Archived BAM Method: Analysis of Staphylococcal Enterotoxins in Food
Bacteriological Analytical Manual
Electrophoretic and Immunoblot Analysis of Staphylococcal Enterotoxins in Food
Immunoblotting can detect Staphylococcus aureus enterotoxin A in food. The method may also be adapted to other toxins in foods.
Staphylococcal enterotoxins (SE), a family of five major serological types of heat stable (1,6,9,11,18, 19,20, 21), emetic enterotoxins (SEA through SEE), are encoded by five genes, which share 50 to 85% homology at the predicted amino acid level (5,12). Enterotoxin A (SEA), a 27 kDa monomeric protein, is an extremely potent gastrointestinal toxin (2,7)and requires very sensitive methods to detect the low levels in foods (ng/g food) that can cause illness.
After antibodies to SEA were produced, immunological testing became the method of choice for SEA detection (4). Radioimmunoassay (13), microslide double diffusion and enzyme-linked immunosorbent assay (ELISA), have been used for testing food samples. ELISA is especially useful, because it is simple, sensitive (0.5 ng/ml), rapid, and available in commercial kits that use distinct antibodies, either polyclonal or monoclonal.
Cross-reaction with unrelated antigens (15,16) or endogenous peroxides in particular foods that react with colorigenic reagents may not be distinguishable from positive results by some methods without extensive controls (17). In addition, heat-treated SEA (in heat processed foods) may give negative results, because heat-treated enterotoxin may aggregate, reducing its reactivity with antibodies. However, it may retain toxicity after heat treatment (1,3).
Methods for analysis of regulatory samples of foods must resolve or avoid "false positive" and "false negative" reactions. Before antibody is applied, the SDS-PAGE immunoblot method, described below, solubilizes and separates proteins, to discriminate cross reactions to heterologous proteins that may occur.
Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) is a common method for protein separation (10,14). An electrical field is applied so that charged molecules migrate through a polyacrylamide matrix to the electrode bearing the opposite charge. The negatively-charged detergent, SDS, denatures and strongly binds proteins. Then, SDS-bound proteins migrate to the positive pole at rates inversely proportional to their molecular weights.
In general, two-part discontinuous gels are used (10). The sample is loaded onto the upper portion (stacking gel), which has a low acrylamide concentration, low pH, and low resolving ability. When a sample runs through the stacking gel, all proteins are concentrated into a narrow band. That narrow band then enters the lower portion (resolving gel) that separates proteins by size. The acrylamide concentration chosen for the resolving gel depends on the sizes of proteins to be separated. Smaller proteins are resolved at higher acrylamide concentrations and vice versa. SEs are 25-30 kDa; 12.5% acrylamide is useful for separating proteins in that range.
Immunoblotting (also known as "Western" blotting) is widely used for analyzing proteins separated by SDS-PAGE. The proteins are transferred from the gel to a membrane. Then, the membrane is probed with an antibody ("primary antibody") against the specific antigen. To detect the antibody-antigen complex, a secondary antibody is used. Usually, this is a polyclonal antibody (e.g. anti-mouse if the primary antibody is a mouse monoclonal) tagged with a biochemically detectable marker. Some common secondary antibody tags are fluorescent molecules (e.g. FITC, rhodamine), horseradish peroxidase, alkaline phosphatase, or biotin. Then, simple colorimetric reactions are carried out to reveal the location of the complex in a band on the membrane at a position corresponding to the molecular weight of the antigen.
Immunoblots for food testing
Immunoblots have two important advantages for food testing. First, even though heat and other treatments during food processing can cause proteins to aggregate, the aggregates are solubilized and unfolded in SDS gels. Other antibody-based methods of food analysis, such as ELISA, do not have an SDS solubilization step. Instead, the sample is applied directly to the antibody, because SDS in the sample would denature the detecting antibody. Second, cross-reacting antigens usually can be distinguished from the desired antigen on the basis of molecular weight in a Western blot. In ELISA, and other assays in which samples are evaluated without separation or purification, cross-reacting antigens increase the background.
- Equipment and materials
- Equipment: Electrophoretic apparatus: Vertical mini gel unit with 8 x 10 cm or 10 x 12 cm plates (Bio-Rad Mini-Protean II or Hoefer SE-260 or equivalent), 1.5 mm spacers, and 1.5 mm 10 well comb. (Wider combs and spacers for larger volumes may be custom-made).
