Introduction

Cancer is defined as a group of diseases characterized by abnormal cell growth associated with pathological manifestations as well as significant morbidity and mortality globally (Choi et al. 2010; Bray et al. 2018). As cancers exhibit dysregulated cell death and inflammatory responses (Hanahan and Weinberg 2000, 2011; Green and Evan 2002), many current therapeutic approaches focus on inducing cancer cell death preferentially (Bernier et al. 2004; Carneiro and El-Deiry 2020). However, evasion of cell death, especially apoptosis, is one of the major mechanisms of primary and adaptive therapeutic resistance in tumors, which in turn leads to poor therapeutic efficacy (Hanahan and Weinberg 2000, 2011; Letai 2008). PANoptosis, a unique inflammatory programmed cell death (PCD) pathway activated by specific triggers and regulated by the PANoptosome, possesses key features of apoptosis, pyroptosis, and necroptosis, but cannot be accounted for by any of the three PCD pathways alone (Nguyen and Kanneganti 2021; Place et al. 2021; Wang and Kanneganti 2021; Gullett et al. 2022; Liu et al. 2022). PANoptosome acts as a molecular scaffold for the contemporaneous engagement of key apoptotic, pyroptotic, and necroptotic machinery. It is a multifaceted multiprotein complex whose main components include proteins such as RIPK1, RIPK3, NLRP3, CASP1, CASP6, CASP8, FADD, and PYCARD (Samir et al. 2020; Zheng and Kanneganti 2020; Zheng et al. 2020; Christgen et al. 2020, 2021; Jiang et al. 2021a, b; Nguyen and Kanneganti 2021; Place et al. 2021; Wang and Kanneganti 2021; Gullett et al. 2022; Liu et al. 2022). Some triggers (including influenza A virus, vesicular stomatitis virus, Listeria monocytogenes, and Salmonella enterica serovar Typhimurium) initiate the assembly of PANoptosome by activating specific sensors, such as ZBP1, and then promote the activation of downstream PCD executioners, including CASP3/CASP7-mediated apoptosis, GSDMD/GSDME-mediated pyroptosis, and RIPK1/MLKL-mediated necroptosis (Malireddi et al. 2019, 2020; Christgen et al. 2020, 2021; Zheng et al. 2020; Jiang et al. 2021a, b; Place et al. 2021; Wang and Kanneganti 2021; Gullett et al. 2022).

Recently, Karki et al. found that the synergism of tumor necrosis factor α (TNF-α) and interferon γ (IFN-γ) triggers caspase-8/FADD-mediated PANoptosis by activating the JAK/STAT1/IRF1 axis in SARS-CoV-2 Infection (Karki et al. 2021a). Subsequently, they identified that ZBP1-mediated PANoptosis disrupts interferon therapeutic efficacy during coronavirus infection, including SARS-CoV-2 and mouse hepatitis virus (MHV) (Karki et al. 2022). Most previous studies on PANoptosis have focused on infectious diseases; however, emerging studies have shown that it is closely related to cancers. For instance, Karki et al. in 2020 identified that interferon regulatory factor 1 (IRF1) as an upstream regulator of PANoptosis to restrict tumorigenesis in colitis-associated colorectal cancer (Karki et al. 2020). In 2021, adenosine deaminase acting on RNA 1 (ADAR1) could restrict ZBP1-mediated PANoptosis activated by combining interferons (IFNs) and nuclear export inhibitors (such as leptomycin B and Selinexor) to promote tumorigenesis (Karki et al. 2021b). Subsequently, Malireddi et al. found that IFN-γ and TNF-α together could trigger PANoptosis of tumor cells and suppress tumor growth in colon and lung cancers, melanoma, and leukemia (Malireddi et al. 2021). In addition, Lin et al. in 2022 demonstrated that phosphorylated cysteine desulfurase (NFS1) could prevent PANoptosis from weakening oxaliplatin-based chemosensitivity in colorectal cancer (Lin et al. 2022). However, the complex correlation between PANoptosis and molecular characteristics, clinical features, and treatment strategies in cancers warrants further exploration.

