HCC tissues and paracancerous tissues were collected from 10 participants diagnosed with HCC at Xiangya Hospital, Central South University (Hunan, China), through imaging, serological, or histopathological examinations. The sex ratio was 5 males to 5 females, with an age range of 43-76 years. Patients were recruited between January 1, 2024, and December 31, 2024. HCC and paracancerous tissues were collected and divided into the Tumor and Normal groups. Exclusion criteria were pathologically confirmed diagnosis of non-HCC and patients with incomplete clinical features. Informed consent was obtained from all participants and approval was obtained from the medical ethics committee of Xiangya Hospital, Central South University (Hunan, China; approval no. 2024052119).

The transformed human liver epithelial-2 (THLE-2) cells (CL-0833) were purchased from Procell Life Science & Technology Co., Ltd. SK-HEP-1 (AW-CCH036) and MHCC-97H (AW-CCH088) human liver cancer cells Huh7 (AW-CCH237), were purchased from Changsha Abiowell Biotechnology Co., Ltd. THLE-2 cells were cultured in THLE-2 Cell Complete Medium (CM-0833; Procell Life Science & Technology Co., Ltd.). Huh7, SK-HEP-1 or MHCC-97H cells were cultured in DMEM (containing NaHCO₃ 1.5 g/l), minimum essential medium or DMEM containing 10% fetal bovine serum and 1% penicillin/streptomycin. Logarithmically grown SK-HEP-1 and MHCC-97H cells were collected and spread inside a six-well plate and cells were processed following wall attachment. First, CDC5L was knocked down and divided into the Control, si-NC (negative control siRNA), si-CDC5L-1 (CDC5L-specific siRNA-1) and si-CDC5L-2 (CDC5L-specific siRNA-2) groups. CDC5L was overexpressed and divided into the si-NC, si-CDC5L, oe-NC (Negative control overexpression vector) and oe-CDC5L (CDC5L overexpression vector) groups. Caspase 3 was knocked down and divided into the si-NC, si-CDC5L, si-CDC5L+si-NC and si-CDC5L+si-Caspase 3 (Caspase 3-specific siRNA) groups. ELAVL1 was knocked down and divided into the Control, si-NC, si-ELAVL1#1 (ELAVL1-specific siRNA#1) and si-ELAVL1#2 (ELAVL1-specific siRNA#2) groups. ELAVL1 was overexpressed and divided into the si-NC, si-ELAVL1, oe-NC and oe-ELAVL1 (ELAVL1 overexpression vector) groups. Furthermore, groups were divided as follows: si-NC, si-CDC5L, si-CDC5L+oe-NC and si-CDC5L+oe-ELAVL1. Finally, GSDME was knocked down and divided into the Control, si-NC, si-GSDME#1 (GSDME-specific siRNA#1) and si-GSDME#2 (GSDME-specific siRNA#2) groups. SK-HEP-1 and MHCC-97H cells were treated with 1 μM CDC5L inhibitor CVT-313 (HY-15339; MedChemExpress) for 48 h (27) and divided into the following groups: Control, CVT-313, CVT-313 + si-NC and CVT-313 + si-GSDME (GSDME-specific siRNA). The transfection of all siRNA and overexpression vectors was carried out using Lipofectamine^® 2000 (cat. no. 11668019; Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. The siRNA was used at a concentration of 50 nM. Transfection was performed at 37°C for 48 h. The time interval between transfection and subsequent experimentation was 48 h. si-CDC5L (HG-SH001253), oe-CDC5L (HG-HO001253), si-ELAVL1-1 (HG-SH001419), oe-ELAVL1 (HG-HO001419), si-Caspase 3 (HG-SH032991), si-GSDME and (HG-SH004403) were obtained from HonorGene. The interference sequences are as follows: si-NC: TTAACGGCGAGAGTTTAGGCC, si-CDC5L-1: GGACAGAATTCTGCAGGAAGC, si-CDC5L-2: GCATAAAGCTGTCCAGAAAGA, si-ELAVL1-1: GGTTTGGGCGGATCATCAACT, si-ELAVL1-2: GAGGCAATTACCAGTTTCAAT, si-Caspase 3: GACAACAGTTATAAAATGGAT, si-GSDME: GAGAGGAATTTCCATCCATTT.

