The presence of a helix breaker in the hydrophobic core
of signal sequences of secretory proteins prevents recognition
by the signal-recognition particle in
Escherichia coli
Hendrik Adams
1
, Pier A. Scotti
2
*, Hans de Cock
1
, Joen Luirink
2
and Jan Tommassen
1
1
Department of Molecular Microbiology and Institute of Biomembranes, Utrecht University, The Netherlands;
2
Department of
Microbiology, Institute of Molecular Biological Sciences, Biocentrum Amsterdam, The Netherlands
Signal sequences often contain a-helix-destabilizing amino
acids within the hydrophobic core. In the precursor of the
Escherichia coli outer-membrane protein PhoE, the glycine
residue at position )10 (Gly
)10
) is thought to be responsible
for the break in the a-helix. Previously, we showed that
substitution of Gly
)10
by a-helix-promoting residues (Ala,
Cys or Leu) reduced the proton-motive force dependency of
the translocation of the precursor, but the actual role of the

with components of the Sec translocon [5,6]. At the onset of
translocation, SecB is released [7] and the preprotein is
translocated by an insertion–deinsertion cycle of SecA into
the SecYEG translocon [8]. Energy for the translocation
process is provided by ATP hydrolysis by SecA [8,9] and by
the proton-motive force (pmf) [9]. At the periplasmic side of
the membrane, leader peptidase removes the signal sequence
from the precursor, and the mature protein is released into
the periplasm [10]. The bacterial SRP-targeting route is
homologous with, but less complex than, the eukaryotic
SRP system [11,12]. The E. coli SRP consists of a single
protein, P48, and a 4.5S RNA, and binds cotranslationally
to hydrophobic sequences [13,14]. The ribosome-nascent
chain (RNC) complex subsequently binds to FtsY and is
targeted to the Sec translocon in the inner membrane
[15,16]. Whereas the SecB route is predominantly used by a
subset of periplasmic and most, if not all, outer-membrane
proteins, inner-membrane proteins are particularly depend-
ent on a functional SRP pathway [17].
We are using outer-membrane protein PhoE as a model
to study protein export. PhoE is targeted via its signal
sequence in a SecB-dependent way to the Sec translocon [3].
Whereas the signal sequence is necessary and, in most cases,
sufficient for translocation across the cytoplasmic mem-
brane, its exact role in the export mechanism is far from
understood. Despite the common function of signal
sequences, i.e. to direct the translocation of the attached
polypeptide chain, there is little sequence homology among
them [18]. Nevertheless, a common structural organization
can be recognized (Fig. 1). Signal sequences are character-

)10
)of
the signal sequence of PhoE was examined [29]. It was
shown that substitution of this residue by a-helix-promoting
residues (Ala, Cys or Leu) reduced the pmf dependency of
the translocation across the inner membrane, but the actual
role of the helix breaker remained obscure. It should be
noted that such substitutions extend the a-helix not just by a
single residue, but, probably, over the entire H domain
(Fig. 1). Whereas the a-helix in the wild-type signal
sequence is too short to span the inner membrane, the
resulting mutant signal sequences would more closely
resemble the membrane-spanning domains of inner-mem-
brane proteins and might therefore be turned into substrates
for the SRP. In this paper, we considered the possibility that
the extended a-helix resulting from the Gly
)10
substitutions
affects the targeting pathway of the precursor.
EXPERIMENTAL PROCEDURES
Reagents and biochemicals
Restriction enzymes were purchased from either Boehringer
Mannheim or Pharmacia. MEGAshortscript T7 transcrip-
tion kit was from Ambion, and [
35
S]methionine and Tran
35
S-label were from Amersham International. Bis(sulfo-
succinimidyl)-suberate (BS
3

