Topology of the Outer Membrane Usher PapC Determined by Site-directed Fluorescence Labeling
2004; Elsevier BV; Volume: 279; Issue: 51 Linguagem: Inglês
10.1074/jbc.m409192200
ISSN1083-351X
AutoresNadine S. Henderson, Stephane Shu Kin So, Cheryl Martin, Ritwij Kulkarni, David G. Thanassi,
Tópico(s)Cell Adhesion Molecules Research
ResumoIn contrast to typical membrane proteins that span the lipid bilayer via transmembrane α-helices, bacterial outer membrane proteins adopt a β-barrel architecture composed of antiparallel transmembrane β-strands. The topology of outer membrane proteins is difficult to predict accurately using computer algorithms, and topology mapping protocols commonly used for α-helical membrane proteins do not work for β-barrel proteins. We present here the topology of the PapC usher, an outer membrane protein required for assembly and secretion of P pili by the chaperone/usher pathway in uropathogenic Escherichia coli. An initial attempt to map PapC topology by insertion of protease cleavage sites was largely unsuccessful due to lack of cleavage at most sites and the requirement to disrupt the outer membrane to identify periplasmic sites. We therefore adapted a site-directed fluorescence labeling technique to permit topology mapping of outer membrane proteins using small molecule probes in intact bacteria. Using this method, we demonstrated that PapC has the potential to encode up to 32 transmembrane β-strands. Based on experimental evidence, we propose that the usher consists of an N-terminal β-barrel domain comprised of 26 β-strands and that a distinct C-terminal domain is not inserted into the membrane but is located instead within the lumen of the N-terminal β-barrel similar to the plug domains encoded by the outer membrane iron-siderophore uptake proteins. In contrast to typical membrane proteins that span the lipid bilayer via transmembrane α-helices, bacterial outer membrane proteins adopt a β-barrel architecture composed of antiparallel transmembrane β-strands. The topology of outer membrane proteins is difficult to predict accurately using computer algorithms, and topology mapping protocols commonly used for α-helical membrane proteins do not work for β-barrel proteins. We present here the topology of the PapC usher, an outer membrane protein required for assembly and secretion of P pili by the chaperone/usher pathway in uropathogenic Escherichia coli. An initial attempt to map PapC topology by insertion of protease cleavage sites was largely unsuccessful due to lack of cleavage at most sites and the requirement to disrupt the outer membrane to identify periplasmic sites. We therefore adapted a site-directed fluorescence labeling technique to permit topology mapping of outer membrane proteins using small molecule probes in intact bacteria. Using this method, we demonstrated that PapC has the potential to encode up to 32 transmembrane β-strands. Based on experimental evidence, we propose that the usher consists of an N-terminal β-barrel domain comprised of 26 β-strands and that a distinct C-terminal domain is not inserted into the membrane but is located instead within the lumen of the N-terminal β-barrel similar to the plug domains encoded by the outer membrane iron-siderophore uptake proteins. All bacterial outer membrane (OM) 1The abbreviations used are: OM, outer membrane; TM, transmembrane; PhoA, alkaline phosphatase; HA, hemagglutination assay; TEV, tobacco etch virus; AMS, 4-acetamido-4′-maleimidylstilbene-2–2′-disulfonic acid; RT, room temperature; OGM, Oregon green 488 maleimide; DDM, dodecylmaltoside; WT, wild type. proteins with known structures span the membrane via a series of transmembrane (TM) antiparallel β-strands arranged to form a β-barrel (1Wimley W.C. Curr. Opin. Struct. Biol. 2003; 13: 404-411Crossref PubMed Scopus (345) Google Scholar, 2Tamm L.K. Arora A. Kleinschmidt J.H. J. Biol. Chem. 2001; 276: 32399-32402Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar, 3Schulz G.E. Curr. Opin. Struct. Biol. 2000; 10: 443-447Crossref PubMed Scopus (246) Google Scholar, 4Koebnik R. Locher K.P. Van Gelder P. Mol. Microbiol. 