Localization of critical domains of EB2 involved in nucleocytoplasmic shuttling of the protein. (A) Schematic representation of EB2 protein. The light gray box represents the NES domain previously identified in the protein (NES B), the dark gray box represents the RNA-binding domain (RBD), the white box represents the REF interacting domain, and the two vertical bars represent the NLS. (B) Heterokaryon assay. EB2 and the various deletion mutants represented in the figure were expressed in HeLa cells from pCI-Flag expression vectors, and a heterokaryon assay was performed. Proteins were then visualized by indirect immunofluorescence, using the anti-Flag M2 MAb (for EB2 and its derivatives). An Alexa Fluor 488-conjugated goat anti-mouse IgG (H+L) antibody was used as the secondary antibody. Cell nuclei were stained with Hoescht 33258. The percentage of heterokaryons in which the proteins were found in both the HeLa and NIH 3T3 cell nuclei was evaluated by counting heterokaryons via conventional microscopy. The percentages are indicated in the right column. External boundaries of deletions are indicated in italics. The black dots at the N termini of mutants NLS.Cter, NLS.ΔB, and NLS.M5 indicate that the SV40 NLS was fused to the N terminus.

The N terminus of EB2 interacts directly with the cellular protein TAP. (A) ³⁵S-labeled TAP was incubated with purified GST or GST-EB2Nter bound to glutathione-Sepharose beads. The bound proteins were analyzed by SDS-PAGE and visualized by autoradiography. RNase treatment was performed where indicated. In lane 1, the equivalent of one-fifth of the TAP-expressing rabbit reticulocyte lysate used in each assay was loaded in the gel (input 1/5). (B) F.EB2Nter was expressed in HeLa cells as indicated. Cell lysates of either mock-transfected cells (lanes 1 to 3) or cells expressing F.EB2Nter (lanes 4 to 6) were immunoprecipitated with M2 anti-Flag MAb resin. The immunoprecipitates were immunoblotted with an anti-TAP rabbit polyclonal antibody (top) or with an anti-Flag polyclonal rabbit antibody (bottom). In lanes 1 and 4, the equivalent of 1/10 of the cell lysate used in each assay was loaded in the gel as an input control.

TAP interacts with a region of EB2 between aa 61 and 140. (A) Schematic representation of EB2Nter and the various peptides and deletion mutants used as GST fusion proteins. External boundaries of deletions are indicated in italics. (B) ³⁵S-labeled TAP was incubated with purified GST or GST fused with the overlapping peptide A, B, or C, depicted in panel A, bound to glutathione-Sepharose beads. The bound proteins were analyzed by SDS-PAGE and visualized by autoradiography. The equivalent of one-fifth of the TAP-expressing rabbit reticulocyte lysate used in each assay was loaded in the gel as an indication of the input (lanes 1, 4, and 7). (C) ³⁵S-labeled TAP was incubated with purified GST, GST-EB2Nter, or the GST-EB2Nter deletion mutants depicted in panel A, bound to glutathione-Sepharose beads. The bound proteins were analyzed by SDS-PAGE and visualized by autoradiography. The equivalent of 1/15 of the TAP-expressing rabbit reticulocyte lysate used in each assay was loaded in the gel as an indication of the input (lane 1).

Several domains of peptide B cooperate for nuclear export. (A) Schematic representation of NLS.B.βgal fusion protein and the various deletion mutants derived from it. The light gray box represent the β-galactosidase open reading frame, the dark gray box represents the SV40 NLS, and the white box represents EB2 aa 61 to 140. The vertical black bar represent NLS1 of EB2. (B) Expression plasmids for the different fusion proteins listed in panel B (left column) were transfected into HeLa cells, and a heterokaryon assay was performed. The percentage of heterokaryons in which the transiently expressed proteins were found in both the HeLa and NIH 3T3 cell nuclei was evaluated by counting the heterokaryons via conventional microscopy. The percentages are indicated in the right column. (C) Amino acid sequence of EB2 peptide B (EBV strain Raji). Amino acids deleted from mutants NLS.B.βgal-Δb3, -Δb5, and -Δb7 are underlined in gray. The locations of the Δ3 and Δ4 deletions are indicated by brackets. Amino acids changed to alanines in the EB2Nter R/A mutant (see panel D) are indicated in bold. (D) ³⁵S-labeled TAP was incubated with purified GST, GST-EB2Nter, or GST-EB2Nter R/A (in which the amino acids indicated in bold in panel C have been mutated to alanines) bound to glutathione-Sepharose beads. The bound proteins were analyzed by SDS-PAGE and visualized by autoradiography. In lane 1, the equivalent of 1/15 of the TAP-expressing rabbit reticulocyte lysate used in each assay was loaded in the gel (input 1/15).

