abstract

The purpose of this study was to develop a highly sensitive and reproducible fluorescence immunoassay (FIA) for the detection of ricin in buffer as well as in ricin spiked blood. For the assay polyclonal antibodies produced in house and commercially available secondary antibodies were used. By fluorescence immunoassay, we were able to detect more than 10 fg/mL of ricin in assay buffer and 20 fg/mL in plasma and red blood cells (RBCs). The assay was performed up to 24 hrs after spiking the toxin in the blood. The coefficient of variation ranged from 4-12 % for plasma and 6-22 % for RBCs. The novelty of this developed assay is the omission of the extraction step, which makes the assay less tedious and time consuming. This newly modified method will be helpful in the timely diagnosis of ricin poisoning in the blood samples. 

 

 

INTRODUCTION

Ricin is considered to be a high threat biohazard due to its high toxicity and easy availability. In the past ricin has been used for the suicidal and homicidal purposes. Perhaps the most published incident in the United Kingdom was the death of Georgi Ivanov Markov in London in 1978 [Knight, 1975]. Ricin was administered to Markov via a tiny pellet injected into the back of his thigh. The pellet was loaded in an umbrella. Use of ricin was strongly implicated, but never proven so far due to lack of availability of suitable techniques [Crompton and Gall, 1980]. More recently on January 5th 2003, six men were arrested in a wood green flat (London) on suspicion of manufacturing the toxin for terrorist purposes. Ricin was also detected in the mail at the white house in Washington D. C. in Nov. 2003. [Shea and Gottron, 2004]

Ricin is one of the most toxic proteinaceous substances on earth and is extracted from the seeds of the castor oil plant, commonly known as Ricinus communis. It is a glycoprotein with molecular weight of 65 000, and consist of two chains A and B of nearly equal size linked together by a disulfide bond [Olsnes and Pihl, 1982], designated as A-S-S-B. The lectin B chain is known to bind to the specific glycoprotein or glycolipids sugar residues present on the cell membrane, and then internalization occurs by endocytosis. The N-glycosidase activity of the catalytic A chain deadenylates the 28S ribosomal RNA, by cleaving the bond between ribose and the adenine at position 4324. This activity prevents the formation of the critical stem loop configuration to which the elongation factor 2 is known to bind during translation. Thus the result is the complete inhibition of the cellular translation with consequent cell death [Batelli, 2004].

Ricin has also been exploited therapeutically. As a part of an immunotoxin, it has been used in the treatment of the cancer patients [Lord et al., 1994]. Thus, for clinical diagnosis, forensic use or to monitor the blood levels in patients, we need to have a highly sensitive and fast method that can detect ricin in minute quantities. The most sensitive method reported to date is that described by Lubelli [Lubelli et al., 2006]. With a newer technique using immuno-PCR these authors are able to detect ricin to as low as 10 fg/mL in the assay buffer, as well as in serum samples. But in immuno-PCR the time required for analysis and ricin detection is longer than that of conventional ELISA techniques.

The previous method used for the detection of ricin from blood cells involves first the extraction of ricin from these cells using lactose [Godal et al., 1981] and then traditional ELISA was performed on the extracted ricin. The main objective of this study was to detect ricin without extraction from the matrix. In the present studies we expected that a very high sensitivity could be achieved with Fluorescence Immuno Assay (FIA). We were able to detect ricin in plasma and RBCs obtained after centrifugation of ricin-spiked blood, even 24 hrs after ricin incubation in blood. We detected ricin from RBCs directly, without extraction, using comparative indirect ELISA.

Materials and methods

Materials

R. communis seeds used for ricin purification were obtained locally. The antirabbit IgG—FITC conjugate (Immunoglobulin G — Fluorescein isothiocyanate) was purchased from Sigma. The other chemicals were of analytical grade from Sigma/Emerck/BDH/SRL.

