TprC/D (Tp0117/131), a trimeric, pore-forming rare outer membrane protein of Treponema pallidum, has a bipartite domain structure

Abstract

Identification of Treponema pallidum rare outer membrane proteins (OMPs) has been a longstanding objective of syphilis researchers. We recently developed a consensus computational framework that employs a battery of cellular localization and topological prediction tools to generate ranked clusters of candidate rare OMPs (D. L. Cox et al., Infect. Immun. 78:5178-5194, 2010). TP0117/TP0131 (TprC/D), a member of the T. pallidum repeat (Tpr) family, was a highly ranked candidate. Circular dichroism, heat modifiability by SDS-PAGE, Triton X-114 phase partitioning, and liposome incorporation confirmed that full-length, recombinant TprC (TprC(Fl)) forms a β-barrel capable of integrating into lipid bilayers. Moreover, TprC(Fl) increased efflux of terbium-dipicolinic acid complex from large unilamellar vesicles and migrated as a trimer by blue-native PAGE. We found that in T. pallidum, TprC is heat modifiable, trimeric, expressed in low abundance, and, based on proteinase K accessibility and opsonophagocytosis assays, surface exposed. From these collective data, we conclude that TprC is a bona fide rare OMP as well as a functional ortholog of Escherichia coli OmpF. We also discovered that TprC has a bipartite architecture consisting of a soluble N-terminal portion (TprC(N)), presumably periplasmic and bound directly or indirectly to peptidoglycan, and a C-terminal β-barrel (TprC(C)). Syphilitic rabbits generate antibodies exclusively against TprC(C), while secondary syphilis patients fail to mount a detectable antibody response against either domain. The syphilis spirochete appears to have resolved a fundamental dilemma arising from its extracellular lifestyle, namely, how to enhance OM permeability without increasing its vulnerability to the antibody-mediated defenses of its natural human host.

Fig 1

TprC forms a β-barrel. (A) Tryptophan fluorescence emission spectra of unfolded and folded TprC^Fl in urea and DDM buffer, respectively. (B) CD spectra of unfolded and folded TprC^Fl (5 μM) in DDM buffer and OmpG (5 μM) in 50 mM NaCl, 10 mM Tris (pH 7.5), 0.2% n-octyl-β-d-glucopyranoside. (C) Heat modifiability of TprC^Fl, E. coli OmpG, 326^βb, and 326^P1-4+. Proteins were stained with GelCode Blue following SDS-PAGE with (+) or without (−) boiling in final sample buffer (SB). Molecular mass standards (kDa) for SDS-PAGE and immunoblot analyses are on the left of each panel. (D) T. pallidum samples (2 × 10⁸ organisms) were dissolved in SB with (+) or without (−) boiling and immunoblotted using rat anti-TprC^Sp antiserum.

Fig 2

TprC is amphiphilic and trimeric and displays pore-forming activity. (A) Ten micrograms each of TprC^Fl, 326^P1-4+, and folded 326^βb was phase partitioned in TX-114 and stained with GelCode Blue following SDS-PAGE. Lanes: aqueous (Aq) and detergent-enriched (Det) phases. (B) Liposomes were reconstituted with 10 μg each of TprC^Fl, 326^P1-4+, or 326^βb followed by sucrose density gradient ultracentrifugation. Fractions were subjected to immunoblotting with antisera directed against TprC^Fl, 326^P1-4+, or 326^βb following SDS-PAGE. Lanes: top fractions (TF) contain liposome-incorporated material whereas middle and bottom fractions (MF and BF, respectively) contain unincorporated material. (C) BN-PAGE and immunoblot analysis of TprC^Fl and T. pallidum lysate solubilized in 0.5% DDM with 50 mM Tris (pH 7.0). Molecular mass standards (kDa) are shown on the left. (D) Quenching of LUVs encapsulating Tb(DPA)₃³⁻ after incubation with 100 nM TprC^Fl, E. coli OmpF, 326^βb, or 326^P1-4+ in 50 mM Tris (pH 7.5), 100 mM NaCl supplemented with 5 mM EDTA. Each bar represents the mean ± standard error of the mean for three independent experiments.

