 33-kDa protein traditionally, and is syn-
thesized with a transit sequence that targets the protein
to the thylakoid lumen [4,5]. PsbO is present in all
oxygen-evolving organisms. It appears to play a central
role in stabilization of the manganese cluster and is
essential for efficient and stable oxygen evolution.

To better understand the role of PsbO, mutants that
lack the psbO gene have been established. PsbO has
been deleted using genetic methods in cyanobacteria
[6,7] and in green algae [8], and such mutants have
been maintained under heterotrophic conditions. In
contrast, no PsbO mutant has been obtained in higher
plants due to the essential role of this protein in pho-
tosynthesis. Previously, however, we isolated an Ara-
bidopsis mutant [9,10] with a defect in the psbO1 gene
Keywords
Arabidopsis thaliana; isoform; oxygen
evolution; psbO1; psbO2
Correspondence
T. Endo
Division of Integrated Life Sciences,
Graduate School of Biostudies, Kyoto
University, Kyoto 606–8502, Japan
Fax: +81 75 7536398
Tel: +81 75 7536381
E-mail:
(Received 24 December 2004, revised 17
February 2005, accepted 1 March 2005)
doi:10.1111/j.1742-4658.2005.04636.x
PsbO protein is an extrinsic subunit of photosystem II (PSII) and has been
proposed to play a central role in stabilization of the catalytic manganese
cluster. Arabidopsis thaliana has two psbO genes that express two PsbO
proteins; PsbO1 and PsbO2. We reported previously that a mutant plant
that lacked PsbO1 (psbo1) showed considerable growth retardation despite
the presence of PsbO2 [Murakami, R., Ifuku, K., Takabayashi, A., Shika-
nai, T., Endo, T., and Sato, F. (2002) FEBS Lett 523, 138–142]. In the pre-

sent study, we characterized the functional differences between PsbO1 and
PsbO2. We found that PsbO1 is the major isoform in the wild-type, and
the amount of PsbO2 in psbo1 was significantly less than the total amount
of PsbO in the wild-type. The amount of PsbO as well as the efficiency of
PSII in psbo1 increased as the plants grew; however, it never reached the
total PsbO level observed in the wild-type, suggesting that the poor activity
of PSII in psbo1 was caused by a shortage of PsbO. In addition, an in vitro
reconstitution experiment using recombinant PsbOs and urea-washed PSII
particles showed that oxygen evolution was better recovered by PsbO1 than
by PsbO2. Further analysis using chimeric and mutated PsbOs suggested
that the amino acid changes Val186 fi Ser, Leu246 fi Ile, and
Val204 fi Ile could explain the functional difference between the two
PsbOs. Therefore we concluded that both the lower expression level and
the inferior functionality of PsbO2 are responsible for the phenotype
observed in psbo1.
Abbreviations
F
m
, maximum fluorescence yield at closed PSII centers; F
0
, minimum fluorescence yield at open PSII centers; F
v
, F
m
) F
0
; PSII,
photosystem II.
FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS 2165
(psbo1), as PsbO proteins in Arabidopsis are encoded

by two genes, psbO1 and psbO2 [11]. On the other
hand, many other plant species have only one psbO
gene. The second gene in Arabidopsis would be a
unique result of a duplication of the genome and chro-
mosomes some 28–48 million years ago, before the
divergence of Arabidopsis from Brassica about 14–24
million years ago [12,13].
Our previous analysis showed that psbo1 exhibited
weak photosynthetic activity and considerable growth
retardation compared with the wild-type [10]. This
observation raised the question of why the psbO2 gene
could not complement the defect in the psbO1 gene,
since PsbO1 and PsbO2 are highly homologous with
regard to the primary structure.
In this study, we characterized the differences in the
accumulation and biochemical activity of these iso-
forms to clarify their functional differences. A detailed
immunoblot analysis using the mutant, psbo1 , and the
wild-type showed that a lower level of PsbO2 would
limit the photosynthetic activity. Additional in vitro
experiments with urea-washed PSII particles reconsti-
tuted with PsbOs revealed that oxygen-evolution was
better recovered by PsbO1 than by PsbO2, though the
two isoforms had a similar binding affinity for the
PSII particles. These results showed that the differ-
ences in both biochemical activity and the level of
accumulation were responsible for the poor photosyn-
thetic activity in psbo1.
We further dissected the functional differences
between PsbO1 and PsbO2, as these isoforms have

quite a similar primary structure, and differ at only 11
amino acids. A previous investigation [5] pointed out
the importance of Cys at positions 29 and 52 of
mature PsbO1 for forming a disulfide bridge [14,15],
and Val at position 148 for producing a b-sheet [16],
as well as carboxylic residues that have been reported
to be involved in the interaction between PsbO and
PSII [17,18]. Moreover, these amino acid residues are
conserved in PsbO2. In addition, the predictions
regarding the structures of Arabidopsis PsbO1 and
PsbO2 based on PSII of Thermosynechococcus elonga-
tus [19] (protein data bank accession number 1S5L)
were similar. Thus, to identify the amino acids respon-
sible for the poor oxygen-evolving activity of PsbO2,
we prepared chimeric proteins derived from PsbO1
and PsbO2. Further, site-directed mutagenesis showed
that the replacement of Val186 of mature PsbO1
with Ser and of Leu246 with Ile decreased the oxygen-
evolving activity, whereas substitution of Val-204 with
Ile increased the oxygen-evolving activity. Based on
these results, we discuss the physiological role of the
duplicated PsbO in Arabidopsis.
Results
PsbO isoform level and the efficiency of PSII
To understand the reason for the loss of photosyn-
thetic activity in psbo1, we measured the protein levels
of PsbOs and the photosynthetic activity in mutant
and wild-type plants. Whereas the wild-type plants
maintained strong photosynthetic activity throughout
their growth, psbo1 plants showed a gradual increase

