This article is mentioned in:

Introduction

Pyroptosis is a recently discovered form of lytic cell death that is characterized by cell swelling and formation of pores and large bubbles on the plasma membrane (1). Pyroptosis is predominantly stimulated by the activation of canonical caspase-1 (Casp1) and noncanonical caspase-4/5/11 (2), leading to the release of N-terminal fragments of pyroptosis execution protein, gasdermin D (GSDMD) (3). GSDMD translocates to the membrane leading to membrane perforation and subsequent infiltration of extracellular content and cell swelling (3). GSDMD is a member of the gasdermin family, which comprises gasdermin A, gasdermin B, gasdermin C, gasdermin E (GSDME) and autosomal recessive deafness type 59 protein (DFNB59) (4). The N domains of the gasdermin family with the exception of DFNB59 are capable of inducing pyroptosis (3,5). In particular, GSDME is cleaved by caspase-3 (Casp3) and switches apoptosis to pyroptosis in several cancers, which is associated with the side effects of chemotherapy and antitumor immunity (6) in breast cancer (7), lung cancer (8), gastric cancer (9) and colorectal cancer (10). Growing evidence suggests that numerous chemotherapeutic drugs, such as cisplatin and etoposide (6), may activate pyroptosis via canonical or noncanonical inflammasome pathways, including TLR and MAPK pathways (11), inhibit cancer progression and overcome apoptosis resistance, which are major causes of cancer resistance (12).

Radiotherapy is an effective nonsurgical treatment for various solid tumors including colorectal cancer (13). Radiotherapy involves multiple pathways that regulate apoptosis, such as Bcl-2 or nuclear factor-κB pathways (14), but the regulatory mechanisms of colorectal cancer after radiation have not been fully elucidated. Pyroptosis is induced in normal tissues and cells, including the bone marrow (15), brain (16), lung (17), intestine (18) and macrophages (19), in response to radiation, as well as in tumor cells in response to chemotherapeutic drugs, such as cisplatin or 5-fluorouracil (8,9). Nevertheless, the effects of radiotherapy on pyroptosis in tumors, especially colorectal cancer and the underlying mechanisms remain unclear.

Long non-coding RNAs (lncRNAs) are non-protein coding transcriptional units that modulate various biological processes, such as gene transcription, pre-mRNA processing and splicing, transport, translation and degradation (20). LncRNAs have been implicated in breast, hepatocellular, colorectal and pancreatic cancer (21). Notably, a growing body of evidence suggests that lncRNAs serve a crucial role in the radiosensitivity of tumors and cells (22,23), especially radiation-induced DNA damage in colorectal cancer. For instance, Zhou et al (24) reported that WWC2 antisense RNA 1 functions as a novel competing endogenous RNA (ceRNA) that regulates fibroblast growth factor 2 expression by sponging microRNA (miR)-16 in radiation-induced intestinal fibrosis. Zou et al (25) reported that OIP5 antisense RNA 1 (OIP5-AS1) regulates radioresistance in colorectal cancer cells by targeting dual specificity YAK1-related kinase via miR-369-3p.

Several reports have demonstrated that lncRNAs regulate the mediators of pyroptosis, including NLR family pyrin domain containing 1 (NLRP1) (26), NLR family pyrin domain containing 3 (NLRP3) (27), Casp1 (28) and Casp3 (29). Nevertheless, the non-coding RNAs that regulate GSDMs, the key execution proteins of pyroptosis have not been identified.

Hence, the present study aimed to investigate the regulatory roles of lncRNAs implicated in ionizing radiation (IR)-induced pyroptosis in colorectal cancer cells by targeting the GSDME. The present study provides novel insights into the molecular mechanisms underlying IR-induced damage in cancer radiotherapy and may lay a theoretical foundation for further therapeutic development in colorectal cancer.

Materials and methods

Cell culture and radiation treatment

293T cells and human colorectal cancer HCT116 cells (American Type Culture Collection) were cultured in Dulbecco's modified Eagle's medium (Gibco; Thermo Fisher Scientific Inc.) containing 10% fetal bovine serum (PAN-Biotech GmbH.), 100 U/ml penicillin and 100 µg/ml streptomycin at 37°C. The cells were irradiated with 0, 8, 12 and 16 Gy at 25°C with a dose rate of 87 cGy/min using a 60Co g-ray source at Beijing Institute of Radiation Medicine (Beijing, China).

miRNA prediction

The potential regulatory miRNAs that bind to the 3′-untranslated region (3′-UTR) of GSDME were screened for using the miRDB (http://mirdb.org/), miRWalk (http://mirwalk.umm.uni-heidelberg.de/), microRNA.org (http://www.microrna.org/) and TargetScan databases (http://www.targetscan.org/).

