Pharmacological CLK inhibition disrupts SR protein function and RNA splicing blocking cell growth and migration in TNBC

Abstract

Background: Dysregulation of alternative splicing plays a pivotal role in tumorigenesis and metastasis in triple-negative breast cancer (TNBC). Serine/arginine-rich (SR) proteins, essential components of the spliceosome, undergo phosphorylation by Cdc2-like kinase (CLK). Here we explored the impact of pharmacological inhibition of CLK using a novel inhibitor, T-025, on the spliceosome complex and transcriptional responses in relation to cell proliferation and migration in TNBC.

Methods: We evaluated the anti-proliferative and anti-migratory efficacy of T-025 in a spectrum of TNBC cell lines. Fluorescent reporter cell lines and flowcytometry were used to determine the effect of T-025 on cell cycle. Deep RNA sequencing was performed to unravel the differentially expressed genes (DEGs) and alternatively spliced genes (ASGs) upon T-025 treatment. Pulldown/MS was used to uncover the impact of T-025 on SRSF7 interactome. Live-cell imaging and photobleaching experiments were conducted to determine the subnuclear localization of SRSF7-GFP and its dynamic mobility.

Results: T-025 exhibited a potent anti-proliferative effect in a spectrum of TNBC cell lines, particularly in highly proliferative cell lines. Treatment with T-025 induced cell cycle arrest in the G1-S phase, resulting in an increased proportion of aneuploidy cells and cells with 4 N DNA. T-025 significantly inhibited cell migration in highly migratory TNBC cell lines. Deep RNA sequencing uncovered numerous DEGs and ASGs upon T-025 treatment, which were significantly enriched in pathways related to cell division, RNA splicing and cell migration. Pulldown/MS showed that SRSF7 interacted more with nuclear-speckle-residing proteins, while less with RNA helicases and polymerases upon T-025 treatment. Enhanced interactions between SRSF7 and other phosphorylated SR proteins localized at nuclear speckles were also observed. Live-cell imaging indicated that T-025 treatment induced the accumulation of SRSF7-GFP at nuclear speckles and nuclear speckles' enlargement, restricting its protein dynamic mobility.

Conclusions: CLK inhibition using T-025 leads to the accumulation of splicing factors at nuclear speckles and stalls their release to splicing sites, resulting in the RNA splicing reprogramming of a large number of genes involved in cell division, migration and RNA splicing. Our findings provide evidence that T-025 could be a promising therapeutic drug for TNBC patients.

Keywords: Alternative splicing; Cdc2-like kinase; Serine/arginine-rich proteins; Triple-negative breast cancer.

Fig. 1

T-025 exhibits an anti-proliferative effect across a broad range of TNBC cell lines. (A) Log2 mRNA expression level of CLK1, CLK2, CLK3 and CLK4 in breast tumor tissue and normal tissue. (B) TNBC and ER-positive tumors using RNA sequencing data derived from TCGA, and (C) basal A, basal B and luminal breast cancer cell lines using published RNA sequencing resources [41]. *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. (D) Cell viability upon T-025 treatment for 72 h analyzed by the SRB assay in six basal A TNBC cell lines: BT20, HCC1806, MDA-MB-468, HCC1937, HCC70 and SUM229PE, (E) ten basal B TNBC cell lines: SKBR7, Hs578T, SUM159PT, HCC38, MDA-MB-231, BT549, SUM1315M02, MDA-MB-436, SUM149PT and MDA-MB-157, and (F) two luminal TNBC cell lines MDA-MB-453 and SUM185PE. The mean ± SD of three biological replicates is shown. 0.1% DMSO was used as a negative control. (G) IC₅₀ of T-025 in eighteen TNBC cell lines calculated in GraphPad Prism, ranked from low to high. (H) 72-hour proliferation rate of eighteen TNBC cell lines using SRB absorption data of cells treated with 0.1% DMSO. (I) Correlation between IC₅₀ of T-025 and proliferation rate in eighteen TNBC cell lines

Fig. 2

T-025 inhibits cell migration in MDA-MB-231 and Hs578T cell lines. (A) Normalized cell migration speed of MDA-MB-231 and (B) Hs578T cell lines treated with 100, 200, 500 and 1000 nM T-025 for 24, 48 and 72 h. 0.1% DMSO was used as a negative control. (C) Single-cell trajectories were used to visualize cell migration upon 500 nM and 1 μM T-025 treatment for 24 h in MDA-MB-231 and Hs578T cell lines. (D) 72-hour proliferation curves of MDA-MB-231 and Hs578T cell line treated with 100, 200, 500 and 1000 nM T-025. *, p < 0.05; **, p < 0.01; ***, p < 0.001; ****, p < 0.0001. (E) Confocal imaging of F-actin staining using rhodamine phalloidin in MDA-MB-231 and Hs578T cell line treated with 1 μM T-025 for 24 h was used to inspect cell morphology changes (scale bar represents 50 μm). (F) Brightfield images captured with the IncuCyte live cell imaging system over 72 h, showcasing MDA-MB-231 and Hs578t cells that were treated with 0.1% DMSO and 1 µM T-025. Scale bar represents 100 μm

