In vitro assay of biological and chemical toxins using antibodies against lethal toxin neutralizing factor

Original paper

IN VITRO ASSAY OF BIOLOGICAL AND CHEMICAL TOXINS USING ANTIBODIES AGAINST LETHAL TOXIN NEUTRALIZING FACTOR

B. V. LIPPS¹

1 Ophidia Products Inc., 11320 South Post Oak, Suite 203, Houston, Texas 77035 USA.

ABSTRACT: Lethal Toxin Neutralizing Factor (N-LTNF), MW 63.0 kDa, was isolated from opossum serum. After trypsin digestion, the active domain of N-LTNF was isolated and sequenced. The synthetic peptide consisting of ten amino acids was designated as LT-10. N-LTNF and LT-10 inhibited the lethality of animal, plant and bacteria toxins when tested on mice non-immunologically. The antibodies against N-LTNF and LT-10 only reacted immunologically with toxins and not with non-toxic substances. Anti-LTNF and anti-LT-10 reacted immunologically by ELISA test with toxins that were not detected by mouse test, such as cholera toxin and digoxin. Anti-LTNF and anti-LT-10 failed to react immunologically with non-toxic substances, such as nerve growth factor and collagen. Currently, mouse bioassay is in use for toxin detection and assay. Binding affinity of IgG from anti-LT-10 showed a linear relationship with mouse bioassay by ELISA detection limit to some toxins only. This may be due to the fact that anti-LTNF and anti-LT-10 detected the toxins that were not lethal to mouse. Thus, anti-LTNF and anti-LT-10 can be useful in assaying toxins as an alternative to mouse bioassay.

KEYWORDS: antibodies, in vitro assay, lethal toxin neutralizing factor, toxins, venoms.

INTRODUCTION

Biological toxins are grouped according to the source: animal, plant, unicellular or one-celled algae, and bacteria. Toxins are diverse and range from well-defined single macromolecules (tetanus, diphtheria, botulinum toxins) to mixtures of complex molecules, such as snake or scorpion venoms or simple chemical entities like digoxin and colchicine (3). Currently, mouse bioassay for toxicity is recommended by the Association of Official Analytical Chemists (AOAC), (5). There is no in vitro test that can recognize all types of toxins collectively. Animal bioassay recognizes toxicity from all types of animal, plant, and bacteria-derived toxins. In addition, mouse bioassay collectively detects toxin lethality, whether it is a single toxin or a mixture of several toxins.

The mouse bioassay developed by Summer (16) measures the lethality of shellfish extracts relative to a pure saxitoxin standard. Although this mouse bioassay is the only currently accepted method for monitoring paralytic shellfish envenoming (PSE) in North America, it has some major disadvantages; for example, it requires a large number of mice, and is therefore expensive and objectionable to animal activists. Furthermore, (13) zinc from oyster tissue has been reported as a causative factor in mouse bioassay. On the other hand, toxins not lethal to mouse go undetected by mouse bioassay.

Recently, Sells et al. (15) reported the use of 4-6-day-old chick embryos for the measurement of lethal effects on non-neurotoxic snake venoms. According to the authors, since the use of eggs less than 10 days of incubation is exempt from UK Home Office Control, this model provides an acceptable alternative for venom-induced hemorrhagic activity.

Numerous investigators have developed immunological tests for assaying various toxins, using polyclonal and monoclonal antibodies. Carlson et al. (1) developed radioimmunoassay (RI) for paralytic shellfish envenoming. Biological toxins have been assayed by several types of ELISA tests using specific antibodies to the desired toxin. Li and Ownby (6) reported the development of an ELISA assay to identify venoms from snakes of the Agkistrodon genus. They used purified 14.5 kDa myotoxin as an antigen that is specific to the snakes of the genus Agkistrodon and its rabbit monovalent antiserum in antigen capture format.

When three tests HPLC, fluorescent method, and two monoclonal antibody test kits were compared for the determination of okadaic acid content of dinoflagellate cells, results were not consistent (14). These investigators concluded that since the outbreaks of diarrhetic shellfish envenoming (DSE) may be caused by okadaic acid, methylokadaic acid, or a combination of these toxins, both ELISA kits may underestimate total toxic effect in shellfish. Draisci et al. (2) compared mouse bioassay, HPLC, and enzyme immunoassay methods for the determination of diarrhetic shellfish envenoming toxins in mussels.

