Modeling Escherichia coli signal peptidase complex with bound substrate: determinants in the mature peptide influencing signal peptide cleavage
© Choo et al; licensee BioMed Central Ltd. 2008
Published: 13 February 2008
Type I signal peptidases (SPases) are essential membrane-bound serine proteases responsible for the cleavage of signal peptides from proteins that are translocated across biological membranes. The crystal structure of SPase in complex with signal peptide has not been solved and their substrate-binding site and binding specificities remain poorly understood. We report here a structure-based model for Escherichia coli DsbA 13–25 in complex with its endogenous type I SPase.
The bound structure of DsbA 13–25 in complex with its endogenous type I SPase reported here reveals the existence of an extended conformation of the precursor protein with a pronounced backbone twist between positions P3 and P1'. Residues 13–25 of DsbA occupy, and thereby define 13 subsites, S7 to S6', within the SPase substrate-binding site. The newly defined subsites, S1' to S6' play critical roles in the substrate specificities of E. coli SPase. Our results are in accord with available experimental data.
Collectively, the results of this study provide interesting new insights into the binding conformation of signal peptides and the substrate-binding site of E. coli SPase. This is the first report on the modeling of a precursor protein into the entire SPase binding site. Together with the conserved precursor protein binding conformation, the existing and newly identified substrate binding sites readily explain SPase cleavage fidelity, consistent with existing biochemical results and solution structures of inhibitors in complex with E. coli SPase. Our data suggests that both signal and mature moiety sequences play important roles and should be considered in the development of predictive tools.
Translocation across the cell membrane requires the presence of short signal sequences termed "signal peptides" that are localized at the amino terminus (N-terminus) of proteins . These N-termini localized signal peptides are subsequently removed from the newly synthesized precursor proteins by type I signal peptidases (SPases) . SPases play essential roles in the viability of bacteria [3, 4], making these enzymes attractive targets for the design of novel antibiotics . Currently, Escherichia coli is by far the most widely used host organism for the bacterial expression of heterologous secreted proteins, especially for therapeutic purposes, with reported yields of 5–10 g/L . Mutations in the signal peptide have been known to affect secretion either by enhancing the processing of the cleavage site or by inhibiting this proteolytic processing . It is well known that besides the signal peptide, the N-terminal region of the mature protein also affects the protein secretion . We are therefore interested in understanding the determinants involved in signal peptide recognition, binding and cleavage.
The E. coli type I SPase is of particular interest in the study of type I SPases, as its active site is relatively accessible at the bacterial membrane surface [5, 9–11]. Although many mutational and biochemical studies have been performed, basic questions such as SPase fidelity and substrate specificity remain unanswered. Signal peptides exhibit limited primary sequence homology, but are well conserved at residues positioned -3 (P3) to -1 (P1) relative to the cleavage site, designated P1-P1' . Comparative analysis of 36 prokaryotic signal peptides reveals that type I SPases specifically recognizes substrates with small neutral residues at both the P3 and P1 positions . P3 is dominated by the presence of alanine, glycine, serine, threonine and valine; while P1 is characterized by alanine, glycine, serine and threonine [12, 13]. Accordingly, the P3 and P1 positions have been proposed to constitute the SPase cleavage site and have been actively applied by various groups for predicting signal peptide cleavage sites [12, 14]. These findings were cited as affirmation of the location of two key determinants within the signal peptide cleavage site. Unfortunately, no solution structures exist that can illustrate precisely how the precursor protein is oriented within the SPase substrate-binding site prior to proteolysis, or the identity of other critical determinants that control substrate specificity .
In this paper, we report the modeling of an E. coli periplasmic dithiol oxidase, DsbA 13–25 in complex with E. coli type I SPase based on the crystal structures of E. coli SPase in complex with β-lactam  and lipopeptide  inhibitors. The DsbA 13–25 precursor protein was selected for this study due to its efficient periplasmic secretion . By threading the P7 to P1' positions against the solved structures of β-lactam  and lipopeptide  inhibitors, our model reveals that precursor protein is bound to E. coli type I SPase with a pronounced twist between positions P3 and P1'. Thirteen subsites S7 to S6' were identified that might be critical to these and other aspects of catalysis. Our model was additionally corroborated by comparative analysis of 107 experimentally validated substrates.