In this study, we focused on PANoptosome core components and comprehensively explored their genomic, epigenomic, and transcriptomic characteristics across 33 cancers. We found that the expression of PANoptosome components was significantly associated with distinct genomic and epigenetic events. Subsequently, we developed a scoring system, the PANoptosome-related potential index (PANo-RPI), to investigate the potential association between PANopotosis and immune signatures. PANo-RPI was highly correlated with many immune response-related pathways and the tumor infiltration of immune cells. Furthermore, the association between PANo-RPI and immunotherapy response was elucidated by reanalyzing multiple immune checkpoint inhibitor (ICI) therapy cohorts, demonstrating that high PANo-RPI was correlated with better immunotherapy response and efficacy. Based on these findings, we hypothesized that small-molecule drugs could potentially activate the assembly of the PANoptosome and induce PANoptosis. This integrated analysis provides a rich resource for understanding PANoptosome biology and preliminarily unveils the potential application of the PANoptosome as a predictive biomarker for immunotherapy response.

Materials and Methods

Data Preparation and Processing

The uniformly normalized pan-cancer dataset, which integrates the Cancer Genome Atlas (TCGA, http://cancergenome.nih.gov/) and Genotype-Tissue Expression (GTEx, http://commonfund.nih.gov/GTEx/) databases, and the corresponding clinical data were downloaded from the UCSC Xena (https://xenabrowser.net/datapages/). The gene expression matrices and clinical data of the immunotherapy cohorts (PRJEB23709 (Gide et al. 2019), GSE91061 (Riaz et al. 2017), GSE100797 (Lauss et al. 2017), and GSE35640 (Ulloa-Montoya et al. 2013)) were downloaded from the BioProject and GEO datasets at NCBI (https://www.ncbi.nlm.nih.gov/).

Bioinformatics Analysis and Online Analysis Platforms

Single nucleotide variation (SNV), copy number variation (CNV), and methylation analyses were conducted using the Gene Set Cancer Analysis ( http://bioinfo.life.hust.edu.cn/GSCA/#/) platform (Liu et al. 2022), we further explored the associations of PANo-RPI with tumor immune cells and stromal cell infiltration using the ESTIMATE method. As shown in Fig. 6B and Table S5, PANo-RPI was positively correlated with the immune score (in 30 out of 33 cancers, Fig. S6A) and stromal score (in 24 out of 33 cancers, Fig. S6B) but negatively correlated with tumor purity (in 29 out of 33 cancers, Fig. S6C). Additionally, we found that PANo-RPI was positively correlated with most chemokines (Fig. S7A) and chemokine receptors (Fig. S7B). These results indicate that the activation of PANoptosome assembly may promote the infiltration of non-tumor cells, especially immune cells. To further infer the recruitment of certain tumor-infiltrating immune cells, the proportions of the immune cell subsets were calculated using ImmuCellAI. Subsequently, we performed Pearson’s correlation analysis and found that PANo-RPI was significantly associated with many immune cell subsets (Fig. 6C and Table S6). For lymphoid lineage cells, PANo-RPI was positively correlated with the infiltration of CD8+T cells, cytotoxic T cells (Tc), exhausted T cells (Tex), helper T-cell 1 (Th1), helper T-cell 2 (Th2), follicular helper T-cell (Tfh), regulatory T cells (Tregs), NK cells, and NKT cells, but negatively correlated with that of naïve CD8+T cells, helper T-cell 17 (Th17), naïve CD4+T cells, and B cells; for myeloid lineage cells, PANo-RPI had a significantly positive correlation with the infiltration of macrophages and dendritic cells (DCs), whereas it was negatively correlated with neutrophils (Fig. 6C). Tumor mutation burden (TMB) and microsatellite instability (MSI) are essential biomarkers in tumor microenvironment (TME), and are considered valid response indicators of immunotherapies in many kinds of cancers (Luchini et al. 2019). Therefore, the association of TMB/MSI with PANo-RPI was evaluated. The radar charts (Fig. 6D, E) showed that PANo-RPI was positively correlated with TMB scores in THCA but negatively correlated with CHOL, READ, PCPG, TGCT, and LUAD (Table S7). PANo-RPI was positively correlated with MSI scores of PRAD and THCA and negatively correlated with those of UCS, ACC, TGCT, PAAD, LUAD, CESC, and STAD (Table S8). The aforementioned results indicate that the PANoptosome, in addition to being related to cell death, is highly likely to be involved in mediating immune signaling regulation and anti-cancer immune responses.

