Demonstration by Mass Spectrometry that Purified NativeTreponema pallidum Rare Outer Membrane Protein 1
(Tromp1) Has a Cleaved Signal PeptideDAVID R. BLANCO,1,2* JULIAN P. WHITELEGGE,3,4 JAMES N. MILLER,1
AND MICHAEL A. LOVETT1,2
Department of Microbiology, Immunology, and Molecular Genetics1 and Department of Medicine,2
School of Medicine, The Pasarow Mass Spectrometry Laboratory,3 and Department of Chemistryand Biochemistry,4 University of California at Los Angeles, Los Angeles, California 90095
Received 11 March 1999/Accepted 10 June 1999
Purified native Tromp1 was subjected to mass spectrometric analysis in order to determine conclusivelywhether this protein possesses a cleaved or uncleaved signal peptide. The molecular masses of Tromp1, threeTreponema pallidum lipoproteins, and a bovine serum albumin (BSA) control were determined by matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry. The molecular masses ofall of the T. pallidum lipoproteins and BSA were within 0.7% of their respective calculated masses. Themolecular mass of Tromp1 was 31,510 Da, which is consistent with a signal-less form of Tromp1, given acalculated mass of unprocessed Tromp1 of 33,571 Da, a difference of 2,061 Da (a 6.5% difference). Purifiednative Tromp1 was also subjected to MALDI-TOF analysis in comparison to recombinant Tromp1 followingcyanogen bromide cleavage, which further confirmed the identity of Tromp1 and showed that native Tromp1was not degraded at the carboxy terminus. These studies confirm that Tromp1 is processed and does notcontain an uncleaved signal peptide as previously reported.
Treponema pallidum subsp. pallidum, the etiologic agent ofvenereal syphilis, has for over a decade been known to possessa unique outer membrane containing an extremely low densityof membrane-spanning surface-exposed protein (17, 20). It isbelieved that these T. pallidum rare outer membrane proteins,termed TROMPs (6), are the only surface-exposed antigens onthis organism and therefore represent the key surface targetsfor protective host immune mechanisms which develop duringsyphilitic infection.
In our previous attempts to identify potential TROMP can-didates, two proteins of 31 and 28 kDa were found to bemarkedly enriched in outer membranes isolated from T. palli-dum (5). The 31-kDa protein, termed Tromp1, was found tohave properties consistent with those of an outer membraneporin protein, including amphiphilicity following phase sepa-ration in the detergent Triton X-114 and electrical conductivitywhen analyzed in planar lipid bilayers (3). It was also deter-mined that recombinant Tromp1, when expressed, exported,and targeted to Escherichia coli outer membranes, also exhib-ited porin activity similar to that measured for native Tromp1(4). Tromp1 has also been found to have 26 to 28% sequenceidentity to adhesin proteins found in the streptococcal family(12), suggesting a potential role as a virulence determinant.
While our studies have shown that Tromp1 is a porin pro-tein, recent studies by Hardham et al. (10) have found thatTromp1 is also part of an operon which possesses similaritiesto ABC transporter systems and that Tromp1, also called TroAin these studies, has 28% sequence identity to periplasmicbinding proteins of these ABC transporter operons. This ap-parent disparity between the demonstrated outer membrane
location and porin activity of Tromp1 and the suggestion thatTroA is a periplasmic binding protein from homology compari-sons is an area of research which is currently being investigated. Ithas also been recently reported by Akins et al. (2) that Tromp1possesses an uncleaved signal peptide, which these investigatorsconclude anchors Tromp1 to the inner membrane and accountsfor its demonstrated hydrophobicity. Because Tromp1 possessesan N terminus blocked to Edman sequencing, the conclusion thatTromp1 is uncleaved was based upon a sodium dodecyl sulfate(SDS)-polyacrylamide gel electrophoresis (PAGE)-based sizecomparison of native Tromp1 and that generated from an in vitrotranslation product. Thus, the findings of these studies have re-sulted in one current view that Tromp1 is not an outer membraneprotein but rather a periplasmic binding protein anchored to theinner membrane by an uncleaved signal peptide.
