US20020058294A1

US20020058294A1 - Method for detecting clostridium botulinum neurotoxin serotypes A, B, E and F in a sample

Info

Publication number: US20020058294A1
Application number: US09/952,078
Authority: US
Inventors: Mark Poli; Victor Rivers
Original assignee: Individual
Current assignee: Individual
Priority date: 2000-09-15
Filing date: 2001-09-14
Publication date: 2002-05-16
Also published as: AU2001290865A1; WO2002023195A3; US6678651B2; CA2452023A1; US20030004710A1; WO2002023195A2

Abstract

Sensitive and specific enzyme-linked immunosorbent assays which detect Clostridium botulinum neurotoxins serotypes A, B, E, and F in a sample are described. The assay is based upon affinity-purified antibodies directed against the C-fragments of each toxin. These assays demonstrate sensitivity close to that of the mouse bioassay without the use of animals and in a much simpler format than other assays of similar sensitivity.

Description

BACKGROUND OF THE INVENTION

The present invention relates to a sensitive and specific enzyme-linked immunosorbent assay for the detection of Clostridium botulinum neurotoxin serotypes A (BoNT A), B (BoNT B), E (BoNT E) and F (BoNT F) in a sample.

The anaerobic bacterium Clostridium botulinum produces seven serologically different toxins. Recognized as the most potent toxins of biological origin, botulinum neurtoxins (BoNTs) are the causative agents of food-borne, infant and wound botulism (Sakaguchi, 1983, Pharmac. Ther. 19, 165-194). The toxins act presynaptically at the neural junction by blocking the release of acetylcholine and thereby causing a flaccid muscular paralysis (Simpson, 1986, Ann. Rev. Pharmacol. Toxicol. 26, 427-453). Paralysis of the respiratory musculature can cause death in untreated patients. All serotypes (MW approximately 150 kDa) consist of two polypeptide subunits joined by an intra chain disulfid bridge and are bound to nontoxic neurotoxin-associated proteins (NAP's) which protect the toxins in the gastrointestinal tract. The mechanism of action is similar for each serotype. The heavy chain (B chain) is the binding subunit, which binds to a receptor on the presynaptic membrane. The light chain (A chain) is the catalytic subunit. Once translocated across the cell membrane, its zinc-dependent protease activity hydrolyzes specific proteins associated with synaptic vesicle docking and acetylcholine release (Schiavo et al., 1994, Sem. Cell. Biol. 5, 221-229).

Currently, the mouse bioassay is the most widely accepted method for detecting BoNT's in serum and foods. This assay has the desired sensitivity (<5 mouse lethal units/mL), but it is cumbersome, time consuming (1-4 days) and involves the use of large numbers of animals (Shone et al., 1985, Appl. Environ. Microbiol. 50, 637-667). Enzyme-linked immunosorbent assays (ELISA's) have been reported by several laboratories (Dezfulian and Bartlett, 1984, J. Clin. Microbiol. 19, 645-648; Shone et al., 1985, supra), but lack the required sensitivit. An enzyme-linked coagulation assay (ELCA) was reported with a sensitivity comparable to the mouse bioassay (Doellgast et al., 1994, J. Clin. Microbiol. 32, 851-853), but this assay relies upon a sophisticated amplification system utilizing a snake venom coagulation factor and is limited by it's complexity and the expense of the reagents.

Therefore, there is a need for a simple, rapid sensitive, and accurate assay for the measurement of BoNT's in samples without the use of animals or complicated and expensive reagents.

SUMMARY OF THE INVENTION

The present invention satisfies the need discussed above. The present invention relates to a simple, sensitive colorimetric capture ELISA for BoNTs with detection limits at or below 1 mouse unit. The assay is reproducible and accurate with negligible cross-reactivity between serotypes. The strength of the assay relies on its novel format and the unique preparation of the antibodies used in the assay. The antibodies are affinity-purified to the heavy chain C-fragment of the toxin. Others have used antibodies which are not affinity purified or which are purified to the whole toxin molecule. We reasoned that since the C-terminal region of the heavy chain is where the binding domain is located, this portion of the molecule should not be covered by associated proteins; if the binding domain was blocked, then the molecule would be precluded from binding to the cell surface and would not be toxic. Thus, the binding region “looks” the same in both the purified and complexed forms. Antibodies to this region should recognize both forms of the toxin. The result of the unique preparation of the antibodies is that they do not cross-react between serotypes, they recognize neutralizing epitopes, and they recognize purified and complexed toxins equally.

