The KKU-213A cell line was established from a 58-year-old male diagnosed with iCCA at the Srinagarind Hospital, Khon Kaen University, Thailand, which the patients’ informed consent and the research protocol (HE621403) was formerly approved by the Ethics Committee for Human Research of Khon Kaen University [24]. KKU-213A is a highly invasive cell line and is OV-associated. TFK-1 was purchased from the cell bank of the RIKEN BioResource Research Center (BRC). This cell line was derived from a 63-year-old male with eCCA [25]. KKU-213A was cultured in Gibco™ Dulbecco’s Modified Eagle Medium (DMEM) (Thermo Scientific, Waltham, MA, USA), while TFK-1 was cultured in Gibco™ Roswell Park Memorial Institute (RPMI) 1640 Medium (Thermo Scientific, Waltham, MA, USA). All media contained 10% (v/v) Gibco™ fetal bovine serum (FBS, Thermo Scientific, Waltham, MA, USA), 100 Unit/mL penicillin and 100 µg/mL streptomycin. Cells were cultured in a humidified incubator (Thermo Scientific, Waltham, MA, USA) at 37 °C with a 5% CO₂ atmosphere. The authenticity of the cell lines was validated using Short Tandem Repeat (STR) DNA profiling. All cell lines were free of mycoplasma and were periodically tested using the Mycoalert assay (Lonza, Rockland, ME, USA).

The two SRPK inhibitors, SRPIN340 and SPHINX31, were purchased from Cayman chemical (Cayman Chemical Company, Ann Arbor, MI, USA). Both were dissolved in dimethyl sulfoxide (DMSO) (Merck KGaA, Darmstadt, Germany) and stored at 4 °C. Cells were treated with various concentration of SRPK inhibitors (10 and 20 µM) for 18 h, with 0.5% (v/v) DMSO used as the vehicle control (0 µM of SRPK inhibitors).

The on-web analytic tool Gene Expression Profiling Interactive Analysis (GEPIA) based on The Cancer Genome Atlas (TCGA) and the Genotype-Tissue Expression (GTEx) datasets (http://gepia.cancer-pku.cn/) was used for survival analysis of SRPK1 and SRPK2 expression in CCA. Cholangiocarcinoma is abbreviated as CHOL in GEPIA, while liver hepatocellular carcinoma is abbreviated as LIHC. The number of CHOL samples in the dataset is 36 tumor tissues and 9 normal tissues, while the number of LIHC samples is 369 tumor tissues and 160 normal tissues. Box-plot analysis was used to illustrate differences in SRPK1 and SRPK2 expression (transcripts per million, TPM) between tumor (T) and normal (N) tissues. The relationship between SRPK1 or SRPK2 expression and the overall survival of CHOL patients was also analysed.

Cells were cultured in 96-well plates (1.6 $\times$ 10⁴ cells per well) for 24 h to reach confluency of adhered cells. After treatment with SRPK inhibitors (0, 10 or 20 µM) for 18 h, the cells were stained and analyzed using the Live/Dead double staining kit (Merck KGaA, Darmstadt, Germany). The cells were first centrifuged at 1500 rpm for 5 min at 25 °C. A 50 µL aliquot of buffer (1 µL of 1 µM Cyto-dye, 1 µL of 1 mg/mL propidium iodide, and 1000 µL of staining buffer) was then added, followed by incubation at 37 °C for 15 min. Fluorescently stained cells were subsequently examined under a fluorescence microscope (Zeiss, Oberkochen, Germany) and the number of dead cells (red fluorescent stain) counted.

Cells were cultured in a 24-well plate (1 $\times$ 10⁵ cells per well) for 24 h and then treated with SRPK inhibitors (0, 10 or 20 µM) for 18 h. An equal volume of cell suspension (5 $\times$ 10⁴ cells) was mixed with 75 µL of MUSE™ AnnexinV and Death Cell Reagent (Cytek Biosciences, Fremont, CA, USA). After 20 min incubation at room temperature in the dark, the cells were injected into capillary flow cells. For analysis, the cells (events) were gated, counted and subsequently collected into four quadrants with positive or negative AnnexinV and positive or negative 7-AAD. The proportions of live cells, dead cells and apoptotic cells (early and late stages) were analyzed as the percentage of total events using the Muse® Cell Analyzer and attached analytical software (Merck KGaA, Darmstadt, Germany).