Ricin purification

Ricin was purified in the laboratory from R. communis seeds as described elsewhere [Kumar et al., 2003]. Briefly, defatted castor seed meal was treated with 5% acetic acid and crude ricin was extracted, which was further purified in a single step by using lactamyl sepharose affinity column. The sepharose 4B was activated according to Hegde method [Hegde, et al., [1991] by introducing an epoxy group, amination of an epoxy activated gel, and subsequent coupling of ligand lactose. Under these conditions the lectins bind to the column matrix, which was then eluted with 0.4 M lactose. The eluted material was further subjected to gel filtration to separate the proteins on the basis of their molecular size. The fractions corresponding to the eluted peak were pooled, dialyzed extensively against phosphate buffered saline (PBS) and lyophilized. Purification using Polyacrylamide Gel Electrophoresis (PAGE) under reduced and non-reduced conditions was performed each time to assess the purity of ricin.

Development of Polyclonal Antibody

The polyclonal anti-ricin antibodies were developed in rabbits by using ricin toxoid produced when the toxin was incubated with 1% formalin at 37 ºC for 21 days. Complete Freund’s Adjuvant was used for the first dose and subsequent doses used incomplete Freund’s Adjuvant. After the booster dose, the animals were bled and antisera were collected.

i). Preparation of toxoid and assessment of residual toxicity

Ricin was treated with 1% formaldehyde for four weeks at 37 ºC for 21 days. Unreacted formaldehyde was removed by desalting on Sephadex G-25 before it was used for immunization. The prepared toxoid was assessed for residual toxicity in vitro and in vivo. For in vitro toxicity assessment, VERO cells (African Green Monkey Kidney epithelial cell line) grown in 24-well plates under routine culture method at 37 ºC in an atmosphere of 95% O₂, 5% CO₂ in Dulbecco's Minimum Essential Media (DMEM) supplemented with 10% Fetal Bovine Serum (including routine supplements and antibiotics) were used. Cells in duplicate were incubated with three concentrations of ricin (10.0, 1.0 and 0.1 μg/mL). Control cells were incubated with growth medium only. After 24 hrs of incubation, we assessed the toxicity using the neutral red viability assay [Borenfreund, et al., 1988].

ii). Immunization and antibody titer

The rabbits were used as the animal model for raising the antibodies against ricin. The animals were immunized with toxoid at the dose of 125 μg/kg body weight through subcutaneous injection; test bleeds were made to assess immune status. Immunization details are given in immunization protocol. The antibody titers of the animals were checked using ELISA. The immunization protocol is shown in the following diagram.

 

IMMUNIZATION PROTOCOL

First injection
(3 week interval)
↓
Second injection - first test bleed for assessment of serum titres
(3 week interval)
↓
Third injection - second test bleed
(3 week interval)
↓
Fourth injection - Third test bleed
(3 week interval)
↓
Booster injection
(1 week interval)
↓
No further injection - Fourth test bleed

Purification of IgG:

The raised ricin antisera were subjected to IgG purification using Bio-Rad IgG purification kit as per the manufacturer’s instruction. Briefly, 3.0 mL serum were first passed through the desalting column (Econo-Pac 10 DG column) and eluted using four mL of application buffer provided with the kit. The four mL of sample collected was then passed through the Econo-Pac serum IgG Purification column. The IgG fraction of serum was then eluted with 20 mL of application buffer. The animals were handled according to the guidelines of CPCSEA and the Institutional Ethical Committee on Animal Experiments approved the experiment.

Western blot analysis

Western blot analysis was performed as described by Chaponi and Migliorini [Chaponi et al., 1999]. Ricin was separated on SDS–PAGE and transferred onto a nitrocellulose membrane filter using electrophoresis and an electro-blotting apparatus (Bio-Rad). After protein transfer, PVDF membrane (0.20 μm, pore size; Pierce Biotechnology, USA) was blocked overnight at 4 ºC with 5% SMP in PBS and incubated 90 min with anti-ricin antibody at 1:2,000 dilution. The treated membrane was incubated with goat anti-rabbit IgG-HRP conjugate (1:50,000) for 90 min at room temperature. The membrane was washed again and developed using an enhanced chemiluminescent detection system (ProteoQwest TM Chemiluminescent Western blotting kit, Sigma) according to manufacturer’s protocol. The image was taken on Pierce CL-XPosureTM X-ray films.