NT1060 F
–
, DlacU169 araD139 rpsL thi relA ptsF25 deoC1 lamBD60 T.J. Silhavy (pers. comm.)
MM152 MC4100 secB::Tn5 [51]
IQ85 Tn10 thiA Dlac araD rpsL rpsE relA secYts24 [51]
CE1513 CE1224 secAts51 leu::Tn10 This study
CE1514 CE1224 Tn10 secYts24 This study
CE1515 CE1224 secB::Tn5 This study
FF283 F
–
, lacDx74 araD139 (araABOIC-leu) D7679 galU galK rpsL ffs::kan/F¢ lac-pro,
lacI
q
Ptac::ffs
[52]
Plasmids
pJP29 Cam
r
, wild-type phoE [30]
pNN5 pJP29 derivative encoding (G-10A)prePhoE [29]
pNN7 pJP29 derivative encoding (G-10C)prePhoE [29]
pNN8 pJP29 derivative encoding (G-10L)prePhoE [29]
pC4Meth101FtsQ-WT Amp
r
, encodes truncated 101FtsQ [13]
pC4Meth94PhoE Amp
r
, encodes truncated 94PhoE [13]
pC4Meth(G-10C)94PhoE pC4Meth94PhoE derivative encoding (G-10C) mutant 94PhoE This study
pC4Meth(G-10L)94PhoE pC4Meth94PhoE derivative encoding (G-10L) mutant 94PhoE This study

plasmid expressing (mutant) phoE from its own promoter,
were grown under phosphate limitation at 30 °Cas
described previously [30]. Cells of the 4.5S RNA conditional
strain FF283 were grown to D
660
¼ 1.0 in Hepes-buffered
synthetic medium, supplemented with 660 l
M
K
2
HPO
4
.
For the depletion of 4.5S RNA, isopropyl b-
D
-thiogalacto-
pyranoside was omitted from the growth medium. To
induce the expression of (mutant) phoE from its own
promoter, cells were collected by centrifugation and washed
with Hepes-buffered synthetic medium with no phosphate
added. The cell pellets were resuspended in the latter
medium at the original absorbance, followed by incubation
at 37 °C for 30 min. For pulse-labeling, cells were incubated
for 45 s with Tran
35
S-label followed by a chase period with
an excess of nonradioactive methionine/cysteine. After
precipitation with 5% (w/v) trichloroacetic acid, radio-
labeled proteins were separated either directly or after
immunoprecipitation with a polyclonal PhoE-specific anti-

linking with 1 m
M
DSSfor10minat25°C, the cross-link
reaction was stopped with quench buffer. Peripheral and
soluble cross-linked complexes were separated from
integral-membrane cross-linked complexes by Na
2
CO
3
extraction as described [35]. Samples were analyzed either
directly or after immunoprecipitation on 12% polyacryla-
mide gels and visualized as described above.
To probe the molecular environment of membrane-
associated RNCs, SRP was reconstituted in vitro from
purified 4.5S RNA and purified hexa-His-tagged P48 as
described [35]. To allow SRP–RNC complex formation
(G-10L)94PhoE and 101FtsQ were synthesized in vitro and
incubated at 25 °C with 350 n
M
reconstituted SRP, and
SRP–RNC complexes were purified from the translation
mixture by centrifugation through a high-salt sucrose
cushion [36]. The SRP–RNC complexes were incubated
with IMVs from strain NT1060 under conditions as
described previously [35]. After cross-linking with 2 m
M
DSS at 25 °C for 10 min, 0.1 vol. quench buffer was added
and incubation was continued on ice for 15 min. Subse-
quently, peripheral and soluble cross-linked complexes were
separated from integral-membrane cross-linked complexes

