Download A novel protein transport system involved in the biogenesis of

Biochimica et Biophysica Acta 1459 (2000) 325^330
www.elsevier.com/locate/bba
A novel protein transport system involved in the biogenesis of bacterial
electron transfer chains
Ben C. Berks a; *, Frank Sargent a , Erik De Leeuw a , Andrew P. Hinsley a ,
Nicola R. Stanley a;b , Rachael L. Jack a;b , Grant Buchanan b , Tracy Palmer a;b
a
Centre for Metalloprotein Spectroscopy and Biology, School of Biological Sciences, University of East Anglia, Norwich, NR4 7TJ, UK
b
Department of Molecular Microbiology, John Innes Centre, Colney, Norwich, NR4 7UH, UK
Received 15 May 2000; accepted 12 June 2000
Abstract
The Tat system is a recently discovered bacterial protein transport pathway that functions primarily in the biosynthesis of
proteins containing redox active cofactors. Analogous transport systems are found in plant organelles. Remarkably and
uniquely the Tat system functions to transported a diverse range of folded proteins across a biological membrane, a feat that
must be achieved without rendering the membrane freely permeable to protons and other ions. Here we review the operation
of the bacterial Tat system and propose a model for the structural organisation of the Tat preprotein translocase. ß 2000
Elsevier Science B.V. All rights reserved.
Keywords: Protein transport; Redox protein; Metalloprotein biosynthesis; Tat pathway; Escherichia coli
1. Introduction
In bacteria the generation of energy by respiratory
or photosynthetic electron transfer chains involves
the cytoplasmic membrane. In order to allow protonmotive force generation by the redox loop mechanism, and to avoid the energetic costs of transporting
substrate molecules (electron donors or terminal electron acceptors) obtained from the extracellular environment into the cytoplasm, a proportion of the redox active components of the electron transfer chains
are always located at the extracellular side of this
membrane. Thus bacterial growth in most environments depends upon the bacterium being able to
* Corresponding author. Fax: +44 (1603) 592250;
E-mail: [email protected]
produce cofactor-containing proteins in the extracellular compartment. Biosynthesis of such extracytoplasmic redox proteins must involve both protein
transport across the cytoplasmic membrane and cofactor insertion into the protein. However, combining these two processes presents a severe biosynthetic
problem to the cell. To understand why this is we
need to consider how a protein without a cofactor is
normally exported from the bacterial cytoplasm.
In Gram-negative bacteria, such as Escherichia
coli, the extracellular compartment is the periplasm.
Proteins are normally exported to the periplasm, or
to the periplasmic side of the cytoplasmic membrane,
by the Sec apparatus [1]. Proteins destined for the
Sec pathway are usually synthesised as precursor
proteins with N-terminal signal peptides that direct
the protein to the Sec system, and that are proteolytically removed from the protein following the trans-
0005-2728 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 5 - 2 7 2 8 ( 0 0 ) 0 0 1 6 8 - 7
BBABIO 44923 7-8-00
326
B.C. Berks et al. / Biochimica et Biophysica Acta 1459 (2000) 325^330
port event. Transport by the Sec system is energised
by a combination of ATP hydrolysis and the transmembrane proton electrochemical gradient. Prior to
transport the substrate protein is kept in an extended
conformation. The protein is then threaded through
the Sec preprotein translocase and only when it
reaches the periplasm does the protein fold. The
problem that biosynthesis of cofactor-containing
periplasmic proteins encounters is the requirement
of the Sec apparatus for an essentially unstructured
precursor protein. In order to stably bind cofactor
the protein must fold, yet if the protein is folded it is
no longer an acceptable substrate for the Sec pathway. One possible way to circumvent this problem is
to transport protein and cofactor across the cytoplasmic membrane separately, and then carry out cofactor insertion in the periplasm. This is the solution
adopted for proteins with some types of cofactor,
primarily those of c-type cytochromes (see article
by Tho«ny-Meyer in this issue [2]). However, for periplasmic proteins containing most other types of cofactor, including iron-sulfur clusters, molybdopterin
cofactors and nucleotide cofactors, the bacterium allows cofactor insertion in the cytoplasm. The cofactor-containing precursor is then targeted to an export
apparatus that is completely distinct from the Sec
system [3] and that has the remarkably property of
being able to transport a folded proteins across a
biological membrane without rendering the membrane freely permeable to protons and other ions.