- Transfer unit, Mini electroblotting unit (Bio-Rad Mini-Trans-Blot or equivalent).
- Power supply: Constant voltage of at least 200V and constant amperage of at least 400 mA (Bio-Rad PowerPac 300 or equivalent). A power supply with timer is recommended.
- Microcentrifuge: A centrifuge that accomodates 1.5 ml (microcentrifuge) tubes and attains speeds of at least 10,000 rpm.
- Homogenizer: An Omni H or other homogenizer for grinding small amounts of food.
- Membrane - Nitrocellulose membranes (similar to Sigma N8142) the same size as gel.
- Miscellaneous small equipment: Rotator or rocking platform for mixing samples, 100 x 15 mm square culture dishes (Falcon 1012 or equivalent), Pyrex baking dish (at least 10 inches square) and 1.5 ml plastic centrifuge (microfuge) tubes.
- Scanner (recommended): A flatbed scanner with resolution of at least 600 DPI (similar to Hewlett Packard 4C).
30% Acrylamide/bis-acrylamide (0.8% bis) pre-mixed solution (Sigma A-3699 or equivalent)
WARNING! ACRYLAMIDE IS NEUROTOXIC. ALWAYS WEAR GLOVES AND OTHER APPROPRIATE PROTECTION WHEN USING.
- BCIP/NBT (5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium), alkaline phosphatase substrate solution (Sigma B-6404 or equivalent).
- Protein molecular weight color markers (6-200 kDa)(Sigma C-3437 or equivalent)
- Purified Staphylococcal enterotoxin A (similar to Sigma S9399 or equivalent
- Molecular biology grade reagents:
- Tris base
- Tween 20 (Polyoxyethylenesorbitan Monolaurate)
- 6N HCl
- Ammonium Persulfate
- non-fat dry milk (similar to Carnation)
- bromphenol blue
- Mouse monoclonal anti-SEA (IGEN Mab 3A #506-022-01)
- Rabbit polyclonal anti-SEA (Sigma S-7656 or equivalent)
- Goat anti-rabbit alkaline phosphatase conjugate (Sigma A2556 or equivalent)
- Goat anti-mouse alkaline phosphatase conjugate (Sigma A4937 or equivalent)
- Preparation of materials
- 1 M Tris, pH 7. Dissolve 121.1 g Tris base in 750 ml H2O. Add 6N HCl to pH 7.0 (approximately 160 ml.). Add H2O to final volume of 1 liters.
- 1M Tris, pH 8 Dissolve 121.1 g Tris base in 750 ml H2O. Add 6N HCl to pH 8.0 (approximately 90 ml.). Add H2O to final volume of 1 liters.
- 20% (W/V) SDS (wear a mask when weighing SDS)
- Gel Buffer A (3M Tris, pH 8.8) Dissolve 181.6 g Tris base in 250 ml H2O. Add 40 ml 6N HCl. Add H2O to final volume of 500 ml.
- Gel Buffer D (0.25M Tris, pH 6.8) Dissolve 15.1 g Tris base in 50 ml H2O. Add 18 ml 6N HCl. Add H2O to final volume of 100 ml.
- 10X Running buffer (250mM Tris; 1.92M glycine; 1.0 % SDS) Dissolve 121 g Tris base and 576 g glycine in 4 l H2O. Stir until completely dissolved. Add 40 g SDS (wear a mask when weighing SDS).
- 10X Western blot transfer buffer (250mM Tris; 1.92M glycine) Add 121 g Tris base and 576 g glycine to 4 l H2O; stir until completely dissolved.
- 1X Western blot transfer buffer. Mix 400 ml 10X Western blot transfer buffer with 2800 ml H2O. Then add 800 ml methanol.
- Tris-Tween blocking buffer (10 mM Tris, pH8; 500 mM NaCl; 0.5% Tween-20) Mix 116 g NaCl with 3 L H2O. Add 40 ml 1.0 M Tris pH 8 and stir until dissolved. Add 20 ml Tween-20 and stir gently. Add water to 4 liters.