improved processing kinetics compared with wild-type
prePhoE in the secB mutant (Fig. 2A). After a 5-min chase
period, hardly any mutant prePhoE was detected anymore,
whereas the vast majority of the wild-type precursor was still
not processed. Together with the previously reported
reduced pmf dependency for translocation of the mutant
precursors [29], our results suggest that the SecB depend-
ency of prePhoE targeting correlates with its DlH
+
dependency for in vitro translocation.
5566 H. Adams et al.(Eur. J. Biochem. 269) Ó FEBS 2002
Of all the precursors tested, the mutant precursor with the
strongest a-helix-promoting residue (Leu) at position )10
appeared to be most efficiently processed in the secB mutant
strain. This mutant precursor was used to verify if
translocation is still dependent on the membrane-embedded
SecYEG complex and on SecA. For this purpose, pulse–
chase experiments were performed in secA51 and secY24
mutant strains at their nonpermissive temperature. In both
strains, processing of the (G-10L)prePhoE protein, like that
of the wild-type precursor, was strongly impaired in
comparison with the processing in the wild-type strain
(Fig. 2B). Apparently, substitution of the glycine residue at
position )10 by an a-helix-promoting residue does not alter
the dependency of the precursor on SecA and SecY,
whereas its SecB dependency is reduced.
Affinity of mutant prePhoE nascent chains for P48
As the SecB dependency of the translocation of the mutant
prePhoE proteins was clearly decreased, we next considered
the possibility that they had become substrates for the

94PhoE protein was also cross-linked to P48 (Fig. 3),
although not as efficiently as (G-10L)94PhoE. In all cases,
antiserum against trigger factor (TF) efficiently precipitated
cross-linked complexes (Fig. 3A,B), confirming the earlier
observation that TF, a cytosolic chaperone, binds to E. coli
nascent polypeptides [13]. Quantification of the data
indicated that the cross-linking efficiency of the mutant
nascent chains was somewhat reduced (Fig. 3B). In conclu-
sion, our results show an increased affinity of the G-10C and
G-10L prePhoE for the P48 component of SRP.
G-10L nascent PhoE interacts with Sec translocon
components
As (G-10L)94PhoE nascent chains apparently have a high
affinity for P48 in vitro, we subsequently examined whether
these nascent chains are targeted to SecY via SRP by
performing cross-linking experiments in vitro in the presence
of IMVs. To obtain a high cross-linking efficiency, recon-
stituted E. coli SRP was added after translation of nascent
(G-10L)94PhoE polypeptides to saturate the RNCs with
SRP. The SRP–RNC complexes were purified over a high-
salt sucrose cushion and incubated with IMVs to allow
targeting. After cross-linking with the membrane-permeable
cross-linking reagent DSS, peripheral and soluble cross-
linked complexes were separated from integral-membrane
cross-linked complexes by Na
2
CO
3
extraction and analyzed
by SDS/PAGE (Fig. 4). In the Na

the restrictive temperature (42 °C), and pulse-labeled at 42 °Cfor45s
with Tran
35
S-label and chased with an excess of unlabeled methionine/
cysteine. Aliquots were removed at the indicated periods and analyzed
as described for panel (A). The precursor and mature forms of the
PhoE proteins are indicated by p and m, respectively.
Ó FEBS 2002 Re-routing a secretory protein via the SRP pathway (Eur. J. Biochem. 269) 5567
of SecA. The  40-kDa product in the Na
2
CO
3
pellet
probably contains proteolytic fragments of the SecA dimer
and monomer cross-linking products, which is in agreement
with earlier reports [38]. The fuzzy  46-kDa product
(Fig. 4A, lane 3) was immunoprecipitated with anti-SecY
serum (Fig. 4B, lane 2), showing that the (G-10L)94PhoE
nascent chains are targeted to the SecYEG translocon.
In the Na
2
CO
3
supernatant, at least three major cross-
linking adducts, of apparent molecular mass  110,  65
and  55 kDa, could be detected (Fig. 4A, lane 5). In
addition, several cross-linking adducts of low molecular
mass (< 30 kDa) were detected. Immunoprecipitation
revealed that the high-molecular-mass adducts represent
cross-linking to SecA (data not shown), TF and P48