in electron transport activity, represented by the chlo-
rophyll fluorescence parameter, F
v
⁄ F
m
(the maximum
photochemical efficiency in PSII) [20,21], as they grew
(Fig. 1).
PsbO1 and PsbO2 can be distinguished on
SDS ⁄ PAGE, since PsbO1 migrated more slowly than
PsbO2 [10]. Immunoblot analysis showed that the
major isoform in the wild-type was PsbO1, which com-
prised about 90% of the total amount of PsbO based
on densitometry of the immunoblot membranes
(Fig. 2A). The amount of PsbO2 was much greater in
psbo1 than in the wild-type (Fig. 2A,B), indicating that
the expression of psbO2 was activated in a compensa-
tory manner. At an early growth stage, the PsbO2 level
in psbo1 accounted for 40% of the total in the wild-
type. The amount of PsbO2 increased considerably in
mature psbo1, and reached about 70% of the total
in the wild-type. This change in the level of PsbO2
coincided with the change in F
v
⁄ F
m
in immature and
mature psbo1 plants; i.e. 0.54 ± 0.05 and 0.68 ± 0.03,
respectively. The accumulation of PsbO2 in mature
plants clearly led to more efficient PSII.

Similarly, immunoblot analysis of other PSII pro-
teins, PsbA and PsbP, showed that the amounts of
0.3
0.4
0.5
0.6
0.7
0.8
0.9
020406080100
da
y
s after
g
ermination
wild-type
psbo1
Fv/Fm
Fig. 1. Potential quantum yield of PSII (F
v
⁄ F
m
) in leaves of the wild-
type (r) and psbo1 (h) during the course of plant growth. Values
represent averages and standard deviations for 10–15 plants.
Functional dissection of PsbO1 and PsbO2 R. Murakami et al.
2166 FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS
PsbA and PsbP were also smaller in psbo1 than in the
wild-type (Fig. 2A,B), whereas the levels of these pro-
teins also increased as psbo1 grew. As PsbO stabilizes

the manganese cluster and PSII core, and it has often
been reported that PsbO deficiency affects the amount
of PsbA and PsbP [22–24], it is very likely that the
amount of PsbOs limits the amounts of PsbA and
PsbP. These results suggest that the amount of PsbOs
in psbo1 is the critical limiting factor for photosyn-
thesis and growth.
Biochemical activities of PsbO1 and PsbO2
in oxygen evolution
As it was clear that the shortage of PsbOs in psbo1 is
the critical limiting factor for photosynthesis, we car-
ried out in vitro experiments using recombinant PsbOs
and urea-washed PSII particles from spinach to clarify
the qualitative differences between PsbO1 and PsbO2.
Arabidopsis PsbO has been reported to be able to bind
urea-washed PSII particles of spinach and to restore
the oxygen-evolving activity as effectively as spinach
PsbO [16,25]. We carried out reconstitution procedures
according to their methods. First, we examined the
oxygen-evolving activity as a function of the PsbOs ⁄
PSII ratio in the assay medium (Fig. 3A). The maxi-
mum activity with PsbO2 was about 80% of that with
PsbO1, whereas the maximum oxygen-evolving activity
reconstituted with both PsbO1 and PsbO2 was
achieved at about 2 PsbOs ⁄ PSII. The restoration curve
obtained with our recombinant PsbOs closely resem-
bled that for spinach PsbO and Arabidopsis PsbO1
reported by Betts et al . [25]. PsbO1 and PsbO2 also
showed a similar affinity for urea-washed PSII parti-
cles (Fig. 3B,C); the binding of both PsbOs with urea-

washed PSII particles was saturated at about 2
PsbOs ⁄ PSII, like the oxygen-evolving activity.
Competition analysis of PsbO1 and PsbO2
To confirm that the two PsbOs have a similar binding
affinity, the competition between them for binding sites
on PSII particles was analyzed. Increasingly larger
amounts of an equimolar mixture of PsbO1 and PsbO2
were incubated with urea-washed PSII particles
(Fig. 4A,B), and the relative amounts of bound-PsbO1
wild-type
y
oun
g
wild-type
mature
psbo1
y
oun
g
psbo1
mature
PsbO
PsbA
PsbP
Relative protein accumulation
140
120
100
80
60

40
20
0
PsbA
PsbP
wild-type
MY
psbo1
MY
wild-type
MY
psbo1
MY
CBB-staining
PsbO1
PsbO2
A
B
Fig. 2. Protein analysis by SDS ⁄ PAGE
(15%). (A) Coomassie brilliant blue-staining
and immunoblot analysis with polyclonal
antibodies against spinach PsbO, PsbP and
PsbA. Antibodies against PsbO detect two
signals with slightly different migrations,
upper band; PsbO1, lower band; PsbO2.
Thylakoid membranes were loaded on a
chlorophyll basis; equivalent to 5 lg of chlo-
rophyll for Coomassie brilliant blue-staining
and 1 lg for immunoblot analysis. The
arrow shows PsbOs in Coomassie brilliant

blue-staining. Y, young plants (leaf size,
about 0.5 cm); M, mature plants (leaf size,
about 2.5 cm). (B) The relationship between
the accumulation of PsbO (black), PsbP
(mid-grey) and PsbA (light grey). The protein
level was quantified by densitometry. Values
are relative to the protein level in the young
wild-type (100%). Standard deviations were
calculated from three measurements.
R. Murakami et al. Functional dissection of PsbO1 and PsbO2
FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS 2167
and bound-PsbO2 with urea-washed PSII particles
were quantified by the densitometry of Coomassie bril-
liant blue-stained proteins on a SDS ⁄ PAGE plate. As
PsbO1 migrated more slowly than PsbO2, the two
could be distinguished in the binding analysis
[10,26–28]. Urea-washed PSII particles incubated with
two moles of PsbOs per mole of PSII particles (one
mole of each protein added, 1 : 1) had similar amounts
of PsbO1 and PsbO2 (approximately PsbO1 ⁄ PsbO2 ¼
51 : 49). The oxygen-evolving activity with a mixture
of one mole each of PsbO1 and PsbO2 was higher
than that with four moles of PsbO2, but lower than
that with four moles of PsbO1. Although the differ-
ence was small, the same results were obtained consis-
tently. As the levels of both proteins were saturated at
about 2 PsbOs ⁄ PSII, we added an excessive amount of
PsbOs per PSII (two moles of each protein, 2 : 2). The
two proteins bound equally with urea-washed PSII
(approximately 50 : 50) and the oxygen-evolving activ-