All these methods have disadvantages that have prevented their widespread implementation, particularly under the regulatory requirements. Mouse bioassay detects a wide range of known and presumably unknown toxins. Therefore, it is unlikely that mouse bioassay will be completely abandoned in favor of other assays until a similarly responsive alternative is found. Lethal Toxin Neutralizing Factor (N-LTNF), MW 63.0 kDa, was isolated from opossum serum (8). After trypsin digestion, the active domain of N-LTNF was separated and sequenced. A synthetic peptide consisting of ten amino acids was designated as LT-10. N-LTNF and LT-10 inhibited lethality of animal, plant, and bacteria toxins when tested on mice non-immunologically (9). This investigation is about antibodies against N-LTNF and LT-10 that reacted immunologically with toxins. The extensive research and results of this investigation could lead to an alternative in vitro test to mouse bioassay for the detection and for quantification of all kinds of biological and chemical toxins. However, mouse bioassay and the in vitro test using anti-LT-10 cannot be strictly compared because toxins that were not lethal to mouse were detected by this antibody. Therefore, in vitro test using anti LT-10 for assaying toxins will be a test by itself.

MATERIALS AND METHODS

The venoms, toxins, and reagents for ELISA used in this work were purchased from Sigma-Aldrich Co. US.

Venoms

The following venoms were used: Crotalus atrox (rattlesnake); Naja kaouthia, (Thailand cobra); Daboia russelli, (common viper); Oxyuranus s. scutellatus, (Australian taipan); Androctonus australis, (scorpion); Apis mellifera, (honeybee); and Astrotia stokesii, (sea snake).

Single toxins

The following toxins were used: botulinum toxin from Clostridium botulinum; cholera toxin from Vibrio cholerae; cobratoxin from Naja kaouthia; holothurin from Actinopyga agrassizi; and ricin from Ricinus communis.

Chemical toxins - Colchicine and digoxin.

Production of polyclonal antibodies to LTNFs (N-LTNF and Syn. LT-10)

Adult Balb/c mice were used for immunization (7,11). The mice were used in compliance with the U. S. Public Health Service Policy on Humane Care and Use of Animals. For the first injection, LTNF or LT-10 was mixed with equal quantities of Freund's complete adjuvant; for the subsequent doses, the antigens were mixed with incomplete Freund's adjuvant. The mice were injected with 100µg/mouse IM four times, two weeks apart. At the end of the immunization period, mice were bled through ophthalmic vein and sera were separated.

Immunological binding of anti-LTNF and ANTI-LT-10 to venoms, biological, and chemical toxins by enzyme-linked immunosorbent assay (ELISA)

ELISA tests were performed in a 96-well microtiter plate described in detail (10,12). The wells were coated with one concentration of venom or toxin as antigen, 10 µg/ml in carbonate-bicarbonate buffer, pH 9.3, and 100 µl/well was added. The plate was incubated overnight at room temperature (RT). After 18 to 24 hours, the plate was washed three times (3X) with phosphate buffered saline (PBS), pH 7.4.

The plate was blocked with 0.25 ml/well of 3% Teleostean gelatin from cold-water fish (Sigma) for 30 min at RT. The plate was washed 3X with PBS and 100 µl/well of threefold diluted anti-LTNF or anti-LT-10 was added; three wells were used for each dilution.

Purification of IgG from mouse anti-LT-10

Anti LT-10 was fractionated on HPLC from Tosoh Co. (Japan) using ionic exchange column from Polymer Laboratories (UK) and Trizma-HCl gradient buffer, pH 7.4. For a single run, 30mg of serum protein was loaded for fractionation. The IgG fraction was collected, dialyzed, and concentrated using apparatus from Spectrum Co. The IgG concentrated fraction was refractionated on HPLC under identical conditions to obtain homogeneous preparation of IgG. The detailed procedure for the purification of IgG from antiserum is described (12).

Determination of mouse lethal dose for venoms and toxins

The lethal dose was determined by IM injection of 0.1ml of venom or toxin at various concentrations in 18g Balb/c mice.

Detection level of venoms and toxins by immunological binding by anti LT-10, IgG

ELISA test was performed as described above, using threefold concentrations of venom or toxin and 10 µg/ml anti-LT-10 IgG. In this case, plate was coated with 100 µl/well of threefold diluted venom or toxins from 33 µg to 0.01 µg. Three wells were used per concentration. The plate was incubated at RT overnight and washed 3X. Ten µg/ml purified anti-LT-10 IgG in 3% gel was added, and the plate was incubated at 37°C for 1 to 2 hours. The plate was washed 3X with PBS and horseradish peroxidase conjugated with mouse IgG made in goat was added and incubated for 1 hour at 37°C. The plate was then washed 3X with PBS and reacted with O Phenylenediamine Dihydrochloride (OPD) for color development. The test was read after 30 min. at 405mm wavelength.

