1024
Biochemical Society Transactions (2005) Volume 33, part 5
The complexity of pathways for protein import
into thylakoids: it’s not easy being green
A. Di Cola, E. Klostermann and C. Robinson1
Department of Biological Sciences, University of Warwick, Gibbet Hill Road, Coventry CV4 7AL, U.K.
Abstract
Numerous proteins are transported into or across the chloroplast thylakoid membrane. To date, two major
pathways have been identified for the transport of luminal proteins (the Sec- and Tat-dependent pathways)
and it is now clear that these protein translocases use fundamentally different transport mechanisms.
Integral membrane proteins are inserted by means of at least two further pathways. One involves the input
of numerous targeting factors, including SRP (signal recognition particle), FtsY and Albino3. Surprisingly,
the other pathway does not involve any of the known chloroplastic targeting factors, and insertion is
energy-independent, raising the possibility of an unusual ‘spontaneous’ insertion mechanism.
Introduction
Chloroplasts perform a variety of biochemical functions
within plant cells, although these organelles are best known
as the site of photosynthesis. Numerous plastid proteins are
essential for the various physiological processes ongoing in
this organelle. Although the chloroplast possesses its own
genome, it encodes only a small proportion of the plastid
proteins and the vast majority are nuclear encoded. These proteins are synthesized as precursors in the cytoplasm and targeted to the different intra-chloroplast destinations via specific routes.
Chloroplasts are formed of a ‘matrioska’ system of membranes separated by soluble intermembrane phases: from the
outside cytosol moving inwards chloroplasts are bound by
a double-membrane envelope enclosing an intermembrane
soluble space and the inner major soluble phase, the stroma.
Many essential pathways operate in the stroma, such as
the Calvin cycle for carbon fixation, amino acid synthesis
and transcription–translation for the synthesis of the chloroplast-encoded-proteins. Within the stroma, a further inner
membrane system constitutes the thylakoid network, where
light is captured and ATP is synthesized. The thylakoid membrane encloses another soluble phase, the thylakoid lumen,
where soluble or loosely membrane-associated photosynthetic proteins are located.
Import into the chloroplast
Cytosolically synthesized chloroplast proteins are imported
post-translationally into the organelle and a default translocation pathway involving distinct machineries in the outer and
inner envelope membranes has been studied in some detail
Key words: chloroplast, protein transport, Sec, signal recognition particle (SRP), thylakoid, twinarginine translocation (Tat).
Abbreviations used: EGFP, enhanced green fluorescent protein; LHCP, light-harvesting chlorophyll-binding protein; SPP, stromal processing peptidase; SRP, signal recognition particle;
Tat, twin-arginine translocation; TPP, thylakoidal processing peptidase.
1
To whom correspondence should be addressed (email [email protected]).
C 2005
Biochemical Society
[1]. Many of the imported proteins remain in the stroma
after import but others are targeted to the thylakoids by
additional pathways that have been studied primarily using
in vitro translocation assays. These studies have shown that
four mainstream pathways are used for protein targeting to
thylakoids, as detailed below.
The first stage for protein transport to the chloroplast is
common to all the routes, involving the synthesis of cytosolic
precursors and translocation across chloroplast envelopes.
Precursors bear an N-terminal transit peptide containing
stromal-targeting information, which is responsible for
directing the precursor across the two chloroplast envelope
membranes. In some cases, the signal peptide is bipartite
and contains additional information for transport across the
thylakoid membrane. The two signal peptides are removed
by two different processing peptidases, one resident in the
stroma (SPP, stromal processing peptidase) and the other in
the thylakoid lumen (TPP, thylakoidal processing peptidase;
for a detailed review, see [1]). Translocation across the
chloroplast envelope is mediated by the Tic/Toc translocase.
The process is ATP- and GTP-dependent and precursors are
translocated in an unfolded state.
Targeting of thylakoid lumen proteins
Essentially, all thylakoid lumen proteins are initially
synthesized with bipartite presequences that contain a transit
peptide and signal peptide in tandem. The removal of the
transit peptide by the SPP exposes the thylakoid signal
peptide and generates a stromal intermediate form that can
undertake one of the two possible pathways to reach the
thylakoid lumen.
