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 helix breaker remained obscure. Here, we considered the possibility that extension of the a-helical structure in the signal sequence resulting from the Gly )10 substitutions affects the targeting pathway of the precursor. Indeed, the mutations resulted in reduced dependency on SecB for tar- geting in vivo. In vitro cross-linking experiments revealed that the G-10L and G-10C mutant PhoE precursors had a dra- matically increased affinity for P48, one of the constituents of the signal-recognition particle (SRP). Furthermore, in vitro cross-linking experiments revealed that the G-10L mutant protein is routed to the SecYEG translocon via the SRP pathway, the targeting pathway that is exploited by integral inner-membrane proteins. Together, these data indicate that the helix breaker in cleavable signal sequences prevents recognition by SRP and is thereby, together with the hydrophobicity of the signal sequence, a determinant of the targeting pathway. Keywords: outer-membrane protein; Sec translocon; SecB; signal-recognition particle; translocation. Most cell envelope proteins of Escherichia coli are translo- cated across or inserted into the cytoplasmic membrane via the membrane-embedded Sec translocon. Targeting of precursor proteins to the translocon can be mediated by components of the Sec pathway or by the signal-recognition particle (SRP) pathway [1,2]. The Sec pathway utilizes a cytosolic chaperone, SecB, which interacts with the mature portion of presecretory proteins [3,4]. The SecB-preprotein complexisthentargetedtoSecA,whichinturninteracts 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- ized by a positively charged N-terminal region (N domain), followed by a 10–15 residues long hydrophobic core (H domain) and a polar C-terminus (C domain) containing the signal-peptidase cleavage site [19]. The importance of a-helix formation in the signal sequence is well documented [20–24]. However, NMR studies on the conformation of signal sequences in a membrane mimetic environment showed that the stable a-helix is disrupted towards the C-terminus of the hydrophobic core [25–27]. Furthermore, a statistical Correspondence to J. Tommassen, Department of Molecular Microbiology, Utrecht University, Padualaan 8, 3584 CH Utrecht, The Netherlands. Fax: + 31 30 2513655, Tel.: + 31 30 2532999, E-mail: J.P.M.Tommassen@bio.uu.nl Abbreviations: SRP, signal-recognition particle; pmf, proton-motive force; RNC, ribosome-nascent chain; BS 3 , bis(sulfosuccinimidyl)- suberate; DSS, disuccinimidyl-suberate; IMV, inverted inner- membrane vesicles; TF, trigger factor. *Present address:IECB-E ´ cole polytechnique ENSCPB, Talence cedex, France. (Received 29 May 2002, revised 10 September 2002, accepted 16 September 2002) Eur. J. Biochem. 269, 5564–5571 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03262.x analysis of signal sequences revealed the common occurrence of a-helix-destabilizing amino acids in the hydrophobic core [28]. In a previous study, the role of the a-helix-breaking glycine residue at position )10 (Gly )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 ) and disuccinimidyl-suberate (DSS) were from Pierce, and oligonucleotides were pur- chased from Isogen Bioscience (Maarsen, the Netherlands). Bacterial strains The E. coli K-12strainsusedinthisstudyarelistedin Table 1. Strains CE1514 and CE1515 were obtained by P1 transduction using strain CE1224 as the recipient and strains IQ85 and strain MM152, respectively, as donor strains. To obtain strain CE1513, strain MM88 was used as Fig. 1. Physical characteristics of the PhoE signal sequence. The signal sequence consists of the positively charged N domain, the hydrophobic H domain and the C-terminal C domain. The a-helix in the H domain is predicted to extend up to the Gly at position )10 in the signal sequence. Introduction of an a-helix-stabilizing residue (Ala, Cys or