ity was as high as the sample; 1 : 1. These results sug-
gested that the two proteins had a similar affinity for
urea-washed PSII.
Next, we examined the effect of increasing the propor-
tion of PsbO1 on the reconstitution of PSII. The ratio of
PsbO1 to total PsbOs bound with urea-washed PSII
particles and the oxygen-evolving activity demonstrated
equal competition between PsbO1 and PsbO2 for PSII
particles (Fig. 4C). These competition analyses were
consistent with the analysis of the reconstitution with
PsbO1 and PsbO2, in terms of similar binding affinity.
Chimeric PsbOs and surface charge
To examine which amino acid of PsbO2 is responsible
for the poor oxygen-evolving activity, we prepared chi-
meric proteins with PsbO1 and PsbO2 in Escherichia
coli (Fig. 5A). For example, the chimeric protein
PsbO1-1-2 has the N-terminal and middle parts of
PsbO1 combined to the C-terminus of PsbO2. Each of
the parts contains about 80 amino acids, and, respect-
ively, has three, five and three amino acid replacements
between PsbO1 and PsbO2.
SDS ⁄ PAGE analysis of chimeric PsbOs showed that
PsbO1-1-2 and PsbO2-1-1 migrated as slowly as
PsbO1, and only PsbO1-2-1 migrated as PsbO2 did
(Fig. 5B). Anion-exchange chromatography was also
performed, as PsbO1 was eluted at a higher NaCl
012345
PsbO1
PsbO2
100

80
60
40
20
0
untreated PSII
urea-washed PSII
0.5
1
2
4
0.5
1
2
4
PsbO1 PsbO2
012345
140
120
100
80
60
40
20
0
PsbO1
PsbO2
A
B
C

Fig. 3. The oxygen-evolving activity of urea-washed PSII reconstitu-
ted with PsbO1 and PsbO2 and the binding affinity for urea-washed
PSII of PsbO1 and PsbO2. (A) The relative oxygen-evolving activity
of urea-washed PSII reconstituted with PsbO1 (r) and PsbO2 (h).
The maximum rate of oxygen evolution (100%) equals the rate
measured for untreated PSII minus that for urea-washed PSII. (B)
Coomassie brilliant blue-staining. Urea-washed PSII reconstituted
with PsbO1 and PsbO2 were separated on SDS ⁄ PAGE (15%). An
amount equivalent to 2 lg of chlorophyll was loaded in each lane.
The arrow shows PsbOs in Coomassie brilliant blue-staining. (C)
Quantification of PsbO1 (r) and PsbO2 (h) bound to urea-washed
PSII particles. Values quantified by densitometry were plotted
against PsbOs added per PSII. Values are relative to the untreated
PSII (100%). Standard deviations were calculated from five meas-
urements.
Functional dissection of PsbO1 and PsbO2 R. Murakami et al.
2168 FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS
concentration than was PsbO2. PsbO1-2-1 again
showed a chromatographic elution profile similar to
PsbO2, whereas the profiles of other chimeric proteins
were similar to that of PsbO1 (data not shown). In
fact, these results were consistent with the theoretical
pI values; the theoretical pI values of PsbO1, PsbO1-1-2
and PsbO2-1-1 were calculated to be 4.96 (calculated
molecular mass, 26565 Da), 4.95 (26567 Da) and 4.96
(26579 Da), respectively, and those of PsbO2 and
PsbO1-2-1 were calculated to be 5.05 (26571 Da) and
5.06 (26555 Da), respectively (to compute pI ⁄ Mw,
). These results suggest that, in
terms of surface charge, PsbO1-2-1 was similar to PsbO2

and the other chimeric PsbOs were similar to PsbO1.
Ppu MI Bam HI
PsbO1
PsbO2
PsbO1-1-2
PsbO2-1-1
PsbO1-2-1
A
B
C
Fig. 5. Oxygen-evolving activity of chimeric proteins derived from
PsbO1 and PsbO2. (A) A model of the chimeric PsbOs. In chimeric
PsbOs, a part of PsbO1 (C-terminus, middle and N-terminus) was
replaced by a corresponding part of PsbO2. Each of the parts con-
tains about 80 amino acids, and, respectively, has three, five and
three amino acid replacements between PsbO1 and PsbO2. (B)
Coomassie brilliant blue-staining on SDS ⁄ PAGE (15%) of PsbO1,
PsbO2 and chimeric PsbOs. They were expressed using a pET-sys-
tem in Escherichia coli and purified by anion-exchange chromato-
graphy. Protein (0.5 lg) was loaded in each lane. (C) Oxygen-evol-
ving activities of urea-washed PSII reconstituted with PsbO1,
PsbO2 and chimeric PsbOs. The maximum rate of oxygen evolu-
tion (100%) equals the rate measured for untreated PSII minus that
for urea-washed PSII.
untreated
PSII
urea-washed
PSII
PsbO
PsbO2