The Tat (twin-arginine translocation)
pathway
Tat substrates are characterized by the presence of an invariant
twin arginine motif located in the N-terminal region of the
Mechanistic and Functional Studies of Proteins
signal peptide. This motif is essential for Tat-mediated translocation and gives the name to the pathway [2]. At the
C-terminus of the peptide, an Ala-Xaa-Ala motif denotes
the cleavage site for TPP allowing the release of the mature
protein in the lumen [3].
One of the major characteristics of this thylakoid translocation system is that the translocation event is not driven
by nucleoside triphosphate hydrolysis, but by the transmembrane pH gradient. For this reason the pathway is also known
as the pH-dependent pathway [4,5]. The second major
and more unusual feature is that this system is capable of
transporting fully folded proteins [6].
It is now known that three proteins (Tha4, Hcf106 and
cpTatC) are constituents of the Tat system in plants [7–9].
Their discovery led to the identification of the homologous
TatABC system in prokaryotes [10]. In bacteria, a number of
periplasmic proteins containing cofactors have been shown
to be Tat substrates, and the fact that these cofactors are
enzymatically inserted only in the cytoplasm, led to the conclusion that Tat passenger proteins must translocate in an
extensively folded form [11]. However, little is known about
the actual translocation mechanism at the present time.
Previously, a detailed in vitro cross-linking study revealed
that in isolated thylakoids, Tat subunits are found in two
distinct subcomplexes: a cpTatC–Hcf106 complex and a Tha4
complex. Tat substrates preferentially bind to the first one
that would work as precursor receptor, this interaction and
the presence of pH triggers the recruitment of Tha4 which
assembles with cpTatC–Hcf106 subcomplex to form the
translocation complex [12,13].
Other aspects of the system remain to be studied in detail
and this applies to the functioning of the system in vivo. A
recent study analysed the targeting of different Tat substrates
in vivo in Chlamydomonas reinhardtii and found that in this
case the Tat system is not dependent on the pH [14]. More
recently, these data have been revisited suggesting that the
transmembrane ψ can also power the Tat translocase [15];
another possibility is that the reconstituted in vitro import
system is lacking some ‘crucial’ yet unidentified factors,
which are normally present in planta. Overall, these findings
have opened a new area of debate where still little is known
and more study is needed.
The Sec pathway
The second mainstream pathway for protein transport
into the thylakoid lumen is the Sec pathway. This is related
to Sec-type translocation pathways in bacteria, which are
responsible for protein export across the plasma membrane
to the periplasm or medium. In bacteria, the Sec complex is
formed by the SecA subunit and the heterotrimeric SecYEG
translocation channel. SecA is an ATPase that acts as the translocation motor pushing segments of precursor proteins across
the channel (reviewed in [16]).
For the chloroplast system, cpSecY [17], cpSecE [18]
and cpSecA have been identified [19–21]. Antibodies against
cpSecY inhibit cpSecA-dependent protein translocation, suggesting that cpSecA and cpSecY work in concert [22] as in
bacteria. cpSecA has been demonstrated to promote ATPdependent transport of OEC 33K and plastocyanin into
thylakoids [21].
The bacterial SecYEG complex transports substrates in an
unfolded state and the energy for translocation comes from
ATP hydrolysis. Analogously, when DHFR (dihydrofolate
reductase) is fused to an Sec signal peptide and assessed for
import into thylakoids in vitro, its translocation is completely
blocked in the presence of folate analogues that are known to
bind to the precursor and stabilize it in a tightly folded form.
These findings strongly suggest that Sec-mediated translocation can occur only when substrates are in an unfolded
state [23]. A more recent observation seems to confirm this
conclusion: EGFP (enhanced green fluroscent protein) targeted to the thylakoid can be efficiently transported via the
Tat pathway, while transport mediated by Sec is very inefficient. The rapid folding of the precursor EGFP in the stroma
appears to prevent its transport through the Sec translocase
[24]. Figure 1 illustrates the salient features of the Tat- and
Sec-dependent pathways.