Cells were cultured (5 $\times$ 10⁴ cells per chamber) in an 8-well cell culture slide for 24 h and then treated with SRPK inhibitors (0, 10 or 20 µM) for 18 h. They were subsequently fixed with 4% (v/v) paraformaldehyde (Merck KGaA, Darmstadt, Germany) for 30 min at room temperature, permeabilized with 0.2% (v/v) Triton-X (Merck KGaA, Darmstadt, Germany) for 5 min, and non-specific binding was blocked by incubation with 1:20 FBS for 20 min. The cells were then incubated overnight at at 4 °C with the following primary antibodies: 1:100 rabbit anti-human Cytochrome C (Cell signaling Technology, Danvers, MA, USA), and 1:1000 mouse anti-human SRSF1 (Thermo scientific, Waltham, MA, USA). This was followed by incubation with the respective secondary antibodies: 1:100 goat anti-rabbit IgG-FITC (Merck KGaA, Darmstadt, Germany), or 1:100 goat anti-mouse IgG-Cy3 (Merck KGaA, Darmstadt, Germany). The cells were then counter stained with a 1:10,000 dilution of 4^′,6-diamidino-2-phenylindole (DAPI) (Merck Millipore, MA, USA) to stain the cell nuclei, and examined under a fluorescence microscope (Zeiss, Oberkochen, Germany).

For the extraction of whole cell lysate proteins, total cells from each treatment group were lysed in radio-immunoprecipitation (RIPA) buffer (Thermo scientific, Waltham, MA, USA) containing phosphatase inhibitor (Merck KGaA, Darmstadt, Germany) and protease inhibitor (Thermo scientific, Waltham, MA, USA). For protein fractionation, cells were lysed in Buffer A (10 mM HEPES KOH pH 7.9, 1.5 mM MgCl₂, 10 mM KCl, 0.1% NP-40, 0.5 mM DTT) and centrifuged at 5000 rpm for 1 min at 4 °C to collect cytoplasmic proteins in the supernatant. The pellets were subsequently extracted with Buffer B (50 mM HEPES KOH pH 7.9, 10% glycerol, 420 mM NaCl, 5 mM MgCl₂, 10 mM KCl, 0.1 mM EDTA, 1 mM DTT), sonicated, and then centrifuged at 15,000 rpm for 15 min, at 4 °C to collect nuclear proteins in the supernatant [26]. The Bradford assay was used to measure the protein concentration of extracted or fractionated samples. Bradford solution (Bio-Rad laboratories, Hercules, CA, USA) (200 µL) was mixed with 1 µL of protein sample (10-fold dilution), or with increasing concentrations (0, 0.1, 0.2, 0.3, 0.4, and 0.5 mg/mL) of bovine serum albumin (BSA; Capricorn scientific, Ebsdorfergrund, Germany) used as a standard. Absorbance was measured at 595 nm to calculate the protein concentration.

Equal amounts of protein from each treatment group were separated using SDS-PAGE (5% stacking gel and 12–15% separating gel), transferred to PVDF membrane (Bio-Rad laboratories, Hercules, CA, USA), and blocked using 5% BSA (Capricorn scientific, Ebsdorfergrund, Germany). The PVDF membranes were incubated overnight at 4 °C with the following primary antibodies: 1:1000 rabbit anti-human cleaved caspase-3 (Cell signaling Technology, Danvers, MA, USA), 1:10,000 rabbit anti-human GAPDH (Merck KGaA, Darmstadt, Germany), 1:1000 mouse anti-phosphoepitope SR protein (Merck KGaA, Darmstadt, Germany), 1:1000 rabbit anti-human beta-actin (Cell signaling Technology, Danvers, MA, USA), 1:10,000 mouse anti-human SRSF1 (Thermo scientific, Waltham, MA, USA), or 1:1000 rabbit anti-human Lamin B1 (Cell signaling Technology, Danvers, MA, USA). They were subsequently probed with HRP-conjugated goat anti-mouse IgG (Merck KGaA, Darmstadt, Germany), or HRP-conjugated goat anti-rabbit IgG (Merck KGaA, Darmstadt, Germany) for 1 h. Protein bands were detected using the enhanced chemiluminescence (ECL) detection system (Bio-Rad laboratories, USA) and imaged by ImageQuantTM LAS 500 (GE Healthcare Life Science, Amersham, UK). The immobilon® ECL ultra-western HRP substrate (Merck KGaA, Darmstadt, Germany) was used to enhance the signal strength of nuclear fractionated proteins samples. The signal intensity of each band was quantified using Image J software (Version 1.54, National Institutes of Health (NIH)’s Center for Information Technology, Bethesda, MD, USA) and the PVDF membranes were imaged using the Amersham ImageQuant™ 800 biomolecular imager (Cytiva, Amersham, UK). Lastly, the signal intensity of each band was quantified by ImageJ software and normalized to GAPDH, beta-actin or Lamin B1 protein expression.