(1:1)
(2:2)
10020 40 60 800
120
0 0.2 0.4 0.6 0.8 1
100
80
60
40
20
0
120
A
B
C
Fig. 4. Competition of PsbO1 and PsbO2 for the reconstitution of
PSII. (A) Reconstitution of PSII with an equimolar mixture of PsbO1
and PsbO2. Reconstituted PSII particles were separated on a 15%
SDS ⁄ PAGE and stained by Coomassie brilliant blue. An amount
equivalent to 2 lg of chlorophyll was loaded in each lane. PsbO1,
PsbO2; 4 mol PsbO1 or PsbO2 per PSII, 1 : 1; 1 mole PsbO1 and
1mole PsbO2 per PSII, 2 : 2; 2 mole PsbO1 and 2 mole PsbO2 per
PSII. (B) Relative oxygen-evolving activity of PSII reconstituted with
an equimolar mixture of PsbO1 and PsbO2. The maximum rate of
oxygen evolution (100%) equals the rate measured for untreated
PSII minus that for urea-washed PSII. (C) Reconstitution of PSII
with PsbOs in which the ratio of PsbO1 to total PsbO was varied.
The relative oxygen-evolving activity (r) and the ratio of bound
PsbO1 : PsbOs (h). The maximum rate of oxygen evolution (100%)
equals the rate measured for untreated PSII minus that for urea-

washed PSII. The ratio of bound PsbO1 : PsbOs was estimated
from the Coomassie brilliant blue-staining on SDS ⁄ PAGE. Standard
deviations were calculated from five measurements.
R. Murakami et al. Functional dissection of PsbO1 and PsbO2
FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS 2169
However, the oxygen-evolving activity of urea-
washed PSII particles reconstituted with PsbO1-2-1
was similar to that of the particles reconstituted with
PsbO1 (Fig. 5C). A lower level of activity was
observed with PsbO1-1-2, a chimera with the C-ter-
minal sequence of PsbO2. Although the difference was
small, the same results were obtained repeatedly. These
results suggest that the difference in surface charge
between PsbO1 and PsbO2 did not affect the difference
in oxygen-evolving activity.
C-Terminal amino acid substitution and
oxygen-evolving activity
PsbO1 and PsbO1-1-2 had three amino acid changes;
Val186 (PsbO1) to Ser (PsbO2), Val204 to Ile, and
Leu246 to Ile (Fig. 6A). To examine which replace-
ment was responsible for the difference in the oxygen-
evolving activity, we prepared mutated PsbO1 in which
an amino acid was substituted for the corresponding
one in PsbO2 (V186S, V204I and L246I). Measure-
ment of the oxygen-evolving activity upon reconstitu-
tion with the spinach PSII core showed that the
activity levels with V186S and L246I were apparently
lower than that with PsbO1 (Fig. 6B), suggesting that
these two amino acids were responsible for the lower
level of activity. Unexpectedly, V204I showed stronger

oxygen-evolving activity than PsbO1.
We next prepared double-mutated proteins with two
amino acid substitutions; i.e. V186S ⁄ V204I, V186S ⁄
L246I and V204I ⁄ L246I. The oxygen-evolving activity
reconstituted with V186S ⁄ V204I and V204I ⁄ L246I was
similar to that with PsbO1; levels were higher than for
V186S and L246I and lower than for V204I (Fig. 6B).
The activity reconstituted with V186S ⁄ L246I was much
weaker than that with V186S, L246I or PsbO2. These
results confirmed that the replacement of Val186 with
Ser and Leu146 with Ile led to a reduction in the level
of oxygen-evolving activity, while the replacement of
Val204 with Ile led to an increase.
Location of amino acids that differ between
PsbO1 and PsbO2
The locations of three amino acids, Val186 Val204 and
Leu246, were predicted using the Thermosynechococcus
elongatus PsbO [19] (protein data bank accession num-
ber; 1S5L) as a template. The prediction suggested that
Val186, Val204 and Leu246 were dispersed in different
secondary structures; i.e. in the a-helix, in the b-sheet
and near the b-sheet, respectively (Fig. 7). The predic-
tion also supported the notion that these replacements
independently affected the oxygen-evolving activity. As
expected, the predicted structure of Arabidopsis PsbO2
was very similar to the structure of PsbO1 (data not
shown), and the substitution of amino acids between
PsbO1 and PsbO2 would not modify the overall struc-
ture of PsbO.
Discussion

The existence of two psbO genes enabled the isolation
of psbo1 which lacked psbO1 expression and showed
weak photosynthetic activity. In this study, we exam-
ined the functional role of two PsbOs in Arabidopsis.
Careful characterization of the PsbOs in psbo1 and the
wild type (Fig. 2A) showed that psbo1 showed a much
greater accumulation of PsbO2 than in the wild type.
This finding suggests a compensational mechanism that
stimulates the expression of PsbO2 when a functional
psbO1 gene is absent. The shortage of PsbOs caused a
photosynthetic defect in psbo1, especially in young
psbo1. However, mature psbo1 exhibited increased lev-
els of PsbO and improved photosynthetic activity, as
estimated from the increased F
v
⁄ F
m
, but the molecular
mechanism of this adaptation to a genetic defect is not
clear.
Our immunoblot analysis also showed that the
accumulation of PsbOs limited the levels of other
PSII proteins as well as the efficiency of PSII. This
result was consistent with findings in early studies
using higher-plant PSII that the extraction of PsbO
from PSII affected the stability of PsbA and the
assembly of other extrinsic proteins [22–24]. In this
respect, the quite different phenotypes observed in
psbo1 and psbO-deletion mutants of green algae and
cyanobacteria might lead to a clearer understanding

of the structure of the oxygen-evolving complex [6,7].
First, although a psbO-deletion mutant of the
Chlamydomonas reinhardtii [8] also had an enhanced
turnover of core PSII polypeptide, PsbA, a psbO-
deletion mutant of cyanobacteria accumulated PSII
reaction centers at nearly normals levels. The differ-
ence in the behavior of the deletion mutants between
higher plants, green algae and cyanobacteria suggests
a difference in the relationship between the extrinsic
proteins. Green algae and higher plants contain PsbP
and PsbQ, whereas cyanobacteria contain PsbV and
PsbU. There is an apparent difference between PsbP
and PsbV, as PsbP cannot bind to the PSII core or
function in the absence of PsbO, but PsbV alone
functions effectively in the absence of PsbO. Second,
PsbP was accumulated at normal levels in the
Chlamydomonas reinhardtii psbO-deletion mutant, sug-
gesting a difference between higher plants and green
algae in the regulation of levels of extrinsic proteins.
Functional dissection of PsbO1 and PsbO2 R. Murakami et al.
2170 FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS
Arabidopsis thaliana O1
Arabidopsis thaliana O2
Nicotiana tabacum
Solanum tuberosum
Lycopersicon esculentum
Pisum sativum
Oryza sativa
Spinacia oleracea
Volvox carteri