Male nude mice aged 4 weeks were supplied by Hunan SJA Laboratory Animal Co., Ltd. A total of 80 mice were used in this study. The mice were housed in a controlled environment with a temperature of 22±2°C, a 12-h light/dark cycle and humidity of 50±10%. The average weight of the mice was approximately 18-20 g at the start of the experiment. The tumor formation model of subcutaneous MHCC-97H cells in nude mice was established. Following a week of acclimatization and rearing, MHCC-97H cells were injected subcutaneously, with different treatment groups receiving corresponding cell types: si-NC, si-CDC5L, oe-NC and oe-CDC5L; si-NC, si-CDC5L, si-CDC5L+si-NC, si-CDC5L+si-Caspase 3; si-NC, si-CDC5L, si-CDC5L+oe-NC and si-CDC5L+oe-ELAVL1. Each nude mouse received an injection of 1×10⁷ cells in a volume of 100 μl, administered into the left axillary region (28). In the si-NC group, mice were subcutaneously injected with si-NC-treated MHCC-97H cells. In the si-CDC5L group, mice were subcutaneously injected with si-CDC5L-treated MHCC-97H cells. In the oe-NC group, mice were subcutaneously injected with oe-NC-treated MHCC-97H cells. In the oe-CDC5L group, mice were subcutaneously injected with oe-CDC5L-treated MHCC-97H cells. In the si-CDC5L+si-NC group, mice were subcutaneously injected with si-CDC5L and si-NC-treated MHCC-97H cells. In the si-CDC5L+si-Caspase 3 group, mice were subcutaneously injected with si-CDC5L and si-Caspase 3-treated MHCC-97H cells. In the si-CDC5L+oe-NC group, mice were subcutaneously injected with si-CDC5L and oe-NC-treated MHCC-97H cells. In the si-CDC5L+oe-ELAVL1 group, mice were subcutaneously injected with si-CDC5L and oe-ELAVL1-treated MHCC-97H cells. Tumor dimensions were assessed twice weekly following implantation. When tumors grew to a diameter of 150-200 mm³, tumor volume and tumor weight were weighted and measured. Tumor volume was tracked and calculated using the formula: V (mm³)=LxW²/2, with L and W denoting the longest and shortest diameters, respectively. The study was terminated 28 days' post-implantation and the tumors were prepared for image capture. To perform sacrifice of mice using cervical dislocation, one hand used a rigid rod to firmly press down on the head and neck area, while the other hand grasped the tail or hind limbs. Then, quickly and forcefully the hindquarters were pulled backwards and upwards to dislocate the cervical vertebrae. Following cervical dislocation, the mice were observed for a ≤5 min to ensure that life was extinct. To determine whether life was extinct, its breathing and heartbeat were observed, while simultaneously checking the pupils and limb reflexes. If the mouse had stopped breathing and its heart had stopped beating following cervical dislocation, the pupils did not react to light and the limbs showed no reflex actions, it was more accurately confirmed that life was extinct.

In addition, following a week of acclimatization, 4-week-old male nude mice were injected subcutaneously with MHCC-97H cells and differently treated cells were injected into corresponding groups: Control, CVT-313, CVT-313 + si-NC and CVT-313 + si-GSDME. The amount of cells injected per nude mouse was 1×10⁷ and a 100-μl volume was injected into the left axilla. Following tumor implantation, tumor measurements were conducted twice a week for observation. When the tumor grew to a suitable size (10 days after planting), the drug was administered in groups according to the tumor volume. Among them, mice in the Control group were subcutaneously injected with normal MHCC-97H cells. Nude mice in the CVT-313 group were subjected to CVT-313 drug intervention. Nude mice in the CVT-313 + si-NC group were injected subcutaneously with transfected si-NC-treated MHCC-97H cells and to CVT-313 drug intervention. Nude mice in the CVT-313 + si-GSDME group were subcutaneously injected with si-GSDME-treated MHCC-97H cells and underwent CVT-313 drug intervention. CVT-313 drug intervention was performed by intraperitoneal injection of CVT-313, administered every other day at a dose of 0.625 mg/Kg in an injection volume of 10 ml/100 g (29). The experiments were terminated after 28 days of tumor seeding and tumors were arranged to be sampled and images captured. All animal experiments were carried out with the approval of the Animal Ethics Committee of Xiangya Medical School, Central South University (approval no. 2024051707) and in compliance with the Declaration of Helsinki (30).