dependent on this pathway in vivo. To test this possibility,
wild-type and the (G-10L)prePhoE were expressed in
FF283 cells which were depleted of 4.5S RNA. After
radioactive labeling of the cells, the PhoE forms were
immunoprecipitated and analyzed by SDS/PAGE (Fig. 5).
Depletion of 4.5S RNA did not result in the accumulation
of precursors of either wild-type prePhoE or (G-10L)pre-
PhoE. Apparently (G-10L)prePhoE translocation is not
dependent on the SRP pathway in vivo.
DISCUSSION
NMR studies of the signal peptides of LamB [25], OmpA
[26] and PhoE [27] showed that the a-helical conformation is
disrupted toward the C-terminus of the hydrophobic core
near a helix-breaking residue, such as Gly
)10
in the case of
prePhoE. Furthermore, a statistical analysis of signal
sequences revealed the common occurrence of helix-break-
ing residues within the hydrophobic core [28], suggesting
that the disruption of the a-helix is a common feature of
signal sequences. In a previous study, it was shown that the
DlH
+
dependency of prePhoE translocation was dramati-
cally reduced when a helix-promoting residue, such as
leucine or cysteine, was substituted for the helix-breaking
Gly
)10
of the signal sequence [29]. Such a substitution is
expected to result in considerable elongation of the a-helix

sequence. Therefore, we considered the possibility that the
Gly
)10
mutations affected the targeting pathway. The
results from the in vivo pulse–chase experiments showed
that targeting of the mutant PhoE precursors is less
dependent on SecB, indicating that they are targeted to
the Sec translocon via another targeting pathway. In vitro
cross-linking with the water-soluble cross-linker BS
3
revealed that the G-10C and G-10L 94PhoE nascent chains
had an increased affinity for the P48 component of SRP
compared with wild-type 94PhoE nascent chains. Further-
more, cross-link experiments with nascent chains in the
presence of IMVs showed SRP-mediated targeting of
(G-10L)94PhoE to the Sec translocon. However, in vivo
pulse–chase experiments revealed normal translocation
kinetics of (G-10L)prePhoE in a 4.5S RNA-depletion
strain. This result is understandable, as the SecB-binding
sites, which are located in the mature domain of the PhoE
precursor [3], are not affected in the G-10L mutant
precursor. Thus, in the absence of SRP, SecB can target
the mutant prePhoE to the SecYEG translocon. Consis-
tently, the processing of the mutant precursors was not
completely SecB independent in a strain expressing SRP
(Fig. 2A). It has been reported previously that the SRP-
targeting pathway is easily overloaded by overexpression of
SRP substrates [17]. Therefore, at the high expression levels
used in these experiments, a proportion of the mutant
prePhoE molecules may still rely on the SecB pathway,

molecular mass marker proteins (MW) are
indicated on the right. Relevant cross-linked
complexes are indicated with arrowheads. (C)
RNCs of wild-type and (G-10L)prePhoE were
synthesized in the presence of IMVs and sub-
sequently incubated with DSS. After quench-
ing, cross-linked products were examined as
described above.
Fig. 5. SRP dependency of (G-10L)prePhoE translocation in vivo.
Wild-type prePhoE and (G-10L)prePhoE were expressed in cells of
strain FF283 either depleted or not depleted of 4.5S RNA. The cells
were pulse-labeled, followed by a chase for the indicated periods. PhoE
proteins were immunoprecipitated, separated by SDS/PAGE and
detected by autoradiography.
Ó FEBS 2002 Re-routing a secretory protein via the SRP pathway (Eur. J. Biochem. 269) 5569
P48. We propose that this additional variable is a-helix
propensity. Apparently, the a-helix propensity of cysteine
compensates for its low hydrophobicity, resulting in a better
interaction of the (G-10C)94PhoE protein with P48.
The mechanism by which secretory proteins are routed
into the SRP-targeting or the SecB-targeting pathways in
E. coli is not fully understood. Although E. coli SRP has
been shown to interact with cleavable signal sequences
in vitro [41,43–46], it is generally assumed that it binds
efficiently, under physiological conditions, only to signal-
anchor sequences, which contain a longer stretch of
consecutive hydrophobic amino acids. Recent studies have
indicated that the hydrophobicity of the targeting signal is
the parameter discriminating between SRP-dependent and
SRP-independent pathways [14]. On the other hand, in vitro