Chlamydomonas reinhardtii
Euglena gracilis
Bigelowiella natans
Synechocystis sp. PCC 6803
Anabaena sp. PCC 7120
Thermosynechococcus elongatus BP-1
Prochlorococcus marinus SS120
Prochlorococcus marinus MIT9313
Synechococcus sp.WH8102
Triticum aestivum
Fritillaria agrestis
*
Synechocystis sp. PCC 6803
Prochlorococcus marinus SS120
Prochlorococcus marinus MIT9313
Synechococcus sp.WH8102
Arabidopsis thaliana O1
Arabidopsis thaliana O2
Nicotiana tabacum
Solanum tuberosum
Lycopersicon esculentum
Pisum sativum
Oryza sativa
Spinacia oleracea
Volvox carteri
Chlamydomonas reinhardtii
Euglena gracilis
Bigelowiella natans
Anabaena sp. PCC 7120
Thermosynechococcus elongatus BP-1

Triticum aestivum
Fritillaria agrestis
*
Arabidopsis thaliana O1
Arabidopsis thaliana O2
Nicotiana tabacum
Solanum tuberosum
Lycopersicon esculentum
Pisum sativum
Oryza sativa
Spinacia oleracea
Volvox carteri
Chlamydomonas reinhardtii
Euglena gracilis
Bigelowiella natans
Synechocystis sp. PCC 6803
Anabaena sp. PCC 7120
Thermosynechococcus elongatus BP-1
Prochlorococcus marinus SS120
Prochlorococcus marinus MIT9313
Synechococcus sp.WH8102
Triticum aestivum
Fritillaria agrestis
*
Relative oxygen-evolving activity
0 20 40 60 80 100 120
PsbO1
PsbO2
V186S
V204I

L246I
186&204
186&246
204&246
AB
Fig. 6. Determination of the amino acid changes responsible for the alteration of protein function. (A) Alignments around the replacements
between PsbO1 and PsbO1-1-2. The alignments were made with
CLUSTAL W. Asterisks indicate substituted residues with Arabidopsis PsbO1
and PsbO2. Arabidopsis PsbO1 and PsbO2 are underlined. (B) Oxygen-evolving activities reconstituted with PsbO1, PsbO2, and mutated-
PsbOs. The maximum rate of oxygen evolution (100%) equals the rate measured for untreated PSII minus that for urea-washed PSII. Stand-
ard deviations were calculated from five measurements.
R. Murakami et al. Functional dissection of PsbO1 and PsbO2
FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS 2171
Interestingly, PsbO1 and PsbO2 showed different
biochemical activity at reconstituting oxygen-evolving
activity with urea-washed PSII isolated from spinach;
the restoration of oxygen-evolving activity with PsbO2
was about 80% of that with PsbO1, whereas the bind-
ing affinities of PsbO1 and PsbO2 were similar. A
recent study on the structure of PSII [19] suggested
that PSII has one copy of PsbO, but early works
[16,25] and our study (Fig. 3) showed that the
maximum oxygen-evolving activity reconstituted with
both PsbO was achieved at about two PsbOs per PSII.
It is not clear why about two PsbOs per PSII are nee-
ded for the maximum activity.
The chimeric PsbO derived from PsbO1 and PsbO2
clearly indicated that the surface charge was not crit-
ical for the different activities of PsbO1 and PsbO2,
whereas negative and positive charges of amino acids

have been reported to be important for the interaction
between PsbO and PSII [5,17,18]. The analysis using
the chimeric PsbOs showed that three amino acid
replacements at the C-terminus affected the restoration
of oxygen-evolving activity.
As predicted in Fig. 7, the three amino acids were
dispersed in PsbO. Interestingly, although Val186 was
not conserved in higher plants, green algae and cyano-
bacteria, its substitution with Ser decreased oxygen-
evolving activities. On the other hand, when Val204 of
PsbO1, which is conserved in higher plants except for
Arabidopsis PsbO2, was substituted with the Ile of
PsbO2, this mutated-protein restored the oxygen-evol-
ving activity better than PsbO1. It should be noted
that the amino acid at this position is substituted with
Ile in both green algae and cyanobacteria. By contrast,
Leu246 is conserved in all higher plants and green
algae, except in PsbO2. PsbO2 has Ile246, which is
substituted with Ile or Val in cyanobacteria, except for
Prochlorococcus marinus which has Lys. The replace-
ment of Leu with Ile was shown to result in a reduc-
tion of the restoration of oxygen-evolving activity. It
has been reported that Leu at position 246 was critical
for PsbO to bind PSII and the restoration of oxygen-
evolving activity in spinach [29]. Although our recon-
stitution experiment with PsbO2 showed a similar
affinity for PSII particles as PsbO1, the importance of
this amino acid residue for oxygen-evolving activity
was consistent with the results of previous works
[29,30]. The similar chemical properties of Leu and Ile

