Automated Botulinum Neurotoxin Detection

Botulinum neurotoxin (BoNT) is undoubtedly the most potent biological toxin known. Intravenously injected it could paralyze and kill a human being at a dose of just 1.2 ng BoNT per kg body weight.1,2 Regardless, this dangerous substance is in high demand and promoted in the form of several medical products that utilize the toxin in safe doses and well-defined application procedures. BOTOX®, a well-known drug for smoothing facial wrinkles (Allergan, Irvine, CA), is indeed based on BoNT and has enjoyed “blockbuster drug” status in recent years.

Apart from cosmetic applications, medical uses of BoNT deal with the treatment of severe underarm sweating, neck and eyelid spasms, crossed eyes, migraine and other primary headaches, and depression. Children suffering from cerebral palsy have been treated with BoNT for spasticity, and the toxin has also been applied in cancer therapy to briefly open tumor blood vessels such that chemotherapy and radiotherapy can destroy tumor cells more efficiently.3,4 However, the widening medical use of BoNT has raised serious concerns among security experts, defense strategists, and public officials, fearing the potential abuse of BoNT for more sinister purposes. In fact, increased international trade with counterfeit BoNT drugs as well as off-label use of BoNT-containing medication or research products have been noted and are of particular concern.5


The toxin itself is a protein that is naturally produced by certain strains of spore-forming Clostridium bacteria, under exclusion of air, and is embedded into a large 900-kDa protein complex that contains several other nontoxic components. Clostridia are soil-dwelling organisms that occasionally make their way into food products. Food poisoning with BoNT is rare, but several cases occur yearly around the globe.

The most common form of human botulism, the disease caused by BoNT intoxication, occurs in newborn children. On average, 80–100 cases of infant botulism per year are documented in the United States. Young babies lack a fully developed intestinal flora. Their intestines can become colonized by ingested Clostridium botulinum spores, which results in the production and release of BoNT inside the gut. Affected babies can become paralyzed and require immediate medical attention including respiratory support and antitoxins (BabyBIG® [Botulism Immune Globulin Intravenous (Human) (BIG-IV)], California Department of Health Services). While most cases involve uncontrollable environmental sources of spores, such as dust particles, honey feeding has been identified as a common and avoidable culprit. Honey sometimes contains low numbers of Clostridium spores that, although harmless for adults, can colonize a baby’s gut.6,7 Most honey jars now carry warning labels that alert consumers not to feed honey to infants under one year of age.

Sensing the toxin

Traditionally, laboratory mice have been used to detect BoNT toxicity in clinical samples. In this “gold standard” type of test, sterile-filtered samples are injected into the peritoneal cavity of mice, which are then observed for signs of botulism. Specific antibodies can also be used to neutralize the toxin in these tests, allowing a more refined analysis that can determine which of the seven BoNT serotypes (A–G) was present. However, BoNT’s intricate mechanism of action allows for other more molecular-based assays that do not require the use of laboratory animals.8 BoNT is a 150-kDa-sized holotoxin protein consisting of a 100-kDa heavy chain, required for neuronal uptake, and a 50-kDa light chain that acts as the toxin’s payload. The latter is a zinc metalloprotease that cleaves certain neuronal proteins required for neurotransmitter release. Once cleaved, the neuron will no longer signal to the connected muscle, resulting in flaccid paralysis.9 The specific zinc metalloprotease activity of BoNT has been exploited for some very sensitive BoNT detection assays. Such assays utilize a BoNT-cleavable amino acid sequence as a substrate in the form of a synthetic peptide or fusion protein.10,11

Detecting BoNT with attomolar sensitivity

The BoNT ALISSA (Assay with a Large Immuno-Sorbent Surface Area) was previously developed in the laboratory at the Beckman Research Institute of the City of Hope, Department of Immunology (Duarte, CA). Just like other endopeptidase assays, ALISSA measures BoNT’s metalloprotease activity. However, before cleaving a specific substrate peptide, BoNT is enriched onto an immunoaffinity matrix that consists of BoNT light chain-specific antibodies bound to a beaded support. The beads are then washed from other sample components including other nonspecific protease contaminants, and reacted with the peptide substrate. The peptide contains a fluorophore and a nonfluorescent quencher moiety on either side of the BoNT cleavage site. Upon interaction with the bead-immobilized BoNT light chain, the peptide substrate is cleaved, releasing a fluorescent peptide fragment (Figure 1a).

Figure 1 – a) Schematic of the ALISSA conversion of fluorogenic substrates by bead- immobilized BoNT light chain (LC). b) Comparison of the manual ALISSA with the “gold standard” live mouse assay. Boxed numbers represent mouse LD50 units (MLD50) used for intraperitoneal injections. ΔRFU = change in relative fluorescence units. (Adapted from Ref. 12.)

