DNA Vaccine-Encoded Flagellin Can Be Used as an Adjuvant Scaffold to Augment HIV-1 gp41 Membrane Proximal External Region Immunogenicity

Abstract

Flagellin's potential as a vaccine adjuvant has been increasingly explored over the last three decades. Monomeric flagellin proteins are the only known agonists of Toll-like receptor 5 (TLR5). This interaction evokes a pro-inflammatory state that impacts upon both innate and adaptive immunity. While pathogen associated molecular patterns (PAMPs) like flagellin have been used as stand-alone adjuvants that are co-delivered with antigen, some investigators have demonstrated a distinct advantage to incorporating antigen epitopes within the structure of flagellin itself. This approach has been particularly effective in enhancing humoral immune responses. We sought to use flagellin as both scaffold and adjuvant for HIV gp41 with the aim of eliciting antibodies to the membrane proximal external region (MPER). Accordingly, we devised a straightforward step-wise approach to select flagellin-antigen fusion proteins for gene-based vaccine development. Using plasmid DNA vector-based expression in mammalian cells, we demonstrate robust expression of codon-optimized full length and hypervariable region-deleted constructs of Salmonella enterica subsp. enterica serovar Typhi flagellin (FliC). An HIV gp41 derived sequence including the MPER (gp41_607-683) was incorporated into various positions of these constructs and the expressed fusion proteins were screened for effective secretion, TLR5 agonist activity and adequate MPER antigenicity. We show that incorporation of gp41_607-683 into a FliC-based scaffold significantly augments gp41_607-683 immunogenicity in a TLR5 dependent manner and elicits modest MPER-specific humoral responses in a mouse model.

Keywords: DNA vaccine; HIV-1; MPER; adjuvant; flagellin; gp41; membrane proximal external region.

Figure 1

Domain structure of full-length flagellin. Image derived from PDB ID: 1UCU [38] visualized with iCn3D (https://www.ncbi.nlm.nih.gov/Structure/icn3d/full.html) and modified to indicate D0, D1, D2 and D3 domains. Residues 89–96 (QRVRELAV) in the D1 domain are encircled by a red dashed line.

Figure 2

FliC constructs expressed in a mammalian system are secreted and maintain Toll-like receptor 5 (TLR5) agonist activity. (A) Schematic representation of flagellin (FliC) constructs; (B) Western blot of cell lysates, supernatants and PNGase F treated supernatants from transiently transfected 293T. Cells were transfected with pVAX (empty vector), FliC, FliC Δ174–400 or FliC Δ220–320. Samples were collected 48 h post transfection. Blots were probed with a mouse anti-FLAG tag antibody to detect FLAG-tagged flagellin proteins. A mouse anti-tubulin antibody, was used to detect tubulin as a loading control; (C) Measurement of relative secretion level of FLAG-tagged proteins using capture ELISA. Supernatants from transiently transfected 293T were applied to anti-FLAG tag antibody coated plates and probed with a horseradish peroxidase (HRP) conjugated mouse anti-FLAG tag antibody. Results shown represent the mean of 4 different experiments where FliC is set to a value of 1. Error bars represent standard error of the mean; (D) Relative TLR5 agonist activity of secreted FliC proteins. Normalized supernatants from transiently transfected 293T were diluted 1:100 and added to HEK-Blue-hTLR5 cells. After 20 h incubation, the quantity of secreted embryonic alkaline phosphatase (SEAP) produced was determined using a colorimetric enzyme assay. Relative quantities, presented as response ratios, are indicative of TLR5 agonist activity. Results shown represent the mean of 4 different experiments where pVAX is set to a value of 1. Error bars represent standard error of the mean.

