The confirmation of the clinical diagnosis of botulism is most effectively achieved by detection of the botulinum toxin in the clinical specimens from patients (61). Blood serum collected from patients before administration of the therapeutic polyvalent antiserum, and feces are routinely tested for botulinum toxin. Constipation from botulism can be an impediment to diagnosis: an enema of sterile water may be required to obtain an adequate fecal sample. Other clinical samples that can be analyzed for botulinum neurotoxins are vomitus, gastric contents, or autopsy specimens submitted from fatal cases.
The food source is also identified by demonstration of botulinum toxin. Failure in detecting preformed toxin in the foods consumed before illness might be presumptive of toxico-infectious botulism—wound (101) or intestinal toxemia botulism (102)—which, however, must be supported by additional epidemiological and laboratory findings. The persistence of botulinum toxin in the feces of both infants and adults clearly indicates intraintestinal toxin production by the toxigenic microorganism and is consistent with intestinal toxemia botulism.
Toxicity testing is not generally performed on foods consumed by babies less than a year old who present with signs of botulism. While some of these foods might be the vehicle of spores, preformed toxin has never been detected, consistent with their composition and processing (103). Environmental samples are not usually analyzed for botulinum toxin, either. However, the demonstration of botulinum toxin in enrichment cultures from environmental, as well as from clinical and food samples not toxic before enrichment, is evidence of the presence of neurotoxigenic spores. These are also identified by testing botulinum toxin production in culture supernatants, since there are no other phenotypic characteristics unique to the species (Fig. 2). Botulinum toxin detection remains the most conclusive test to demonstrate contamination with C. botulinum spores and is fundamental in investigations of all forms of botulism.
The large variety of antigenic structures and the extreme lethality of botulinum toxins require highly versatile methods capable of detecting very low amounts of substances. Early detection of botulinum neurotoxin is necessary for the avoidance of contaminated food by the public. The conventional method for the detection of botulinum toxin involves the use of laboratory animals, which is impractical, costly, and time consuming, since up to 4 days are necessary for responses. In Europe, animal testing is restricted by law (104). Many alternative in vitro tests for botulinum toxin have been devised. Some of the newest tests are comparable in sensitivity and specificity to the standard in vivo assay. The advantages and disadvantages of these methodologies will be briefly addressed.
The mouse bioassay is universally accepted as the method of choice for the detection of botu-linum toxin because of its high specificity and sensitivity (2 LD50/mL): one mouse lethal dose is equivalent to approximately 20 pg of crystalline botulinum toxin (105). The same procedure is followed with all types of samples, except that solid samples require a preliminary extraction with gelatin phosphate buffer to solubilize the botulinum toxin. Test fluids are subjected to either centrifu-gation or membrane filtration to exclude the presence of microorganisms and consequent nonspecific deaths due to infections in mice. Diagnostic polyvalent (A-F) and monovalent botulinum antisera are used in the mouse neutralization tests. These antisera are only available from a few suppliers because limited demand makes their industrial production infeasible (3). Small volumes of the sample fluids are injected intraperitoneally (IP) into pairs of mice, either as such or mixed with polyvalent antiserum. Trypsinization has been reported to enhance the low toxicity of some samples. Mice are observed at time intervals for 4 days. Those injected with the untreated or trypsinized toxic fluids show the first symptoms of botulism (in sequence: ruffled fur, weakness of limbs, gasping for breath) within 10 hours. Death due to respiratory paralysis generally occurs 24 hours after injection. A longer time can elapse depending on the amount and type of toxin administered. Mice receiving samples mixed with botulinum polyvalent antitoxin survive. Occasionally, the polyvalent antiserum does not prevent death in mice. In these cases, one of the following hypotheses must be considered: (a) contamination of samples with lethal nonbotulinum substances; (b) excess of toxin in the sample; (c) botulinum toxin not as yet recognized. Two- or fivefold dilutions of the samples are generally sufficient to rule out the first two. The heat lability of the toxin can be demonstrated by injecting another pair of mice with samples previously heated at 100°C for 10 minutes; serum samples cannot be heated due to coagulation.
Once toxicity has been demonstrated, the mouse neutralization test with monospecific botuli-num antitoxins can be performed to identify toxin type: mice will be "protected" by the antitoxin type-specific for the botulinum toxin involved. The simultaneous presence of two different botuli-num toxins may rarely account for excessive toxicity of samples, and a mixture of two monovalent antitoxins will be necessary to achieve complete neutralization (19).