Thylakoid membrane proteins
The SRP (signal recognition particle) pathway
The SRP pathway is used to direct a subset of hydrophobic proteins into the thylakoid membrane. In eukaryotes,
prokaryotes and archaea a related SRP pathway mediates
co-translational protein transport and insertion into the
membranes (reviewed in [25]). In plants, the SRP pathway
is best known for the post-translational targeting and membrane insertion of nuclear-encoded LHCPs (light-harvesting
chlorophyll-binding proteins). For an extended review on
chloroplast SRP, see [26]. In short, cpSRP54, a protein homologue of mammalian/bacterial SRP54 subunits, together
with cpSRP43, a novel protein unique to this plant posttranslational SRP pathway, forms the cpSRP that, in contrast
with all other known SRPs, contains no RNA component. It
binds to LHCP to form a ‘transit complex’, the soluble form
of the hydrophobic LHCP for transport in the stroma. Lhcb1,
Lhcb4.1 and Lhcb5 have been shown to use this pathway.
For recent developments concerning the interactions within
the transit complex, see [27,28]. Integration of LHCP also
requires GTP, cpFtsY and Alb3. cpFtsY is thought to act as
a receptor and form a complex with cpSRP and functional
Alb3 [29]. The interaction of cpSRP with cpFtsY seems
to promote GTP binding to the GTPases cpSRP54 and
cpFtsY, which target the complex to Alb3 in the thylakoid
membrane. Alb3 is homologous with the bacterial YidC and
mitochondrial Oxa1 proteins that are known to assist in the
insertion of membrane proteins into the plasma membrane
or inner mitochondrial membrane respectively. cpSRP43 has
no function in targeting to Alb3, but may be involved in the
regulation of the GTPase cycle to minimize GTP hydrolysis
[28,29]. Besides cpSRP and cpFtsY, cpSecY was localized in
Alb3 complexes [29,30]. Studies suggest that it is not involved
in post-translational LHCP integration [22,29]. But in a
C 2005
Biochemical Society
1025
1026
Biochemical Society Transactions (2005) Volume 33, part 5
Figure 1 Model of Tat and Sec targeting pathways for thylakoid
Figure 2 Models for the SRP and ‘spontaneous’ targeting
lumen proteins
Thylakoid lumen proteins are synthesized in the cytosol with a bipartite
pathway for thylakoid membrane proteins
The SRP-dependent pathway is used by LHCPs such as Lhcb1. After
presequence. After import into the chloroplast in an unfolded state,
the first targeting signal is cleaved off and the proteins are then transported across the thylakoid membrane by means of either a Sec- or
import into the chloroplast, the substrate interacts with cpSRP, which
comprises SRP54 and SRP43 subunits, and its partner protein, FtsY.
These factors direct Lhcb1 to the membrane and insertion occurs in
Tat-specific signal peptide. Proteins such as plastocyanin on the Sec
pathway interact with SecA in the stroma, which hydrolyses ATP and
effectively pushes the preprotein through a membrane-bound translocon
a process that depends on membrane-bound Alb3. Other membrane
proteins such as PsbW and PsaK are inserted by pathways that do
not require any known protein targeting apparatus, and these may
in an unfolded state. The translocon is known to contain SecY and SecE
but is otherwise poorly characterized. It is also unclear whether the
substrate protein remains unfolded in the stromal phase, perhaps with
therefore insert spontaneously. One pathway is used by a small number
of thylakoid membrane proteins (to date, CFoII, PsbX, PsbW and PsbY)
that bear cleavable signal-type peptides that somehow aid insertion.
the aid of (unidentified) chaperone molecules, or whether it is actively
unfolded at the membrane. Substrates on the Tat pathway, for example
PsbP or PsbQ, have been shown to refold in the stroma and are not
The other is used by the majority of thylakoid membrane proteins (PsaK
is shown as an example) and the protein inserts into the membrane as
the mature-size form.
believed to interact with any dedicated Tat components until they reach
the membrane. Here, they are transported by a membrane-bound Tat
system that contains Tha4, Hcf106 and TatC. The Tat system is able to
translocate substrates in a fully folded state and the thylakoidal pH
is essential for translocation, at least in isolated thylakoids and intact
isolated chloroplasts.
