NTR–cargo direct binding.
The direct binding of the candidate cargoes to the NTRs was analyzed by bead halo assay. From well-characterized proteins that have not been reported as cargoes, (i) proteins ranked high (within the top 15% in the 2nd-Z-rankings or 4% in the 3rd-Z-rankings), around presumptive cutoffs (within about top 15–25% in the 2nd-Z-rankings), or lower and (ii) highly ranked proteins that are suspected as indirect cargoes or false positives based on their well-known features, e.g., PMPCA, GALE, UAP1, NQO2, EEF1A2, RAB2A, RAB8A, S100A4, S100A6, S100A13, and S100P, were selected and analyzed. Proteins in (i) verify the cargo identification and cutoff setting, and proteins in (ii) serve for finding indirect cargoes and false positives. The negative rate of these bead halo assays should be higher than the true overall false positive rate of the SILAC-Tp, because proteins in (ii) were selected preferentially. GST or GST-NTR was attached to glutathione-Sepharose beads, mixed with an extract of E. coli expressing GFP or a GFP-fusion protein, and observed by fluorescence microscopy. Q69L-Ran, which inhibits the NTR–cargo functional binding, was added as appropriate. The contrast of the bead fluorescence between the GST and GST-NTR indicates the binding, and the inhibition of this binding by Q69L-Ran certifies the specificity of the binding; ++ or +, positive; ± or –, negative. Summary of the results (p2–5): The results are summarized in both 2nd- and 3rd-Z-rank order. The 2nd-Z-15% and 3rd-Z-4% cargoes are indicated by cyan, and positive binding (++ or +) is indicated by blue. Trn-1 (p6–8): The GFP-fusion proteins were divided into five groups (A–E) according to the expression levels. Because GFP binds weakly to Trn-1, the concentrations of GFP (control) and GFP-fusion proteins were equalized within each group, and the binding was observed in the same conditions. The images are comparable within a group. Trn-2 (p9): GFP weakly binds to Trn-2, and the concentrations of GFP and GFP-fusion proteins were equalized. The images are comparable. Proteins whose ranks differed substantially between the Trn-1 and Trn-2 Z-ranking were assayed. Imp-13 (p10–13): GFP does not bind to Imp-13, and GFP was added to the control mixture at the highest concentration. Three images (GST, GST-Imp-13, and GST-Imp-13 + Q69L-Ran) for each GFP-fusion protein were acquired under identical conditions, and the background intensities and dynamic ranges were equalized. Trn-SR (p14–16): GFP weakly binds to Trn-SR, and the procedures were similar to those used for Trn-1. The GFP-fusion proteins were divided into four groups (A–D), and the images are comparable within a group. Imp-α/β (p17–20): GFP does not bind to Imp-α or -β, and the procedures were similar to those used for Imp-13. GST-Imp-α2 lacks the N-terminal Imp-β-binding domain. Western blotting (p21–22): The GFP-fusion proteins in the E. coli extracts were relatively quantified by Western blotting using an anti-GFP antibody (Roche). The extracts containing the amounts of protein (ng) indicated at the bottoms were loaded. The arrowheads indicate the expected full-length products. The GFP-moieties including those of the partial products were quantified by chemiluminescence. GFP was used as the standard. The Western blots were replicated more than three times. Accessions and sequences (p23–27): The cDNAs were cloned from a HeLa cDNA library by PCR. The accession numbers of the proteins are listed. If the sequence of a used protein is different from that in the database, the deleted, substituting, or inserted amino acids are indicated by the colors. The sequences that matched perfectly are not presented.