might explain their similar binding affinities. It has
also been reported that the C-terminal peptide of PsbO
(15 amino acids) competed with mature PsbO to bind
the PSII core, suggesting that the C-terminus of PsbO
plays an important role in the interaction with an
integral part of the PSII core [31]. Identification of the
importance of the three amino acid residues in PsbO,
especially V204I, could be useful for understanding the
efficiency of PSII.
In vivo and in vitro experiments have shown that
both a shortage of the PsbOs and a lower level of oxy-
gen-evolving activity resulted in the poor photosyn-
thetic activity and retarded growth in psbo1. In most
higher plants, such as spinach, pea and rice, PsbO is
encoded by only a single gene [26,32]. The existence of
a duplicated psbO gene in Arabidopsis should help us to
understand the molecular selection of duplicated genes.
L246I
V186S
V186S
V204I
L246I
A
B
Fig. 7. The location of three amino acid substitutions between
PsbO1 and PsbO2. (A) Stereoview of Thermosynechococcus elong-
atus PsbO. The corresponding positions of Val186, Val204 and
Leu246 in Thermosynechococcus elongatus PsbO were predicted
by SwissModel ( />html) and are shown in red and indicated by an arrow. (B) Stereo-
view of PsbO, PsbA and Mn clusters in Thermosynechococcus

elongatus. PsbO: green, PsbA: blue, Mn clusters: pink, The posi-
tions corresponding to Val-186, Val-204 and Leu-246 are shown in
red and indicated by an arrow. These figures were generated with
P
YMOL ( />Functional dissection of PsbO1 and PsbO2 R. Murakami et al.
2172 FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS
In Arabidopsis, psbQ was also encoded by two nuc-
lear genes (At4g21280 and At4g05180) with about
82% similarity at the gene level. On the other hand,
psbP (At1g06680) has a similar but nonfunctional gene
(At2g30790) (about 84% similarity at the gene level); it
encodes a much smaller protein than PsbP, suggesting
that the isogene of psbP recently lost its function. Simi-
larly, other subunits in PSII encoded by nuclear genes,
such as psbR and psbS, are encoded by a single gene.
The investigation of PsbQ isogenes should provide
another clue as to the physiological role of duplicated
genes for extrinsic proteins in the oxidation of water.
Although the duplicated genes would initially have
had the same functions, the accumulation of mutations
over time has resulted in either the functional loss of one
copy or functional divergence between them (our data
and [33]). It is not clear how much functional differenti-
ation exists between PsbO1 and PsbO2 and whether the
inferior psbO2 will lose its activity. At present, the data-
base (Brassica Genome Gateway, rc.
ac.uk) suggests that Brassica rapa and Brassica napus
have a psbO1-like gene (accession number; Brassica
rapa, BQ791144 and Brassica napus, AF139818) and a
psbO2-like gene (accession number; Brassica rapa,

BG543314 and Brassica napus, CD821869), indicating
that any potential evolutionary pressure to remove the
psbO2 gene is not strong. Indeed, PsbO2 might have
some physiological significance; for example, the poor
oxygen-evolving activity with PsbO2 might be advanta-
geous under excessive light conditions.
Conclusions
The psbO2 gene could not complement the defect in
the psbO1 gene for both quantitative and qualitative
reasons: a shortage of PsbOs and the poor oxygen-
evolving activity of PsbO2. The functional difference
between PsbO1 and PsbO2 in the restoration of
oxygen-evolving activity was ascribed to three amino
acid replacements at the C-terminus; Val186 in PsbO1
for Ser in PsbO2, Val204 for Ile, and Leu246 for Ile.
Experimental procedures
Plant growth and measurements of chlorophyll
fluorescence
Seeds of the wild-type and psbo1 were sown in soil after
cold treatment for 24 h at 4 °C. Seeds were obtained from
Lehle Seeds (Round Rock, TX, USA). They were grown at
23 ± 0.5 °C under white light (30 lmolÆphotons ⁄ m
)2
Æs)
with a 9-h light (from 09:00–18:00 h): 15-h dark cycle for
2 months.
The chlorophyll fluorescence parameter (F
m
) F
0

) ⁄ F
m
(¼ F
v
⁄ F
m
), was measured using a PAM2000 chlorophyll
fluorometer (Waltz, Effeltrich, Germany). The minimum
fluorescence at the open PSII centers, F
o
, was determined
by measuring light. F
m
(the maximum chlorophyll fluores-
cence at closed PSII centers in the dark) was measured by
applying a 1-s pulse of saturating white light.
Immunoblot analysis
Thylakoid membranes were isolated from wild-type and
psbo1 mutant leaves according to the method of Endo et al.
[34]. Protein samples were prepared in sample buffer
[50 mm Tris ⁄ HCl pH 6.8, 5% sodium dodecylsulfate (SDS),
5% glycerol, and 5% 2-mercaptoethanol], and analyzed by
15% SDS ⁄ PAGE to evaluate visually the level of protein
expression with Coomassie brilliant blue staining [35]. Thyl-
akoid membranes were loaded on a chlorophyll basis. The
protein analysis with respect to the protein content showed
the results similar to that on a chlorophyll basis. The thy-
lakoid proteins were also transferred to a polyvinylidene
difluoride membrane with a semidry blotting system [36],
and detected with rabbit antiserum against spinach PsbO,

kindly provided by the late A. Watanabe of the University
of Tokyo, or rabbit antiserum against spinach D1 protein
(PsbA), provided by Y. Yamamoto (Division of Integrated
Life Science, Kyoto University, Japan). The chlorophyll
concentration was measured by spectrophotometry in 80%
acetone [37].
The amounts of PsbO, PsbA and PsbP in plants were
quantified from immunoblots and bound PsbOs to urea-
washed PSII were quantified from Coomassie brilliant blue-
stained plates. Image analysis was performed using the
public-domain software nih-image (version 1.62).
Preparation of recombinant PsbOs
The recombinant PsbO1 and PsbO2 were expressed and
purified as described previously [38]. The expression vectors
for the chimeric proteins were produced by digestion of
psbO1- and psbO2-pET 21d + with BamHI, NotIor
PpuMI and by the ligation of resultant fragments. Site-
directed mutagenesis was performed according to the proto-
col for the Quick-change Site-directed Mutagenesis Kit
(Stratagene) and a Technical Review for Long polymerase-
chain reaction using KOD plus (TOYOBO, Osaka, Japan).
All mutant constructs were confirmed by DNA sequencing.
These mutant proteins were purified similar to the original
PsbO1 and PsbO2.
Reconstitution of PSII with recombinant PsbOs
Photosystem II membranes were isolated from spinach
leaves purchased at a local market according to the method
R. Murakami et al. Functional dissection of PsbO1 and PsbO2
FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS 2173
of Ghanotakis et al. [39]. Native extrinsic proteins in PSII