Plasmids, miRNA mimics and small interfering (si)RNAs

miRNA expression plasmids were generated by cloning pre-miRNA and flank sequences into pcDNA3.0 (Invitrogen; Thermo Fisher Scientific Inc.) as described previously (30) and used in luciferase reporter assays and reverse transcription-quantitative PCR (RT-qPCR) experiments. miR-375, miR-379, and miR-448 mimics were obtained from Shanghai GenePharma Co., Ltd. GSDME was silenced using a siRNA targeting human GSDME (siGSDME) (31). Nuclear paraspeckle assembly transcript 1 (NEAT1) was silenced using siRNA targeting human NEAT1 isoform-1 (siNEAT1-1), isoform-2 (siNEAT1-2) or both isoforms (siNEAT1-1+2) (32). The pre-miRNA cloning, PCR primer, miRNA mimics and siRNA sequences are listed in Table SI. The scrambled negative control RNAs for miRNA mimics and siRNA were obtained from Shanghai GenePharma Co., Ltd. The control vector pcDNA3.0 was used for luciferase reporter assays and RT-qPCR analysis of miRNAs.

Transfection

293T or HCT116 cells were seeded at a density of 3-5×105 cells/well in 6-well plates and cultured to 50% confluence at 37°C. Cells were then transfected with miRNA mimics or siRNAs at 100 pmol/well using Lipofectamine 2000® reagent (Invitrogen; Thermo Fisher Scientific Inc.) according to the manufacturer's instructions. The RT-PCR and western blotting experiments were performed at 48 h post-transfection, or the HCT116 cells were irradiated at 12 h post-transfection followed by subsequent experiments as indicated.

Luciferase reporter assays

The GSDME 3′-UTR reporter plasmid was generated by inserting the 3′-UTR fragment of human GSDME into pGL3-basic vector (Promega Corporation) using XbaI/NdeI. The GSDME 3′-UTR reporter plasmids with a mutation in either one or both of the two miR-448-binding sites were generated using the QuickMutation kit (Beyotime Institute of Biotechnology). 293T cells were seeded at a density of 1×105 cells/well in 24-well plates. The next day, cells were transfected with 100 ng luciferase reporter plasmid, 1 ng Renilla luciferase expression plasmid (pRL-TK; Promega Corporation) and 400 ng miR-370, -375, -376, -379, -448, -452 expression plasmid or control vector pcDNA3.0 plasmid using Lipofectamine 2000® reagent (Invitrogen; Thermo Fisher Scientific Inc.) according to the manufacturer's instructions. After 48 h, the cells were lysed and assessed following the Dual-Luciferase Reporter Assay System (Promega Corporation) protocol. Luminescence measurements were obtained using an FB12 luminometer (Berthold Detection Systems GmbH). Relative luciferase activities were calculated by normalizing the ratio of firefly/Renilla luciferase to that of control cells.

Cell viability assay

HCT116 cells were seeded at a density of 5×103 cells/well in a 96-multiwell plate and irradiated or transfected with GSDME siRNA, NEAT1 siRNAs or miR-448 mimics. HCT116 cells were cultured for 72 h and subsequently analyzed using the Cell Counting Kit-8 assay (Dojindo Molecular Technologies, Inc.) according to the manufacturer's instructions. Briefly, 100 μl mixture (CCK-8 reagent:medium, 1:10) was added into each well and the cells were incubated at 37°C for 1.5 h, then absorbance at 450 nm was measured using a microplate reader (LabSystems Multiskan MS; Thermo Fisher Scientific Inc.).

Pyroptotic cell observation

HCT116 cells were irradiated at 0, 8, 12 and 16 Gy, and subsequently cultured for 48 h. Pyroptotic cells were observed under a phase-contrast microscope (Nikon Corporation) and five fields per group were assessed.

Lactate dehydrogenase (LDH) release assay

LDH released from cultured HCT116 cells into media was measured using the CytoTox 96 Non-Radioactive Cytotoxicity Assay (Promega Corporation) according to the manufacturer's instructions. LDH release was calculated as % LDH release=(sample LDH activity-background LDH)/(total LDH activity-background LDH) ×100%.

Flow cytometry

HCT116 cells were irradiated for 24 h, harvested, dissociated into single cell suspensions with 0.25% trypsin at 37°C for 5 min and stained with APC-conjugated Annexin V and 7-aminoactinomycin D (7-AAD) using the Annexin V-APC staining kit (cat. no. 62700-80; BioGems International, Inc.) according to the manufacturer's instruction. Cells were analyzed by flow cytometry using a FACSCalibur (BD Biosciences). Data were further analyzed using FlowJo v10.0 (FlowJo, LLC).

