Cell lines and cell culture
Six HCC cell lines (Huh7, HepG2, Hep3B, MHCC97H, MHCC97L, and MHCC-LM3) were purchased from ATCC (Rockville, Maryland). Cells were continuously cultured in Dulbecco’s modified Eagle’s medium (Sigma–Aldrich, USA) containing 10% fetal bovine serum (FBS) (Gibco, New York, USA), 100 IU/ml penicillin G and 100 μg/ml streptomycin (Sigma–Aldrich, USA). All cell lines were incubated at 37 °C in a humidified atmosphere with 5% CO2.
Plasmid construction, RNA interference and transfection
SRSF10-shRNA, plasmids for SRSF10 overexpression with and without Flag-tag, CDC25A-overexpressing plasmids tagged with and without GFP-Flag, and CDC25A-shRNA targeting various sites were purchased from GeneChem Biotechnology Company (Shanghai, China). The shRNA sequences and primer information are described in more detail in Supplementary Table 1.
The 24/48 h transfection and knockdown efficiency were assessed by immunofluorescence and western blot analysis. The knockdown efficiency was determined using two different RNAi, and the sequence with the higher result was selected for further experiments.
qRT–PCR and RNA sequencing
Total RNA was extracted from cells using an RNeasy Mini kit (Invitrogen, Life Technologies, Carlsbad, CA, USA). Complementary DNA (cDNA) synthesis was performed with a GeneAmp RNA PCR kit (Life Technologies) using 1 µg RNA per sample. qPCRs were set up using the iScriptTM two-step RT–PCR kit with SYBR Green (Invitrogen) and the targeting gene primers (see Supplementary Table 1).
RNA array was performed on an Illumina HiSeq machine with paired-end sequencing under the support of Kangchen Biotech (200,233; Shanghai, China).
Reagents and antibodies
The chemical reagents rabusertib (CHK1-specific antagonist also known as LY2603618; cat. #s2626), and MG132 (proteasome inhibitor; cat. #s2619) were obtained from Selleckchem (Houston, TX, USA). CHX (cycloheximide, protein synthesis inhibitor; cat. #2112 s) was obtained from Cell Signaling Technology (Beverly, Massachusetts, USA). Commercially available antibodies for western blotting, immunoprecipitation, immunohistochemistry, and immunofluorescence staining are described in more detail in Supplementary Table 2.
Mutagenesis
Mutation and truncation of SRSF10 and CDC25A were performed by PCR-based methods. The CDSs of SRSF10 (NM_054016.4) and CDC25A (NM_001789) were synthesized and subcloned into the vector pEGFP-C1 (Clontech, 5' BglII—3' BamHI). Point mutation of CDC25A in Lys150 and Lys169 was introduced to arginine (R), while its Ser178 was replaced by alanine (A) or glutamic acid (E). For SRSF10, deleting the 11–84 aa sequence led to a mutation lacking the RNA recognition motif (RRM), also known as SRSF10 (ΔRRM1). Both SRSF10 and CDC25A wild-type and mutants were validated by DNA sequencing and western blotting with the help of Kangchen Biotech (200,233; Shanghai, China).
Cell proliferation, cell cycle, and invasion assays
Cell proliferation was analyzed using a commercial CCK-8 assay kit (#C0038, Beyotime) and EdU detection kit (#C0075S, Beyotime). Fluorescence-activated cell sorting (FACS) was used to assess the cell cycle with a PI staining kit (#R40432, Sigma). Cell invasion was assessed by the transwell assay with a 6-well insert device (8 μm pore size; Corning Life Sciences, Bedford, MA) and Biocoat Matrigel (BD Biosciences) according to the manufacturer’s instructions.
Cycloheximide (CHX)-based protein stability assay
Cells were treated with 10 μM cycloheximide (CHX) for the indicated periods (0 h, 1 h, 2 h, 4 h, and 8 h) to block protein synthesis. MG132 (20 μM) was also administered to inhibit the proteasome before harvesting. Endogenous and exogenous CDC25A protein expression was then assayed as described previously [19].
mRNA decay assay
To measure mRNA stability, 5 g/ml actinomycin D (Sigma Aldrich, USA) was added to cells to inhibit transcription, followed by incubation for the different times indicated. Total RNA was extracted at each time point and quantitated by RT‒PCR. The transcript levels were plotted to create the appropriate nonlinear regression curves using a one-phase decay equation. Exponential fitting curves were determined to quantify the RNA decay rate constant (y = a*e−kt; where k is the decay rate constant, y is the relative amount of RNA, and t is the time). The rate of mRNA turnover was estimated according to the half-life t1/2 = ln(2)/k. The transcript 18S rRNA, which does not decay over time, was detected as a control.
Protein extraction, immunoprecipitation and western blot analysis
Protein extracts were obtained by washing the cells once with PBS and resuspending them in lysis buffer. Tissues were lysed using a cell disruptor and a complete protease inhibitor cocktail (Cat. #5,056,489,001, Roche).