Comparative indirect ELISA assay

A comparative indirect ELISA was designed for detection of ricin. The ELISA was performed in 96 wells polystyrene black plates. Ricin was coated on plate (1µg /100uL), diluted in carbonate bicarbonate buffer pH 9.6, and incubated at 37 ºC for 5 hrs. The unbound ricin was removed by washing the plate 3 times (5 min each) with 0.05% (v/v) PBS-tween-20 and was blocked with 200 µL of 3% BSA in PBS for 3 hrs at 37 ºC. We optimized blocking at various time intervals (30 to 240 min) with different blocking agents (different percentages of skimmed milk powder and BSA).

Rabbit blood was taken directly from the heart through into a tube containing heparin. 200 µL of heparinized rabbit blood was dispensed from Eppendorf pipettes and 50 µL of ricin solution (in varying concentrations starting from 2 µg to 2 fg) was added. The blood was then incubated at 37 ºC on rocker plates for different time intervals, i.e., 1, 2, 4, 16 and 24 hrs. After incubation the blood was centrifuged for 10 min at 3000 rpm. The plasma and RBCs were separated. Equal amounts (100 µL) of rat or rabbit anti-ricin IgG were added to plasma and RBCs separately. (No difference was observed between the binding of the rat and rabbit IgGs, and, rabbit IgG was preferred for further studies.) The plasma and RBCs were then incubated at 37 ºC for 2 hrs with rabbit IgG. After incubation for 2 hrs at 37 ºC, the RBCs were centrifuged at 3000 rpm for 10 min, and, 100 µL of the RBC’s supernatant and 100 µL of plasma were added to the ricin pre-coated plate, which was then incubated overnight at 4 ºC. The following day the plate was washed three times with Phosphate Buffered Saline Tween-20 (PBST) and 100 µL of anti-rabbit FITC conjugate (1:1000 diluted in PBST) was added to the wells of plate, which was then incubated at 37 ºC for one hour. The plate was washed four times with PBST and the fluorescence was read at 485/528 nm. In all experiments, antigen and antibody negative control were also measured.

Calculation of results (statistical analysis)

A plot of the ricin concentration against the fluorescence (rf) values was drawn. The mean, standard error and coefficient of variation were calculated for each group of four samples for all the three matrices used, i.e., assay buffer, plasma and RBCs.

Results

In the present study, we purified ricin by a single step method using lactamyl sepharose affinity column. The purified ricin was converted to its toxoid by incubation in formalin and animals were immunized. The antibody titer was studied after each bleeding. The antibody titer was determined using ELISA. The purified IgG antibody titer was found to be 1:12800 by ELISA (Figure 1). Western blot analysis was performed and it showed a high specificity of IgG with ricin (Figure 2).

During our studies we observed that bovine serum albumin blocking gives better results as compared to skimmed milk powder. An increase of blocking time from 120 to 180 min reduces the background noise significantly. Excellent results were obtained when the plates were blocked for 180 min for assay buffer, plasma and blood cell supernatant. The fluorescence was read by calibrating the instrument at different sensitivities starting from 25 to 175. At a high sensitivity of 150 to 175, the difference between positive and negative controls increases significantly.

