Fig 2: RTK-dependent ERRα activity in breast cancer cells.(a) Box plot depicting the % of nuclei positive ERRα by immunohistochemical staining in human breast tumour samples classified in various subtypes. Significance determined by one-way analysis of variance (ANOVA; P=0.0487). (b) Heatmap showing the intensities of ERRα binding of peaks determined by ChIP-seq in SKBr3 cells on 90 min stimulation with EGF (100 μM) or HRG (100 μM) or vehicle (veh), and clustered according to overlapping or growth factor-specific peaks by Homer. Below: western blot of ERRα and RPLP0 as loading control. (c) Average binding intensity of ERRα recruitment in the region ±500 bp around the centre of the peaks bound in all conditions (common sites) or induced on growth factor treatment (GF-induced). Statistical significance calculated by one-way ANOVA. (d) Enrichment of canonical pathway analysis classified according to –log(P value) generated by IPA for target genes displaying ERRα recruitment at ±20 kb from their transcriptional start site, in serum-starved conditions or on growth factor treatment. (e) Gene set enrichment of KEGG pathways enriched on depletion of ERRα in SKBr3 cells and specifically observed on treatment of cells with growth factors (false discovery rate<25%). siRNA-mediated depletion of ERRα is monitored by western blot analysis in all the conditions studied (upper left panel). (f) Standard ChIP analysis of genomic enrichment of ERRα recruitment to growth factor-reprogrammed sites in SKBr3 cells on treatment with vehicle or lapatinib. (g) Western blot depicting the expression of ERRα in SKBr3, BT-474 and NIC-5231 breast cancer cells on treatment with lapatinib. (h) Relative mRNA expression of ESRRA, a known target gene of ERRα in SKBr3 cells on lapatinib treatment. (i) Expression of ERRα and ubiquitin in SKBr3 cells treated with lapatinib or veh in the presence or not of the proteasome inhibitor MG-132. All statistical significance is calculated for results in lapatinib-treated cells relative to vehicle (veh) using three independent replicates. Error bars represent s.e.m., statistical significance is calculated using two-tailed unpaired t-test; *P<0.05; **P<0.01.

Fig 3: FMRP and nucleolin interact in both cytosolic high molecular weight and nuclear low molecular weight complexes.(A, B) Native FMRP protein complexes were fractionated by loading the light membrane (A) and nucleolar (B) fractions on a superose 6 size exclusion chromatography column. The absorbance of the column eluent at 280 nm (A280) was plotted against the elution volume (ml). Different FMRP binding partners and markers are shown, including histone H3 and GAPDH as negative controls in the endomembrane and nucleolar fractions, respectively. The elution positions of standard proteins employed include thyroglobuline (669 kDa), ferritin (440 kDa), aldolase (158 kDa), ovalbumin (75 kDa), carbonic anhydrase (29 kDa), ribonuclease (13.7 kDa), and aprotinin (6.5 kDa). The peak fractions, as indicated by a solid line, were subjected to SDS-PAGE and immunoblotting using antibodies against FMRP (71 kDa), nucleolin (76 kDa), CYFIP2 (146 kDa), RPLP0 (34 kDa) and eIF5 (58 kDa) LM, light membrane; Nu, nucleoli. The molecular mass of the peak fractions is indicated above the peaks. (C) Interaction of FMRP with CYFIP, nucleolin, eIF5, and RPLP0 as analyzed by co-immunoprecipitation. Endogenous FMRP was immunoprecipitated from HeLa cell lysates using an anti-FMRP antibody before and after RNase treatment. FMRP co-precipitated with nucleolin, RPLP0, eIF5, and CYFIP2. Interaction with the latter two proteins was sensitive to RNase treatment. Proteins were visualized by using antibodies against FMRP, nucleolin, eIF5, CYFIP2 and RPLP0. N-WASP and GAPDH were used as a negative IP controls. IP, immunoprecipitation; TCL, total cell lysate. (D) Direct interaction between FMRP and nucleolin. GST pull-down experiments were conducted by mixing bacterial lysate expressing His-tagged FMRP fl (upper panel) or FMRP Nterm (lower panel) with different GST-fused nucleolin proteins (RRM1&2, aa 284–466; RRM3&4, aa 467–644; RRM3&4-RGG, aa 499–710; RGG, aa 645–710) immobilized on GSH sepharose beads. Proteins retained on the beads were resolved by SDS-PAGE and processed for Western blot using a monoclonal antibody against FMRP. Mixed samples before performing pulldown (PD) analysis were used as input control. (E) Low-affinity interaction between the FMRP Nterm and the nucleolin RGG. Fluorescence polarization assay was used as a tool for monitoring the interaction of the FMRP Nterm (increasing concentrations as indicated) with the IAEDANS-labeled fluorescent RGG (0.5 μM) (open circles). As negative controls, FMRP Nterm was titrated into IAEDANS alone (0.5 μM) (closed circles). The inset depicts the displacement of FMRP Nterm from IAEDANS-labeled fluorescent RGG by increasing concentrations of unlabeled RGG and the synthetic peptide construct 5(KPR)TASP.

Fig 4: FMRP is localized at various intracellular sites in HeLa cells.Confocal laser scanning microscopy (cLSM) images of HeLa cells depicting endogenous FMRP (green channel) costained with various cytosolic (A) and nuclear (B) markers (red channel), including antibodies against CYFIP2, RPLP0 (ribosomal proteins), nucleolin (nucleolar marker), MTCO2 (mitochondrial protein), NUP62 (nucleoporins), lamin B1 (nuclear intermediate filament proteins), and calreticulin (endoplasmic reticulum marker). Detection of Na+/K+-ATPase and phalloidin staining were used to detect the cellular membrane and F-actin, respectively. DNA was stained by using DAPI (blue channel). Boxed areas in the merged panels depict enlarged areas of interest. Scale bar: 10 μm.

Fig 5: FMRP shows a diverse subcellular distribution pattern in HeLa cells as revealed by subcellular fractionation analysis.(A) Experimental cell fractionation procedure employing several differential centrifugation steps. Cells were fractionated into six distinct fractions, including heavy membrane (plasma membrane and rough endoplasmic reticulum), light membrane (polysomes, golgi apparatus, smooth endoplasmic reticulum), cytoplasm (cytoplasm and lysosomes), enriched nuclear membrane (containing rough endoplasmic reticulum), nucleoplasm, and nucleoli. S, supernatant; P, pellet. (B) FMRP is largely absent in the cytoplasm and nucleoplasm and predominantly localizes to solid compartments. The protein concentrations were normalized in all fractions with exception of the nucleoplasm due its low protein content as compared to the other fractions. In each lane, 5 μg proteins were loaded except for the nucleoplasm, where one μg was used. In addition to FMRP and its binding partner CYFIP, the fractions were analyzed by using different subcellular marker, including Gαq/11, Na+/K+-ATPase and Rac1 (plasma membrane), EEA2 (endosomes), GAPDH (cytoplasm), eIF5 and RPLP0 (ribosomes and rough ER). Nuclear markers included histone H3 and lamin B1. Nucleolin was used as nucleolar marker. (C) Detection of FMRP in mitochondria. The presence of FMRP in isolated mitochondrial fraction was analyzed by SDS-PAGE and immunoblotting, using antibodies against FMRP, two mitochondrial proteins MTCO2 and ACAT1, the cytosolic GAPDH as well as the nuclear proteins lamin B1, histone H3 and nucleolin. Equal protein amounts of the mitochondrial fraction and the total cell lysate were used.