Content uploaded by Christiane Funk
Author content
All content in this area was uploaded by Christiane Funk
Content may be subject to copyright.
AtFtsH6 is involved in the degradation of the
light-harvesting complex II during high-light
acclimation and senescence
Agnieszka Z
˙elisko*
†
, Maribel Garcı´a-Lorenzo*, Grzegorz Jackowski
†
, Stefan Jansson
‡
, and Christiane Funk*
§
*Departments of Biochemistry and Plant Biology and ‡Umeå Plant Science Centre, Umeå University, S-901 87 Umeå, Sweden; and †Department of Plant
Physiology, Adam Mickiewicz University, Institute of Experimental Biology, Al. Niepodlegos´ci 14, 61-713, Poznan´ , Poland
Edited by George H. Lorimer, University of Maryland, College Park, MD, and approved August 9, 2005 (received for review April 27, 2005)
Degradation of the most abundant membrane protein on earth,
the light-harvesting complex of Photosystem II (LHC II), is highly
regulated under various environmental conditions, e.g., light
stress, to prevent photochemical damage to the reaction center.
We identified the LHC II degrading protease in Arabidopsis thaliana
asaZn
2ⴙ
-dependent metalloprotease, activated by the removal of
unknown extrinsic factors, similar to the proteolytic activity di-
rected against Lhcb3 in barley. By using a reversed genetic ap-
proach, the chloroplast-targeted protease FtsH6 was identified as
being responsible for the degradation. T-DNA KO A. thaliana
mutants, lacking ftsH6, were unable to degrade either Lhcb3
during dark-induced senescence or Lhcb1 and Lhcb3 during high-
light acclimation. The A. thaliana ftsH6 gene has a clear orthologue
in the genome of Populus trichocarpa. It is likely that FtsH6 is a
general LHC II protease and that FtsH6-dependent LHC II proteol-
ysis is a feature of all higher plants.
membrane protein 兩photosynthesis 兩protease
During evolution, cells have developed a complex system of
molecular chaperones and proteases to control protein
quality and turnover and to prevent protein damage or minimize
its adverse effects on cell metabolism. Many factors trigger the
degradation of proteins, including changes in environmental
conditions, genetic mutations, and limitations to the availability
of cofactors. Despite the multitudinous examples of proteolytic
processes that take place inside the chloroplasts of higher plants
and the necessity of their regulation for cell viability, our
knowledge of the biochemical identity of chloroplast proteases,
their substrates and physiological significance is, to date, very
limited (1). For example, many questions concerning the turn-
over regulation of the two of the most common proteins on earth
[the soluble ribulose-1,5-bisphosphate-carboxylase兾oxygenase
and the membrane protein LHC II, the apoprotein of the main
light-harvesting complex of Photosystem II (PS II)] remain
unanswered.
LHC II is located in the thylakoid membrane, where it collects
energy from sunlight and transfers it, in the form of excitation
energy, to the PS II reaction center. Its structure and function
have been studied extensively. The excitation energy transfer
within LHC II occurs on a time scale of femtoseconds (2, 3) and
depends on the type, orientation, and exact position of the
pigments associated with the protein moiety. Cr ystallographic
data have revealed eight chlorophyll (chl) a, six chl b, two luteins,
one neoxanthine, and one violaxanthine in the protein scaffold
(4–6). The functional unit of LHC II is a trimer, representing
various permutations of Lhcb1–3 apoproteins, each of ⬇25 kDa.
The genes coding for the three apoproteins are typically found
as multiple copies in the genomes of higher plants (7). In
Arabidopsis thaliana, for example, there are five copies of lhcb1
and three of lhcb2, but only one copy of the lhcb3 gene (8). Lhcb1
accounts for ⬇60% of the total LHC II apoprotein content,
whereas Lhcb2 and Lhcb3 account for ⬇30% and 10%, respec-
tively (9). Lhcb1–3 apoproteins are similar enough to form
homo- and heterotrimers in various combinations. LHC II
apoproteins are highly conserved; thus, they can be expected to
have distinct functional roles. Indeed, LHC II populations with
different Lhcb1兾Lhcb2 ratios differ significantly with respect to
their PS II efficiency (10) and cation-dependent thylakoid
adhesion-promoting activity (11).
Although the main function of LHC II is energy collection and
transfer, it also is involved in the distribution of excitation energy
between PS II and Photosystem I (PS I) (12) and in preventing
damage to the photosynthetic machinery when there is too much
light (13). All of these functions necessitate a very efficient and
highly effective regulation of LHC II apoproteins during short-
and long-term acclimation to different environments (14–16).