In order to begin to address the controversy surroundingTromp1, we have isolated and purified the hydrophobic formof native Tromp1 for mass spectrometric analysis to determineconclusively whether this protein possesses a cleaved or un-cleaved signal peptide. The findings reported here demon-strate conclusively that Tromp1 has a cleaved signal peptideand is therefore not anchored to the inner membrane by anuncleaved signal peptide as previously reported (2). Also im-plicit from these findings is that the hydrophobicity of Tromp1is not due to an uncleaved signal peptide, as reported previ-ously, but is rather an inherent property of this protein, whichwe believe is consistent with its outer membrane location.
Isolation and purity of native Tromp1 and three other T. pal-lidum hydrophobic proteins. Native hydrophobic forms ofTromp1, the 47-kDa lipoprotein, the MglB homolog lipoprotein(41 kDa), and the TmpC lipoprotein (35 kDa) were isolated fromapproximately 2 3 1011 T. pallidum cells as follows. T. pallidumsubsp. pallidum (Nichols strain) was extracted in phosphate-buff-ered saline (PBS; pH 7.2) from 20 intratesticularly infected rab-bits as previously described (13). Approximately 800 ml of trepo-nemal extract was centrifuged twice at 400 3 g for 10 min each
* Corresponding author. Mailing address: Department of Microbi-ology, Immunology, and Molecular Genetics, School of Medicine, Uni-versity of California at Los Angeles, Los Angeles, CA 90095. Phone:(310) 206-6510. Fax: (310) 206-3865. E-mail: [email protected].
time in order to pellet gross tissue debris and then at 20,000 3 gfor 30 min in order to pellet the treponemes. The treponemalpellet was washed in 200 ml of PBS and then recentrifuged at20,000 3 g for 30 min. The final treponemal pellet was resus-pended in 26 ml of ice-cold PBS to which was added 4 ml ofice-cold 10% hydrogenated Triton X-114 (Calbiochem, San Di-ego, Calif.) to yield a final detergent concentration of 2%. Thesuspension was next incubated on a rocker at 4°C for 2 h in orderto solubilize the outer membrane. After this incubation, the sus-pension was centrifuged at 20,000 3 g in order to remove T.pallidum protoplasmic cylinders and the supernatant was re-moved and warmed to 37°C for 5 min, which resulted in cloudformation of the detergent. The suspension was then centrifugedat 3,000 3 g in order to yield separated hydrophobic (bottom) andaqueous (top) phases. The hydrophobic phase was recovered(approximately 2 ml), extracted twice with 40 ml of warmed PBS,and then centrifuged as described above. The final extracted hy-drophobic phase (approximately 2 ml) was then combined with 30ml of ice-cold acetone, incubated for 2 h at 4°C, and then centri-fuged at 10,000 3 g for 30 min to recover precipitated protein.The protein pellet was then subjected to two-dimensional SDS-PAGE as previously described (5). After electrophoresis, proteinsin the gel were transferred to a polyvinylidene difluoride (PVDF)membrane (Millipore, Bedford, Mass.) as previously described(18), stained with 0.2% amido black in water for 20 min, anddestained by using distilled water. Protein spots corresponding toTromp1, the 47-kDa lipoprotein, the MglB lipoprotein, and theTmpC lipoprotein were identified based upon their respectivemolecular weights and isoelectric points as previously described(15).
In order to elute the proteins from the membrane, the sec-tion of the PVDF membrane corresponding to each of theseproteins was cut out and placed in 250 ml of 2% SDS–1%hydrogenated Triton X-100–50 mM Tris, pH 9.0 (1). The sus-pensions were incubated on a shaker for 16 h at room temper-ature and then centrifuged at 13,000 3 g for 10 min. Samplesfrom each of the supernatants were analyzed by SDS-PAGE(12% polyacrylamide) and identified by silver staining (Bio-Rad). The total amount of protein recovered from each of theeluted samples was estimated to be 3 to 5 mg.
As shown in Fig. 1, 1/25 (approximately 200 ng) of eachsample analyzed showed a single band corresponding to amolecular mass consistent with that of Tromp1 (31 kDa), the47-kDa lipoprotein, the 41-kDa MglB homolog lipoprotein,and the 35-kDa TmpC lipoprotein. In addition, Tromp1 wasfurther analyzed by two-dimensional immunoblotting usingspecific anti-Tromp1 serum and enhanced chemiluminescenceas previously described (4). As shown in Fig. 2, the amidoblack-stained immunoblot showed a doublet at 31 kDa (Fig.2A) corresponding to the molecular mass and pI (6.6) reportedpreviously for Tromp1 (3). The stained 31-kDa doublet re-acted specifically with anti-Tromp1 serum (Fig. 2B), confirm-ing that the isolated protein was Tromp1.