Leu) at position )10 results in extension of the a-helical core region as indicated. The leader peptidase cleavage site is depicted with an arrow. Table 1. Bacterial strains and plasmids used in this study. Ts, temperature sensitive. Cam r and Amp r , resistance to chloramphenicol and ampicillin, respectively. Designation Relevant characteristics Description/reference Strains CE1224 F – , thr leu D(proA-proB-phoE-gpt) his thi argE lacY galK xyl rpsL supE ompR [49] MC4100 F – , DlacU169 araD139 rpsL thi relA [50] MM88 F – , DlacU169 araD139 thiA rpsL relA leu::Tn10 secAtsA51 B. Oudega (pers. comm.) 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 Ó FEBS 2002 Re-routing a secretory protein via the SRP pathway (Eur. J. Biochem. 269) 5565 the donor and CE1224 as the recipient in a P1 transduction experiment. Plasmid construction Plasmid pJP29 and derivatives carrying mutations in the PhoE signal-sequence-encoding region and other plasmids are listed in Table 1. Plasmid pC4Meth94PhoE was used to generate truncated phoE mRNA, encoding a 94-residue PhoE polypeptide exposing the signal sequence just outside the ribosome [13]. Plasmid pC4Meth101FtsQ-WT was used to generate truncated FtsQ mRNA, encoding a 101-residue FtsQ polypeptide exposing the signal-anchor domain just outside the ribosome. To compensate for the loss of methionines from the deleted domains of the proteins, both constructs contain a C-terminal tetra-methionine tag sequence for labeling. To introduce the Cys and Leu mutations for the Gly )10 residue into pC4Meth94PhoE, the EcoRI/BamHI fragment of the plasmid was replaced by PCR fragments created using the PhoE forward primer (5¢-GCCGGAATTCTAATATGAAAAAGAGCACTCT GGC-3¢) and the 94PhoE reverse primer (5¢-CGCGGA TCCTTTTTGCTGTCAGTATCAC-3¢), pNN7 and pNN8 as the templates, respectively, and Pfu polymerase. The resulting plasmids were designated pC4Meth(G-10C)94 PhoE and pC4Meth(G-10L)94PhoE, respectively. In vivo pulse–chase experiments Cells of strain CE1224 or its derivatives each containing a 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- serum [31] by SDS/PAGE [32] and visualized by autoradio- graphy. In vitro transcription, translation, targeting and cross-linking analysis To generate truncated mRNA, plasmids (Table 1) encoding truncated nascent chains were linearized and transcribed as described previously [13]. The resulting mRNAs were translated in vitro in a lysate of strain MC4100 as described [13,33]. The mixture containing RNCs was chilled on ice and treated with 1 m M BS 3 at 25 °C for 10 min before addition of 0.1 vol. quench buffer (1 M glycine/0.1 M NaHCO 3 , pH 8.5). After incubation for 20 min at 0 °C, cross-linked products were immunoprecipitated [34], and the precipitates were analyzed by SDS/PAGE (12% gels). Radiolabeled proteins were visualized with a Phosphor- Imager 473 (Molecular Dynamics) and quantified using the Imagequant software (Molecular Dynamics). To test the targeting of wild-type prePhoE RNCs, truncated mRNAs were translated in the presence of inverted inner-membrane vesicles (IMVs) [33] from strain MC4100. After cross- 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 by Na 2 CO 3 extraction as described [35]. Samples were analyzed either directly or after immunoprecipitation on 12% or 15% gels and visualized as described above. RESULTS SecB dependency of the targeting of mutant prePhoE By the substitution of an a-helix-promoting residue (Leu, Ala or Cys) for the helix-breaking Gly )10 of the signal sequence of PhoE, the a-helix is expected to be extended considerably (Fig. 1). As these mutant signal sequences resemble more closely the membrane-spanning domains of integral-membrane proteins, the mutations might affect the targeting route of the precursors to the SecYEG translocon. This possibility was first tested in vivo in pulse–chase experiments. The processing kinetics of the wild-type and mutant PhoE proteins were