Fig 2: Transportin-2 binds to ERa via two similar PY-motifs in the AF-1 domain. (A) Schematic representation of the ERa-mNLS deletion mutants and the ERa-mNLS-mPYs. (B) GST pulldown assays were performed with recombinant GST and GST-TNPO2, using HeLa lysates expressing 3 × HA-ERa-mNLS, 3 × HA-ERa?1-mNLS, or 3 × HA-ERa?2-mNLS (left panel), and 3 × HA-ERa-mNLS or 3 × HA-ERa-mNLS-mPYs (right panel). Full-length images of western blots are shown in Supplementary Fig. 12. (C) Immunofluorescence images of 3 × HA-ERa?1-mNLS, 3 × HA-ERa?2-mNLS, 3 × HA-ERa-mNLS, or 3 × HA-ERa-mNLS-mPYs in HeLa cells, in the presence of 10 nM E2 for 3 h. The cells were immunostained for HA (green) and stained with DAPI (blue). Scale bars, 20 µm. The subcellular distribution (%) of the 3 × HA-ERa-mNLS mutants was determined from their nuclear, nuclear and cytoplasmic, and cytoplasmic distribution, calculated from > 300 HA-positive cells. Each data point represents the average of results obtained from three independent experiments and the error bars denote the standard deviation. (D) Immunofluorescence images of HeLa cells co-transfected with 3 × HA-ERa-mNLS and mCherry-M9, in the presence or absence of 10 nM E2, for 2 h. The cells were immunostained for HA (green) and stained with DAPI (blue). Scale bars, 20 µm.

Fig 3: TNPO2 is necessary for the E2-induced cytoplasmic localisation of ERa-mNLS, but does not function as a nuclear export receptor. (A) The effect of knockdown of the importin-ß family, calreticulin, or hikeshi on the subcellular distribution of 3 × HA-ERa-mNLS. HeLa cells were transfected with siRNAs against the importin-ß family, hikeshi, and calreticulin. The next day, the cells were transfected with plasmids expressing 3 × HA-ERa-mNLS. The cells were subsequently treated with 10 nM E2 for 2 h and immunostained for HA. The subcellular distribution (%) of 3 × HA-ERa-mNLS was determined from the nuclear, nuclear and cytoplasmic, and cytoplasmic distribution, calculated from > 300 transfected cells. Each data point represents the average of results obtained from three independent experiments and the error bars denote the standard deviation. (B) Immunofluorescence images of 3 × HA-ERa-mNLS in HeLa cells transfected with TNPO2-specific siRNAs, in the presence or absence of 10 nM E2 for 2 h. The cells were immunostained for HA. Scale bars, 20 µm. (C) TNPO2 recognises 3 × HA-ERa-mNLS in an E2-dependent and RanGTP-sensitive manner. GST pulldown assays were performed with recombinant GST, GST-importin-ß1 and GST-TNPO2, using HeLa lysates expressing 3 × HA-ERa-mNLS, in the presence or absence of 2 µM recombinant 6 × His-RanQ69L. GST-importin-ß1 was used as the negative control. #, non-specific band. CBB stain was used as the loading control for the reaction. The average band intensity of 3 × HA-ERa-mNLS bound to GST-TNPO2 in the presence of E2 was set to 1. Each data point represents the average of four independent experiments and the error bars denote the standard deviation. **P < 0.01, ***P < 0.001 as determined by one-way ANOVA followed by Tukey’s multiple comparison test. Full-length images of western blots and CBB staining are shown in Supplementary Fig. 11.

Fig 4: A model for the nucleocytoplasmic shuttling of ERa, regulated by the importin family. Following cytoplasmic translation, ERa translocates into the nucleus by active transport mediated by importin-a4/ß (1). ERa can also shuttle between the nucleus and cytoplasm by passive diffusion (2). Simultaneously, nuclear export receptor(s) export ERa from the nucleus (3). In the cytoplasm, TNPO2 shields the NLS from importin-a and increases the cytoplasmic retention of ERa in an E2-dependent manner (4).

Fig 5: TNPO1/2 and the ß-catenin nuclear localization signal (NLS) are required for Wnt signaling in vivo.(A) Depletion of tnpo1 and tnpo2 using two different pairs of non-overlapping sgRNAs represses gfp expression in Xenopus tropicalis Tg(pbin7Lef-dGFP) embryos at stage 10. Key used to quantify embryos with whole mount in situ hybridization (WMISH) signal (blue – normal gfp signal nd red – reduced gfp signal). Uninjected control (UIC) embryos were used as a negative control. (B) siRNA mediated TNPO1 and/or TNPO2 knockdown reduces luciferase activity in mouse embryonic fibroblasts that harbor a stable integration of luciferase under the control of TCF/LEF promoters. Wnt signaling was activated by human ß-catenin-GFP overexpression. Control siRNA and GFP were used as negative controls. Experiments were performed in triplicate. (C) Schematic diagram of three ß-catenin constructs used in the double axis assay in Xenopus laevis. ** indicates P687A, M688A substitutions (top left). Dorsal views of X. laevis embryos with anterior to the top (bottom left). Dotted lines indicate the embryonic axis, and the white arrows indicate the head. Histogram of the percent of embryos with secondary axes from three independent replicates. p-Values are from Fisher’s exact test (A) and (C) and unpaired two-tailed t-test (B) where ns is p>0.05, p<0.05 (*), 0.0021 (**), 0.0002 (***), and p<0.0001 (****).The data is uploaded as Figure 6—source data 1. Figure 6—source data 1.Source data related to Figure 6.