Clinical specimens were categorized into HCC tissues and adjacent non-tumor tissues. Data were obtained from The Cancer Genome Atlas (TCGA, https://xenabrowser.net/datapages/?cohort=TCGA%20Pan-Cancer%20(PANCAN)&removeHub=https%3A%2F%2Fxena.treehouse.gi.ucsc.edu%3A443) data (369 tumor/50 normal). Box plots of differentially expressed CDC5L were drawn. Kaplan-Meier survival analysis was applied to assess the survival outcomes of patients with HCC in the high- and low-CDC5L groups. Gene Set Enrichment Analysis (GSEA; https://www.gsea-msigdb.org/gsea/index.jsp) was applied to explore Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway regulation. Bioinformatics analysis was conducted to the expression of evaluate Caspases 1, 3, 6, 7, 8, 9 and 10 in HCC. Venn diagrams displayed the intersection of CDC5L binding proteins and Caspase 3 interacting RNA-binding proteins (RBPs). Bioinformatics analysis was conducted to measure 22 RBPs levels in HCC. The binding of ELAVL1 with Caspase 3 was shown based on the University of California, Santa Cruz (UCSC) Genome Browser (http://genome.ucsc.edu/cgi-bin/hgTracks?db=hg38&lastVirtModeType=default&lastVirtModeExtraState=&virtModeType=default&virtMode=0&nonVirtPosition=&position=chr4%3A184627598%2D184627850&hgsid=2947477028_2O6buwujau0adLCDAGdFsuwkAUi0).

Based on previous studies, the viability of SK-HEP-1 and MHCC-97H cells was evaluated using the Cell Counting Kit-8 (CCK-8) assay. Briefly, cells were seeded in 96-well plates at a density of 5,000 cells per well and incubated. The CCK-8 reagent (10 μl) was added to each well, and the plates were incubated for 2 h at 37°C. Absorbance was measured at 450 nm using a microplate reader. A clone formation assay was performed to assess the proliferative capacity of SK-HEP-1 and MHCC-97H cells (31). The migration and invasion of SK-HEP-1 and MHCC-97H cells was assessed through Transwell assays (32). For migration and invasion assays, 2×10⁴ cells were seeded in the upper chamber of Transwell inserts with or without Matrigel coating. Transwell inserts were precoated with Matrigel and incubated at 37°C for 30 min. After 24 h, cells that migrated or invaded through the membrane were fixed with methanol and stained with 0.1% crystal violet at room temperature for 5 min. The number of cells was counted in five random fields under a light microscope.

CDC5L, Caspase 7, Caspase 9, Caspase 6, Caspase 8, Caspase 1, Caspase 3, Caspase 10, GSDME and CDC5L mRNA levels were quantified with RT-qPCR. Total RNA was isolated from cells at a density of 80-90% confluence using the TRIzol^® Total RNA Extraction Kit (cat. no. 15596026; Thermo Fisher Scientific, Inc.) and its concentration and purity were evaluated. Subsequently, mRNA was reverse-transcribed into cDNA using mRNA Reverse Transcription Kit (cat. no. CW2569; CWBIO) according to the manufacturer's protocol. Target genes levels were determined on ABI 7900 system (Applied Biosystems; Thermo Fisher Scientific, Inc.) using Ultra SYBR Mixture (CW2601; CWBIO) and calculated by 2^−ΔΔCq method (33), with β-actin serving as internal reference gene. The PCR cycling conditions were as follows: Initial denaturation at 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec, 60°C for 30 sec, and 72°C for 30 sec. The experiments were replicated three times. The primer sequences used are listed in Table I.

