Engineering antibody fragments that neutralize scorpion toxins and protect from the venom of Androctonus australis. 

Philippe Billiald

Muséum national d’Histoire naturelle, USM505, 57, rue Cuvier, 75231 Paris cedex 05

Poisoning by scorpion venom is a major public health hazard in many tropical regions and passive immunotherapy, which was discovered in 1890, remains the only specific treatment. No real progress have been made over this long period, and improved antivenom production, tolerance and efficiency are all required (Toxicon 2003, 41, 541-557). Antivenoms are usually heterologous polyclonal F(ab)’2 or Fab fragments prepared by proteolysis from immunized animal serum. These antibodies can not be produced in a reproducible manner and do not diffuse easily from the vascular compartment following injection. Their preparation remains tedious and their use may induce side effects involving early anaphylactic shock and late serum sickness. The replacement of polyclonal antibody fragments with monoclonal mouse or human antibodies is probably not feasible for several reasons, including instability of immortalized human cell lines and cost of production. Antibody engineering now provides the opportunity to create novel therapeutic antibody fragments with various formats and multiple valencies which are potentially effective in detoxification.

Over the last few years, we have engineered a panel of recombinant antibody fragments that actively neutralise the most potent neurotoxic polypeptides (Aah I, II and III) of the venom of the North African scorpion A. australis venom. These toxins target the potential-dependent sodium channels and are responsible for almost all the toxicity of the venom when injected to mammals. scFv 4C1 (VH and VL antibody domains connected by a short flexible peptide) was fully functional and actively neutralised toxin AahII but its production yield was low (FEBS Lett., 1999, 442, 183-188). Another scFv retained the affinity and specificity of whole IgG 9C2 that participates in the neutralisation of the toxic effect of AahI and AahIII. This scFv was then used as a building block to engineer a recombinant bivalent diabody (dimer of scFv) which has the same size as an Fab fragment but with two antigen-binding sites. This diabody is fully functional and has improved pharmacokinetic properties for therapeutic applications. It protects mice from the AahI toxin in conditions that mimic a natural envenomation with a capacity of 100 to 200 LD50 / mg (Cell. Mol. Life Sci., 2003, 60, 617-628; Toxicon 2004, 43, 233-241). Finally, undergoing studies also indicate that bispecific diabodies may be valuable tools to protect from the whole venom of the scorpion Androctonus australis, thereby opening new prospects for the treatment of envenomations.