MALDI mass spectrometry of purified native Tromp1. Toobtain an accurate measurement of the molecular mass ofTromp1, its mass spectrum was recorded by using matrix-as-sisted laser desorption ionization (MALDI) time-of-flight (TOF)mass spectrometry. MALDI-TOF mass spectrometry was per-formed by using a Voyager RP machine (Perceptive Biosystems,Framingham, Mass.) operating in linear mode. The three other T.pallidum membrane lipoproteins and bovine serum albumin(BSA) were also analyzed as controls. For sample preparation, 0.3ml of chloroform-methanol-precipitated protein (21) (1 to 10pmol) dissolved in 60% HCOOH (2 to 4 ml) was mixed with 0.5ml of a 20-mg/ml matrix solution (2,5-dihydroxybenzoic acid; Al-
drich, Milwaukee, Wis.) in 50% HCOOH–50% isopropanol anddried immediately on the MALDI plate.
The measured masses of the lipoproteins and BSA, as well astheir calculated masses based upon their gene sequences andposttranslational modifications, are presented in Table 1. Rea-sonable agreement between the measured and calculated masseswas obtained for the three control lipoproteins and BSA, withdiscrepancies of 0.7% or less. Mass accuracy of 0.1% is expectedfor the MALDI-TOF technique using internal calibration, as weused for these experiments, and thus it was concluded that theproteins had been modestly modified during the isolation process.Methionine oxidation and acrylamide adducts of cysteine areexpected on proteins exposed to SDS-PAGE. Nevertheless, themeasured masses support the structural assignment of N-palmi-toyl, S-[2,3-bis(palmitoyloxy)-propyl] to the N-terminal cysteine ofthese lipoproteins and support translational termination at thefirst stop codon in the case of the 47-kDa lipoprotein.
As shown in Fig. 3, the MALDI-TOF mass spectrum ofTromp1 showed strong ions for the singly charged protein(M1H1) and the doubly charged ion (M12H1); the triplycharged ion was also visible (M13H1). The measured mass ofTromp1 was found to be, whether in the presence or absenceof an internal BSA standard, 31,510 Da (Fig. 3), which is indisagreement with the calculated mass of 33,571 Da for anunprocessed translation product (6.5% error). This mass de-termination for Tromp1 was consistent and reproducible fol-lowing several separate experiments. A much closer agreementbetween the measured and calculated masses was obtained fora cleaved Tromp1 translation product (Table 1). A bettermatch between masses is achieved if the THA leader peptidaseI cleavage site in the Tromp1 signal peptide is invoked, al-though the calculated mass (31,182 Da) is lower than themeasured mass (1% error). Arbitrary cleavage N terminal toThr20 of the THA motif improves the match considerably(0.1% error). These results demonstrate that the nativeTromp1 protein is cleaved to size from its original translationproduct. However, the level of accuracy in this analysis pre-cludes assignment of the true N terminus and, formally, cleav-age of the C terminus must also be considered.
MALDI mass spectrometric comparison of native and recom-binant Tromp1. Native Tromp1 and a signal-less form of recom-
FIG. 1. SDS-PAGE, followed by silver staining, of isolated native Tromp1,the 47-kDa lipoprotein (47-kDa LP), the 41-kDa Mglb homolog lipoprotein(Mglb), and the 35-kDa TmpC lipoprotein (TmpC). Approximately 200 ng ofeach isolated protein was combined with sample buffer containing 2% SDS and5% 2-mercaptoethanol and boiled prior to electrophoresis. The values on the leftindicate the positions of molecular weight standards (103).
binant Tromp1 were treated with CNBr, and the fragments wereprobed by MALDI-TOF mass spectrometry in order to comparethe observed peptide maps with theoretical ones based upon theknown gene sequence and potential cleavage sites.
