Abstract

   Early diagnosis and disease management based on risk stratification
   have a very positive impact on colon adenocarcinoma (COAD) prognosis.
   It is of positive significance to further explore risk stratification
   of COAD patients and identify predictive molecular biomarkers.
   PANoptosis is defined as a form of inflammatory cell death regulated by
   PANoptosome, with common features of pyroptosis, apoptosis and
   necroptosis. The role of PANoptosis in COAD has not been fully studied.
   In this study, we analyzed significant differences in the expression of
   PANoptosis-related gene (PRG) features in COAD. Subsequently, the
   PANoptosis associated lncRNAs (PALs) associated with PRGs were analyzed
   by LASSO algorithm and multivariate Cox analysis, and PALs related to
   the prognosis of COAD were selected. Based on the expression patterns
   of prognostic PAL features, we performed unsupervised consensus cluster
   analysis to categorize COAD samples into distinct PAL molecular
   subtypes and investigate their associated immune infiltration
   characteristics. We subsequently constructed PAL score model based on
   prognostic characteristics and verified its independent prognostic
   value for COAD. The nomogram diagnostic model was established to
   confirm the prognostic value of PAL scoring system again. Pathway
   enrichment analysis, somatic mutation profiling, and drug sensitivity
   analysis were employed to comprehensively assess the clinical value of
   the PAL score. Additionally, qRT-PCR was used to further validate the
   abnormal expression of the selected targets in COAD. Our results
   provide a new idea for clinical risk stratification and new evidence
   for the role of PANoptosis in COAD.

Supplementary Information

   The online version contains supplementary material available at
   10.1007/s12672-025-01838-3.

   Keywords: Colon adenocarcinoma, PANoptosis, lncRNA, Immune
   infiltration, Prognostic

Introduction

   Despite advances in disease management, colon adenocarcinoma (COAD)
   remains one of the leading causes of death worldwide [[36]1]. Early
   diagnosis and disease management based on risk stratification had a
   very positive impact on COAD prognosis [[37]2, [38]3]. Therefore,
   developing accurate risk assessment models and identifying predictive
   molecular biomarkers is urgently needed for COAD patients [[39]4].

   Long non-coding RNAs (lncRNAs) are involved in a variety of biological
   processes [[40]5]. LncRNAs can interact with RNA, DNA, and proteins to
   form complexes, thereby regulating gene expression through various
   mechanisms, such as modulating transcription, mRNA stability, and
   translation [[41]6, [42]7]. In addition, lncRNAs can also regulate the
   level of chemical reactions by acting as scaffolds or decoys for
   proteins or RNA, or adjust gene expression by influencing the structure
   of chromatin [[43]5, [44]8]. LncRNAs also play a crucial role in the
   occurrence and progression of COAD [[45]9]. In COAD, multiple
   biological functions including cell proliferation, apoptosis, cell
   metastasis and invasion, cell cycle, epithelial-mesenchymal transition
   (EMT), cancer stem cells (CSCs), and drug resistance have been reported
   to be associated with lncRNAs [[46]10].

   PANoptosis, a unique form of programmed cell death (PCD) discovered in
   recent years, is defined as an inflammatory cell death regulated by the
   PANoptosome. It shares key characteristics of pyroptosis, apoptosis,
   and necroptosis [[47]11]. The key components of PANoptosome include
   RIPK1, ASC, RIPK3, CASP6, ZBP1 and CASP1 [[48]12]. PANoptosome assembly
   is mediated by key molecular motifs collectively known as death fold
   domains, and defects in its components have been implicated in a range
   of human diseases, including carcinogenesis process [[49]13]. The
   potential prognostic value of PANoptosis for COAD has been reported
   [[50]14]. PANoptosis is regulated by IRF1 to prevent colorectal cancer
   [[51]15]. However, the role of PANoptosis in COAD is not fully studied.

