Pyroptosis is a type of programmed cell death characterized by the activation of inflammatory caspases and the cleavage of gasdermin proteins. Pyroptosis can suppress tumor development and induce antitumor immunity, and activating pyroptosis is a potential treatment strategy for cancer. To uncover approaches to harness the anticancer effects of pyroptosis, we aimed to identify regulators of pyroptosis in cancer. A CRISPR-Cas9 screen identified that loss of USP48, a deubiquitinating enzyme, significantly inhibited cell pyroptosis. USP48 promoted pyroptosis by stabilizing gasdermin E (GSDME). USP48 bound GSDME and removed K48-linked ubiquitination at positions K120 and K189. Clinical tissue testing confirmed that the expression of USP48 positively correlated with GSDME and pyroptosis-related factors. Single-cell sequencing showed that the functions of T cells and tumor-associated macrophages in the tumor microenvironment were inhibited after USP48 knockout. Finally, overexpression of USP48 enhanced the therapeutic efficacy of programmed cell death protein 1 inhibitors in tumors in mouse models. Together, these findings define a pyroptosis regulation pathway and indicate that pharmacologic activation of USP48 may provide an effective strategy to sensitize cancer cells to pyroptosis and improve response to immunotherapy.

Significance:

USP48 promotes pyroptosis by deubiquitinating GSDME and enhances antitumor immunity, indicating that increasing USP48 activity may be a future therapeutic strategy for treating cancer.

Pyroptosis is a type of programmed cell death mediated by the gasdermin family of proteins, which has been discovered in recent years (1). Different proteases or granzymes mediate the cleavage of the gasdermin family proteins to release the N-terminal domain, which is then transported to the cell membrane to form multimeric pores, causing changes in membrane permeability, and leads to cell swelling and membrane rupture (2–4). Pyroptosis is an important immune response of the body, which plays an important role in antagonizing infection and endogenous danger signals (5). It is widely involved in the occurrence and development of tumors, infectious diseases, metabolic diseases, nervous system–related diseases, and atherosclerotic diseases (6). In recent years, studies have proved that pyroptosis could also shrink tumors and inhibit cells proliferation.

Recent studies have found that gasdermin E (GSDME, also known as DFNA5), which is mutated in familial aging-associated hearing loss, can be cleaved by activated caspase-3, thereby converting non-inflammatory apoptosis into pyroptosis in GSDME-expressing cells (7, 8). Caspase-3 could not be activated by self-splicing or autocatalysis. Caspase-3 was cleaved by granzyme B or caspase-10 at the D175 site. Then, p20 and p11 subunits were composed that induced the activation of caspase-3 (8). In many cancers, the expression of GSDME is suppressed (9). Studies have shown that reduced GSDME levels are associated with decreased 5-year survival and metastasis in many cancer patients, which suggests that GSDME may be a tumor suppressor (10). The evidence that GSDME may be used as a tumor suppressor also includes promoter DNA methylation in many primary cancers that can inactivate GSDME; GSDME inhibits colony formation, proliferation and invasiveness of gastric cancer, melanoma, and colorectal cancer cells; and, the loss of GSDME has been shown to abrogate the effectiveness of some chemotherapeutic drugs (11–15). In addition, the role of GSDME in antitumor immunity has also received more attention. Recent studies have revealed that the expression of GSDME enhances the number and function of tumor-infiltrating natural killer (NK) cells and CD8+ T lymphocytes (16). Granzyme B from NK cells can also activate caspase-independent pyroptosis in target cells by directly cleaving GSDME at the same site as caspase-3 (17). Therefore, GSDME in tumor tissues acts as a tumor suppressor by activating pyroptosis and enhancing antitumor immunity.

Ubiquitin is an evolutionarily conserved 8.5 kDa protein, which is covalently linked to the N-terminal or lysine residue of the substrate protein through the sequence reaction of ubiquitin activating enzyme (E1), ubiquitin coupling enzyme (E2) and ubiquitin ligase (E3; ref. 18). The ubiquitin chain can be formed by connecting ubiquitin with lysine 48 of ubiquitin itself, resulting in the degradation of substrates by proteasomes. Other types of polyubiquitin (polyUb) chain ligation and monoubiquitination can lead to degradation or nondegradable effects on substrates, such as repositioning, promoting protein–protein interactions and regulating signal events (19). Ubiquitination is regulated by more than 1,000 human proteins, accounting for about 4% of the proteome, including two E1s, about 40 E2s, more than 600 E3s and about 100 deubiquitinating enzymes (20). It has been reported that there are more than 8,000 ubiquitination sites on thousands of proteins (21). Studies have shown that ubiquitin-related proteins play a key regulatory role in many biological processes, such as the cell cycle, DNA damage, apoptosis, and autophagy (22–25). However, the roles and molecular mechanisms of ubiquitin-related proteins in pyroptosis have not been well identified.

The latest progress of CRISPR-Cas9 gene screening technology makes it possible to measure gene importance, cancer cell dependence, and genetic interactions in human cells in a high-throughput manner (26). Here, we screened and identified USP48, a ubiquitinase that plays an important role in regulating cell pyroptosis by applying the CRISPR-Cas9 gene screening technology, combined with the detection of related secretory factors. In terms of molecular mechanism, USP48 stabilizes its expression by causing deubiquitination of GSDME, achieving its regulatory effect on pyroptosis. Clinical tumor tissue testing confirmed that the expression of USP48 has a significant positive correlation with GSDME. The single-cell sequencing results showed that the immune microenvironment in the tumor tissues of the mice after USP48 gene knockout was significantly changed, the functions of T cells and tumor-associated macrophages (TAM) in the tumor microenvironment are affected to varying degrees after pyroptosis is damaged. Finally, the tumor formation experiment in mice confirmed that overexpression of USP48 could effectively improve the therapeutic effect of programmed cell death protein 1 (PD-1) inhibitors.

Cell cultures

HeLa, HEK293T, L929, and Pan02 cell lines were purchased from ATCC. All four cells were subcultured using DMEM (Gibco), containing 10% FBS (Gibco) and 0.1% penicillin–streptomycin solution, in a cell incubator at 37°C and 5% CO2 saturated humidity. When the cell confluence reaches about 80%, the cells are digested with trypsin for subculture. All cell lines were examined with the short tandem repeat profiling by vendors. Mycoplasma infection of all cell lines was tested periodically.

Retroviral infection and overexpression of USP48 and GSDME

The siRNA used for the knockdown of USP48 in the cells was purchased from GenePharma. The siRNAs were transfected into cells using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Lentivirus-mediated overexpression and knockdown of USP48 and GSDME against cells were purchased from GeneChem. Cells (30% confluence) were incubated in medium containing optimal dilutions of lentivirus mixed with polybrene. After 48 hours of transfection, cells were subjected to puromycin selection (5 mg/mL) to obtain stably transfected cells.

Tissue specimens

The collection of adjacent normal tissue specimens from a standard distance (3 cm) from the edge of the tumor tissue removed from patients with pancreatic ductal adenocarcinoma (PDAC) or hepatocellular carcinoma (HCC) undergoing surgical resection was performed. All clinical specimens used in this study were histopathologically and clinically diagnosed. To use these clinical data for research purposes, the consent and approval of the Institutional Research Ethics Committee of the Second Hospital of Shandong University was obtained in advance. The research complied with all relevant ethical norms involving human participants. The research protocol complied with the ethical standards of the Declaration of Helsinki and the patient's written informed consent was obtained according to the Declaration of Helsinki.

Animals and animal model

Kras G12D;PDX1-Cre mice were crossed with USP48flox/flox mice to obtain pancreas-specific knockout mice. All animal experiment procedures followed the guidelines of the NIH and the guidelines of the Second Hospital of Shandong University. The Usp48flox/flox mice were generated by Cyagen Biosciences Inc. USP48 is located on chromosome 4 of mice, and CRISPR/Cas9 technology was used to design single-guide RNA (sgRNA) and ssDNA to obtain Usp48 gene conditional knockout mice by high-throughput electrotransfer of fertilized eggs. All involved mice were C57/B6 mice. All animals were randomly numbered. After data collection, genotypes were revealed, and animals were randomly assigned to each group. During data collection and analysis, the group assignment was unknown to the researchers. The sample size and inclusion criteria of each group were confirmed with adequate power based on the literature and our previous experience. None of the mice with the appropriate genotype were excluded from this study or used in any other experiments. Mice did not undergo prior treatment or procedures. All mice were fed a standard chow diet ad libitum and housed under standard conditions (controlled temperature, humidity, and 12-hour light-dark cycle) in a pathogen-free facility with animals under veterinary supervision, with no more than 5 mice per cage. All mice were reared in the Animal Experiment Center of the Second Hospital of Shandong University. All experimental procedures were performed in accordance with the Guide for the Care and Use of Laboratory Animals published by the NIH (8th edition, 2011) and approved by the Institutional Animal Care and Use Committee of the Second Hospital of Shandong University.

CRISPR knockout pooled lentiviral sgRNA libraries

A combined library of 384 unique sgRNA or gRNA sequences can disrupt (or “knock out”) 96 deubiquitination related genes in the entire population of cells in a single experiment. The cell population is then screened for the release of lactate dehydrogenase (LDH), so the specific gene driving that phenotype can be identified. The CRISPR Knockout Pooled Lentiviral sgRNA Library was purchased from Dharmacon.

