Efficient single tobamoviral vector-based bioproduction of broadly neutralizing anti-HIV-1 monoclonal antibody VRC01 in Nicotiana benthamiana plants and utility of VRC01 in combination microbicides

Abstract

Broadly neutralizing monoclonal antibodies (bnMAbs) may offer powerful tools for HIV-1 preexposure prophylaxis, such as topical microbicides. However, this option is hampered due to expensive MAb biomanufacturing based on mammalian cell culture. To address this issue, we developed a new production system for bnMAb VRC01 in Nicotiana benthamiana plants using a tobamovirus replicon vector. Unlike conventional two-vector-based expression, this system was designed to overexpress full-length IgG1 from a single polypeptide by means of kex2p-like enzyme recognition sites introduced between the heavy and light chains. An enzyme-linked immunosorbent assay (ELISA) revealed that gp120-binding VRC01 IgG1 was maximally accumulated on 5 to 7 days following vector inoculation, yielding ~150 mg of the bnMAb per kg of fresh leaf material. The plant-made VRC01 (VRC01p) was efficiently purified by protein A affinity followed by hydrophobic-interaction chromatography. ELISA, surface plasmon resonance, and an HIV-1 neutralization assay demonstrated that VRC01p has gp120-binding affinity and HIV-1-neutralization capacity virtually identical to the human-cell-produced counterpart. To advance VRC01p's use in topical microbicides, we analyzed combinations of the bnMAb with other microbicide candidates holding distinct antiviral mechanisms in an HIV-1 neutralization assay. VRC01p exhibited clear synergy with the antiviral lectin griffithsin, the CCR5 antagonist maraviroc, and the reverse transcriptase inhibitor tenofovir in multiple CCR5-tropic HIV-1 strains from clades A, B, and C. In summary, VRC01p is amenable to robust, rapid, and large-scale production and may be developed as an active component in combination microbicides with other anti-HIV agents such as antiviral lectins, CCR5 antagonists, and reverse transcriptase inhibitors.

Fig 1

Single tobamoviral vector-based overexpression of VRC01 in N. benthamiana. (A) Schematic representation of two VRC01 vector constructs, L-(KP6pp)-H and H-(KP6pp)-L. RB and LB, right and left borders, respectively; P, Arabidopsis Act2 promoter; RdRP, RNA-dependent RNA polymerase with artificial introns (28); M, movement protein; LC and HC, light and heavy chains, respectively; KP6pp, amino acid peptide containing two kex2p recognition sites at the N and C termini; UTR, untranslated region. (B) Quantification of VRC01 accumulation in leaf extracts from L-(KP6pp)-H and H-(KP6pp)-L constructs at 5 and 7 dpi. The amounts, in milligrams per kg of fresh leaf material, of VRC01 were quantified by gp120 ELISA. Data are expressed as means ± standard errors of the means (SEM) of biological triplicates. (C) Nonreducing SDS-PAGE analysis of crude extracts of N. benthamiana leaves expressing the L-(KP6pp)-H construct VRC01 (VRC01p). At 7 dpi, leaf proteins expressing VRC01 were extracted with 100 mM phosphate (pH 6.0), 100 mM sodium chloride, 40 mM ascorbic acid buffer. Lane 1, control leaf extract; lane 2, VRC01p extract. The arrow indicates the position of assembled VRC01 at ∼150 kDa. (D) Nonreducing SDS-PAGE analysis of purified VRC01p. The bnMAb was purified by protein A (lane 1) followed by a Phenyl HP resin (lane 2) as described in Materials and Methods. A total of 10 μg of purified protein was loaded into each well. After the two-step purification, the H₂L₁ or Fab-Fc by-product (∼100 kDa, present after protein A purification) was efficiently removed, and VRC01p was purified to >98%.