2000; 37: 239-253Crossref PubMed Scopus (908) Google Scholar). Proteins found in the OM of mitochondria and chloroplasts also adopt similar β-barrel folds (5Paschen S.A. Waizenegger T. Stan T. Preuss M. Cyrklaff M. Hell K. Rapaport D. Neupert W. Nature. 2003; 426: 862-866Crossref PubMed Scopus (347) Google Scholar). Only six to nine residues are required for a β-strand to cross the central non-polar region of the lipid bilayer (4Koebnik R. Locher K.P. Van Gelder P. Mol. Microbiol. 2000; 37: 239-253Crossref PubMed Scopus (908) Google Scholar), and only half of these need to be hydrophobic since only alternating residues face into the bilayer. This makes prediction of TM β-strands by sequence analysis difficult. Furthermore protocols typically used to map the topology of α-helical membrane proteins are not applicable to OM proteins. For example, alkaline phosphatase (PhoA) fusions are commonly used to identify periplasmic regions of inner membrane proteins (6Manoil C. Mekalanos J.J. Beckwith J. J. Bacteriol. 1990; 172: 515-518Crossref PubMed Google Scholar). However, PhoA cannot be used to map OM proteins reliably as the large PhoA insertion folds in the periplasm and disrupts proper assembly of proteins in the OM (7Newton S.M. Klebba P.E. Michel V. Hofnung M. Charbit A. J. Bacteriol. 1996; 178: 3447-3456Crossref PubMed Scopus (38) Google Scholar). Smaller insertions are generally tolerated in surface and periplasmic loops of β-barrel proteins, and introduction of epitopes and protease cleavage sites has been used to map OM proteins (7Newton S.M. Klebba P.E. Michel V. Hofnung M. Charbit A. J. Bacteriol. 1996; 178: 3447-3456Crossref PubMed Scopus (38) Google Scholar, 8Koebnik R. Braun V. J. Bacteriol. 1993; 175: 826-839Crossref PubMed Google Scholar, 9Guédin S. Willery E. Tommassen J. Fort E. Drobecq H. Locht C. Jacob-Dubuisson F. J. Biol. Chem. 2000; 275: 30202-30210Abstract Full Text Full Text PDF PubMed Scopus (84) Google Scholar, 10Merck K.B. de Cock H. Verheij H.M. Tommassen J. J. Bacteriol. 1997; 179: 3443-3450Crossref PubMed Google Scholar, 11Murphy C.K. Kalve V.I. Klebba P.E. J. Bacteriol. 1990; 172: 2736-2746Crossref PubMed Google Scholar). However, the mapping of such insertions in whole bacteria may be critically affected by the membrane or protein structure, and mapping of periplasmic regions requires disruption of the OM with the potential for misleading results (7Newton S.M. Klebba P.E. Michel V. Hofnung M. Charbit A. J. Bacteriol. 1996; 178: 3447-3456Crossref PubMed Scopus (38) Google Scholar, 9Guédin S. Willery E. Tommassen J. Fort E. Drobecq H. Locht C. Jacob-Dubuisson F. J. Biol. Chem. 2000; 275: 30202-30210Abstract Full Text Full Text PDF PubMed Scopus (84) Google Scholar, 10Merck K.B. de Cock H. Verheij H.M. Tommassen J. J. Bacteriol. 1997; 179: 3443-3450Crossref PubMed Google Scholar, 11Murphy C.K. Kalve V.I. Klebba P.E. J. Bacteriol. 1990; 172: 2736-2746Crossref PubMed Google Scholar). To overcome these difficulties, we adapted a site-directed fluorescence labeling protocol for topology mapping of OM proteins. Our method only requires insertion or substitution of single cysteine residues and permits analysis of OM proteins in their native state in intact bacteria. We used this protocol to map the topology of the PapC usher protein, which is required for assembly and secretion of P pili in uropathogenic Escherichia coli (12Dodson K.W. Jacob-Dubuisson F. Striker R.T. Hultgren S.J. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 3670-3674Crossref PubMed Scopus (171) Google Scholar, 13Norgren M. Baga M. Tennent J.M. Normark S. Mol. Microbiol. 1987; 1: 169-178Crossref PubMed Scopus (81) Google Scholar). P pili are composite structures built from multiple pilus subunits (14Bullitt E. Makowski L. Nature. 1995; 373: 164-167Crossref PubMed Scopus (191) Google Scholar, 15Kuehn M.J. Heuser J. Normark S. Hultgren S.J. Nature. 1992; 356: 252-255Crossref PubMed Scopus (264) Google Scholar). The PapG adhesin, located at the pilus tip, binds to Galα(1–4)Gal moieties present in kidney glycolipids, and P pili are key virulence factors for the development of pyelonephritis (16Soto G.E. Hultgren S.J. J. Bacteriol. 1999; 181: 1059-1071Crossref PubMed Google Scholar, 17Bock K. Breimer M.E. Brignole A. Hansson G.C. Karlsson K-A. Larson G. Leffler H. Samuelsson B.E. Strömberg N. Svanborg-Edén C. Thurin J. J. Biol. Chem. 1985; 260: 8545-8551Abstract Full Text PDF PubMed Google Scholar). P pili are assembled by the chaperone/usher secretion pathway, which is responsible for biogenesis of a superfamily of surface structures associated with virulence (18Thanassi D.G. Saulino E.T. Hultgren S.J. Curr. Opin. Microbiol. 1998; 1: 223-231Crossref PubMed Scopus (159) Google Scholar). Pilus subunits cross the inner membrane in an unfolded form via the Sec general secretory pathway (19Pugsley A. Microbiol. Rev. 1993; 57: 50-108Crossref PubMed Google Scholar, 20Van den Berg B. Clemons Jr., W.M. Collinson I. Modis Y. Hartmann E. Harrison S.C. Rapoport T.A. Nature. 2004; 427: 36-44Crossref PubMed Scopus (994) Google Scholar) and then must interact with the PapD chaperone in the periplasm. The chaperone acts by a mechanism termed donor strand complementation that couples subunit folding with the simultaneous capping of subunit interactive surfaces (21Barnhart M.M. Pinkner J.S. Soto G.E. Sauer F.G. Langermann S. Waksman G. Frieden C. Hultgren S.J. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7709-7714Crossref PubMed Scopus (150) Google Scholar, 22Choudhury D. Thompson A. Stojanoff V. Langermann S. Pinkner J. Hultgren S.J. Knight S.D. Science. 1999; 285: 1061-1066Crossref PubMed Scopus (524) Google Scholar, 23Sauer F.G. Fütterer K. Pinkner J.S. Dodson K.W. Hultgren S.J. Waksman G. Science. 1999; 285: 1058-1061Crossref PubMed Scopus (331) Google Scholar). Chaperone-subunit complexes must next target to the PapC usher in the OM for pilus biogenesis. The usher serves as a pilus assembly platform as well as a secretion channel (13Norgren M. Baga M. Tennent J.M. Normark S. Mol. Microbiol. 1987; 1: 169-178Crossref PubMed Scopus (81) Google Scholar, 24Thanassi D.G. Saulino E.T. Lombardo M-J. Roth R. Heuser J. Hultgren S.J. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 3146-3151Crossref PubMed Scopus (144) Google Scholar, 25Valent Q.A. Zaal J. de Graaf F.K. Oudega B. Mol. Microbiol. 1995; 16: 1243-1257Crossref PubMed Scopus (28) Google Scholar, 26Klemm P. Christiansen G. Mol. Gen. Genet. 1990; 220: 334-338Crossref PubMed Scopus (80) Google Scholar). Pilus assembly is thought to occur at the periplasmic face of the usher concomitant with secretion of the pilus fiber through the usher to the cell surface (18Thanassi D.G. Saulino E.T. Hultgren S.J. Curr. Opin. Microbiol. 1998; 1: 223-231Crossref PubMed Scopus (159) Google Scholar). Chaperone-subunit complexes initially target to an N-terminal region of the usher consisting of the first 124–139 residues of the mature N terminus (27Ng T.W. Akman L. Osisami M. Thanassi D.G. J. Bacteriol. 