Western blotting was applied to assess CDC5L, GSDME, Gasdermin E N-terminal fragment (GSDME-N), Caspase 1, 3, 6, 7, 8, 9 and 10, KH RNA binding domain containing, signal transduction associated 1 (KHDRBS1), nudix hydrolase 21 (NUDT21), heterogeneous nuclear ribonucleoprotein A2/B1 (HNRNPA2B1), RNA binding motif protein X-linked (RBMX), heterogeneous nuclear ribonucleoprotein C (HNRNPC) and ELAVL1 expressions. Total proteins were extracted using RIPA buffer (cat. no. AWB0136; Changsha Abiowell Biotechnology Co., Ltd.). The protein concentration was determined using the BCA Protein Assay Kit (cat. no. 23225; Thermo Fisher Scientific, Inc.). A total of 30 μg of protein was loaded per lane. The proteins were separated via SDS-PAGE using a 10% gel and transferred onto nitrocellulose membranes. Membranes were blocked with 5% skimmed milk at room temperature for 1.5 h, followed by incubation with primary antibodies overnight at 4°C. The primary antibodies used were: CDC5L (cat. no. 12974-1-AP; 1:1,000; Proteintech Group, Inc.), GSDME (cat. no. ab215191; 1:1,000; Abcam), GSDME-N (cat. no. ab222408; 1:1,000; Abcam), Caspase-7 (cat. no. ab25900; 1:1,000; Abcam), Caspase-9 (cat. no. ab2013; 1:2,000; Abcam), Caspase-6 (cat. no. ab52951; 1:20,000; Abcam Plc), Caspase-8 (cat. no. ab227430; 1:1,000; Abcam), Caspase-1 (cat. no. ab286125; 1:1,000; Abcam), Caspase 3 (cat. no. 19677-1-AP; 1:1,000; Proteintech Group, Inc.), Caspase-10 (cat. no. ab2012; 1:2,000; Abcam), KHDRBS1 (cat. no. 10222-1-AP; 1:3,000; Proteintech Group, Inc.), NUDT21 (cat. no. 10322-1-AP; 1:5,000; Proteintech Group, Inc.), HNRNPA2B1 (cat. no. 67445-1-Ig; 1:8,000; Proteintech Group, Inc.), RBMX (cat. no. ab250272; 1:1,000; Abcam), HNRNPC (cat. no. ab133607; 1:20,000; Abcam), ELAVL1 (cat. no. ab238528; dilution, 1:1,000; Abcam) and β-actin (cat. no. AWA80002; 1:5,000; Changsha Abiowell Biotechnology Co., Ltd.). Membranes were incubated with HRP goat anti-mouse/rabbit IgG (cat. no. SA00001-1/SA00001-2; 1:5,000/1:6,000; Proteintech Group, Inc.) at room temperature for 1.5 h. Next, membranes were treated with an ECL reagent (cat. no. AWB0005, Changsha Abiowell Biotechnology Co., Ltd.) for 1 min and then subjected to analysis using an imaging system (Bio-Rad ChemiDoc MP, Bio-Rad Laboratories, Inc.). The protein levels were quantified by Quantity One 4.6.6 software (Bio-Rad Laboratories, Inc.), with β-actin serving as internal control.

Lactate Dehydrogenase (LDH), IL-1β and IL-18 levels were assessed using LDH Assay Kit (cat. no. A020-2; Nanjing Jiancheng Bioengineering Institute), IL-1β Assay Kit (human; cat. no. KE00021; Proteintech Group, Inc.; mouse; cat. no. KE10003; Proteintech Group, Inc.) and IL-18 Assay Kit (human: cat. no. KE00193; Proteintech Group, Inc.; mouse; cat. no. CSB-E04609m; Wuhan Huamei Biotech Co., Ltd.) according to the instructions.

In cells with overexpression/knockdown of ELAVL1, the half-life of Caspase 3 mRNA was assessed using RT-qPCR following treatment with 5 μg/ml Act D for 0, 2, 4, 6, 8 and 10 h. In addition, in cells where CDC5L was knocked down and ELAVL1 was overexpressed, RT-qPCR was employed to measure Caspase 3 mRNA expression following treatment with 5 μg/ml Act D for 0, 2, 4, 6, 8 and 10 h, in order to determine the half-life of Caspase 3 mRNA.

Hoechst 33342 and PI staining was performed to observe pyroptosis. Following the corresponding time of the aforementioned treatments, cells in each group were washed once with PBS, followed by the addition of 1 ml of cell staining buffer. Subsequently, 5 μl of Hoechst 33342 staining solution and 5 μl of PI staining solution were added to the cells. Cells were then gently mixed and incubated for 30 min at 4°C. The staining status of each group of cells was viewed under a microscope and then the fluorescence status of each group was viewed under a microscope and images captured.