The ‘spontaneous’ pathway for thylakoid
membrane insertion
recently described co-translational cpSRP pathway [26]
mediating transport of plastid-encoded proteins, elongating
nascent chloroplast-encoded D1 protein has been shown to
bind to ribosome-associated cpSRP54, and cpSecY transiently
interacts with isolated D1 elongation intermediates [30a].
Further candidates for co-translational protein transport
via cpSRP are the chloroplast-encoded proteins D2, CP43,
PSI-A and CF0 III which have been shown to interact with
Alb3 in a yeast two-hybrid system especially adapted to
detect membrane protein interactions [31]. C. reinhardtii
possesses two homologues of the alb3 gene. In an Alb3.1
deletion mutant it was found that D1 protein was inserted
into the thylakoid membrane while assembly into the reaction
centre complex was delayed, suggesting a role in complex
assembly rather than as an integrase [32].
C 2005
Biochemical Society
So far, we have described thylakoid protein targeting pathways involving interactions with proteinaceous machineries.
A fourth group of thylakoid proteins appears to behave
differently. Proteins belonging to this group are all integral
membrane proteins, and their main feature is that they do not
require any known protein transport machinery, nor energy,
for insertion into thylakoid membranes. This is an unusual
pathway, which is unique to chloroplasts, and the absence of
any known energy requirement or essential targeting factor
suggests the possibility of a spontaneous insertion mechanism
[33].
Examples of proteins that insert ‘spontaneously’ are the
W and X subunits of Photosystem II (PsbW and PsbX),
PsaK from Photosystem I [34] and the SecE subunit (cpSecE)
[35]. The first two bear a bipartite presequence resembling
luminal proteins, but have been shown to insert into thylakoid
membranes in the absence of SRP, NTPs or a functional
Sec machinery. Furthermore, proteolysis of thylakoids or
inactivation of Alb3 does not influence their insertion [36].
Mechanistic and Functional Studies of Proteins
It has been proposed that after the removal of the transit
peptide, the remaining N-terminal propeptide containing a
hydrophobic region is necessary for the formation of a ‘helical
hairpin’-type loop intermediate in the thylakoid membrane,
upon interaction with the lipid bilayer of the thylakoid
membrane [33]. Figure 2 depicts the operation of the SRPdependent and ‘spontaneous’ insertion mechanisms.
18
19
20
21
22
23
References
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Jarvis, P. and Robinson, C. (2004) Curr. Biol. 14, R1064–R1077
Chaddock, A.M., Mant, A., Karnauchov, I., Brink, S., Herrmann, R.G.,
Klosgen, R.B. and Robinson, C. (1995) EMBO J. 14, 2715–2722
Shackleton, J.B. and Robinson, C. (1991) J. Biol. Chem. 266,
12152–12156
Cline, K., Ettinger, W.F. and Theg, S.M. (1992) J. Biol. Chem. 267,
2688–2696
Mould, R.M. and Robinson, C. (1991) J. Biol. Chem. 266, 12189–12193
Clark, S.A. and Theg, S.M. (1997) Mol. Biol. Cell 8, 923–934
Motohashi, R., Nagata, N., Ito, T., Takahashi, S., Hobo, T., Yoshida, S.
and Shinozaki, K. (2001) Proc. Natl. Acad. Sci. U.S.A. 98, 10499–10504
Voelker, R. and Barkan, A. (1995) EMBO J. 14, 3905–3914
Walker, M.B., Roy, L.M., Coleman, E., Voelker, R. and Barkan, A. (1999)
J. Cell Biol. 147, 267–276
Bogsch, E.G., Sargent, F., Stanley, N.R., Berks, B.C., Robinson, C. and
Palmer, T. (1998) J. Biol. Chem. 273, 18003–18006
Weiner, J.H., Bilous, P.T., Shaw, G.M., Lubitz, S.P., Frost, L., Thomas,
G.H., Cole, J.A. and Turner, R.J. (1998) Cell (Cambridge, Mass.) 93,
93–101
Cline, K. and Mori, H. (2001) J. Cell Biol. 154, 719–729
Mori, H. and Cline, K. (2002) J. Cell Biol. 157, 205–210
Finazzi, G., Chasen, C., Wollman, F.A. and de Vitry, C. (2003) EMBO J.