This Sec-independent protein transport pathway is
usually called the Tat (twin arginine translocation)
system [4] because targeting to the pathway involves
signal peptides containing two consecutive and invariant arginine residues. An alternative Mtt designation
has also been employed [5]. Here we brie£y review
the major features of the bacterial Tat transport
pathway as they are currently understood and propose a model for the structural organisation of the
components of the Tat preprotein translocase. More
detailed reviews of the Tat pathway can be found in
[6^8].
It has recently become apparent that the bacterial
Tat system is structurally and mechanistically analogous to the vpH-dependent thylakoid import pathway of plant chloroplasts [9^13]. In addition, the
mitochondrial genomes of all higher plants encode
a protein homologous to one component of the bac-
terial Tat pathway suggesting that the mitochondria
of plants also possess a Tat-like protein transport
system [5,14].
2. Targeting to the Tat pathway
The signal peptides directing proteins to the Tat
export pathway have a tripartite structure resembling
that of Sec signal peptides. A basic (n-) region is
followed by a hydrophobic (h-) region and then a
c-region containing the recognition site for the protease that will remove the signal peptide following
transport. However, several features distinguish Tat
and Sec signal peptides. While Sec signal peptides
show no amino acid sequence conservation, Tat signal peptides have a consensus Ser-Arg-Arg-Xaa-PheLeu-Lys amino acid sequence motif (with Xaa representing a polar amino acid) at the n-region/h-region
boundary [8,15]. Within the Tat consensus motif the
arginine residues are invariant and are normally essential for transport by the Tat pathway [15^18].
Recent data, however, indicate that in certain signal
peptide contexts conservative substitution of a lysine
for one of the arginine residues still allows weak Tat
targeting activity [15]. The non-arginine residues of
the consensus motif are each present in at least 50%
of Tat signal peptides. Mutating these non-arginine
consensus residues, in particular substitutions at the
consensus phenylalanine position, can lead to severe
defects in Tat transport suggesting that the entire
consensus motif is important for Tat targeting [15].
In addition to the presence of a consensus motif, Tat
signal peptides di¡er from Sec signal peptides in possessing a substantially less hydrophobic h-region [17].
The reduced hydrophobicity of Tat signal peptide hregions has been demonstrated to be critical both for
targeting to the Tat pathway and for avoiding e¤cient interaction with the Sec machinery [17]. Tat
signal peptides often contain basic residues in the
c-region [11]. Recent mutagenesis data suggest that
the presence of, or sign of, the charge in the c-region
of Tat signal peptides is not a major determinant in
targeting to the Tat pathway [15]. However, positively charged residues are not normally found in
the vicinity of the signal peptide processing site of
Sec signal peptides, and when such residues are introduced experimentally transport e¤ciency is re-
BBABIO 44923 7-8-00
B.C. Berks et al. / Biochimica et Biophysica Acta 1459 (2000) 325^330
duced (e.g. [19]). It is thus possible that the positively
charged c-regions found in many Tat signal peptides
are an additional determinant that prevents the signal peptide functionally interacting with the Sec
pathway [17]. A ¢nal di¡erence between Tat and
Sec signal peptide is their length. Sec signal peptides
generally comprise no more that 26 amino acids
while Tat signal peptides are normally 26^60 amino
acids long [8]. The additional length of Tat signal
peptides is mainly due to their larger n-regions [17].
The functional signi¢cance of the greater length of
Tat signal peptides is currently unclear.
Tat signal peptides undoubtedly act as Tat-speci¢c
targeting determinants but could they have additional functions? It is likely that the cell has a mechanism
to prevent substrate proteins accessing the Tat pathway until cofactor insertion has been completed or,
in the case of hetero-oligomeric proteins exported
using a single signal peptide, a multisubunit complex
has been formed [3,6]. The most obvious way to
achieve such control of transport is for the signal
peptide to be sequestered by speci¢c protein-protein
interactions, either with conformations of the substrate protein that are no longer available in the fully
folded mature protein, or by protein-speci¢c chaperones that can sense the folding state of the substrate
protein. A particularly attractive idea is that the putative signal peptide-binding chaperones could be the
same proteins that are involved in, and are therefore
released following, cofactor insertion.