For quantitative determination of the toxin level, serial dilutions of the samples are injected intraperitoneally into groups of four mice. Survival and death are recorded after 4 days and the median lethal dose (LD50) is calculated according to Reed and Muench (106). Intravenous injections shorten the times for determination of LD50 values and require fewer animals, since a single dilution is tested and values are extrapolated from standard curves (107). Lethality of a sample may also be determined as minimum lethal dose (MLD), which is the ultimate dilution that causes death of all animals injected.
When a large number of samples has to be analyzed, as, for instance, in microbial risk evaluation studies for food quality assurance programs, a fast alternative for the detection of botulinum toxins is offered by in vitro assays. Methods currently available have recently been reviewed by Hatheway and Ferreira (105). Most of them depend on the immunogenic properties of the toxins, i.e., on their reaction to type-specific antibodies. They can be broadly divided into four categories: precipitation assays, agglutination assays, and radio- (RIA) and enzyme (mostly ELISA) immunoassays.
Precipitation assays rely on the formation of immune complexes that can be observed directly in agar gels as in immunodiffusion (108), capillary tube diffusion (109), toxic colony overlay (110) assays, or after molecular separation as in countercurrent immunoelectrophoresis (111). Although very quick and easy to perform, the minimum sensitivity reported for these assays—100 mouse IP LD50/mL—is too low for analysis of food and clinical samples. Moreover, problems of cross-reactivity between different serotypes are encountered which can cause false results.
Agglutination assays are based on the ability of red blood cells or latex particles coated with toxin or antitoxin to agglutinate with the corresponding antitoxin or toxin. In the presence of the test sample, the cells or latex particles are cross-linked and can be easily visualized (112,113). Cross-reaction between type A and B toxins, due to the use of polyclonal antibodies raised against the large toxin complexes that share nontoxic molecules, interferes with the sensitivity of the tests.
Radioimmunoassays were originally developed to increase the sensitivity of earlier tests (114). However, RIA were abandoned in favor of EIA or ELISA methods that are based on the same principles, but avoid the hazards of radioisotopes.
Several enzyme-linked immunosorbent assay variants have been designed for the detection of botulinum toxin, e.g., sandwich (115), double-sandwich (116,117), and immunoblot (118) procedures. Improvements have been achieved through production of antibodies against more purified toxins, production of monoclonal antibodies, enhanced detection systems, simplified and faster schemes, increased reproducibility, and automation.
One recent commercially available ELISA, the enzyme-linked coagulation assay (119), has a high sensitivity (<1 LD50/mL) that is achieved through amplified signals based on the cascade reaction of blood coagulation. This test entails three stages. In stage I, the botulinum neurotoxin binds to horse IgG specific for the neurotoxin; the antibody is conjugated to Russell's viper venom factor X-activating enzyme. In stage II, a mixture of coagulation factors is added: if the complex is present, thrombin will be formed. Then, alkaline phosphatase labeled fibrinogen is added and the presence of thrombin will cause the hydrolysis of fibrinogen to fibrin. In stage III, fibrin is detected by addition of a substrate for alkaline phosphatase. This assay has successfully been applied to the detection of the toxin in foods with results absolutely comparable to the mouse bioassay (120), but its applicability to patients sera samples is still to be verified.
Optical fiber-based biosensors have recently been proposed as a fast, specific, and sensitive immunoassay for the detection of botulinum neurotoxins (121). Antitoxins are immobilized on the surface of fibers to capture neurotoxin. A second fluorescently labeled antibody is added and the fluorescence emission generated upon laser excitation is detected by a photodiode. Despite the high speed of responses, its use in food analyses will depend on the cost and availability of technical equipment.
A major drawback of immunoassays is that they detect antigenicity rather than toxicity. Hence, limits of detection should be expressed in terms of toxin mass, rather than biological activity, and conversion tables for each toxin type would be convenient for this purpose.
Innovative approaches are based on the detection of the recently elucidated zinc endopeptidase activity of botulinum neurotoxins by the use of synaptic vesicles and synthetic peptides (122,123). These methods measure the biological activity of the botulinum toxin rather than its immunogenicity and are suitable for both diagnostic and pharmacological assessments.
Was this article helpful?