Engineering filamentous phage carriers to improve focusing of antibody responses against peptides

Abstract

The filamentous bacteriophage are highly immunogenic particles that can be used as carrier proteins for peptides and presumably other haptens and antigens. Our previous work demonstrated that the antibody response was better focused against a synthetic peptide if it was conjugated to phage as compared to the classical carrier, ovalbumin. We speculated that this was due, in part, to the relatively low surface complexity of the phage. Here, we further investigate the phage as an immunogenic carrier, and the effect reducing its surface complexity has on the antibody response against peptides that are either displayed as recombinant fusions to the phage coat or are chemically conjugated to it. Immunodominant regions of the minor coat protein, pIII, were removed from the phage surface by excising its N1 and N2 domains (Delta3 phage variant), whereas immunodominant epitopes of the major coat protein, pVIII, were altered by reducing the charge of its surface-exposed N-terminal residues (Delta8 phage variant). Immunization of mice revealed that the Delta3 variant was less immunogenic than wild-type (WT) phage, whereas the Delta8 variant was more immunogenic. The immunogenicity of two different peptides was tested in the context of the WT and Delta3 phage in two different forms: (i) as recombinant peptides fused to pVIII, and (ii) as synthetic peptides conjugated to the phage surface. One peptide (MD10) in its recombinant form produced a stronger anti-peptide antibody response fused to the WT carrier compared to the Delta3 phage carrier, and did not elicit a detectable anti-peptide response in its synthetic form conjugated to either phage carrier. This trend was reversed for a different peptide (4E10(L)), which did not produce a detectable anti-peptide antibody response as a recombinant fusion; yet, as a chemical conjugate to Delta3 phage, but not WT phage, it elicited a highly focused anti-peptide antibody response that exceeded the anti-carrier response by approximately 65-fold. The results suggest that focusing of the antibody response against synthetic peptides can be improved by decreasing the antigenic complexity of the phage surface.

Figure 1

Pictorial depiction of wild-type and engineered filamentous phage coat proteins. (A) Domain structure of WT pIII. The bracket indicates the surface-exposed N1 and N2 domains of pIII that were removed, leaving (B) the Δ3 variant comprising four N-terminal residues, part of the second glycine-rich linker (G2) and the C-terminal (CT) domain to allow phage assembly into particles containing a single ssDNA genome. (C) N-terminal amino acid sequence of mature pVIII from f1 phage, f8-5 phage, and the modified Δ8 phage variant showing alterations to CBRs.

Figure 1

Pictorial depiction of wild-type and engineered filamentous phage coat proteins. (A) Domain structure of WT pIII. The bracket indicates the surface-exposed N1 and N2 domains of pIII that were removed, leaving (B) the Δ3 variant comprising four N-terminal residues, part of the second glycine-rich linker (G2) and the C-terminal (CT) domain to allow phage assembly into particles containing a single ssDNA genome. (C) N-terminal amino acid sequence of mature pVIII from f1 phage, f8-5 phage, and the modified Δ8 phage variant showing alterations to CBRs.

Figure 2

Average half-maximal anti-f8-5, Δ3, Δ8, or Δ8Δ3 Ab titers after one, three, and five immunizations. Error bars show standard error of the arithmetic mean. ANOVA was used to compare the significance between immunization groups at the first, third and fifth immunizations (p-values for 1st, 2nd, 3rd immunizations: < 1×10⁻⁵). Note that the Y-axis is log scaled.

Figure 3

Immunization with modified phage variants redirects the Ab response to different sets of epitopes. Pooled sera from groups of 5 mice, immunized three times with WT f8-5 phage (A), or phage variants Δ3 (B), Δ8 (C), or Δ8Δ3 (D) were used to probe equivalent amounts of all four phage variants in western blots. The phage variants being probed are named along sample lanes at the top of each blot, and the immunizing phage are named on the bottom. The arrow indicates an unusual band produced by immunization with Δ8 and Δ8Δ3 phage.

Figure 4

Average half-maximal Ab titers against phage (f8-5 or Δ3), peptide and SPDP crosslinker produced by mice immunized two and four times with phage bearing recombinant or chemically-conjugated MD10 or 4E10_L peptides. Immune sera from: (A) phage displaying recombinant MD10 peptide; (B) MD10 synthetic peptide conjugated to phage via the SPDP crosslinker; (C) phage displaying recombinant 4E10_L^A peptide; (D) 4E10_L^B synthetic peptide conjugated to phage via the SPDP crosslinker; (E) SPDP crosslinker conjugated to phage and blocked with cysteine. Error bars show standard error of the arithmetic mean for each immunization group. Note that the Y-axis is log scaled.

Figure 5

Boxplot showing (A) ratios of anti-peptide to anti-phage half-maximal Ab titers calculated for individual mice or (B) ratios of anti-SPDP crosslinker to anti-phage half-maximal Ab titers calculated for individual mice immunized two and four times with phage bearing recombinant or chemically-conjugated MD10 or 4E10_L peptides. Asterisks denote instances where titer ratios could not be determined due to undetectable anti-peptide, anti-SPDP or anti-phage Ab responses. Note that the Y-axis is log scaled.

Figure 5

Boxplot showing (A) ratios of anti-peptide to anti-phage half-maximal Ab titers calculated for individual mice or (B) ratios of anti-SPDP crosslinker to anti-phage half-maximal Ab titers calculated for individual mice immunized two and four times with phage bearing recombinant or chemically-conjugated MD10 or 4E10_L peptides. Asterisks denote instances where titer ratios could not be determined due to undetectable anti-peptide, anti-SPDP or anti-phage Ab responses. Note that the Y-axis is log scaled.