Llama single domain antibodies specific for the 7 botulinum neurotoxin serotypes as heptaplex immunoreagents

Abstract

Background: There are currently 7 known serotypes of botulinum neurotoxin (BoNT) classified upon non-cross reactivity of neutralizing immunoglobulins. Non-neutralizing immunoglobulins, however, can exhibit cross-reactivities between 2 or more serotypes, particularly mosaic forms, which can hamper the development of highly specific immunoassays, especially if based on polyclonal antisera. Here we employ facile recombinant antibody technology to subtractively select ligands to each of the 7 BoNT serotypes, resulting in populations with very high specificity for their intended serotype.

Methods and findings: A single llama was immunized with a cocktail of 7 BoNT toxoids to generate a phage display library of single domain antibodies (sdAb, VHH or nanobodies) which were selected on live toxins. Resulting sdAb were capable of detecting both toxin and toxin complex with the best combinations able to detect 100s-10s of pg per 50 microL sample in a liquid bead array. The most sensitive sdAb were combined in a heptaplex assay to identify each of the BoNT serotypes in buffer and milk and to a lesser extent in carrot juice, orange juice and cola. Several anti-A(1) sdAb recognized A2 complex, showing that subtype cross-reactivity within a serotype was evident. Many of our sdAb could act as both captor and tracer for several toxin and toxin complexes suggesting sdAb can be used as architectural probes to indicate BoNT oligomerisation. Six of 14 anti-A clones exhibited inhibition of SNAP-25 cleavage in the neuro-2A assay indicating some sdAb had toxin neutralizing capabilities. Many sdAb were also shown to be refoldable after exposure to high temperatures in contrast to polyclonal antisera, as monitored by circular dichroism.

Conclusions: Our panel of molecularly flexible antibodies should not only serve as a good starting point for ruggedizing assays and inhibitors, but enable the intricate architectures of BoNT toxins and complexes to be probed more extensively.

Figure 1. Seroconversion of Snoop the llama after being immunized with the 7 serotypes of BoNT toxoids.

Antibody capture from serum after the 6^th immunization was monitored on the seven serotypes of a) toxoid, b) toxin and c) toxin complexes versus a control antigen, bovine serum albumin (BSA).

Figure 2. Capturing the llama anti-BoNT repertoire by phage display.

Polyclonal phage ELISA of round 2 selected phage populations analyzed on each a) toxin or b) toxin complex indicates the potential degree of specificity for the serotype of toxin upon which the population was selected.

Figure 3. Deducing the sensitivity of selected anti-BoNT llama single domain antibodies.

Lower limits of detection of the best antibody pairs on cognate toxin and toxin complex for each serotype a) A, b) B, c) C, d) D, e) E, f) F, and g) G. To provide a non-specific background value for each plot, the mMFI of the pairs employed on 1e+5 pg/well (i.e. 10x the top concentration used in this titration) of non-cognate serotypes are provided as follows: A toxin, 2.3; A complex, 2.0; B toxin, 1.3; B complex, 2.0; C toxin, 8.0; C complex, 3.3; D toxin, 32.3; D complex, 4; E toxin, 2.8; E complex; 3.5; F toxin, 4.5; F complex, 7.5; G toxin, 2.5; G complex, 2.0.

Figure 4. Heptaplex assays using selected pairs of anti-BoNT captors and tracers in various food matrices.

Assays were challenged with 1e+4 pg per well of each of the toxin or toxin complexes in a) PBSTB, b) 2% reduced fat milk, c) orange juice d) carrot juice and e) cola.

Figure 5. In vitro tissue culture assay to discover potential inhibitory activities of our sdAb clones.

Neuro-2A assay demonstrating the ability of some of the anti-A BoNT sdAb to inhibit the process of BoNT intoxication by reducing the intracellular cleavage of SNAP-25. + indicates toxin without sdAb and – indicates no toxin and no sdAb.