However, despite much research, very little is known about the
proteases involved in the regulation of LHC II apoproteins. It
has been suggested that a light-activated serine兾cysteine-type
protease is involved in LHC II degradation during bean chlo-
roplast development (17, 18) and in the degradation of LHC II
trimers that have preaccumulated in mature bean thylakoid
membranes (19). Another serine兾cysteine-type proteolytic en-
zyme has been implicated in the ATP-dependent degradation of
LHC II apoproteins during high-light acclimation of spinach
plants (14, 20, 21). Both enzymes are claimed to be peripherally
attached to the stromal side of the thylakoid membrane. Fur-
thermore, an overexpressed peptide with the N-terminal se-
quence of pea Lhcb1 has been shown to be degraded in vitro by
a stromal serine-type glutamyl endopeptidase and by a cysteine-
type enzyme attached to the thylakoid membrane (22). It has
also been suggested that the SppA protease might be involved in
the degradative regulation of LHC II apoproteins (23). However,
the LHC II protease(s) have not been conclusively identified.
Recently, we identified a metalloprotease activity responsible
for Lhcb3 degradation in barley during dark-induced senescence
(24). This integral thylakoid membrane protease has a pH
optimum at 7.8, requires either Zn
2⫹
or Mg
2⫹
ions for activation,
and is stimulated by ATP. Although the protease is present in
control leaves, Lhcb3 degradation takes place only during se-
nescence, perhaps after removal of as yet unidentified factors on
the stromal side of the thylakoid membrane (24). A reversed
genetic approach was used here to demonstrate that FtsH6 is a
protease involved in LHC II degradation.
Materials and Methods
Plant Material. WT, a mutant lacking FtsH5 (var1), and T-DNA
KO mutants lacking FtsH6 (SALK㛭012429) and FtsH11
(SALK㛭533047) of A. thaliana (var. Colombia) were grown
under short-day conditions (8 h light兾16 h dark with a 23°C兾16°C
temperature cycle) at a light intensity of 150
mol䡠m
⫺2
䡠s
⫺1
.
This paper was submitted directly (Track II) to the PNAS office.
Abbreviations: chl, chlorophyll; HL, high light intensity.
§To whom correspondence should be addressed. E-mail: christiane.funk@chem.umu.se.
© 2005 by The National Academy of Sciences of the USA
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0503472102 PNAS
兩
September 20, 2005
兩
vol. 102
兩
no. 38
兩
13699–13704
PLANT BIOLOGY
Mutant seeds were obtained from the Nottingham Arabidopsis
Stock Centre. Senescence of A. thaliana leaves was induced by
covering individual leaves with foil, while keeping the rest of the
plant under normal light conditions for 2, 4, and 6 days.
A. thaliana plants were acclimated to elevated irradiance by
incubation at 600
mol䡠m
⫺2
䡠s
⫺1
for 24, 48, and 72 h. Nonsenesc-
ing and nonacclimated plants were used as controls. A homozy-
gous mutation in ftsH5 (var1) gives rise to variegated plants.
Homozygous mutations were confirmed by using PCR with the
forward primer TCT TTC CCT CTC TCC AAC AT and the
reverse primer GCT TTT TCG CAA CCT CGT AA for the
ftsH6 deletion mutant, and the forward primer TCT CTC TCT
TTA T TT CTC TT and the reverse primer TCT GCC TGT GCT
ATC TAA AT for the ftsH11 deletion mutant. Amplification of
the insertion was performed by using the primer GCG TGG
ACC GCT TGC TGC AAC T.
In Vitro
Lhcb1 and Lhcb3 Degradation System. Thylakoids from
fresh, senescing and, acclimated leaves were isolated as de-
scribed in ref. 25, adding 0.2% BSA to the homogenization and
hypotonic buffers. The thylakoids were washed for 30 min by
using 0.5 M NaCl, then resuspended at a concentration of 0.6 mg
of chl per ml in ‘‘optimal incubation medium’’ (24) and incubated
in darkness at 25°C for 0–6 h. To inhibit metalloproteases,
thylakoids were incubated for6hinthepresence of 0.2 mM
phosphoramidon.
Determination of Divalent Metal Cations Essential for Proteolytic
Activity. Thylakoids were preincubated for4hinthepresence of
5 mM EDTA at 0°C, then washed in an incubation medium
without Mg
2⫹
and Na
2⫹
ions. After suspension at a concentra-
tion of 0.6 mg of chl per ml in an incubation medium containing
0.02 mM Zn
2⫹
, they were incubated in darkness at 25°C for 6 h.