Results and discussion
Substrate binding site
In our model, the bound precursor protein makes significant contact with E. coli SPase I from S7 to S6'. Models described earlier only focused on the P3-P1' segment and did not analyze in full the different substrate binding pockets on either side of the scissile bond. In particular, the S2 subsite was formerly proposed as the S1 subsite by Paetzel et al. [9, 19] as it largely overlaps with the latter. In contrast to the analysis by Paetzel et al. , our model reveals that the Ser18 (P2) side chain is not solvent exposed but is completely buried at this location. The ability of S3'/S4' to alter their electrostatic requirements by varying side chain conformations (Figure 3) may help explain the propensity to find substrates with charged amino acids at these positions (discussed in detail in Substrate Specificity).
Substrate binding conformation
Ten positions for hydrogen bonding were identified supporting high affinity binding between E. coli SPase and DsbA 13–25. These include Ser18 (P2) O...Ser88 NH, Ser18 (P2) O...Ser88 OG, Ala19 (P1) N...Ser88 OG, Gly89 N...Gln21 (P2') OE1, Ala19 (P1) N...Ser90 OG, Ser90 OG...Ala20 (P1') O, Lys145 NZ...Ala19 (P1) O, Gln194 NE2...Asp24 (P5') OD2, Ser206 OG...Asp24 (P5') OD2 and Arg282 NH1...Glu23 (P4') OE1. Our model suggests that the enzyme-substrate contact points extend all the way from P7 to P6' of the DsbA precursor protein.
The orientation of DsbA 13–25 side chains within the active site (P7-P6') of E. coli SPase adopts the pattern : ↓ • • • ↓ ↓ ↓ • • ↓ ↓ • • (where ↓ represents a side chain oriented towards the binding site and • represents a side chain oriented away or across the binding site). Specifically, the P3-P1' segment adopts the pattern: ↓ ↓ ↓ •, with the side chains of P3, P2 and P1 oriented towards the binding groove thereby supporting the stringent selectivity criteria in this region. The side chain of P1' alone is oriented differently, in accord with the observed variability in this position. A similar conformation was obtained for the precursor sequence OmpA 15–27 [21, 22] H2N-FATVAQAΔATSTKK-COOH (P1-P1' cleavage site indicated by Δ) in complex with E. coli type 1 SPase (data not shown). Here again, the P3-P1' side chains of OmpA adopt the orientation ↓ ↓ ↓ •, while the model proposed by Paetzel et al.  and Ekici et al. , adopts the pattern ↓ • ↓ •, with the side chain of P2 not pointing towards the binding groove. The disparity between our model and Paetzel et al.  may be attributed to the selection of different template structures where the structures of the covalently bound peptide inhibitor complex and the analogous enzyme LexA were used to guide the P1 and P3 to P6 positions of the later , while the coordinates of P7 to P1' for our model were guided by the solved structures of β-lactam  and lipopeptide  inhibitors in complex with E. coli SPase. In our models, the P2 side chains in the bound DsbA and OmpA models are hydrogen-bonded to the catalytically important SPase I residue, Ser88 . The twist in the backbone conformation in the region P3-P1' is representative of the transition state, with three critical hydrogen bonds conserved in our model and the bound β-lactamase and lipoprotein inhibitors, with the atoms Ser88 Oγ, Ser90 Oγ and Lys145 Nζ important for catalytic activity.
We have developed a structural model for E. coli SPase I complex, with bound DsbA and OmpA propeptides, consistent with existing biochemical data and solution structures of E. coli type I SPase-inhibitor complexes [16, 17]. The developed models provided an opportunity to examine the bound structure of E. coli type I SPase complex that have been difficult to solve experimentally. It appears that both signal and mature moiety sequences play direct role in catalysis. This work advances our understanding of the molecular mechanism governing signal peptide specificities and SPase fidelity, it also serves as a useful guide for designing suitable signal and mature peptide sequences for enhancing heterologous protein expression using E. coli as the host organism.