compared in a secB null mutant strain. Previously, it was demonstrated that introduction of an a-helix-stabilizing residue (Ala, Cys or Leu) instead of the Gly )10 did not result in dramatic differences in the processing kinetics of prePhoE in wild-type cells [29]. As the export of wild-type PhoE is SecB dependent [3], its precursor strongly accumulated in a secB mutant (Fig. 2A). Interestingly, the mutant precursors showed considerably 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 SRP pathway. To determine whether components of the SRP pathway are indeed involved in the targeting of (G-10L)prePhoE to the translocon, in vitro cross-linking studies were performed. Previously, Valent et al.[13] analyzed the interaction of nascent prePhoE protein with soluble proteins in an E. coli lysate. Nascent PhoE 94-mer extended with a tetra-methionine tag-sequence (94PhoE) was synthesized in an E. coli lysate and treated with the water-soluble cross-linker BS 3 . Whereas, in these experi- ments, nascent chains of integral inner-membrane proteins could be cross-linked to the P48 component of SRP, this was not the case for nascent 94PhoE [13]. To investigate whether substitution of the Gly )10 residue by an a-helix- stabilizing residue resulted in a higher affinity for P48, (G-10L)94PhoE and 94PhoE were synthesized and tested for cross-linking to P48 present in the E. coli lysate. Whereas hardly any cross-linked 94PhoE could be immu- noprecipitated with anti-P48 antibodies, strong cross-link- ing of (G-10L)94PhoE to P48 was observed (Fig. 3A). To determine whether the improved cross-linking of (G-10L) 94PhoE to P48 was due solely to the increased hydropho- bicity of this mutant signal sequence, similar cross-linking experiments were also performed for the (G-10C)94PhoE mutant PhoE protein. Even though cysteine has an even lower hydrophobicity than glycine on the consensus hydrophobicity scale of Eisenberg et al. [37], the (G-10C) 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 2 CO 3 pellet, at least two major (G-10L)94PhoE cross-linked complexes could be detected, one at  110 kDa and one at  46 kDa (Fig. 4A, lane 3). The 110-kDa complex could be immunoprecipitated with antiserum directed against SecA, indicating that it is a complex of the radiolabeled (G-10L)94PhoE and SecA (Fig. 4B, lane 1). In addition, cross-linking adducts of  220 kDa and  40 kDa were also immunoprecipitated from the Na 2 CO 3 pellet with anti-SecA serum. We assume that the  220-kDa product corresponds to cross-linked complexes between (G-10L)94PhoE and the dimeric form Fig. 2. In vivo processing kinetics of prePhoE and mutant prePhoE proteins in sec mutants. (A) Cells of secB mutant strain CE1515 carrying plasmid pJP29 encoding wild-type PhoE (WT) or derivatives were grown under phosphate limitation to express PhoE. The cells were pulse-labeled, followed by a chase for the indicated periods. PhoE proteins were immunoprecipitated, separated by SDS/PAGE followed by autoradiography. G-10A (G-10A)prePhoE; G-10C (G-10C)prePhoE; G-10L (G-10L)prePhoE. (B) SecAts51 and sec- Yts24 derivatives of CE1224 or their isogenic wild-type parental strain (wt) carrying plasmids pJP29 or pNN8, encoding prePhoE or (G-10L)prePhoE, respectively, were grown under phosphate limitation for 3 h at the permissive temperature (30 °C), subsequently for 2 h at 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 (Fig. 4B, lane 5 and 6), respectively. The identity of the low- molecular-mass adducts is unknown. As the signal sequence of 94PhoE has no affinity for P48 (Fig. 3), and the SecB- binding sites in the mature domain are not exposed from the ribosome in RNCs of 94PhoE, these RNCs cannot be targeted to the translocon. Consistently, no cross-linking adducts similar to those obtained with (G-10L)94PhoE were obtained, when 94PhoE and (G-10L)94PhoE nascent chains were incubated with IMVs after cross-linking with DSS (Fig. 4C, lanes 