were extracted by incubation in urea-washed buffer (50 mm
Mes–NaOH (pH 8.3) and 3 m urea), to give urea-washed
PSII particles. Recombinant PsbOs were mixed with urea-
washed PSII particles and incubated at room temperature
(25 °C) for 1 h in the dark to reconstitute PSII [23]. The
concentration of PSII was estimated on the basis of chl
concentrations and a stoichiometry of 250 chlorophylls per
PSII complex.
Photosystem II activity was measured as oxygen
evolution at 25 °C with a Clark-type Oxygen electrode
(Hansatech, UK) in the presence of 2 mm DCBQ (dichloro-
p-quinone) as the electron acceptor for PSII. Red satura-
tion actinic light, at an intensity of 2000 lmolÆphotons ⁄
m
)2
Æs, was provided by an incandescent lamp that was used
in conjunction with an HA50 heat-absorbing filter and an
R-60 red optical filter (Kenko, Tokyo, Japan).
Alignment of PsbOs
Similarity was evaluated using the clustal w program
[40–42]. The alignment parameters used were: protein mass
matrix: BLOSUM series, gap opening penalty (GOP): 10.0,
gap extension penalty (GEP): 0.05 and Delay divergent
sequences: 40%.
Acknowledgements
This study was supported in part by a COE Scientific
Research Grant from the Ministry of Education, Cul-
ture, Sports, Science and Technology of Japan.
References
1 Barber J (1998) Photosystem two. Biochim Biophys Acta

1365, 269–277.
2 Rochaix JD & Erickson J (1988) Function and assembly
of photosystem II: genetic and molecular analysis.
Trends Biochem Sci 13, 56–59.
3 Debus RJ (1992) The manganese and calcium ions of
photosynthetic oxygen evolution. Biochim Biophys Acta
1102, 269–352.
4 Bricker TM & Frankel LK (1998) The structure and
function of the 33 kDa extrinsic protein of photosystem
II: a critical assessment. Photosynth Res 56, 157–173.
5 Seidler A (1996) The extrinsic polypeptide of photosys-
tem II. Biochim Biophys Acta 1277, 35–60.
6 Burnap RL & Sherman LA (1991) Deletion mutagenesis
in Synechocystis sp. PCC6803 indicates that the Mn-sta-
bilizing protein of photosystem II is not essential for O
2
evolution. Biochemistry 30, 440–446.
7 Mayes SR, Cook KM, Self SJ, Zhang Z & Barber J
(1991) Deletion of the gene encoding the Photosystem
II, 33 kDa protein from Synechocystis sp. PCC 6803
does not inactivate water-splitting but increases
vulnerability to photoinhibition. Biochim Biophys Acta
1060, 1–12.
8 Mayfield SP, Bennoum P & Rochaix JD (1987) Expres-
sion of the nuclear encoded OEE1 protein us required
for oxygen evolution and stability of photosystem II
particles in Chlamydomonas reinhardtii. EMBO J 6,
313–318.
9 Shikanai T, Munekage Y, Shimizu K, Endo T &
Hashimoto T (1999) Identification and characterization

of Arabidopsis mutants with reduced quenching of
chlorophyll fluorescence. Plant Cell Physiol 40,
1134–1142.
10 Murakami R, Ifuku K, Takabayashi A, Shikanai T,
Endo T & Sato F (2002) Characterization of an Arabi-
dopsis thaliana mutant with impaired psbO, one of
genes encoding 33-kDa protein of oxygen-evolving
complex, which shows retarded growth. FEBS Lett 523,
138–142.
11 The Arabidopsis Genome Initiative (2000) Analysis of
the genome sequence of the flowering plant Arabidopsis
thaliana. Nature 408, 796–815.
12 Bowers JE, Chapman BA, Rong J & Paterson AH
(2003) Unravelling angiosperm genome evolution by
phylogenetic analysis of chromosomal duplication
events. Nature 422, 433–438.
13 Ermolaeva MD, Wu M, Eisen JA & Salzberg SL (2003)
The age of the Arabidopsis thaliana genome duplication.
Plant Mol Biol 51, 859–866.
14 Burnap RL, Qian M, Shen JR, Inoue Y & Sherman LA
(1994) Role of disulfide linkage and putative intermole-
cular binding residues in the stability and binding
of the extrinsic manganese-stabilizing protein to the
photosystem II reaction center. Biochemistry 33,
13712–13718.
15 Betts SD, Ross JR, Hall KU, Pichersky E & Yocum
CF (1996) Functional reconstitution of photosystem II
with recombinant manganese-stabilizing protein contain-
ing mutations that remove the disulfide bridge. Biochim
Biophys Acta 1274, 135–142.

