Identification and Functional Assessment of the First Placental Adhesin of Treponema pallidum That May Play Critical Role in Congenital Syphilis

Abstract

Syphilis is a global, re-emerging sexually transmitted infection and congenital syphilis remains a major cause of adverse pregnancy outcomes due to bacterial infection in developing nations with a high rate of fetus loss. The molecular mechanisms involved in pathogenesis of the causative agent, Treponema pallidum subsp. pallidum remain poorly understood due to the difficulties of working with this pathogen, including the inability to grow it in pure culture. To reduce the spread of syphilis, we must first increase our knowledge of the virulence factors of T. pallidum and their contribution to syphilis manifestations. Tp0954 was predicted to be a surface lipoprotein of T. pallidum. Therefore, we experimentally demonstrated that Tp0954 is indeed a surface protein and further investigated its role in mediating bacterial attachment to various mammalian host cells. We found that expression of Tp0954 in a poorly adherent, but physiologically related derivative strain of the Lyme disease causing spirochete Borrelia burgdorferi B314 strain promotes its binding to epithelial as well as non-epithelial cells including glioma and placental cell lines. We also found that Tp0954 expression facilitates binding of this strain to purified dermatan sulfate and heparin, and also that bacterial binding to mammalian cell lines is mediated by the presence of heparan sulfate and dermatan sulfate in the extracellular matrix of the specific cell lines. These results suggest that Tp0954 may be involved not only in initiating T. pallidum infection by colonizing skin epithelium, but it may also contribute to disseminated infection and colonization of distal tissues. Significantly, we found that Tp0954 promotes binding to the human placental choriocarcinoma BeWo cell line, which is of trophoblastic endocrine cell type, as well as human placental tissue sections, suggesting its role in placental colonization and possible contribution to transplacental transmission of T. pallidum. Altogether, these novel findings offer an important step toward unraveling syphilis pathogenesis, including placental colonization and T. pallidum vertical transmission from mother to fetus during pregnancy.

Keywords: Borrelia burgdorferi; Tp0954; Treponema pallidum subspecies pallidum; congenital syphilis; heterologous expression system; placenta binding adhesin; surrogate system.

Figure 1

Expression of Tp0954 in B. burgdorferi B314 strain and detection on the surface of transformed spirochetes. (A) Schematic of the construct expressing the Tp0954 coding region driven by the B. burgdorferi ospC promoter (pTp-Bb). (B) Tp0954 expression in B314(pTp-Bb) and not vector transformed B314(V) strain was detected by RT-PCR using Tp0954 primers. Positive control RT-PCR is also shown in which primers were used to amplify the ospC gene of B. burgdorferi. RT-PCR products were absent when reverse transcriptase was not included in the reaction indicting specificity of RT-PCR. (C) Detection of three Tp0954 protein bands (marked by horizontal lines) in T. pallidum (Tp) Nichols strain as well as Tp0954 transformed B314 strain, B314(pTp-Bb) using antibodies against 281 N-terminal amino acids of mature Tp0954 protein (anti-Tp0954N) suggest differential post-translational processing of full size Tp0954 protein of ~50 kD size. *Non-specific binding with cross reactivity of antibodies with unrelated B314 proteins. (D) Limited proteinase K treatment of intact spirochetes digested only surface Tp0954 versions and B. burgdorferi OspC protein but not periplasmic flagellar protein. (E,F) IFA using anti-Tp0954N primary antibodies followed by TRITC-labeled secondary antibodies using unpermeabilized spirochetes confirmed the surface localization of Tp0954 only in Tp0954 transformed B314 and not in the control, B314(V), strain. (G,H) The lack of staining of flagella in unpermeabilized spirochetes by anti-FlaB monoclonal antibodies treatment followed by anti-mouse Alexa fluor 488 antibodies indicate that integrity of outer membrane of B314 was maintained during IFA procedure. (I,J) Staining of B. burgdorferi flagella after permeabilization with methanol indicates the specific reactivity of anti-FlaB antibodies. All spirochetes present in each microscopic field are detected by co-labeling with DNA stain, DAPI and superimposed images are shown in (E–J). Bar indicates 20 μ size.