BRIEF DESCRIPTION OF THE DRAWINGS

These and other features, aspects, and advantages of the present invention will become better understood with reference to the following description and appended claims, and accompanying drawings where: [0006]
FIGS. 1A and 1B. Colorimetric ELISAs for BoNT serotypes A and B. Assays were performed as described in Materials and methods in assay buffer (open symbols) or 10% human serum (solid symbols). Points are means of triplicate determinations. Where not visible, error bars are within the symbol. [0007]
FIGS. 2A and 2B. Colorimetric ELISAs for BoNT serotypes E and F. Assays were performed as described in Materials and Methods in assay buffer (open symbols) or 10% human serum (solid symbols). Points represent means of triplicate determinations. Where not visible, error bars are within the symbol. [0008]
FIGS. 3A and 3B. Comparison of ELISA detection of purified neurotoxin (solid symbols) and toxin complexed with associated nontoxic proteins (open symbols) for BoNT A and B. Assays were performed as described in Materials and Methods. Points are means of triplicate determinations. Where not visible, error bars are within the symbols. [0009]
FIGS. 4A and 4B. Comparison of ELISA detection of purified neurotoxin BoNT E and F (open symbols) and toxin complexed with associated nontoxic proteins (solid symbols). Assays were performed as described in Materials and Methods. Points represent means of triplicate determinations. Where not visible, error bars are within the symbols. [0010]
FIG. 5. Comparison of trypsinized and non-trypsinized BoNT E. Purified BoNT E, either trypsinized (solid squares) or non-trypsinized (open squares), and trypsinized BoNT E complex (solid circles) were analyzed by ELISA according to Materials and Methods. Trypsinized (activated) complex and trypsinized purified toxin were used at equal mouse unit concentrations; non-trypsinized BoNT E was used at used at equal mass to trypsinized BoNT E. Trypsinizing purified toxin destroys antigenic determinants, while NAPs significantly protect the complexed toxin. Points represent means of triplicate determinations.[0011]

