Inhibitors of CLK protein kinases suppress cell growth and induce apoptosis by modulating pre-mRNA splicing

Abstract

Accumulating evidence has demonstrated the importance of alternative splicing in various physiological processes, including the development of different diseases. CDC-like kinases (CLKs) and serine-arginine protein kinases (SRPKs) are components of the splicing machinery that are crucial for exon selection. The discovery of small molecule inhibitors against these kinases is of significant value, not only to delineate the molecular mechanisms of splicing, but also to identify potential therapeutic opportunities. Here we describe a series of small molecules that inhibit CLKs and SRPKs and thereby modulate pre-mRNA splicing. Treatment with these small molecules (Cpd-1, Cpd-2, or Cpd-3) significantly reduced the levels of endogenous phosphorylated SR proteins and caused enlargement of nuclear speckles in MDA-MB-468 cells. Additionally, the compounds resulted in splicing alterations of RPS6KB1 (S6K), and subsequent depletion of S6K protein. Interestingly, the activity of compounds selective for CLKs was well correlated with the activity for modulating S6K splicing as well as growth inhibition of cancer cells. A comprehensive mRNA sequencing approach revealed that the inhibitors induced splicing alterations and protein depletion for multiple genes, including those involved in growth and survival pathways such as S6K, EGFR, EIF3D, and PARP. Fluorescence pulse-chase labeling analyses demonstrated that isoforms with premature termination codons generated after treatment with the CLK inhibitors were degraded much faster than canonical mRNAs. Taken together, these results suggest that CLK inhibitors exhibit growth suppression and apoptosis induction through splicing alterations in genes involved in growth and survival. These small molecule inhibitors may be valuable tools for elucidating the molecular machinery of splicing and for the potential development of a novel class of antitumor agents.

Figure 1. Cpd-1, Cpd-2, and Cpd-3 inhibit the phosphorylation of SR proteins.

(A) Chemical structures of Cpd-1, Cpd-2, and Cpd-3. (B) MDA-MB-468 cells were treated with Cpd-1, Cpd-2, or Cpd-3 for 3 h. Immunoblot analyses were performed for phospho-SR isoforms (arrows). The positions of standard molecular weight markers are indicated on the left. The identity of each isoform was validated by analyzing cells transfected with the respective cognate SR siRNA (data not shown). (C) MDA-MB-468 cells were treated with Cpd-1 for 3 h. Cell lysates were treated with 3 U of CIP in the presence or absence of phosphatase inhibitors (PhosSTOP phosphatase inhibitor cocktail; Roche) for 1 h at 37°C. Immunoblot analyses were performed for SR and phospho-SR isoforms (arrows). (D) MDA-MB-468 cells were treated with Cpd-1, Cpd-2, or Cpd-3 for 6 h at the indicated concentrations. Nuclear speckles were visualized with an anti-SC35 antibody and detected using fluorescence microscopy. The data shown are representative of two to three independent experiments.

Figure 2. Alterations of S6K pre-mRNA splicing induced by Cpd-1, Cpd-2, and Cpd-3.

(A) MDA-MB-468 cells were treated with Cpd-1, Cpd-2, or Cpd-3 for 72 h. Expression of S6K was detected by RT-PCR. ACTB mRNA expression was evaluated as an internal control. (B) Schematic representation of the alternatively spliced forms of S6K mRNAs upon treatment with the compounds. Each box represents an exon. Yellow boxes represent exons for the kinase domain of S6K and red boxes represent introns including a novel exon. (C) MDA-MB-468 cells were treated with Cpd-1, Cpd-2, or Cpd-3 for 6 h at the indicated concentrations. Quantitative RT-PCR analyses were performed for the expression of S6K mRNA exons 6–8 (exon 7 skipped). The data represent means ± SD from three independent analyses. (D) Cell viability after treatment with Cpd-1, Cpd-2, or Cpd-3. MDA-MB-468 cells were treated with each compound for 72 h at the indicated concentrations.

Figure 3. Correlation between splicing activity and cell growth by CLK1 and CLK2.