In order to investigate interaction between CDC5L and ELAVL1, Co-IP experiments were conducted. SK-HEP-1 and MHCC-97H cells were collected for protein extraction. The cells were lysed in a lysis buffer containing protease inhibitors at 4°C for 30 min. The lysate was cleared by centrifugation at 14,000 × g for 10 min at 4°C. The protein concentration was determined using the BCA protein assay kit (cat. no. 23225; Thermo Fisher Scientific, Inc.). For each immunoprecipitation reaction, 500 μg of protein lysate was incubated with 2 μg of CDC5L antibody (cat. no. 12974-1-AP; Proteintech Group, Inc.) in a total volume of 500 μl of lysis buffer at 4°C overnight. Subsequently, 30 μl of Protein A/G agarose beads (cat. no. sc-2003; Santa Cruz Biotechnology, Inc.) added to IP mixture and allowed to incubate for 2 h at 4°C. The beads were washed three times with 1 ml of lysis buffer to remove nonspecific binding. The entire immunoprecipitate was eluted by boiling in 50 μl of 2X SDS-PAGE loading buffer for 10 min at 95°C. The eluted proteins were separated by SDS-PAGE using a 10% polyacrylamide gel and transferred onto nitrocellulose membrane. Membranes were blocked with 5% skimmed milk at room temperature for 1.5 h. The membrane was probed with primary antibodies against CDC5L (cat. no. 12974-1-AP; dilution, 1:1,000; Proteintech Group, Inc.) and ELAVL1 (cat. no. AWA48828; dilution, 1:1,000; Changsha Abiowell Biotechnology Co., Ltd.) overnight at 4°C. Membranes were incubated with HRP goat anti-rabbit IgG (cat. no. SA00001-2; 1:6,000; Proteintech Group, Inc.) for 1.5 h. Next, membranes were treated with an ECL reagent (cat. no. AWB0005, Changsha Abiowell Biotechnology Co., Ltd.) for 1 min and then subjected to analysis using an imaging system (Bio-Rad ChemiDoc MP, Bio-Rad Laboratories, Inc.). The protein levels were quantified by Quantity One 4.6.6 software (Bio-Rad Laboratories, Inc.).

Cell RNA-binding proteins were immunoprecipitated and fluorescence quantification was performed to detect whether CDC5L (cat. no. 12974-1-AP; Proteintech Group, Inc.) or ELAVL1 (cat. no. 11910-1-AP; Proteintech Group, Inc.) binds to Caspase 3. Briefly, cells were lysed using RIP lysis buffer at 4°C for 30 min. The lysate was cleared by centrifugation at 14,000 × g for 10 min at 4°C. For each immunoprecipitation reaction, 100 μl of cell lysate was incubated with 50 μl of magnetic beads coated with Protein A (according to the manufacturer's protocol) in RIP buffer for 2 h at 4°C. The beads were washed three times with RIP buffer. The isolated RNA was subjected to RT-qPCR analysis, with total RNA used as input controls. Primer sequences were H-Caspase 3, TGG CAACAGAATTTGAGTCCT forward and H-Caspase 3, ACCATCTTCTCACTTGGCAT reverse.

Data analysis was conducted using GraphPad Prism 8.0 software (Dotmatics). Results are presented as the mean ± standard deviation. For comparisons between two groups, Student's unpaired t-test was used, whereas for multiple group comparisons, one-way ANOVA followed by Tukey's post-hoc test was applied. P<0.05 was considered to indicate a statistically significant difference.

First, CDC5L expression was analyzed in HCC based on the TCGA database. CDC5L exhibited high expression levels in the Tumor group compared with the Normal group (Fig. 1A). In addition, Kaplan-Meier survival analysis demonstrated that the overall survival time of patients with HCC in the high-CDC5L group was shorter, as compared with the low-CDC5L group (Fig. 1B). GSEA analysis showed that CDC5L was mainly enriched in Transcriptional Misregulation In Cancer [Normalized Enrichment Score (NES)=1.45; P<0.001; Fig. 1C]. Moreover, the functional pathways of highly expressed CDC5L involved the MAPK signaling pathway, PI3K-Akt signaling pathway, Wnt signaling pathway and others (Table SI). Next, western blot analysis further revealed CDC5L expression in clinical tissues. CDC5L expression was elevated in the HCC group compared with the paracancerous tissue group (Fig. 1D). In addition, CDC5L expression was further validated at the cellular level. CDC5L expression was elevated in human liver cancer cells (Huh7, SK-HEP-1 and MHCC-97H cells) compared with THLE-2 cells. Among them, CDC5L expression was most markedly elevated in SK-HEP-1 and MHCC-97H cells (Fig. 1E and F) and was therefore selected for subsequent experimental studies. The present results indicated that CDC5L expression was upregulated in HCC.