SDS兾PAGE, Immunoblotting, and Integration Densitometry. SDS兾
PAGE was performed as described in ref. 26, using 14%
acrylamide. For the in vivo experiments, samples were loaded at
a concentration of 3
g of chl per lane. In the in vitro experi-
ments, the same amount of sample was loaded, corresponding to
3
g of chl of the time point T0. After electrophoresis, the
polypeptides were electrotransferred to nitrocellulose sheets and
immunostained with monospecific anti-Lhcb1 or anti-Lhcb3
antibodies, as described in ref. 27. They were made visible by
using enhanced chemiluminescence. For quantification of the
immunostained bands, the films were scanned by using GEL-PRO
ANALYZER 3.2 (Media Cybernetics, Silver Spring, MD). For each
individual immunoblot, the range of linearity of the immunore-
sponse was checked, and only those immunoblots in which the
sample dilution factor remained within the range of proportion-
ality in all integrated areas were taken into consideration.
chl Determination. chl concentration was assayed according to the
method in ref. 28.
Results
Senescence-Dependent Degradation of Lhcb3. During senescence in
barley, Lhcb3 has been shown to be the main proteolytic target
of all light-harvesting proteins (29). To compare the degradation
of Lhcb3 during senescence in A. thaliana to that described for
barley, several leaves of A. thaliana plants were covered with
aluminum foil. During 6 days of dark-induced senescence of the
A. thaliana leaves, chl disappeared gradually as indicated by
yellowing of these leaves (Fig. 1A). chl abundance was reduced
to 95%, 52%, and 47% of initial levels after 2, 4, and 6 days of
senescence, respectively (Fig. 1B). Simultaneously, the abun-
dance of Lhcb3 in vivo dropped significantly, to 77% and 46% of
the initial amount after 2 and 6 days of senescence, respectively
(Fig. 1B). Loading of the gels was performed according to chl
Fig. 1. A. thaliana senescence assay. Plants with covered leaves were grown
under normal light conditions for 2, 4, and 6 days. They were then photo-
graphed after the covers were removed. (A) Senescing leaves are marked with
arrows. (B) Degradation kinetics of chl and Lhcb3 during dark-induced senes-
cence. (C–E) Effect of phosphoramidon (C), endogenous Zn2⫹ions (D), and
NaCl (E), a factor releasing extrinsic thylakoid membrane proteins on Lhcb3
degradation in thylakoid membranes, isolated from leaves after 2 days of
senescence and incubated in vitro for 0–6 h. Mean values of four to six
experiments are presented. The abundance of Lhcb3 at 0 h was taken as 100%.
13700
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0503472102 Z
˙elisko et al.
amount in the sample; the real Lhcb3 degradation in the leaf on
fresh weight basis therefore is more pronounced (only 21% left).
Because gels are more precisely loaded on an equal chl basis, we
will in the following express LHC II degradation on a chl basis;
as a biochemical assay, this would produce the most reliable
results. It has to be kept in mind that the true values of LHC II
degradation, on fresh weight or per leaf basis, is larger.
To study the senescence-induced degradation of Lhcb3 fur-
ther, we used the in vitro system for incubating thylakoids
isolated from senescing barley leaves that is described in ref. 24.
When thylakoids from A. thaliana leaves that had been senescing
for 2 days were isolated and incubated in vitro for 6 h, a fraction
of Lhcb3 was found to have been degraded (data not shown). To
confirm that, in A. thaliana leaves, the Lhcb3-degrading protease
is a zinc-dependent metalloprotease, integral to the thylakoid
membrane, thylakoids were incubated in vitro in the presence of
phosphoramidon, an inhibitor of metalloproteases. This treat-
ment inhibited two-thirds of the Lhcb3 degradation (Fig. 1C),
confirming that the Lhcb3 reduction in senescing A. thaliana
leaves is due to degradation by a metalloprotease.
To identify the metal essential for the activity of the A. thali-
ana Lhcb3-degrading protease, thylakoids isolated from senesc-
ing leaves were preincubated in the presence of EDTA to remove
divalent metal cations. After removal of EDTA, the thylakoids
were incubated either with or without Zn
2⫹
(Fig. 1D). In the
absence of the metal, no anti-Lhcb3 proteolytic activity was
observed; however, Zn
2⫹
ions were able to restore Lhcb3
degradation activity to levels similar to the control. To deter-
mine the location of the Lhcb3-degrading protease in the
thylakoid, membranes isolated from senescing leaves were
washed with NaCl (0.5 M) to release extrinsic thylakoid mem-
brane proteins. Washing of thylakoids caused a 4-fold acceler-
ation of Lhcb3 degradation, indicating that the degrading pro-
tease was neither removed nor inactivated (Fig. 1E) and that it
is integral to the thylakoid membrane. However, other factors兾
proteins, shielding this protease, might have been removed by
the salt wash, causing enhanced activation of the protease. Taken
together, these data show that the protease degrading Lhcb3
during senescence of A. thaliana leaves is similar to the protease
found in barley leaves (24).
Degradation of Lhcb1 During Acclimation to High Light Intensity (HL).