EB2Nter behaves like a dominant-negative mutant of EB2. (A) Schematic representation of pDM128/PL reporter plasmid. The CAT coding sequence is inserted into intronic sequences. The positions of the 5′ and 3′ splice sites (SS) are indicated. In the absence of EB2, only the spliced mRNA generated from pDM128/PL is exported efficiently to the cytoplasm (no CAT expression). The presence of EB2 allows export of the unspliced mRNA (and consequently CAT expression) by a mechanism which involves binding of EB2 to the RNA and recruitment of the TAP-p15 complex via REF. Both the RNA-binding domain (indicated by the small black area) and the REF interaction domain have been localized to the C-terminal half of EB2. A direct interaction between EB2Nter and the TAP-p15 complex is also indicated. (B) Dose-dependent inhibition of EB2-mediated mRNA export by EB2Nter. pDM128/PL was transfected into HeLa cells either alone or together with 0.5 μg of expression plasmid for F.EB2 and increasing amounts (0.01 to 0.5 μg) of expression plasmid for F.EB2Nter, as indicated in the figure. The efficiency of unspliced mRNA export was evaluated by quantification of CAT expression by CAT ELISA (top), and the amounts of F.EB2 and F.EB2Nter proteins expressed in each assay were visualized by Western blotting using the M2 anti-Flag MAb (bottom). The data shown are derived from a single experiment but are representative of four experiments. (C) Relative activities of EB2Nter deletion mutants in EB2-mediated mRNA export inhibition assay. pDM128/PL was transfected into HeLa cells either alone or with expression plasmids for F.EB2 (0.5 μg), F.EB2Nter (0.5 μg), and F.EB2Nter internal deletion mutants (up to 3 μg), as indicated. The F.EB2Nter-Δ4 and -ΔB mutants are both tagged with the SV40 NLS. The efficiency of unspliced mRNA export was evaluated by quantification of CAT expression by CAT ELISA (top). The levels of expression of F.EB2, F.EB2Nter, and F.EB2Nter mutants were evaluated by Western blotting using the M2 anti-Flag MAb (bottom). The data shown are derived from a single experiment but are representative of four independent experiments. (D) Relative activities of EB2Nter and the EB2Nter R/A mutant in EB2-mediated mRNA export inhibition assay. pDM128/PL was transfected into HeLa cells either alone (lane 1) or with expression plasmids for F.EB2 (0.5 μg) (lanes 1 to 3), F.EB2Nter (0.5 μg) (lane 3), and F.EB2Nter R/A (2 μg) (lane 4). The efficiency of unspliced mRNA export was evaluated by quantification of CAT expression, using CAT ELISA (left). The levels of expression of F.EB2, F.EB2Nter, and F.EB2Nter R/A were evaluated by Western blotting using the M2 anti-Flag MAb (right).

Overexpression of TAP rescues EB2-mediated export. pDM128/PL was transfected into HeLa cells either alone or together with expression plasmids for F.EB2 (0.5 μg), F.EB2Nter (0.1 and 1 μg), and F.TAP (1.5 μg), as indicated. The efficiency of unspliced mRNA export was evaluated by quantification of CAT expression by CAT ELISA (top), and the levels of expression of F.EB2, F.EB2Nter, and F.TAP proteins were visualized by Western blotting using the M2 anti-Flag MAb (bottom). The data shown are derived from a single experiment but are representative of four experiments. Statistical tests were performed using data collected from these four experiments. CAT expression was significantly increased when F.TAP was overexpressed at both EB2Nter concentrations used (P < 0.05). *, unspecific band.