RESULTS

Results of table 1 showed that: 1. anti-LTNF and anti-LT-10 reacted immunologically with venoms, biological, and chemical toxins and failed to react with non-toxic substances, such as NGF, chemical collagen, and creatine. Therefore, the immunological reaction was specific to toxins; 2. immunological reaction was stronger with anti LT-10 than anti-LTNF. ELISA titer for O. s. scutellatus venom was 600 with anti-LTNF versus 24300 with anti-LT-10. Stronger reactions were observed with both anti-LT-10 for botulinum toxin and the chemical colchicine.

Table 1.
Immunological binding of anti-LTNF and anti-LT-10 to venoms, toxins, and chemicals by ELISA test.

Table 2. Comparison between detection limit of venoms and toxins by immunological binding by IgG from anti-LT-10 and mouse bioassay.

The results showed that in some cases toxicity of venoms and toxins was to some extent proportional to the ELISA detection limit. The mouse lethal dose for C. atrox venom was 300µg and ELISA detection limit was 3.7µg. Mouse lethal dose for sea snake Astrotia stokesii was 4µg and its ELISA detection limit was 0.03µg. Mouse lethal dose for botulinum toxin was 1µg and ELISA detection level was 0.04µg, whereas for holothurin the mouse lethal dose was 200µg and its detection level by anti-LT-10 was 1.8µg. Lower the ELISA detection limit meant higher the toxicity. The toxins that were not lethal to mice, such as cholera toxin and digoxin showed the presence of toxicity by this technique using anti-LT-10. Mouse lethal dose for honeybee and colchicine was 15µg and the detection limit was 2.0µg. On the other hand, the detection level for digoxin by anti-LT-10 IgG was 0.02, ten times more toxic than in honeybee venom or colchicine, but was unable to kill mice with doses up to 500µg. Therefore, this in vitro test cannot be strictly compared to mouse bioassay. This is itself a new alternative in vitro test to assay toxins.

DISCUSSION

Lethal Toxin Neutralizing Factor (N-LTNF), MW 63.0 kDa, was isolated from opossum serum. After trypsin digestion, the active domain of N-LTNF was separated and sequenced. The synthetic peptide consisting of ten amino acids having the sequence: Leu-Lys-Ala-Met-Asp-Pro-Thr-Pro-Pro-Leu was designated as LT-10. N-LTNF and LT-10 inhibited lethality of animal, plant, and bacteria toxins when tested on mice non-immunologically (8,9). However, antibodies against N-LTNF and LT-10 only reacted immunologically with toxins and failed to react with non-toxic substances, such as NGF and collagen. Anti-LTNF and anti-LT-10 also reacted with cholera toxin and digoxin, a chemical toxin that was not lethal to mice. Thus, anti-LTNF and anti-LT-10 were capable of detecting toxins irrespective of their lethality to mice.

The mechanism behind this unusual property of anti-LTNF and anti LT-10 to react immunologically with toxins is currently not understood. The non-immunological activity of anti-LTNFs may be comparable to the non-specific phenomenon observed in the Wasserman test, where non-specific (cardiolipin) reacts with antibodies specific to syphilis (4). In this case, unrelated antibodies against LTNFs had specific immunological reactions with toxins. This may be partially explained by the three dimensional structures of various snake and scorpion venom toxins, which appear to have a small similar conserved region. Currently, three-dimensional structures for other toxins are not known. The fact that antibodies against LT-10 reacted immunologically with venom toxins may suggest that the active site for toxicity may be ten or less amino acids. This unique property of anti-LTNFs to bind to toxins can be utilized in perfecting in vitro assays for various types of venoms and toxins. The use of anti-LTNF can be extended for assaying potency of antivenoms and anti-toxins as an alternative to the only available mouse bioassay currently in use.

The results of ELISA binding between anti-LT-10 and venoms or toxins was an immunological reaction verified by immunoprecipitation tests (IP). IP is a very crude immunological test requiring a high concentration of antigen and antibody. The sensitivity of IP to ELISA is 1:10,000 to 1:100,000. IP tests were performed using cobratoxin, crotoxin, digoxin, NGF, collagen, creatine, and synthetic LT-10 at the concentration of 1mg/ml reacting with 1:1 diluted anti-LT-10. Single band of precipitin (ppt) was shown with cobratoxin, crotoxin, and not with others. (NGF at this concentration gave positive reaction with 1:1 specific anti-NGF). This indicated that the biological toxins must have a concentration 1mg/ml to give IP positive. NGF being nontoxic failed to react with anti LT-10 by IP and ELISA tests. However, LT-10 and colchicine at the concentration of 5mg/ml gave ppt with 1:1 diluted anti-LT-10 and collagen and creatine that are not lethal to mouse failed to do so. Collagen and creatine failed to react with anti-LT-10. Digoxin was not tested because it was not soluble beyond 0.5mg/ml. Therefore, the ELISA binding reaction of anti-LT-10 with biological and chemical toxins was a specific immunological test.