(A) Scatter plots comparing in vitro kinase inhibition with cellular splicing alteration of S6K after treatment with Cpd-1, Cpd-2, Cpd-3, and a series of other compounds. The X-axis shows the pIC₅₀ [−log₁₀(IC₅₀)] value for CLK1, CLK2, SRPK1, SRPK2, or SRPK3 inhibition in cell-free enzymatic assays. The Y-axis shows the splicing induction activity defined as the drug concentration (μM) that induced 10% of the copy number of the aberrantly spliced S6K mRNA without skipping exon 7 compared with the copy number of the canonical mRNA (Rc0.1). R²: coefficient of determination; blue arrow: Cpd-1; red arrow: Cpd-2; orange arrow: Cpd-3. (B) Scatter plot analysis comparing GI₅₀ values (concentration required to inhibit growth by 50%) on the X-axis with Rc0.1 values on the Y-axis for the same compounds shown in (A). Blue arrow: Cpd-1; red arrow: Cpd-2; orange arrow: Cpd-3. (C, D) MDA-MB-468 cells were transfected with CLK1 siRNA, CLK2 antisense oligonucleotide (ASO), or control Non-Silencing siRNA (NS) at the indicated concentrations. Cells were harvested after 24 and 48 h. The data represent means ± SD from three independent experiments. (C) The expression levels of CLK1 and CLK2 were measured by quantitative RT-PCR. (D) ASO and siRNA transfection experiments were performed to identify the kinases that altered the splicing pattern of S6K pre-mRNA. RT-PCR analyses of the expression levels of S6K mRNA exons 6–7 (canonical mRNA) and exons 6–8 (aberrantly spliced mRNA lacking skipped exon 7).

Figure 4. Induction of alternative splicing by CLK inhibitors.

(A) Schematic representation of the alternative splicing induced by the CLK inhibitors. RNA-seq reads were aligned to exon junctions and classified as exon skipping (ES)-type splicing or alternative donor/acceptor (ADA)-type splicing. The junction sequences generated in cells treated with an inhibitor are shown as red boxes. (B) Numbers of genes and events with expression of alternatively spliced transcripts induced by Cpd-2 based on the RNA-seq data. MDA-MB-468 cells were treated with 5 μM Cpd-2 for 24 h. The genes or events with expression of ES-type splicing and ADA-type splicing were further classified into genes and events whose transcripts generated frameshifts. (C) Gene Ontology analyses of genes that expressed frameshifted transcripts induced by exon skipping after treatment with Cpd-1 for 24 h. The functional categories (Y-axis) and corresponding P-values (X-axis) are shown.

Figure 5. Rapid decay of mRNA splicing variants in cells treated with CLK inhibitors.

(A) MDA-MB-468 cells were treated with Cpd-1, Cpd-2, or Cpd-3 for 24 h at the indicated concentrations. Total RNA was analyzed by RT-PCR. (B) Schematic representation of the RNA decay assay. MDA-MB-468 cells were treated with each compound for 6 h and cellular RNA was labeled for 1 h. After removal of the labeling reagents, the cells were cultured for a further 24 h. The cells were harvested before, harvest (1), and after, harvest (2), the chase and the labeled mRNA was purified. (C) Loss of labeled mRNAs of the indicated genes after the 24 h chase. The ratios of the amounts of the mRNAs at harvest (2) to those at harvest (1) were calculated for the individual splice isoforms using quantitative RT-PCR. The cells were treated with 50 μM Cpd-1 (left) or 5 μM Cpd-2 (right) before RNA labeling. The data represent means ± SD from three independent analyses. Statistical analyses were performed using an unpaired Student’s t-test (**P < 0.01; ***P < 0.001). s.v.: splicing variant. (D) Cpd-2 was added to MDA-MB-468 cells for 24 h at the indicated concentrations. Immunoblot analyses were performed. (E) MDA-MB-468 cells were treated with CLK inhibitors for 48 h. The cellular DNA contents were determined by flow cytometry. Representative data for three independent experiments are shown.