Fig 3

TprC expressed by T. pallidum associates with the peptidoglycan sacculus. (A) TX-114 phase partitioning of T. pallidum lysates (1 × 10⁹ organisms) without (−) or with (+) preincubation with 2% DDM. Lanes: whole cells (WC), TX-114-insoluble material (Ins), aqueous phase (Aq), and detergent-enriched phase (Det). (B) Extensively washed TX-114-insoluble material visualized in negatively stained whole mounts by transmission electron microscopy. (C) Immunoblot analysis of TX-114-insoluble material with anti-TprC^Sp or immune rabbit serum (IRS). Molecular mass standards (in kilodaltons) are shown at left.

Fig 4

TprC is expressed at low abundance and is accessible to proteinase K in motile treponemes. (A) Quantitative immunoblot analysis of TprC expressed in T. pallidum. T. pallidum lysates in amounts indicated were immunoblotted with anti-TprC^Sp antiserum followed by densitometric analysis; a standard curve generated from densitometric values obtained for graded amounts of TprC^Fl was used to determine the copy number of TprC per T. pallidum cell. Molecular mass standards (kDa) are indicated on the left. (B) Immunoblot analysis of TprC, detected by the anti-TprC^Sp antiserum, in motile treponemes (1.0 × 10⁸ T. pallidum organisms/lane) treated for 1 h with graded concentrations of proteinase K (PK). (C) PK accessibility of TprC and two periplasmic controls in intact and detergent-lysozyme-treated organisms incubated with (+) or without (−) 10 μg/ml of PK. Each lane represents 1.0 × 10⁸ T. pallidum organisms immunoblotted with antisera to TprC^Sp, TP0453, or TroA. (D) Live imaging of intact treponemes with and without incubation with 100 μg/ml of PK. Asterisks and arrowheads indicate flexing and wave form propagation, respectively. Time scale is 0 to 0.12 s.

Fig 5

TprC antibodies have opsonic activity. (A) Percentages of rabbit peritoneal macrophages containing internalized treponemes. Results shown are means ± standard errors of the means for three independent experiments. (B) Representative micrographs of macrophages incubated with motile treponemes and the indicated sera. Each image is a composite of bright-field images, nuclei stained with DAPI, and FITC-labeled T. pallidum. (C) Percentages of treponemes recovered following incubation of rabbit peritoneal macrophages for 4 h with the indicated sera. Results shown are means ± standard errors of the means for three independent experiments. P values of <0.05 (Student's t test) were considered significant.

Fig 6

TprC^C contains the β-barrel of TprC and forms a trimer. (A to C) CD spectroscopy, heat modifiability, and TX-114 phase partitioning of TprC^N and TprC^C. Aq and Det, aqueous and detergent-enriched phases, respectively. (D) Quenching of Tb(DPA)₃³⁻ encapsulated within LUVs following incubation with TprC^Fl, TprC^N, TprC^C, and E. coli OmpF. (E) SDS-PAGE of unboiled TprC^N and TprC^C (asterisk indicates TprC^C trimer). Numbers at left in panels B and E are molecular masses in kilodaltons.

Fig 7

T. pallidum-infected rabbits, but not humans, mount an antibody response against TprC. (A) Reactivities of normal rabbit serum (NRS), immune rabbit serum (IRS; representative of 3 different animals), normal human serum (NHS), and pooled human secondary syphilitic sera (HSS) against folded, unboiled TprC^Fl; 326^P1-4+; and folded, unboiled 326^βb. (B) Immunoblot reactivities of individual patient sera. (C) Immunoblot reactivities of the anti-TprC^Fl antiserum and IRS used for the opsonophagocytosis assays in Fig. 5 against TprC^N and folded, unboiled TprC^C. Each immunoblot assay was performed using 100 ng of recombinant protein. Numbers at left are molecular masses in kilodaltons.

Fig 8

Topological models of TprC, TprI, and TprF. (A) Comparison of domain architectures. (B) TprC/D and TprI are proposed to consist of trimers with identical β-barrels and N-terminal periplasmic domains that directly or indirectly tether the barrels to the peptidoglycan (PG) sacculus. TprF is entirely periplasmic. TprC^Sp is depicted as periplasmic based on its solubility in aqueous buffer. TprI shares with TprF a putative periplasmic fragment not present in TprC.