3. Tat pathway components
Studies in E. coli have identi¢ed two operons, tatABCD and tatE, coding for components of the Tat
pathway (Fig. 1). Strains with in-frame deletions in
tatB, in tatC, or in a combination of both tatA and
tatE, are completely defective in the translocation of
precursor proteins with twin-arginine signal peptides
[4,14,20]. Consistent with a role in transmembrane
protein translocation these four genes encode integral
membrane proteins [21]. TatA and TatE are more
than 50% identical in sequence and both in turn exhibit weak sequence similarity to TatB. The TatA/B/
E proteins have a predicted structure in which an Nterminal transmembrane helix is followed at the cytoplasmic side of the plasma membrane by an am-
327
Fig. 1. Organisation of the E. coli tat operons. A putative
stem-loop structure lies between the tatC and tatD genes. Arrows indicate transcripts associated with the tatA locus [21].
phipathic helix and then a sequence-variable region.
TatC is predicted to contain six transmembrane helices with no extensive extramembranous domains. It
has recently been demonstrated that neither the
water-soluble tatD gene product, nor two other
E. coli genes encoding TatD homologues are required for protein export [21].
Genomic analysis of a wide range of bacterial species suggests that a minimum of two copies of TatA/
B/E-like proteins and one copy of a TatC-like protein is required to form a functional Tat system [6,7].
Recent data suggest that the plant thylakoid vpHdependent transport system also utilises at least two
TatA/B/E-type proteins [22].
4. A model for the structural organisation of the Tat
preprotein translocase
Since the essential Tat components are all integral
membrane proteins we speculate that these proteins
come together to form the Tat preprotein translocase. Fig. 2 depicts a model for the organisation of
the Tat proteins within the translocase complex. The
model was formulated on the basis of the arguments
outlined in this section.
We reasoned that the translocase must, at some
stage within the transport cycle, provide a water¢lled channel across the membrane through which
the folded substrate proteins are transported. Since
the structures of many Tat substrates are known, we
can deduce that this channel must attain a diameter
BBABIO 44923 7-8-00
328
B.C. Berks et al. / Biochimica et Biophysica Acta 1459 (2000) 325^330
î when fully open [6]. The channel
of at least 60 A
should be surrounded by proteinaceous elements,
presumably membrane-spanning K-helices. The number of transmembrane helices required to form a
î is considerpore with an internal diameter of 60 A
ably greater than the number of predicted transmembrane helices in the known Tat components [6]. This
suggests that the channel must be formed from multiple copies of at least one Tat subunit. We have
recently co-puri¢ed E. coli TatA and TatB in detergent solution as a complex with a molecular mass of
around 600 kDa (F. Sargent, T. Palmer and B.C.
Berks, unpublished observations). The mass of this
complex is similar to that inferred for the plant vpHdependent pathway translocase [23]. TatA is present
in a large molar excess over TatB in the puri¢ed E.
coli TatAB complex. We therefore suggest that TatA
is the major channel-forming subunit of the Tat
translocase. TatE is functionally interchangeable
with TatA [4,20] and so would presumably also
have a channel forming function. However, our puri¢cation of an exclusively TatAB-containing complex suggests that there are probably separate TatA
and TatE based translocases.
In addition to providing a transport channel for
the substrate, the translocase must have a mechanism
for identifying that substrate. Since targeting to the
Tat pathway involves the recognition of a highly
conserved amino acid sequence motif within precursor signal peptides it is likely that the translocase
makes very speci¢c protein-protein interactions with
the Tat signal peptide. We guess that the amino acids
in the translocase that form the binding site for the
twin-arginine consensus motif should be strongly
conserved during evolution. TatA/B/E family proteins show only limited sequence conservation, with
a glycine residue at the junction between the predicted transmembrane and amphipathic helices the
only invariant residue. We therefore think it unlikely
that these proteins provide the signal peptide binding
site. In contrast, TatC proteins show substantial sequence conservation, and so we have proposed that
TatC proteins contain the site of signal peptide recognition [6]. Note that a single copy of the TatC
subunit per translocase would be su¤cient to accomplish the proposed signal peptide binding function.