DETAILED DESCRIPTION OF THE INVENTION

A capture ELISA method comprises the use of two (monoclonal or polyclonal) antibodies to the same antigen with two different epitopes, one of which is conjugated with biotin. The antigen containing supernatant is reacted with the first antibody and washed with a buffer solution. The antibody linked antigen is then reacted with the second biotinylated antibody and then washed to remove the excess antibody. The antibody-antigen-antibody/biotin is then cross-linked with avidin optionally linked to a chromogenic enzyme and washed. Finally, a substrate is reacted with the avidin or the chromogenic enzyme and the development of color product is measured. [0012]
In one embodiment, the present invention provides a method for detecting BoNT comprising: [0013]
(a) reacting a sample suspected of having an antigen with a first antibody, Ab(A); [0014]
(b) washing the reaction mixture to remove unreacted antibody; [0015]
(c) reacting the Ab(A) linked antigen with a second antibody which is conjugated with biotin, Ab(B); [0016]
(d) washing the reaction mixture to remove unreacted antibody; [0017]
(e) reacting Ab(A)-antigen-Ab(B) with avidin linked to a chromogenic enzyme; [0018]
(f) washing the reaction mixture; [0019]
(g) reacting the chromogenic enzyme; [0020]
(h) measuring optical density of reaction and calculating concentration of antigen using a standard curve. [0021]
A sample includes any solution which is suspected of containing BoNT such as animal serum including human serum, food samples, and dirt samples. [0022]
The antibody against BoNT can be of any isotype, such as IgA, IgG or IgM, Fab fragments, or the like. The antibody may be a monoclonal or polyclonal and produced by methods as generaly described in Harlow and Lane, [0023] Antibodies, A Laboratory Manual, Cold Spring Harbor Laboratory, 1988, incorporated herein by reference. Any monoclonal which has the desired specificity and affinity can potentially be used once the conditions for use of the antibody are tested. Preferably, polyclonal antibodies are used since polyclonal antibodies can recognize different epitopes of BoNT protein thereby enhancing the sensitivity of the assay. Polyclonal antibodies for use in the assay can be any polyclonal available commercially with the requisite affinity and specificity. It is preferable that the antibodies are affinity purified to the C-fragment of BoNT as described in the examples below so that the antibody may recognize purified and complexed toxin equally. The antibodies used to measure BoNTs in the examples below were serotype specific polyclonal antibodies purified from hyperimmune horse serum. Any serum from any other source could potentially work as long as the affinity of the antibody to the antigen is high enough to result in detection, i.e. the detection limit of the assay will be dictated by the affinity of the antibody. Most preferably, the antibody is affinity-purified against the C-fragment of BoNT. Antibodies were affinity purified over a column with heavy-chain C-fragment of specific serotypes to produce Ab(A). The affinity purified antibodies were labeled with biotin to produce biotinylated antibody Ab(B). Any other label can be used which provides high affinity between the label and the detection molecule. Care must be taken that the coupling of the BoNT to the label will not inactivate the enzyme. Also, it is important that the background be low since high background would adversly effect the detection limit.
The antibody can be applied to the solid support by direct or indirect means. Indirect bonding allows maximum exposure of the toxin binding sites to the assay solutions since the sites are not themselves used for binding to the support. The solid support can be a any phase used in performing immunoassays, including dipsticks, membranes, absorptive pads, beads, microtiter wells, test tubes, and the like. Preferred are microtiter plates which produce low background noise and can bind a high amount of capture antibody in the wells. These factors affect signal-to-noise ratios and hence assay performance. In addition, most plate readers are set up for 96-well plates as described in Materials and Methods below but other plates can be used if a washer/reader is adjusted or designed for it. If a plate is used, the anti-BoNT antibody is bound to the plate using directions from the manufacturer or any other method known to the investigator. [0024]
The solid support is preferably non-specifically blocked after binding the BoNT antibodies to the solid support. Non-specific blocking of surrounding areas can be with Superblock™ blocking reagent from Pierce (Rockford, Ill.), whole or derivatized bovine serum albumin, or albumin from other animals, whole animal serum, casein, non-fat milk, and the like. [0025]
The sample is applied onto the solid support with bound BoNT-specific antibody such that the BoNT protein will be bound to the solid support through said antibodies. Excess and unbound components of the sample are removed and the solid support is preferably washed so the antibody-antigen complexes are retained on the solid support. The solid support may be washed with a washing solution which may contain a detergent such as Tween-20, Tween-80, or any other washing solution as long as it does not denature proteins and inferfere with protein/protein interactions. [0026]
After the BoNT has been allowed to bind to the solid support, a second antibody which reacts with BoNT is applied. The second antibody may be labeled, preferably with a visible label. The labels may be soluble or particulate and may include dyed immunoglobulin binding substances, simple dyes or dye polymers, dyed latex beads, dye-containing liposomes, dyed cells or organisms, or metallic, organic, inorganic, or dye solids. The labels may be bound to the BoNT antibodies by a variety of means that are well known in the art. In some embodiments of the present invention, the lables may be enzymes that can be coupled to a signal producing system. Examples of visible labels include alkaline phosphatase, beta-galactosidase, horseradish peroxidase, and biotin. The steptavidin-phosphatase could be substituted by any streptavidin or avidin-linked indicator system, for example streptavidin-peroxide, Europium, or chemilunescence and its respective developing substrate or indicator, generically referred to as substrate. Many enzyme-chromogen or enzyme-substrate-chromogen combinations are known and used for enzyme-linked assays. [0027]
