Fig 2: Short term hypoxia effects on VEGF mRNA splicing are concomitant with intranuclear redistribution of SAFB1/2 and its dissociation from the nuclear matrix.A Schematic representation of VEGFA (left panel) indicating the alternatively spliced forms of exon 8 resulting in the production of pro-angiogenic VEGFa (red) or anti-angiogenic VEGFb form (green). Arrows show Proximal Splice Site (PSS; red) and Distal Splice Site (DSS; green) selected by the splicing machinery to produce the indicated spliced forms. mRNA fold change levels (right panel) of total VEGFA (magenta) or its alternatively spliced forms (VEGFa: red; VEGFb: green) were determined by RT-PCR in MCF-7 cells incubated at 21% or 1% O2 for 2 h. Results are shown as ratios in relation to the total VEGFA levels in normoxia and represent the mean of three independent experiments performed in triplicates ± s.d. (n = 9; *P < 0.05).B Immunofluorescence microscopy images (left panels) of MCF7 cells grown on coverslips at 21% or 1% O2 for 2 h using antibodies against SAFB1 or SAFB2 (As indicated). Nuclei were stained with DAPI (Scale bars: 10 μM). Graphs (right panels) represent the quantification of SAFB1 (upper panels) or SAFB2 (lower panels) foci/cell, their total area/cell and average size from a total of 80–100 individual cells from two independent experiments in each condition ±s.e.m. (***P < 0.001; **P < 0.01).C Immunoblotting analysis of subcellular fractions (TE: Total Extracts, CY: Cytoplasmic fraction, NP: Soluble nucleoplasmic fraction, RN: Soluble nuclear fraction after RNase treatment, DN: Soluble nuclear fraction after DNase treatment, NM: Nuclear Matrix fraction; see also Fig.ure S1C) of MCF-7 cells incubated at 21% or 1% O2 for 2 h (as indicated) using antibodies against the indicated proteins. SF2, Lamin B, and actin were used as fractionation controls for RNAse treated, nuclear matrix and cytoplasmic fraction respectively.D Quantitative analysis of immunoblots (shown in C) depicting the SAFB1 or SAFB2 ratio (as indicated) between RT (RNase treated) and NM (Nuclear Matrix) fractions after the indicated treatment. Columns represent the mean of three independent experiments ±s.e.m. (n = 3; *P < 0.05, **P < 0.01).

Fig 3: Hypoxic dissociation of SAFB1/2 from the nuclear matrix is fast, HIF-independent and reversible upon reoxygenation.A Western blot analysis of proteins recovered in a soluble fraction (left panels) and a total cell extract (right panels) of MCF7 cells incubated under 1% O2 for the indicated time. Graphs represent the quantitative analysis of immunoblots (soluble fractions) depicting the SAFB1 or SAFB2 ratio over actin (as indicated) after the indicated treatment. Columns represent the mean of three independent experiments ±s.e.m. (n = 3; *P < 0.05, **P < 0.01).B Western blot analysis of proteins recovered in a soluble fraction (left panels) and a total cell (right panels) of MCF7 cells incubated under 1% O2 for 2 h followed by re-oxygenation for the indicated time. Graphs represent the quantitative analysis of immunoblots (soluble fractions) depicting the SAFB1 or SAFB2 ratio over actin (as indicated) after the indicated treatment. Columns represent the mean of two independent experiments ±s.e.m. (n = 2; *P < 0.05, **P < 0.01).C Immunoblotting of soluble proteins of MCF7 cells incubated with 1 mM DMOG for the indicated time.D Immunoblotting of soluble proteins of MCF7 cells overexpressing Flag–HIF–1α under normoxia.E Western blotting of soluble proteins of MCF7 cells incubated at 21% or 1% O2 for 2 h after silencing HIF-1α expression by siRNA (as indicated).F Western blotting of soluble proteins of MCF7 cells incubated at 21% or 1% O2 for 2 h with or without 1 μΜ actinomycin D (as indicated).Images are representative of two (B–F) or three (A) independent experiments.

Fig 4: Short-term hypoxia abolishes the association of SAFB1/2 with essential components of splicing machinery and chromatin organization.A Volcano plots of proteins showing significantly altered association with SAFB1 (left panel) or SAFB2 (right panel) in MCF7 cells kept in 21% O2 or 1% O2 after mass spectrometry analysis of anti-SAFB1 or anti-SAFB2 immunoprecipitates. Graphs depict the difference between the two samples expressed as log2(x) versus their statistical significance expressed as –Log10(p-value) from two biological replicates (P < 0.05). SAFB1/2 levels are shown in purple.B Venn diagram showing the number of proteins exhibiting decreased association with either SAFB1 (magenta) or SAFB2 (yellow) under 1% O2.C-E STRING analysis depicting the different clusters of proteins (as indicated) showing decreased association with either SAFB1 (C), SAFB2 (D), or both SAFB1/2 (E), and under 1% O2. Biological functions are indicated by color as shown. Gene names corresponding to proteins further verified by immunoblotting are indicated in color.F Protein extracts (INPUTS) or anti-SAFB1 (left panels)/anti-SAFB2 (right panels) immunoprecipitates (IP) of MCF7 cells grown at 21% or 1% O2 for 2 h were analyzed using antibodies against the indicated proteins. Images are representative of two independent experiments.G Schematic representation of SAFB1 and SAFB2 domains (as indicated; SAP: SAF-A/B, Acinus and PIAS, HD: Homology Domain, RRM: RNA Recognition Motif, RE/RG: Arginine-Glutamate/Arginine-Glycine) and SAFB1/2 peptides identified by mass spectrometry (Green). Methionine residues (oxidized shown in Red or not shown as vertical Magenta lines) are also indicated in their respective positions.