1–4). To investigate whether the cross- linking adducts of (G-10L)94PhoE that were obtained are similar to the cross-linking adducts with a known substrate of the SRP pathway, FtsQ was used as a model. This class II membrane protein, with a short N-terminal cytoplasmic tail [39], was synthesized as a slightly longer nascent chain (101 residues) than (G-10L)94PhoE to expose properly its signal- anchor domain. Indeed, 101FtsQ interacted properly with SecY and SecA (Fig. 4A, lane 8; Fig. 4B, lane 3 and 4). Furthermore, the same cross-linking efficiency was obtained for P48 (Fig. 4B, lane 8) as was observed for the (G-10L) prePhoE (Fig. 4B, lane 6), but TF was hardly cross-linked if at all (Fig. 4B, compare lane 5 and 7). In conclusion, these results show that (G-10L)94PhoE nascent chains are correctly targeted to the SecY protein in the translocon via the SRP pathway. SRP dependency of (G-10L)prePhoE in vivo As the experiments described above show that (G-10L) prePhoE is targeted in vitro to the Sec translocon via the SRP pathway, it was of interest to determine whether it is 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 in the signal sequence. Consistent with a considerable conformational change, these substitutions resulted in a higher electrophoretic mobility of the mutant precursors compared with that of wild-type prePhoE [29] (see also Fig. 2A), suggesting a more compact conformation of the Fig. 3. Cross-linking of soluble E. coli proteins to PhoE nascent chains and mutant derivatives. (A) [ 35 S]methionine-labeled nascent 94PhoE or mutant derivatives were synthesized in an E. coli lysate and treated with the homo-bifunctional chemical cross-linker BS 3 . After quench- ing, both P48- and TF-cross-linked complexes were immunoprecipi- tated with antisera directed against P48 and TF, analyzed on SDS/ PAGE and visualized with a PhosphorImager. (B) Quantification of data presented in panel (A), after correction for translation efficiency. The highest amounts of immunoprecipitated cross-linked nascent chains were obtained for (G-10L)prePhoE in the case of P48 cross- linked complexes and for WT prePhoE in the case of the TF cross- linked complexes. These amounts were set to 100%, and the relative cross-linking efficiencies of the other prePhoE forms to TF and P48 are shown. 5568 H. Adams et al.(Eur. J. Biochem. 269) Ó FEBS 2002 signal sequence. In addition, CD measurements on synthetic signal peptides showed a considerable increase in the a-helical content by the G-10L substitution [40]. Because of the extension of the a-helix, the mutant signal sequences more closely resemble the signal-anchor sequences of integral-membrane proteins than does the wild-type signal 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, because of overloading of the SRP pathway. The re-routing of (G-10L)prePhoE to the Sec translocon via the SRP instead of the SecB pathway could be explained by the increased hydrophobicity of the hydrophobic core of the mutant signal sequence, because hydrophobicity was previ- ously reported to be an important variable in the interaction with SRP [14,41,42]. However, the hydrophobicity of cysteine is even slightly lower than that of glycine [37]. Therefore, the cross-linking of (G-10C)prePhoE to P48 indicates that another variable, in addition to hydropho- bicity, contributes to the interaction of signal sequences with Fig. 4. Targeting of SRP–RNCs to the Sec translocon. [ 35 S]Methionine-labeled (G-10L)94PhoE or 101FtsQ was incubated with 350 n M reconstituted SRP. SRP–RNCs were purified and targeted to IMVs as des- cribed in Experimental procedures. The cross- linker DSS was used to analyze SRP–RNC interactions. After quenching, peripherally bound and soluble proteins were separated from the inner membranes by carbonate extraction. Samples were either (A) directly or (B) after immunoprecipitation (IP) with the indicated antisera, subjected to SDS/PAGE, and