RT-qPCR

Total RNA was extracted from cultured HCT116 cells using TRIzol® (Invitrogen; Thermo Fisher Scientific Inc.). cDNA was reverse transcribed using the ImProm-II Reverse Transcription System (Promega Corporation) according to the manufacturer's instructions. Quantitative polymerase chain reaction (PCR) amplification was performed using the SYBR green method (Takara Bio, Inc.) using the Bio-Rad IQTM5 Multicolor Real-time PCR Detection System (Bio-Rad Laboratories, Inc.). PCR primer sequences were obtained from OriGene Technologies Inc. or designed using NCBI Primer-BLAST (https://www.ncbi.nlm.nih.gov/tools/primer-blast/). Primer sequences were listed in Table SI. The reaction conditions were as follows: 95°C for 1 min, then 40 cycles of 10 sec at 95°C, 30 sec at 55°C and 30 sec at 72°C. The threshold cycle (Ct) values of miRNAs were normalized to those of U6, and those of the remaining genes were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Differential expression was calculated according to the 2−ΔΔCq method (33).

Western blotting

Proteins were extracted from HCT116 cells using the radioimmunoprecipitation assay buffer with proteinase inhibitors (Thermo Fisher Scientific, Inc.) Protein concentration was determined using the bicinchoninic acid method. Subsequently, ~30 μg protein was separated by SDS-PAGE on 12% gels and transferred to a polyvinylidene fluoride membrane. The membrane was blocked with 8% nonfat milk in TBS-0.1% Tween-20 at 25°C for 2 h and then incubated with anti-GSDMD (1:2,000; cat. no. ab210070; Abcam), anti-GSDME (1:2,000; cat. no. ab215191; Abcam) or the loading control anti-GAPDH (1:2,000; cat. no. AC001; ABclonal Biotech Co., Ltd.) antibodies at 4°C overnight. After washing, the HRP-conjugated secondary antibodies (1:5,000; cat. no. SA00001-2; ProteinTech Group, Inc.) were incubated at 25°C for 2 h. The protein bands were visualized using an enhanced chemiluminescent kit (MilliporeSigma). Experiments were performed at least in duplicate and similar results were obtained.

Immunohistochemistry

Immunostaining images of GSDME in normal colon and colon adenocarcinoma tissues were obtained from the Human Protein Atlas (http://www.proteinatlas.org). The examples of the pictures were selected from the normal (male, patient id: 1857; female patient id: 1423) and colon adenocarcinoma (male, patient id: 3408; female patient id: 1958) tissues, which were stained with antibody CAB019306. The images are available at v.20.1. proteinatlas.org.

Gene expression omnibus (GEO) and the cancer genome atlas (TCGA) data analyses

LncRNA NEAT1 and miR-448 expression data were obtained from PubMed GEO (https://www.ncbi.nlm.nih.gov/geo/) (GSE65292, GSE8704, GSE43310, GSE137013) and NCBI Sequence Read Archive (SRA, http://www.ncbi.nlm.nih.gov/Traces/sra) (PRJNA644335) databases or from supplementary data from Rogers et al (34). Expression and overall survival analyses for NEAT1 and GSDME in colon adenocarcinoma (COAD) from TCGA database (http://gepia.cancer-pku.cn) were performed using the Gene Expression Profiling Interactive Analysis online tool (http://gepia.cancer-pku.cn/) (35). The threshold for defining high/low gene expression levels was set using the following criteria: 50% of the patients were categorized into the high-expression group and the remaining 50% into the low-expression group.

Statistical analyses

Student's unpaired two-tailed t-test or one-way analysis of variance followed by Tukey's post hoc test was used to assess experimental data using GraphPad Prism 8 software (GraphPad Software, Inc.). The Kaplan-Meier analysis and log-rank test was used for overall survival analysis. The data are presented as means ± standard error of the mean from three biological replicates. P<0.05 was considered to indicate a statistically significant difference.

Results

IR-induced GSDME-mediated pyroptosis in HCT116 cells

To elucidate the mechanisms underpinning IR-induced damage in colonic epithelium, the effects of IR on human colorectal carcinoma HCT116 cells were investigated at 0, 8, 12 and 16 Gy. Cell viability analysis revealed that IR dose-dependently inhibited cell growth by ~50% at 72 h post-irradiation (P<0.01) (Fig. 1A). Flow cytometry analysis revealed that IR dose-dependently increased the percentage of dead cells (double positive for 7-AAD and Annexin V, which provides an approximate quantitative estimate of pyroptotic cells) (36) at 48 h post-irradiation (P<0.05) (Fig. 1B). Pyroptotic cells were observed using a phase-contrast microscope; a lower number of dilated cells was observed under 12 Gy and a higher number under 16 Gy compared with those in the control group, which showed that pyroptotic cell percentage may increase with dose (Fig. 1C). LDH released into media from irradiated HCT116 cells exhibited a significant gradual increase in response to dose of irradiation (P<0.05) (Fig. 1D). Western blotting analysis revealed that full-length and cleaved GSDMD expression remained unchanged with increasing radiation dose (Fig. 1E). In contrast, full-length GSDME protein expression was upregulated and cleaved GSDME protein expression increased continuously with increasing radiation dose (Fig. 1E).