Fig. 6
figure 6

The correlation between PANo-RPI and cancer immunity. A Correlation between PANo-RPI and the enrichment scores of gene sets developed by Mariathasan et al. B Correlation between PANo-RPI and immune scores, stromal scores, and tumor purity. C Correlation between PANo-RPI and the proportions of the immune cell subsets. D, E Correlation between PANo-RPI and D TMB and E MSI

Full size image

The Potential of PANo-RPI in Predicting Immunotherapy Response and Efficacy

Considering the distinct association between PANo-RPI and anti-cancer immunity, we assessed whether PANo-RPI was associated with the response and efficacy of cancer immunotherapies. Four immunotherapy datasets (PRJEB23709, GSE91061, GSE100797, and GSE35640) were used in this study. We evaluated the PANo-RPI of non-responders and responders and found that the PANo-RPI of responders was higher than that of non-responders in all four immunotherapy cohorts (P < 0.05) (Fig. 7A–D). Subsequently, we separated patients into high and low PANo-RPI groups using the corresponding median level and counted the occupation ratios of immunotherapeutic responses (complete response, CR; partial response, PR, stable disease, SD; and progressive disease, PD) in the two groups. The stacked bar plots show that the objective tumor response (CR+PR) was significantly improved in the high PANo-RPI group (Fig. 7E–H). Additionally, survival analysis indicated that after immunotherapies, patients with high PANo-RPI had a significantly longer survival time than those with low PANo-RPI in GSE91061 (Log-rank P = 0.003) and GSE100797 (Log-rank P = 0.004) (Fig. 7I, J). Although not statistically significant, the PRJEB23709 cohort showed a similar trend (Fig. 7K). Furthermore, we calculated the aggregated scores of antigen processing (MHC), effector cells (EC), suppressor cells (SC), and CP (checkpoints) using Immunophenogram (Charoentong et al. 2017). The results showed that PANo-RPI was positively correlated with MHC and EC but negatively correlated with SC and CP (Fig. 7L). Additionally, the immunophenoscore (IPS) is considered a superior predictor of the ICI response (Charoentong et al. 2017). It was significantly correlated with PANo-RPI in three out of the four immunotherapy cohorts in our study (Fig. 7L). These results imply that PANo-RPI has remarkable potential for predicting immunotherapy response and efficacy in many cancers. Since the number and functions of CD8+T cells are major factors associated with immunotherapeutic activity, we assessed the association between PANo-RPI and CD8+T-cell infiltration. As expected, correlation analysis demonstrated that PANo-RPI had a significant positive correlation with the CD8+T-cell proportion in PRJEB23709 (Spearman: r = 0.389, P < 0.001), GSE91061 (Spearman: r = 0.695, P < 0.001), GSE100797 (Pearson: r = 0.545, P = 0.005), and GSE35640 (Spearman: r = 0.577, P < 0.001) (Fig. 7M–P). Taken together, the aforementioned analyses revealed that the PANoptosome may potentiate the immune response and efficacy by remodeling the tumor immune microenvironment (TIME), such as recruiting immune cells and maintaining the differentiation and functions of CD8+T cells.