To determine whether the senescence-related metalloprotease
degrading Lhcb3 was responsible for degradation of LHC II
under other conditions as well, proteolysis of Lhcb1 during HL
acclimation was investigated. Lhcb1 is the LHC protein that most
rapidly is degraded during HL acclimation (16). A. thaliana
leaves respond to HL conditions by a gradual loss of chl; the chl
content decreased to 79%, 78%, and 66% of the initial value
after 24, 48, and 72 h of HL acclimation, respectively (Fig. 2A).
The decline in chl was accompanied by the disappearance of
Lhcb1, its abundance on chl basis dropping to 76%, 62%, and
44% of the initial value after 24, 48, and 72 h, respectively (Fig.
2A), corresponding to 60%, 48%, and 29% on fresh-weight basis.
When thylakoids were isolated from A. thaliana leaves accli-
mated to HL for 24 h and incubated in vitro for 6 h, 31% of Lhcb1
was found be degraded (Fig. 2B). Phosphoramidon inhibited this
reduction; only 5% of Lhcb1 were degraded in its presence (Fig.
2B). Although the absence of Zn
2⫹
inhibited the reduction in
Lhcb1 levels, degradation resumed after reincubation of thyla-
koids in the presence of Zn
2⫹
(83% of initial value, 77% in the
control) (Fig. 2C). Thus, HL acclimation-related Lhcb1 reduc-
tion is, like senescence-induced Lhcb3 degradation, caused by a
Zn
2⫹
metalloprotease. As demonstrated for Lhcb3, Lhcb1 deg-
radation was accelerated after washing the thylakoids with NaCl
(Fig. 2D), supporting the notion that the degrading protease is
an integral membrane protein, which is shielded by extrinsic
factors.
Identification of the Protease Involved in Degradation of Lhcb1
and兾or Lhcb3. Based on the biochemical properties of the Lhcb1
and Lhcb3 degrading protease, suitable candidates for this
enzyme are products of A. thaliana genes coding for chloroplast-
targeted homologs of eubacterial FtsH protease and transmem-
brane processing metallopeptidases. Both of these types of
proteins are zinc-dependent proteases, integrally associated with
the thylakoid membrane (30, 31).
Fig. 2. Lhcb1 degradation during acclimation to elevated irradiance.
A. thaliana plants were grown in HL (600
mol䡠m⫺2䡠s⫺1) for 24, 48, and 72 h,
and the following phenomena were monitored. (A) Degradation kinetics of
chl and Lhcb1 during acclimation to HL. (B–D) Effect of phosphoramidon (B)
and of endogenous Zn2⫹ions (C) and NaCl (D), a factor releasing extrinsic
thylakoid membrane proteins on Lhcb1 degradation in thylakoid membranes,
isolated from leaves acclimated to HL for 24 h and incubated in vitro for 0–6
h. Mean values of four to six experiments are presented. The abundance of
Lhcb1 at 0 h was taken as 100%.
Z
˙elisko et al. PNAS
兩
September 20, 2005
兩
vol. 102
兩
no. 38
兩
13701
PLANT BIOLOGY
In Escherichia coli, FtsH is a membrane-bound, ATP-
dependent metalloprotease that belongs to the AAA-protein
family (32). In the genome of A. thaliana there are 12 genes
encoding FtsH proteases, products of 9 of which have proved to
be located in the chloroplast (33). FtsH1, FtsH2, FtsH5, and
FtsH8 have been identified as integral thylakoid proteins, ori-
ented so that their ATP-binding domain and catalytic zinc-
binding site face the stroma (30, 34).
Using a reversed genetic approach, we examined whether a
chloroplast-targeted FtsH is involved in LHC II degradation.
Because FtsH1, FtsH2, FtsH5, and FtsH8 have partially over-
lapping functions, we tested a loss-of function mutation of one
of these proteases, FtsH5. We also identified A. thaliana T-DNA
KO mutants for ftsH6 and ftsH11; seeds from these lines were
germinated and plants selfed to obtain plants homozygous for
the mutation. Thus, homozygous plants lacking FtsH5 (var1,
At5g42270), FtsH6 (At5g15250), and FtsH11 (At5g53170) were
obtained. These mutants, as well as WT plants, were grown
under standard conditions, and two LHC II protease assays were
performed. ⌬FtsH6 and ⌬FtsH11 plants did not show any visible
change in the plant phenotype [not in growth, development, or
pigmentation (data not shown)]; however, lack of the ftsH5 gene
product caused, as described in ref. 35, leaf variegation. Using
the LHC II protease assay, var1 (⌬FtsH5) as well as ⌬FtsH11
mutants were able to degrade Lhcb3 during senescence (Fig. 3
Aand B) and Lhcb1 during acclimation (Fig. 3 Aand C). The
minor impact on Lhcb1 degradation in the ⌬FtsH5 mutant
during high-light acclimation (only 80% of Lhcb1 is degraded
compared with W T) is consistent with its proposed role in
degrading photodamaged subunits in PS II (36). Although the
role of FtsH11 is not known, our data show that it has no major
function in LHC II degradation. However, in the ⌬FtsH6
mutant, the level of Lhcb3 did not change during senescence
(Fig. 3 Aand B). Clearly, therefore, FtsH6 is involved in Lhcb3
degradation. In addition, the level of Lhcb1 did not decline in the
⌬FtsH6 mutant during HL acclimation (Fig. 3 Aand C).