IHC

The specimens were formalin-fixed, paraffin-embedded sections, and the protein to be detected was evaluated by IHC staining. Two independent observers scored the proportion of positively stained tumor cells and the staining intensity. The cut-off values for high and low expression of the protein of interest are selected on the basis of the measurement of heterogeneity using the log-rank test on overall survival.

Multiplex immunofluorescence

Tissues were fixed overnight in 4% PFA at 4°C and embedded in paraffin. After preparation into sections, rehydration was carried out through ethanol series. Antigen retrieval was performed using a microwave oven in 10 mmol/L sodium citrate (pH 6). Sections were blocked with 10% goat serum and incubated with primary antibody (1:200) at 4°C overnight. Then, the sections were thoroughly washed and incubated with secondary antibody (1:1,000) and DAPI (0.4 μg/mL) for 1 hour at room temperature. Sections were thoroughly washed and mounted in an antifade fluorescent mounting medium.

Western blotting and coimmunoprecipitation

The cells were treated with RIPA buffer containing 1x protease inhibitor and phosphatase inhibitor, and the product was collected. The expression of corresponding proteins was detected by SDS-PAGE gel electrophoresis, and Western blotting and band quantification were performed using the ChemiDoc MP Imaging System. For co-immunoprecipitation (co-IP), the cell lysate (1 mL) was mixed with 10 μL of anti-MYC, anti-Flag, or anti-HA antibody and 20 μL of protein G magnetic beads at 4°C overnight. Then, the IP beads were washed three times with lysis buffer and then heated in SDS loading buffer at 97°C for 5 minutes. The product was used for SDS-PAGE and Western blotting analysis.

In vitro ubiquitination assay

USP48 and GSDME proteins were expressed using a TNT fast coupled transcription/translation system (Promega). A ubiquitination kit (Boston Biochem) was used to perform ubiquitination analysis according to the manufacturer's recommended protocol.

qRT-PCR

Total RNA was isolated from cells cultured with TRIzol reagent (Invitrogen) as directed. According to the manufacturer's instructions for the reagent, the PrimeScript Reverse Transcription Kit (TaKaRa) was used to reverse the RNA (1 μg) to cDNA (20 μL). The experiment was performed at least three times and repeated twice. Endogenous GAPDH was used as a standardized control. Quantitative analysis was performed by comparing CT values.

Flow cytometry analysis

A single-cell suspension from mouse pancreatic and tumor tissues was prepared. The cells were incubated with an Fc blocking agent and stained with the following fluorescent dye-conjugated antibodies or isotypes for 30 minutes at 4°C: mouse anti-CD4, mouse anti-FOXP3, mouse anti-CD25, mouse anti-CD8, mouse anti-CD3, small mouse anti-F4/80, mouse anti-CD11b, and mouse anti-NK1.1 (BioLegend). The data was immediately acquired by the FACSAria SORP flow cytometer (BD Biosciences) and analyzed using FlowJo software.

SYTOX Green uptake and time-lapse microscopy

The cells were seeded in a 96-well plate overnight and treated with 10 μmol/L raptinal in the presence of 2.5 μmol/L SYTOX Green for 2 hours. A microplate reader was used to continuously record the fluorescence at 528 nm after excitation at 485 nm every 10 minutes. For time-lapse microscopy, cells seeded in 35-mm glass bottom dishes overnight were treated with 10 μmol/L raptinal in complete DMEM containing 2.5 μmol/L SYTOX Green and imaged using a Zeiss 880 laser scanning confocal microscope within an environmental chamber maintained at 37°C and 5% CO2.

LDH detection

The detection of LDH was performed using the CytoTox 96 Non-Radioactive Cytotoxicity Assay Kit (Promega) according to the manufacturer's recommended protocol.

Detection of IL18 and IL1β

The ELISA kit purchased from Invitrogen (Invitrogen) was used to process the cells according to the protocols provided by the supplier. Then, a microplate reader was used to detect and record the results.

RNA sequencing library construction for 10x genomics single-cell 5′ sequencing

Pancreatic tissues from 4 mice (two KrasG12D;PDX1-Cre and two KrasG12D;Usp48flox/flox;PDX1-Cre) were taken and prepared into single-cell suspensions. The single-cell suspensions were subsequently stained with antibody (CD45-BV421, 0.5 mL/106 cells) and then sorted into centrifuge tubes pre-wetted with FBS using a FACSAria SORP flow cytometer (BD Biosciences). Dead cells were excluded by propidium iodide. A single-cell 5′ RNA sequencing (RNA-seq) library for immunized (CD45+) cells was constructed for each of the four samples separately, according to the protocol of the Chromium Single Cell 5′ Library Construction Kit (10x Genomics). The data generated in this study has also been deposited in NCBI-Sequence Read Archive (SRA) database (Bioproject accession: PRJNA783473).

Xenograft tumor studies

The cells transfected with the corresponding lentivirus or plasmid were made into a cell suspension with a density of 1×107/mL. C57/B6 mice aged 8 to 12 weeks were selected and 100 μL of the cell suspension was subcutaneously injected into the left armpit. Tumor growth was observed according to the experimental needs condition.

Statistical analysis

Each experiment was performed at least three times. The sample size for all other graphs was indicated as “n = x” above the graphs. Statistical analysis was performed using GraphPad Prism 8 software (GraphPad Software). The two-tailed Student t test was used to determine the significance of the differences between the two independent samples. Pearson correlation analysis was performed to determine the correlation between the two variables. Survival differences were determined using the Kaplan–Meier method and the Log-rank test. P values < 0.05 were considered statistically significant. All P values are indicated in the graphs (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; n.s., not significant). This study selects a representative experimental result from three or more independent experiments to present.

Data availability

The data presented in this study is provided in the main article and Supplementary Data files. Data generated by single-cell RNA-seq have been deposited in NCBI-SRA database (Bioproject accession: PRJNA783473, https://dataview.ncbi.nlm.nih.gov/object/PRJNA783473?reviewer=29otk289l0g9l9pbkt07a7nme8).

USP48 is involved in the regulation of pyroptosis

Pyroptosis is a newly identified programmed cell death, which is mainly mediated by inflammasomes through the activation of a variety of caspases, resulting in the shear and multimerization of a variety of gasdermin family members, which caused cell perforation and cell death (27). Raptinal is a fast caspase-3 activator, and activated caspase-3 can induce pyroptosis by cleaving GSDME. We observed the morphologic changes of human cervical cancer HeLa cells after raptinal treatment by microscope, and found that pyroptotic ballooning cell membranes appeared in HeLa cells, confirming that raptinal can induce pyroptosis in HeLa cells (Fig. 1A). When cells undergo pyroptosis, pores are formed on the surface of the cell membrane, leading to the release of cytoplasmic contents, such as LDH and inflammatory cytokines.To find related genes involved in the regulation of pyroptosis, we conducted CRISPR-Cas9 gene screening in HeLa cells, using the secretion of LDH after raptinal treatment as a readout, and the obtained genes were further screened and verified by single siRNA screening technology (Fig. 1B). The screening library consists of 384 unique sgRNAs (Supplementary Table S1), knocking down the expression of 96 deubiquitination related genes. According to the fold change of LDH secretion after sgRNA transfection, 36 genes were selected as the main genes. Among them, the silencing of 17 candidate genes increased the secretion of LDH by more than two-fold, while the silencing of the other 19 genes inhibited the secretion of LDH (Fig. 1C and D). To further screen the obtained genes, we designed four independent siRNAs for each gene. The results showed that the interference of USP48 had the significant inhibitory effect on the release of LDH in cells (Fig. 1E).

Figure 1.

USP48 is involved in the regulation of pyroptosis. A, Pyroptotic morphology after raptinal treatment of HeLa cells. B, Two-stage screening strategy diagram of the CRISPR Knockout Pooled Lentiviral sgRNA Library and siRNA verification. C, The fold change of LDH secretion in HeLa cells transfected with 384 targeted sgRNAs after being treated with 10 μmol/L raptinal for 1 hour. The targeted genes that significantly increased or decreased LDH secretion are highlighted in red or green, respectively. D, Fold changes of LDH secretion in HeLa cells transfected with siRNAs for secondary validation, followed by treatment with raptinal as in C. Data are normalized to untreated HeLa cells (dashed line). E, qRT-PCR verifies the transfection efficiency of si-NC and USP48-specific siRNA in HeLa cells. F, The LDH detection kit was used to analyze the changes in LDH levels in HeLa cells transfected with siRNA (si-NC) or siRNA specific to USP48 (si-USP48) after treatment with raptinal. G, An ELISA experiment was used to analyze the expression changes of IL1β in 293T cells transfected with siRNA (si-NC) or siRNA specific to USP48 (si-USP48) after treatment with raptinal. H, An ELISA experiment was used to analyze the expression changes of IL18 in 293T cells transfected with siRNA (si-NC) or siRNA specific to USP48 (si-USP48) after treatment with raptinal. I, Time-lapse microscopy observed the morphologic changes and the absorption of SYTOX Green of HeLa cells transfected with siRNA (si-NC) or USP48-specific siRNA (siUSP48–1) after treatment with raptinal. J, Quantitative analysis chart of SYTOX Green absorption. K, Kinetics of caspase-3 and HMGB1 release by immunoblotting of cell lysates and culture supernatants. Data are shown as mean ± SD of three independent experiments. **, P < 0.01; ***, P < 0.001. Scale bar, 20 μm.

Figure 1.