The ricin coating on ELISA plate for comparative indirect ELISA assay was confirmed by performing the colorimetric ELISA. Ricin binds with the plate in a similar manner as it was coated in the plate (data not shown). The detection of ricin using comparative indirect ELISA was standardized initially in Phosphate Buffered Saline (PBS) using competitive ELISA. We were able to detect ricin to the femtogram level in PBS (Figure 1). In blood, we detected ricin by comparative indirect ELISA at different time periods of ricin incubation. Ricin in varying concentration (starting from 2 µg to 2 fg) was added to blood, which was then incubated for 1, 2, 4, 16, and 24 hrs. After incubation plasma and RBCs were separately tested for the presence of ricin. After one hour of incubation of ricin in blood, we were able to detect 200—20 fg/mL ricin in the plasma (CV: -6 to 12 %), however in RBCs we were able to detect ricin only up to 200—20 ng/mL (CV: -4 to 16%) (Figure 2). After two hours incubation, the results obtained were almost same as those from one hour (Figure 3). However with increases in time of ricin incubation in blood of 4 hours and 16 hrs, only minute quantities of ricin (20 fg/mL) were detected in plasma and RBCs (Figure 4) with CVs that ranged from 4%—18% for plasma and 6—22% for RBCs. After 24 hrs incubation, 20 fg/mL ricin were detected in RBCs (CV: – 7 to 25%). However in plasma, only 200 ng—20 ng/mL of ricin were detectable (CV: – 6 to 12%) (Figure 6).

The results clearly indicate that initially ricin binds to RBCs, and therefore less ricin is available in the plasma. Ricin, a lectin, strongly binds with RBCs and it may be released from RBCs as the incubation time increases and thus available for detection. During each time point, those test samples having fluorescence significantly lower than that of antigen negative control samples were considered positive. The developed assay was repeated several times and also by different individuals. The sensitivity of assay in blood and RBCs was well within reported limits of detection. The results of this study showed that ricin can be detected in plasma and RBCs in the lower femtogram range. In this study, we could not detect ricin in RBC ghosts (i.e., in erythrocyte membranes that remain intact after hemolysis) at different time intervals of incubation.

Discussion

Ricin is a highly toxic plant protein that belongs to a group of proteins called ribosome-inactivating proteins (RIP). There are many reports about extraction of ricin by individuals for criminal or terrorist activities as the purification of ricin is not very difficult. Ricin may vary in degree of glycosylation between different castor bean plant species as well as within the same plant as a result of multigenic expression. The term ricin applies to all the isoforms, including the toxic mutant, but not to the individual side chains. In other words, ricin is characterized by the generalized structure formula A-S-S-B. All the forms of ricin that have A-S-S-B, whether they are active or inactive, are called ricin. Ricin and ricin A-chain is used in the preparation of an immunotoxin for the treatment of cancer. Thus we need a sensitive immunoassay that can detect minute quantities of ricin in biological fluids, in case of ricin exposure, and also can monitor the toxin in biological fluids of patients undergoing immunotoxin therapy.

Different assays have been described for the detection of ricin in solution by various authors including radioimmunoassay [Godal et al., 1981] that detected ricin in picogram level in the blood. Poli [Poli et al., 1994] reported a chemiluminescence ELISA that can detect ricin from diluted serum and urine samples. A sandwich ELISA, colloidal gold based immuno-chromatographic assay and hydrogel based protein microchips has also been reported [Shyu et al., 2002a; Shyu et al., 2002b; Rubina et al., 2005]. Cook [Cook et al., 2006] described an antigen capture sandwich ELISA, by which they have done retrospective identifications of ricin in animals tissues following various routes of exposure. Malcolm and Brian [Malcom et al.,1997] detected ricin from diluted human plasma after extraction of ricin in extraction buffer containing lactose.

In the present study our main objective was to detect ricin from whole blood and to omit the extraction step. Because no matter how efficient the extraction method is, 100% extraction of ricin from blood cells is quite impossible. At a molar concentration ranging from 0.25 to 0.4 M lactose only 50% ricin can be extracted from tissues with two to three repetitions are required [Griffith et al., 1986]. The percentage of ricin extracted definitely affected the sensitivity of assay. By omitting the extraction step not only the sensitivity of the assay has been increased, but also the time and labor has been reduced and the assay has been made simpler than before.