Therefore, we conclude that the FtsH6 protease is involved in
both senescence-induced degradation and HL acclimation, and
that it degrades both Lhcb3 and Lhcb1 during these processes.
If FtsH6 is a key protease for LHC II degradation during
senescence and acclimation, the decrease in chl content during
senescence and HL acclimation should, as a consequence, be
retarded in the ⌬FtsH6 mutant. Indeed, when chl levels were
measured after 6 days of dark-induced senescence, WT plants
had lost 53% of their chl, whereas ⌬FtsH6 plants had only lost
33%. Similarly, 6 days of HL acclimation resulted in a 34% loss
of chl in WT plants but only a 20% loss in ⌬FtsH6 mutants. The
loss of chl in the absence of FtsH6 is due to degradation of other
chl-binding proteins (LHCI proteins and the chl abinding
proteins of PS I and PS II).
To study the degradation of Lhcb proteins in vivo,WTand
FtsH6 mutant plants were subjected to HL for 24 h. Immuno-
labeling on thylakoids isolated from these leaves shows degra-
dation of Lhcb1 as well as Lhcb3 in WT, whereas both these
proteins resist degradation in ⌬FtsH6 plants (Fig. 4). The
apparent accumulation of the LHCII apoproteins is due to
loading of the SDS兾PAGE on chl basis.
Given its important role in LHC II turnover, the FtsH6
protease should have been conserved during evolution. We
identified the genes coding for FtsH proteases in the recently
sequenced genome of Populus trichocarpa (http:兾兾genome.jgi-
psf.org兾Poptr1兾Poptr1.home.html). By comparing the FtsH
gene families of A. thaliana and Populus (12 vs. 16 genes), we
were able to identify a clear FtsH6 orthologue in Populus
(fgenesh4㛭pg.C㛭LG㛭XVII000398; M.G.-L., S.J., and C.F., un-
published work). The genes coding for FtsH6 in the annual plant
and the tree showed a 78% similarity at the protein level. This
resemblance is reflected in the overall sequence comparison
(data not shown). The ATP-binding motifs, the zinc-binding
motif, and the second region of homology, which are essential for
proteolytic activity of the enzyme, are extremely conserved.
Furthermore, the conser ved lumenal domain identified in ref. 37
was found in FtsH6 of both A. thaliana and Populus. This
sequence seems to be restricted to FtsH homologues from
organisms performing oxygenic photosynthesis and in A. thali-
ana is present in FtsH2 (VAR2), FtsH8 (Chr1), and FtsH6
(Chr5) (37).
Because of a lack of a specific anti-FtsH6 antibody, we were
not able to study the expression of this protease in vivo under
Fig. 3. Degradation of Lhcb proteins in WT and FtsH5, FtsH6, and FtsH11
mutants. (A) Degradation of Lhcb proteins in WT and FtsH5, FtsH6, and FtsH11
mutants. (B) Degradation of Lhcb3 in thylakoid membranes isolated from senesc-
ing leaves. (C) Degradation of Lhcb1 in thylakoid membranes isolated from leaves
acclimated to HL. Values represent the means of four to six experiments. The
degradation of proteins (Lhcb1 or Lhcb3) in WT was taken as 1.
Fig. 4. In vivo degradation of Lhcb1 and Lhcb3 during HL acclimation in WT
and FtsH6 mutant. The abundance of proteins was analyzed in thylakoids
isolated from control leaves (0 h) as well as leaves acclimated to HL (600
mol䡠m⫺2䡠s⫺1for 24 h) by using antibodies directed against Lhcb1 and Lhcb3,
respectively.
13702
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0503472102 Z
˙elisko et al.
senescence or high-light acclimation. Therefore, to obtain more
information about the expression characteristics of the A. thali-
ana FtsH6 gene, we investigated the public DNA microarray data
available at www.genevestigator.ethz.ch. It appears that the
FtsH6 protease is mainly expressed in leaves, although its
expression level is rather low. Expression is not restricted to HL
conditions or senescence, although in older and senescing leaves
the expression seems to be higher than in young leaves. The
highest expression was found in developing rosettes. Further-
more, the expression has been shown to be up-regulated during
ethylene treatment, zinc treatment, senescence, heat stress, and
osmotic or salt stress, but no change has been observed during
cold stress.