Preferably, the affinity purified BoNT antibody is used as a second antibody to bind the BoNT-antibody Ab(A) complex. The second antibody is preferably labeled with biotin producing a biotinylated BoNT antibody, Ab(B). After binding Ab(B) to the Ab(A)-BoNT complex, the plates are washed and neutravidin-linked alkaline phosphatase is added. After binding of the neutravidin to the Ab(B)-BoNT-Ab(A) complex, enzyme substrate p-nitrophenyl phosphate was added and the color allowed to develop. [0028]
Accumulated label or color may be detected by optical detection devices such as reflectance analyzers, video image analyzers and the like. The visible intensity of accumulated label could correlate with the concentration of BoNT in the sample. The correlation between the visible intensity of accumulated label and the amount of BoNT may be made by comparison of the visible intensity to a set of reference standards. Preferably, the standards have been assayed in the same way as the unknown sample, and more preferably alongside the sample, either on the same or on a different solid support. The concentration of standards to be used can range from about 0.1 ng of BoNT per ml of solution, up to about 10 mg of BoNT per ml of solution Toxicon 34(9), 975-985). Preferably, several different concentrations of BoNT are used so that quantitating the unknown by comparison of intensity of color is more accurate. Additionally, BoNT complexed with nontoxic-neurotoxin associated proteins can be used to establish a standard curve. If one is trying to detect complexed toxin, one should use complexed toxin in the standard curve. This gives the best accuracy. Naturally-occurring BoNT intoxications are all caused by complexed toxin. The fact that this assay recongizes both forms the free and complexed forms of the toxin is one of its strengths. [0029]
As evident to a person with ordinary skill in the art, it may be necessary to undergo one or more serial dilutions of the sample such that the level of BoNT in the sample can be compared to one of the set standards. The BoNT measurement is then corrected for the dilution factor. [0030]
All the materials and reagents required for assaying BoNT according to the present invention can be assembled together in a kit. This generally will comprise one or more solutions containing a known concentration of BoNT, a washing solution, a solution of a chromogen which changes color or shade by the action of the enzyme directly or indirectly through action on a substrate, a anti-BoNT conjugated to a label such that it could be detected and an anti-BoNT antibody without a label, pipettes for the transfer of said solutions, test tubes for said solutions, and a solid support carrying on the surface thereof a polyclonal antibody to BoNT. The kit may also contain one or more solid support having an anti-BoNT antibody for use in assaying one or more samples simultaneously or individually, and the necessary reagent required to develop the label. It is also preferable that the BoNT used for standards, whether free or complexed with a substrayte, be provided so that it could be assayed fresh along with the unknown sample. Such kits will comprise distinct containers for each individual reagent. The reagents may be supplied from storage bottles or one or more of the test tubes may be prefilled with the reagents or controls. [0031]
The components of the kit may also be provided in dried or lyophilized forms. When reagents or components are provided as a dried from, reconstitution generally is by the addition of a suitable solvent. It is envisioned that the solvent also may be provided in another container means. [0032]
In any of the above test kits, all solutions and antibodies for detection of one BoNT can be provided in one kit. Alternatively, solutions and antibodies for more than one serotype can be provided in one kit. In addition, a test plate can be prepared such that different parts of the plate can be used to detect a different serotype of BoNT. [0033]
The kits of the present invention also will typically include a means for containing the reagents such as vials or tubes in close confinement for commercial sale such as, e.g. injection or blow-molded plastic containers into which the desired vials are retained. [0034]
The following examples are included to demonstrate an embodiment of the invention. It should be appreciated by those of skill in the art that in light of the present disclosure, many changes can be made in the specific embodiment disclosed and still obtain a like or similar result without departing from the spirit and scope of the invention. [0035]
The following MATERIALS AND METHODS were used in the examples that follow. [0036]
Toxins [0037]
Purified [0038] Clostridium botulinum neurotoxin serotypes A and B (1.6×10⁸and 7.0×10⁷mouse i.p. LD₅₀'s/mg, respectively) and neurotoxins A and B complexed with NAPs (4.5×10⁷and 1.4×10⁷mouse i.p. LD₅₀'s/mg, respectively) were purchased from the University of Wisconsin Food Research Institute (Madison, Wis.). Purified C. botulinum neurotoxin serotypes E and F (4.5×10⁷and 8.0×10⁶mouse i.p. LD₅'s/mg, respectively) and neurotoxins E and F complexed with NAPs (2×10⁶and 4×10⁶) mouse i.p. LD₅₀'s/mg, respectively) were purchased from METAbiologics, Inc, Madison, Wis. Stock solutions (10 μg/mL), were kept at 40° C. in sterile buffer [50 mM sodium acetate pH 4.2, 2% gelatin, and 3% bovine serum albumin (BSA)]. Working dilutions were prepared immediately before use. Heavy-chain C-fragments were purchased from Ophidian Pharmaceuticals (Madison, Wis.) and stored at 40° C. before use.
Preparation of affinity columns [0039]
Cyanogen bromide-activated Sepharose 4B (Pharmacia, Uppsala, Sweden) was hydrated in 1 mM HCl for 4 hr at room temperature with gentle stirring, and then washed with coupling buffer (0.1 M NaHCO[0040] ₃, 0.5 M NaCl, pH 8.3). Heavy-chain C-fragment (10 mg) dissolved in 5 mL Coupling buffer was added to 3 g of washed beads and slowly rotated for 5 hr at room temperature. After centrifugation for 15 min at 3,000×g, the pellet was resuspended in 15 mL of blocking buffer (0.1 M NaHCO₃, 0.2 M glycine, pH 8.0) and gently rotated overnight at 40° C. The coupled gel was then poured into a column and the beads were allowed to settle. After washing with three column volumes of wash buffer (0.1 M NaC₂H₃O₂, 0.5 M NaCl, pH 4.0) followed by three column volumes of coupling buffer, the column was equilibrated in phosphate-buffered saline (PBS) containing 0.9% benzoyl alcohol as a bacteriostatic agent and refrigerated.