RNA samples were isolated from harvested cells in each treatment group using the E.Z.N.A.® Total RNA Kit 1 (Omega Bio-Tek, Inc., Norcross, GA, USA). The RNA concentration was measured by NanoDrop 2000 (Thermo scientific Wilmington, DE, USA) and 1 µg was used to synthesize complementary DNA (cDNA) with the iScript™ Reverse Transcription Supermix (Bio-Rad Laboratories, Hercules, CA, USA) according to the manufacturer’s instructions. Quantitative PCR was performed to determine the mRNA expression level of target genes. Reaction mixtures (20 µL) were prepared using 200 ng cDNA, 0.4 µM of each of the forward and reverse primers, and 1X MyTaqTM HS Red Mix (Bioline Reagents Limited, London, UK). Sequences for the PCR primers of each target gene are shown in Table 1. Thermocycling conditions consisted of 5 min pre-denaturation at 95 °C, 40 cycles of 95 °C for 30 sec, 62 °C for 30 sec and 72 °C for 30 sec for cycling amplification, and 72 °C for 5 min for final extension. PCR amplification products were mixed with 6$\times$ fluorescent DNA staining reagent (Novel Juice, Gene DireX, Taoyuan, Taiwan) and analysed by 3% agarose gel electrophoresis. Band intensities were evaluated by Amersham ImageQuant™ LAS500 (GE Healthcare Life Science, Amersham, UK) and quantitated using Image J software, with beta-actin used for semi-quantitative normalization.

Data were presented as the mean $\pm$ standard deviation (SD) of triple biological replicates. The statistical significance of differences between two groups was determined with the Student’s t-test, with p 0.05 considered to indicate statistical significance.

Intrahepatic CCA is a malignant epithelial neoplasm characterized by biliary differentiation within the liver. The anatomical origins of intrahepatic CCA overlap with those of liver cancer, thus limiting the application of differential screening by ultrasound and Computed Tomographic scan. Furthermore, the cancer registry database of the National Cancer Institute of Thailand reports that combined liver and bile duct cancer is the most common cancer type in the Thai population. The expression level (TPM counts) of SRPK1 and SRPK2 in tumor (T) and normal adjacent (N) tissues from the TCGA databases were compared using the GEPIA on-line web tool. In CHOL, both SRPK1 and SRPK2 expression were significantly higher in T compared to N tissue (Fig. 1A). In LIHC, the average expression of both SRPK1 and SRPK2 was higher in T compared to N tissues, but the difference was not statistically significant (Fig. 1B). The correlation of gene expression with the overall survival of CHOL patients was also analyzed. High SRPK1 expression was associated with shorter overall survival (50% survival after 10–15 months vs. 40 months for low expression), with a hazard ratio for death of 2.3 (Fig. 1C). High SRPK2 expression was not significantly related to overall survival (Fig. 1D).

Fig. 1.

Expression of SRPK1 and SRPK2 (transcripts per million, TPM), and correlation with the survival of CHOL patients (TCGA dataset, GEPIA analysis). Box-plot analysis comparing the expression levels of SRPK1 (A) and SRPK2 (B) in the cancer tissue of CHOL and LIHC patients. Tumor tissue (T; red box), normal adjacent tissue (N; grey box). Overall survival of CHOL patients according to high and low expression of SRPK1 (C), and SRPK2 (D). High SRPK1 expression was associated with a significantly worse survival rate (HR = 2.3) (C), whereas high SRPK2 expression showed no significant association with survival (D). *p 0.05. CHOL, cholangicarcinoma; LIHC, Liver hepatocellular carcinoma; TCGA, The Cancer Genome Atlas; GEPIA, Gene Expression Profiling Interactive Analysis; SRPK, serine/arginine protein kinase; HR, hazard ratio.

We next studied the effect of SRPK inhibitors on the induction of CCA cell death. The differential live/dead cell count was monitored by dual staining. In viable cells, the esterase enzyme catalyzes calcein into calcein-am, which emits green fluorescence. Moreover, PI passes through the nuclear membrane and binds to the DNA of dead cells, emitting red fluorescence. In both KKU-213A (Fig. 2A) and TFK-1 (Fig. 2B) cells, the confluence of stained viable cells decreased following treatment with SRPK inhibitors (as observed by fluorescence microscopy, 20$\times$ and 40$\times$), whereas the number of red fluorescent cells increased compared to the vehicle control. Both SRPK inhibitors caused a dose-dependent increase in the number of dead KKU-213A cells, especially SPHINX31 which induced cell death even at 10 µM (Fig. 2C). For TFK-1 cells, the number of red fluorescent cells increased significantly in all treatment groups, but the dose-dependent effect was observed only with SPHINX31 treatment (Fig. 2D).