Figure 3

Residues QRVRELAV (89–96) of FliC are required for TLR5 agonist activity. (A) Schematic representation of FliC constructs; (B) Western blots of cell lysates and supernatants from transiently transfected 293T. Cells were transfected with pVAX, FliC, FliC Δ174–400, FliC Δ220–320, FliC Δ89–96, FliC Δ89–96 Δ174–400 or FliC Δ89–96 Δ220–320. Blots were probed with a mouse anti-FLAG tag antibody to detect FLAG-tagged flagellin proteins. A mouse anti-tubulin antibody, was used to detect tubulin as a loading control; (C) Relative TLR5 agonist activity of secreted FliC proteins. Normalized supernatants from transiently transfected 293T were diluted 1:100 and added to HEK-Blue-hTLR5 cells. After 20 h incubation, the quantity of SEAP produced was determined using a colorimetric enzyme assay. Relative quantities, presented as response ratios, are indicative of TLR5 agonist activity. Results shown represent the mean of 4 different experiments where pVAX is set to a value of 1. Error bars represent standard error of the mean; (D) Induction of IL-1β transcription detected by RT-PCR and gel electrophoresis. Normalized supernatants from transiently transfected 293T were diluted 1:100 and added to THP-1 or J774A.1 cells. Cells were incubated for 2 h then harvested and RNA was obtained for RT-PCR with IL-1β specific primers. Cells treated with LPS or with supernatant from empty vector transfected cells were used as induced and non-induced controls, respectively. RT-PCR with GAPDH specific primers was used as an internal control. To exclude potential DNA carry-over and amplification or contamination, control reactions omitting the RT enzyme (-RT) or omitting input RNA (negative control) are also shown.

Figure 4

FliC Δ174-400 with gp41_607-683 inserted at its N-terminus is secreted and maintains TLR5 agonist activity. (A) Schematic representation of flagellin constructs; (B) Western blot of cell lysates and supernatants from transiently transfected 293T. Cells were transfected with pVAX, FliC, FliC Δ174–400, FliC Δ220–320, gp41_607–683 FliC, gp41_607–683 FliC Δ174–400, gp41_607–683 FliC Δ220–320, FliC Δ174-[gp41_607–683]-400, FliC Δ220-[gp41_607–683]-320, FliC gp41_607–683, FliC Δ174–400 gp41_607–683 or FliC Δ220–320 gp41_607–683. Samples were collected 48 h post transfection. Blots were probed with a mouse anti-FLAG tag antibody to detect FLAG-tagged flagellin proteins. A mouse anti-tubulin antibody, was used to detect tubulin as a loading control; (C) Measurement of relative secretion level of FLAG-tagged proteins using capture ELISA. Supernatants from transiently transfected 293T were applied to anti-FLAG tag antibody coated plates and probed with an HRP conjugated mouse anti-FLAG tag antibody. Results shown represent the mean of 3 different experiments where FliC is set to a value of 1. Error bars represent standard error of the mean; (D) Relative TLR5 agonist activity of secreted FliC proteins. Normalized supernatants from transiently transfected 293T were diluted 1:100 and added to HEK-Blue-hTLR5 cells. After 20 h incubation, the quantity of SEAP produced was determined using a colorimetric enzyme assay. Relative quantities, presented as response ratios, are indicative of TLR5 agonist activity. Results shown represent the mean of 3 different experiments where pVAX is set to a value of 1. Error bars represent standard error of the mean.

Figure 4

FliC Δ174-400 with gp41_607-683 inserted at its N-terminus is secreted and maintains TLR5 agonist activity. (A) Schematic representation of flagellin constructs; (B) Western blot of cell lysates and supernatants from transiently transfected 293T. Cells were transfected with pVAX, FliC, FliC Δ174–400, FliC Δ220–320, gp41_607–683 FliC, gp41_607–683 FliC Δ174–400, gp41_607–683 FliC Δ220–320, FliC Δ174-[gp41_607–683]-400, FliC Δ220-[gp41_607–683]-320, FliC gp41_607–683, FliC Δ174–400 gp41_607–683 or FliC Δ220–320 gp41_607–683. Samples were collected 48 h post transfection. Blots were probed with a mouse anti-FLAG tag antibody to detect FLAG-tagged flagellin proteins. A mouse anti-tubulin antibody, was used to detect tubulin as a loading control; (C) Measurement of relative secretion level of FLAG-tagged proteins using capture ELISA. Supernatants from transiently transfected 293T were applied to anti-FLAG tag antibody coated plates and probed with an HRP conjugated mouse anti-FLAG tag antibody. Results shown represent the mean of 3 different experiments where FliC is set to a value of 1. Error bars represent standard error of the mean; (D) Relative TLR5 agonist activity of secreted FliC proteins. Normalized supernatants from transiently transfected 293T were diluted 1:100 and added to HEK-Blue-hTLR5 cells. After 20 h incubation, the quantity of SEAP produced was determined using a colorimetric enzyme assay. Relative quantities, presented as response ratios, are indicative of TLR5 agonist activity. Results shown represent the mean of 3 different experiments where pVAX is set to a value of 1. Error bars represent standard error of the mean.