- Loading buffer (250mM Tris, pH 7; 4% SDS; 20% glycerol; 10% -mercaptoethanol; 0.05% bromphenol blue: Mix 12.5 ml 1.0 M Tris pH 7 with 10 ml glycerol and 5 ml -mercaptoethanol. Add H2O to 40 ml. Add 10 ml 20% SDS and 25 mg bromphenol blue.
- 10% (w/v) APS (Ammonium persulfate) - prepare fresh weekly; cover in aluminum foil.
Commercially Prepared gels are available through suppliers (depending on apparatus).
- Assemble the casting tray - The gel is cast between two glass plates. Follow the manufacturer's instructions and assemble the plates for casting two gels, using the 1.5 mm spacers. Always cast two gels; the second can be stored, used as a duplicate or as a backup for the first.
- Prepare lower gel - Always wear gloves when handling acrylamide solutions Prepare the resolving (lower) gel (Table 1) in a 50 ml tube. Add acrylamide last. The amount of APS and TEMED is higher than normally recommended, so the solution polymerizes rapidly. If you find it hard to work fast enough, half the amount of APS and TEMED.
Table 1. Lower (resolving) gel (for two 0.75mm gels)
% Gel 12.5% 15% 16% 17% Buffer A 2.8 ml 2.8 ml 2.8 ml 2.8 ml H2O 1.5 ml 0.83 ml 0.58 ml 0.33 ml 20% SDS 50 µl 50 µl 50 µl 50 µl 10% APS 30 µl 30 µl 30 µl 30 µl TEMED 30 µl 30 µl 30 µl 30 µl Acrylamide solution 3.1 ml 3.75 ml 4 ml 4.25 ml Total 7.5 ml
- Pour lower gel - Quickly pour the gel solution between the glass plates to 2/3 of the height. Immediately overlay (GENTLY) with 300 µl H2O or isobutanol. Note, however, that isobutanol may react with the plastic of the apparatus
- Prepare gel to add stacking gel - The gel will polymerize within 15 minutes, forming a clearly visible interface between the water and the gel. The gel can now be used immediately or covered with Saran Wrap and stored (at 4°C) for later use. Right before using the gel, pour off the water (or isobutanol).
- Prepare and pour stacking (upper) gel - Prepare the stacking gel solution (table 2), pour to top of plates and insert 1.5 mm comb. After the gel polymerizes, GENTLY remove the comb and wash the wells with H2O. Drain the wells carefully.
Table 2. Stacking gel (for two 0.75mm gels)
% Gel 4.5% Buffer D 0.625 ml H2O 1.5 ml 20% SDS 30 µl 10% APS 20 µl TEMED 20 µl Acrylamide solution 0.375 ml Total 2.57 ml
- Checking the gels - assemble the electrode unit (follow the manufacturer's instructions) and fill the upper buffer with H2O to test for leaks.
- Sample Preparation
- Mushroom samples - Homogenize at least 1 g mushrooms (follow sampling procedure) with the homogenizer. Add an equal amount (w/v) of the can brine and homogenize again. Transfer 300 µl (approximately) to a 1.5 ml tube. Volume measurements may be inaccurate because these samples are very viscous. It may help to cut the micropipette tip to a larger aperture, or add approx.300 mg to a preweighed tube. Add an equal volume of Loading Buffer. Heat at 90°C for 2 min and centrifuge for 1 min.
- Purified SEA (positive control) - Make a 1 µg/ml stock solution in H2O. For 1 ng add 1 µl to 19 µl H2O, and add an equal volume of Loading Buffer. Heat at 90°C for 2 min and centrifuge for 1 min.
- Running the gel
- Loading - Load gels with 40 µl (sample size depends on combs and spacer use; wider combs and spacers for larger volumes are available through the FDA upon request) of the sample per well. Whenever possible, skip one lane between samples, to minimize cross-contamination. Be sure to apply the positive control several lanes from any test samples. Place 5 µl (25 µg protein) of prestained (color) MW marker in a nearby well to monitor the progress of the run, the effectiveness of transfer and the size of the bands.
- Running the gel - Pour the running buffer into the outer chamber and then GENTLY add buffer to the upper chamber, using a pipette so as not to disturb the samples. When attaching the leads, VERIFY correct electrical polarity. Wrong electrical orientation is the most common mistake in SDS-PAGE. Run at 150 V for 1.5 h (one gel) or 100 V for 2.2 h (two gels), or until the bromphenol blue reaches the bottom of the resolving gel. When running two gels, make sure not to overheat the apparatus (lower the voltage if necessary).