   In this study, we screened PANoptosis associated lncRNAs (PALs) that
   were associated with the prognosis of COAD. Based on the expression
   patterns of prognostic PAL features, COAD samples were classified into
   distinct PAL molecular subtypes, which were then analyzed for
   variations in immune infiltration characteristics. Then we constructed
   a PAL scoring model based on the prognostic characteristics, and
   verified the prognostic value of the PAL scoring system by establishing
   a nomogram diagnostic model. Pathway enrichment analysis, somatic
   mutation characteristics and drug sensitivity analysis were used to
   evaluate the clinical value of PAL score. qRT-PCR was used to further
   validate the abnormal expression of the selected target in COAD. Our
   results provide a new idea for clinical risk stratification and a new
   basis for the role of PANoptosis in COAD.

Materials and methods

Transcriptome data collection and preprocessing

   In this study, transcriptome data for COAD were obtained from the
   publicly accessible The Cancer Genome Atlas (TCGA) database. Using the
   Perl programming language, we extracted and integrated transcriptome
   files for each sample. Gene annotation for each sample was performed
   using the Ensemble human genome browser GRCh38.p13 annotation file
   within the Perl framework. Concurrently, clinical baseline data for
   COAD samples were extracted from the TCGA database using Perl. After
   excluding samples lacking clinical survival information, a total of 41
   normal samples and 446 COAD samples with complete clinical baseline
   data were collected for subsequent analyses. Copy number variation
   (CNV) frequency files were downloaded from the UCSC Xena database
   ([52]https://xenabrowser.net/datapages/). The STRING database was
   utilized to predict the interaction network among genes.

Identification and prognostic value analysis of lncRNAs associated with
PANoptosis

   Based on previously reported literature, we collected a panel of 14
   PANoptosis-related genes (PRGs) (Supplementary Table 1) [[53]16]. Using
   the Pearson algorithm, we calculated the potential associations between
   these 14 PRGs and lncRNAs, identifying PANoptosis-associated lncRNAs
   (PALs). A filtering threshold was set at |r|> 0.5 and p < 0.05.
   Utilizing the "survival" R package, we performed single-factor Cox
   analysis to evaluate the association of PALs with clinical prognosis in
   COAD and computed hazard ratios (HR) and p-values for each PAL
   variable. Subsequently, employing the "glmnet" R package, we
   constructed a LASSO regression model to further identify key PAL
   variables. Furthermore, using multi-factor Cox analysis, we identified
   PALs with independent prognostic value and calculated the risk scores
   for each variable. The potential associations between PALs and the 14
   PRGs were evaluated using the "ggplot2" R package.

Identification of unsupervised molecular subtypes of PALs

   Based on PAL expression features identified with independent prognostic
   value, we employed the "ConsensusClusterPlus" script to explore
   molecular subtype characteristics of PALs in COAD samples. Using the
   k-means clustering algorithm, COAD samples were classified into
   subgroups ranging from k = 2 to 9, and the optimal model parameters
   were determined to define PAL molecular subgroups. The "ggplot2" script
   was used to visualize an unsupervised Principal Component Analysis
   (PCA) model, examining distribution patterns among different molecular
   subgroups. Clinical survival curves for PAL molecular subgroups were
   plotted using the "survival" script.

Assessment of immune infiltration status features

   Based on the transcriptomic features of samples, we utilized the
   "ESTIMATE" script to calculate immune scores, stromal scores, ESTIMATE
   scores, and tumor purity for each sample. These four metrics were used
   to evaluate the immune infiltration status of the samples.
   Additionally, using the "GSVA" script and known markers for 23 immune
   cell types, we conducted ssGSEA analysis to quantitatively assess the
   infiltration proportions of these 23 immune cell types in each sample.
   Using the Pearson correlation algorithm, we employed the "ggplot2"
   script to compute correlations between independent prognostic variables
   of PALs and the 23 immune cell types.