Figure 1 Irradiation-induced pyroptosis in colorectal cancer HCT116 cells. HCT116 cells were irradiated with 8.0, 12.0 and 16.0 Gy. (A) Cell viability was analyzed using a Cell Counting Kit-8 assay at 72 h post-irradiation. ****P<0.0001 for normal control vs. 8.0 Gy; &&P<0.01 for 8.0 Gy vs. 12.0 Gy. (B) Flow cytometry analysis of irradiated HCT116 cells at 48 h. (C) Pyroptotic cells with large bubbles were observed at 48 h. The arrows indicate pyroptotic cells. The number of pyroptotic cells per field is indicated. (D) LDH released in media was measured at 24 h. (E) Total protein was extracted at 24 h. The expression and activation of GSDMD and GSDME were analyzed using western blotting. Glyceraldehyde 3-phosphate dehydrogenase was used as a loading control. (F) Caspase-3, -9, Bcl-2, Bcl-xl and Bax expression were analyzed using RT-qPCR in HCT116 cells at 48 h. GAPDH was used as the control. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001. OD, optical density; RT-q, reverse transcription-quantitative; GSDMD, gasdermin D; GSDME, gasdermin E; LDH, lactate dehydrogenase; NC, negative control, comprising nonirradiated cells.

It has previously been reported that radiation may induce apoptosis by selectively active caspase-9 and caspase-3/7 and both proapoptotic and antiapoptotic Bcl-2 family proteins (37). The expression of apoptosis-related genes in HCT116 cells was assessed after irradiation. The results demonstrated that the expression levels of Caspase-3 and Caspase-9 were significantly increased at -12 Gy (P<0.05), and those of the proapoptotic gene Bax were significantly increased at all doses (P<0.01), but that of antiapoptotic genes Bcl-2 and Bcl-xl was unchanged (Fig. 1F). These results indicated IR may induce caspases- and Bax-mediated apoptosis and GSDME-mediated pyroptosis in HCT116 cells.

GSDME is a target of miR-448

Next, the upstream factors of GSDME in human colorectal carcinoma cells were investigated based on the literature suggesting that miRNAs regulate cellular function by binding to the 3′-UTR of target genes (38). In this regard, 4 databases (including miRDB, miRWalk, microRNA.org and TargetScan) were screened and 6 candidates (miR-370, miR-375, miR-376, miR-379, miR-448 and miR-452) were obtained (Fig. 2A). 293T cells were then transfected with these miRNA plasmids and these miRNAs were highly expressed compared with cells transfected with control plasmid pcDNA3.0 (Fig. 2B). Transcriptional activity of GSDME was analyzed using GSDME 3′-UTR luciferase reporter assays, and miR-375, miR-379 and miR-448 were revealed as potentially targeting GSDME, but miR-370, miR-376 and miR-452 were excluded because their luciferase activities remained unchanged (Fig. 2C). The role of potential miRNAs were assessed by analyzing the expression of GSDME using RT-qPCR and western blotting, which revealed that miR-448 mimic significantly inhibited the expression of GSDME at the mRNA and protein levels compared with those in cells transfected with the mimic control miRNA (P<0.05), and siGSDME also decreased GSDME expression (P<0.0001) (Fig. 2D and E). As GSDME contained 2 miR-448 binding sites in its 3′-UTR sequences, mutant GSDME 3′-UTR luciferase reporter plasmids with mutations in either or both binding sites were generated (Fig. 2F). Mutations of either or both miR-448 binding sites rescued decreased luciferase activity to control levels (Fig. 2G), suggesting that GSDME was a downstream target of miR-448.

