Fig 2: XE7 influences alternative splice site selection of SRp20 in synergy with ASF/SF2 in HEK293 cells in vivo. (A) Structure of the SRp20 minigene. Indicated are: alternative exon (grey) and introns (horizontal lines). Lines connecting the exons indicate the splicing patterns, either including or omitting the alternative exon. (B) XE7 promotes exon 4 exclusion (i) and addition of ASF/SF2 leads to a further decrease in the form with exon 4 included (ii). An increasing amount of pCMV-HA-XE7 (0–3 µg) was cotransfected with 1 µg of the minigene. Parental vector was added to ensure that similar amounts of cDNAs were transfected. One µg of EGFP-ASF/SF2 was added in (ii). The ethidium bromide stained agarose gels are from representative experiments. Sizes of the bands are shown in bp on the right side of the gels. M = 100 bp DNA marker. (C) Splicing results of three independent experiments using pCMV-HA-XE7 (i) and of 3 independent experiments using pCMV-HA-XE7 with or without (w/o) adding 1 µg ASF/SF2 to the splicing reaction (ii). Error bars indicate ± S.E.

Fig 3: Localization of XE7 in the nucleus (A) Endogenous localization of XE7 in HeLa cells. XE7 was detected with anti-XE7 rabbit polyclonal antibody (i). XE7 oligopeptide antigen (5 µg/ml) competes out XE7 (ii). (B) Colocalization of endogenous XE7 and SR proteins in Calu-6 cells using anti-XE7 and anti-SR. Yellow in the merged image indicates colocalization of XE7 with SR-proteins. (C) Effect of blocking transcription. HeLa cells transfected with pCMV-HA-XE7 were treated with a-amanitin (50 µg/ml) for 8 h, followed by addition of anti-HA and anti-SC35 antibodies. Yellow in the merged image indicates colocalization of XE7 with SC35 in transcriptionally inactive cells. The nucleus was stained with DAPI. Scale bars: 5 µm.

Fig 4: Interaction of XE7 with ZNF265 and ASF/SF2 in yeast two-hybrid assays. Yeast were cotransformed with DNA-binding-plasmids (vector pGBK = bait) and activation-plasmids (vector pGAD or pACT =prey) encoding the proteins indicated, and colonies were grown on SD (-A/-H/-L/-W) dropout plates prior to ß-galactosidase assay. p53 and T-antigen interaction was used as a positive control, and T-antigen and LamC as negative control. ZNF265SF = ZNF265 short form, ZNF265LF = ZNF265 long form. XE7?RS = XE7 with the RS-rich region deleted.

Fig 5: Interaction of XE7 with ZNF265SF and with ASF/SF2 by coimmunoprecipitation. (A) Coimmunoprecipitation of XE7 and ZNF265SF. HeLa cells were transfected with pCMV-HA-XE7 and pCMV-Myc-ZNF265SF, and Calu-6 cells with pCMV-HA-XE7 and pEGFP-ZNF265SF and cell extracts (500 µg) immunoprecipitated with 5 µg antibodies or control IgG (negative control) (lanes 3 and 6) using Protein A agarose. Ten microlitres of the complexes were resolved on SDS–PAGE gel, blotted and the presence of XE7 was confirmed with anti-HA antibody (lane 2). The presence of ZNF265SF was confirmed with a GFP antibody (lane 4). Positive control was cell extract only (lanes 1 and 5). (B) Coimmunoprecipitation of XE7 and ASF/SF2. HeLa cells were transfected with pCMV-HA-XE7 and EGFP-ASF/SF2 and cell extracts (500 µg) were immunoprecipitated with 5 µg antibodies or control IgG (lanes 3 and 6) using Protein A agarose. Ten microlitres of the complexes were resolved on SDS–PAGE gel, blotted, and the presence of HA-XE7 was confirmed with anti-HA (lane 1). Positive control was cell extract alone (lane 2). The presence of ASF/SF2 was shown when incubating the cell extract with anti-GFP in the immunoprecipitation, followed by detection of the blot with the same antibody (lane 5, upper band). Positive control for ASF/SF2 was cell extract alone (lane 4).