Use of Epivolve phage display to generate a monoclonal antibody with opsonic activity directed against a subdominant epitope on extracellular loop 4 of Treponema pallidum BamA (TP0326)

Abstract

Introduction: Syphilis, a sexually transmitted infection caused by the spirochete Treponema pallidum (Tp), is resurging globally. Tp's repertoire of outer membrane proteins (OMPs) includes BamA (β-barrel assembly machinery subunit A/TP0326), a bipartite protein consisting of a 16-stranded β-barrel with nine extracellular loops (ECLs) and five periplasmic POTRA (polypeptide transport-associated) domains. BamA ECL4 antisera promotes internalization of Tp by rabbit peritoneal macrophages.

Methods: Three overlapping BamA ECL4 peptides and a two-stage, phage display strategy, termed "Epivolve" (for epitope evolution) were employed to generate single-chain variable fragments (scFvs). Additionally, antisera generated by immunizing mice and rabbits with BamA ECL4 displayed by a Pyrococcus furiosus thioredoxin scaffold (PfTrx^BamA/ECL4). MAbs and antisera reactivities were evaluated by immunoblotting and ELISA. A comparison of murine and rabbit opsonophagocytosis assays was conducted to evaluate the functional ability of the Abs (e.g., opsonization) and validate the mouse assay. Sera from Tp-infected mice (MSS) and rabbits (IRS) were evaluated for ECL4-specific Abs using PfTrx^BamA/ECL4 and overlapping ECL4 peptides in immunoblotting and ELISA assays.

Results: Each of the five mAbs demonstrated reactivity by immunoblotting and ELISA to nanogram amounts of PfTrx^BamA/ECL4. One mAb, containing a unique amino acid sequence in both the light and heavy chains, showed activity in the murine opsonophagocytosis assay. Mice and rabbits hyperimmunized with PfTrx^BamA/ECL4 produced opsonic antisera that strongly recognized the ECL presented in a heterologous scaffold and overlapping ECL4 peptides, including S2. In contrast, Abs generated during Tp infection of mice and rabbits poorly recognized the peptides, indicating that S2 contains a subdominant epitope.

Discussion: Epivolve produced mAbs target subdominant opsonic epitopes in BamA ECL4, a top syphilis vaccine candidate. The murine opsonophagocytosis assay can serve as an alternative model to investigate the opsonic potential of vaccinogens. Detailed characterization of BamA ECL4-specific Abs provided a means to dissect Ab responses elicited by Tp infection.

Keywords: BamA ECL4; Pyrococcus furiosus thioredoxin; Treponema pallidum; monoclonal antibody; opsonic antibody; outer membrane protein; subdominant epitope; syphilis.

Figure 1

Schematic of mAb generation using Epivolve. A peptide incorporating a non-native amino acid at a desired site is used to pan a scFv phage library for peptide binders. Phages that bind the modified, but not the native, peptide undergo AXM mutagenesis (21) to generate phages that recognize the native peptide +/− the modified peptide with high affinity.

Figure 2

BamA ECL4 is a target antigen for mAb generation. (A) Ribbon diagram for the structural model of Tp BamA (TP0326) depicting the β-barrel, ECL4, and the five periplasmic polypeptide-transport-associated (POTRA) domains. ECLs 4, 6, and 7 form a dome that occludes the barrel opening. (B) Sequences of the three overlapping ECL4 peptides (S1–S3) used for Epivolve. The modified aspartic acid residue in each peptide is represented in black. A light blue line indicates predicted linear B-cell epitopes in S2 and S3. Asterisk indicates an additional residue added to centrally position the modified residue in the S1 peptide. Blue arrow above the sequence indicates the glutamine to leucine substitution in the Mexico A strain of T. pallidum (10). (C) Representative titration ELISAs using native and modified peptides, pre- and post-error-prone PCR.

Figure 3

Heavy and light chains used to form full-length mAbs. (A) Five heavy and three light chains (HC and LC, respectively) were fused to a mouse IgG2 constant domain to form five distinct full-length mAbs. A multiple sequence alignment of the (B) HCs and (C) LCs. The amino acid residues with a consensus identity of over 51% are represented by gray shading, while lowercase letters indicate amino acids with a mismatched identity. HC and LC CDRs predicted using abYsis (34) are denoted by cyan or pink shading, respectively. Asterisk (*) indicates a non-conserved residue in all three LC CDR3s.