Hyperimmunization of horses: Animals were immunized with at least three doses of monovalent botulinum toxoid, one serotype (A,B,C,D,E,F, or G) per horse, and subsequently boosted with botulinum toxin of the same serotype. Alum precipitated toxoid in Ribi adjuvant or toxin in saline was administered subcutaneously in multiple sites in the neck and flank to increase absorption of antigenic materials, and to minimize the potential for focal inflammation (18-21 ga needle). Only the first injection of toxoid was administered with an adjuvant. Subsequent boosters consisted of toxoid or botulinum toxin only. Injections of toxoid or, toxin were given within six inches of a potential site of blood withdrawl. Horses were confined to the USAMRIID Large Animal Research Facility (LARF) barn or an immediately adjacent paddock for at least one hour after administration of toxoid or toxin for observation for adverse reactions. An anaphylaxis kit was immediately available and horses were monitored during the observation period. When the animals are plasmapheresed, they were kept in the LARF barn overnight to facilitate observation during the post bleeding period. [0041]
The immunization schedule was one of the following for each of the animals: [0042]
Immunization Schedule (first six horses) [0043]
Week 0: 2.0 mg of toxoided neurotoxin *—and one dose of tetanus toxoid (three horses will receive Complete Freunds Adjuvant (CFA) and three additional horses receiving Ribi adjuvant with botulinum toxoid—see adjuvant, IV.C.5. Only the primary injection will contain an adjuvant) [0044]
Day 4: Bleed for lymphocyte stimulation assay (10 ml) [0045]
Week 2: 0.5 mg of toxoided neurotoxin ** [0046]
Week 4: Bleed for titer (10 ml) [0047]
Week 5: 0.5 mg of toxioded neurotoxin ** [0048]
[0049] Week 5+3, 6, & 9 days: Bleed for titer & IgG/IgM ratio (10 ml)
[0050] Week 5+12 days: Bleed for IgG/IgM ratio and titer
[0051] Week 5+16 days: Begin plasmapheresis every 3 days for 12 bleeds
(8-12 liters per bleed). [0052]
After 12 plasmaphereses, rest for 1 week and boost, entering the schedule again at week 5 (52 day cycle). [0053]
Immunization Schedule: (sixty nine horses) [0054]
Week 0: 2.0 mg of toxoided neurotoxin * [0055]
Week 2: 0.5 mg of toxoided neurotoxin ** [0056]
Week 4: Bleed for titer (10 ml) [0057]
Week 5: 0.5 mg of toxioded neurotoxin ** [0058]
[0059] Week 5+12 days: Bleed for IgG/IgM ratio titer
[0060] Week 5+16 days: Begin plasmapheresis every 3 days for 12 bleeds
(8-12 liters per bleed) [0061]
After 12 plasmaphereses, rest for 1 week and boost (52 day cycle) [0062]
Animals were bled (10 ml) and their plasma examined for anti-botulinum antibody with a [0063] mouse neutralization assay 12 days after the second immunization of the initial series. Horses that had not mounted an adequate antibody response were evaluated and a determination was made as to whether to continue or to remove the horse from the protocol. As serological results from early groups become available, schedules were modified for remaining animals.
Affinity purification of Anti-BoNT antibodies [0064]
Antibodies against specific serotypes were purified from hyperimmune horse serum. The serum was first precipitated by gradually adding an equal volume of 90%-saturated ammonium sulfate followed by gentle stirring for 2 hr at room temperature. The solution was then sedimented at 3000×g for 15 min. The pellet was taken up in the original volume of PBS, and then dialyzed against three to five changes of PBS to remove residual ammonium sulfate. This crude antibody fraction (50 mL) was circulated twice over the appropriate C-fragment affinity column. The column was then washed with three column volumes of PBS and the BoNT-specific antibody eluted with 0.1 M glycine, pH 3.0, and immediately neutralized with 0.1 volumes of 10×PBS. The eluted fraction was dialyzed extensively against distilled water, lyophilized and stored at 40° C. [0065]
Biotinylation of affinity-purified antibody [0066]
Affinity-purified antibody (10 mg) was dissolved in 1 mL of 10 mM PBS, pH 7.4. To this solution was added 22.2 μL of a freshly prepared solution of 20 mg/mL NHS-LC-biotin (Pierce, Rockford, Ill.) in PBS. The vial was wrapped in foil to protect from light and rotated slowly for 30 min at room temperature. The solution was then dialyzed overnight in the dark against 4 L of 20 mM ammonium acetate at 40° C. Aliquots of 0.5 mg biotinylated antibody were transferred to amber vials. To each vial was added 0.5 mL of a solution containing 6 mg/mL each of radioimmunoassay-grade BSA and gelatin. The vials were then lyophilized and stored at 40° C. Individual vials were reconstituted with 0.5 mL of Superfreeze Conjugate StabilizerTM (Pierce) before use. The reconstituted product was stored at −200° C. [0067]
Assay [0068]
Microtiter plates (Immulon-4, Dynatech Laboratories, Chantilly, Va.) were coated overnight at 40° C. with 100 μL/well of a solution containing 4 ug/ml (BoNT A and B), 5 μg/mL (BoNT E) or 2.5 μg/mL (BoNT F) affinity-purified antibody in coating buffer (0.1 M Na[0069] ₂CO₃, pH 9.6). The remaining sites of absorption were blocked by adding 150 μL/well of Superblock™ blocking reagent (Pierce) for 30 min at 37° C. and then were washed four times with wash buffer (PBS/0.05% Tween 20). Standard curves were prepared by diluting BoNT stock solutions with the appropriate volumes of assay buffer (60 mM PBS, 0.1% BSA, 0.1% Tween 20, and 0.01% sodium azide) or the appropriate human serum matrix. Standards and unknowns (100 μL/well) were incubated for at least 1 hr at 37° C. and the plates washed as above. The appropriate dilution of biotinylated antibody in assay buffer (1:200 for BoNT-A, 1:150 for BoNT-B, 1:250 for BoNT E, 1:400 for BoNT F, 100 μL/well) was added and the plates again were incubated and washed as above. Neutravidin-linked alkaline phosphatase (Pierce, Rockford, Ill., 2 mg/mL in PBS) diluted in assay buffer (1:2,000 for BoNT-A, 1:2,800 for BoNT-B, 1:2,800 for BoNT E, 1:1,500 for BoNT F, 100 μL/well) was then added and the plates were incubated 30 min at 37° C. After a final wash, substrate (para-nitrophenyl phosphate, 1 mg/mL in 1 M Tris, 0.03% MgCl2, pH 9.8) was then added (100 μL/well) and the color allowed to develop for 20-30 min. The optical density at 405 nm was then read on a Bio-Tek 311 Microtiter Plate Reader (Bio-Tek Instruments, Winooski, Vt.). Standard curves were constructed by plotting the absorbance values (mean of triplicate wells) against toxin concentrations, and unknown concentrations were determined from the linear regression equation.