A signal-less form of recombinant Tromp1, corresponding tothe signal peptide cleavage site of alanine-phenylalanine-gly-cine (AFG), was generated as follows. An N-terminal primer(59-CGCCATATGAGCAAGGATGCCGCAGCAGAC-39; theunderlined region is the tromp1 gene sequence) correspondingto signal peptide cleavage after AFG was generated containingan NdeI restriction endonuclease site at the 59 end (GibcoBRL, Gaithersburg, Md.). This construct results in a singlemethionine residue placed ahead of the site of signal peptidecleavage for the purpose of translation. A C-terminal primerconsisting of 59-CGCGGATCCCTAGCGAGCCAACGCAGCAA-39 and corresponding to the end of the tromp1 gene wasgenerated containing a BamHI restriction endonuclease site atthe 59 end. A PCR was performed with these primers as de-scribed previously (4). The tromp1 PCR product was ligatedinto pET17b (Novagen, Inc.), which had previously been di-gested with NdeI and BamHI. The resulting construct wastransformed into E. coli BL21 DE3/pLysE (Novagen, Inc.)using cells made competent by CaCl2. Expression and fast-performance liquid chromatography purification of recombi-nant Tromp1 were performed as described previously (4).
For CNBr treatment of native and recombinant Tromp1,chloroform-methanol-precipitated protein pellets were dis-solved in 35 ml of a 1-g/ml saturated solution of CNBr (Sigma,St. Louis, Mo.) in 90% formic acid (Fisher, Fair Lawn, N.J.) towhich was added 15 ml of water. A final CNBr/Met molar ratioof 500 was exceeded in all digestions. CNBr digestion wascarried out for 4 h in the dark at room temperature. Thereaction mixture was then dried (SpeedVac), redissolved in250 ml of 0.1% trifluoroacetic acid (Pierce, Rockford, Ill.), anddried again. For MALDI-TOF mass spectrometry, 0.3 ml of theCNBr-digested samples (1 to 2 pmol) dissolved in 60%HCOOH (2 to 4 ml) was mixed with 0.5 ml of a 10-mg/ml matrixsolution (a-hydroxycinnamic acid; Aldrich) in 0.1% trifluoro-acetic acid–70% CH3CN and dried immediately on theMALDI plate. Bovine insulin was used as a standard.
Table 2 summarizes the results of MALDI-TOF mass spec-trometric analysis of CNBr-treated recombinant and nativeTromp1. Central to the interpretation of the mapping experi-ments is the appearance of peptides derived from the N and Ctermini of the protein. The C-terminal peptide (peptide 8) wasobserved in both cases, providing conclusive evidence that anyprocessing was not at the C terminus. Internal peptides (pep-tides 3, 4, and 6) were well represented in both proteins,proving that the protein analyzed was Tromp1. In addition, theset of the native Tromp1 peptides used to search the NationalCenter for Biotechnology Information protein databases wasfound to match only Tromp1 and TroA, confirming that thepeptides analyzed were from Tromp1. In native Tromp1, N-terminal data set peptides with masses agreeing well with thoseof peptide 2 and hybrid peptide 2/3 were observed, whereasonly peptide 2 was observed in the recombinant protein. In thecase of native Tromp1, the masses of the peptides that wereobserved support a model with cleavage between residues 19and 20 generating a novel set of CNBr fragments that were notobserved in the map of the uncleaved protein.
The results presented here, in contrast to the previous reportby Akins et al. (2), conclusively demonstrate that Tromp1 hasa processed and cleaved signal peptide. Mass spectrometryanalysis of native Tromp1 resulted in a molecular mass of31,510 Da, consistent with a processed form of this proteingiven a calculated mass of unprocessed Tromp1 of 33,571 Da,a difference of 2,061 Da, which is the average size of a 19-residue signal peptide (19). This result was also reproduciblewhether an internal BSA protein control was included in orexcluded from the sample containing Tromp1. Further, massspectrometry peptide analysis of CNBr-treated Tromp1 con-
FIG. 2. Two-dimensional immunoblot analysis of isolated native Tromp1.Approximately 500 ng of isolated native Tromp1 was separated in the firstdimension by denaturing isoelectric focusing (IEF) and then in the seconddimension by SDS-PAGE. Proteins were then transferred to a PVDF membrane.(A) Amido black-stained PVDF membrane showing the detection of onlyTromp1. (B) Membrane in panel A probed with specific anti-Tromp1 serum,confirming the detection of Tromp1. The values on the left indicate the positionsof molecular weight standards (103).