Immobilization of BoNT appears to stabilize its metalloprotease activity and has a profound accelerating effect on BoNT’s enzyme kinetic properties. Maximum conversion rates were 18-fold higher for bead-immobilized BoNT than for free BoNT. Because a single immobilized BoNT molecule can convert two billion substrate molecules per hour, the ALISSA is exquisitely sensitive. In fact, half attomolar limit of detection levels (0.5 × 10–18 mol/L) have been demonstrated in toxin-spiked human blood serum, which is equivalent to the detection of 300 toxin molecules in a 1-mL sample volume. Furthermore, the BoNT ALISSA was 4–5 orders of magnitude more sensitive than the live mouse bioassay (Figure 1b).12

BoNT ALISSA on microcolumns

Although extremely sensitive, execution of the bead-based BoNT ALISSA can be a tedious task. Great care needs to be taken to not lose any beads during the stringent incubation and wash steps. In the manual ALISSA, antibody-conjugated Protein A Sepharose (GE Healthcare, Piscataway, NJ) beads were handled on disposable fritted spin columns. Such bead handling is not very amenable to automation, with the exception of magnetic bead processing, which has not led to comparable sensitivity levels.

ALISSA’s enzymatic conversion can also be performed on other types of solid support matrix. Therefore, a microcolumn-based version of the BoNT ALISSA was developed. The Mass Spectrometric Immuno Assay (MSIA) from Thermo Fisher Scientific (Waltham, MA) utilizes antibody-conjugated monolithic microcolumns inside of disposable pipet tips for the affinity enrichment of targeted antigens. The MSIA tips were originally developed for mass spectrometric readout,13,14 but can also be adapted to ELISA-like immunoassays with other means of detection, such as optical readers.

Robotizing the BoNT ALISSA

Toxin enrichment on microcolumns is best performed through repeated liquid sample aspiration and dispense steps that bring toxin molecules in close contact with specific antibodies conjugated to the MSIA columns. The same requirement applies to subsequent wash steps, enzyme conditioning, and the BoNT-specific enzymatic reaction. Because such repeated pipetting would guarantee a bad case of thumb arthritis for a lab technician, the procedure is best suited for a robotic pipetting system.

The Versette Pipetting Workstation (Thermo Fisher Scientific) was used successfully to conduct the entire ALISSA procedure within a few hours. The system is compact and occupies only a small 27 × 22-in. footprint on the lab bench, at a height of 26½’’, plus a computer. The Versette is equipped with a 96-tip pipetting head that can be configured freely by using either all or only the required number of tip positions. ALISSA includes standard curves and multiple sample replicates; hence 48 tips per run are typically required. The Versette is designed with a split, two-layered stage table that moves horizontally in the x- and y-directions while the pipetting head moves only vertically on the z-axis. The 96 syringe pistons are individually sealed and operate simultaneously inside the head compartment. The setup provides six reagent positions for bulk volume container troughs, deep-, or regular-well microtiter plates (Figure 2).

Figure 2 – Configuration of the Versette Pipetting Workstation for the BoNT ALISSA. A typical ALISSA BoNT bot run consisted of prerun conditioning of BoNT-specific MSIA columns with phosphate-buffered saline, followed by a 2.5-hr incubation that performs 1000 aspirate/dispense cycles with BoNT-containing serum samples, six washes with an immunoprecipitation (IP) wash buffer, six washes with a conditioning buffer, 1000 cycles with the reactive fluorogenic peptides, followed by aspiration of eluent and slow elution back into the MTP with the fluorogenic peptides. At the completion of the run, an automated e-mail was sent by the Versette control software, Control Mate, to alert the user. The plate was read on a VICTOR™ 3 Multilabel Plate Reader (PerkinElmer, Shelton, CT) in fluorimeter mode.

To perform a completely automatic, hands-off ALISSA run, the technicians placed BoNT-containing serum samples into a 1.1-mL/well 96-well polypropylene microtiter plate (MTP), immunoprecipitation wash buffers and conditioning buffers into separate rows on a 2.0-mL/well plate, and the fluorogenic peptide substrates into a black well MTP sealed with a sticky aluminum foil seal. The seal’s purpose was to protect the fluorogenic peptides from light and evaporation. It can easily be pierced by the MSIA tips on the Versette. The eluent, 50% DMSO with sodium carbonate solution (0.2 M), was provided in a 125-mL trough container. It was used to elute remaining fluorescent product from the MSIA columns after the enzymatic reaction had been completed. All reagents and buffer compositions, except the eluent, are essentially the same as previously described.12,15 The lower deck housed a 250-mL polypropylene container as a waste collector and a second one served as a “tip tipper.” The tip tipper consisted of a piece of Whatman blotting paper (GE Healthcare) that was held inside the container on top of its vertical splash guards. Its purpose was to remove foamy air bubbles that form on the tips when dispensing detergent-containing wash buffers, followed by a column-drying air gap, into the waste collector. Thereby, the tip tipper prevents cross-contamination between individual samples.