Immunoprecipitation was performed using protein G-agarose (Millipore, Temecula, CA). Blots were then developed with ECL western blotting reagents (Pierce Biotechnology, Rockford, IL). The signal intensity was quantified with ImageJ (National Institutes of Health, Bethesda, MD).
Immunohistochemistry
Tissues were obtained from the specimens of HCC patients who met the following criteria: 1) informed consent was provided; 2) the patients were clinically and pathologically diagnosed with HCC; 3) the patients did not undergo any neoadjuvant therapy before surgery; and 4) adjacent nonneoplastic liver tissues were available for comparison. A total of 74 paired HCC specimens embedded in paraffin were cut into 5-µm slides. The sections were counterstained as previously described [20].
The degree of immunostaining was reviewed and scored separately by two independent pathologists who were blinded to the histopathological features of the samples. Based on the proportion of tumor cells, the following scores were assigned: 0, no tumor cells; 1, < 10% tumor cells; 2, 10 − 35% tumor cells; 3, 35 − 75% tumor cells; and 4, > 75% tumor cells. The staining intensity was scored as follows: 1, no staining; 2, weakly stained (light yellow); 3, moderately stained (yellow brown); and 4, strongly stained (brown). The staining index was determined by multiplying the staining intensity score by the tumor cell proportion score. A staining index ≥ 6 was considered to indicate high expression of SRSF10, while an index < 6 indicated low expression.
Immunofluorescence staining
Cells grown on coverslips were stained as previously described [20]. The antibodies used are listed in Supplementary Table 2. Images were obtained using a confocal microscope (TCSSP8, Leica Microsystems) equipped with an acousto-optical beam splitter, a 405-nm laser (for DAPI), an argon laser (488 nm for Alexa 488), and a diode-pumped solid-state (DPSS) laser (561 nm).
RNA immunoprecipitation (RIP) and crosslinking immunoprecipitation q-PCR (CLIP-qPCR)
Immunoprecipitation targeting CDC25A pre-mRNA was performed using Magna RIP™ (Catalog No. 17–700, MilliporeSigma, USA) according to the manufacturer’s protocol. The SRSF10 antibody or IgG-immunoprecipitated RNA extracts were reverse-transcribed using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA), and the quantification of target RNA was performed by RT‒qPCR).
In the CLIP study of Flag-SRSF10, we tested the recognition of CDC25A. Three primer pairs for exon 5, exon 6, and exon 7 were designed to evaluate precipitated RNA levels as previously described [21]. The sequences of these primers are presented in more detail in Supplementary Table 1.
Tumor xenograft models
Five-week-old male BALB/c nude mice were randomly divided into six groups. Specifically, xenograft tumors were generated by subcutaneously inoculating the right axillary fossa with 200 μl (1 × 106 cells) of SRSF10, CDC25A(△E6), CDC25A(L), sh-CDC25A(△E6), or sh-CDC25A(FL) in HepG2 cells and control cells. The size of the palpable tumors was recorded every 3 days by measuring the tumor length (L, the longest diameter) and width (W, the shortest diameter), which were recorded six consecutive times. All mice were sacrificed after 35 days. The tumor volume (V) was calculated according to the formula V = 1/2 (L × W2).
Bioinformatic analyses
The survival, differential expression, and correlation of the candidate gene were assessed using the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn), the starBase Pan-Cancer Analysis Platform (http://starbase.sysu.edu.cn/panCancer.php), the Cancer Genome Atlas (TCGA) database (https://cancergenome.nih.gov/), Cancer RNA-Seq Nexus (CRN) database (http://syslab4.nchu.edu.tw/), and NCBI’s Gene Expression Omnibus database (https://www.ncbi.nlm.nih.gov/geo/). PhosphoSitePlus (https://www.phosphosite.org/) was applied to predict the PTM sites. The Ensembl Genome Browser (https://www.ensembl.org/) was used to blast RNA sequences of CDC25A variants. Gene set enrichment analysis (GSEA) of SRSF10- and CDC25A-relevant gene signatures was performed with GSEA software v.2.0 according to a reported protocol [22].
Statistical analyses
Statistical analyses were performed using SPSS 19.0 software (version 19.0; SPSS Inc., Chicago, IL). The results are displayed as the mean ± SD or mean ± SEM, and two-group comparisons were evaluated using Student’s t test.
The relationship between SRSF10 expression and clinicopathological characteristics was analyzed using the χ2 test. On the basis of the correlation between the IHC score and patient survival, the cutoff point of each dataset subgroup was determined using the survminer R package. The “surv-cut point” function, which repeatedly tested all potential cut points to find the maximum rank statistic, was applied to the IHC score. Patients were then divided into high- and low-score groups on the basis of maximally selected log-rank statistics to reduce the batch effect of calculation. In the univariate survival analysis, cumulative survival curves were calculated according to the Kaplan–Meier method, and the survival curves were analyzed using the log-rank test.