Figure 2 GSDME is a target of miR-448. (A) miRNAs identified by screening of miRDB, miRWalk, microRNA.org and TargetScan databases. (B) 293T cells were transfected with GSDME 3′-UTR luciferase reporter plasmid and miRNA vectors, the expression of each miRNA was verified by RT-qPCR. (C) Luciferase activity was measured using a dual-luciferase reporter assay. (D) Selection of potential regulatory miRNAs (miR-375-3p, miR-379-3p and miR-448) was verified with RT-qPCR. GSDME siRNA was used as a positive control. (E) 293T cells were transfected with GSDME siRNA and selected miRNA mimetics. GSDME expression was analyzed using western blotting. (F) Potential binding sites between miR-448 and GSDME were predicted by miRBase. (G) Either or both of 2 miR-448 binding sites in GSDME 3′-UTR reporter plasmids were mutated followed by transfection into 293T cells. A luciferase activity assay was subsequently performed. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001, compared with the pcDNA3.0 group unless otherwise indicated. RT-q, reverse transcription-quantitative; GSDME, gasdermin E; miR, microRNA; si, small interfering; NC, negative control; MUT, mutant; WT, wildtype; UTR, untranslated region; normal, untransfected cells.

Identification of lncRNAs targeting miR-448/GSDME in HCT116 cells

To examine the hypothesis regarding ceRNA and the association between miR-448 and GSDME, the potential upstream lncRNAs of miR-448 were investigated. Several lncRNAs have been reported to act as ceRNAs that regulate miR-448, including metastasis-associated lung adenocarcinoma transcript 1 (MALAT1), NEAT1, OIP5-AS1, Pvt1 oncogene (PVT1) and small nucleolar RNA host gene 1 (SNHG1) (Table I) (39-43). The expression of these lncRNAs was investigated in HCT116 cells according to radiation dose and over time. Given that IR dose-dependently increased pyroptosis, it was hypothesized that the expression of these lncRNAs would be significantly upregulated if they were involved in the regulation of GSDME expression via miR-448.

Table I List of miR-448 related ceRNAs.

Table I

List of miR-448 related ceRNAs.

Author, year	lncRNA	Full name (Aliases)	miRNA	Target	Tissue	Function	(Refs.)
Bamodu et al 2016	MALAT1	Metastasis Associated Lung Adenocarcinoma Transcript 1	miR-448	KDM5B	Breast cancer	Promotes aggressive breast cancer	(39)
Jiang et al 2018	NEAT1	Nuclear Paraspeckle Assembly Transcript 1	miR-448	ZEB1	Breast cancer	Contribute progression	(40)
Deng et al 2018	OIP5-AS1	OIP5 Antisense RNA 1	miR-448	Bcl-2	Lung adenocarcinoma	As oncogene	(41)
Zhao et al 2018	PVT1	Pvt1 Oncogene	miR-448	/	Pancreatic cancer	Promotes proliferation and migration	(42)
Pei et al 2018	SNHG1	Small Nucleolar RNA Host Gene 1	miR-448	IDO	Breast cancer	Immune escape of breast cancer	(43)

[i] ceRNA, competing endogenous RNA; miRNA, microRNA; ZEB1, zinc finger E-box binding homeobox 1; IDO, indoleamine 2,3-dioxygenase; KDM5B, lysine demethylase 5B.

The expression of MALAT1 exhibited a trend to be upregulated with radiation dose, but this did not reach significance (Fig. 3A). OIP5-AS1 expression was downregulated with IR dose, whereas the expression of PVT1 and SHNG1 was not regulated in response to IR dose (Fig. 3A). Notably, NEAT1 expression was significantly upregulated with an increase in IR dose (P<0.01) (Fig. 3A). IR time-dependently increased NEAT1 expression (P<0.001), whereas no significant temporal effects of IR on other lncRNAs were observed (Fig. 3B). NEAT1 expression in microarray or sequencing data from the PubMed GSE or SRA database and the literature and identified that NEAT1 expression was increased in human blood (44), monkey thymus (45), mouse embryonic fibroblast cells (46), mouse bone marrow (47) and mouse spleen (48) after irradiation (Fig. 3C-H), suggesting that NEAT1 expression was ubiquitously upregulated in response to IR.

Figure 3 Induction of miR-448 ceRNAs in HCT116 cells by radiation. (A) HCT116 cells were irradiated with 8.0, 12.0 and 16.0 Gy. The expression of MALAT1, nuclear NEAT1, OIP5-AS1, PVT1 and SHNG1 was analyzed at 48 h post-irradiation using RT-qPCR. (B) HCT116 cells were irradiated with 12.0 Gy. The expression of lncRNAs MALAT1, NEAT1, OIP5-AS1, PVT1 and SHNG1 was analyzed at 24, 48 and 72 h using RT-qPCR. (C-H) NEAT1 expression in human blood, monkey thymus, mouse embryonic fibroblast cells, mouse bone marrow and mouse spleen based on the PubMed GEO (GSE65292, GSE8704, GSE43310, GSE137013) and PubMed Sequence Read Archive database (PRJNA644335) or supplementary data from Rogers et al (34). *P<0.05, **P<0.01, ***P<0.001, compared with the NC group. lncRNA, long non-coding RNA; MEF, mouse embryonic fibroblast; GEO, Gene Expression Omnibus; RT-q, reverse transcription-quantitative; BM, bone marrow; MALAT1, metastasis-associated lung adenocarcinoma transcript 1; NEAT1, nuclear paraspeckle assembly transcript 1; OIP5-AS1, OIP5 antisense RNA 1; PVT1, Pvt1 oncogene; SNHG1, small nucleolar RNA Host Gene 1; ceRNA, competing endogenous RNA; NC, negative control, comprising nonirradiated cells or as defined in databases.