Figure 2

Tp0954-expressing bacteria bind to various mammalian cell lines. (A,D) Radiolabeled Tp0954-expressing B314 and plasmid transformed control strains were used to evaluate binding to mammalian epithelial cell lines. B314(pTp-Bb) bound at a 5-fold higher level to Vero cells (A), and 6-fold higher level to HEK293 (293) cells to (D) as compared B314 transformed by the empty pJSB175 vector, B314(V). (B,C) Binding assay conducted followed by washing and bioluminescence measurement confirmed that B314(pTP-Bb) binds at significantly higher level to 293 and C6 glioma cells. (E) B314(pTp-Bb) bound 3.6-fold higher level to the C6 glioma neuronal cells compared to control strain, while (F) B314(pTp-Bb) binds relatively weakly to endothelial EA.hy926 cell line. (G) Increase in binding of B314(pTp-Bb) was 3.5-fold higher than B314(V) to placental BeWo cell line. Statistically significant difference was determined by unpaired two-tailed t-test with Welch's correction (****P < 0.0001).

Figure 3

Tp0954 mediated binding to purified GAGs. (A) Homology of Tp0954 epitopes with VAR2CSA domains of P. falciparum PfEMP1 protein. (B) Recombinant Tp0954 protein bound at significantly higher levels to dermatan sulfate than other chondroitin sulfates. (C) Binding to heparin, the most negatively charged biological molecule, enhanced in the most pronounced manner on expression of Tp0954 in B314 strain. Tp0954-expressing B314 binding increased minimally to empty wells (PBS) used as negative control (not shown). (D–F) Binding of B314(pTp-Bb) increased in the most pronounced manner compared to control B314(V) strain to dermatan sulfate (chondroitin sulfate B) among three types of chondroitin sulfates examined. Statistically significant difference was determined by unpaired two-tailed t-test with Welch's correction (****P < 0.0001).

Figure 4

Heparan sulfate and dermatan sulfate as ECM components on specific mammalian cells are required for Tp0954-expressing spirochetes binding. (A,C) Low level binding of B314(V) strain were observed to both 293 and C6 glioma cells and removal of GAGs by specific lyases reduced binding further. (B) Pretreatment of 293 with chondroitinase ABC and not by a combination of heparinase I and III (HI and HIII) inhibited binding of radiolabeled Tp0954-expressing B314 strain, indicating contribution of dermatan sulfate to adherence on these cells, while (D) removal of both heparan sulfate and dermatan sulfate significantly reduced binding of B314(pTp-Bb) to C6 cell line. GAG-lyases are labeled as: NT-No Treatment, Chon-ABC-chondroitinase ABC, and HI+ HII-heparinase I + III. Statistically significant difference was determined by unpaired two-tailed t-test with Welch's correction (*P < 0.05, ****P < 0.0001, ns, not significant).

Figure 5

B314 gains ability to bind to human placental tissue sections after Tp0954 expression. (A,B) B314(pTp-Bb) strain that expressed Tp0954 on the spirochete surface efficiently binds to placental tissue sections as detected by bioluminescence measurement by IVIS-50 after addition of D-luciferin substrate for luciferase enzyme present in these spirochetes. Control B314(V) strain does not bind to placental tissue significantly confirming poor adherence ability of this strain to mammalian cells. (C) A broad overview of binding of control B314(V) labeled with FITC conjugated anti-B. burgdorferi antibodies by IFA showed barely detectable spirochetes on placental tissue section that was visualized by labeling with wheat germ agglutinin conjugated Alexa fluor 647 shown by red fluorescence, and (D) poor binding by these control spirochetes was confirmed by observation of sections at higher magnification. (E) A significant increase in binding to placental tissue section after Tp0954 expression in B314 strain was detected by observation of green fluorescent spirochetes using FITC-conjugated antibodies. (F) Spirochetes labeled against B. burgdorferi can be more clearly observed at higher magnification depicting significantly higher level of binding of B314(pTp-Bb) compared to B314(V) shown in (D). (C–F) Bar on top panels indicate 100 μ and bottom panels depict 20 μ size.

Figure 6

IFA of T. pallidum co-cultured with low passage rabbit epithelial Sf1Ep cells. (A) IFA using primary polyclonal mouse anti-Tp0954N antibodies followed by TRITC-conjugated secondary antibodies using unpermeabilized T. pallidum SS14 strain co-cultured with Sf1Ep cells showed punctate staining confirming the surface localization of Tp0954. (B) The lack of staining of flagella by anti-FlaA in unpermeabilized T. pallidum indicate that integrity of outer membrane spirochetes was maintained during IFA procedure, and (C) staining of flagellar protein after permeabilization indicate antibodies reactivity and periplasmic location of flagella. Bar indicates 20 μ size.