Construction of PAL scoring system model and independent prognostic analysis

   Based on the independent prognostic features and risk coefficients of
   PAL expression, we calculated PAL scores for each COAD sample using the
   formula: PAL Score = Prognostic Variable Expression Feature * Risk
   Coefficient. The "caret" script was used to divide COAD samples into
   training and validation sets in a 7:3 ratio based on PAL’s independent
   prognostic features. The samples were then stratified into two distinct
   PAL score subgroups using the optimal cutoff value derived from
   clinical survival prognosis. By combining clinical baseline data with
   PAL scores, we conducted univariate and multivariate Cox analyses using
   the "survival" script to evaluate the independent prognostic value of
   various clinical pathological indicators and PAL scores. ROC curves for
   1-year, 3-year, and 5-year outcomes for different PAL score cohorts
   were plotted using the "survivalROC" script to assess their diagnostic
   performance. A nomogram diagnostic model was constructed based on
   clinical pathological variables and PAL score models using the "rms"
   script. The consistency between the nomogram diagnostic model and
   actual survival rates was visualized with the "regplot" script. The
   concordance index between PAL scores and clinical pathological
   variables was assessed using "pec," "rms," "dplyr," and "survival"
   scripts. Finally, decision curves for the nomogram diagnostic model and
   other variables were plotted using the "ggDCA" R script to evaluate the
   precision of each variable.

Analysis of differential gene expression and somatic mutation features in PAL
score subgroups

   Using a differential threshold of |fold change|> 2 and p < 0.05, we
   applied the "limma" script to identify genes differentially expressed
   between PAL score subgroups. Gene Ontology (GO) and Kyoto Encyclopedia
   of Genes and Genomes (KEGG) enrichment analyses were then performed on
   these differentially expressed genes using the "clusterProfiler" script
   to explore potential molecular mechanisms. Somatic mutation data (MAF
   files) for COAD were retrieved from the TCGA database and processed
   with Perl. The "maftools" R script was used to analyze somatic mutation
   features across the PAL score subgroups. Immune therapy response data
   for COAD samples, including PD1 and CTLA4 responses, were sourced from
   the TICA database, and the "limma" script was utilized to evaluate the
   response of PAL subgroups to immune therapies. Additionally, using data
   from the GDSC database, we employed the "pRRophetic" R package to
   predict potential responses to small molecule targeted drugs in
   different PAL score subgroups.

Cell culture

   NCM460 cells (human normal colon epithelial cells) were obtained from
   InCell Corporation (San Antonio, TX, USA). Cells were cultured in
   Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F-12,
   Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with
   10% fetal bovine serum (FBS, Gibco), 100 U/mL penicillin, and 100 μg/mL
   streptomycin (Gibco). Cells were maintained in a humidified incubator
   at 37 °C with 5% CO₂. The culture medium was changed every 2–3 days,
   and cells were subcultured at a ratio of 1:3 when reaching 80–90%
   confluence. HCT116 cells (human colon cancer cell line) were purchased
   from the American Type Culture Collection (ATCC, Manassas, VA, USA).
   Cells were cultured in RPMI 1640 medium (Gibco) supplemented with 10%
   fetal bovine serum (FBS, Gibco), 100 U/mL penicillin, and 100 μg/mL
   streptomycin (Gibco). Cells were maintained in a humidified incubator
   at 37 °C with 5% CO₂. The culture medium was changed every 2–3 days,
   and cells were cultured at a ratio of 1:4 when reaching 80–90%
   confluence.

Quantitative real-time PCR (qRT-PCR) analysis

   qRT-PCR was performed to analyze gene expression levels. Total RNA was
   isolated from the cultured cells using a commercially available RNA
   extraction kit following the manufacturer's instructions. The
   concentration and purity of RNA were determined using a
   spectrophotometer. Subsequently, cDNA synthesis was carried out using a
   reverse transcription kit according to the manufacturer's protocol. The
   qRT-PCR reactions were prepared using a master mix that included cDNA,
   gene-specific primers, and SYBR Green PCR Master Mix. The reactions
   were conducted in triplicate on a real-time PCR machine, following the
   appropriate cycling conditions. GAPDH or β-actin was used as an
   internal control for normalization. Relative gene expression levels
   were calculated using the 2^-ΔΔCt method. Primers pairs are shown in
   Supplementary Table 2. All experiments were conducted in accordance
   with standard laboratory protocols and best practices.

Data analysis

   In this study, all data preprocessing and analyses were conducted using
   R version 4.1.0 and Perl. Statistical differences between two groups
   were assessed using the Wilcoxon rank-sum test. For comparisons
   involving more than two groups, statistical differences were evaluated
   using ANOVA. Pearson correlation analysis was employed to assess
   correlation coefficients between variables. Cellular experimental data
   were statistically analyzed using GraphPad Prism 9.0 software (GraphPad
   Software, San Diego, CA, USA). Experimental data are presented as
   mean ± standard deviation (Mean ± SD). Each experiment was performed
   independently at least three times. A p-value less than 0.05 was
   considered statistically significant, with significance levels denoted
   as *P < 0.05, **P < 0.01, and ***P < 0.001. All p-values were
   two-tailed.