2004; 186: 5321-5331Crossref PubMed Scopus (70) Google Scholar, 28Nishiyama M. Vetsch M. Puorger C. Jelesarov I. Glockshuber R. J. Mol. Biol. 2003; 330: 513-525Crossref PubMed Scopus (66) Google Scholar). The complexes subsequently form stable interactions with the usher C terminus, which is required for subunit assembly into pili and secretion to the cell surface (29Saulino E.T. Thanassi D.G. Pinkner J.S. Hultgren S.J. EMBO J. 1998; 17: 2177-2185Crossref PubMed Scopus (133) Google Scholar, 30Thanassi D.G. Stathopoulos C. Dodson K.W. Geiger D. Hultgren S.J. J. Bacteriol. 2002; 184: 6260-6269Crossref PubMed Scopus (69) Google Scholar). Pilus assembly occurs by a mechanism termed donor strand exchange in which chaperone-subunit interactions are replaced by subunit-subunit interactions (21Barnhart M.M. Pinkner J.S. Soto G.E. Sauer F.G. Langermann S. Waksman G. Frieden C. Hultgren S.J. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7709-7714Crossref PubMed Scopus (150) Google Scholar, 31Sauer F.G. Pinkner J.S. Waksman G. Hultgren S.J. Cell. 2002; 111: 543-551Abstract Full Text Full Text PDF PubMed Scopus (217) Google Scholar, 32Zavialov A.V. Berglund J. Pudney A.F. Fooks L.J. Ibrahim T.M. MacIntyre S. Knight S.D. Cell. 2003; 113: 587-596Abstract Full Text Full Text PDF PubMed Scopus (223) Google Scholar). Donor strand exchange allows subunits to undergo a topological transition to a lower energy state, presumably providing the driving force for pilus biogenesis at the usher. Mature PapC contains 809 amino acids and has a molecular mass of 88.2 kDa. Computer modeling of the PapC topology predicted 24 TM β-strands, and analysis of PapC by circular dichroism confirmed a largely β-sheet secondary structure (30Thanassi D.G. Stathopoulos C. Dodson K.W. Geiger D. Hultgren S.J. J. Bacteriol. 2002; 184: 6260-6269Crossref PubMed Scopus (69) Google Scholar). The usher C terminus forms a distinct domain that is not required for proper folding of the β-barrel in the OM (30Thanassi D.G. Stathopoulos C. Dodson K.W. Geiger D. Hultgren S.J. J. Bacteriol. 2002; 184: 6260-6269Crossref PubMed Scopus (69) Google Scholar). PapC assembles into an oligomeric complex and forms channels 2–3 nm in diameter (24Thanassi D.G. Saulino E.T. Lombardo M-J. Roth R. Heuser J. Hultgren S.J. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 3146-3151Crossref PubMed Scopus (144) Google Scholar), large enough to allow secretion of folded pilus subunits. Detailed knowledge of the arrangement of the usher in the lipid bilayer is essential for understanding the molecular mechanisms of pilus assembly and secretion across the bacterial OM. Based on our topology mapping results and experiments showing that the usher C terminus is tightly associated with the OM, we propose that the usher folds with an N-terminal β-barrel domain comprised of 26 TM β-strands and that the C terminus is located within the barrel channel in a fashion analogous to the plug domains found in OM iron-siderophore uptake proteins (33Ferguson A.D. Chakraborty R. Smith B.S. Esser L. van der Helm D. Deisenhofer J. Science. 2002; 295: 1715-1719Crossref PubMed Scopus (304) Google Scholar, 34Ferguson A.D. Hofmann E. Coulton J.W. Diederichs K. Welte W. Science. 1998; 282: 2215-2220Crossref PubMed Scopus (668) Google Scholar, 35Locher K.P. Rees B. Koebnik R. Mitschler A. Moulinier L. Rosenbusch J.P. Moras D. Cell. 1998; 95: 771-778Abstract Full Text Full Text PDF PubMed Scopus (458) Google Scholar, 36Buchanan S.K. Smith B.S. Venkatramani L. Xia D. Esser L. Palnitkar M. Chakraborty R. van der Helm D. Deisenhofer J. Nat. Struct. Biol. 1999; 6: 56-63Crossref PubMed Scopus (492) Google Scholar). Strains and Plasmids—E. coli strain DH5α (37Grant S.G. Jessee J. Bloom F.R. Hanahan D. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 4645-4649Crossref PubMed Scopus (924) Google Scholar) was used for plasmid manipulations; strain AAEC185 (38Blomfield I.C. McClain M.S. Eisenstein B.I. Mol. Microbiol. 