In the model shown in Fig. 2 TatB is assigned the
role of physically connecting, and thus mediating
communication between, the TatA and TatC subunits. We know from our puri¢cation of a TatAB
complex that the TatA and TatB subunits must
physically interact. Since TatB has overall structural
Fig. 2. A model for the structural organisation of the Tat preprotein translocase of E. coli. A cross-section through the translocase is
depicted. A substrate protein is shown bound to TatC by its signal sequence.
BBABIO 44923 7-8-00
B.C. Berks et al. / Biochimica et Biophysica Acta 1459 (2000) 325^330
similarity to TatA we reason that TatB could act as a
pseudo-TatA subunit and integrate itself into the
TatA channel structure. We also know that the
TatC protein is completely undetectable in a strain
unable to produce TatB suggesting that TatB stabilises, and thus makes protein-protein interactions
with, TatC [20]. Further, we know that increasing
the ratio of TatB molecules to other Tat components
prevents the Tat system from functioning [20]. A
possible interpretation of this observation that is
consistent with the TatB function proposed here is
that, as the concentration of TatB in the cell increases, the TatB molecules bound to the TatC subunits are no longer also bound to TatA subunits.
Thus the physical connection between precursor
binding site and channel is broken.
5. Bioenergetics
The mechanism of the Tat translocase is constrained in two ways by the proton electrochemical
gradient (vp) across the cytoplasmic membrane.
First, by analogy to the vpH-dependent thylakoid
transporter, protein translocation by the bacterial
Tat translocase is likely to be energised exclusively
by vp, and the e¡ects of uncouplers on Tat transport
in whole bacterial cells are consistent with this idea
[3,17,18]. However, nothing is currently know about
the molecular basis of the coupling of protein transport in the Tat system to vp. Second, the mechanism
of protein translocation by the Tat system must prevent substantive ion leak across the cytoplasmic
membrane during the transport process to avoid
uncoupling the cell. This constraint may prevent
mechanisms which envisage a dynamic pore that
packs around the substrate protein since it is
hard to see how such packing could be su¤ciently
tight to prevent the movement of protons. Instead, it
is more likely that the translocase has two protonimpermeable gates that are never open simultaneously.
Clearly our studies of this remarkable and mechanistically unique protein transport pathway are at an
early stage. A full understanding of the Tat system
will not only be of high intellectual interest but will
be fundamental to our understanding of the biogenesis of biological electron transport chains.
329
Acknowledgements
We wish to thank the many colleagues with whom
we have discussed aspects of the Tat system. Work in
the authors' laboratories is supported by the Biotechnology and Biological Sciences Research Council, the
Royal Society, the University of East Anglia and the
Commission of the European Community.
References
[1] A.P. Pugsley, The complete general-secretory pathway in
Gram-negative bacteria, Microbiol. Rev. 57 (1993) 50^108.
[2] Tho«ny-Meyer, Haem-polypeptide interactions during cytochrome c maturation, Biochem. Biophys. Acta 1459 (2000)
315^323 (this issue).
[3] C.-L. Santini, B. Ize, A. Chanal, M. Mu«ller, G. Giordano,
L.-F. Wu, A novel Sec-independent periplasmic protein
translocation pathway in Escherichia coli, EMBO J. 17
(1998) 101^112.
[4] F. Sargent, E. Bogsch, N.R. Stanley, M. Wexler, C. Robinson, B.C. Berks, T. Palmer, Overlapping functions of components of a bacterial Sec-independent protein export pathway, EMBO J. 17 (1998) 3640^3650.
[5] J.H. Weiner, P.T. Bilous, G.M. Shaw, S.P. Lubitz, L. Frost,
G.H. Thomas, J.A. Cole, R.J. Turner, A novel and ubiquitous system for membrane targeting and secretion of cofactor-containing proteins, Cell 93 (1998) 93^101.
[6] B.C. Berks, F. Sargent, T. Palmer, The Tat protein export
pathway, Mol. Microbiol. 35 (2000) 260^274.