Results: Proof-of-concept

A dilution series of BoNT light chain A (LCa) in human serum demonstrates the performance of the assay (Figure 3). Sub-femtomolar BoNT concentrations can be detected in 0.9-mL samples after ~8 hr run time.

Figure 3 – ALISSA BoNT bot data obtained with BoNT light chain (LC) type A spiked into pooled human serum. TPEN = tetrakis-(2-pyridylmethyl)ethylenediamine (TPEN, 100 μM) is a potent zinc metalloprotease inhibitor and was added in control runs to validate BoNT-produced signals. Each bar represents data from three ALISSA experiments.

Conclusion and outlook

The robotized ALISSA has been performing repeatedly in a robust and reliable fashion. Future improvements are expected to reduce the overall run time by optimizing pipetting cycles, while retaining the method’s high level of sensitivity. The ALISSA method can be seen as a platform technology that is extendable to the detection of other enzymes, provided that suitable antibody and substrate combinations are available. Examples of previous (manual) ALISSA adaptations include detection of other BoNT serotypes and the zinc metalloprotease anthrax lethal factor.15


  1. Gill, D.M. Bacterial toxins: a table of lethal amounts. Microbiol. Rev. 1982, 46(1), 86–94.
  2. Arnon, S.S.; Schechter, R. et al. Botulinum toxin as a biological weapon: medical and public health management. Jama  2001, 285(8),1059–70.
  3. Cooper, G. Therapeutic Uses of Botulinum Toxin; Humana: Totowa, NJ, 2007; pp xiv; 238 pp.
  4. Ansiaux, R.; Gallez, B. Use of botulinum toxins in cancer therapy. Exp.Opin. Invest. Drugs  2007, 16(2), 209–18.
  5. Coleman, K.; Zilinskas, R.A. Fake botox, real threat. Sci. Amer. 2010, 302(6), 84–9.
  6. Brook, I. Infant botulism. J. Perinatol. 2007, 27(3), 175–80.
  7. Koepke, R.; Sobel, J. et al. Global occurrence of infant botulism, 1976–2006. Pediatrics  2008, 122(1), e73–82.
  8. Čapek, P.; Dickerson, T. Sensing the deadliest toxin: technologies for botulinum neurotoxin detection. Toxins 2010, 2(1), 242010,53.
  9. Pellizzari, R.; Rossetto, O. et al. Tetanus and botulinum neurotoxins: mechanism of action and therapeutic uses. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences  1999, 354(1381), 259–68.
  10. Ruge, D.R.; Dunning, F.M. et al. Detection of six serotypes of botulinum neurotoxin using fluorogenic reporters. Anal. Biochem. 2011, 411(2), 200–9.
  11. Wang, D.; Baudys, J. et al. Improved detection of botulinum neurotoxin serotype A by Endopep-MS through peptide substrate modification. Anal. Biochem. 2013, 432(2), 115–23.
  12. Bagramyan, K.; Barash, J.R. et al. Attomolar detection of botulinum toxin type A in complex biological matrices. PLoS ONE 2008, 3(4), e2041.
  13. Nedelkov, D,; Kiernan, U.A. et al. Volumetric mass spectrometry protein arrays. Meth. Molec. Biol. 2007, 382, 333–43.
  14. Trenchevska, O.; Kamcheva, E. et al. Mass spectrometric immunoassay for quantitative determination of protein biomarker isoforms. J. Proteome Res. 2010, 9(11), 5969–73.
  15. Bagramyan, K.; Kalkum, M. Ultrasensitive detection of botulinum neurotoxins and anthrax lethal factor in biological samples by ALISSA. Meth. Molec. Biol.  2011, 739, 23–36.

Karine Bagramyan, Ph.D., is Staff Scientist; Aneela Reddy, B.S., is Research Associate; and Markus Kalkum, Ph.D., is Associate Professor, Beckman Research Institute of the City of Hope, Department of Immunology, 1500 E. Duarte Rd., Duarte, CA 91010, U.S.A.; tel.: 626-301-8301; e-mail:; Dobrin Nedelkov, Ph.D., is R&D Manager/Site Leader at Thermo Fisher Scientific; Tempe, AZ, U.S.A. The authors thank Drs. James Marks and Jianlong Lou from the University of California, San Francisco, for providing monoclonal anti-BoNT/A LC antibodies. This work was supported by the National Institutes of Health through grants AI096169-01 and AI065359-05.