Humanization and characterization of an anti-ricin neutralization monoclonal antibody

Abstract

Ricin is regarded as a high terrorist risk for the public due to its high toxicity and ease of production. Currently, there is no therapeutic or vaccine available against ricin. D9, a murine monoclonal antibody developed previously in our laboratory, can strongly neutralize ricin and is therefore a good candidate for humanization. Humanization of D9 variable regions was achieved by a complementarity-determining region grafting approach. The humanized D9 (hD9) variable regions were further grafted onto human heavy and light chain constant regions to assemble the complete antibody gene. A foot-and-mouth-disease virus-derived 2A self-processing sequence was introduced between heavy and light chain DNA sequences to cleave the recombinant protein into a functional full-length antibody molecule from a single open reading frame driven by a single promoter in an adenoviral vector. After expression in mammalian cells and purification, the hD9 was demonstrated to have equimolar expression of the full-length antibody heavy and light chains. More importantly, the hD9 exhibited high affinity to ricin with K(D) of 1.63 nM, comparable to its parental murine D9 (2.55 nM). In a mouse model, intraperitoneal (i.p.) administration of hD9, at a low dose of 5 µg per mouse, 4 hours after the i.p. challenge with 5×LD50 ricin was found to rescue 100% of the mice. In addition, administered 6 hours post-challenge, hD9 could still rescue 50% of the mice. The hD9 has the potential to be used for prophylactic or therapeutic purposes against ricin poisoning.

Figure 1. Humanization of D9 Fv by CDR-grafting.

Residues are numbered according to Kabat. CDRs are marked with boxes and FRs are between boxes. Key FR residues are marked with *. Two key FR residues in D9 VH, which were kept in hD9 VH are marked with green.

Figure 2. Molecular model of D9 Fv.

FRs in VH and VL are shown in green and orange. CDRs in VH and VL are shown in red and blue.

Figure 3. VH44-N interaction with four VL residues in VH-VL interface.

VH44-N was an interchain packing residue, located in the VH-VL interface, interacting with VL87-Y, VL98-F, VL99-G, and VL 100-A. VH and VL are shown in green and orange.

Figure 4. VL82a-L interaction with VH65-D of VH CDR2.

VH82a-L was a Vernier Zone residue to support VH CDR2 by interacting with VH65-D of VH CDR2. VH and VL are shown in green and orange.

Figure 5. SDS-PAGE analysis of recombinant hD9.

Lane 1 is a molecular marker. Lanes 2 and 4 are control human IgG1 and hD9 in non-reducing condition. Lanes 3 and 5 are control human IgG1 and hD9 in reducing condition.

Figure 6. In vitro binding affinity analysis by SPR.

SPR sensorgram of the kinetics of association and dissociation of a range of concentrations from 0 to 500 nm of hD9 (A) or D9 (B) to immobilized ricin.

Figure 7. In vivo post-exposure therapeutic protection assay.

Ricin was given at the dose of 5×LD50 to mice by i.p route. hD9 or D9 at the dose of 5 µg was administered i.p. at 2, 4, or 6 hours after ricin challenge and then mouse survival rate was monitored for 7 days.

Figure 8. Half-life in mouse serum.

hD9 or D9 at the dose of 5 µg was administered by the i.p. route into mice. Sera were collected at different time points to calculate hD9 or D9 concentrations using an immunoassay. The hD9 or D9 remaining in sera is expressed as percentages plotted against time in days on the figure.