Fig. 3

T-025 impairs the cell cycle G1-S transition, increasing the population of cells containing 4 N DNA. (A) Representative confocal images show MDA-MB-231 and Hs578T FUCCI reporter cells treated with 1 µM T-025 for 0, 12, 24 and 48 h. Cell cycle phases were marked by red (G1), yellow (G1-S transition), green (S-G2-M) and blue nuclei (stained by Hoechst; M-G1). (B) Quantifications of cell cycle phases in MDA-MB-231 and (C) Hs578T treated with 50, 100, 200, 500 and 1000 nM T-025 for 24 and 48 h. The mean ± SD of three biological replicates is shown. 0.1% DMSO was used as a negative control. (D) Flow cytometry shows the distribution of cells containing 2 N, 2–4 N and 4 N DNA content, as well as aneuploidy cells in MDA-MB-231 and Hs578T cells treated with 200, 500 and 1000 nM T-025 for 48 h. (E) Quantification of the percentage of cells containing 2 N (light blue), 2–4 N (medium blue) and 4 N DNA (dark blue) and (F) aneuploid population. The mean ± SD of three biological replicates is shown. 0.1% DMSO was used as a negative control. *, p < 0.05; **, p < 0.01. ***, p < 0.001; ****, p < 0.0001

Fig. 4

T-025 modulates alternative splicing of multiple genes related to RNA splicing, cell division and migration. (A) Volcano plot reveals significantly downregulated and upregulated genes in MDA-MB-231 cells treated with 1 µM T-025 for 24 h. 0.1% DMSO was used as a negative control. Red points represent upregulated genes with log₂FC > 1.0 and an adjusted p-value of < 0.01. Blue points represent downregulated genes with log₂FC < 1.0 and an adjusted p-value of < 0.01. Grey points represent genes with no significant difference. (B) GO annotation analysis shows the top 20 biological processes which enriched in the significantly upregulated and downregulated genes. (C) The number of different ASEs affected by T-025. The number of ASEs with average count > 10,|ΔPSI| >0.1 and FDR < 0.01 were counted and categorized by AS type and|ΔPSI|. (D) The number of ASGs affected by T-025 in the MDA-MB-231 cells. Genes with or without a significant difference in mRNA expression level based on DESeq2 analysis were labelled as blue and purple respectively. (E) GO annotation analysis shows the top 20 biological processes enriched in the ASGs affected by exon skipping with|ΔPSI| >0.4 and FDR < 0.01. (F) Example of Sashimi plot to visualize the alternatively spliced exon of STAG1 and CENPE (middle exon was alternatively spliced). The genomic reads densities (measured in RPKM) and junction reads (plotted as arcs) in each condition are shown. (G) Number of alternative spliced events of CDC25 family members CDC25A, B and C with|ΔPSI| >0.1 and FDR < 0.01. (H) The top 15 most important hub genes of the PPI network of focal adhesion-related genes was identified by CytoHubba in Cytoscape based on MCC algorithm. Significantly downregulated genes were labeled as blue

Fig. 5

T-025 treatment alters SRSF7 interactions with nuclear speckles proteins and RNA machinery. (A) Western blot (WB) analysis of input samples for GFP pulldown/MS in MDA-MB-231 GFP and SRSF7-GFP expressing cells upon 1 µM T-025 treatment for 24 h. Tubulin served as the loading control. (B) GFP pulldown/MS analysis shows significantly upregulated and downregulated proteins in the SRSF7 interactome in MDA-MB-231 SRSF7-GFP expressing cells post 1 µM T-025 treatment for 24 h (|Log2FC|>1, FDR < 0.05). 0.1% DMSO was used as a negative control. Red points represent upregulated SRSF7 interactors with log2FC > 1.0 and an FDR of < 0.05. Blue points represent downregulated SRSF7 interactors with log2FC < 1.0 and an FDR of < 0.05. Gray points represent SRSF7 interactors with no significant difference. Experiments were performed in four independent biological replicates. (C) Gene ontology (GO) annotation analysis shows the biological process, cellular component, and molecular function of the upregulated and (D) downregulated proteins in the SRSF7 interactome. The top 10 terms with significant enrichment (FDR < 0.05) are shown. (E) WB of the input extracts from MDA-MB-231 GFP and SRSF7-GFP expressing cells post 1 µM T-025 treatment for 24 h. Tubulin was used as the loading control. (F) GFP pulldown/WB demonstrates the SRSF7 interactors of interest using GFP-Trap agarose beads. (G) Quantification of the expression level of SRSF7 interactors of interest in GFP pulldown/WB. Mean + SD of three independent biological replicates. Significance was determined using the Student T-test. *, p < 0.05; **, p < 0.01. (H) Distribution of upregulated and (I) downregulated SRSF7 interactors in MDA-MB-231 cells across different spliceosomal subcomplexes