Fig. 7
figure 7

Potential of PANo-RPI to predict immunotherapy response. A–D Comparison of PANo-RPI levels between responders (R) and non-responders (NR) in immunotherapy cohorts from A PRJEB23709, B GSE91061, C GSE100797 and D GSE35640 datasets. Rate of clinical response (complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD)) to immunotherapies in high or low PANo-RPI groups in the E PRJEB23709, F GSE91061, G GSE100797 and H GSE35640 cohorts. I–K Kaplan–Meier curves for low and high PANo-RPI groups in I GSE91061, J GSE100797 and K GSE35640 cohorts. L Correlation heatmap between PANo-RPI and MHC, EC, CP, SC, and IPS scores in the four immunotherapy cohorts. M–P Correlation scatterplots between CD8+T-cell infiltrations and PANo-RPI in M PRJEB23709, N GSE91061, O GSE100797, and P GSE35640 cohorts

Full size image

Potential Small-Molecule Drugs Based on PANoptosome Components

Strategies that can activate PANoptosome assembly and induce tumor cell apoptosis warrant further in-depth exploration. In this study, we focused on mining small-molecule drugs that may directly target the PANoptosome. We analyzed the correlation between the mRNA expression of each PANoptosome gene and the half-maximal inhibitory concentration (IC50) of each small-molecule compound in two drug-related databases (GDCS and CTRP). Figure 8A, B summarize the correlation between gene expression and sensitivity to GDSC and CTRP drugs (top 30) in pan-cancer analysis. These results showed that the IC50 values of most of the top 30 drugs were significantly negatively correlated with the expression of RIPK3, CASP8, NLRP3, PYCARD, and CASP1, which implied that these small-molecule drugs might be potential activators of the PANoptosome. Notably, four small-molecule drugs, PIK-93, SNX-2112, AZD7762, and selumetinib, showed a significant correlation with the PANoptosome components in both the GDSC and CTRP databases (Fig. 8A, B). Among these, SNX-2112 and AZD7762 induce apoptosis in several tumor cells (Landau et al. 2012; Cheng et al. 2021; Ozgiray et al. 2022). To further evaluate the binding affinities of PIK-93, SNX-2112, AZD7762, and selumetinib for each PANoptosome component, molecular docking analysis was performed. The binding poses of the four candidate drugs and their corresponding interactions with protein residues of NLRP3, RIPK3, RIPK1, CASP6, CASP8, and CASP1, as well as the corresponding binding energy for each interaction, were obtained using Autodock Vina v.1.2.2 (Figs. 9, S8 and Table S8). We found that each candidate primarily bonded to these protein residues via hydrogen bonds and electrostatic interactions. In addition, the majority of these binding energies were less than − 5 kcal/mol, indicating stable binding between the four small-molecule drugs and the PANoptosome components. Among these, SNX-2112 and AZD7762 showed significantly stable binding with NLRP3 (Fig. 9A and G) and RIPK3 (Fig. 9B and H), with corresponding binding energies of less than − 8 kcal/mol. These findings provide insights into the mechanism of PANoptosome activation and its value in clinical applications.

Fig. 8
figure 8

Potential small-molecule drugs based on PANoptosome components. A, B Correlation heatmaps depicting the relationships between the expression PANoptosis-related genes and the sensitivity of small-molecule drugs in GDSC and CTRP databases

Full size image
Fig. 9
figure 9

Molecular docking of small-molecule drugs and PANoptosome Components. A–F Molecular docking simulations of SNX-2112 with A NLRP3, B RIPK3, C RIPK1, D CASP6, E CASP8 and F CASP1. G–L Molecular docking simulations of AZD7762 with G NLRP3, H RIPK3, I RIPK1, J CASP6, K CASP8 and L CASP1. The left cartoons depict the crystal structures of small-molecule compounds and their respective targets. The three-dimensional structures on the right illustrate the binding pockets. Hydrogen bonds are represented by blue dashed lines, and Pi stacking is indicated by orange or green bonds