Currently, there is no reliable in vitro test for assaying biological and chemical toxins comparable to mouse bioassay that recognizes the wholesomeness of toxins. Development of in vitro test for assaying biological toxins will be very advantageous. Some laboratories do not have animal facilities or experienced personnel to work with animals. Moreover, animals are expensive and their maintenance is problematic. Considering all these problems, a well-developed standardized in vitro technique to assay biological toxins will have tremendous research and commercial potential. Introduction of the novel antibody, anti-LT-10, will help to understand the structure and the mode of action of biological and chemical toxins.

Received March 20, 2001

Accepted June 5, 2001

• ¹
  CARLSON RE., LEVER ML., LEE BW., GUIRE PE. Development of immunoassays for paralytic shellfish poisoning. In: RAGELIS EP. Ed. Sea Food Toxins. Washington: American Chemical Society, 1984: 181-3.
• ²
  DRAISCI R., CROCI L., GIANNETTI L., COZZI L., STACCHINI A. Comparison of mouse bioassay, HPLC and enzyme immunoassay methods for determining diarrhetic shellfish poisoning in mussels. Toxicon, 1994, 32, 1379-84.
• ³
  HATHEWAY CL., FERREIRA JL. Detection and identification of Clostridium botulinum neurotoxins. In: SINGH BL., TU AT. Eds. Natural Toxins. New York: Plenum Press, 1996: 481-98.
• ⁴
  HEIMOFF LL. The diagnosis of syphilis. Bull. New York Acad. Med., 1976, 52, 863-7.
• ⁵
  HORWITZ W. Ed. Official method of analysis Washington: Association of Official Analytical Chemists, 1990: 881-2.
• ⁶
  LI Q., OWNBY CL. Development of an enzyme-linked immunosorbent assay (ELISA) for identification of venoms from snakes in Agkistrodon genus. Toxicon, 1994, 32, 1315-25.
• ⁷
  LIPPS BV. Biological and immunological properties of nerve growth factor from snake venoms. J. Nat. Toxins, 1998, 7, 121-30.
• ⁸
  LIPPS BV. Anti-lethal factor from opossum serum is a potent anti-dote for animal, plant and bacterial toxins. J. Venom. Anim. Toxins, 1999, 5, 56-66.
• ⁹
  LIPPS BV. Small synthetic peptides inhibit the lethality in mice for toxins derived from animal, plant and bacteria. J. Venom. Anim. Toxins, 2000, 6, 77-86.
• ¹⁰
  LIPPS BV. Production of polyclonal antibodies against cobratoxin, botulinum toxin and ricin without altering their toxicity or use of adjuvant. J. Nat. Toxins, 2001, 10, 27-32.
• ¹¹
  LIPPS BV., KHAN AA. Antigenic cross reactivity among the venoms and toxins from unrelated diverse sources. Toxicon, 2000, 38, 973-80.
• ¹²
  LIPPS BV., KHAN AA. The presence of pharmacological substances: myoglobin and histamine in venoms. J. Venom. Anim. Toxins, 2001, 7, 45-55.
• ¹³
  MCCULLOCH AW. BOYD RK., DEFREITAT ASW., FOXALL RA., JAMIESON WD., LAYCOCK MV., QUILLIAN MA., WRIGHT JLC., BOYKO VJ., MCLAREN JW., MEIDENA MR., POLKINGTON R., ARSENAULT E., RICHARD DJA. Zinc from oyster tissue as causative factor in mouse death in official bioassay for paralytic shell fish poison. J. Assoc. Off. Anal. Chem., 1989, 72, 384-6.
• ¹⁴
  MORTON SL., TINDALL DR. Determination of okadaic acid content of diniflagellate cells: a comparison of the HPLC-fluorescent method and two monoclonal antibody ELISA kits. Toxicon, 1996, 34, 947-54.
• ¹⁵
  SELLS PG., IOANNOU P., THEAKSTON RDG. A humane alternative to the measurement of the lethal effects (LD50) of non-neurotoxic venoms using hens eggs. Toxicon, 1998, 36, 985-91.
• ¹⁶
  SUMMER H., MEYER KF. Paralytic shell-fish poisoning. Arch. Pathol., 1937, 24, 560-98.

Publication Dates

• Publication in this collection
  16 Sept 2002
• Date of issue
  2002

History

• Accepted
  05 June 2001
• Received
  20 Mar 2001