[7] L.-F. Wu, B. Ize, A. Chanal, Y. Quentin, G. Fichant, Bacterial twin-arginine signal peptide-dependent protein translocation pathway: evolution and mechanism, J. Mol. Microbiol. Biotechnol. 2 (2000) 179^189.
[8] B.C. Berks, A common export pathway for proteins binding
complex redox cofactors?, Mol. Microbiol. 22 (1996) 393^
404.
[9] A.M. Settles, A. Yonetani, A. Baron, D.R. Bush, K. Cline,
R. Martienssen, Sec-independent protein translocation by
the maize Hcf106 protein, Science 278 (1997) 1467^1470.
[10] H. Mori, K. Cline, A signal peptide that directs non-Sec
transport in bacteria also directs e¤cient and exclusive transport on the thylakoid delta pH pathway, J. Biol. Chem. 273
(1998) 11405^11408.
[11] M. Wexler, E. Bogsch, R.B. Klo«sgen, T. Palmer, C. Robinson, B.C. Berks, Targeting signals for a bacterial Sec-independent export system direct plant thylakoid import by the
vpH pathway, FEBS Lett. 431 (1998) 339^342.
[12] K. Keegstra, K. Cline, Protein import and routing systems
of chloroplasts, Plant Cell 11 (1999) 557^570.
[13] C. Robinson, this issue.
[14] E. Bogsch, F. Sargent, N.R. Stanley, B.C. Berks, C. Robinson, T. Palmer, An essential component of a novel bacterial
BBABIO 44923 7-8-00
330
[15]
[16]
[17]
[18]
B.C. Berks et al. / Biochimica et Biophysica Acta 1459 (2000) 325^330
protein export system with homologues in plastids and mitochondria, J. Biol. Chem. 273 (1998) 18003^18006.
N.R. Stanley, T. Palmer, B.C. Berks, The twin arginine consensus motif of Tat signal peptides is involved in Sec-independent protein targeting in Escherichia coli, J. Biol. Chem.
275 (2000) 11591^11596.
A. Dreusch, D.M. Bu«rgisser, C.W. Heizmann, W.G. Zumft,
Lack of copper insertion into unprocessed cytoplasmic nitrous oxide reductase generated by an R20D substitution in
the arginine consensus motif of the signal peptide, Biochim.
Biophys. Acta 1319 (1997) 311^318.
S. Cristöbal, J.-W. de Gier, H. Nielsen, G. von Heijne, Competition between Sec- and Tat-dependent protein translocation in Escherichia coli, EMBO J. 18 (1999) 2982^2990.
D. Halbig, T. Wiegert, N. Blaudeck, R. Freudl, G.A.
Sprenger, The e¤cient export of NADP-containing glucose-fructose oxidoreductase to the periplasm of Zymomonas
mobilis depends both on an intact twin-arginine motif in the
signal peptide and on the generation of a structural export
signal induced by cofactor binding, Eur. J. Biochem. 263
(1999) 543^551.
[19] P. Li, J. Beckwith, H. Inouye, Alteration of the amino terminus of the mature sequence of a periplasmic protein can
severely a¡ect protein export in Escherichia coli, Proc. Natl.
Acad. Sci. USA 85 (1988) 7685^7689.
[20] F. Sargent, N.R. Stanley, B.C. Berks, T. Palmer, Sec-independent protein translocation in Escherichia coli: a distinct
and pivotal role for the TatB protein, J. Biol. Chem. 274
(1999) 36073^36082.
[21] M. Wexler, F. Sargent, R.L. Jack, N.R. Stanley, E. Bogsch,
C. Robinson, B.C. Berks, B.C., T. Palmer, TatD is a cytoplasmic protein with DNase activity. No requirement for
TatD family proteins in Sec-independent protein export,
J. Biol. Chem. 275 (2000) 16717^16722.
[22] H. Mori, E.J. Summer, X. Ma, K. Cline, Component speci¢city for the thylakoid Sec and delta pH-dependent protein
transport pathways, J. Cell Biol. 146 (1999) 45^55.
[23] J. Bergho«fer, R.B. Klo«sgen, Two distinct translocation intermediates can be distinguished during protein transport by
the TAT (vpH) pathway across the thylakoid membrane,
FEBS Lett. 460 (1999) 328^332.
BBABIO 44923 7-8-00