Full size image

Discussion

In this study, we scrutinized the genetic, epigenetic, and transcriptional profiles of eight core PANoptosome component genes across 33 different cancers. A pronounced divergence was noted in the molecular characteristics of PANoptosome component genes between tumors and normal tissues. Certain PANoptosome genes exhibited a heightened SNV frequency and showed significant correlations with CNV and methylation. Furthermore, these genetic mutations and epigenetic modifications had a discernible impact on the expression of PANoptosome genes. This underscores the pivotal role of genetic alterations in driving cancer progression. Consequently, these findings not only lay the groundwork for an in-depth exploration of PANoptosome-related mechanisms but also enhance our comprehension of the intricacies associated with the participation of PANoptosome in tumorigenesis and development.

Building on the expression profiles of PANoptosome component genes, we devised PANo-RPI for evaluating the intrinsic PANoptosome assembly potential in cancers. Our observations revealed that the majority of cancers exhibited elevated PANo-RPI levels, particularly in MESO, LAML, DLBC, and THYM. Moreover, numerous cancers, including BRCA, CHOL, ESCA, HNSC, KIRC, KIRP, STAD, THCA, BLCA, and CESC, displayed higher PANo-RPI levels than their non-tumor counterparts. In parallel, we established a correlation between high PANo-RPI and improved prognosis in patients with ACC, BRCA, KICH, KIRP, MESO, SARC, READ, STAD, THCA, SKCM, and UVM. These findings suggest that, in comparison to normal cells, tumor cells exhibit increased intrinsic PANoptosis, while extrinsic PANoptosis is more likely triggered by external stimuli such as clinical treatments. However, a high PANo-RPI was associated with an unfavorable outcome in patients with HNSC, GBM, KIRC, LAML, LUAD, LGG, LUSC, OV, PAAD, and TGCT. The rapid and uncontrolled proliferation of cancer cells, coupled with their tendency to evade cell death, may offer an explanation for these contradictory results (Koo et al. 2015; Reina-Campos et al. 2018).

Targeting genes or pathways associated with the inflammatory response and resha** the tumor microenvironment (TME) from a “cold” to a “hot” state holds promise as a strategy to enhance the effectiveness of immunotherapy (Duan et al. 2020; Wei et al. 2020). As an immunogenic cell death pathway, PANoptosis plays a crucial role in anti-tumor immunity and tumor suppression (Liu et al. 2022). However, the mechanisms facilitating tumor cell PANoptosis remain poorly understood. In this study, we identified some small-molecule drugs that may induce PANoptosis, such as SNX-2112 and AZD7762. These two drugs have previously been reported to induce apoptosis (Okawa et al. 2009; Wang et al. 2018; Hu et al. 2019). Consequently, SNX-2112 and AZD7762 will be employed in follow-up experiments to further explore their association with PANoptosis and assess their clinical applications in cancer treatments.

Certain types of precision medicine approaches have the potential to directly induce PANoptosis, effectively eliminating tumor cells through cytotoxic effects while bolstering the anti-tumor immune response. However, the application of targeted PANoptosis therapy, especially when combined with immunotherapy strategies like immune checkpoint inhibitors, may induce acute inflammation in any organ system, with pneumonitis representing one of the most severe immune-related adverse events (irAEs) (Spagnolo et al. 2022). In addition to short-term adverse events, there are lingering concerns regarding long-term adverse events resulting from therapies, such as idiopathic pulmonary fibrosis triggered by changes in immunologic and environmental factors (Karampitsakos et al. 2023a, b). Therefore, the simultaneous goal of maximizing the effectiveness of PANoptosis-based approaches while minimizing adverse events is a crucial consideration for the future development of new cancer treatment strategies.