USP48 is involved in the regulation of pyroptosis. A, Pyroptotic morphology after raptinal treatment of HeLa cells. B, Two-stage screening strategy diagram of the CRISPR Knockout Pooled Lentiviral sgRNA Library and siRNA verification. C, The fold change of LDH secretion in HeLa cells transfected with 384 targeted sgRNAs after being treated with 10 μmol/L raptinal for 1 hour. The targeted genes that significantly increased or decreased LDH secretion are highlighted in red or green, respectively. D, Fold changes of LDH secretion in HeLa cells transfected with siRNAs for secondary validation, followed by treatment with raptinal as in C. Data are normalized to untreated HeLa cells (dashed line). E, qRT-PCR verifies the transfection efficiency of si-NC and USP48-specific siRNA in HeLa cells. F, The LDH detection kit was used to analyze the changes in LDH levels in HeLa cells transfected with siRNA (si-NC) or siRNA specific to USP48 (si-USP48) after treatment with raptinal. G, An ELISA experiment was used to analyze the expression changes of IL1β in 293T cells transfected with siRNA (si-NC) or siRNA specific to USP48 (si-USP48) after treatment with raptinal. H, An ELISA experiment was used to analyze the expression changes of IL18 in 293T cells transfected with siRNA (si-NC) or siRNA specific to USP48 (si-USP48) after treatment with raptinal. I, Time-lapse microscopy observed the morphologic changes and the absorption of SYTOX Green of HeLa cells transfected with siRNA (si-NC) or USP48-specific siRNA (siUSP48–1) after treatment with raptinal. J, Quantitative analysis chart of SYTOX Green absorption. K, Kinetics of caspase-3 and HMGB1 release by immunoblotting of cell lysates and culture supernatants. Data are shown as mean ± SD of three independent experiments. **, P < 0.01; ***, P < 0.001. Scale bar, 20 μm.

Close modal

Ubiquitination is an important item in posttranslation. Its most well-known function is to guide the degradation of proteins. Ubiquitination is reversible and can be reversed by a large group of proteases called deubiquitinating enzyme (DUB; ref. 28). USP48 is an important member of the USP family of DUBs (29). To further clarify the role of USP48 in pyroptosis, we silenced the expression of USP48 by transfecting siRNA into the HeLa cell line (Fig. 1D) and treated the cells with raptinal to detect the effect of USP48 silencing on LDH, IL1β, and IL18 in HeLa cells (Fig. 1F-H). The results showed that the silence of USP48 could significantly inhibit the release of LDH, IL1β, and IL18 in HeLa cells. In addition, we also observed the morphologic changes and the absorption of SYTOX Green in HeLa cells transfected with siRNA (si-NC) or USP48-specific siRNA (si-USP48) after raptinal treatment by time-lapse microscopy, and the results showed that USP48 silencing significantly inhibited raptinal-induced pyroptosis and SYTOX Green absorption (Fig. 1I and J). The above results were further confirmed in L929 cells, which is a mouse fibroblast cell line (Supplementary Fig. S1A–S1G). In addition, we also found that USP48 silencing inhibited the raptinal-induced release of HMGB1 (Fig. 1K; Supplementary Fig. S1H). HMGB1 is a multifunctional protein, and extracellular HMGB1 is related to inflammation. Studies have shown that inflammasome-mediated pyroptosis promotes the release of HMGB1 (30). In addition, upregulating the expression of USP48 in 293T cells also obtained consistent results with the above (Supplementary Fig. S1I–S1L). In summary, we have preliminarily confirmed that USP48 may be involved in the regulation of pyroptosis.

USP48 regulates the expression of GSDME

To elucidate the specific mechanism of USP48 regulating cell pyroptosis, We examined the expression of USP48 in six different human-derived cell lines and found that USP48 was lowly expressed in 293T cells and highly expressed in HeLa cells (Fig. 2A). On the basis of this result, we stably overexpressed USP48 in 293T cell lines and stably knocked down USP48 in HeLa cell lines, respectively (Fig. 2B), and performed proteomic sequencing on 293T cells overexpressing USP48 or transfected with an empty plasmid. (Fig. 2C). It was found that the expressions of 112 proteins were upregulated, and the expressions of 258 proteins were negatively correlated with the overexpression of USP48 (Supplementary Fig. S2A and S2B; Supplementary Table S2). In addition, the results also showed that USP48 was involved in the regulation of various cell biological processes, including pyroptosis and cell metabolism (Supplementary Fig. S2C and S2D). To further identify the proteins physically associated with USP48, we used anti-Flag affinity purification mass spectrometry to identify potential USP48-interacting proteins in 293T cells expressing Flag-tagged USP48 and found 59 proteins that could directly interact with USP48 (Supplementary Table S3). After combined analyses of both mass spectrometry and proteomics results, we obtained 8 related proteins (Fig. 2D and E). To further verify the above results, We tested the changes in expression levels of these eight proteins after overexpression of USP48 and knockdown of USP48, and found that the GSDME had more significant and stable changes compared with the other several proteins (Fig. 2F and G). In addition, we examined the expression of GSDMD, a key protein for pyroptosis, and the results showed that the change in the expression level of USP48 had no significant effect on the protein expression of GSDMD. After examining the expression of GSDME and GSDMD in the above six cell lines separately, we also further confirmed that the expression of GSDME was positively correlated with that of USP48, but the expression status of GSDMD and USP48 was not significantly correlated (Fig. 2H).To further confirm the regulatory effect of USP48 on GSDME, we also verified it in human breast cancer cell line MCF7 and human pancreatic cancer cell line Panc-1, and consistent results were obtained. Notably in that overexpression of the catalytically inactive mutant USP48 (USP48/C98A) could not affect the expression of GSDME (Fig. 2I).

Figure 2.

GSDME is a downstream target protein of USP48. A, Detection of USP48 expression in 6 different human tumor cell lines. B, Stable overexpression of USP48 in 293T and stable knockdown of USP48 in HeLa cells. C, The heat map summarizes the difference in protein expression in 293T cells with normal expression of USP48 and overexpression of USP48. D, The Flag-USP48 pulldown product from 293T cells was separated by SDS-PAGE, visualized by silver staining, and the protein interacting with USP48 was identified by mass spectrometry. E, Joint analysis of proteomics results and mass spectrometry sequencing results to get an intersection. F and G, Verification of the results obtained in C by immunoblotting. H, Detection of GSDME and GSDMD expression in 6 different human tumor cell lines. I, Western blotting of USP48 and GSDME in MCF7 and Panc-1. J, Lysates from cells expressing the control or the indicated constructs were immunoprecipitated (IP) with IgG or anti-USP48 and then immunoblotted with anti-Flag. K, Lysates from cells expressing the control or the indicated constructs were immunoprecipitated with IgG or anti-GSDME and then immunoblotted with anti-HA. L, Western blotting of USP48 and GSDME in raptinal-treated 293T cells and HeLa cells transfected with control vector, USP48 or sh-USP48. Data are shown as mean ± SD of three independent experiments. ***, P < 0.001.

Figure 2.

GSDME is a downstream target protein of USP48. A, Detection of USP48 expression in 6 different human tumor cell lines. B, Stable overexpression of USP48 in 293T and stable knockdown of USP48 in HeLa cells. C, The heat map summarizes the difference in protein expression in 293T cells with normal expression of USP48 and overexpression of USP48. D, The Flag-USP48 pulldown product from 293T cells was separated by SDS-PAGE, visualized by silver staining, and the protein interacting with USP48 was identified by mass spectrometry. E, Joint analysis of proteomics results and mass spectrometry sequencing results to get an intersection. F and G, Verification of the results obtained in C by immunoblotting. H, Detection of GSDME and GSDMD expression in 6 different human tumor cell lines. I, Western blotting of USP48 and GSDME in MCF7 and Panc-1. J, Lysates from cells expressing the control or the indicated constructs were immunoprecipitated (IP) with IgG or anti-USP48 and then immunoblotted with anti-Flag. K, Lysates from cells expressing the control or the indicated constructs were immunoprecipitated with IgG or anti-GSDME and then immunoblotted with anti-HA. L, Western blotting of USP48 and GSDME in raptinal-treated 293T cells and HeLa cells transfected with control vector, USP48 or sh-USP48. Data are shown as mean ± SD of three independent experiments. ***, P < 0.001.

Close modal

GSDME is a newly discovered important protein involved in the regulation of pyroptosis and an important tumor suppressor. In cells expressing GSDME, GSDME can be cleaved and activated by caspase-3, resulting in cell pyroptosis (8). Subsequently, we further confirmed the physical interaction between USP48 and GSDME by co-IP (Fig. 2J and K). Subsequently, we confirmed the regulatory effect of USP48 on GSDME through IHC and Western blotting in human pancreatic cancer and paraneoplastic tissues, as well as human liver cancer and paraneoplastic tissues, and confirmed the positive correlation between USP48 and GSDME (Supplementary Fig. S3A–S3H). In addition, we also found that USP48 did not affect the cleavage of GSDME (Fig. 2L). In summary, we found that USP48 could regulate the expression of GSDME (but does not affect its cleavage) and has a direct interaction with GSDME.