cross-linked complexes were visualized with a PhosphorImager. The positions of 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 cross-linking studies have revealed that the binding of TF to a sequence within the first 125 amino-acid residues of pro- OmpA (but beyond the signal peptide) excluded the association of the precursor to SRP [47]. This observation led to the proposal that secretory precursors are targeted to the SecB pathway when they emerge from the ribosome by means of their preferential recognition by TF. However, we found that a single amino-acid substitution (G-10L or G-10C) in the signal sequence of PhoE results in a high affinity for P48, even though TF is still bound to the G-10L PhoE precursor. Therefore, TF binding apparently does not prevent the binding to P48, although we cannot exclude the possibility that different (G-10L)prePhoE or (G-10C)pre- PhoE molecules bind to either TF or P48, but not to both at the same time. In the case of the 101FtsQ substrate, TF was not cross-linked efficiently whereas P48 was, in accordance with previous observations [35]. In general, our results are in agreement with the reported binding of TF to secretory precursors [47], but the basis for routing of secretory proteins to the SecB pathway appears not to be the exclusion of SRP by TF. More likely, the helix breaker present in the wild-type prePhoE signal sequence prevents interaction with SRP, whereas the hydrophobic core of the mutant signal sequences adopts a longer a-helical structure, which is recognized by SRP as a substrate. It is interesting to note that the natural signal sequences of at least some secreted proteins of Gram-positive bacteria, which do not possess a SecB pathway and might therefore be entirely dependent on the SRP pathway for protein secretion, also contain an extended a-helix and have functional characteristics similar 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 protein. J. Mol. Biol. 224, 369–379. 4. Knoblauch, N.T., Rudiger, S., Schonfeld, H.J., Driessen, A.J., Schneider-Mergener, J. & Bukau, B. (1999) Substrate specificity of the SecB chaperone. J. Biol. Chem. 274, 34219–34225. 5. Manting, E.H., van der Does, C. & Driessen, A.J. (1997) In vivo cross-linking of the SecA and SecY subunits of the Escherichia coli preprotein translocase. J. Bacteriol. 179, 5699–5704. 6. van der Does, C., Manting, E.H., Kaufmann, A., Lutz, M. & Driessen, A.J. (1998) Interaction between SecA and SecYEG in micellar solution and formation of the membrane-inserted state. Biochemistry 37, 201–210. 7. Fekkes, P., van der Does, C. & Driessen, A.J. (1997) The mole- cular chaperone SecB is released from the carboxy-terminus of SecA during initiation of precursor protein translocation. EMBO J. 16, 6105–6113. 8. Economou, A., Pogliano, J.A., Beckwith, J., Oliver, D.B. & Wickner, W. (1995) SecA membrane cycling at SecYEG is driven by distinct ATP binding and hydrolysis events and is regulated by SecD and SecF. Cell 83, 1171–1181. 9. Duong, F. & Wickner, W. (1997) The SecDFyajC domain of preprotein translocase controls preprotein movement by regulat- ing SecA membrane cycling. EMBO J. 16, 4871–4879. 10. Dalbey, R.E. & Wickner, W. (1985) Leader peptidase catalyzes the release of exported proteins from the outer surface of the Escherichia coli plasma membrane. J. Biol. Chem. 260, 15925– 15931. 11. Rapoport, T.A., Jungnickel, B. & Kutay, U. (1996) Protein transport across the eukaryotic endoplasmic reticulum and bac- terial inner membranes. Annu. Rev. Biochem. 65, 271–303. 12. Johnson, A.E. & van Waes, M.A. (1999) The translocon: a 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. 21. Briggs, M.S. & Gierasch, L.M. (1984) Exploring the conforma- tional roles of signal sequences: synthesis and conformational analysis of k receptor protein wild-type and mutant signal pep- tides. Biochemistry 23, 3111–3114. 