Discussion
We report here that FtsH6, a metalloprotease belonging to the
FtsH family, is involved in the degradation of Lhcb1 during HL
acclimation and Lhcb3 during dark induced senescence. It seems
likely that FtsH6 is a general LHC II protease that degrades
various LHC II apoproteins when they are not needed, not just
during HL acclimation and senescence. Because the biochemical
features of the protease in barley and A. thaliana are similar and
the A. thaliana ftsH6 gene has such a clear orthologue in Populus,
we believe that FtsH6-dependent LHC II proteolysis is a feature
of all higher plants.
A phylogenetic and biochemical analysis of the A. thaliana
ftsH genes indicates the existence of four pairs of closely
related genes: ftsH1 and ftsH5,ftsH2 and ftsH8,ftsH7 and ftsH9,
and the mitochondrial ftsH3 and ftsH10 genes (38). The FtsH6
sequence is related to the ‘‘VAR1兾VAR2’’ group (consisting
of the FtsH1, -2, -5, and -8 proteins), whose members have
partially redundant functions and are involved in the biogen-
esis of chloroplasts and the degradation of photodamaged
PsbA protein (33, 37–42). From an evolutionary perspective,
it seems likely that PsbA degradation was the original function
of FtsH6 and that, after ancient gene duplication, the FtsH6
gene was neofunctionalized and evolved to recognize other
substrates, namely LHC II apoproteins. The set of loss-of-
function mutants now available makes it possible to examine
whether FtsH6 affects other photosynthesis-related proteins;
this is quite likely because other chloroplast proteases seem to
act on several substrates. Clp protease, for example, degrades
both PsbO (43) and polypeptides of the cytochrome b
6
f
complex (44), whereas DegP1 protease degrades both plasto-
cyanine and PsbO (45). Similarly, the chloroplast processing
enzyme (CPE) removes transit peptides of numerous chloro-
plast polypeptides (46).
It should be noted that our data do not directly prove that
LHC II apoproteins are the substrates of FtsH6 in vivo. Although
this is the simplest explanation of our data, it is still possible that
FtsH6 acts on another substrate, somehow activating the ‘‘real’’
LHC II protease. It is also possible that FtsH6 cooperates with
other FtsH proteases, coexisting with them in an enzymatic
heterocomplex; the occurrence of heterocomplexes comprising
FtsH2 and FtsH5 has been demonstrated experimentally (33).
FtsH6 may cooperate with members of other families of chlo-
roplast proteases as well; according to current thinking, the
proteolysis of photodamaged PsbA is an example of the coop-
erative action of proteases belonging to different families. It has
been suggested that DegP2 (a member of the DegP兾HtrA family
of serine proteases) performs the primary cleavage event, after
which the two breakdown products are removed by one or more
members of the FtsH VAR1兾VAR2 group (reviewed in ref. 1).
A similar scenario might also describe the degradation of LHC II
apoproteins.
Several researchers have reported other protease activities
involved in LHC II degradation in vivo and in vitro. These
proteases are localized peripherally at either the stroma
thylakoid membranes (19) or at the outer surface of the grana
thylakoids (14, 20, 38). Their main substrate during chl defi-
ciency conditions (17, 18) and during acclimation to HL (14,
20, 38) was found to be the LHC II monomer. LHC II trimer
(19) or LHC II N-terminal peptide (22) preaccumulated in
mature chloroplasts. Some of these protease activities may
cooperate with FtsH6 in the degradation of Lhcb1 and Lhcb3.
In E. coli, FtsH was found to be able to degrade both soluble
and membrane proteins, initiating proteolysis at either the N
terminus or C terminus (47). The substrate must be partially
unfolded, because FtsH lacks a robust unfoldase activity (48).
A feature of the energy-dependent proteolytic reaction is that
it occurs in a progressive manner along the polypeptide chain
(49). It is, therefore, also possible that during LHC II degra-
dation, a (serine- or cysteine-type) protease cleaves Lhcb1 or
Lhcb3 before FtsH6 continues the degradation. However, the
kinetics of FtsH6-catalyzed, HL-dependent degradation of
LHC II seem to be quicker than those of the activity reported
in ref. 20, in which was observed ATP-dependent degradation
of the LHC II monomer only after a lag phase of 48–72 h.
These data suggest that FtsH6-mediated proteolysis is the
primary step, but it is also possible that there are multiple
proteolytic activities that can act on LHC II. If, in the future,
mutants lacking other proteolytic activities are isolated, the
relative importance of these proteases in LHC II degradation
can be addressed.