NEAT1 regulates GSDME expression

To confirm the functional association between NEAT1 and GSDME, first, the expression of GSDME protein was found to be higher in normal colon compared with COAD tissue in The Human Protein Atlas database (Fig. 4A). Next, NEAT1 and GSDME expression and the overall survival rate in COAD was assessed using TCGA data, which revealed a positive association between NEAT1 and GSDME levels and the advanced stages of COAD (Fig. 4B). Patients with COAD with higher NEAT1 and GSDME expression had lower overall survival, suggesting that they are positively associated in COAD (Fig. 4C). Secondly, NEAT1 was knocked down in HCT116 cells by transfecting cells with siRNAs against isoform NEAT1-1, NEAT1-2 or combined isoforms NEAT1-1+2. Treatment with siNEAT1-1, siNEAT1-2 and siNEAT1-1+2 significantly decreased NEAT1 expression compared with treatment with the negative control (P<0.05) (Fig. 4D), which indicated that siRNAs against NEAT1 were effective. However, GSDME expression was not affected by siNEAT1-1 or siNEAT1-2 treatment, but was downregulated following transfection of cells with siNEAT1-1+2 (P<0.01) (Fig. 4E). miR-448 expression was also confirmed to be upregulated in HCT116 cells transfected with siNEAT1-1+2 compared with those transfected with negative control siRNA (P<0.01) (Fig. 4F). Western blots revealed that full-length GSDME protein expression was downregulated following treatment with siNEAT1-1+2, but remained unchanged following treatment with siNEAT1-1 and siNEAT1-2 compared with normal control levels (Fig. 4G). In addition, the activated form of the N-terminal fragment of cleaved GSDME was not detected (Fig. 4G). This downregulation could be rescued by co-transfection of miR-448 mimetic with siNEAT1-1+2 in HCT116 cells (P<0.01) (Fig. 4H), supporting the involvement of miR-448 in the regulation of GSDME by NEAT1.

Figure 4 lncRNA NEAT1 upregulates GSDME via miR-448. (A) GSDME expression in human COAD was obtained from Human Protein Atlas (http://www.proteinatlas.org). The examples of the pictures by sex and patient id are shown and the images are available at v.20.1 www.proteinatlas.org. (B) Expression of lncRNA NEAT1 and GSDME in the major stages of COAD. (C) Overall survival analysis of NEAT1 and GSDME in COAD from the Cancer Genome Atlas database using the Gene Expression Profiling Interactive Analysis online tool. (D-G) HCT116 cells were transfected with siGSDME, siNEAT1-1, siNEAT1-2, siNEAT1-1+2, or negative control siRNA. After 48 h, (D) NEAT1 expression and (E) GSDME expression were analyzed using RT-qPCR. GAPDH was used as a control. (F) miR-448 expression was analyzed using RT-qPCR in HCT116 cells after transfection with siNEAT1-1, siNEAT1-2 and siNEAT1-1+2. (G) Expression of full-length and cleaved GSDME was analyzed using western blotting. GAPDH was used as a loading control. (H) GSDME mRNA expression was analyzed using RT-qPCR in HCT116 cells after transfection with siGSDME, siNEAT1-1+2, or co-transfection with siNEAT1-1+2 and miR-448 mimetics. *P<0.05, **P<0.01 compared with the NC group unless otherwise indicated. RT-q, reverse transcription-quantitative; GSDME, gasdermin E; miR, microRNA; si, small interfering; NEAT1, nuclear paraspeckle assembly transcript 1; lncRNA, long non-coding RNA; COAD, colon adenocarcinoma; HR, hazard ratio; NC, negative control (comprising cells transfected with NC siRNA); normal, untransfected cells.