Precursor protein sequence data
The SPdb database http://proline.bic.nus.edu.sg/spdb/ was used as a source for the E. coli type I SPase substrates. A set of 107 precursor protein sequences was extracted and used in the current analysis where sequences with 80% sequence identity were extracted using the CD-HIT software  to reduce redundancy and bias .
The atomic coordinates of E. coli type I SPase were extracted from the Protein Databank (PDB) entry 1B12  which has a 1.95 Å resolution structure. Atomic coordinates for E. coli type I SPase-bound β-lactam  and lipopeptide  inhibitors were retrieved from PDB entries 1B12 and 1T7D respectively. The structures were relaxed by conjugate gradient minimization, using the Internal Coordinate Mechanics (ICM) software package .
The coordinates for the P7 to P1' positions of DsbA 13–25 were obtained by threading against the crystallographic structures of E. coli type I SPase-bound inhibitors [16, 17]. The P7 to P3 positions were taken from the structure of E. coli type I SPase-bound lipopeptide inhibitor  by substituting the atoms N1, C2, C5, O6, N7, C8, C10, O11, N12, C13, C14, O15, N16, C18, C26, O27, N28, C29, C31, O32 and N33 with DsbA 13–17 main-chain atoms; while coordinates for the P2 to P1' positions were guided by the solution structure of the E. coli type I SPase-bound β-lactam  inhibitor based on the atoms N4, C5, C6, C3, C9, O10, C15, C16, O17, C18, O19, C20. A flexible docking using biased Monte-Carlo procedure [29–31] that incorporates the Rapid Exact-Boundary ELectrostatics (REBEL) algorithm for evaluation of the electrostatic solvation energy  was subsequently performed to sample different conformations and orientations of P2' to P6' positions with respect to the receptor. In each iteration, a random move in the P2' to P6' positions of the ligand was performed and new conformations were selected based on the Metropolis criterion with a temperature of 5000 K [31, 33]. The simulation was terminated after 20,000 energy evaluations  and the results analyzed for consistency.
Intermolecular hydrogen bonds
The number of intermolecular hydrogen bonds was calculated using HBPLUS  in which hydrogen bonds are defined in accordance with the standard geometric parameters.
This article has been published as part of BMC Bioinformatics Volume 9 Supplement 1, 2008: Asia Pacific Bioinformatics Network (APBioNet) Sixth International Conference on Bioinformatics (InCoB2007). The full contents of the supplement are available online at http://www.biomedcentral.com/1471-2105/9?issue=S1.
- Mullins C, Meyer H-A, Hartmann E, Green N, Fang H: Structurally related SPC1p and SPC2p of yeast signal peptidase complex and functionally distinct. J Biol Chem 1996, 271: 29094–29099. 10.1074/jbc.271.46.29094View ArticlePubMedGoogle Scholar
- von Heijne G: Signal Peptidases. Molecular Biology Intelligence Unit. Austin: RG Landes Company; 1994.Google Scholar
- Date T: Demonstration by a novel genetic technique that leader peptidase is an essential enzyme in Escherichia coli . J Bacteriol 1983, 154: 76–83.PubMed CentralPubMedGoogle Scholar
- Klug G, Jager A, Heck C, Rauhut R: Identification, sequence analysis, and expression of the lepB gene for a leader peptidase in Rhodobacter capsulatus . Mol Gen Genet 1997, 253: 666–673. 10.1007/s004380050370View ArticlePubMedGoogle Scholar
- Black MT, Bruton G: Inhibitors of bacterial signal peptidase. Curr Pharm Des 1998, 4: 133–154.PubMedGoogle Scholar
