Fig 2: XGboost model predicts the optimal functional SSO binding site to modulate the splicing of NEDD4Le13 in TNBC.(A) SF-specific AS effect probability scores as determined by SHAP analysis. PCBP1 is included as a negative example. Black boxes in the top track indicate the binding sites of the SSOs designed and tested to promote NEDD4Le13 exon inclusion. Gray boxes in the top track indicate the positions of SSOs not predicted to regulate exon inclusion, used as negative controls. Blue heatmaps indicate eCLIP binding from ENCODE cell lines, whereas darker blue indicates a higher binding score. Scale bar: 50 nucleotides. (B) Probability distribution of protein–protein interaction scores between HNRNPL and the rest of the SFs in the Akerman et al, 2015 dataset. P value and probability are highlighted for SRSF7. As described by Akerman et al, 2015, the probabilities are predicted using a Bayesian probability model. The p value was calculated using a Mann–Whitney test. (C,D) HNRNPL and SRSF7 expression levels across the subtypes of breast cancer in TCGA. Basal tumor samples n = 148, Her2+ tumor samples n = 70, Luminal tumor samples n = 666, Normal-type tumor samples n = 30, Solid tissue normal samples n = 108. P values were calculated using a Wilcox test. The box plot center line indicates the median value, upper and lower box edges correspond to the 25th and 75th percentiles, and the whiskers extend to 1.5 * interquartile range from the box edge.

Fig 3: SSO-0205 modulates NEDD4Le13 inclusion, causes cancer cell migratory response to TGFβ, and decreases cell proliferation.(A) PCR measuring NEDD4L isoforms in MDA-MB-231 and MCF10A cells in response to TGFβ stimulation (0, 3, or 6 h) after SSO-0205 treatment (400 nM) (24 h). (B) Normalized RNA-seq counts for HNRNPL and SRSF7 in MCF10A and MDA-MB-231 cells showing significantly higher expression of these SFs in MDA-MB-231 cells. Statistical differences calculated with DESeq2; *≤1 × 10−85 (n = 2 biological replicates). (C) Cell proliferation as measured by Cell TiterGlo® 96-well assay. MDA-MB-231 (top graph) and MCF10A (bottom graph) cells were treated with lipofectamine alone as a control or with the indicated SSO for 48 h (400 nM) (n = 4 biological replicates). (D) Cell cycle phase analysis measured by propidium iodide flow cytometry (G1/G0; S phase and G2). The cell cycle phase for each cell line is represented by % (top graphs). One-way ANOVA analysis was performed to compare differences between control and SSO-0205 treatment, MDA-MB-231 p value = 0.0018, MCF10A p value = 0.1596. A representative Propidium Iodide histogram plot for MDA-MB-231 is shown in the lower panel. Cells were treated with lipofectamine alone or with SSO-0205 (400 nM) for 24 h (n = 3 biological replicates). (E) qPCR quantifying expression levels of the indicated cell cycle-related genes in MDA-MB-231 (top line) or MCF10A (bottom line) cells treated with SSO or lipofectamine alone as a control for 24 h followed by 0, 3, or 6 h treatment with hrTGFβ (n = 3–4 biological replicates). (F) qPCR quantifying expression levels of the indicated migration/invasion-related genes in MDA-MB-231 (top line) or MCF10A (bottom line) cells treated with SSO (400 nM) or lipofectamine as a control for 24 h followed by 0, 3, or 6 h treatment with hrTGFβ (n = 3–4 biological replicates). (G) Transwell migration assay on the indicated cell types (MDA-MB-231 top graph and MCF10A bottom graph), cells were treated with lipofectamine or SSO-0205 (1 µM) for 24 h, followed by overnight TGFβ treatment in the Transwell chamber (n = 2–3 biological replicates). Mean and Standard deviation are represented in C, E–G. Statistical differences were calculated by Student’s t-test vs the corresponding time point at the control (Ctr) group; *≤0.05; **≤0.01; ***≤0.001; #≤0.1. Source data are available online for this figure.

Fig 4: Kock-down of the HNRNPL and SRSF7 promote NEDD4Le13 inclusion in the TNBC cancer cell line.(A) Western blots measuring siHNRNPL (left panel) and siSRSF7 (right panel) transfection efficiency by protein decreased in MCF10A and MB231 cells treated with corresponding siRNA alone or in combination for 70 h at 20 nM. (B) Agarose gel for PCR product measuring NEDD4L isoforms in MDA-MB-231 and MCF10A cells in response to siRNA treatments. (C) Quantification of agarose gels (n = 3 biological replicates). Mean and Standard deviation are represented. Statistical differences between each siRNA treatment group (inclusion or skipping) vs corresponding siControl were calculated by Student’s t-test; *≤0.05; #≤0.1. Source data are available online for this figure.