Results

Expression and mutation landscape features of PRGs in COAD

   To elucidate the potential roles of PRGs in COAD, we utilized the
   "limma" script to analyze the differential expression of 14 PRGs
   between normal and tumor tissues. Differential expression analysis
   revealed significant differences in 10 out of the 14 PRGs.
   Specifically, ZBP1, NLRP3, RIPK1, RIPK3, and TNFAIP3 were markedly
   downregulated in tumor tissues, while CASP8, FADD, MAP3K7, RBCK1, and
   PSTPIP2 were significantly overexpressed (Fig. [54]1A). Copy number
   variation (CNV) analysis of the PRG signature indicated significant
   amplifications in ZBP1, PYCARD, NLRP3, and TNFAIP3, whereas RBCK1
   exhibited notable copy number deletions in COAD (Fig. [55]1B). Mutation
   landscape analysis revealed that a substantial proportion of PRGs
   harbored mutations in COAD, with mutation rates of 5% for NLRP3, 4% for
   CASP8, 3% for TNFAIP3, and 3% for RNF31 (Fig. [56]1C). Analysis of
   protein–protein interaction (PPI) networks depicted significant
   interactions among the 14 PRGs (Fig. [57]1D). These findings highlight
   notable expression differences and genomic alterations in the PRG
   signature in COAD, emphasizing their potential roles in the
   carcinogenesis process.

Fig. 1.

   [58]Fig. 1
   [59]Open in a new tab

   Differential Expression Analysis and Mutation Landscape Features of PRG
   Signature. A Differential expression analysis of PRGs between normal
   and tumor tissues in COAD. B Copy number variation analysis of PRG
   signature in COAD. C Mutation landscape features of PRG signature in
   COAD. D Protein–protein interaction (PPI) network diagram illustrating
   potential interactions among the 14 PRG signatures

Identification and prognostic analysis of PANoptosis-associated lncRNA

   Using Pearson correlation analysis, we identified PALs signature
   significantly associated with the PRGs and evaluated their correlation
   with clinical prognosis in COAD. Setting a correlation threshold (r)
   greater than 0.5, we identified and obtained 993 PALs associated with
   PRGs. The Sankey diagram in Fig. [60]2A illustrates the mutual
   associations between the PRG signature and PALs. By integrating PAL
   expression features with clinical survival baseline data in COAD, we
   evaluated the prognostic value of PALs. Univariate Cox analysis
   identified 35 PALs significantly associated with clinical prognosis in
   COAD, including one protective factor and 34 risk factors
   (Fig. [61]2B). Using LASSO algorithm analysis, we further narrowed down
   the prognostic PAL variables and identified 8 feature variables
   (Fig. [62]2C). Subsequently, through multi-factor Cox analysis, we
   evaluated the independent predictive value of these 8 feature variables
   for predicting clinical prognosis in COAD and identified 6 PAL
   variables with independent prognostic value. A correlation heatmap
   (Fig. [63]2D) showed significant associations between these 6
   independent prognostic PALs and the 14 PRGs.

Fig. 2.

   [64]Fig. 2
   [65]Open in a new tab

   Identification of PANoptosis-Related lncRNAs in COAD. A Sankey diagram
   illustrating the mutual associations between PALs and the PRG
   signature. Correlation threshold set at |r|> 0.5. B Univariate Cox
   analysis evaluating the prognostic value of PALs in COAD. C LASSO
   analysis used to select prognostic PAL variables. D Correlation
   analysis showing the relationship between independent prognostic PAL
   variables and the PRG signature