Figure 4

Immunoreactivity profiles of BamA ECL4 mAbs. (A) Titration ELISAs of the mAbs against native and irrelevant peptides (magenta and grey, respectively) and PfTrx^BamA/ECL4 and PfTrx ^Empty (blue and black, respectively). (B) mAb AUC values calculated for native peptide and PfTrx^BamA/ECL4. (C) Immunoblot reactivities against graded nanogram amounts of PfTrx ^BamA/ECL4 (21.4 kDa) or the BamA β-barrel (44.9 kDa). Based on the MW, PfTrx^BamA/ECL4 and BamA β-barrel were loaded at a 2:1 ratio, with 1 ng corresponding to 0.047 and 0.022 picomoles, respectively. PfTrx^Empty (200 ng) and nonspecific mAb (IGX6939) served as specificity controls. Immunoblots were done using mAbs at 4 μg/ml. (D) A rat polyclonal BamA ECL4 (residues 568–602) antiserum (10) (diluted 1:1,000) was used as a comparator. (E) Immunoblot reactivity of IGX7141 mAb (4 μg/ml) against graded nanogram amounts of Nichols (top) and Mexico A PfTrx^BamA/ECL4 (bottom) variants.

Figure 5

Identification of an opsonic BamA ECL4 mAb. (A) Immunoblot reactivities of pooled sera (diluted 1:1,000) from five mice hyperimmunized with PfTrx^BamA/ECL4 against graded nanogram amounts of TbpB-LCL^BamA/ECL4. (B) ELISA reactivity of murine PfTrx^BamA/ECL4 antisera or NMS with native S1, S2, and S3 peptides represented as AUC values. (C) Freshly extracted Tp were pre-incubated with 10% heat-inactivated NMS, pooled MSS, mouse antisera to PfTrx^BamA/ECL4, PfTrx^Empty, TP0751 or Tpp17, or 10 μg/ml of the individual mAbs followed by incubation with murine BMDMs for 4 h at an MOI 10:1. Phagocytic indices were determined as described in Materials and methods. Asterisks show significant differences with p-values of ≤0.05, ≤0.01, or <0.0001. (D) Each representative confocal micrograph is a composite of 9–12 consecutive Z-stack planes with labeling of Tp, plasma membranes, and nuclei shown in green, red, and blue, respectively. (E) Immunoblot reactivity of pooled MSS and NMS (diluted 1:250) against 200 ng of PfTrx^BamA/ECL4. ELISA reactivity (AUC values) of pooled MSS against (F) PfTrx^BamA/ECL4 and (G) the S1, S2, and S3 peptides.

Figure 6

Absence of antibodies to the subdominant BamA ECL4 epitope in immune rabbit serum. (A) Immunoblot reactivities of sera (diluted 1:1,000) from rabbits hyperimmunized with PfTrx^BamA/ECL4 against graded nanogram amounts of TbpB-LCL^BamA/ECL4. (B) ELISA reactivity of rabbit PfTrx^BamA/ECL4 antisera or NMS with native S1, S2, and S3 peptides represented as AUC values. (C) Freshly extracted Tp were pre-incubated with 10% heat-inactivated NRS, five individual IRS, or sera from rabbits hyperimmunized with PfTrx^BamA/ECL4, PfTrx^Empty, TP0751, or Tpp17 followed by incubation with rabbit peritoneal macrophages for 4 h at an MOI 10:1. Phagocytic indices were determined as described in Materials and methods. Asterisks show significant differences with p-values of ≤0.05, ≤0.01, or ≤0.001. (D) Each representative confocal micrograph is a composite of 9–12 consecutive Z-stack planes with labeling of Tp, plasma membranes, and nuclei shown in green, red, and blue, respectively. (E) Immunoblot reactivity of individual IRS and NRS (diluted 1:250) against 200 ng of PfTrx^BamA/ECL4. ELISA reactivity (AUC values) of IRS against (F) PfTrx^BamA/ECL4 and (G) the S1, S2, and S3 peptides.