22, 807–815
Theg, S.M., Cline, K., Finazzi, G. and Wollman, F.A. (2005)
Trends Plant Sci. 10, 153–154
Dalbey, R.E. and Chen, M. (2004) Biochim. Biophys. Acta 1694, 37–53
Laidler, V., Chaddock, A.M., Knott, T.G., Walker, D. and Robinson, C.
(1995) J. Biol. Chem. 270, 17664–17667
24
25
26
27
28
29
30
30a
31
32
33
34
35
36
Schuenemann, D., Amin, P., Hartmann, E. and Hoffman, N.E. (1999)
J. Biol. Chem. 274, 12177–12182
Nakai, M., Goto, A., Nohara, T., Sugita, D. and Endo, T. (1994)
J. Biol. Chem. 269, 31338–31341
Voelker, R., MendelHartvig, J. and Barkan, A. (1997) Genetics 145,
467–478
Yuan, J.G., Henry, R., McCaffery, M. and Cline, K. (1994) Science 266,
796–798
Mori, H., Summer, E.J., Ma, X.Y. and Cline, K. (1999) J. Cell Biol. 146,
45–56
Hynds, P.J., Robinson, D. and Robinson, C. (1998) J. Biol. Chem. 273,
34868–34874
Marques, J.P., Schattat, M.H., Hause, G., Dudeck, I. and Klosgen, R.B.
(2004) J. Exp. Bot. 55, 1697–1706
Koch, H.G., Moser, M. and Muller, M. (2003) Rev. Physiol.
Biochem. Pharmacol. 146, 55–94
Schunemann, D. (2004) Curr. Genet. 44, 295–304
Funke, S., Knechten, T., Ollesch, J. and Schunemann, D. (2005)
J. Biol. Chem. 280, 8912–8917
Goforth, R.L., Peterson, E.C., Yuan, J., Moore, M.J., Kight, A.D., Lohse,
M.B., Sakon, J. and Henry, R.L. (2004) J. Biol. Chem. 279, 43077–43084
Moore, M., Goforth, R.L., Mori, H. and Henry, R. (2003) J. Cell Biol.
162, 1245–1254
Klostermann, E., Helling, I.D., Carde, J.P. and Schunemann, D. (2002)
Biochem. J. 368, 777–781
Zhang, L.X., Paakkarinen, V., Suorsa, M. and Aro, E.M. (2001)
J. Biol. Chem. 276, 37809–37814
Pasch, J.C., Nickelsen, J. and Schunemann, D. (2005)
Appl. Microbiol. Biotechnol., doi:10.1007/s00253-005-0029-3
Ossenbuhl, F., Gohre, V., Meurer, J., Krieger-Liszkay, A., Rochaix, J.D.
and Eichacker, L.A. (2004) Plant Cell 16, 1790–1800
Robinson, C. and Mant, A. (1997) Trends Plant Sci. 2, 431–436
Mant, A., Woolhead, C.A., Moore, M., Henry, R. and Robinson, C.
(2001) J. Biol. Chem. 276, 36200–36206
Steiner, J.M., Kocher, T., Nagy, C. and Loffelhardt, W. (2002)
Biochem. Biophys. Res. Commun. 293, 747–752
Woolhead, C.A., Thompson, S.J., Moore, M., Tissier, C., Mant, A.,
Rodger, A., Henry, R. and Robinson, C. (2001) J. Biol. Chem. 276,
40841–40846
Received 26 July 2005
C 2005
Biochemical Society
1027