1991; 5: 1439-1445Crossref PubMed Scopus (130) Google Scholar), which lacks the chromosomal fim genes coding for type 1 pili, was used for hemagglutination assays (HAs); and strain SF100 (39Baneyx F. Georgiou G. J. Bacteriol. 1990; 172: 491-494Crossref PubMed Google Scholar), which lacks the OmpT protease, was used for analysis of the PapC constructs and for the topology mapping experiments. Strain XL-3 was used for analysis of the OxlT constructs (40Ye L. Jia Z. Jung T. Maloney P.C. J. Bacteriol. 2001; 183: 2490-2496Crossref PubMed Scopus (30) Google Scholar). Strains were grown in LB broth containing appropriate antibiotics at 37 °C with aeration. Plasmids used in this study are listed in Supplemental Table I. PapC expression was induced at an A600 of 0.6 for 1 h with 0.1% l-arabinose, and the ΔpapC pap operon from plasmid pMJ2 was induced with 0.05 mm isopropyl β-d-thiogalactoside. Plasmids expressing the single cysteine OxlT variants G49C or D104C were induced at an A600 of 0.6 for 1 h with 1 mm isopropyl β-d-thiogalactoside. Plasmids designated pNH (Supplemental Table I), encoding single cysteine substitution or insertion mutants of PapC, were derived from plasmid pMJ3, which encodes WT PapC with a C-terminal hexahistidine tag (His tag), using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA). Primers used for mutagenesis are listed in Supplemental Table I. All mutations were verified by sequencing. Random insertion of tobacco etch virus (TEV) protease sites into PapC was performed using the TnTAP transposon as described previously (41Ehrmann M. Bolek P. Mondigler M. Boyd D. Lange R. Proc. Natl. Acad. Sci. U. S. A. 1997; 64: 13111-13115Crossref Scopus (42) Google Scholar). Briefly pMM1 was transformed into DH5α/pMJ3, and a single Ampr, Tetr colony was selected for overnight growth and plasmid isolation. Plasmids were digested with NheI and transformed into DH5α, selecting for Ampr, Kanr, and Tets. The colonies were then screened for PhoA activity using 5-bromo-4-chloro-3-indolyl phosphate indicator plates. Blue colonies were selected and screened by PCR for insertion of TnTAP into papC using the primer UPSTRMPAPC (5′-GAAGCGCTGGATTACACCCTCAG-3′), which binds upstream of the papC open reading frame, and primer TNTAPPHOA (5′-GCAGTAATATCGCCCTGAGCAGC-3′), which binds to phoA within the TnTAP transposon. Plasmids containing TnTAP insertions in papC were isolated, and the insertion junctions were determined by sequencing using the TNTAPPHOA primer. Finally the phoA and neo reporter genes of TnTAP were removed by digesting the plasmids with NotI and religating to produce the plasmids designated pHG (Supplemental Table I), which code for PapC proteins containing in-frame insertions of the 24-residue sequence LTLIHKFENLYFQSAAAILVYKSQ (the TEV protease recognition site is underlined). HA—Analysis of pilus biogenesis by HA was performed as described previously (30Thanassi D.G. Stathopoulos C. Dodson K.W. Geiger D. Hultgren S.J. J. Bacteriol. 2002; 184: 6260-6269Crossref PubMed Scopus (69) Google Scholar) using AAEC185/pMJ2 (ΔpapC pap operon) complemented with pMON6235Δcat (vector), pMJ3 (WT PapC), or one of the PapC mutants. Heat-modifiable Mobility Shift Assay—Expression and folding of the various PapC mutants in the OM was analyzed using a heat-modifiable mobility shift assay as described previously (27Ng T.W. Akman L. Osisami M. Thanassi D.G. J. Bacteriol. 2004; 186: 5321-5331Crossref PubMed Scopus (70) Google Scholar). Digestion of TnTAP Insertion Mutants of PapC with TEV Protease— PapC mutants containing TnTAP insertions were tested for susceptibility to cleavage by TEV protease as described previously (41Ehrmann M. Bolek P. Mondigler M. Boyd D. Lange R. Proc. Natl. Acad. Sci. U. S. A. 1997; 64: 13111-13115Crossref Scopus (42) Google Scholar) using whole bacteria or isolated OM. OM was isolated as described previously (30Thanassi D.G. Stathopoulos C. Dodson K.W. Geiger D. Hultgren S.J. J. Bacteriol. 