TABLE 1. Molecular masses of T. pallidum proteins measured byMALDI-TOF mass spectrometry
ProteinMolecular mass (Da)
% DifferenceCalculated Measureda
Uncleaved Tromp1 33,571 31,510 6.5Tromp1 cleaved between
firmed that the sample tested was Tromp1 and that it was notdegraded from the carboxy terminus. In addition, three T.pallidum lipoproteins, isolated the same way as Tromp1 fromthe hydrophobic phase of a Triton X-114 detergent extract,were also analyzed in order to confirm the validity of themolecular mass results obtained for T. pallidum proteins iso-lated by this method. The molecular masses obtained for thethree lipoproteins were all within 0.7% of their calculatedvalues given the addition for each of these lipoproteins of acovalent association of glycerol and three molecules of the fattyacid palmitate. Therefore, mass spectrometry analysis, which pro-vides the most reliable assessment of molecular mass, has proventhat native Tromp1 possesses a cleaved signal peptide.
Of the three potential leader peptidase I cleavage sitespresent at the carboxy terminus of the Tromp1 signal peptide,which include (from the N to the C terminus) threonine-histi-dine-alanine (THA), alanine-phenylalanine-glycine (AFG),and alanine-alanine-alanine (AAA), theoretical cleavage fol-lowing THA results in a calculated molecular mass of themature protein having the closest agreement with the massspectrometry result of native Tromp1 (1% error). THA is alsothe cleavage site predicted for the Tromp1 signal peptide bythe SignalP analysis program (14). However, given that a massaccuracy of 0.1% is expected for the MALDI-TOF technique,prediction of the cleavage site based upon this level of accuracywould place the cleavage site N terminal to the threonine of
FIG. 3. Molecular mass spectrum of native Tromp1. The mass spectrum was recorded with a Perceptive Biosystems Voyager RP with laser intensity sufficient toachieve efficient ionization. The singly and doubly charged ions of BSA were used to calibrate the mass spectrometer, and similar results were obtained whether aninternal calibration was used or not. dhb, 2,5-dihydroxybenzoic acid matrix solution.
TABLE 2. Peptide mapping by MALDI-TOF mass spectrometry of native and recombinant Tromp1 after CNBr cleavage
Peptide no.
Calculated mass(es) (Da) of CNBr-cleavedTromp1 peptide(s)a
Measured mass (Da) of CNBr-cleavedTromp1 peptide(s) (% difference)
a The Peptide Mass program (9a) was used to calculate average peptide masses (M1H1) based upon natural isotopic abundance. Limited partial cleavage isconsidered, and internal methionine sulfoxide was used to calculate the mass rather than methionine delivered by Peptide Mass (thus, hybrids are 16 Da larger thanthe sizes reported by this program for peptides with C-terminal homoserine lactone). Peptides are numbered from the N to the C terminus.
b Recombinant Tromp1 was constructed without a signal peptide and expressed with a start Met following the putative cleavage motif of alanine-phenylalanine-glycine (AFG).
c Calculated masses of the N-terminal peptide of signal-less Tromp1 cleaved at AFG (1,727 and 3,718 Da), THA (1,931 and 3,923 Da), and N terminal to Thr20 (2,241and 4,232 Da).
d Percent difference was calculated against peptide 2 cleaved at AFG plus the N-terminal Met (1,858 Da).e ND, not detected.
the THA motif. It is unlikely that this is the actual N terminusof the mature protein because this would require the upstreamsequence of threonine-glycine-phenylalanine to be recognizedas the cleavage motif, which is not a classic leader peptidase Icleavage recognition site (19).
Purified native Tromp1 and a signal-less form of recombi-nant Tromp1 were also analyzed in this study by mass spec-trometry following CNBr cleavage. The intact-protein massdata combined with the CNBr mapping data provide over-whelming evidence that the observed protein is indeed Tromp1and that again it is cleaved to a significantly shorter length thanthat predicted from the uncleaved protein. CNBr mappingdata can be accommodated with cleavage between amino acids19 and 20 or thereabout, bearing in mind the N-terminal mod-ification implied by blockage to Edman sequencing. Definitivestructural assignment of the N terminus of Tromp1 requiresfurther biochemical analyses. Central to these analyses will beelectrospray-ionization mass spectrometry, which has recentlybeen applied to membrane proteins (23) and has, indeed, beenused to measure the molecular weight of recombinant Tromp1(22). Resolution by electrospray-ionization mass spectrometryshould solve the heterogeneity of Tromp1 apparent in gels.