EXAMPLE 1

Standard curves: background, linearity, and detection limits [0070]
Affinity-purifying the horse anti-BoNT sera resulted in a 10-fold increase in specific activity, as measured by mouse neutralization assay (data not shown). The combined effects of highly-purified capture antibody plus exquisite specificity and affinity of the neutravidin/biotin linkage used to couple the chromogenic enzyme to the biotinylated secondary antibody resulted in low background and high specific absorbance. Background in each assay was typically 0.01-0.02 absorbance units or less (data not shown) and was not subtracted from standard curves. For serotypes A and B, standard curves were linear over the range of 0.1-10 ng/mL (10 pg-1 ng/well) and did not differ significantly between assay buffer or 10% human serum (FIGS. 1A and 1B). The detection limit of each assay as described is approximately 0.2 ng/mL (20 pg/well), where absorbance readings were typically twice background. Accurate quantitation was possible at about 0.5 ng/mL (50 pg/well), where absorbance readings were typically 3-5 times background and variations between triplicate wells typically 5-10%. Linearity, as measured by the correlation coefficient (r[0071] ²) of the regressed line, ranged from 0.993-0.999 for all assays.
Background in the BoNT E assay was very low, typically 0.01 absorbance units or less (data not shown). Background in the BoNT F assay was higher, typically 0.1-0.2 absorbance units. The reason for this is unknown, but we were unable to reduce this background by extensively optimizing reagent concentrations without also reducing the slope of the standard curve. It is possible that there was some kind of low-level cross-reactivity of the anti-BoNT F antibody with the blocking reagents, but substituting gelatin, BSA or casein for the SuperblockTM did not appreciably lower the background. In neither assay was the background subtracted from the standard curves. For both serotypes, standard curves were linear over the operating range of the assay and did not differ significantly between assay buffer or 10% human serum (FIGS. 2A and 2B). The limit of accurate quantitation for each assay as described was approximately 0.5 ng/mL (50 pg/well) for BoNT E. Due to the higher background, the quantitation limit for BoNT F was 2 ng/mL (200 pg/well). Variation between triplicate wells was typically 5-10%. Linearity, as measured by the correlation coefficient (r[0072] ²) of the regressed line, ranged from 0.994-0.997 for all assays.

EXAMPLE 2

Inter- and intra-assay variation [0073]
To measure inter- and intra-assay variation (repeatability and reproducibility), assay buffer and 10% human serum in assay buffer were spiked with purified BoNT at three concentrations within the standard curve. These solutions were divided into aliquots and kept frozen at −200° C. Aliquots were thawed immediately before analysis. [0074]
To measure intra-assay variability (reproducibility), five assays were performed in a single day. Separate standard curves were prepared independently from the stock solution for each assay. The results of this experiment for BoNT A and B are shown in Table 1. Both accuracy and precision were excellent; deviation from the expected values was 0-5%, and the standard deviations were typically 2-6% of the mean. The results of this experiment for BoNT E and F are shown in Table 2. Both accuracy and precision were again excellent; deviation from the expected values was 0-2.5%, and the standard deviations were typically 2-5% of the mean. [0075]

To measure inter-assay variability (repeatability), one assay was performed on 5 consecutive days. The results of this experiment with BoNT A and B are shown in Table 3 and for BoNT E and F in Table 4. Again, both accuracy and precision were excellent. Deviation from the expected values and standard deviations were similar to those in the intra-assay variation experiment.

TABLE 1


Intra-assay variability (reproducibility) of the BoNT ELISA = s. Five
independent assays were performed by the same operator in a single day.
Separate standard curves were prepared for each assay. Results are
expressed as ng/mL (±SD). Analytes were the appropriate media spiked
with purified BoNT as described in the text. Correlation coefficients
(r²) describe the linear regression lines fitted to the standard
curves.

	0.5 ng/mL	2.5 ng/mL	8.0 ng/mL	mean r²

BoNT A (assay buffer)	0.5 (±.03)	2.5 (±.17)	8.4 (±.58)	.997 (±.002)
BoNT A (10% serum)	0.5 (±.01)	2.5 (±.10)	8.2 (±.32)	.995 (±.002)
BoNT B (assay buffer)	0.5 (±.02)	2.6 (±.13)	8.4 (±.53)	.997 (±.001)
BoNT B (10% serum)	0.5 (±.01)	2.5 (±.07)	8.0 (±.24)	.997 (±.002)

TABLE 2


Intra-assay variability (reproducibility) of the BoNT ELISAs. Five
independent assays were performed by the same operator in a single
day. Separate standard curves were prepared for each assay.
Results are expressed as ng/mL (±SD). Analytes were the
appropriate media spiked with purified BoNT as described in the
text. Correlation coefficients (r²) describe the linear
regression lines fitted to the standard curves.