to those of the G-10L mutant PhoE [48]. In conclusion, our
results indicate that the helix breaker in cleavable signal
sequences prevents recognition by SRP, and it appears that
besides hydrophobicity the a-helix propensity of the hydro-
phobic core of the signal sequence helps to determine the
targeting pathway.
ACKNOWLEDGEMENTS
We would like to thank Elaine Eppens and Margot Koster for helpful
discussions and interest in the work, and Nico Nouwen for construction
of strain CE1513. Our thanks also go to William Wickner and Arnold
Driessen for providing antibodies against SecY and SecA, respectively.
Further, we thank Bauke Oudega for providing strain MM88, and
Tom Silhavy for his gift of strain NT1060. Finally, we thank Malene
Urbanus for her efforts with the cross-linking experiments. This work
was supported by EU grant HPRN-CT-2000-00075 from the European
Community.
REFERENCES
1. Fekkes, P. & Driessen, A.J. (1999) Protein targeting to the
bacterial cytoplasmic membrane. Microbiol. Mol. Biol. Rev. 63,
161–173.
2. Mu
¨
ller, M., Koch, H.G., Beck, K. & Scha
¨
fer, U. (2001) Protein
traffic in bacteria: multiple routes from the ribosome to and across
the membrane. Prog. Nucleic Acids Res. Mol. Biol. 66, 107–157.
3. de Cock, H., Overeem, W. & Tommassen, J. (1992) Biogenesis of
outer-membrane protein PhoE of Escherichia coli. Evidence for
multiple SecB-binding sites in the mature portion of the PhoE

dynamic gateway at the ER membrane. Annu. Rev. Cell. Dev. Biol.
15, 799–842.
13. Valent, Q.A., de Gier, J.W., von Heijne, G., Kendall, D.A., ten
Hagen-Jongman, C.M., Oudega, B. & Luirink, J. (1997) Nascent
membrane and presecretory proteins synthesized in Escherichia
coli associate with signal-recognition particle and trigger factor.
Mol. Microbiol. 25, 53–64.
14. Lee, H.C. & Bernstein, H.D. (2001) The targeting pathway of
Escherichia coli presecretory and integral membrane proteins is
specified by the hydrophobicity of the targeting signal. Proc. Natl.
Acad. Sci. USA 98, 3471–3476.
15. Luirink, J., ten Hagen-Jongman, C.M., van der Weijden, C.C.,
Oudega, B., High, S., Dobberstein, B. & Kusters, R. (1994) An
alternative protein targeting pathway in Escherichia coli:studies
on the role of FtsY. EMBO J. 13, 2289–2296.
16. de Leeuw, E., Poland, D., Mol, O., Sinning, I., ten Hagen-
Jongman, C.M., Oudega, B. & Luirink, J. (1997) Membrane as-
sociation of FtsY, the E.coliSRP receptor. FEBS Lett. 416,
225–229.
17. Ulbrandt, N.D., Newitt, J.A. & Bernstein, H.D. (1997) The E.coli
signal-recognition particle is required for the insertion of a subset
of inner-membrane proteins. Cell 88, 187–196.
18. Watson, M.E. (1984) Compilation of published signal sequences.
Nucleic Acids Res. 12, 5145–5164.
19. von Heijne, G. (1985) Signal sequences. The limits of variation.
J. Mol. Biol. 184, 99–105.
5570 H. Adams et al.(Eur. J. Biochem. 269) Ó FEBS 2002
20. Emr, S.D. & Silhavy, T.J. (1983) Importance of secondary struc-
ture in the signal sequence for protein secretion. Proc. Natl. Acad.
Sci. USA 80, 4599–4603.

