Then, cells were incubated with FITC-conjugated anti-rabbit IgG 1 : 50 for 1 h at 37 °C. Fluorescence at 525 nm was measured in a microplate reader Spectramax M2e (Molecular Devices, Sunnyvale). Conidia macerated with liquid nitrogen were used as a sample of the total (extra- and intracellular) GAPDH protein. Conidia without

an immunolabeling treatment were used as the negative control. Conidial suspensions of a wild-type (WT) green fluorescent protein expressing M. anisopliae (2 × 107 conidia mL−1) were used to treat insect wings by immersion for 20 s. The wings see more from Dysdercus peruvianus disinfected previously in 37% H2O2 were placed on the surface of 0.7% water agar and incubated for 8 h at 28 °C to induce conidial swelling and germination. The conidia were counted in five objective fields under a fluorescence microscope and recorded with three replicates of wings. The conidia were counted before VE821 and after washing in 0.05% Tween 20 for 30 s. The following treatments were performed: (1) before immersion of the wings in the conidial suspension – preincubation with bovine serum albumin (BSA) (25 μg mL−1) for 1 h at 37 °C and preincubation with recombinant

GAPDH (25 μg mL−1; from M. anisopliae, Supporting Information, Appendix S1) for 1 h at 37 °C; (2) conidial suspension was treated before immersion of the wings – with anti-CHI2 antisera (1 : 100) for 1 h at 37 °C and anti-GAPDH antiserum (1 : 100, produced with P. brasiliensis GAPDH). Experiments were in triplicate; the means and SEs were determined. Statistical analysis was performed using a t-test. P values of 0.0001 or less were considered statistically significant. The ORF from gpdh1 (GenBank accession number EF050456) predicts a 338 amino acid protein with an estimated MW of 36 kDa and ID-8 a theoretical pI of 8.26. In silico protein domain analysis found no domain other than the expected NAD-binding domain (from Val4 to Cys151) and the C-terminal domain (from Leu156 to Tyr313) typical of GAPDH (Figs 1 and S1; Appendix S2). The putative M. anisopliae GAPDH sequence

had high identity and similarity values with fungal counterparts (Table S1), and a phylogenetic tree was built (Fig. S2) showing a distribution consistent with other orthologs and one intron at a conserved position (Ridder & Osiewacz, 1992; Templeton et al., 1992; Jungehulsing et al., 1994). The M. anisopliae gpdh1 gene is a single copy (Fig. 1a). To characterize possible isoforms of the GAPDH in M. anisopliae, cell extracts were analyzed by 2-D gel electrophoresis [Fig. 1b (A)]. The immunodetection of native GAPDH isoforms was performed using anti-GAPDH P. brasiliensis polyclonal antiserum. The Western blot revealed three reactive isoforms, with pIs of 6.6, 6.8 and 7.0 [Fig. 1b (B)]. A protein with a molecular mass of 36 kDa and pI 7.0 was excised from 2-D gel electrophoresis blots of M. anisopliae mycelial protein extracts.