- Georgiou G, Segatori L: Preparative expression of secreted proteins in bacteria: status report and future prospects. Curr Opin Biotechnol 2005, 16: 538–45. 10.1016/j.copbio.2005.07.008View ArticlePubMedGoogle Scholar
- Martoglio B, Dobberstein B: Signal sequences: more than just greasy peptides. Trends Cell Biol 1998, 8: 410–5. 10.1016/S0962-8924(98)01360-9View ArticlePubMedGoogle Scholar
- Andersson H, von Heijne G: A 30-residue-long "export initiation domain" adjacent to the signal sequence is critical for protein translocation across the inner membrane of Escherichia coli . Proc Natl Acad Sci USA 1991, 88: 9751–9754. 10.1073/pnas.88.21.9751PubMed CentralView ArticlePubMedGoogle Scholar
- Paetzel M, Dalbey RE, Strynadka NCJ: The structure and mechanism of bacterial type I signal peptidases. A novel antibiotic target. Pharmacol Ther 2000, 87: 27–49. 10.1016/S0163-7258(00)00064-4View ArticlePubMedGoogle Scholar
- Wolfe PB, Wickner W, Goodman JM: Sequence of the leader peptidase gene of Escherichia coli and the orientation of leader peptidase in the bacterial envelope. J Biol Chem 1983, 258: 12073–12080.PubMedGoogle Scholar
- Wolfe PB, Zwizzinski C, Wickner W: Purification and characterization of leader peptidase from Escherichia coli . Methods Enzymol 1983, 97: 40–57.View ArticlePubMedGoogle Scholar
- Fikes JD, Barkocy-Gallagher GA, Klapper DG, Bassford PJ Jr: Maturation of Escherichia coli maltose-binding protein by signal peptidase I in vivo: sequence requirement for efficient processing and demonstration of an alternative cleavage site. J Biol Chem 1990, 265: 3417–3423.PubMedGoogle Scholar
- von Heijne G: A new method for predicting signal sequence cleavage sites. Nucleic Acids Res 1986, 14: 4683–4690. 10.1093/nar/14.11.4683PubMed CentralView ArticlePubMedGoogle Scholar
- Folz RJ, Notwehr SF, Gordon JI: Substrate specificity of eukaryotic signal peptidase. Site-saturation mutagenesis at position -1 regulates cleavage between multiple sites in human (delta pro) apolipoprotein A-II. J Biol Chem 1988, 263: 2070–2078.PubMedGoogle Scholar
- Karla A, Lively MO, Paetzel M, Dalbey R: The identification of residues that control signal peptidase cleavage fidelity and substrate specificity. J Biol Chem 2005, 280: 6731–6741. 10.1074/jbc.M413019200View ArticlePubMedGoogle Scholar
- Paetzel M, Dalbey RE, Strynadka NCJ: Crystal structure of a bacterial signal peptidase in complex with a β-lactam inhibitor. Nature 1998, 396: 186–190. 10.1038/25403View ArticlePubMedGoogle Scholar
- Paetzel M, Goodall JJ, Kania M, Dalbey RE, Page MGP: Crystallographic and biophysical analysis of a bacterial signal peptidase in complex with a lipopeptide-based inhibitor. J Biol Chem 2004, 279: 30781–30790. 10.1074/jbc.M401686200View ArticlePubMedGoogle Scholar
- Perna NT, Plunkett G 3rd, Burland V, Mau B, Glasner JD, Rose DJ, Mayhew GF, Evans PS, Gregor J, Kirkpatrick HA, Posfai G, Hackett J, Klink S, Boutin A, Shao Y, Miller L, Grotbeck EJ, Davis NW, Lim A, Dimalanta ET, Potamousis KD, Apodaca J, Anantharaman TS, Lin J, Yen G, Schwartz DC, Welch RA, Blattner FR: Genome sequence of enterohaemorrhagic Escherichia coli O157:H7. Nature 2001, 409: 529–533. 10.1038/35054089View ArticlePubMedGoogle Scholar
- Paetzel M, Dalbey RE, Strynadka NCJ: Crystal structure of a bacterial signal peptidase apo-enzyme: Implications for signal peptide binding and the ser-lys dyad mechanism. J Biol Chem 2002, 277: 9512–9519. 10.1074/jbc.M110983200View ArticlePubMedGoogle Scholar