Competition for TAP's interaction with EB2Nter by 9G8 alters EB2's shuttling efficiency. (A) Schematic representation of the 9G8 SR protein and the C-terminal deletion mutant 9G8Nter. 9G8Nter retained both the TAP interaction domain and sequences necessary for 9G8 shuttling. (B) 9G8Nter inhibits EB2-dependent export of the unspliced mRNA generated from pDM128/PL (Fig. 6A). HeLa cells were transfected with pDM128/PL either alone or together with expression plasmids for F.EB2 (0.5 μg) and F.9G8Nter (0.5 μg), as indicated. The efficiency of unspliced mRNA export was evaluated by quantification of CAT expression by CAT ELISA (top). The levels of expression of F.EB2 and F.9G8Nter mutants were visualized by Western blotting using the M2 anti-Flag MAb (bottom). Data from a representative experiment out of four independent experiments are presented in the figure. (C) 9G8Nter competition for TAP binding interferes with EB2Nter nuclear export. HeLa cells were cotransfected with pCI.Myc.EB2Nter (0.5 μg) and pCI.F.9G8Nter (0.5 μg) expression plasmids, and a heterokaryon assay was performed. Proteins were visualized by indirect immunofluorescence, using the 9E10 anti-Myc antibody (for detection of Myc.EB2Nter) (a and d) and an anti-Flag polyclonal antibody (Sigma) (for detection of F.9G8Nter) (b and e). An Alexa Fluor 488-conjugated goat anti-mouse IgG (H+L) antibody (Invitrogen) (a and d) and a Fluorolink Cy3-labeled goat anti-rabbit antibody (Amersham) (b and e) were used as secondary antibodies. Cell nuclei were stained with Hoescht 33258 (Sigma) (c and f). NIH 3T3 cell nuclei are indicated by white arrows. A minimum of 20 heterokaryons of each sort (either expressing F.9G8Nter or not) were observed: only one example of each is presented in the figure. The phenotype presented in panel d was found in 75% of heterokaryons expressing 9G8Nter but in only 10 to 15% of heterokaryons that did not express 9G8Nter.

Stimulation of pDM128 unspliced mRNA export by MS2-B and MS2-BC fusion proteins. (A) Schematic representation of pDM128-4xMS2 reporter plasmid. pDM128-4xMS2 is a modified version of pDM128/PL, depicted in Fig. 5, in which four RNA-binding sites for the MS2 coat protein have been introduced within the intron, immediately downstream of the CAT open reading frame. MS2 fusion proteins can thus be tethered to RNA via the MS2 binding sites, allowing the mRNA export properties of the MS2 fusions to be tested. (B) pDM128-4xMS2 was transfected into HeLa cells either alone or together with an expression plasmid for HA.MS2, HA.MS2-B (peptide B; aa 61 to 140), HA.MS2-BC (peptide BC; aa 61 to 200), or F.EB2Nter. The efficiency of unspliced mRNA export was evaluated by quantification of CAT expression by CAT ELISA (top), and the amounts of MS2 fusion proteins and F.EB2Nter expressed in each assay were evaluated by Western blotting (bottom) using either an anti-HA MAb (for the MS2 fusions) or the anti-Flag MAb M2. The data shown are derived from a single experiment but are representative of five experiments. Statistical tests were performed using data collected from these five experiments. Both HA.MS2-B and HA.MS2-BC were significantly different from the HA.MS2 control (P < 0.05), whereas F.EB2Nter was not (P > 0.1).

Model of recruitment of TAP to EB2-containing mRNPs. (A) EB2 first interacts with its target mRNA and recruits REF into the mRNP. (B) TAP recruited by REF onto the mRNP interacts with EB2 at a low affinity. (C) TAP is transferred to the mRNA following its interaction with REF and changes its conformation, thereby interacting with EB2 at a higher affinity.