Figure 5

FliC Δ174–400 gp41_607–683 and gp41_607–683 have similar gp41 antigenicity. (A) Western blot of cell lysates, supernatants and PNGase F treated supernatants from transiently transfected 293T. Cells were transfected with pVAX, FliC Δ174–400, FliC Δ174–400 gp41_607–683 or gp41_607–683. Samples were collected 48 h post transfection. Blots were probed with a mouse anti-FLAG tag antibody to detect FLAG-tagged flagellin proteins. A mouse anti-tubulin antibody, was used to detect tubulin as a loading control; (B) Schematic representation of 4E10 and 10E8 binding sites; (C) Binding of MPER-specific broadly neutralizing antibodies (bnAB) to FliC Δ174–400 gp41_607–683 and gp41_607–683. Normalized supernatants from transiently transfected 293T were added to streptavidin coated plates. Captured antigen was probed with either 10E8 or 4E10 antibodies, then an AP-conjugated mouse anti-human IgG secondary antibody was used to detect antigen-bound 10E8 and 4E10. Binding was measured using a fluorescent AP substrate assay and further normalized to levels of captured antigen in each well. Results shown represent the mean of 3 different experiments where gp41_607–683 is set to a value of 1. Error bars represent standard error of the mean.

Figure 6

FliCΔ174–400 augments HIV-1 gp41_607–683 immunogenicity. (A) Detection of HIV-1 gp41_607–683 binding antibodies in vaccinated mice. Female BALB/c mice (10 mice per experimental arm) were injected intramuscularly with either pVAX (empty vector) or DNA vaccines FliCΔ174–400 gp41_607–683 or gp41_607–683 (50 μL in each hind leg at a concentration of 1 μg/μL). Two weeks following the 4th vaccination, mouse serum (1:200 dilution) was analyzed for a gp41_607–683 specific IgG response. Each point represents the mean absorbance value obtained from individual mouse serum (each analyzed in duplicate). Horizontal bars represent average values per vaccine group and error bars represent +/− standard error of the mean. Significance was determined using an unpaired t-test with ** indicating a p value < 0.01; (B) Detection of antibodies binding C-terminal residues of HIV-1 gp41_607–683. Mouse sera from the mice vaccinated with FliCΔ174–400 gp41_607–683 were analyzed for an MPER-specific IgG response by comparing binding to gp41_607–683 (containing the entire MPER) and gp41_607–671 (lacking the twelve C-terminal residues of MPER). Vertical bars graphs represent the mean absorbance value obtained for pooled sera (n = 10 mice) analyzed in duplicate. Error bars represent standard error of the mean.

Figure 7

Deletion of FliC residues 89–96 results in a reduced adjuvant effect on HIV-1 gp41_607–683 immunogenicity. Female BALB/c mice (10 mice per experimental arm) were injected intramuscularly with either pVAX (empty vector) or DNA vaccines FliCΔ174–400 gp41_607–683 or FliC Δ89–96 Δ174–400 gp41_607–683 (50 μL in each hind leg at a concentration of 1 μg/μL). Two weeks following the 4th vaccination, mouse serum (1:200 dilution) was analyzed for a gp41_607–683 specific IgG response. Each point represents the mean absorbance value obtained from individual mouse serum (each with two replicates per mouse). Horizontal bars represent average values per vaccine group and error bars represent standard error of the mean. Significance was determined using an unpaired t-test with ** indicating a p value < 0.01.