	0.5 ng/mL	2.5 ng/mL	8.0 ng/mL	mean r²

BoNT E	0.5 (±0.03)	2.6 (±0.13)	7.9 (±0.17)	0.997 (±0.001)
(assay
buffer)
BoNT E	0.5 (±0.03)	2.5 (±0.05)	7.8 (±0.32)	0.996 (±.002)
(10% serum)

	4.0 ng/mL	10.0 ng/mL	16.0 ng/mL	mean r²

BoNT F	4.2 (±0.04)	10.1 (±0.07)	15.7 (±0.22)	0.994 (±0.002)
(assay
buffer)
BoNT F	4.1 (±0.06)	10.2 (±0.04)	16.0 (±0.17)	0.994 (±0.003)
(10% serum)

TABLE 3


Inter-assay variability (repeatability) of the BoNT ELISA = s. Five
independent assays were performed on separate days by the same operator.
Results are expressed as ng/mL (±SD). Analytes were the appropriate
media spiked with purified BoNT as described in the text. Correlation
coefficients (r²) describe the linear regression lines fitted to
the standard curves.

	0.5 ng/mL	2.5 ng/mL	8.0 ng/mL	mean r²

BoNT A (assay buffer)	0.5 (±.02)	2.5 (±.17)	8.5 (±.18)	.998 (±.002)
BoNT A (10% serum)	0.5 (±.01)	2.5 (±.09)	7.8 (±.17)	.998 (±.002)
BoNT B (assay buffer)	0.5 (±.05)	2.4 (±.11)	8.3 (±.49)	.997 (±.001)
BoNT B (10% serum)	0.5 (±.02)	2.5 (±.09)	7.9 (±.45)	.996 (±.002)

TABLE 4


Inter-assay variability (repeatability) of the BoNT ELISAs. Five
independent assays were performed on separate days by the same operator.
Results are expressed as ng/mL (±SD) Analytes were the appropriate
medium spiked with purified BoNT as described in the text. Correlation
coefficients (r²) describe the linear regression lines fitted to the
standard curves.

	0.5 ng/mL	2.5 ng/mL	8.0 ng/mL	mean r²

BoNT E	0.5 (±.02)	2.4 (±.08)	7.8 (±.18)	.997 (±.001)
(assay buffer)
BoNT E	0.5 (±.02)	2.5 (±.05)	7.9 (±.30)	.995 (±.001)
(10% serum)

	4.0 ng/mL	10.0 ng/mL	16.0 ng/mL	mean r²

BoNT F	4.2 (±.05)	10.0 (±.09)	15.7 (±.20)	.996 (±.002)
(assay buffer)
BoNT F	4.2 (±.05)	10.1 (±.12)	16.0 (±.09)	.997 (±.001)
(10% serum)

EXAMPLE 3

Cross-reactivity between serotypes [0080]
BoNT serotypes exhibit 30-60% sequence identity (Oguma et al., 1995, Microbiol. Immunol. 39, 161-168; Singh et al., 1996, Toxicon 34, 267-275). However, serotype-specific antisera have been reported to elicit little or no cross-reactivity (Sakaguchi, 1983, Pharmac. Ther. 19, 165-194; Kozaki et al., 1989, In: Simpson, L. L. (Ed.), Botulinum Neurotoxins and Tetanus Toxin. Academic Press, San Diego, pp. 301-818). The antibodies used in these assays were affinity-purified against the 50 kD C-fragment of the heavy chains of BoNT A, B, E and F. Each assay was <1% cross-reactive against the other serotype and also serotypes A and B (data not shown), suggesting that most common epitopes among serotypes must reside elsewhere on the protein chains. [0081]