USP48 affects pyroptosis by regulating the expression of GSDME

Due to the key role of GSDME in activating cell pyroptosis, we speculate that GSDME also plays an important role in the regulation of cell pyroptosis by USP48. To confirm this hypothesis, we established a 293T cell line that simultaneously expressed USP48 and short hairpin RNA (shRNA) targeting GSDME (Fig. 3A). As expected, the downregulation of GSDME substantially rescued the promotion of overexpression of USP48 on the production of LDH (Fig. 3B), IL1β (Fig. 3C), and IL18 (Fig. 3D) in 293T cells after treatment with raptinal. Conversely, in HeLa cell lines expressing both shUSP48 and GSDME (Fig. 3E), it was also confirmed that overexpression of GSDME could rescue the inhibitory effect of USP48 knockdown on the production of LDH (Fig. 3F), IL1β (Fig. 3G), and IL18 (Fig. 3H) in 293T cells after treatment with raptinal. Subsequently, we also used time-lapse microscopy to observe the morphology and absorption of SYTOX Green in the 293T cell line expressing USP48 and shRNA targeting GSDME after treatment with raptinal and consistent results were obtained (Fig. 3I and J). In addition, we also obtained results consistent with the above in L929 cells (Supplementary Fig. S4A–S4I). Interestingly, although the study confirmed that GSDME could be cleaved by caspase-3 and induce pyroptosis, our results suggest that USP48 did not achieve regulatory effects on GSDME and pyroptosis through caspase-3. Activating or overexpressing caspase-3 in 293T cells with USP48 knockdown did not affect the ability of USP48 to regulate GSDME and pyroptosis, suggests that USP48 regulates pyroptosis by affecting the expression of GSDME rather than the cleavage of it by caspase-3 (Supplementary Fig. S5A–S5I). In conclusion, we confirmed that USP48 may promote the expression of GSDME in a way that was independent of caspase-3, thereby promoting the occurrence of pyroptosis.

Figure 3.

USP48 regulates pyroptosis through GSDME. A, Western blotting of USP48 and GSDME in 293T cells transfected with different combinations of control vector, USP48 vector, and/or sh-GSDME. B, Measurement of LDH production in transformed cells described in A. C, Measurement of IL1β production in transformed cells described in A. D, Measurement of IL18 production in transformed cells described in A. E, Western blotting of USP48 and GSDME in HeLa cells transfected with different combinations of control vector, shUSP48 vector, and/or GSDME. F, Measurement of LDH production in transformed cells described in E. G, Measurement of IL1β production in transformed cells described in E. H, Measurement of IL18 production in transformed cells described in E. I and J, Time-lapse microscopy observed the morphologic changes and the absorption of SYTOX Green in 293T cells transfected with different combinations of control vector, USP48 vector, and/or sh-GSDME. Data are shown as mean ± SD of three independent experiments. **, P < 0.01. Scale bar, 20 μm.

Figure 3.

USP48 regulates pyroptosis through GSDME. A, Western blotting of USP48 and GSDME in 293T cells transfected with different combinations of control vector, USP48 vector, and/or sh-GSDME. B, Measurement of LDH production in transformed cells described in A. C, Measurement of IL1β production in transformed cells described in A. D, Measurement of IL18 production in transformed cells described in A. E, Western blotting of USP48 and GSDME in HeLa cells transfected with different combinations of control vector, shUSP48 vector, and/or GSDME. F, Measurement of LDH production in transformed cells described in E. G, Measurement of IL1β production in transformed cells described in E. H, Measurement of IL18 production in transformed cells described in E. I and J, Time-lapse microscopy observed the morphologic changes and the absorption of SYTOX Green in 293T cells transfected with different combinations of control vector, USP48 vector, and/or sh-GSDME. Data are shown as mean ± SD of three independent experiments. **, P < 0.01. Scale bar, 20 μm.

Close modal

USP48 prevents the degradation of GSDME by deubiquitinating it

The above results have preliminarily confirmed that USP48 exerts its regulation of pyroptosis through physical interaction with GSDME, but the specific mechanism is still unclear. USP48 is an important member of the deubiquitinating enzyme family. Studies have confirmed that it is involved in regulating the stability of Gli1, TRAF2, Mdm2, and many other proteins (31–34). Therefore, we speculate that USP48 deubiquitinates GSDME, thereby inhibiting the proteasome degradation of GSDME. The ubiquitin-proteasome pathway is an important protein degradation regulatory system in cells.

To further confirm whether USP48 affects the stability of GSDME, we found that the proteasome-specific inhibitor MG132 could effectively reverse the impact of USP48 knockdown on GSDME (Fig. 4A) and through cycloheximide chase analysis to evaluate the potential of USP48 in regulating GSDME protein turnover rate, which shows that the reduction of the USP48 level is obviously related to the shortening of the half-life of GSDME (Fig. 4B). Interestingly, we found that the change of USP48 expression had no effect on GSDME at the RNA level (Supplementary Fig. S6A and S6B), which was also consistent with USP48 affecting the protein stability of GSDME. In view of the above observations, we established 293T cells with Dox-induced wild-type USP48 (USP48/WT) and catalytically inactive mutant USP48 (USP48/C98A). It was found that only wild-type USP48 gradually increased GSDME levels in a Dox dose-dependent manner (Fig. 4C), while no significant changes in GSDME protein levels were detected in cells expressing USP48/C98A (Fig. 4D). Through further experiments, we found that the overexpression of wild-type USP48 significantly reduced the ubiquitination level of GSDME, while 293T cells expressing USP48/C98A had no effect on the ubiquitination of GSDME (Fig. 4E). In contrast, knockdown of USP48 led to the accumulation of ubiquitinated GSDME (Fig. 4F). We also further confirmed this result by in vitro ubiquitination experiments (Supplementary Fig. S6C). The above results were further confirmed in L929 cells (Supplementary Fig. S6D–S6F). Besides, we also found that overexpression of catalytically inactive mutant USP48 in 293T cells had no effect on pyroptosis (Supplementary Fig. S6G–S6J).

Figure 4.

USP48 inhibits the degradation of GSDME by deubiquitinating it. A, In the absence or presence of the proteasome inhibitor MG132 (50 μg/mL), the levels of USP48 and GSDME proteins in cells transfected with control or USP48 shRNA were detected by Western blotting. B, Within the specified hours, in the absence or presence of the protein synthesis inhibitor cycloheximide (CHX; 50 μg/mL), Western blotting was used to detect USP48 and GSDME in HeLa cells transfected with control or USP48 shRNA protein levels. C and D, Western blotting to measure USP48 and GSDME protein levels in cells with Dox-inducible expression of wild-type USP48 (USP48/WT; C) and a catalytically inactive mutant of USP48 (USP48/C98A; D). E, Lysates from cells expressing GSDME, USP48, and HA-Ub were pulled down with anti-Myc and then immunoblotted with anti-HA (top). Input is immunoblotting with anti-Myc and anti-Flag (bottom). β-Actin was used as a loading control. F, Lysates from cells expressing GSDME, shUSP48, and HA-Ub were pulled down with anti-Flag and then immunoblotted with anti-HA (top). Input is immunoblotting with anti-USP48 and anti-Flag (bottom). β-Actin was used as a loading control. G, Lysates from cells expressing GSDME, USP48, and ubiquitin mutants were pulled down with anti-Myc and then immunoblotted with anti-HA (top). Input is immunoblotting with anti-Myc and anti-Flag (bottom). β-Actin was used as a loading control. H, A series of GSDME ubiquitination site mutants. I, The cell lysate expressing the designated GSDME ubiquitination site mutant was pulled down with anti-Myc and then immunoblotted with anti-HA (top). Input is immunoblotting with anti-Myc (bottom). β-Actin was used as a loading control. J, The lysine site associated with GSDME was mutated to arginine. K, Using anti-Myc to pull down the lysate of cells expressing the specified GSDME mutant and then using anti-HA for immunoblotting (top). Input is immunoblotting with anti-Myc (bottom). β-Actin was used as a loading control. Data are shown as mean ± SD of three independent experiments. **, P < 0.01.

Figure 4.

USP48 inhibits the degradation of GSDME by deubiquitinating it. A, In the absence or presence of the proteasome inhibitor MG132 (50 μg/mL), the levels of USP48 and GSDME proteins in cells transfected with control or USP48 shRNA were detected by Western blotting. B, Within the specified hours, in the absence or presence of the protein synthesis inhibitor cycloheximide (CHX; 50 μg/mL), Western blotting was used to detect USP48 and GSDME in HeLa cells transfected with control or USP48 shRNA protein levels. C and D, Western blotting to measure USP48 and GSDME protein levels in cells with Dox-inducible expression of wild-type USP48 (USP48/WT; C) and a catalytically inactive mutant of USP48 (USP48/C98A; D). E, Lysates from cells expressing GSDME, USP48, and HA-Ub were pulled down with anti-Myc and then immunoblotted with anti-HA (top). Input is immunoblotting with anti-Myc and anti-Flag (bottom). β-Actin was used as a loading control. F, Lysates from cells expressing GSDME, shUSP48, and HA-Ub were pulled down with anti-Flag and then immunoblotted with anti-HA (top). Input is immunoblotting with anti-USP48 and anti-Flag (bottom). β-Actin was used as a loading control. G, Lysates from cells expressing GSDME, USP48, and ubiquitin mutants were pulled down with anti-Myc and then immunoblotted with anti-HA (top). Input is immunoblotting with anti-Myc and anti-Flag (bottom). β-Actin was used as a loading control. H, A series of GSDME ubiquitination site mutants. I, The cell lysate expressing the designated GSDME ubiquitination site mutant was pulled down with anti-Myc and then immunoblotted with anti-HA (top). Input is immunoblotting with anti-Myc (bottom). β-Actin was used as a loading control. J, The lysine site associated with GSDME was mutated to arginine. K, Using anti-Myc to pull down the lysate of cells expressing the specified GSDME mutant and then using anti-HA for immunoblotting (top). Input is immunoblotting with anti-Myc (bottom). β-Actin was used as a loading control. Data are shown as mean ± SD of three independent experiments. **, P < 0.01.