28. Shinde, U.P., Guru Row, T.N. & Mawal, Y.R. (1989) Export of
proteins across membranes: the helix reversion hypothesis. Biosci.
Rep. 9, 737–745.
29. Nouwen, N., de Kruijff, B. & Tommassen, J. (1996) DlH
+
dependency of in vitro protein translocation into Escherichia
coli inner-membrane vesicles varies with the signal sequence core
region composition. Mol. Microbiol. 19, 1205–1214.
30. Bosch, D., Leunissen, J., Verbakel, J., de Jong, M., van Erp, H. &
Tommassen, J. (1986) Periplasmic accumulation of truncated
forms of outer-membrane PhoE protein of Escherichia coli K-12.
J. Mol. Biol. 189, 449–455.
31. Bosch, D., de Boer, P., Bitter, W. & Tommassen, J. (1989) The role
of the positively charged N-terminus of the signal sequence of
E.coliouter-membrane protein PhoE in export. Biochim. Biophys.
Acta 979, 69–76.
32. Lugtenberg, B., Meijers, J., Peters, R., van der Hoek, P. & van
Alphen, L. (1975) Electrophoretic resolution of the Ômajor outer-
membrane proteinÕ of Escherichia coli K-12 into four bands. FEBS
Lett. 58, 254–258.
33. de Vrije, T., Tommassen, J. & de Kruijff, B. (1987) Optimal
posttranslational translocation of the precursor of PhoE protein
across Escherichia coli membrane vesicles requires both ATP and
the proton-motive force. Biochim. Biophys. Acta 900, 63–72.
34. Ro
¨
misch, K., Webb, J., Lingelbach, K., Gausepohl, H. &
Dobberstein, B. (1990) The 54-kD protein of signal-recognition
particle contains a methionine-rich RNA binding domain. J. Cell
Biol. 111, 1793–1802.

(2000) SRP-dependent co-translational targeting and SecA-de-
pendent translocation analyzed as individual steps in the export of
a bacterial protein. EMBO J. 19, 6419–6426.
43. Powers, T. & Walter, P. (1997) Co-translational protein targeting
catalyzed by the Escherichia coli signal-recognition particle and its
receptor. EMBO J. 16, 4880–4886.
44. Valent, Q.A., Kendall, D.A., High, S., Kusters, R., Oudega, B. &
Luirink, J. (1995) Early events in preprotein recognition in E.coli:
interaction of SRP and trigger factor with nascent polypeptides.
EMBO J. 14, 5494–5505.
45. Bernstein, H.D., Zopf, D., Freymann, D.M. & Walter, P. (1993)
Functional substitution of the signal-recognition particle 54-kDa
subunit by its Escherichia coli homologue. Proc. Natl. Acad. Sci.
USA 90, 5229–5233.
46. Luirink, J., High, S., Wood, H., Giner, A., Tollervey, D. &
Dobberstein, B. (1992) Signal-sequence recognition by an
Escherichia coli ribonucleoprotein complex. Nature (London) 359,
741–743.
47. Beck, K., Wu, L.F., Brunner, J. & Mu
¨
ller, M. (2000) Dis-
crimination between SRP- and SecA/SecB-dependent substrates
involves selective recognition of nascent chains by SRP and trigger
factor. EMBO J. 19, 134–143.
48. Adams, H., Scotti, P.A., Luirink, J. & Tommassen, J. (2002) De-
fective translocation of a signal sequence mutant in a prlA4 sup-
pressor strain of Escherichia coli. Eur. J. Biochem. 269, 5572–5580.
49. Tommassen, J., van Tol, H. & Lugtenberg, B. (1983) The ultimate
localization of an outer-membrane protein of Escherichia coli
K-12 is not determined by the signal sequence. EMBO J. 2, 1275–