EXAMPLE 4

Recognition of toxins with associated nontoxic proteins [0082]
These assays were developed with toxin standards comprised of highly purified toxins devoid of associated proteins. Naturally occurring toxins, however, are typically complexed with NAPs which serve to protect the toxins from acidic and/or proteolytic degradation in the gastrointestinal tract. These proteins could conceivably block antigenic sites and prevent recognition of the toxins by the antibodies. Therefore, we evaluated whether these assays could detect native BoNTs with their naturally occurring NAPs. [0083]
Because NAPs can vary in both molecular weight and in mass ratio to the purified toxin, it was impossible to compare complexed toxins to the purified toxins on a mass basis. To avoid this problem, we used activity units (mouse LD[0084] ₅₀) as our method of direct comparison. The results of this experiment are shown in FIGS. 3A and 3B for BoNT A and B and FIGS. 4A and 4B for BoNT E and F. The assays appeared to recognize the complex similarly for BoNT B, and only slightly less than the native toxin (approx 25%) for BoNT A. This amount of variability is consistent with the fact that both toxin preparations in each assay were quantified using the mouse bioassay, where quantitation variability is typically +10%.
The assay for BoNT F recognized the complex and neurotoxin approximately equally. However, the assay for BoNT E suggested that the antibody recognized the complex significantly better than the purified neurotoxin. These results were counterintuitive because the antibodies were produced against purified neurotoxin, and affinity-purified against purified C-fragment. While complexation with NAPs might easily decrease specific recognition by antibodies due to masking of antigenic determinants, it was difficult to postulate a viable mechanism whereby specific recognition was increased. BoNT E is different than other serotypes. All clostridial neurotoxins are synthesized and expressed as single polypeptide chains of about 1300 aminoacids. Subsequent proteolytic “nicking” produces the heavy and light chains linked by disulfide bridges. This nicking step is required for toxicity. In most serotypes, this proteolytic activation step occurs intracellularly with active toxin being expressed by the bacterial strain involved. With BoNT E, however, no intracellular proteolysis occurs; inactive toxin is expressed by the bacterial culture and extracellular proteolytic activation is required. Commercially available BoNT E is usually artificially “nicked” by mild treatment with trypsin. This step increases toxicity by approximately 10-fold. We postulated that this trypsinization step, in addition to activating the neurotoxin, might also destroy antigenic determinants and render the toxin less recognizable by our antibodies. [0085]
To test this hypothesis, we purchased from our commercial supplier, METAbiologics, Inc., activated (trypsinized) and unactivated BoNT E neurotoxin. We then tested each, along with the complexed form, in the BoNT E ELISA. Results are shown in FIG. 5. The complexed toxin and the non-activated toxin were recognized similarly by the assay. Although the minor differences in quantitation may have been real, they are not inconsistent with errors involved in determining activity by mouse bioassay (McLellan et al., 1996, Toxicon 34(9), 975-985). In contrast, the activated neurotoxin was barely recognized at all. This suggests two very important things. First, trypsinizing the purified neurotoxin destroyed critical antigenic determinants required for antibody recognition. Second, the complexed neurotoxin appears to have been largely protected from this effect, while still being effectively activated. Evidently, the NAPs were functioning in the trypsinization step in the same way they function in the gut. That is, they allowed proteases to “nick” the toxin at the appropriate site(s) for activation, but protected the molecule from further proteolysis that might prove destructive. This information also suggests that the activation step should precede the final purification steps in commercial preparations. That the purified and the complexed toxin should be recognized equally is not unexpected. As for other serotypes (Szylagyi, et al., 2000), the C-fragment of BoNT E is believed to contain regions that are intimately involved with receptor binding (Clayton, et al., 1995). Should the accessibility of these regions be blocked by NAPs, binding would be hindered, with a resulting loss of toxicity. Thus, binding of antibodies purified against the C-fragment is less likely to be sterically hindered by NAPs than that of antibodies directed against the whole toxin or toxoid. [0086]
We present here simple, sensitive and accurate colorimetric capture ELISAs for BoNT neurotoxins type E and F in assay buffer and 10% human serum. These assays demonstrate sensitivities similar to that of the mouse bioassay, yet offer significant savings in both money and time while eliminating the use of animals. Because the antibodies were affinity-purified against the C-fragments of each toxin, interference by NAPs was minimal. In vitro activation of BoNT E by treatment with trypsin destroys antigenic determinants and may be a confounding factor in the development of immunological assays. [0087]

Claims

What is claimed is:

1. A method for detecting BoNT antigen in a sample comprising:

(a) reacting a sample suspected of having an antigen with a first antibody, Ab(A);

(b) washing the reaction mixture to remove unreacted antibody;

(c) reacting the Ab(A) linked antigen with a second antibody which is conjugated with biotin, Ab(B);

(d) washing the reaction mixture to remove unreacted antibody;

(e) reacting Ab(A)-antigen-Ab(B) with avidin linked to a chromogenic enzyme;

(f) washing the reaction mixture;

(g) reacting the chromogenic enzyme;

(h) measuring optical density of reaction and calculating concentration of antigen using a standard curve.

2. The method of claim 1 wherein said BoNT antigen is chosen from the group consisting of BoNT serotype A, BoNT serotype B, BoNT serotype F, and BoNT serotype E.

3. The method of claim 2 wherein said sample is chosen from the group consisting of: animal serum, food and dirt.

4. The method of claim 3 wherein the standard is assayed simultaneously along with the sample.

5. The method of claim 2 wherein said first antibody is bound to a solid support.

6. The method of claim 5 wherein said solid support is chosen from the group consisting of microtiter plate, membrane, and dipstick.

7. The method of claim 2 wherein said antibody is affinity-purified to heavy chain C-fragment of BoNT.

8. A test kit for the assay of BoNT antigen comprising:

one or more solutions containing a known concentration of free BoNT or complexed BoNT to serve as a standard;

a solution of a anti-BoNT antibody, and biotinylated anti-BoNT antibody affinity purified to the C-fragment of BoNT;

a chromogen which changes color or shade by action with biotin; and, optionally,

a solid support.

9. The kit according to claim 8 wherein said BoNT is chosen from the group consisting of BoNT serotype A, BoNT serotype B, BoNT serotype E, and BoNT serotype F.

10. A test kit of claim 8 wherein said kit further comprises test tubes for said solutions.