Fig 2: Binding of IGF2BP2 to its mRNA targets is affected by RPSAP52 knockdown. a IGF2BP2 was immunoprecipitated from A673 extracts and the pulled-down RNA was analyzed by RT-PCR. The IGF2BP2 and NRAS mRNAs were used as positive controls. Gel images represent semi-quantitative RT-PCR, whereas data on graphs represent means of two independent RT-qPCR analysis ±SD. b Immunoprecipitation of IGF2BP2 from A673 extracts followed by western blot of retrieved proteins. 10% of total extract prior to IP was loaded as control (input). c IGF2BP2 was immunoprecipitated in control or RPSAP52-depleted A673 cells, and the retrieved proteins and RNAs were isolated and analyzed by western blot (left) or RT-qPCR (right), respectively. Graphs correspond to means from two replicates ±SD. d Analysis of IGF2BP2 binding targets from iCLIP-seq experiments in control (scr) or depleted cells (sh4 B11). The absolute number of peaks mapping to 3'UTR regions were 1762 (scr) and 622 (sh4), and to intronic regions were 973 (scr) and 846 (sh4). Asterisks correspond to P-values < 2.2e–16 (two-tailed Fisher’s tests). e Above: heatmap of genes with differential iCLIP counts on their 3'UTR. Results from two experiments are shown. Below: differential enrichment of these genes according to GO biological process categories (top ten are shown). f Above: Venn diagram showing the relation between genes with significant iCLIP peaks present on their 3'UTR regions in the control (scr) or RPSAP52-depleted (sh4) sample. Below: top ten GO enrichment categories (Biological process) for the same genes in the scr or sh4 sample. g UCSC Genome Browser view of LIN28B 3'UTR with the read coverage from IGF2BP2 iCLIP experiment. Previous IGF2BP2-CLIP data positions are shown in red, and predicted let-7 binding sites are indicated by the arrows. h UCSC Genome Browser view of HMGA2 3'UTR with the read coverage from IGF2BP2 iCLIP experiment. Position of significant peaks and CITS are shown above the profiles, and predicted let-7 binding sites are indicated by the arrows. Source data are provided as a Source Data file

Fig 3: RPSAP52 is abundantly expressed in sarcoma and regulates the LIN28B/let-7 balance. a Upper graph: Pearson coefficient between RPSAP52 expression levels and CGI methylation in the TCGA sarcoma cohort indicates a negative correlation. Lower graph: Pearson’s index indicates a weaker association between RPSAP52 and HMGA2 expression levels in the same cohort. b Upper graphs: RT-qPCR analysis to estimate HMGA2 and RPSAP52 expression levels in a panel of Ewing’s sarcoma and rhabdomyosarcoma cell lines. Expression is relative to GUSB mRNA levels. Graphs represent the mean ±SD of three independent RNA extractions. Lower panel: semi-quantitative RT-PCR analysis of expression in the same cell lines. The two RPSAP52 isoforms are indicated, and the higher-migrating band depicted by an asterisk contains an additional exonic sequence (encompassing coordinates chr12:66,169,917–66,170,002 (hg19)), detected in those cells lines with the highest expression of RPSAP52. c RNA pull-down assays confirm the interaction of RPSAP52 with IGF2BP2 and HNRNPQ in A673 cell extracts. Different truncated fragments of RPSAP52 were assayed as indicated, and the band identified by MS and corresponding to IGF2BP2 and HNRNPQ is indicated on the protein gel (left). Western blot to test the association between RPSAP52 RNA and IGF2BP2 and HNRNPQ proteins (middle panel). The drawing summarizes the data obtained from the pull-downs (right). d Total RNA from A673 stable clones constitutively expressing sh1 or sh4 shRNA sequences was analyzed by RT-qPCR to assess HMGA2 and RPSAP52 transcripts levels (lower graph) or let-7 miRNAs levels (upper graph). Graphs represent the mean ±SD of three independent replicates. e Western blot on A673 clones to analyze protein levels upon stable knockdown of RPSAP52 transcripts. f Growth-inhibitory effect of RPSAP52 knockdown in A673 mice tumor xenografts. Upper graph: tumor volume (n = 10) was monitored over time. Mean values are shown ±SEM. Lower graph: tumors were excised and weighed at 25 days. The photograph shows the relative size of all tumors extracted. Scale bar = 10 mm. Source data are provided as a Source Data file

Fig 4: RPSAP52 expression influences proliferative cellular programs and is a prognosis factor. a Left: Volcano plot indicating differential expression (green = down, red = up) between control (scr) and RPSAP52-depleted (sh4) A673 cells. The vertical green lines correspond to 2.0-fold up and down, respectively, and the horizontal green line represents a P-value of 0.05 (two-tailed unpaired t-test). Right: enriched GO terms for shRNA-RPSAP52-affected genes. The y axis shows GO terms and the x axis shows statistical significance (two-tailed Fisher’s exact test). b RT-qPCR analysis of candidate genes altered in RPSAP52-depleted A673 cells. Clones stably expressing two different shRNAs were analyzed. Graphs represent the mean ±SD of three replicates (two-tailed unpaired student t-test, ***P < 0.001, ****P < 0.0001). c Above: in the TCGA sarcoma cohort, Kaplan–Meier analysis of overall survival indicates that patients with high RPSAP52 expression levels have poorer prognosis than cases with low expression. Below: HMGA2 expression has no prognostic value in the same cohort. Significance of the log-rank test is shown. d Kaplan–Meier analysis of overall survival in the sarcoma cohort from TCGA, indicating that patients with a hypermethylated HMGA2/RPSAP52 promoter display better prognosis. e Summary of the results in the context of HMGA2/IGF2BP2/let-7 axis. RPSAP52 positively regulates HMGA2 expression through both transcriptional and post-transcriptional mechanisms. Binding of RPSAP52 to IGF2BP2 in the cytoplasm might promote downregulation of let-7 levels by LIN28B-dependent and independent mechanisms. This binding could also modulate the formation of mRNPs for a number of IGF2BP2 mRNA targets, thereby directing their translation efficiency. Source data are provided as a Source Data file