- Stop the gel, dissemble the electrophoresis unit, and remove the top glass plate of the gel.
- Assembling the transfer apparatus - Use gloves when handling nitrocellulose filters. Assemble the transfer unit according to manufacturer's instructions, in a baking dish filled with cold transfer buffer to avoid air bubbles.
- Transferring proteins - Add ice to the unit's cooling reservoir to keep it cool while running. Connect the electric lead and VERIFY correct electrical polarity. Electroblot at 400 mA for one and half hour (you may need to change the ice). [Consult your manual].
- Dissemble the unit - Confirm successful transfer of colored MW markers. Discard gel and put membrane into a baking dish.
- Block the membrane - Incubate in Tris-Tween Blocking Buffer for 20 min with gentle shaking in a square culture dish; use at least 20 ml solution.
- Primary antibody - Add anti-SEA in Tris-Tween Blocking Buffer. If using the monoclonal anti-SEA from IGEN, dilute 1:300. Use at least 10 ml of solution (20 ml for two membranes). For SIGMA anti SEA dilute 1:1000. Shake gently for 2 hours.
- Washing - Wash for 20 minutes in 20 ml or more of Tris-Tween Blocking Buffer with gentle shaking.
- Secondary antibody - Incubate 1-2 hours with the secondary antibody. If a mouse monoclonal was used as the primary antibody, use an anti-mouse alkaline phosphatase conjugate diluted 1:1000 in 10 ml Tris-Tween Blocking Buffer.
- Washing and color development - Wash vigorously three times for 20 min with at least 20 ml Tris-Tween Blocking Buffer. Add 10 ml BCIP/ NBT color reagent for detection. Watch for signal development and for background to determine the correct incubation time empirically for each sample and membrane (approximately 10 min).
- Densitometry (recommended)
- Dry the membrane and then scan the blot to quantitate the signal. Set the scan mode to 256 gray scale black and white photograph scanning. The approximate size of the file (in TIFF format) of a mini gel blot is 1.4 Mb. While contrast and brightness can be adjusted to improve the data for presentation, this will affect the quantitation of the image. Quantitation should always be performed with the raw data compared to a standard curve of known amounts of toxin. While there is no established method for immunoblot quantitation, the bands can be quantitated using NIH Image software (public domain software for MacIntosh).
- Data presentation. Test samples should be recorded using the form suggested here. This form contains all information necessary for Western blot data analysis.
Example 1- Western immunoblotting of food contaminated with SEA
Western blotting was tested for the ability to detect SEA in foods that are commonly associated with food poisoning. Each sample was homogenized, spiked with purified SEA (2 ng/40 µl), and applied directly to the gel.
SEA was detectable in each sample, regardless of which food was present (Fig. 1). Undiluted milk samples distorted SEA mobility (data not shown), but ten-fold diluted milk samples ran correctly. Heterologous antigens cross-reacted in several samples, because polyclonal anti-SEA antibodies reacted with components from the food matrix. For example, the antibodies recognized a 66 KDa protein in milk, whether or not SEA is present in milk (Lanes 2 and 3, Fig. 1). This unrelated band did not affect the assay for SEA, because SEA is determined by the intensity of the 27 kDa band, detected only in the "spiked" sample.
Figure 1. Western immunoblots of foods contaminated with SEA. Food samples were homogenized and spiked with purified SEA. The sample (40 µl) was then applied directly to the gel and assayed by Western blot. Milk, potato salad and meat product with or without SEA were tested. Lane 1 -- Protein Standards; Lane 2 -- milk; Lane 3 -- milk+SEA; Lane 4 -- potato salad; Lane 5 -- potato salad+SEA; Lane 6 -- meat; Lane 7 - meat+SEA.
Example 2 - Detection of SEA in heat-treated mushrooms by Western blotting
Canned foods are problematic for ELISA because ELISA often fails to detect heat-treated SEA. Canned mushrooms were used to see if Western blots can detect heat-treated SEA in food.