2002; 184: 6260-6269Crossref PubMed Scopus (69) Google Scholar). Samples were subjected to SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Osmonic Inc., Gloucester, MA). Cleavage of PapC was monitored by blotting with anti-PapC or anti-His tag (Covance, Richmond, CA) antibodies followed by an alkaline phosphatase-conjugated secondary antibody (Sigma). The blots were developed with 5-bromo-4-chloro-3-indolyl phosphate-nitro blue tetrazolium substrate (KPL, Gaithersburg, MD). Site-directed Fluorescence Labeling—This protocol was adapted from the method used by Ye and coworkers (40Ye L. Jia Z. Jung T. Maloney P.C. J. Bacteriol. 2001; 183: 2490-2496Crossref PubMed Scopus (30) Google Scholar) to map the inner membrane protein OxlT. A 20-ml culture of SF100 expressing WT PapC or one of the PapC mutants was harvested, resuspended in 2.6 ml of buffer A (100 mm potassium sulfate, 50 mm potassium phosphate, pH 8.0), and divided into two equal aliquots. Freshly prepared 4-acetamido-4′-male-imidylstilbene-2–2′-disulfonic acid (AMS, Molecular Probes, Eugene, OR) was added to one aliquot to 100 μm final concentration, and the aliquots were incubated at room temperature (RT, 25 °C) for 7 min. Both aliquots were washed three times with buffer A, and then Oregon green 488 maleimide (OGM, Molecular Probes) was added to 40 μm final concentration to both aliquots. The samples were incubated in the dark at RT for 15 min, the reaction was quenched by addition of 6 mm β-mercaptoethanol, and the aliquots were washed three times with buffer A. The bacteria were then resuspended in 1 ml of 20 mm Tris-HCl (pH 8.0) containing protease inhibitors for isolation of OM as described previously (27Ng T.W. Akman L. Osisami M. Thanassi D.G. J. Bacteriol. 2004; 186: 5321-5331Crossref PubMed Scopus (70) Google Scholar). The OM pellet was resuspended in 20 mm Tris (pH 8.0), 0.3 m NaCl, 0.5% dodecylmaltoside (DDM, Anatrace, Maumee, OH) and solubilized overnight by rocking at 4 °C. After spinning out insoluble material (16,100 × g, 30 min, 4 °C), imidazole was added to 20 mm final concentration, and the mixture was rocked for 30 min at RT with 50 μl of 50% nickel-nitrilotriacetic acid resin (Qiagen, Chatsworth, CA) to bind the His-tagged PapC. The nickel-nitrilotriacetic acid resin was then washed three times with 20 mm Tris (pH 8.0), 0.3 m NaCl, 20 mm imidazole, 0.1% DDM. The supernatant was removed, 25 μl of 2× SDS-PAGE sample buffer was added to the resin, the sample was incubated at 95 °C for 10 min, and 18 μl was loaded on a 10% gel. After electrophoresis, the fluorescence profile was obtained by scanning with a STORM fluorescence imaging system (Amersham Biosciences) using the blue fluorescent chemifluorescence mode (excitation wavelength of 450 ± 30 nm). The same gel was then stained with Coomassie Brilliant Blue R-250, scanned using a GS-710 densitometer (Bio-Rad), and analyzed with Quantity One software (Bio-Rad). The fluorescence value for each PapC band was normalized to its protein level determined from Coomassie staining. Then the ratio of the normalized OGM signal without AMS pretreatment to the normalized OGM signal with AMS pretreatment was calculated for each PapC to give the -fold reduction in OGM labeling due to pretreatment with AMS. Control PapC constructs containing cysteines at known surface and periplasmic locations were included in all assays; this was important to monitor the quality of the AMS reagent. All PapC mutants were analyzed at least twice. As an alternative to AMS, Alexa Fluor 594 C5 maleimide (Molecular Probes) was used in the above protocol at a final concentration of 100 μm and an incubation time of 15 min. The specificity of AMS as a blocking agent for surface-exposed residues was tested using the OxlT mutants G49C and D104C, which contain periplasmic cysteines (40Ye L. Jia Z. Jung T. Maloney P.C. J. Bacteriol. 