Fig. 6

T-025 enhances the interaction of SRSF7 with a wide range of phosphorylated SR/SR-like proteins. (A) GFP pulldown/WB shows the interaction of phosphorylated SR proteins with SRSF7 in MDA-MB-231 cells treated with 1 µM T-025 for 24 h using anti-phospho-epitope SR proteins antibody, clone 1H4. (B) A volcano plot illustrates the SRSF7 phospho-interactome, as determined by GFP pulldown/MS. The experiments were performed in four independent biological replicates. Significance was determined using the Student’s T-test. Red points represent upregulated phospho-sites with log2FC > 1.0 and an FDR of < 0.05. Blue points represent downregulated phospho-sites with log2FC < 1.0 and an FDR of < 0.05. Grey points represent phospho-sites with no significant difference. (C) A bar graph displays the log2FC of phospho-sites from different SRSFs enriched in the SRSF7 phospho-interactome upon 1 µM T-025 treatment. (D) GO annotation analysis reveals the biological process, cellular component, and molecular function of the upregulated phospho-proteins in the SRSF7 phospho-interactome. The top 10 terms with significant enrichment (p- value < 0.05) are shown. (E) Cytoscape-ClueGo gene network interaction analysis on upregulated phospho-proteins in the SRSF7 phospho-interactome. (F) The top 10 most important hub genes of the protein-protein (PPI) network of upregulated phospho-proteins, identified by CytoHubba in Cytoscape based on MCC algorithm

Fig. 7

T-025 induces accumulation of SRSF7-GFP at nuclear speckles over a 24-hour period. (A) Confocal images show the subnuclear localization of SRSF7-GFP in Hs578T BAC-SRSF7-GFP reporter cells treated with 1 µM T-025 for 24 h. 0.1% DMSO was used as a negative control. (B) Quantification of counts and (C) occupied area of accumulated SRSF7-GFP in speckles, and (D) the ratio of GFP integrated intensity between speckles and the nucleoplasm. Error bars represent the measurements from 9 images captured per group. **, p < 0.01; ****, p < 0.0001. (E) Representative time-lapse confocal images show the SRSF7-GFP accumulation in speckles in Hs578T BAC-SRSF7-GFP reporter cells treated with 1 µM T-025 in 24 h. (F) Quantifications of counts and (G) occupied area of accumulated SRSF7-GFP in speckles, (H) nucleoplasm GFP intensity, (I) the ratio of GFP integrated intensity between speckles and the nucleoplasm and (J) the intensity of total GFP in nuclei. All measurements were normalized to the first imaging time point. (K) Immunofluorescence microscopy images show the localization of SRSF7-GFP and the nuclear speckles marker, SC-35 in MDA-MB-231 and Hs578T BAC-SRSF7-GFP reporter cells treated with 1 µM T-025 for 24 h

Fig. 8

T-025 reduces the dynamic movement of SRSF7-GFP accumulated at nuclear speckles. (A) Representative time-lapse confocal images display frames from a photobleaching experiment in MDA-MB-231 and Hs578T BAC-SRSF7-GFP reporter cells treated with 1 µM T-025 for 24 h. 0.1% DMSO was used as a negative control. (B) The graph shows the normalized fluorescence intensity of SRSF7-GFP in the unbleached and bleached areas of MDA-MB-231 and (C) Hs578T BAC-SRSF7-GFP reporter cells treated with 1 µM T-025 for 24 h. Bars represent the mean ± SEM, n = 16–23 cells for each condition. (D) Representative frames from a photobleaching experiment in Hs578T SRSF7-GFP expressing cells with 1 µM T-025 treatment for 24 h are shown. The half-bleached area is labelled with a red asterisk, and the nuclear speckle and nearby nucleoplasm area were labelled with green and purple asterisks, respectively, in the images. (E) The graph shows the normalized fluorescence intensity of SRSF7-GFP in the bleached area (red), nuclear speckle (green) and nearby nucleoplasm area (purple) in Hs578T SRSF7-GFP expressing cells treated with 1 µM T-025 for 24 h. Bars represent the mean ± SEM, n = 11 cells for each condition