The contents of a can of mushrooms (113 g in 500 ml flask) were inoculated with an overnight culture (106 cells /ml) of S. aureus (ATCC13565), then cells were grown for 6 h at 37C with shaking. Samples were taken hourly to measure bacterial growth and SEA production. Each sample was autoclaved at 121°C for 20 min to simulate canning and then assayed by Western blot. As shown in Figure 2. SEA was detected in contaminated mushrooms at 130 min (lane 3), at mid-log phase. Although there are additional cross-reacting bands, they have different molecular weights from SEA, and do not affect the analysis. There is no 27 kDa band in the uninfected control (lane 1).
Figure 2. Detection of SEA produce by S. aureus (ATCC No. 13565) grown on mushrooms. Lane 1 -- sample after 0 min; Lane 2 -- 60 min; Lane 3 -- 130 min ; Lane 4 -- 180 min; Lane 5 -- 240 min; Lane 6 -- 300 min.
|Slow or no polymerization of the gel||APS is old, OR APS, TEMED or acrylamide were left out|
|No tracking dye observed||Wrong polarity|
|Smile effect or gel overheats||High voltage leads to excessive heat.|
|Sample floats in the well or diffuses out of well||Wrong concentration of glycerol in the loading dye|
|No transfer||Wrong polarity or problem with transfer buffer.|
|Transfer OK but no signal in the positive control, membrane turns purple||problem with antibodies.|
|Transfer OK but no signal in the positive control, membrane is colorless||problem with detection reagents|
|High background||problems with blocking. increase time and/or add 0.5%-1% non-fat dry milk to the Blocking Buffer|
|Diffuse /distorted marker bands||Too little SDS|
|Distorted SEA band in samples||Too much protein in sample (overloading). Dilute sample or use chromatography or immunoprecipitation to remove major proteins|
Limitations of Western blotting:
Western blotting has some limitations, which are important to recognize when applying the method to food analysis.
- First, inactive and active SE are nearly indistinguishable by Western blotting (or any other antibody-based method).
- Second, only small sample volumes (30-50 µl) can be loaded onto a gel (Wider combs and spacers for larger volumes are available through the FDA upon request)., which may limit the sensitivity of the method. Preparative methodology (tube gels and preparative electrophoresis), which is under development, may overcome this limitation. The present technology with small samples is nevertheless extremely sensitive. When compared directly with ELISA using contaminated mushroom samples, Western blotting was as sensitive as ELISA with native samples and much more sensitive with heat-denatured samples.
A third limitation of Western blots is that cross-reactive bands potentially could co-migrate with the antigen. Cross-reactivity is an inherent problem with all immunological methodology, because antibodies recognize small regions of proteins and similar epitopes may occur in other proteins. This is a major concern in ELISA and other methods in which the proteins are not separated. It is a smaller concern with Western blotting because the proteins are separated, but false positives are still a potential problem.
There are several ways to minimize this problem. One is to increase the specificity of the reaction by using monoclonal antibodies. Alternatively, several independently isolated antibodies and control samples of uncontaminated similar food can be used to determine whether the bands represent toxin or unrelated antigens.
- Finally, it is important to note that Western blots have a linear response over a broad range of toxin concentrations. However, at very low levels, the signal is not linear.
|FDA ELECTROPHORETIC ANALYSIS REPORT|
|Gel: % acrylamide||spacer mm||sample vol. ml||apparatus:|
|Running volts||time: min||remarks:|
|Transfer mA||time: min||remarks:|
|Blocking:||time: min||volume: ml||remarks:|
|Antibody 1:||time: min||volume: ml||supplier:||dilution:|
|Antibody 2:||time: min||volume: ml||supplier:||dilution:|
|Wash buffer:||vol : ml||wash#1: min||wash#2: min||wash#3: min|
|Developing:||vol: ml||time min||Background:|
1. Anderson J.E., R.R. Beelman, and S. Doores 1986. Persistence of serological and biological activities of staphylococcal enterotoxin A in canned mushrooms. J. Food Protect. 59:1292-1299.
2. Archer, D.L., and F.E. Young. 1988. Contemporary issues: disease with a food vector. Clin. Microbiol. Rev. 1:377-398.
3. Bennett, R.W. 1992. The biomolecular temperment of staphylococcal enterotoxin in thermally processed food J. Assoc. Off. Agric. Chem. 75:6-12.