Close modal

Ubiquitin can posttranslationally modify proteins. The binding of a single ubiquitin molecule to a target protein is referred to as monoubiquitination, and additional ubiquitin portions can be spliced to that initial ubiquitin to form polyUb chains. These polyUb chains can be linked by isopeptide bond formation between the C-terminus and the lysine of the target protein (K6, K11, K27, K29, K33, K48, and K63). Two of the most fully characterized forms of polyubiquitination occur by attachment to lysine 48 (K48) or lysine 63 (K63; ref. 35). We also found that USP48 affects K48-linked ubiquitin in GSDME, but it has no effect on K63-linked ubiquitination (Fig. 4G). Then, we predicted the ubiquitination sites of GSDME through the website (http://plmd.biocuckoo.org/index.php) and verified the seven predicted sites (K30, K39, K120, K161, K189, K240, and K248) through in vitro ubiquitination experiments. To examine the K48-linked ubiquitination sites, HEK293T cells were cotransfected with the seven Myc tagged mutants of GSDME and HA-ubiquitin K48. An in vitro ubiquitination assay showed that ubiquitin K48 ubiquitinated GSDME at the K30, K120 and K189 sites (Fig. 4H and I). Next, to identify the target residues of GSDME modified by USP48, we constructed corresponding point mutants at positions K30, K120, and K189 of GSDME (Fig. 4J), HEK293T cells were cotransfected with the GSDME mutant labeled with Myc, HA-ubiquitin K48, and Flag-USP48. The results showed that USP48 inhibited the K48-linked ubiquitination of GSDME at K120 (Fig. 4K). Taken together, these results indicate that USP48 prevents the degradation of GSDME by inhibiting K48-linked ubiquitination at position K120 of GSDME.

USP48 affects antitumor immunity by regulating the expression of GSDME

Our above results have initially confirmed that USP48 can promote the expression of GSDME, thereby promoting the occurrence of pyroptosis in cells. Pancreatic cancer is one of the most common malignant tumors of digestive tract, and PDAC is one of the most common pathologic types. It has been shown that approximately 90% of patients with PDAC carry mutations in the KRAS gene, and crossing LSL-KrasG12D mice with pancreatic-specific cre (Pdx1-cre) mice can conditionally mutate the KRAS gene in pancreatic tissue and increase the probability of PDAC. In the previous research, our group has made an in-depth exploration on the occurrence and progress of PDAC, has a certain research foundation, and successfully constructed a mouse model of pancreatic spontaneous tumor (36). To further demonstrate the functionality of USP48, a group of KRASG12D; PDX1-Cre (KC) mice were crossed with Usp48 systemic knockout mice (Usp48flox/flox) to establish a USP48-deficient mouse model of spontaneous pancreatic tumorigenesis (KUC). The expression of USP48 and GSDME in the tissues was then detected by IHC and immunofluorescence using pancreatic tissue from the above mice and laboratory-preserved liver tissue from mice lacking USP48 specificity (obtained by crossing USP48fl/fl mice with Alb-cre mice; ref. 29), and the results showed that the expression levels of USP48 were positively correlated with the expression of GSDME (Fig. 5AF). Western blotting was used to detect the expression of USP48 and GSDME in the liver and pancreas of the above mice, consistent results were obtained. Furthermore, loss of USP48 in mouse tissues also had no effect on GSDMD expression (Fig. 5G and H). The above results further confirmed the regulatory effect of USP48 on GSDME in vivo. It is worth noting that compared with KC mice, the loss of USP48 significantly promoted the occurrence of pancreatic tumors, which was consistent with our previous findings in HCC (29). KrasG12D;USP48flox/flox;PDX1-Cre (KUC) mice had already developed acinar ductal metaplasia at about 3 months and obvious cancer occurred in about 5 months. The deletion of USP48 also significantly reduced the survival time of mice (Supplementary Fig. S7A–S7G). The function and mechanism of USP48 in regulating pancreatic tumorigenesis need to be further studied.

Figure 5.

USP48 is involved in regulating antitumor immunity. A, IHC was used to detect the expression of USP48 and GSDME in liver cancer tissues of Alb-Cre; USP48flox/flox and Alb-Cre; USP48WT mice. B, Immunofluorescence was used to detect the expression of USP48 and GSDME in liver cancer tissues of Alb-Cre; USP48flox/flox and Alb-Cre; USP48WT mice. C, Quantification of B. D, IHC was used to detect the expression of USP48 and GSDME in pancreatic tissues of KC mice and KUC mice. E, Immunofluorescence was used to detect the expression of USP48 and GSDME in pancreatic tissues of KC mice and KUC mice. F, Quantification of E. G, Western blotting was used to detect the expression of USP48, GSDME, and GSDMD in liver cancer tissues of Alb-Cre; USP48flox/flox mice and Alb-Cre; USP48WT mice. H, Western blotting was used to detect the expression of USP48, GSDME and GSDMD in pancreatic tissues of KC mice and KUC mice. I, Multicolor immunofluorescence to detect the distribution of CD8+ T cells in the pancreatic tissues of mice. J, Multicolor immunofluorescence to detect the distribution of CD4+ T cells in the pancreatic tissues of mice. K, Flow cytometry assay to detect the proportion of NK cells, CD8+ T cells, TAMs, and Tregs in the pancreatic cancer tissues of KC mice and KUC mice. Data are shown as mean ± SD of three independent experiments. ***, P < 0.001; n.s., not significant. Scale bar, 20 μm.

Figure 5.

USP48 is involved in regulating antitumor immunity. A, IHC was used to detect the expression of USP48 and GSDME in liver cancer tissues of Alb-Cre; USP48flox/flox and Alb-Cre; USP48WT mice. B, Immunofluorescence was used to detect the expression of USP48 and GSDME in liver cancer tissues of Alb-Cre; USP48flox/flox and Alb-Cre; USP48WT mice. C, Quantification of B. D, IHC was used to detect the expression of USP48 and GSDME in pancreatic tissues of KC mice and KUC mice. E, Immunofluorescence was used to detect the expression of USP48 and GSDME in pancreatic tissues of KC mice and KUC mice. F, Quantification of E. G, Western blotting was used to detect the expression of USP48, GSDME, and GSDMD in liver cancer tissues of Alb-Cre; USP48flox/flox mice and Alb-Cre; USP48WT mice. H, Western blotting was used to detect the expression of USP48, GSDME and GSDMD in pancreatic tissues of KC mice and KUC mice. I, Multicolor immunofluorescence to detect the distribution of CD8+ T cells in the pancreatic tissues of mice. J, Multicolor immunofluorescence to detect the distribution of CD4+ T cells in the pancreatic tissues of mice. K, Flow cytometry assay to detect the proportion of NK cells, CD8+ T cells, TAMs, and Tregs in the pancreatic cancer tissues of KC mice and KUC mice. Data are shown as mean ± SD of three independent experiments. ***, P < 0.001; n.s., not significant. Scale bar, 20 μm.

Close modal

Pyroptosis is a natural immune response (37). Previous studies have shown that GSDME inhibits tumor growth in mice by promoting cell antitumor immunity. Compared with control mice, mice expressing GSDME had more NK cells and CD8+ cytotoxic T killer cells in tumors and expressed more toxic proteins and cytokines (16). Therefore, we speculate that USP48 may exert its antitumor effect by promoting the expression of GSDME and then promoting the antitumor immune response of cells. To confirm the hypothesis, we explored the effect of the loss of USP48 on the immune cell population in PDAC tissues. The immunofluorescence results showed that the tumor-infiltrating CD8+ and CD4+ T cells were significantly reduced in KUC mice (Fig. 5I and J). The proportion of NK cells and CD8+ cytotoxic T killer cells was significantly reduced, but the proportion of TAMs and Treg cells was significantly increased (Fig. 5K). Together, these data imply that USP48 can affect the antitumor immunity of cells by regulating the expression of GSDME.

Single-cell sequencing experiments clarify the regulatory effect of USP48 on antitumor immunity

We performed single-cell sequencing in pancreatic tissues of KC and KUC mice to determine the antitumor immune function of USP48. Single-cell RNA-seq was performed using the 10x chromium method. Consistent with the enrichment approach, all cells were positive for the CD45 gene (PTPRC) irrespective of treatment (Fig. 6A). We applied the size clustering algorithm as a quality control indicator for cells. Figure 6B shows the cell distribution in pancreatic tumor tissues of KC and KUC mice. Then, we applied the size clustering algorithm as the quality control index of the cells, divided the cells into groups, and defined five cell populations that capture the TAMs, dendritic cells (DC), T cells, monocyte-1 cells, and monocyte-2 cells (Fig. 6B and C). By comparing the differences of cell subgroups in KC and KUC pancreatic cancer tissues, we found that there were significant differences in the number of cells in TAMs and monocyte-2 cells. Specifically, the downregulation of USP48 expression significantly increased the number of TAMs and monocyte-2 cells (Fig. 6D). We showed the difference of marker genes in each cell subgroup through heat maps and dot maps (Cd3d = T cells; Cd14 = monocytes; C1qb = TAMs; S100a3 and Thbs1 = monocytes; and Clec9a = DCs; Fig. 6E and F) and obtained the number and proportion of each cell subgroup in KC and KUC pancreatic cancer tissues (Fig. 6G).