During our studies we found that the plasma constituents do not interfere with the assay and with the antigen detection limit; it is same for both the assay buffer (PBS) and for plasma. We also observed that blocking time plays an important role when performing the assay with plasma. During the initial phase of studies, the background noise was high when using plasma and blood cell supernatant, but the noise decreased by increasing the blocking time from 60 min to 180 min.

Conclusions

The main focus of this study was to detect ricin without extraction from the matrix. In our in vitro studies we demonstrate that a comparative indirect ELISA was highly sensitive to measurable quantities of ricin available in the plasma or released from RBCs at different time intervals. The detection of ricin was successful up to 24 hrs after exposure. The sensitivity of the assay was sufficient enough to determine toxin in plasma above 20 fg/mL up to 16 hrs after exposure and in RBCs up to 24 hrs of exposure. The scenario may be different in vivo as ricin be distributed in various organs and blood cell components soon after exposure. Therefore, to detection of ricin from serum/plasma of individuals exposed to ricin will be quite difficult at lethal dose, estimated for humans to be 1 to 10 µg per kg body weight following inhalation or injection. (See for example http://www.jmedcbr.org/Issue_0301/Suntres/Suntres_1205.html) Hence, more efforts are needed for the detection of ricin at its lethal exposure dose.

Acknowledgements

Authors are thankful to Dr. R. Vijayaraghavan, Director, Defence research and Development Establishment, Gwalior, for his keen interest and providing necessary facilities for this study. We further declare that the performed experiments comply with the current laws in India.

 

Editor Notes:

According to the Science Advisory Board of the OPCW, as stated in February 2008, the following definition of ricin is suggested. (Available at: www.archiviodisarmo.it/siti/sito_archiviodisarmo/upload/documenti/70478_OPCW_2008_Seconda_Conferenza_di_revisione.pdf)

“All forms of ricin originating from Ricinus communis, including any possible variations in the structure of the molecule arising from natural processes or manmade modification, are to be considered ricin as long as they conform to the basic ‘native’ bipartite molecular structure of ricin (A-S-S-B) that is required for mammalian toxicity. Once the inter-chain S-S bond is broken or the protein denatured, it is no longer ricin.”

 

 

figures

 

Figure 3. Detection of ricin in assay buffer after 1 hr of incubation. 50 µL of ricin (in varying concentrations starting from 2 ug to 2 fg) was added to 200µL buffer. An equal amount of rabbit antiricin IgG were added to buffer and incubated at 37 ºC for 2 hrs. After incubation 100 µL of buffer was added to the ricin pre-coated plate. Next day antirabbit FITC conjugate was added and fluorescence was read at 485/528 nm. In all experiments, antigen and antibody negative control were also kept. Ricin in assay buffer after 1 hr incubation (-•-) and control without ricin (•). [go to text reference]

 

Figure 4. Detection of ricin in plasma and RBCs after 1 hr of incubation. 50 µL of ricin (in varying concentrations starting from 2 ug to 2 fg) was added to 200µL blood. After addition of ricin blood was incubated at 37 ºC on rocker for 1hr. After incubation plasma and RBCs were separated. An equal amount (100 µL) of rabbit antiricin IgG were added to plasma and RBCs separately and incubated at 37 ºC for 2 hrs. After incubation RBCs were centrifuged at 3000 rpm for 10 min and 100 µL of RBCs supernatant and 100 µL of plasma were added to the ricin pre-coated plate. Next day antirabbit FITC conjugate was added and fluorescence was read at 485/528 nm. In all experiments, antigen and antibody negative control were also kept. Ricin after 1 hr incubation in plasma (-•-) and RBCs (-•-). Control plasma without ricin (•) and Control RBCs without ricin (•). [go to text reference]

 

