The plant toxin ricin found in the Ricinus communis beans poses a potential bioterrorism threat. Ricin is highly dangerous, with minimum human lethal doses at 0.1-1 µg/kg associated with nonspecific symptoms of intoxication. Robust methods for detecting contaminated environmental or clinical samples at trace level are crucial. Mass spectrometry provides specific and sensitive detection of ricin after selective isolation by immunoprecipitation (IP). Both proteotypic peptides and the depurination activity on a small RNA substrate (adenine release assay) are then monitored on the same sample. This work aims at optimizing the two protocols to lower the limit of detection for ricin.

Isolation of the toxin from milk was done using a KingFisher Duo Prime robot and a ricin monoclonal antibody (mAb) coupled to magnetic beads. The recovered samples were digested by trypsin into peptides or incubated with an RNA substrate (5’-RNA(GCGCGdAGAGCGC)-3’) to release adenine. Peptides and adenine were injected onto C18 reversed-phase and HILIC columns, respectively. The signal was acquired on a Q-Exactive instrument operated in the PRM mode.

To improve the recovery of IP, six types of magnetic beads (SerA, PrtG, Tosylactivated, Sheep-Anti Mouse (SAM), Protein A, and G) were coated with the ricin mAb and compared. The SAM beads provided the best results using ~6 times less mAb than the Tosylactivated beads. Different washing buffers were also evaluated but no significant difference in signal was observed. HILIC and reversed-phase columns were tested for adenine detection. HILIC improved adenine retention (RT 4 min). Finally, milk samples spiked with 20 ng/mL of ricin showed four and three-fold improvements for proteomic and activity-based identifications, compared to previous conditions, respectively. Improving peptide recovery after IP is investigated to detect ricin at concentrations below 1 ng/mL.