It has been found that the nonphosphorylated form of LHC
II is the substrate for degradation (14), and that loss of pigments
destabilizes the apoprotein (50). The N-terminal domain of LHC
II was found to be essential for recognition of the protease (21).
These findings are consistent with our earlier data showing that
the protease degrading Lhcb3 is present in sufficient quantities
even under conditions where LHC II is stable, but either the
substrate or protease has to be activated in some way (24).
Enhanced degradation by FtsH6 has been shown after salt-
washing of the thylakoid membrane. The extrinsic proteins,
shielding the stromal part of FtsH6 (or Lhcb1 and -3), might be
the regulating factors. Their removal, either by dissociation or by
another protease, might trigger LHC II degradation. In vivo, this
trigger must in some way sense the ‘‘surplus’’ of antenna size.
Both our data (24) and those of other authors (14, 20, 21)
indicate that the substrate availability (or structure), rather than
the presence of the protease per se, determines the proteolytic
rate in chloroplasts.
Because of the lack of a specific anti-FtsH6 antibody, we
were unable to directly investigate the expression of FtsH6 in
WT plants during HL acclimation or senescence. According to
published data, the expression level of FtsH6 is low, both at the
mRNA and protein level (51), in plants grown under normal
conditions. It has even been suggested that it is a pseudogene
(38). Interestingly, ⌬FtsH6 mutants have previously been
analyzed and found not to be impaired in their recovery from
photoinhibition in vivo (33). We show in this study that FtsH6
is involved in the degradation of LHC II, during both HL
acclimation and senescence. The protease appears to be
constitutively present in small amounts in the thylakoid,
apparently not increasing in quantity during conditions when
LHC II degradation occurs. It is possible that this is a general
theme in chloroplast proteolysis; perhaps the turnover of many
chloroplast proteins is mainly determined by the structure of
the substrate, rather than the availability of the protease itself.
Further studies with mutants lacking the different chloroplast
proteases may verify this hypothesis.
We thank the Swedish Research Council, the Swedish Foundation for
Strategic Research, the Swedish Research Council for Environment,
Agricultural Sciences, and Spatial Planning, and the Carl Tryggers
Foundation for financial support.
Z
˙elisko et al. PNAS
兩
September 20, 2005
兩
vol. 102
兩
no. 38
兩
13703
PLANT BIOLOGY
1. Adam, Z. & Clarke, A. K. (2002) Trends Plant Sci. 7, 451–456.
2. Peterman, E. J., Gradinaru, C. C., Calkoen, F., Borst, J. C., van Grondelle, R.
& van Amerongen, H. (1997) Biochemistry 36, 12208 –12215.
3. Croce, R., Muller, M. G., Bassi, R. & Holzwarth, A. R. (2001) Biophys. J. 80,
901–915.
4. Kuhlbrandt, W., Wang, D. N. & Fujiyoshi, Y. (1994) Nature 367, 614 –621.
5. Liu, Z., Yan, H., Wang, K., Kuang, T., Zhang, J., Gui, L., An, X. & Chang, W.
(2004) Nature 428, 287–292.
6. Standfuss, J., Terwisscha van Scheltinga, A. C., Lamborghin i, M. & Kuhl-
brandt, W. (2005) EMBO J. 24, 919–928.
7. Jansson, S. (1994) Biochim. Biophys. Acta 1184, 1–19.
8. Jansson, S. (1999) Trends Plant Sci. 4, 237–241.
9. Jackowski, G., Kacprzak, K. & Jansson, S. (2001) Biochim. Biophys. Acta 1504,
340–345.
10. Walters, R. G. & Horton, P. (1999) Photosynth. Res. 61, 77– 89.
11. Jackowski, G. & Kluck, E. (1993) Photosynthetica 29, 401–407.
12. Allen, J. F. & Forsberg, J. (2001) Trends Plant Sci. 6, 317–326.
13. Bassi, R. & Caffarri, S. (2000) Photosynth. Res. 64, 243–256.
14. Yang, D. H., Webster, J., Adam, Z., Lindahl, M. & Andersson, B. (1998) Plant
Physiol. 118, 827–834.
15. Rintamaki, E., Martinsuo, P., Pursiheimo, S. & Aro, E.-M. (2000) Proc. Natl.
Acad. Sci. USA 97, 11644 –11649.
16. Jackowski, G., Olkiewicz, P. & Z
˙elisko, A. (2003) J. Photochem. Photobiol. B
70, 163–170.
17. A nastassiou, R. & Argyroudi-Akoyunoglou, J. H. (1995) Photosynth. Res. 43,
241–250.
18. Tziveleka, L. A. & Argyroudi-Akoyunoglou, J. H. (1998) Plant Physiol. 117,
961–970.
