Botulinum neurotoxins (BoNTs) are the deadliest substances known, with intravenous human lethal doses estimated at 1-3 ng/kg1,2. Seven structurally similar serotypes of BoNT, labeled A through G, exist, each consisting of a heavy chain domain responsible for cell binding, uptake, and translocation into the cytosol and a light chain that encodes a zinc endopeptidase3-5. The exquisite toxicity of BoNT results from, in part, its specific binding and entry into motor neurons at the neuromuscular junction6. Once inside the neuron, the light chain endopeptidase specifically cleaves one or more of the soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) proteins required for vesicle fusion, inhibiting neurotransmitter release and leading to flaccid paralysis7-14. Commonly known as the disease "botulism," paralysis of the diaphragm and intercostal muscles by BoNT ultimately results in respiratory failure and death unless early diagnosis and treatment are received.
Human foodborne botulism is most commonly associated with BoNT serotypes A, B, E, and F (BoNT/A, BoNT/B, etc.) and usually results from the ingestion of contaminated food15,16; although, several cases of wound botulism were reported amongst intravenous drug users17,18. In the United States, infant botulism resulting from the ingestion of Clostridium spores by children under the age of one is the most common form of botulism19-21. However, foodborne BoNT outbreaks resulting from improper home canning and food processing were reported in both the United States and abroad. Between 2000-2009, at least 338 cases of foodborne botulism were reported worldwide including six fatalities22. The ability to rapidly and sensitively detect foodborne botulism outbreaks is a critical indication that could aid early diagnosis23,24. Furthermore, detection methods that allow cost-effective and routine food testing will lead to improved food security.
BoNT's neuronal specificity and long biological half-life also makes it a potent therapeutic. In the United States, BoNT-based drugs are approved by the Food and Drug Administration for the treatment of cosmetic conditions and neuromuscular-related disorders including glabellar lines, cervical dystonia, migraine headaches, overactive bladder, and strabismus. Numerous "off-label" applications are documented, including high-dose treatments for severe muscle dysfunction25-28. Accurate toxin quantification is critical for correct dosing, as underdosing may lead to ineffective treatment while overdosing puts patients at risk of potentially harmful side effects. Unfortunately, no standardized potency assay protocol is shared across manufacturers, resulting in unit definition inequalities between BoNT-based drug products29-31.
The standard test for BoNT is the mouse bioassay in which BoNT-containing samples are injected intraperitoneally into mice and the numbers of deaths recorded over 1-7 days16,32,33. The mouse bioassay is very sensitive with limits of detection (LOD) of 5-10 pg BoNT/A34; however, ethical concerns over animal use, the high cost of training personnel and maintaining animal facilities, long assay times, and the lack of standardized protocols resulted in calls to develop standardized, animal-free BoNT testing and quantification methods35-39. Recently, several alternate BoNT quantification methods were developed that offer mouse or near-mouse bioassay sensitivity40-49. These methods commonly use fluorescence, mass-spectrometry, or immunological methods and offer assay times considerably shorter than the mouse bioassay without animal use. Mass-spectrometry approaches combined with immunological techniques were shown to detect and quantify BoNT contained in food and other complex samples; however, personnel training requirements and specialized equipment limit these assays50-55. Most other alternate assays are not readily applicable to complex sample testing or lack the throughput required for routine BoNT testing. The highly variable nature of food sample viscosity, pH, salt content, and matrix constituents presents an especially difficult challenge when trying to develop in vitro assay methods with sensitivity to match the extreme potency of BoNT. Furthermore, even simple and relatively benign buffer systems, such as those resulting from resuspension of BoNT-based drug products, contain salt, albumin, and sugar stabilizers (i.e. excipients) that significantly impact in vitro BoNT potency56. Toxin purification is required for accurate activity testing of all but the simplest of samples56-59.
The BoTest Matrix assays were designed for rapid, high-throughput, and consistent quantification of BoNT from highly complex samples using equipment commonly found in research laboratories56,60. These assays use paramagnetic beads covalently linked to serotype-specific anti-BoNT antibodies to bind and sequester BoNT out of a sample and then remove interfering matrix compounds by washing. Following washing, bound BoNT proteolytic activity is then quantified in an optimized reaction buffer using a reporter compatible with the BoNT serotype being tested. These reporters are fluorogenic proteins consisting of a N-terminal cyan fluorescent protein (CFP) moiety and a C-terminal yellow fluorescent protein derivative (Venus) moiety linked by a BoNT substrate, SNAP25 residues 141-206 or synaptobrevin residues 33-94 constituting the BoTest A/E or B/D/F/G reporters, respectively45. Reporter cleavage by BoNT is monitored using Förster resonance energy transfer (FRET). When the reporter is intact, excitation of CFP results in FRET to Venus, quenching CFP emission and exciting Venus emission. Cleavage of the reporter by BoNT prevents FRET, leading to an increase in CFP emission and decrease in Venus emission. BoNT activity can then be quantitatively measured using the ratio of the CFP and Venus emissions. LOD below 3 pg are possible from a wide range of foods using a high-throughput 96-well plate format56. Increased sensitivity can be obtained using larger sample volumes since the assay allows concentration of the toxin on the bead surface.
The BoTest Matrix assays for BoNTs A, B, E, and F were developed and tested with food, pharmaceutical, and environmental samples56,60. Here, we describe procedures for executing these assays for the detection of BoNT in low complexity (e.g. pharmaceutical, BoNT in buffer) and high complexity (e.g. food, environmental) samples. Specific processing methods for several sample types are addressed in this protocol and sample types not described here can usually be adapted using a combination of the presented methods. The protocol was developed and tested with BoNT/A but is adaptable to other BoNT serotypes using their respective assays as demonstrated elsewhere56,60.