Next, CDC5L was knocked down. Compared with the si-NC group, CDC5L levels were reduced in the si-CDC5L-1 and si-CDC5L-2 groups. Among them, the si-CDC5L-2 group exhibited the most obvious reduction in CDC5L expression (Fig. 2A) and was therefore selected for subsequent experimental studies. Furthermore, CDC5L was overexpressed. For cells transfected with oe-CDC5L alone, RT-qPCR results showed a significant increase in CDC5L mRNA expression, indicating that oe-CDC5L transfection successfully achieved overexpression of CDC5L (Fig. 2B). CDC5L knockdown repressed cell viability, proliferation, migration and invasion, but CDC5L overexpression facilitated cell viability, proliferation, migration and invasion (Fig. 2C-E). In addition, CDC5L knockdown promoted LDH levels, PI positive rate, GSDME and GSDME-N levels and IL-18 and IL-1β levels. CDC5L overexpression suppressed LDH levels, PI positive rate, GSDME and GSDME-N levels and IL-18 and IL-1β levels (Fig. 3A-D). Moreover, the effects of low and high CDC5L expression was verified in HCC in vivo. It was found that, following CDC5L knockdown, the tumor volume was gradually reduced, along with reductions in tumor size and weight. Conversely, following CDC5L overexpression, the tumor volume gradually increased and both tumor size and weight grew larger (Fig. 4A-C). In addition, IL-18 and IL-1β levels were elevated following CDC5L knockdown. By contrast, these cytokine levels were diminished following CDC5L overexpression (Fig. 4D). CDC5L knockdown promoted GSDME and GSDME-N levels but repressed CDC5L levels. CDC5L overexpression suppressed GSDME and GSDME-N levels but elevated CDC5L levels (Fig. 4E). These results indicated that CDC5L regulated HCC pyroptosis.

Based on GeneCards search results, Caspases 1-10 were seen to be involved in the progression of HCC. The Caspase family genes common to both the high and low expression groups of CDC5L were Caspases 1, 2, 3, 6, 7, 8, 9 and 10. According to GeneCards search results, among them, Caspases 1, 3, 6, 7, 8, 9 and 10 mediated the process of pyroptosis (Fig. 5A). The expression of Caspases 1, 3, 6, 7, 8, 9 and 10 was validated using RT-qPCR and western blotting. Caspase 3, a key protein involved in both pyroptosis and apoptosis, exerted tumor cytotoxicity via GSDME upon activation (34). Combining the results of mRNA and protein, it was observed CDC5L knockdown substantially elevated Caspase 3 expression, but CDC5L overexpression result in a markedly reduction in Caspase 3 expression (Fig. 5B). Therefore, Caspase 3 was selected for subsequent experiments. RIP was further used to detect the binding of CDC5L to Caspase 3 mRNA. The results showed that CDC5L did not bind to Caspase 3 mRNA (Fig. 5C). Next, Caspase 3 was used to interfere. For cells transfected with si-Caspase 3 alone, RT-qPCR results showed that compared with the si-NC group, the expression of Caspase 3 mRNA was markedly reduced in both the si-Caspase 3#1 and si-Caspase 3#2 groups. Among them, the si-Caspase 3#2 group exhibited the most pronounced reduction in Caspase 3 mRNA expression and was used for subsequent experiments. This indicated that si-Caspase 3 transfection successfully achieved knockdown of Caspase 3 (Fig. 5D). Compared with the si-NC group, si-CDC5L group exhibited a reduced the expression of CDC5L and increased the expression of Caspase 3. Upon the additional knockdown of Caspase 3, its expression decreased, while CDC5L expression remained largely unchanged (Fig. 5E). Cellular function experiments showed that, compared with si-NC group, cell viability, proliferation, migration and invasion capabilities in the si-CDC5L group were diminished. Following the further knockdown of Caspase 3, cell viability, proliferation, migration and invasion capabilities increased (Figs. 5F and 6A-B). In addition, compared with the si-NC group, LDH, IL-18 and IL-1β levels, PI positive rate and GSDME and GSDME-N protein levels in the si-CDC5L group increased. Following the further knockdown of Caspase 3, LDH, IL-18 and IL-1β levels decreased, PI positive rate decreased and GSDME and GSDME-N protein levels decreased (Figs. 6C-E and S1). Furthermore, the effects of interfering with CDC5L and Caspase 3 were verified on HCC in vivo. It was found that, following the knockdown of histone deacetylase 6 (HDAC6), tumor volume gradually decreased and tumor size and tumor weight also decreased. Following the further knockdown of Caspase 3, tumor volume gradually increased and tumor size weight also increased (Fig. 7A-C). In addition, following the knockdown of HDAC6, LDH, IL-18 and IL-1β levels in the tissue supernatant increased. Following the further knockdown of Caspase 3, LDH, IL-18 and IL-1β levels decreased (Fig. 7D). Following the knockdown of HDAC6, CDC5L expressions in the tumor tissue decreased and Caspase 3, GSDME and GSDME-N expression increased. Following the further knockdown of Caspase 3, there was no significant change in CDC5L expression and Caspase 3, GSDME and GSDME-N expression decreased (Fig. 7E). These results indicated that CDC5L regulated pyroptosis in HCC in a Caspase 3-dependent manner.