Fig 2

Comparison of VRC01p with VRC01_HEK. (A) gp120 ELISA using recombinant gp120 proteins from virus strains Q769.h5 (clade A), SF162 (clade B), and DU156 (clade C). The assay was performed in triplicate, and each data point represents the mean ± SEM. The half-maximal effective concentrations (EC₅₀s) of VRC01_HEK and VRC01p were determined by nonlinear regression analysis (GraphPad Prism 5.0) to be 0.28 and 0.35 μg/ml, respectively, for gp120_Q769.h5; 4.24 and 4.11 μg/ml, respectively, for gp120_SF162; and 15.1 and 67.9 μg/ml, respectively, for gp120_DU156. (B) SPR. Each VRC01 protein was captured on a sensor chip via an anti-human IgG (Fc) antibody of IgG1 isotype, and various concentrations of recombinant gp120 from SF162 were used as analytes. For each protein, the assay was performed in triplicate. A representative sensorgram obtained with VRC01p is shown. The fitted curves, based on the 1:1 binding kinetics, represent the concentrations of gp120 (50, 16.66, 5.55, 1.85, and 0.617 μg/ml from top to bottom). The equilibrium dissociation constants, K_D, for VRC01_HEK and VRC01p were determined to be 3.82 ± 0.60 and 3.97 ± 0.74 nM, respectively. Data are expressed as means ± standard deviations (SD) of experimental triplicate analysis. (C) Anti-HIV-1 activity of VRC01_HEK and VRC01p against Env-pseudotyped SF162 virus in HOS-CD4-CCR5⁺ cells. Percent neutralization was calculated by dividing the luminescence of sample wells by that of virus-only control wells. IC₅₀s for each VRC01 were determined by nonlinear regression analysis (GraphPad Prism 5.0; Table 1). A representative graph is shown, with each data point representing the mean ± SEM of experiments done in triplicate. No statistical difference was found based on average IC₅₀s for VRC01_HEK and VRC01p.

Fig 3

HIV-1 neutralization by VRC01p, GRFT, and the VRC01p-GRFT combination. The clade A Q769.h5, clade B SF162 and SS1196.1, and clade C Du156 and ZM214M.PL15 pseudoviruses were used. The starting concentrations for each neutralizing agent were as follows: 2 μg/ml (or 13.3 nM) for Q769.h5, 20 μg/ml (or 133.3 nM) for SF162, 10.54 μg/ml (or 70.3 nM) for SS1196.1, 1.08 μg/ml (or 7.2 nM) for Du156, and 1.54 μg/ml (or 10.3 nM) for ZM214M.PL15 for neutralization by VRC01p; and 16.4 nM (Q769.h5), 20 nM (SF162), 0.244 nM (SS1196.1), 0.096 nM (Du156), and 20 nM (ZM214M.PL15) for neutralization by GRFT. The effect of the combination of VRC01p and GRFT against each virus was tested by dilution of the two agents starting at a fixed 1:1 ratio proportional to their IC₅₀s. The graphs, plotted in GraphPad Prism 5.0, are representative of three independent experiments each done in triplicate. Each data point represents the mean percentage inhibition ± SEM of experimental triplicates.

Fig 4

HIV-1 neutralization by VRC01p, MVC, and the VRC01p-MVC combination. The clade A Q769.h5, clade B SF162 and SS1196.1, and clade C Du156 and ZM214M.PL15 pseudoviruses were used. The starting concentrations for each neutralizing agent were as follows: 2 μg/ml (or 13.3 nM) for Q769.h5, 20 μg/ml (or 133.3 nM) for SF162, 10.54 μg/ml (or 70.3 nM) for SS1196.1, 1.08 μg/ml (or 7.2 nM) for Du156, and 1.54 μg/ml (or 10.3 nM) for ZM214M.PL15 for neutralization by VRC01p; and 0.2 μM (Q769.h5), 1 μM (SF162), 0.544 μM (SS1196.1), 0.588 μM (Du156), and 0.5 μM (ZM214M.PL15) for MVC. The effect of the combination of VRC01p and MVC against each virus was tested by dilution of the two agents starting at a fixed 1:1 ratio proportional to their IC₅₀s. The graphs, plotted in GraphPad Prism 5.0, are representative of three independent experiments. Each data point represents the mean percentage inhibition ± SEM of experiments done in triplicate.

Fig 5

HIV-1 neutralization by VRC01p, TFV, and the VRC01p-TFV combination. The clade A Q769.h5, clade B SF162 and SS1196.1, and clade C Du156 and ZM214M.PL15 pseudoviruses were used. The starting concentrations for each neutralizing agent were as follows: 2 μg/ml (or 13.3 nM) for Q769.h5, 20 μg/ml (or 133.3 nM) for SF162, 10.54 μg/ml (or 70.3 nM) for SS1196.1, 1.08 μg/ml (or 7.2 nM) for Du156, and 1.54 μg/ml (or 10.3 nM) for ZM214M.PL15 for neutralization by VRC01p; and 30 μM (Q769.h5), 100 μM (SF162), 27.2 μM (SS1196.1), 41.6 μM (Du156), and 100 μM (ZM214M.PL15) for neutralization by TFV. The effect of the combination of VRC01p and TFV against each virus was tested by dilution of the two agents starting at a fixed 1:1 ratio proportional to their IC₅₀s. The graphs, plotted in GraphPad Prism 5.0, are representative of three independent experiments. Each data point represents the mean percentage inhibition ± SEM of experiments done in triplicate.