Figure 6.

Single-cell analysis of KC and KC;USP48−/− CD45+ cells in pancreatic cancer. A, KC (n = 2) and KC; USP48−/− (n = 2) pancreatic cancer tissues were collected for digestion to obtain single cells, CD45-labeled magnetic beads were used for analysis, and sequenced after 10x Genomics library construction. B, UMAP displayed cell distribution in KC (n = 2) and KC; USP48−/− (n = 2). C, Group annotation of the obtained single cells. D, Comparison of KC (n = 2) and KC;USP48−/− (n = 2) differences in cell subgroups in pancreatic cancer tissues (it can be clearly seen that there are obvious differences in the number of TAMs and monocyte-2 cells). E, Marker genes in each cell subgroup (as a basis for grouping). F, Dot diagram showing each subgroup marker gene (Cd3d, T cells; Cd14, monocytes; C1qb, ATMs; S100a3 and Thbs1, monocytes; Clec9a, DCs). G, The proportion of each subgroup of cells in KC (n = 2) and KC; USP48−/− (n = 2) pancreatic cancer tissues.

Figure 6.

Single-cell analysis of KC and KC;USP48−/− CD45+ cells in pancreatic cancer. A, KC (n = 2) and KC; USP48−/− (n = 2) pancreatic cancer tissues were collected for digestion to obtain single cells, CD45-labeled magnetic beads were used for analysis, and sequenced after 10x Genomics library construction. B, UMAP displayed cell distribution in KC (n = 2) and KC; USP48−/− (n = 2). C, Group annotation of the obtained single cells. D, Comparison of KC (n = 2) and KC;USP48−/− (n = 2) differences in cell subgroups in pancreatic cancer tissues (it can be clearly seen that there are obvious differences in the number of TAMs and monocyte-2 cells). E, Marker genes in each cell subgroup (as a basis for grouping). F, Dot diagram showing each subgroup marker gene (Cd3d, T cells; Cd14, monocytes; C1qb, ATMs; S100a3 and Thbs1, monocytes; Clec9a, DCs). G, The proportion of each subgroup of cells in KC (n = 2) and KC; USP48−/− (n = 2) pancreatic cancer tissues.

Close modal

To identify the immune cell subpopulations regulated by USP48, we first performed a subpopulation analysis of T cells in KC and KUC pancreatic cancer tissues (Fig. 7A). It was found that the proportion of exhaustible T (Tex) cells and Treg cells in KUC pancreatic cancer tissues increased significantly, indicating that the reduction of USP48 expression inhibited the antitumor immunity (Fig. 7B). We also used UMAP to show the expression of standard marker genes for T-cell subgroup classification and the ratio of T-cell subgroups in KC and KUC pancreatic cancer tissues, further confirming the above results. (Figure 7C and D). Subsequently, we used multicolor immunofluorescence experiments to detect the distribution of Tex cells and Treg cells in pancreatic cancer tissues of KC and KUC mice, and the results obtained were consistent with the sequencing results (Fig. 7E and F).

Figure 7.

Knockout of USP48 increases the proportion of depleted T cells, Tregs, and TAMs infiltrating pancreatic cancer. A, KC and KC; USP48−/− pancreatic cancer tissue T-cell subgroup analysis. B, UMAP shows the difference between KC and KC; USP48−/− pancreatic cancer T-cell subpopulation. C, UMAP shows the standard marker gene for T-cell subgroup classification. D, Histogram shows the proportions of KC and KC; USP48−/−pancreatic cancer cells of tissue T-cell subsets. E, Multicolor fluorescent staining of KC and KC; USP48−/− Tex cell levels in pancreatic cancer tissues. F, Multicolor fluorescent staining of KC and KC; USP48−/− Treg cells in pancreatic cancer tissues. G, UMAP showing KC and KC; USP48−/− TAMs in pancreatic cancer tissues. H, Violin image showing specific marker genes in TAMs. I, Multicolor fluorescent staining of KC and KC; USP48−/− TAM cell levels in pancreatic cancer tissues. **, P < 0.01; ***, P < 0.001. Scale bar, 20 μm.

Figure 7.

Knockout of USP48 increases the proportion of depleted T cells, Tregs, and TAMs infiltrating pancreatic cancer. A, KC and KC; USP48−/− pancreatic cancer tissue T-cell subgroup analysis. B, UMAP shows the difference between KC and KC; USP48−/− pancreatic cancer T-cell subpopulation. C, UMAP shows the standard marker gene for T-cell subgroup classification. D, Histogram shows the proportions of KC and KC; USP48−/−pancreatic cancer cells of tissue T-cell subsets. E, Multicolor fluorescent staining of KC and KC; USP48−/− Tex cell levels in pancreatic cancer tissues. F, Multicolor fluorescent staining of KC and KC; USP48−/− Treg cells in pancreatic cancer tissues. G, UMAP showing KC and KC; USP48−/− TAMs in pancreatic cancer tissues. H, Violin image showing specific marker genes in TAMs. I, Multicolor fluorescent staining of KC and KC; USP48−/− TAM cell levels in pancreatic cancer tissues. **, P < 0.01; ***, P < 0.001. Scale bar, 20 μm.

Close modal

In the above results, we have initially found that knockdown of USP48 significantly increased the proportion of TAM subgroups (Fig. 7G and H). Therefore, further analysis of TAMs was performed, and the same results were obtained. Multicolor immunofluorescence experiments also confirmed that the lack of USP48 increased the distribution of TAMs in pancreatic cancer tissues (Fig. 7I). In summary, we applied single-cell RNA-seq to confirm that USP48 deletion inhibited the antitumor immunity of pancreatic cancer cells.

Moreover, we applied the GSDME knockdown mouse pancreatic cancer cell line Pan02 and constructed a pancreatic tumor model in C57/B6 mice by in situ injection, and multicolor fluorescence assay and flow cytometric assay obtained the expected results that GSDME knockdown suppressed the antitumor ability of pancreatic cancer cells. This finding tentatively confirms that GSDME may play a key role in regulating the antitumor immunity of pancreatic cancer cells by USP48, but the exact mechanism remains to be further investigated (Supplementary Fig. S8A–S8F).

USP48-GSDME modulates the sensitivity of mice to anti–PD-1 immunotherapy

To further clarify the role of USP48 in antitumor immunity, Pan02 cells overexpressing USP48 or empty vectors were subcutaneously implanted into C57/B6 mice. The mice were treated with aPD-1 (an anti–PD-1 antibody) on days 7, 11 and 15, and the growth of the mice was continuously observed and killed on day 30 (Fig. 8A). PD-1 antibody is an important immunotherapy. It has a good therapeutic effect in the treatment of more than 10 malignant tumors, such as lung cancer, lymphoma, liver cancer, gastric cancer, and pancreatic cancer (38). The results showed that, compared with the empty vector, overexpression of USP48 significantly inhibited the growth of tumors in mice, but overexpression of the USP48 mutant (USP48/C98A) had no inhibitory effect on tumor growth in mice in vivo. Overexpression of USP48 also significantly improved the sensitivity of mice to anti–PD-1 treatment (Fig. 8BE). To clarify its mechanism, we expressed shGSDME or scramble in Pan02 cells overexpressing USP48 and injected them subcutaneously into C57/B6 mice. Consistent with the expected results, knocking down the expression of GSDME in Pan02 mice overexpressing USP48 can effectively reverse the inhibitory effect of USP48 overexpression on tumor growth and the promotion of anti–PD-1 treatment sensitivity (Fig. 8F and I).

Figure 8.

USP48-GSDME affects the sensitivity of mice to anti–PD-1 immunotherapy. A, Experimental flow chart. B and F, Tumor photos after subcutaneous implantation of Pan02 cells under different treatments. Each group contains 10 mice. C and G, Tumor weight after 30 days of subcutaneous implantation of Pan02 cells under different treatments. Each group contains 10 mice. D and H, Volume changes of tumors within 30 days after subcutaneous implantation of Pan02 cells under different treatments. Each group contains 10 mice. E and I, Survival statistics of C57/B6 mice after subcutaneous implantation of Pan02 cells under different treatments. Each group contains 10 mice. J and K, Flow cytometry assay to detect the proportion of TAMs and NK cells in the tumor tissue. Data are shown as mean ± SD of three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001; n.s., not significant.

Figure 8.

USP48-GSDME affects the sensitivity of mice to anti–PD-1 immunotherapy. A, Experimental flow chart. B and F, Tumor photos after subcutaneous implantation of Pan02 cells under different treatments. Each group contains 10 mice. C and G, Tumor weight after 30 days of subcutaneous implantation of Pan02 cells under different treatments. Each group contains 10 mice. D and H, Volume changes of tumors within 30 days after subcutaneous implantation of Pan02 cells under different treatments. Each group contains 10 mice. E and I, Survival statistics of C57/B6 mice after subcutaneous implantation of Pan02 cells under different treatments. Each group contains 10 mice. J and K, Flow cytometry assay to detect the proportion of TAMs and NK cells in the tumor tissue. Data are shown as mean ± SD of three independent experiments. *, P < 0.05; **, P < 0.01; ***, P < 0.001; n.s., not significant.