- Tong JC, Tan TW, Ranganathan S: Modeling the structure of bound peptide ligands to major histocompatibility complex. Protein Sci 2004, 13: 2523–2532. 10.1110/ps.04631204PubMed CentralView ArticlePubMedGoogle Scholar
- Carlos JL, Paetzel M, Brubaker G, Karla A, Ashwell CM, Lively MO, Cao G, Bullinger P, Dalbey RE: The role of the membrane spanning domain of type I signal peptidase in cleavage site selection. J Biol Chem 2000, 275: 38813–38822. 10.1074/jbc.M007093200View ArticlePubMedGoogle Scholar
- Ekici OD, Karla A, Paetzel M, Lively MO, Pei D, Dalbey RE: Altered -3 substrate specificity of E. coli signal peptidase 1 mutants as revealed by screening a combinatorial peptide library. J Biol Chem 2007, 282: 417–425. 10.1074/jbc.M608779200View ArticlePubMedGoogle Scholar
- Paetzel M, Goodall JJ, Kania M, Dalbey RE, Page MG: Crystallographic and Biophysical Analysis of a Bacterial Signal Peptidase in Complex with a Lipopeptide-based Inhibitor. J Biol Chem 2007, 279: 30781–30790. 10.1074/jbc.M401686200View ArticleGoogle Scholar
- Paetzel M, Strynadka NCJ: Signal peptide cleavage in the E. coli membrane. CSBMCB Bulletin 2001.Google Scholar
- van Roosmalen ML, Geukens N, Jongbloed JDH, Tjalsma H, Dubois J-YF, Bron S, van Dijl JM, Anńe J: Type I signal peptidases of Gram-positive bacteria. Biochim Biophys Acta 2004, 1694: 279–297. 10.1016/j.bbamcr.2004.05.006View ArticlePubMedGoogle Scholar
- Choo KH, Tan TW, Ranganathan S: SPdb – a signal peptide database. BMC Bioinformatics 2005, 6: 249. 10.1186/1471-2105-6-249PubMed CentralView ArticlePubMedGoogle Scholar
- Li W, Godzik A: Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences. Bioinformatics 2006, 22: 1658–1659. 10.1093/bioinformatics/btl158View ArticlePubMedGoogle Scholar
- Kajava AV, Zolov SN, Pyatkov KI, Kalinin AE, Nesmeyanova MA: Processing of Escherichia coli alkaline phosphatase. Sequence requirements and possible conformations of the -6 to -4 region of the signal peptide. J Biol Chem 2002, 277: 50396–50402. 10.1074/jbc.M205781200View ArticlePubMedGoogle Scholar
- Abagyan R, Totrov M, Kuznetsov D: ICM – a new method for protein modeling and design: Applications to docking and structure prediction from the distorted native conformation. J Comp Chem 1994, 15: 488–506. 10.1002/jcc.540150503View ArticleGoogle Scholar
- Abagyan R, Totrov M: Ab initio folding of peptides by the optimal-bias Monte Carlo minimization procedure. J Comput Phys 1999, 151: 402–421. 10.1006/jcph.1999.6233View ArticleGoogle Scholar
- Fernández-Recio J, Totrov M, Abagyan R: Soft protein-protein docking in internal coordinates. Protein Sci 2002, 11: 280–291. 10.1110/ps.19202PubMed CentralView ArticlePubMedGoogle Scholar
- Totrov M, Abagyan R: Rapid boundary element solvation electrostatics calculations in folding simulations: Successful folding of a 23-residue peptide. Biopolymers 2001, 60: 124–133. 10.1002/1097-0282(2001)60:2<124::AID-BIP1008>3.0.CO;2-SView ArticlePubMedGoogle Scholar
- Metropolis NA, Rosenbluth AW, Rosenbluth NM, Teller AH, Teller E: Equation of State Calculations by Fast Computing Machines. J Chem Phys 1953, 21: 1087–1092. 10.1063/1.1699114View ArticleGoogle Scholar
- McDonald IK, Thornton JM: Satisfying hydrogen bonding potential in proteins. J Mol Biol 1994, 238: 777–793. 10.1006/jmbi.1994.1334View ArticlePubMedGoogle Scholar
- Crooks GE, Hon G, Chandonia JM, Brenner SE: WebLogo: a sequence logo generator. Genome Res 2004, 14: 1188–1190. 10.1101/gr.849004PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.