In the previous study by Akins et al. (2), their conclusion thatTromp1 possesses an uncleaved signal peptide was used toexplain the hydrophobicity of Tromp1 and suggest its anchor-ing to the inner membrane. In contrast, our latest studies haveindicated that the hydrophobicity of Tromp1 is due to reasonsother than an uncleaved signal peptide. Further, our previousstudies showed that treatment of T. pallidum with low concen-trations of the nonionic detergent Triton X-114, which com-pletely solubilizes the T. pallidum outer membrane withoutsolubilizing the inner membrane (7, 16), resulted in the com-plete release of Tromp1 with no residual detection of Tromp1in the inner membrane protoplasmic cylinder complex (3).Such findings are consistent with the idea that this hydrophobicprotein has an outer membrane origin rather than being an-chored to the inner membrane.
It was also reported by Akins et al. (2), who used a purifiedrecombinant, that Tromp1 did not show any porin activity bythe liposome swelling assay (8). Similarly, we have also foundthat with the exception of recombinant Tromp1 targeted to E.coli outer membranes (4), no soluble recombinant form ofTromp1 tested has shown porin activity when planar lipid bi-layers were used. It should be emphasized, however, that this isin direct contrast to purified native Tromp1, which we havefound to have consistent and reproducible porin activity fol-lowing isolation by isoelectric focusing (3) and more recentlyby fast-performance liquid chromatography (9). We believe thatthe difference in this demonstrable porin activity between thenative and recombinant proteins may be due to conformation.
In summary, the findings presented in this study conclusivelyshow that native Tromp1 does, indeed, possess a processedsignal peptide and is therefore an exported protein. Thesefindings, therefore, indicate that Tromp1 is not anchored to theinner membrane and support the possibility that Tromp1 is abona fide outer membrane protein of T. pallidum.
We thank Xiao-Yang Wu for his excellent technical assistance. Wealso thank Kym Faull for his support and encouragement.
This work was supported by U.S. Public Health Service grants AI-21352 and AI-12601 to M. A. Lovett and AI-37312 to J. N. Miller.Funds from the UCLA Jonsson Comprehensive Cancer Center (P30-A16042) were used toward the purchase of the MALDI-TOF massspectrometer used.
REFERENCES
1. Aebersold, R. 1993. Internal amino acid sequence analysis of proteins afterin situ protease digestion on nitrocellulose, p. 100–101. In P. T. Matsudaira(ed.), A practical guide to protein and peptide purification for microsequenc-ing, 2nd ed. Academic Press, Inc., New York, N.Y.
2. Akins, D. R., E. Robinson, D. Shevchenko, C. Elkins, D. L. Cox, and J. D.Radolf. 1997. Tromp1, a putative rare outer membrane protein, is anchoredby an uncleaved signal sequence to the Treponema pallidum cytoplasmicmembrane. J. Bacteriol. 179:5076–5086.
3. Blanco, D. R., C. I. Champion, M. M. Exner, H. Erdjument-Bromage,R. E. W. Hancock, P. Tempst, J. N. Miller, and M. A. Lovett. 1995. Porin activityand sequence analysis of a 31-kilodalton Treponema pallidum subsp. pallidumrare outer membrane protein (Tromp1). J. Bacteriol. 177:3556–3562.
4. Blanco, D. R., C. I. Champion, M. M. Exner, E. S. Shang, J. T. Skare,R. E. W. Hancock, J. N. Miller, and M. A. Lovett. 1996. RecombinantTreponema pallidum rare outer membrane protein 1 (Tromp1) expressed inEscherichia coli has porin activity and surface antigenic exposure. J. Bacte-riol. 178:6685–6692.
5. Blanco, D. R., K. Reimann, J. Skare, C. I. Champion, D. Foley, M. M. Exner,R. E. W. Hancock, J. N. Miller, and M. A. Lovett. 1994. Isolation of the outermembranes from Treponema pallidum and Treponema vincentii. J. Bacteriol.176:6088–6099.
6. Blanco, D. R., E. M. Walker, D. A. Haake, C. I. Champion, J. N. Miller, andM. A. Lovett. 1990. Complement activation limits the rate of in vitro trepo-nemicidal activity and correlates with antibody-mediated aggregation ofTreponema pallidum rare outer membrane protein (TROMP). J. Immunol.144:1914–1921.