Fig. 2.

Effects of SRPIN340 and SPHINX31 on the induction of CCA cell death. Fluorescence microscopy images of live/dead cell dual staining of CCA cells treated with the SRPK inhibitors SRPIN340 and SPHINX31 (20$\times$ and 40$\times$ magnification). Viable cells stain green fluorescent and dead cells stain red fluorescent. KKU-213A (A), and TFK-1 (B). Counts for red fluorescent cells are shown in the bar graph, with the Y-axis for the dead cell count in KKU-213A (C), and TFK-1 (D) cells. Statistical analysis was performed on data from three independent experiments. Scale bar = 100 µm. *p 0.05 and **p 0.01. CCA, Cholangiocarcinoma; DMSO, dimethyl sulfoxide.

The dead cell phenotype of CCA cells following treatment with SRPK inhibitors was investigated by flow cytometry after AnnexinV/7-AAD staining. Treated cells were stained and classified into four staining quadrants. High annexinV staining but low 7-AAD staining was gated as early apoptosis, whereas positive staining for both annexinV and 7-AAD was gated as late-stage apoptosis. For KKU-213A, an increasing number of early apoptotic cells was observed after treatment with the SRPK inhibitors, especially with 20 µM SPHINX31 (Fig. 3A). For TFK-1, dose-dependent effects were observed with both SRPIN340 and SPHINX31, with a $>$50% apoptotic cell population observed after treatment with 20 µM SPHINX31 (Fig. 3B).

Fig. 3.

Effect of SRPIN340 and SPHINX31 on CCA cell apoptosis. AnnexinV/7-AAD staining coupled with flow cytometry was used to gate apoptotic cell populations after SRPIN340 or SPHINX31 treatment of KKU-213A (A) and TFK-1 (B) cells. The percentage of cells in each staining quadrant was calculated and is shown in the bar graph. Early apoptosis is represented as grey color, and late apoptosis as black color. Immunofluorescence (IF) was used to track mitochondrial leakage of cytochrome C, thus allowing the evaluation of apoptotic induction in KKU-213A (C) and TFK-1 (D) cells following treatment with SRPIN340 or SPHINX31. The white arrows indicate the diffuse pattern of cytoplasmic cytochrome C. Western blot analysis of cleaved caspase-3 was performed to assess the induction of caspase-dependent apoptosis in KKU-213A (E) and TFK-1 (F) cells treated with SRPIN340 or SPHINX31. Statistical analysis was performed on data from three independent experiments. Scale bar = 100 µm. **p 0.01, and ***p 0.001.

Cytochrome C is a mitochondrial protein that is released into the cytosol to form the apoptosome, which subsequently activates caspase enzymes. IF and confocal microscopy were used here to assess the intracellular diffusion of cytochrome C, with green fluorescence representing the localization of cytochrome C and merging with DAPI to mark the nuclei. Untreated controls revealed perinuclear staining with cytoplasmic spots (granular pattern) representing cytochrome C localized in the mitochondria. Following treatment with SRPIN340 or SPHINX31, a diffuse cytoplasmic staining pattern was observed for cytochrome C in KKU-213A (Fig. 3C) and TFK-1 (Fig. 3D) cells. Moreover, lower confluencies were observed for the treated cells in a dose-dependent-manner.

Caspase-3 is a pro-apoptotic protein. Once cleaved, it is responsible for the majority of proteolysis that occurs during apoptosis. The expression of cleaved caspase-3 was determined here by Western blotting. The results showed trends for increased levels of cleaved caspase-3 relative to controls following the treatment of KKU-213A and TFK-1 cells with SRPK inhibitors (Fig. 3E,F). Quantification of the protein band intensity revealed that treatment with 20 µM SPHINX31 significantly increased the level of cleaved caspase-3 expression in TFK-1 cells (Fig. 3F).

SRPKs are responsible for the phosphorylation of SRSFs, thereby inducing the nuclear localization of SRSFs and their ability to act as splicing factors. We investigated the effect of SRPK inhibitors on the phosphorylation of SRSF proteins in CCA cells by performing Western blot analysis with an anti-phosphoepitope SR protein antibody. These phosphoepitopes are present in each of the SRSF1-SRSF12 domains, but can be distinguished by differences in the molecular weight (MW) of the various pSRSFs.

