SDS-polyacrylamide gel electrophoresis of venom and protein identification

The proteomic analysis was performed independently on Pc and Pl wasp venom (Supplementary Fig. S1). Venom apparatus were dissected from 50 individuals per sample and glands were collected in 25–50 μl of Ringer’s saline supplemented with a protease inhibitor cocktail (Sigma). Glands were opened to release the venom and centrifuged for 5 min at 500 × g to remove residual tissues.

For 1D SDS-PAGE, samples were mixed with 4× Laemmli buffer containing β-mercaptoethanol (v/v) and boiled for 5 min. Proteins were then separated on a 6–16% linear gradient SDS-PAGE and the gel was silver stained as previously described^7,13.

For isoelectric focusing (IEF), samples were prepared by boiling the protein solution for 5 min with 4% (v:v) of a denaturing solution (0.15 M dithioerythritol, 10% SDS). After cooling, the samples were mixed with an equal volume of a solution containing 9.2 M urea, 0.1 M dithioerythritol, and 2% CHAPS. IEF was performed using slab gel. Slab gels were made on glass tubes 14 cm in length (1.5 mm internal diameter) that were filled with 4% acrylamide, 9.2 M urea, 2% ampholytes [1% pH 3–10 (Pharmacia) and 1% pH 2–11 (Servalytes)], and 2% CHAPS. Isoelectric focusing was run in two steps: a first run at 20 mA, 0.1 W/tube, 700 V for a total of 10,000 V/h, followed by a second run at 20 mA, 0.1 W/tube, 3,000 V for a total of 2,000 V/h. For the second dimension, 6–16% linear gradient SDS-PAGE was used. IEF gels were incubated with 4x Laemmli buffer containing ß-mercaptoethanol and loaded on top of the 1D SDS-PAGE. After separation, proteins were silver stained as previously described^7,13.

Identification of proteins by mass spectrometry was performed on 1D bands and 2D spots excised from the gels as previously described^7,13. MS/MS data analysis was performed with the Mascot software (http://www.matrixscience.com) licensed in house using the combined Pc and Pl unisequences and non-redundant NR (NCBI). Data validation criteria were (i) one peptide with individual ion score above 50 (the Mascot significant identity threshold corresponding to p <0.05 is 45 in our case) or (ii) at least two peptides of individual ion score above 20 (corresponding to 1% probability that a peptide spectrum match is a random event). The Mascot score was calculated as -10*LOG10(P). The calculated FDR (based on an automatic decoy database search) ranged from 0 to 1.4% depending of the individual gel analysis. Mascot analysis was performed with a fragment ion mass tolerance of 0.30 Da and a parent ion tolerance of 0.30 Da. Carbamidomethyl of cysteine was specified in Mascot as a fixed modification, and oxidation of methionine as a variable modification. The maximum missed cleavage allowed was set to 2.