“…To determine the specificity of the interaction between TAP and the selected proteins, we performed in vitro binding assays+ [ 35 S]methionine-labeled putative TAP partners were synthesized in vitro in rabbit reticulocyte lysates and assayed for binding to glutathione agarose beads coated with either GST-TAP, GST alone, or various TAP fragments fused to GST+ Binding of RanBP2, p205, and KIAA0169 was not further investigated+ In preliminary experiments we noticed that some of the selected proteins bound to GST-TAP only in the presence of nuclear extracts+ This was the case for CRM1, importin-b, Nup88, Nup93, and RanGAP1 (data not shown)+ Additional experiments indicate that TAP/CRM1 interaction was bridged by CAN and probably by other nucleoporins present in the column (data not shown)+ Because TAP-mediated export is distinct from the CRM1 export pathway (Pasquinelli et al+, 1997;Saavedra et al+, 1997;Zolotukhin & Felber, 1997;Bogerd et al+, 1998;Otero et al+, 1998), we conclude that TAP/CRM1 interaction is not relevant for TAP function+ However, our results indicate that the two proteins share common binding sites at the NPC (see below)+ Binding of in vitro-synthesized transportin, CAN, Nup98, hGle2, and E1B-AP5 to GST-TAP neither required nor was stimulated by addition of nuclear extracts (Figs+ 3, 4, and 5 and data not shown) suggesting that these interactions may be direct, and thus were characterized further+ When GST pull-down assays were performed with various fragments of TAP, hGle2 interacted with the C-terminal domain of TAP (TAP371-619) but not with its N-terminal domain (TAP1-372) (Fig+ 3A, lanes 3-5)+ The C-terminal domain of TAP also interacted with all nucleoporins analyzed in this study (Fig+ 7)+ In contrast, E1B-AP5 associates with the N-terminus (TAP1-372) but not with the C-terminal domain of TAP (TAP371-619; Fig+ 3B, lanes 4 and 5)+ Shorter N-terminal TAP fragments (TAP1-265 and TAP1-185) exhibited a reduced binding affinity for E1B-AP5 (Fig+ 3B, lanes 6 and 7), whereas deletion of TAP's first 60 amino acids severely impaired its binding to E1B-AP5 (TAP61-372; Fig+ 3B, lane 8)+ This observation was supported further by the GST pull-down assay shown in Figure 3C+ In this assay, various TAP deletion mutants synthesized in vitro were tested for their ability to bind to an E1B-AP5 fragment expressed in E. coli as a GST fusion+ This E1B-AP5 fragment (101-619) binds TAP with the same efficiency as the full-length protein (data not shown)+ Deletion of TAP residues 22-55 severely impaired its binding to E1B-AP5 (Fig+ 3C, lanes 4-6), whereas short deletions downstream of position 55 had no effect+ Together, our results indicate that fragment 1-372 represents the E1B-AP5 binding domain of TAP+ Within this domain, the first 55 amino acids appear to contribute substantially to the affinity of the interaction+ Although we can- not rule out that the interaction of TAP with hGle2 and E1B-AP5 is mediated by RNA or by a factor present in the reticulocyte lysate, we consider this possibility unlikely as these interactions were neither prevented by microccocal nuclease treatment nor stimulated by including increasing amounts of HeLa extracts, reticulocyte lysate, or RNA in the binding reactions (data not shown and Fig+ 3D)+ Because the E1B-AP5-binding domain of TAP overlaps with its CTE-binding domain, we next tested the effect of the CTE RNA on TAP/E1B-AP5 interaction+ Full-length TAP and an alanine scan mutant (TAP295) that no longer interacts with the CTE RNA …”