The apoptotic cell populations and apoptotic protein activation in TFK-1 cells were investigated following treatment with SRPIN340 and SPHINX31 (Fig. 3B,D,F). Decreased expression of proteins with an approximate MW of 32, 35 and 40 kDa was observed following treatment with 10 µM and 20 µM SRPIN340, and with 10 µM SPHINX31, relative to vehicle controls (Fig. 4A). These protein bands were predicted to be pSRSF1/pSRSF12, pSRSF2/pSRSF7, and pSRSF5/pSRSF10, respectively, according to their MW. SRPIN340 significantly decreased the phosphorylation of pSRSF5/pSRSF10, pSRSF2/pSRSF7 and pSRSF1/pSRSF12 in a dose-dependent manner. Treatment with 10 µM SPHINX31 significantly decreased the phosphorylation of pSRSF2/pSRSF7 and pSRSF1/pSRSF12, with the trend increasing at 20 µM SPHINX31 (Fig. 4B).

Fig. 4.

Inhibitory effects of SRPIN340 and SPHINX31 on the phosphorylation and nuclear translocation of splicing factors in TFK-1 cells. Western blot analysis of SRSFs with anti-phosphoepitope SR protein antibody in TFK-1 cells treated with SRPIN340 or SPHINX31 (A). The band intensity for each phospho-SRSF was calculated and the SRSFs were identified by their predicted molecular weight (B). Immunofluorescence (IF) staining of the predominant SRSF1 is represented as red fluorescence, while the cell nuclei are marked by DAPI staining (C). Western blot analysis of SRSF1 protein derived from cellular fractionation revealed the accumulation of SRSF1 following treatment with SRPK inhibitors (D). Statistical analysis was performed on data from three independent experiments. *p 0.05, **p 0.01, ***p 0.001. SRSF, serine/arginine rich-splicing factor; DAPI, 4^′,6-diamidino-2-phenylindole.

SRSF1 is the predominant SRSF expressed in various cancer types. Of note, pSRSF1/pSRSF12 was significantly decreased after treatment of TFK-1 cells with SRPK inhibitors (Fig. 4A,B). Therefore, a concentration of 20 µM of SRPK inhibitor was chosen to investigate the translocation of SRSF1 from the cytoplasm to the nucleus. IF and Western blotting were employed to study the effect of SRPK inhibitors on this translocation,

Under fluorescence microscopy, red fluorescence represents the localization of SRSF1 and is merged with DAPI staining to mark the nucleus. SRSF1 was mainly localized in the area with DAPI-marked nuclei in the vehicle control cells, but was dispersed into the cytoplasmic area after treatment with 20 µM SRPIN340 or SPHINX31 (Fig. 4C). To confirm the cytoplasmic vs. nuclear accumulation of SRSF1, Western blotting of SRSF1 protein was performed after subcellular protein fractionation. Relative to the vehicle control, higher SRSF1 levels were found in the cytoplasmic fraction and lower levels in the nuclear fraction after treatment with SRPK inhibitor, especially for 20 µM SPHINX31 (Fig. 4D).

After showing that treatment with SRPK inhibitors reduced SRSF phosphorylation (Fig. 4A) and the nuclear translocation of SRSFs (Fig. 4B), we next investigated changes in gene splicing errors. SRPK inhibitor treatment of TFK-1 cells clearly induced apoptosis (Fig. 3B,D,F). We therefore investigated splicing errors in the apoptotis-related genes BIN1, MCL1 and BCL2. The results showed that treatment with SRPIN340 or SPHINX31 significantly decreased the level of anti-apoptotic BIN1+12A, but not BIN1 all-forms (Fig. 5A,B), possibly explaining the restoration of cellular apoptosis. Furthermore, the levels of pro-apoptotic MCL-1S (Fig. 5C,D) and BCL-xS (Fig. 5E,F) isoforms were increased, thus enabling the cellular apoptosis of TKF-1 cells.

Fig. 5.

Effect of SRPIN340 and SPHINX31 on the correction of splicing errors in apoptosis-related genes in TFK-1 cells. Reverse transcription-polymerase chain reaction (RT-PCR) was performed to quantify splicing of the apoptosis-related genes BIN1, MCL1 and BCL2. The level of the anti-apoptotic BIN1+12A isoform was compared to that of BIN1_All, comprising all isoforms (A), with the normalized intensity shown in the bar graph (B). The level of pro-apoptotic MCL-1S was also compared to that of MCL-1L (C), with the normalized intensity shown in the bar graph (D). The level of pro-apoptotic BCL-xS was compared to that of BCL-xL (E), with the normalized intensity shown in the bar graph (F). Statistical analysis was performed on data from three independent experiments. *p 0.05, **p 0.01, and ***p 0.001. NTC, No template control.