Close modal

Subsequently, we performed flow cytometry in mouse tumors, and the results were consistent with the previous results, namely, overexpression of USP48 can reduce the proportion of TAMs in mouse tumor tissues, while increasing the level of NK cells and knocking down GSDME can reverse this result (Fig. 8J and K). These results further confirmed the important role of the USP48-GSDME pathway in antitumor immunity and discovered its key regulatory role in tumor immunotherapy.

Protein ubiquitination is a strictly controlled process and is reversible. It can be reversed by a large group of proteases called deubiquitinating enzymes. Most deubiquitinating enzymes can break down and release ubiquitin from the substrate protein for editing, and some deubiquitinating enzyme are related to editing or processing ubiquitin-like proteins and binding proteins (39). USP48 is a deubiquitinating enzyme expressed in almost all human tissues. It has been shown that USP48 has a key role in the development and progression of numerous diseases. Cetkovska and colleagues found a stabilizing effect of USP48 on Mdm2 in osteosarcoma cells, U2OS and H1299 lung cancer cells, and overexpression of Mdm2 was associated with loss of p53 tumor suppressor activity in several human cancers (33). In addition, in glioblastoma, USP48 can activate Gli-dependent transcription by stabilizing the Gli1 protein, with implications for cell proliferation and tumorigenesis (31). Our previous study also demonstrated that USP48 can play a key regulatory role in hepatocarcinogenesis and development of liver cancer by regulating the stability of SIRT6 (29). However, the regulation of USP48 in pyroptosis and antitumor immunity has not been investigated.

Pyroptosis, a type of nonclassical apoptotic pathway, is a mechanism of programmed necrosis in inflammatory cells (40). The concept of pyroptosis was first proposed by Cookson and Brennan in 2001 to describe a caspase-1–dependent mode of cell death in inflammatory cells (41). However, for a long period of time, the mechanism of pyroptosis has not made great progress. Until 2015, Academician Shao Feng from the Beijing Institute of Life Sciences reported that the process of pyroptosis would be accompanied by the cleavage of GSDMD by caspase, and the cleaved GSDMD would form holes in the cell membrane, leading to the release of IL1β, IL18, and other inflammatory factors (27). Up to this point, the mechanism of pyroptosis had been studied clearly. In this study, we applied CRISPR-Cas9 high-throughput screening technology and found that USP48 had a significant regulatory effect on pyroptosis.

GSDME is also an important member of the gasdermin family. Previously, it was shown that the promoter of GSDME was methylated in a variety of cancer cells, and this epistatic modification inhibited its expression in cancer cells. In addition, it was also found that GSDME could inhibit the development of a variety of cancers, including breast cancer and melanoma (42, 43). In 2017, Shao Feng and colleagues revealed its key role in pyroptosis and found that GSDME could be cleaved and activated by caspase-3, which in turn converts apoptosis into pyroptosis (7). Interestingly, our results found that USP48 could affect the expression of GSDME (and thus pyroptosis) through physical interaction with GSDME and does not affect the activation and cleavage of GSDME by caspase-3.

Our study revealed the regulatory role of USP48 on pyroptosis and its molecular mechanism. We found that USP48 could reduce the degradation of GSDME by deubiquitinating it, and ultimately promoting the development of pyroptosis. Unlike immunosilencing apoptosis, pyroptosis is characterized by cell membrane rupture in which numerous cytokines and danger signaling molecules are released. The occurrence of pyroptosis is usually accompanied by the activation of immune system and inflammatory reaction (44). In 2020, Judy Lieberman's group at Harvard Medical School and Boston Children's Hospital found that granzyme B from NK cells could directly cleave GSDME and activate pyroptosis, which occurs to further activate the antitumor immune response and inhibit tumor growth (16). It has also been shown that USP48 is a nuclear DUB regulated by casein kinase-2, which together with the COP9 signalosome controls UPS-dependent activation of NF-κB/RelA turnover in the nucleus. The decisive role of NF-κB in cellular processes including inflammation, immunity and cell survival (45). In view of these studies, we speculate that USP48 also has a regulatory role on cellular antitumor immunity, which was confirmed by both single-cell sequencing technology and flow cytometry. We found that USP48 could regulate the ratio and number of TAMs, NK cells, Tregs and other immune cells.

The tumor immune microenvironment is a highly complex system. It has also become an important research hotspot in recent years. With the deepening of immunotherapy research, the efficacy of immunotherapy has been continuously improved. PD-1 inhibitors are a new class of drugs that block PD-1 and are used to treat certain types of cancer by activating the immune system to attack tumors (46, 47). We found that changes in USP48 expression significantly modulated the efficacy of PD-1 inhibitors by constructing a xenograft tumor model in C57/B6 mice, and increased USP48 expression significantly improved the therapeutic effect of PD-1 inhibitors in mouse tumors.

In conclusion, our study identified the key regulatory role of USP48 on pyroptosis and elucidated its molecular mechanism, in addition to the role of USP48 in antitumor immunity and immunotherapy. Therefore, specifically targeting the USP48 or USP48-GSDME axis may be a potential future therapeutic strategy. Nevertheless, our study leaves much to be desired, as we have only described the regulatory role of USP48 in antitumor immunity and immunotherapy. But, the specific molecular mechanisms need to be further elucidated.

No disclosures were reported.

Y. Ren: Validation, investigation, writing–original draft. M. Feng: Data curation, formal analysis. X. Hao: Writing–review and editing. X. Liu: Validation. J. Li: Data curation. P. Li: Visualization. J. Gao: Software. Q. Qi: Funding acquisition. L. Du: Writing–review and editing. C. Wang: Supervision. Q. Wang: Resources. Y. Wang: Supervision, project administration, writing–review and editing.

This research was supported by grants from National Natural Science Foundation of China (82172350, 81874040, and 82002228), Key Research and Development Program of Shandong (2019GSF108218 and 2019GSF108139), Taishan Scholars Climbing Program of Shandong Province (tspd20210323), Young Taishan Scholars Program of Shandong Province (tsqn201909176 and tsqn201909177), Natural Science Foundation of Shandong Province (ZR2020QH280 and ZR2020MH007), and Qilu Young Scholars Program of Shandon University.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