7. Cunningham, T. M., E. M. Walker, J. N. Miller, and M. A. Lovett. 1988.Selective release of the Treponema pallidum outer membrane and associatedpolypeptides with Triton X-114. J. Bacteriol. 170:5789–5796.
8. Davidson, A. L., and H. Nikaido. 1991. Purification and characterization ofthe membrane-associated components of the maltose transport system fromEscherichia coli. J. Biol. Chem. 266:8946–8951.
9. Exner, M. M., D. R. Blanco, J. N. Miller, and M. A. Lovett. Unpublisheddata.
9a.ExPASy Proteomics Tools Website. 7 April 1999, revision date. [Online.]http://www.expasy.ch/tools/peptide-mass.html. [29 June 1999, last date ac-cessed.]
10. Hardham, J. M., L. V. Stamm, S. F. Porcella, J. G. Frye, N. Y. Barnes, J. K.Howell, S. L. Mueller, J. D. Radolf, G. M. Weinstock, and S. J. Norris. 1997.Identification and transcriptional analysis of a Treponema pallidum operonencoding a putative ABC transport system, an iron-activated repressor pro-tein homolog, and a glycolytic pathway enzyme homolog. Gene 197:47–64.
11. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4. Nature (London) 227:680–685.
12. Lowe, A. M., P. A. Lambert, and A. W. Smith. 1995. Cloning of an Entero-coccus faecalis endocarditis antigen: homology with adhesins from some oralstreptococci. Infect. Immun. 63:703–706.
13. Miller, J. N., S. J. Whang, and F. P. Fazzan. 1963. Studies on immunity inexperimental syphilis. II. Treponema pallidum immobilization (TPI) antibodyand the immune response. Br. J. Vener. Dis. 39:199–203.
14. Nielsen, H., J. Engelbrecht, S. Brunak, and G. von Heijne. 1997. Identifica-tion of prokaryotic and eukaryotic signal peptides and prediction of theircleavage sites. Protein Eng. 10:1–6.
15. Norris, S. J., and The Treponema pallidum Polypeptide Research Group.1993. Polypeptides of Treponema pallidum: progress toward understandingtheir structural, functional, and immunologic roles. Microbiol. Rev. 57:750–779.
16. Radolf, J. D., N. R. Chamberlain, A. Clausell, and M. V. Norgard. 1988Identification and localization of integral membrane proteins of virulentTreponema pallidum by phase partitioning with the nonionic detergent Tri-ton X-114. Infect. Immun. 56:490–498.
17. Radolf, J. D., M. V. Norgard, and W. W. Shulz. 1989. Outer membraneultrastructure explains the limited antigenicity of virulent Treponema palli-dum. Proc. Natl. Acad. Sci. USA 86:2051–2055.
18. Towbin, H., T. Staehelin, and J. Gordon. 1979. Electrophoretic transfer ofproteins from polyacrylamide gels to nitrocellulose sheets: procedure andsome applications. Proc. Natl. Acad. Sci. USA 76:4350–4354.
19. Von Heijne, G. 1986. A new method for identifying secretory signal se-quences and for predicting the site of cleavage between a signal sequenceand the mature exported protein is described. Nucleic Acids Res. 14:4683–4690.
20. Walker, E. M., G. A. Zampighi, D. R. Blanco, J. N. Miller, and M. A. Lovett.1989. Demonstration of rare protein in the outer membrane of Treponemapallidum subsp. pallidum by freeze-fracture analysis. J. Bacteriol. 171:5005–5011.
21. Wessel, D., and U. I. Flugge. 1984. A method for the quantitative recovery ofprotein in dilute solution in the presence of detergents and lipids. Anal.Biochem. 138:141–143.
22. Whitelegge, J. P., and K. F. Faull. Unpublished data.23. Whitelegge, J. P., C. Gundersen, and K. F. Faull. 1998. Electrospray-ioniza-
tion mass spectrometry of intact intrinsic membrane proteins. Protein Sci.7:1423–1430.
![Modulation of bacterial outer membrane vesicle …...Outer membrane vesicles (OMVs) bud from the outer membrane (OM) of Gram-negative bacteria [1-4]. These spherical particles are](https://static.documents.pub/doc/80x56/5f0965c97e708231d426a4d6/modulation-of-bacterial-outer-membrane-vesicle-outer-membrane-vesicles-omvs.jpg)