2001; 183: 2490-2496Crossref PubMed Scopus (30) Google Scholar). The OxlT mutants were assayed using AMS as a blocking agent according to our protocol described above or using methanethiosulfonate ethyltrimethylammonium (Biotium Inc., Hayward, CA) as a blocking agent (2 mm final concentration for 10 min) as described previously (40Ye L. Jia Z. Jung T. Maloney P.C. J. Bacteriol. 2001; 183: 2490-2496Crossref PubMed Scopus (30) Google Scholar). PapC mutants that did not label with OGM in intact bacteria were subsequently analyzed by labeling with OGM after solubilization with DDM or denaturation with SDS (42Fu D. Maloney P.C. J. Biol. Chem. 1998; 273: 17962-17967Abstract Full Text Full Text PDF PubMed Scopus (35) Google Scholar). For labeling of solubilized PapC, OM was extracted with DDM as described above, but before the addition of imidazole samples were incubated for 15 min at RT with 40 μm OGM. The reaction was quenched with 6 mm β-mercaptoethanol, and protein was purified and analyzed as described above. For labeling of denatured PapC, purified protein was isolated as described above but boiled in sample buffer without β-mercaptoethanol. OGM was added to 1 mm, and the samples were incubated for 15 min at RT. The samples were then subjected to SDS-PAGE and analyzed as described above. Association of the PapC C Terminus with the OM—OM was isolated from SF100/pMJ3 as described previously (30Thanassi D.G. Stathopoulos C. Dodson K.W. Geiger D. Hultgren S.J. J. Bacteriol. 2002; 184: 6260-6269Crossref PubMed Scopus (69) Google Scholar). A final concentration of 50 μg/ml l-1-tosylamido-2-phenylethyl chloromethyl ketone-treated trypsin (Sigma) was added to 0.2 mg OM (as determined by the bicinchoninic acid protein assay, Pierce). After incubation on ice for 2 min, digestion was terminated by addition of 1 ml of 20 mm HEPES (pH 7.5), 0.3 m NaCl, 1 mm phenylmethylsulfonyl fluoride. The OM was pelleted by centrifugation (80,000 × g, 1 h, 4 °C) and resuspended in 1 ml of 20 mm HEPES (pH 7.5), 0.3 m NaCl, 1 mm phenylmethylsulfonyl fluoride. Samples were treated on ice for 1 h with NaOH (0.01, 0.05, or 0.1 m), urea (4 or 6 m), 1 m NaCl, or 0.2 m Na2CO3 or were sonicated (4 min, 30 s on, and 30 s off) in a water bath sonicator (Misonix). The membrane and soluble fractions were then separated by centrifugation (80,000 × g,1h, 4 °C). The supernatant fractions were precipitated by addition of trichloroacetic acid. Both the membrane and trichloroacetic acid pellets were treated with SDS-PAGE sample buffer at 95 °C for 10 min, resolved by SDS-PAGE, and blotted with an anti-His tag antibody as described above. Identification of Two Surface Loops of PapC by TEV Protease Cleavage—As a first approach to map the topology of the usher, we generated random insertions of the TEV protease recognition sequence in PapC using the TnTAP transposon (41Ehrmann M. Bolek P. Mondigler M. Boyd D. Lange R. Proc. Natl. Acad. Sci. U. S. A. 1997; 64: 13111-13115Crossref Scopus (42) Google Scholar). TnTAP was designed specifically for analysis of OM proteins and results in a final in-frame insertion of a 24-residue sequence carrying the TEV protease cleavage site. We isolated 29 unique TnTAP insertions in PapC that were stably expressed in the OM (Fig. 1). Approximately one-third of the insertions were isolated more than once with a pronounced hotspot for insertion following residue Val-76 (V76TEVins) in a large periplasmic loop of PapC (Fig. 1). This loop also contained the highest concentration of TnTAP