The earlier results in the present study indicated that CDC5L downregulated Caspase 3 and reduced the stability of Caspase 3 mRNA. It was proposed in the present study that CDC5L might influence the stability of Caspase 3 mRNA through specific RBPs. Therefore, starting from CDC5L, a search for interacting proteins was performed and starting from Caspase 3 mRNA a search for interacting RBPs was performed; the RBPs were determined by finding the intersection of these two ways. First, the Venn diagram revealed the intersection between CDC5L binding proteins and Caspase 3 interacting RBPs, with a total of 22 (Fig. 8A). Bioinformatics analysis of the expression of these 22 RBPs in HCC was conducted to screen significant RBPs (Fig. 8B). Next, the expression of markedly altered RBP proteins KHDRBS1, NUDT21, HNRNPA2B1, RBMX, HNRNPC and ELAVL1 was validated in clinical tissues. As compared with the Normal group, KHDRBS1, NUDT21, HNRNPA2B1, RBMX, HNRNPC and ELAVL1 levels were elevated in the Tumor group. Among them, the expression of ELAVL1 was the most prominently increased, so it was selected for further analysis (Fig. 8C). Subsequently, ELAVL1 was knocked down and overexpressed. Compared with the si-NC group, ELAVL1 levels in the si-ELAVL1#1 and si-ELAVL1#2 groups were repressed, with the most significant decrease observed in the si-ELAVL1#2 group. As compared with the oe-NC group, ELAVL1 expression in the oe-ELAVL1 group was markedly elevated (Fig. 8D). The present study confirmed that si-ELAVL1 and oe-ELAVL1 were successfully transfected. Based on the UCSC Genome Browser, it was shown that ELAVL1 bound with Caspase 3 (Fig. 8E). RIP further confirmed the binding of ELAVL1 to Caspase 3 mRNA. Following the knockdown of ELAVL1, the binding of ELAVL1 to Caspase 3 mRNA was weakened. Following the overexpression of ELAVL1, the binding of ELAVL1 to Caspase 3 mRNA was enhanced (Fig. 8F). Furthermore, ELAVL1 knockdown inhibited the half-life of Caspase 3 mRNA following treatment with 5 μg/ml Act D for 0, 2, 4, 6, 8 and 10 h. ELAVL1 overexpression extended the half-life of Caspase 3 mRNA following treatment with 5 μg/ml Act D for 0, 2, 4, 6, 8 and 10 h (Fig. 8G). In addition, ELAVL1 knockdown promoted Caspase 3, LDH, IL-1β, IL-18, GSDME and GSDME-N levels. ELAVL1 overexpression repressed Caspase 3, LDH, IL-1β, IL-18, GSDME and GSDME-N levels (Fig. 9A-C). In addition, ELAVL1 knockdown promoted cell pyroptosis, while ELAVL1 overexpression inhibited cell pyroptosis (Fig. 10).