Figure 5. Detection of ricin in plasma and RBCs after 2 hrs of incubation. 50 µL of ricin (in varying concentrations starting from 2 ug to 2 fg) was added to 200µL blood. After addition of ricin blood was incubated at 37 ºC on rocker for 2 hr. After incubation plasma and RBCs were separated. An equal amount (100 µL) of rabbit antiricin IgG were added to plasma and RBCs separately and incubated at 37 ºC for 2 hrs. After incubation RBCs were centrifuged at 3000 rpm for 10 min and 100 µL of RBCs supernatant and 100 µL of plasma were added to the ricin pre-coated plate. Next day antirabbit FITC conjugate was added and fluorescence was read at 485/528 nm. In all experiments, antigen and antibody negative control were also kept. Ricin after 2 hrs incubation in plasma (-•-) and RBCs (-•-). Control plasma without ricin (•) and Control RBCs without ricin (•). [go to text reference]

 

Figure 6. Detection of ricin in plasma and RBCs after 4 hrs of incubation. 50 µL of ricin (in varying concentrations starting from 2 ug to 2 fg) was added to 200µL blood. After addition of ricin blood was incubated at 37 ºC on rocker for 4 hrs. After incubation plasma and RBCs were separated. An equal amount (100 µL) of rabbit antiricin IgG were added to plasma and RBCs separately and incubated at 37 ºC for 2 hrs. After incubation RBCs were entrifuged at 3000 rpm for 10 min and 100 µL of RBCs supernatant and 100 µL of plasma were added to the ricin pre-coated plate. Next day antirabbit FITC conjugate was added and fluorescence was read at 485/528 nm. In all experiments, antigen and antibody negative control were also kept. Ricin after 4 hrs of incubation in plasma (-•-) and RBCs (-•-). Control plasma without ricin (•) and Control RBCs without ricin (•). [go to text reference]

 

Figure 7. Detection of ricin in plasma and RBCs after 16 hrs of incubation. 50 µL of ricin (in varying concentrations starting from 2 ug to 2 fg) was added to 200µL blood. After addition of ricin blood was incubated at 37 ºC on rocker for 16 hrs. After incubation plasma and RBCs were separated. An equal amount (100 µL) of rabbit antiricin IgG were added to plasma and RBCs separately and incubated at 37 ºC for 2 hrs. After incubation RBCs were centrifuged at 3000 rpm for 10 min and 100 µL of RBCs supernatant and 100 µL of plasma were added to the ricin pre-coated plate. Next day antirabbit FITC conjugate was added and fluorescence was read at 485/528 nm. In all experiments, antigen and antibody negative control were also kept. Ricin after 16 hrs incubation in plasma (-•-) and RBCs (-•-). Control plasma without ricin (•) and Control RBCs without ricin (•). [go to text reference]

 

Figure 8. Detection of ricin in plasma and RBCs after 24 hrs of incubation. 50 µL of ricin (in varying concentrations starting from 2 ug to 2 fg) was added to 200µL blood. After addition of ricin blood was incubated at 37 ºC on rocker for 24 hrs. After incubation plasma and RBCs were separated. An equal amount (100 µL) of rabbit antiricin IgG were added to plasma and RBCs separately and incubated at 37 ºC for 2 hrs. After incubation RBCs were centrifuged at 3000 rpm for 10 min and 100 µL of RBCs supernatant and 100 µL of plasma were added to the ricin pre-coated plate. Next day antirabbit FITC conjugate was added and fluorescence was read at 485/528 nm. In all experiments, antigen and antibody negative control were also kept. Ricin after 24 hrs of incubation in plasma (-•-) and RBCs (-•-). Control plasma without ricin (•) and Control RBCs without ricin (•).[go to text reference]

references

 

Batelli, M.G. (2004). Cytotoxicity and toxicity to animals and humans of ribosome-inactivating proteins, Mini Rev.Med Chem 4, 513-521.