22. Batenburg, A.M., Demel, R.A., Verkleij, A.J. & de Kruijff, B. (1988) Penetration of the signal sequence of Escherichia coli PhoE protein into phospholipid model membranes leads to lipid-specific changes in signal peptide structure and alterations of lipid orga- nization. Biochemistry 27, 5678–5685. 23. McKnight, C.J., Briggs, M.S. & Gierasch, L.M. (1989) Functional and nonfunctional LamB signal sequences can be distinguished by their biophysical properties. J. Biol. Chem. 264, 17293–17297. 24. Keller, R.C., Killian, J.A. & de Kruijff, B. (1992) Anionic phos- pholipids are essential for a-helix formation of the signal peptide of prePhoE upon interaction with phospholipid vesicles. Biochemistry 31, 1672–1677. 25. Wang, Z., Jones, J.D., Rizo, J. & Gierasch, L.M. (1993) Mem- brane-bound conformation of a signal peptide: a transferred nuclear Overhauser effect analysis. Biochemistry 32, 13991–13999. 26. Rizo, J., Blanco, F.J., Kobe, B., Bruch, M.D. & Gierasch, L.M. (1993) Conformational behavior of Escherichia coli OmpA signal peptides in membrane mimetic environments. Biochemistry 32, 4881–4894. 27. Chupin, V., Killian, J.A., Breg, J., de Jongh, H.H., Boelens, R., Kaptein, R. & de Kruijff, B. (1995) PhoE signal peptide inserts into micelles as a dynamic helix-break-helix structure, which is modulated by the environment. A two-dimensional 1 HNMR study. Biochemistry 34, 11617–11624. 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. 35. Valent, Q.A., Scotti, P.A., High, S., de Gier, J.W., von Heijne, G., Lentzen, G., Wintermeyer, W., Oudega, B. & Luirink, J. (1998) The Escherichia coli SRP and SecB-targeting pathways converge at the translocon. EMBO J. 17, 2504–2512. 36. High, S., Flint, N. & Dobberstein, B. (1991) Requirements for the membrane insertion of signal-anchor type proteins. J. Cell Biol. 113, 25–34. 37. Eisenberg, D., Schwarz, E., Komaromy, M. & Wall, R. (1984) Analysis of membrane and surface protein sequences with the hydrophobic moment plot. J. Mol. Biol. 179, 125–142. 38.Scotti,P.A.,Valent,Q.A.,Manting,E.H.,Urbanus,M.L., Driessen, A.J., Oudega, B. & Luirink, J. (1999) SecA is not required for signal-recognition particle-mediated targeting and initial membrane insertion of a nascent inner-membrane protein. J. Biol. Chem. 274, 29883–29888. 39. Carson, M.J., Barondess, J. & Beckwith, J. (1991) The FtsQ protein of Escherichia coli: membrane topology, abundance, and cell division phenotypes due to overproduction and insertion mutations. J. Bacteriol. 173, 2187–2195. 40. van Dalen, A., Killian, A. & de Kruijff, B. (1999) DW stimulates membrane translocation of the C-terminal part of a signal sequence. J. Biol. Chem. 274, 19913–19918. 41. Kim, J., Rusch, S., Luirink, J. & Kendall, D.A. (2001) Is Ffh required for export of secretory proteins? FEBS Lett. 505, 245– 248. 42. Neumann-Haefelin, C., Scha ¨ fer, U., Mu ¨ ller, M. & Koch, H.G. (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– 1279. 50. Casadaban, M.J. (1976) Transposition and fusion of the lac genes to selected promoters in Escherichia coli using bacteriophage k and l. J. Mol. Biol. 104, 541–555. 51. Shiba, K., Ito, K., Yura, T. & Cerretti, D.P. (1984) A defined mutation in the protein export gene within the spc ribosomal protein operon of Escherichia coli: isolation and characterization of a new temperature-sensitive secY mutant. EMBO J. 3, 631–635. 52. Ribes, V., Ro ¨ misch, K., Giner, A., Dobberstein, B. & Tollervey, D. (1990) E.coli4.5S RNA is part of a ribonucleoprotein particle that has properties related to signal-recognition particle. Cell 63, 591–600. Ó FEBS 2002 Re-routing a secretory protein via the SRP pathway (Eur. J. Biochem. 269) 5571 . The presence of a helix breaker in the hydrophobic core of signal sequences of secretory proteins prevents recognition by the signal- recognition particle. to the Sec translocon via the SRP instead of the SecB pathway could be explained by the increased hydrophobicity of the hydrophobic core of the mutant signal