1.
Kovacs
SB
,
Miao
EA
.
Gasdermins: effectors of pyroptosis
.
Trends Cell Biol
2017
;
27
:
673
84
.
2.
Zhou
Z
,
He
H
,
Wang
K
,
Shi
X
,
Wang
Y
,
Su
Y
, et al
.
Granzyme a from cytotoxic lymphocytes cleaves GSDMB to trigger pyroptosis in target cells
.
Science
2020
;
368
:
eaaz7548
.
3.
Humphries
F
,
Shmuel-Galia
L
,
Ketelut-Carneiro
N
,
Li
S
,
Wang
B
,
Nemmara
VV
, et al
.
Succination inactivates gasdermin D and blocks pyroptosis
.
Science
2020
;
369
:
1633
7
.
4.
Ding
J
,
Wang
K
,
Liu
W
,
She
Y
,
Sun
Q
,
Shi
J
, et al
.
Pore-forming activity and structural autoinhibition of the gasdermin family
.
Nature
2016
;
535
:
111
6
.
5.
Wang
Q
,
Wang
Y
,
Ding
J
,
Wang
C
,
Zhou
X
,
Gao
W
, et al
.
A bioorthogonal system reveals antitumor immune function of pyroptosis
.
Nature
2020
;
579
:
421
6
.
6.
Xia
X
,
Wang
X
,
Cheng
Z
,
Qin
W
,
Lei
L
,
Jiang
J
, et al
.
The role of pyroptosis in cancer: pro-cancer or pro-"host"?
Cell Death Dis
2019
;
10
:
650
.
7.
Wang
Y
,
Gao
W
,
Shi
X
,
Ding
J
,
Liu
W
,
He
H
, et al
.
Chemotherapy drugs induce pyroptosis through caspase-3 cleavage of a gasdermin
.
Nature
2017
;
547
:
99
103
.
8.
Jiang
M
,
Qi
L
,
Li
L
,
Li
Y
.
The caspase-3/GSDME signal pathway as a switch between apoptosis and pyroptosis in cancer
.
Cell Death Discov
2020
;
6
:
112
.
9.
De Schutter
E
,
Croes
L
,
Ibrahim
J
,
Pauwels
P
,
Op de Beeck
K
,
Vandenabeele
P
, et al
.
GSDME and its role in cancer: from behind the scenes to the front of the stage
.
Int J Cancer
2021
;
148
:
2872
83
.
10.
Ibrahim
J
,
De Schutter
E
,
Op de Beeck
K
.
GSDME: a potential ally in cancer detection and treatment
.
Trends Cancer
2021
;
7
:
392
4
.
11.
Tan
G
,
Huang
C
,
Chen
J
,
Zhi
F
.
HMGB1 released from GSDME-mediated pyroptotic epithelial cells participates in the tumorigenesis of colitis-associated colorectal cancer through the ERK1/2 pathway
.
J Hematol Oncol
2020
;
13
:
149
.
12.
Zhang
X
,
Zhang
P
,
An
L
,
Sun
N
,
Peng
L
,
Tang
W
, et al
.
Miltirone induces cell death in hepatocellular carcinoma cell through GSDME-dependent pyroptosis
.
Acta Pharm Sin B
2020
;
10
:
1397
413
.
13.
Lu
Y
,
Hou
K
,
Li
M
,
Wu
X
,
Yuan
S
.
Exosome-delivered LncHEIH promotes gastric cancer progression by upregulating EZH2 and stimulating methylation of the GSDME promoter
.
Front Cell Dev Biol
2020
;
8
:
571297
.
14.
Cai
J
,
Yi
M
,
Tan
Y
,
Li
X
,
Li
G
,
Zeng
Z
, et al
.
Natural product triptolide induces GSDME-mediated pyroptosis in head and neck cancer through suppressing mitochondrial hexokinase-ΙΙ
.
J Exp Clin Cancer Res
2021
;
40
:
190
.
15.
Shen
X
,
Wang
H
,
Weng
C
,
Jiang
H
,
Chen
J
.
Caspase 3/GSDME-dependent pyroptosis contributes to chemotherapy drug-induced nephrotoxicity
.
Cell Death Dis
2021
;
12
:
186
.
16.
Zhang
Z
,
Zhang
Y
,
Xia
S
,
Kong
Q
,
Li
S
,
Liu
X
, et al
.
Gasdermin E suppresses tumor growth by activating antitumor immunity
.
Nature
2020
;
579
:
415
20
.
17.
Liu
Y
,
Fang
Y
,
Chen
X
,
Wang
Z
,
Liang
X
,
Zhang
T
, et al
.
Gasdermin E–mediated target cell pyroptosis by CAR T cells triggers cytokine release syndrome
.
Sci Immunol
2020
;
5
:
eaax7969
.
18.
Dikic
I
,
Wakatsuki
S
,
Walters
KJ
.
Ubiquitin-binding domains: from structures to functions
.
Nat Rev Mol Cell Biol
2009
;
10
:
659
71
.
19.
Ikeda
F
,
Dikic
I
.
Atypical ubiquitin chains: new molecular signals. ‘Protein modifications: beyond the usual suspects' review series
.
EMBO Rep
2008
;
9
:
536
42
.
20.
Ernst
A
,
Avvakumov
G
,
Tong
J
,
Fan
Y
,
Zhao
Y
,
Alberts
P
, et al
.
A strategy for modulation of enzymes in the ubiquitin system
.
Science
2013
;
339
:
590
5
.
21.
Harris
IS
,
Endress
JE
,
Coloff
JL
,
Selfors
LM
,
McBrayer
SK
,
Rosenbluth
JM
, et al
.
Deubiquitinases maintain protein homeostasis and survival of cancer cells upon glutathione depletion
.
Cell Metab
2019
;
29
:
1166
81
.
22.
Gutierrez-Diaz
BT
,
Gu
W
,
Ntziachristos
P
.
Deubiquitinases: pro-oncogenic activity and therapeutic targeting in blood malignancies
.
Trends Immunol
2020
;
41
:
327
40
.
23.
Ashkenazi
A
,
Bento
CF
,
Ricketts
T
,
Vicinanza
M
,
Siddiqi
F
,
Pavel
M
, et al
.
Polyglutamine tracts regulate beclin 1–dependent autophagy
.
Nature
2017
;
545
:
108
11
.
24.
Sharma
A
,
Alswillah
T
,
Singh
K
,
Chatterjee
P
,
Willard
B
,
Venere
M
, et al
.
USP14 regulates DNA damage repair by targeting RNF168-dependent ubiquitination
.
Autophagy
2018
;
14
:
1976
90
.
25.
Lork
M
,
Verhelst
K
,
Beyaert
R
.
CYLD, A20 and OTULIN deubiquitinases in NF-κB signaling and cell death: so similar, yet so different
.
Cell Death Differ
2017
;
24
:
1172
83
.
26.
Zhou
Y
,
Zhu
S
,
Cai
C
,
Yuan
P
,
Li
C
,
Huang
Y
, et al
.
High-throughput screening of a CRISPR/Cas9 library for functional genomics in human cells
.
Nature
2014
;
509
:
487
91
.
27.
Shi
J
,
Zhao
Y
,
Wang
K
,
Shi
X
,
Wang
Y
,
Huang
H
, et al
.
Cleavage of GSDMD by inflammatory caspases determines pyroptotic cell death
.
Nature
2015
;
526
:
660
5
.
28.
Harrigan
JA
,
Jacq
X
,
Martin
NM
,
Jackson
SP
.
Deubiquitylating enzymes and drug discovery: emerging opportunities
.
Nat Rev Drug Discov
2018
;
17
:
57
78
.
29.
Du
L
,
Li
Y
,
Kang
M
,
Feng
M
,
Ren
Y
,
Dai
H
, et al
.
USP48 is upregulated by Mettl14 to attenuate hepatocellular carcinoma via regulating SIRT6 stabilization
.
Cancer Res
2021
;
81
:
3822
34
.
30.
Frank
D
,
Vince
JE
.
Pyroptosis versus necroptosis: similarities, differences, and crosstalk
.
Cell Death Differ
2019
;
26
:
99
114
.
31.
Zhou
A
,
Lin
K
,
Zhang
S
,
Ma
L
,
Xue
J
,
Morris
SA
, et al
.
Gli1-induced deubiquitinase USP48 aids glioblastoma tumorigenesis by stabilizing Gli1
.
EMBO Rep
2017
;
18
:
1318
30
.
32.
Li
S
,
Wang
D
,
Zhao
J
,
Weathington
NM
,
Shang
D
,
Zhao
Y
.
The deubiquitinating enzyme USP48 stabilizes TRAF2 and reduces E-cadherin–mediated adherens junctions
.
FASEB J
2018
;
32
:
230
42
.
33.
Cetkovská
K
,
Šustová
H
,
Uldrijan
S
.
Ubiquitin-specific peptidase 48 regulates Mdm2 protein levels independent of its deubiquitinase activity
.
Sci Rep
2017
;
7
:
43180
.
34.
Uckelmann
M
,
Densham
RM
,
Baas
R
,
Winterwerp
HHK
,
Fish
A
,
Sixma
TK
, et al
.
USP48 restrains resection by site-specific cleavage of the BRCA1 ubiquitin mark from H2A
.
Nat Commun
2018
;
9
:
229
.
35.
Nakamura
N
.
Ubiquitin system
.
Int J Mol Sci
2018
;
19
:
1080
.
36.
Li
J
,
Zhan
H
,
Ren
Y
,
Feng
M
,
Wang
Q
,
Jiao
Q
, et al
.
Sirtuin 4 activates autophagy and inhibits tumorigenesis by upregulating the p53 signaling pathway
.
Cell Death Differ
2022
.
37.
Tang
R
,
Xu
J
,
Zhang
B
,
Liu
J
,
Liang
C
,
Hua
J
, et al
.
Ferroptosis, necroptosis, and pyroptosis in anticancer immunity
.
J Hematol Oncol
2020
;
13
:
110
.
38.
Baxi
S
,
Yang
A
,
Gennarelli
RL
,
Khan
N
,
Wang
Z
,
Boyce
L
, et al
.
Immune-related adverse events for anti–PD-1 and anti–PD-L1 drugs: systematic review and meta-analysis
.
BMJ
2018
;
360
:
k793
.
39.
Mevissen
TET
,
Komander
D
.
Mechanisms of deubiquitinase specificity and regulation
.
Annu Rev Biochem
2017
;
86
:
159
92
.
40.
Shi
J
,
Gao
W
,
Shao
F
.
Pyroptosis: gasdermin-mediated programmed necrotic cell death
.
Trends Biochem Sci
2017
;
42
:
245
54
.
41.
Cookson
BT
,
Brennan
MA
.
Pro-inflammatory programmed cell death
.
Trends Microbiol
2001
;
9
:
113
4
.
42.
Zhao
P
,
Wang
M
,
Chen
M
,
Chen
Z
,
Peng
X
,
Zhou
F
, et al
.
Programming cell pyroptosis with biomimetic nanoparticles for solid tumor immunotherapy
.
Biomaterials
2020
;
254
:
120142
.
43.
Erkes
DA
,
Cai
W
,
Sanchez
IM
,
Purwin
TJ
,
Rogers
C
,
Field
CO
, et al
.
Mutant BRAF and MEK inhibitors regulate the tumor immune microenvironment via pyroptosis
.
Cancer Discov
2020
;
10
:
254
69
.
44.
Hou
J
,
Hsu
JM
,
Hung
MC
.
Molecular mechanisms and functions of pyroptosis in inflammation and antitumor immunity
.
Mol Cell
2021
;
81
:
4579
90
.
45.
Schweitzer
K
,
Naumann
M
.
CSN-associated USP48 confers stability to nuclear NF-κB/RelA by trimming K48-linked Ub-chains
.
Biochim Biophys Acta
2015
;
1853
:
453
69
.
46.
Hugo
W
,
Zaretsky
JM
,
Sun
L
,
Song
C
,
Moreno
BH
,
Hu-Lieskovan
S
, et al
.
Genomic and transcriptomic features of response to anti–PD-1 therapy in metastatic melanoma
.
Cell
2016
;
165
:
35
44
.
47.
Cortellini
A
,
Tucci
M
,
Adamo
V
,
Stucci
LS
,
Russo
A
,
Tanda
ET
, et al
.
Integrated analysis of concomitant medications and oncological outcomes from PD-1/PD-L1 checkpoint inhibitors in clinical practice
.
J Immunother Cancer
2020
;
8
:
e001361
.