4. Bergdoll, M.S., M.J. Suargalla, and G.M. Dack. 1959. Staphylococcal enterotoxin. Identification of specific precipitating antibody with enterotoxin-neutralizing property. J. Immunol. 83:334-338.
5. Bergdoll, M.S. 1972. The enterotoxin, In: The Staphylococci ed. Jay Cohen, Wiley Interscience, pp. 301-331.
6. Denny, C.B., J.Y. Humber, and C.W. Bohrer. 1971. Effect of toxin concentration on the heat inactivation of staphylococcal enterotoxin A in beef bouillon and in phosphate buffer. Appl Microbiol 21: 1064-1066
7. Evenson, M.I., M.W. Hinds, R.S. Bernstein, and M.S. Bergdoll. 1988. Estimation of human dose of staphylococcal enterotoxin A from a large outbreak of staphylococcal food poisoning involving chocolate milk. Int. J. Food Microbiol. 7:311-316.
8. Frieman, S.M., J.R. Tumang, and M.K. Crow. 1993. Microbial superantigens as etiopathogenic agents in autoimmunity. Rheum. Dis. Clin. North Am. 19:207-222.
9. Fung, D.Y., D.H. Steinberg, R.D. Miller, M.J. Kurantnick, and T.F. Murphy. 1973. Thermal inactivation of staphylococcal enterotoxins B and C. Appl Microbiol 26: 938-942.
10. Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227:680-685.
11. Lee I.C., K.E. Stevenson, and L.G. Harmon. 1977. Effect of beef broth protein on the thermal inactivation of staphylococcal enterotoxin B1. Appl Environ Microbiol 33: 341-344
12. Marrack P., and J. Kappler. 1990. The staphylococcal enterotoxins and their relatives. Science 248:705-711
13. Miller, R.A., R.F. Reiser, and S.M. Bergdoll. 1976. The detection of staphylococcal enterotoxin A, B, C, D, and E in food by radioimmunoassay, using staphylococcal cells containing protein A as immunosorbent. Appl. Environ. Microbiol. 36:421-426.
14. Orden, J. A., J. Goyache, and J. Hernandez, A. Domenech, G. Suarez, and E. Gomez-Lucia. 1992. Applicability of an immunoblot technique combined with a semiautomated electrophoresis system for detection of staphylococcal enterotoxins in food extracts. Appl Environ Microbiol 58: 4083-4085
15. Park, C. E., M. Akhtar, and M. K. Rayman. 1992. Nonspecific reactions of a commercial enzyme-linked immunosorbent assay kit (TECRA) for detection of staphylococcal enterotoxins in foods. Appl Environ Microbiol 58: 2509-2512
16. Park, C. E., M. Akhtar, and M. K. Rayman. 1993. Simple solutions to false-positive staphylococcal enterotoxin assays with seafood tested with an enzyme-linked immunosorbent assay kit (TECRA). Appl. Environ. Microbiol. 59:2210-2213.
17. Park, C. E., M. Akhtar, and M. K. Rayman. 1994. Evaluation of a commercial enzyme immunoassay kit (RIDASCREEN) for detection of staphylococcal enterotoxins A, B, C, D, and E in foods. Appl. Environ. Microbiol. 60:677-681.
18. Read, Jr., R.B., and J.G. Bradshaw .1966. Staphylococcal enterotoxin B thermal inactivation in milk. J Dairy Sci 49: 202-203.
19. Read Jr., R.B., and J.G. Bradshaw 1966. Thermal inactivation of staphylococcal enterotoxin B in veronal buffer. Appl Microbiol 14: 130-132.
20. Schwabe, M., S.Notermans, R. Boot, S.R. Tatini, and J. Kramer. 1990. Inactivation of staphylococcal enterotoxins by heat and reactivation by high pH treatment. Int. J. Food Microbiol. 10:33-42.
21. Tibana, A., K.M. Rayman, M. Akhtar, and R. Szabo. 1987. Thermal stability of staphylococcal enterotoxin A, B, and C in a buffered system. J. Food Protect. 50:239-242.
Hypertext Source: Detection of Staphylococcus aureus Enterotoxin A in food by Western Electrophoretic and Immunoblot Analysis of Staphylococcal Enterotoxins in Food, Bacteriological Analytical Manual, 8th Edition, Revision A, 1998. Chapter 13b.
Author: Avraham Rasooly