16 Betts SD, Ross JR, Pichersky E & Yocum CF (1997)
Mutation Val235Ala weakens binding of the 33-kDda
manganese stabilizing protein of photosystem II to one
of two sites. Biochemistry 36, 4047–4053.
17 Rivas JDL & Heredia P (1999) Structural predictions
on the 33 kDa extrinsic protein associated to the oxy-
gen-evolving complex of photosynthetic organism.
Photosynth Res 61, 11–21.
18 Miura T, Shen JR, Takahashi S, Kamo M, Nakamura
E, Ohta H, Kamei A, Inoue Y, Domae N et al. (1997)
Identification of domains on the extrinsic 33-kDa pro-
tein possibly involved in electrostatic with photosystem
II complex by means of chemical modification. J Biol
Chem 272, 3788–3798.
Functional dissection of PsbO1 and PsbO2 R. Murakami et al.
2174 FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS
19 Ferreira KN, Iverson TM, Maghlaoui K, Barber J &
Iwata S (2004) Architecture of the photosynthetic
oxygen-evolving center. Science 303, 1831–1838.
20 Schreiber U & Neubauer C (1987) The Polyphasic
rise of chlorophyll fluorescence upon onset of strong
continuous illumination: II. partial control by the
photosystem II donor side and possible ways of inter-
pretation. Z Naturforsch 42C, 1255–1264.
21 Oxborough K (2004) Imaging of chlorophyll a fluores-
cence: theoretical and practical aspects of an emerging
technique for the monitoring of photosynthetic perform-
ance. J Exp Bot 55, 1195–1205.
22 Miyao M & Murata N (1983) Role of the 33-kDa poly-
peptide in preserving Mn in the photosynthetic oxygen-

evolution system and its replacement by chloride ions.
FEBS Lett 164, 375–378.
23 Murata N & Miyao M (1985) Extrinsic membrane pro-
teins in the photosynthetic oxygen evolving complex.
Trends Biochem Sci 10, 122–124.
24 Yamamoto Y, Ishikawa Y, Nakatani E, Yamada M,
Zhang H & Wydrzynski T (1998) Role of and extrinsic
protein of photosystem II in the turnover of the reac-
tion center-binding protein D1 during photoinhibition.
Biochemistry 37, 1565–1574.
25 Betts SD, Hachigian TM, Pichersky E & Yocum CF
(1994) Role of disulfide linkage and putative intermole-
cular binding residues in the stability and binding of the
extrinsic manganese-stabilizing protein to the photosys-
tem II reaction center. Plant Mol Biol 26, 117–130.
26 Schubert M, Petersson UA, Haas BJ, Funk C, Schroder
WP & Kieselbach T (2002) Proteome map of the chlor-
oplast lumen of Arabidopsis thaliana. J Biol Chem 277,
8354–8365.
27 Kieselbach T, Bystedt M, Hynds P, Robinson C &
Schro
¨
der WP (2000) A peroxidase homologue and novel
plastocyanin located by proteomics to the Arabidopsis
chloroplast thylakoid lumen. FEBS Lett 480, 271–276.
28 Peltier JB, Emanuelsson O, Kalume DE, Ytterberg J,
Friso G, Rudella A, Liberles DA, So
¨
derberg L, Roe-
pstorff P, von Heijne G & Wijk KJ (2002) Central func-

tions of the lumenal and peripheral thylakoid proteome
of Arabidopsis determined by experimentation and gen-
ome-wide prediction. Plant Cell 14, 211–236.
29 Betts SD, Lydarkis-Simantris N, Ross JR & Yocum CF
(1998) The carboxyl-terminal tripeptide of the manga-
nese-stabilizing protein is required for quantitative
assembly into photosystem II and for high rates of oxy-
gen evolution activity. Biochemistry 37, 14230–14236.
30 Lydakis-Simantiris N, Betts SD & Yocum CF (1999)
Leucine 245 is a critical residue for folding and function
of the manganese stabilizing protein of photosystem II.
Biochemistry 38, 15528–15535.
31 Shutova T, Villarejo A, Zeits B, Klimov V, Gillbro T,
Samuelsson G & Renger G (2003) Comparative studies
on the properties of the extrinsic manganese-stabilizing
protein from higher plants and of a synthetic peptide of
its C-terminus. Biochim Biophys Acta 1604, 95–104.
32 Peltier JB, Friso G, Kalume DE, Roepstorff P, Nilsson
F, Adamska I & Wijk KJ (2000) Proteomics of the
chloroplast: systematic identification and targeting ana-
lysis of lumenal and peripheral thylakoid proteins. Plant
Cell 12, 319–341.
33 Gu Z, Steinmetz LM, Gu X, Scharfe C, Davis RW &
Li WH (2003) Role of duplicate genes in genetic robust-
ness against null mutations. Nature 421, 63–66.
34 Endo T, Mi H, Shikanai T & Asada K (1997) Donation
of electrons to plastquinone by NAD(P)H dehydrogen-
ase and by ferredoxin-quinone reductase in spinach
chloroplasts. Plant Cell Physiol 38, 1272–1277.
35 Leammli UK (1970) Cleavage of structural proteins

during the assembly of the head of bacteriophage T4.
Nature 227, 680–685.
36 Sambrook J, Fritsch EF & Maniatis T (1989) Molecular
Cloning: A Laboratory Manual. 2nd edn. Cold Spring
Harbor, NY.
37 Mackinnery G (1941) Absorption of light by chloro-
phyll solutions. J Biol Chem 140, 315–322.
38 Murakami R, Ifuku K, Takabayashi A, Shikanai T,
Endo T & Sato F (2004) Photosynthesis: fundamental
aspects to global perspectives. Proceedings of the 13th
International Congress on Photosynthesis. Article Num-
ber: ul388.
39 Ghanotakis D, Babcock GT & Yocum CF (1984) Struc-
tural and catalytic properties of the oxygen-evolving
complex. Biochim Biophys Acta 765, 388–398.
40 Thompson JD, Higgins DG & Gibson TJ (1994) CLUS-
TAL W: improving the sensitivity of progressive multi-
ple sequence alignment through sequence weighting,
position-specific gap penalties and weight matrix choice.
Nucleic Acids Res 22, 4673–4680.
41 Higgins DG, Thompson JD & Gibson TJ (1996) Using
CLUSTAL for multiple sequence alignments. Methods
Enzymol 266, 383–402.
42 Saitou N & Nei M (1987) The neighbor-joining method:
a new method for reconstructing phylogenetic trees.
Mol Biol Evol 4, 406–425.
R. Murakami et al. Functional dissection of PsbO1 and PsbO2
FEBS Journal 272 (2005) 2165–2175 ª 2005 FEBS 2175