insertions. In other regions of PapC, notably the C terminus, we were unable to isolate insertions. The insertion sites in PapC define regions flexible enough to accommodate the 24-residue TnTAP sequence. The non-random distribution of insertions was thus as least partly attributable to the screening procedure, which selected for stable, in-frame insertions in PapC (41Ehrmann M. Bolek P. Mondigler M. Boyd D. Lange R. Proc. Natl. Acad. Sci. U. S. A. 1997; 64: 13111-13115Crossref Scopus (42) Google Scholar). To confirm that the PapC mutants containing TnTAP insertions were folded properly in the OM, we took advantage of the fact that PapC, in common with other OM β-barrel proteins, is resistant to denaturation by SDS unless incubated at high temperatures and thus exhibits a characteristic heat-modifiable mobility on SDS-PAGE (24Thanassi D.G. Saulino E.T. Lombardo M-J. Roth R. Heuser J. Hultgren S.J. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 3146-3151Crossref PubMed Scopus (144) Google Scholar). Each of the TnTAP insertion mutants exhibited heat-modifiable mobility, indicating proper folding of the PapC β-barrel (data not shown). We next assessed functionality of the TnTAP insertion mutants by testing their ability to complement a ΔpapC pap operon for assembly of adhesive pili by HA. The majority of the insertion mutants (22 of 29, Fig. 1) were defective for pilus biogenesis, having HA titers of 0–8 compared with a WT HA titer of 64 (Supplemental Table II). The functional defects found with the TnTAP insertion mutants are not surprising given the large size of the inserted sequence. Although most of the TnTAP insertion mutants were not functional for pilus biogenesis, we proceeded with topology mapping studies as all were able to adopt a correct β-barrel fold in the OM. TEV protease was added to whole bacteria or to isolated OM, and cleavage of the usher was monitored by immunoblot analysis using anti-His tag or anti-PapC antibody. Three of the TnTAP insertion mutants (A124TEVins, R138TEVins, and S194TEVins) were clearly cleaved in whole bacteria, indicating a surface location (Fig. 2). These insertions occurred in two N-terminal loops predicted to be surface-exposed by a prior modeling study (30Thanassi D.G. Stathopoulos C. Dodson K.W. Geiger D. Hultgren S.J. J. Bacteriol. 2002; 184: 6260-6269Crossref PubMed Scopus (69) Google Scholar). These two loops are also the largest surface loops present in the β-barrel region of PapC (Fig. 1, see below). Three additional TnTAP insertion mutants (G30TEVins, A69TEVins, and D74TEVins) were cleaved by TEV protease in isolated OM but not in whole bacteria, suggesting a periplasmic location (data not shown). However, we could not discriminate against the possibility that these insertions instead were located at the surface but only became exposed upon isolation of the OM. This is a major limitation of topology mapping protocols that require membrane disruption to reach internal sites. None of the remaining TnTAP insertions in PapC was cleaved by TEV protease in either whole bacteria or OM (data not shown). This was surprising given that the insertion mutants were stably expressed, and the large size of the inserted sequence should favor exposure of the cleavage site. The TEV protease may not have been able to recognize the cleavage sites due to local conformational effects, or the sites may have been sterically blocked by protein or membrane structure (7Newton S.M. Klebba P.E. Michel V. Hofnung M. Charbit A. J. Bacteriol. 1996; 178: 3447-3456Crossref PubMed Scopus (38) Google Scholar, 9Guédin S. Willery E. Tommassen J. Fort E. Drobecq H. Locht C. Jacob-Dubuisson F. J. Biol. Chem. 2000; 275: 30202-30210Abstract Full Text Full Text PDF PubMed Scopus (84) Google Scholar, 10Merck
Referência(s)