Furthermore, compared with the si-NC group, ELAVL1 expression in the si-CDC5L group was decreased. Compared with the oe-NC group, ELAVL1 expression in the oe-CDC5L group was markedly elevated (Fig. 11A). Co-IP verified the binding of CDC5L to ELAVL1 protein (Fig. 11B). In addition, following CDC5L knockdown, Caspase 3 expression increased. Following CDC5L overexpression, Caspase 3 expression decreased (Fig. 11C). RIP further verified the binding of ELAVL1 to Caspase 3 mRNA. Following CDC5L knockdown, the binding of ELAVL1 to Caspase 3 mRNA was enhanced. Following CDC5L overexpression, the binding of ELAVL1 to Caspase 3 mRNA was weakened (Fig. 11D). Next, the effects of interfering with CDC5L and overexpressing ELAVL1 on HCC pyroptosis was explored in vitro and in vivo. In vitro, CDC5L knockdown suppressed ELAVL1 expression. Upon further overexpression of ELAVL1, its expression increased (Fig. 11E). CDC5L knockdown promoted Caspase 3 expression. Upon further overexpression of ELAVL1, Caspase 3 expression decreased (Fig. 11F). RIP verified the binding stability of ELAVL1 protein with Caspase 3 mRNA. CDC5L knockdown enhanced the binding stability of ELAVL1 protein with Caspase 3 mRNA. Upon further overexpression of ELAVL1, it inhibited the binding stability of ELAVL1 protein with Caspase 3 mRNA (Fig. 12A). In addition, CDC5L knockdown promoted an increase in LDH, IL-1β, IL-18, GSDME and GSDME-N levels. Upon further overexpression of ELAVL1, it suppressed LDH, IL-1β, IL-18, GSDME and GSDME-N levels (Fig. 12B-C). Furthermore, CDC5L knockdown promoted cell pyroptosis. Upon further overexpression of ELAVL1, it inhibited cell pyroptosis (Fig. 12D). In addition, CDC5L knockdown extended the half-life of Caspase 3 mRNA following treatment with 5 μg/ml Act D for 0, 2, 4, 6, 8, 10 h. Overexpression of ELAVL1 inhibited the half-life of Caspase 3 mRNA following treatment with 5 μg/ml Act D for 0, 2, 4, 6, 8, 10 h (Fig. 12E). In vivo, it was found that following CDC5L knockdown, tumor volume gradually decreased and tumor size and weight were also reduced. Following further overexpression of ELAVL1, the tumor volume gradually increased and the tumor size and weight also increased (Fig. 13A-C). In addition, CDC5L knockdown promoted an increase in LDH, IL-1β, IL-18, CAPS3, GSDME and GSDME-N levels. Following further overexpression of ELAVL1, it suppressed CAPS3, LDH, IL-1β, IL-18, GSDME and GSDME-N levels (Fig. 13D - Forward). These results indicated that CDC5L competitively bound to ELAVL1 to inhibit the binding of ELAVL1 with Caspase 3 mRNA, thereby regulating pyroptosis in HCC.

Finally, the effect of CDC5L inhibitor CVT-313 and GSDME interference on HCC pyroptosis and tumor progression was investigated the in vitro and in vivo. RT-qPCR was performed on cells transfected with si-GSDME alone and the results showed that compared with the si-NC group, the expression of GSDME mRNA was markedly reduced in both the si-GSDME#1 and si-GSDME#2 groups. Among them, the si-GSDME#1 group exhibited the most pronounced reduction in GSDME mRNA expression and was used for subsequent experiments. This indicated that si-GSDME transfection successfully achieved knockdown of GSDME (Fig. 14A). At the cellular level, compared with the Control group, GSDME, GSDME-N and Caspase 3 levels were elevated, while those of CDC5L and ELAVL1 were reduced in the CVT-313 group. Following the additional knockdown of GSDME, GSDME and GSDME-N levels decreased, whereas Caspase 3, CDC5L and ELAVL1 levels remained largely unchanged (Fig. 14B-C). Cellular function experiments revealed that, as compared with the Control group, cell viability decreased and proliferation, migration and invasion capabilities were reduced in the CVT-313 group. Upon further knockdown of GSDME, cell viability increased and proliferation, migration and invasion capabilities increased (Figs. 14D-E and 15A). In addition, as compared with the Control group, LDH levels in the cell supernatant and pyroptosis increased in the CVT-313 group. Upon further knockdown of GSDME, LDH levels in the cell supernatant decreased and pyroptosis reduced (Fig. 15B-C). At the animal level, relative to the Control group, tumor volume decreased, tumor size reduced and tumor weight was alleviated in the CVT-313 group. Upon further knockdown of GSDME, tumor volume gradually increased and tumor size and weight also increased (Fig. 16A-C). In addition, as compared with the Control group, IL-1β, IL-18, LDH, Caspase 3, GSDME and GSDME-N levels increased, while CDC5L and ELAVL1 levels decreased in the CVT-313 group. upon further knockdown of GSDME, IL-1β, IL-18, LDH, GSDME and GSDME-N levels decreased, with no significant changes in CDC5L, ELAVL1 and Caspase 3 levels (Fig. 16D-E). These results indicated that CDC5L/ELAVL1 silencing regulated Caspase 3/GSDME to promote HCC pyroptosis and repress tumor progression.