NEAT1 promotes radioresistance by enhancing pyroptosis via miR-448/GSDME

To further investigate the regulation of lncRNAs/miR-448/GSDME in IR-induced pyroptosis in HCT116 cells, HCT116 cells were irradiated with increasing doses. A gradual decrease in expression of miR-448 was observed (P<0.001; Fig. 5A), which was negatively associated with the expression of NEAT1 and GSDME in irradiated HCT116 cells (Figs. 3A and 5B). Based on the PubMed GEO database (34,47), the expression of miR-448 was downregulated in the mouse lung, but not in the heart or bone marrow suggesting that miR-448 may be regulated in a tissue-specific manner in response to IR (Fig. 5C and D). Similarly, TCGA data analysis demonstrated that GSDME exhibited a tissue-specific differential expression between tumors and normal tissues (Fig. S1). To confirm the regulation of NEAT1 and miR-448 in IR-induced pyroptosis, HCT116 cells were transfected with siNEAT1-1+2 and miR-448 mimetics before irradiation. Knockdown of NEAT1 increased miR-448 expression (P<0.01) and resulted in a decrease in LDH release and rescued cell viability in radiated HCT116 cells compared with irradiated cells transfected with negative control RNA (P<0.05) (Fig. 5E-G); similar results were obtained from overexpression of miR-448, suggesting a regulatory role of NEAT1/miR-448 on viability of human colorectal cancer cells via pyroptosis (P<0.05) (Fig. 5F-G). The expression of GSDME mRNA and full-length protein was downregulated, in accordance with the changes in morphology and LDH release following transfection with siNEAT1 and miR-448 mimetics. Expression of the cleaved N-terminal fragment of GSDME was also decreased and the ratios of GSDME-N vs. GSDME-F were not significantly different (Fig. 5H-I). Taken together, lncRNA NEAT1 might sponge miR-448 and promote radioresistance by enhancing IR-induced pyroptosis in human colorectal cancer HCT116 cells via regulating the expression, but not activation, of GSDME.

Figure 5 NEAT1 promotes radioresistance by enhancing pyroptosis via modulating miR-448/GSDME. (A and B) HCT116 cells were irradiated with 8.0, 12.0 and 16.0 Gy. Expression of miR-448 and GSDME were analyzed using RT-qPCR. (C and D) miR-448 expression in the mouse heart and lung after irradiation from the Gene Expression Omnibus dataset GSE147242 and in mouse bone marrow from GSE137013. (E-I) HCT116 cells were transfected with siNEAT1-1+2, miR-448 mimetics or negative control siRNA and then irradiated with 12.0 Gy at 12 h post-transfection. (E) Expression of miR-448 was analyzed using RT-qPCR at 48 h post-irradiation. (F) LDH released in media was measured at 24 h. (G) Cell viability was analyzed using a Cell Counting Kit-8 assay at 72 h. (H) Expression of GSDME was analyzed using RT-qPCR at 48 h. (I) Expression and activation of GSDME protein were analyzed at 48 h using western blotting. GAPDH was used as a loading control. *P<0.05, **P<0.01, ***P<0.001 compared with the NC group unless otherwise indicated. RT-q, reverse transcription-quantitative; IR, ionizing radiation; LDH, lactate dehydrogenase; BM, bone marrow; GSDME, gasdermin E; miR, microRNA; si, small interfering; NEAT1, nuclear paraspeckle assembly transcript 1; miR, microRNA; NC, negative control [comprising nonirradiated cells or animals in (A-D) or cells transfected with NC siRNA in (F-I)].

Discussion

The present study demonstrated that in addition to enhancing apoptosis via caspases and the bcl-2 family, IR dose-dependently induced pyroptosis in human colorectal cells and identified GSDME as a target of miR-448. In the present study, miR-448 activity was rescued by mutations in either or both of the 2 miR-448 binding sites in GSDME 3′-UTR. In addition, the present study demonstrated that miR-448 inhibited GSDME-mediated pyroptosis and highlighted that the lncRNA NEAT1 as a potential upstream driver of miR-448. Notably, the present study observed that lncRNA NEAT1 knockdown regulated radiosensitivity through IR-induced pyroptosis and viability by regulating the expression, but not activation of full-length GSDME. Collectively, the findings of the present study indicated the dose-dependent effects of IR on pyroptosis in human colorectal carcinoma HCT116 cells via GSDME activation and underscored the critical role of lncRNA NEAT1 in IR-induced pyroptosis via the regulation of miR-448 and expression, but not cleavage of GSDME.