Borenfreund, E.; Babich, H. and Martin-Alnwick, N. Comparisons of two in vitro cytotoxicity assays-the neutral red (NR) and tetrazolium MTT Tests. Toxicity In vitro, 1988, 2, 1-6.

Chaponi, L. and Migliorini, P. (1999). In: Antibody usage in the lab : immunoblotting. Springer-Verlag Press. pp 32-46.

Cook, D.L., Jonathan, D and Griffiths, G.D. (2006). Retrospective identification of ricin in animal tissues following administration by pulmonary and oral routes. Toxicology 223, 61-70.

Crompton, R and Gall, D. (1980). Georgi Markov-Death in a pellet, Medico-legal. J 48, 48-62.

Godal, A., Olsnes, S. and Pihl, A. (1981). Radioimmunoassay of abrin and ricin in blood. J. Toxicol. Environ. Health 8, 409-417.

Griffith, G.D., Newman, H. and Gee, D.J. (1986). Identification and quantification of ricin toxin in animal tissues using Elisa, J Forensic Sci. Soc 26, 349-358.

Hegde, R., Maiti, T.K. and Poddar, S.K. (1991). Purification and characterization of three toxins and two agglutinins from Abrus precatorius seed by using lactamyl-sepharose affinity chromatography. Analytical Biochemistry 194, 101-109.

Knight, B. (1975). Ricin-A potent homicidal poison. British Medical Journal 1, 350-351.

Kumar, O., Sugendran, K and Vijayaraghavan, R. (2003). Oxidative stress associated hepatic and renal toxicity induced by ricin in mice. Toxicon 41, 333-338.

Leith, A.G., Griffith, G.D. and Green, M.A. (1988). Quantification of ricin toxin using a highly sensitive avidin/biotin enzyme linked immunosorbent assay. J of Forensic Sci. Soc 28, 227-236.

Lord, J.M., Roberts, L.M.and Robertus, J.D. (1994). Ricin structure, mode of action, and some current application, FASEB J 8, 201-208.

Lubelli, C., Chatgilialoglu, A., Bolognesi, A., Strocchi, P., Colombatti M. and Stirpe, F. (2006). Detection of ricin and other ribosome-inactivating proteins by an immunopolymerase chain reaction assay, Analytical Biochemistry 355, 102-109.

Malcolm, A and Brian, P. (1997). A sensitive immunosorbent assay for the detection of ricin in blood, Defence Science and Technology Organisation (DSTO), Report No. DSTO-TR-0572.

Olsnes, S. and Pihl, A. (1982). Toxic lectins and related proteins. In:Molecular action of toxins and viruses. Cohen and Van Heyingen (Eds), Elsevier Biomedical Press, New York. pp. 51-105.

Poli, M.A., Rivera, V.R., Hewetson, J.F. and Merill, G.A. (1994). Detection of ricin by colorimetric and chemiluminescence ELISA, Toxicon 32, 1371-1377.

Rubina, A.Y., Dyukova, V.I., Dementieva, E.I., Stomakhin, A.A., Nesmeyanov, V.A., Grishin, E.V and Zasedatelev A.S. (2005). Quantitative immunoassay of biotoxins on hydrogel based protein microchips. Anal. Biochem. 340, 317-329.

Shyu, R.H., Chiao, D.J., Liu, H.W. and Tang, S.S. 2002a. Monoclonal antibody based enzyme immunoassay for detection of ricin. Hybrid Hybridol. 21, 96-73.

Shyu, R.H., Shyu, H.F., Liu, H.W. and Tang, S.S. (2002b). Colloidal gold based immunochromatographic assay for detection of ricin. Toxicon. 40, 255-258.

Shea, D. and Gottron, F., Ricin: Technical Background and Potential Role in Terrorism, Congressional Research Service, The Library of Congress, http://www.fas.org/irp/crs/RS21383.pdf, updated February 2004, accessed on October 28, 2008.