19. Georgakopoulos, J. H., Sokolenko, A., Arkas, M., Sofou, G., Herrmann, R. G.
& Argyroudi-Akoyunoglou, J. H. (2002) Biochim. Biophys. Acta 1556, 53–64.
20. Lindahl, M., Yang, D.-H. & Andersson, B. (1995) Eur. J. Biochem. 231,
503–509.
21. Yang, D. H., Paulsen, H. & Andersson, B. (2000) FEBS Lett. 466, 385–388.
22. Forsberg, J., Strom, J., Kieselbach, T., Larsson, H., Alexciev, K., Engstrom, A.
& Akerlund, H. E. (2005) Physiol. Plant. 123, 21–29.
23. Lensch, M., Herrmann, R. G. & Sokolenko, A. (2001) J. Biol. Chem. 276,
33645–33651.
24. Z
˙elisko, A. & Jackowski, G. (2004) J. Plant Physiol. 161, 1157–1170.
25. Harrison, M. A. & Melis, A. (1992) Plant Cell Physiol. 33, 627–637.
26. Laemmli, U. K. (1970) Nature 227, 680– 685.
27. Jansson, S., Stefansson, H., Nystrom, U., Gustafsson, P. & A lbertsson, P. A.
(1997) Biochim. Biophys. Acta 1320, 297–309.
28. Arnon, D. (1949) Plant Physiol. 24, 1–15.
29. Rosiak-Figielek, B. & Jackowski, G. (2000) Aust. J. Plant Physiol. 27, 245–251.
30. Lindahl, M., Tabak, S., Cseke, L., Pichersky, E., Andersson, B. & Adam, Z.
(1996) J. Biol. Chem. 271, 29329–29334.
31. Lew is, A. P. & Thomas, P. J. (1999) Protein Sci. 8, 439–442.
32. Langer, T. (2000) Trends Biochem. Sci. 25, 247–257.
33. Sakamoto, W., Zaltsman, A., Adam, Z. & Takahashi, Y. (2003) Plant Cell 15,
2843–2855.
34. Sinvany-Villalobo, G., Davydov, O., Ben-Ari, G., Zaltsman, A., Raskind, A. &
Adam, Z. (2004) Plant Physiol. 135, 1336–1345.
35. Sakamoto, W. (2003) Genes Genet. Syst. 78, 1–9.
36. Sakamoto, W., Tamura, T., Hanba-Tomita, Y., Sodmergen & Murata, M.
(2002) Genes Cells 7, 769–780.
37. Bailey, S., Thompson, E., Nixon, P. J., Horton, P., Mullineaux, C. W., Robinson,
C. & Mann, N. H. (2002) J. Biol. Chem. 277, 2006–2011.
38. Yu, F., Park, S. & Rodermel, S. R. (2004) Plant J. 37, 864– 876.
39. Chen, M., Choi, Y., Voytas, D. F. & Rodermel, S. (2000) Plant J. 22, 303–313.
40. Takechi, K., Sodmergen, K., Murata, M., Motoyoshi, F. & Sakamoto W. (2000)
Plant Cell Physiol. 41, 1334–1346.
41. Ostersetzer, O. & Adam, Z. (1997) Plant Cell 9, 957–965.
42. Lindahl, M., Spetea, C., Hundal, T., Oppenheim, A. B., Adam, Z. & Andersson,
B. (2000) Plant Cell 12, 419– 431.
43. Halperin, T., Ostersetzer, O. & Adam, Z. (2001) Planta 213, 614 –619.
44. Majeran, W., Wollman, F. A. & Vallon, O. (2000) Plant. Cell 12, 137–149.
45. Chassin, Y., Kapri-Pardes, E., Sinvany, G., Arad, T. & Adam, Z. (2002) Plant
Physiol. 130, 857–864.
46. Richter, S. & Lamppa, G. K. (1998) Proc. Natl. Acad. Sci. USA 95, 7463–7468.
47. Chiba, S., Akiyama, Y. & Ito, K. (2002) J. Bacteriol. 184, 4775–4782.
48. Herman, C., Prakash, S., Lu, C. Z., Matouschek, A. & Gross, C. A. (2003) Mol.
Cell,11, 659– 669.
49. Lee, C., Schwartz, M. P., Prakash, S., Iwakura, M. & Matouschek (2001) Mol.
Cell 7, 627–637.
50. Paulsen, H. (2001) in Advances in Photosynthesis: Regulation of Photosynthesis,
eds. Aro, E.-M. & Andersson, B. (Kluwer Academic, Dordrecht, The Neth-
erlands), Vol. 11, pp. 219–233.
51. Adam, Z. & Ostersetzer, O. (2001) Biochem. Soc. Trans. 29, 427– 430.
13704
兩
www.pnas.org兾cgi兾doi兾10.1073兾pnas.0503472102 Z
˙elisko et al.