Previous studies have reported that pyroptosis is induced in the normal intestine or colon via activation of Casp1/GSDMD in ulcerative colitis injury (49-51) or in response to IR (18,52). A study demonstrated that pyroptosis in human colorectal cancer cells is induced by chemotherapeutic drugs, such as lobaplatin via Casp3/GSDME activation rather than Casp1/GSDMD (10). However, GSDMD has also been implicated in regulation of pyroptosis in colon cancer (53). In this regard, low expression of GSDMD was associated with poor colorectal cancer prognosis and lipopolysaccharide induced GSDMD-mediated pyroptosis. Additionally, overexpression of GSDMD in HT29 cells reduced cell survival and induced cell death (48). Hence, both GSDMD and GSDME are thought to regulate pyroptosis in colorectal cancer cells depending on pathological state or type of stimuli (10,53). The findings of the present study indicated that GSDME, but not GSDMD, induced pyroptosis in human colorectal cancer HCT116 cells in response to IR, which resembles the regulatory mechanism of pyroptosis in chemotherapy (6). This finding of the present study provides a novel mechanism of IR-induced damage, in addition to apoptosis and necrosis in cancer radiotherapy.

ceRNAs serve crucial roles in multiple physiological and pathological processes (24,40,54). ceRNAs have been reported to regulate pyroptosis regulators. For instance, HOX transcript antisense RNA/miR-455-3p regulated NLRP1-mediated inflammasome activation and X inactive specific transcript/miR-133-3p or maternally expressed 3/miR-7a-5p regulated NLRP3-mediated inflammasome activation (55,56). In the present study, pyroptosis execution protein GSDME was first identified as a target of miR-448 and secondly, it was identified that ceRNA NEAT1/miR-448 may regulate IR-induced GSDME-mediated pyroptosis and viability in human colorectal cancer cells. Other lncRNAs may also sponge miR-448, but they were not upregulated by IR, suggesting the absence of regulatory effects on miR-448 and IR-induced pyroptosis in the present study. Notably, expression data in the present study demonstrated that IR-triggered miR-448 downregulation, but not NEAT1 upregulation, which may be tissue-specific, with consideration of tissue-specific differential expression of GSDME in tumors. The results of the present study indicated a complex regulation of GSDME by NEAT1/miR-448 in cancer radiotherapy.

Previous studies have shown that NEAT1 knockdown repressed proliferation, induced apoptosis (57,58), or promoted the fluorouracil sensitivity in colorectal cancer cells (59,60), but inhibited radioresistance by inducing G0/G1 arrest and apoptosis in cervical cancer (61) or enhanced radioresistance by depressing apoptosis in nasopharyngeal carcinoma (62). The findings of the present study suggested that NEAT1 knockdown was involved in suppression of IR-induced pyroptosis, which may contribute to enhancing radioresistance and a provided new approach for understanding the mechanism of NEAT1 in cancer therapy. NEAT1 is induced by genotoxic stress such as DNA damage and is upregulated by p53 (63), which may act as a transcriptional regulator of multiple genes, including puma and p21, that are associated with cell survival and cancer progression (64,65). Hence, IR might regulate NEAT1 via p53-mediated pathways and subsequently affect pyroptosis and viability in human colorectal cancer cells via the miR-448/GSDME axis.

GSDME cleavage is key for cell membrane rupture and pyroptosis (11). The results of the present study, demonstrated that GSDME cleavage and full-length GSDME expression were inhibited by siNEAT1 or miR-448 overexpression. The level of GSDME expression in cancer cells is a critical determinant of its function and modulates the decision to switch between apoptosis and pyroptosis in cancer cells (66). Although the decrease in expression of N-terminal fragment of GSDME was too small in the present study to definitively conclude that NEAT1/miR-448 directly regulated the cleavage of GSDME, it can still be proposed that the cleavage of GSDME is affected by NEAT1/miR-448 via reduced total expression of GSDME.

The present study has certain shortcomings. For example, more colorectal cancer cell lines should be used to investigate the cell variability and whether NEAT1 expression is high in all tissues should be assessed. Moreover, GSDME is critical for inducing a switch from apoptosis to pyroptosis (6), which can both be triggered by IR (14); therefore, NEAT1 might be involved in this switching process. Thus, further experiments will be required to address these questions.

In summary, the present study demonstrated IR dose-dependently induced pyroptosis and identified the pyroptosis execution protein GSDME as a target of miR-448. lncRNA NEAT1 sponged miR-448 and regulated IR-induced pyroptosis and cell viability in human colorectal cancer cells via regulating the expression, but not activation of GSDME. The findings of the present study elucidated a novel pyroptosis-related mechanism by which NEAT1 participates in colorectal cancer progression and laid theoretical foundations for further therapeutic development.

Supplementary Data

Availability of data and materials

The analyzed data generated during the study are available from the corresponding author upon reasonable request.

Authors' contributions

CG, FS and XZ conceived and designed the experiments. FS, JD, JZ and HF performed the experiments and data analysis. CG wrote the manuscript and it was revised for important intellectual content by XZ. CG and FS confirmed the authenticity of all the raw data. All authors read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Acknowledgements

Not applicable.

Abbreviations:

References