PAL molecular subtype characterization based on prognostic PAL signatures

   Based on the prognostic PAL signature's expression features, we
   performed unsupervised consensus clustering analysis to classify COAD
   samples into distinct PAL molecular subtypes. According to the optimal
   clustering parameters, COAD samples were categorized into three PAL
   subgroups: PAL Subgroup A with 146 samples, PAL Subgroup B with 204
   samples, and PAL Subgroup C with 96 samples (Fig. [66]3A). PCA plots
   demonstrated significant separation among the three PAL molecular
   subgroups, indicating distinct characteristics between them
   (Fig. [67]3B). Clinical prognosis analysis indicated that PAL Subgroup
   B had poorer clinical outcomes compared to Subgroups A and C,
   suggesting that Subgroups A and C may offer potentially better clinical
   survival benefits (Fig. [68]3C). Using ESTIMATE algorithm, we evaluated
   the immune infiltration status characteristics among different PAL
   molecular subgroups. The assessment showed that in PAL Subgroup C,
   associated with the best prognosis, stromal score and ESTIMATE score
   were significantly lower, while tumor purity was markedly higher,
   indicating a lower immune infiltration status potentially associated
   with clinical survival benefits in COAD (Fig. [69]3D–G). Using ssGSEA
   algorithm, we quantitatively analyzed the relative abundance of 23
   immune cell infiltrates within PAL subgroups. Notably, in the PAL
   Subgroup C with the best clinical prognosis, a significant
   downregulation was observed in the proportions of most immune cell
   infiltrates, such as activated B cells, activated CD8 T cells,
   activated dendritic cells, underscoring the close association between
   immune activation status and adverse prognosis in COAD (Fig. [70]3H).
   Visualization heatmap results revealed differential expression patterns
   of the 6 independent prognostic PALs across different PAL molecular
   subgroups and clinical features (F[71]ig. [72]3I). Based on these
   findings, we conclude that the independent prognostic features derived
   from the PAL score accurately classify COAD samples into distinct
   molecular subtypes, with immune activation status potentially playing a
   key role in the adverse prognosis of COAD.

Fig. 3.

   [73]Fig. 3
   [74]Open in a new tab

   Molecular Subtype Features and Immune Infiltration Analysis of PAL in
   COAD. A Consensus clustering analysis of PAL molecular subtypes based
   on prognostic features. B PCA plot analysis of PAL molecular subgroups.
   C Clinical survival outcome analysis of PAL molecular subtypes. D–G
   ESTIMATE algorithm revealing the immune infiltration status of PAL
   molecular subgroups. H Quantitative analysis of 23 immune cell
   infiltrates in PAL molecular subtypes using ssGSEA. I Differential
   expression patterns of PAL prognostic features across different
   clinical features and PAL subgroups

Construction of PAL score models based on prognostic features

   We developed a systematic PAL score model based on prognostic
   expression features of PALs, calculating PAL scores for each COAD
   sample. Using optimal clinical survival cutoffs, we stratified COAD
   samples into two PAL score subgroups across the complete, training, and
   validation sets of the PAL score model (Fig. [75]4A–C). Survival
   analysis across the three independent PAL score models revealed
   significantly better overall survival rates in the low PAL score
   subgroup compared to the high PAL score subgroup, indicating a
   potential association of high PAL scores with adverse prognosis in COAD
   (Fig. [76]4D–F). PCA analysis based on prognostic expression features
   of PAL successfully differentiated PAL score subgroups, emphasizing
   their independence and variability (F[77]ig. [78]4G–I). These findings
   suggest that the PAL score model, based on independent prognostic
   features of PAL, accurately predicts clinical outcomes in COAD,
   facilitating stratification of COAD samples into distinct risk
   subgroups.

Fig. 4.

   [79]Fig. 4
   [80]Open in a new tab

   Establishment and Independence Analysis of PAL Score System Models in
   COAD. A–C Construction and stratification of PAL score subgroups in the
   complete PAL score model, training set PAL score model, and validation
   set PAL score model. D–F Clinical prognostic outcome analysis of COAD
   across three independent PAL score models. G–I Unsupervised PCA
   analysis based on prognostic features of PAL in COAD

Analysis of the independent prognostic value of the PAL score system

   By integrating the PAL score system with clinical-pathological
   features, we were able to further assess the independence and stability
   of PAL scores in predicting clinical outcomes in COAD. Through
   univariate and multivariate Cox regression analyses, incorporating both
   clinical-pathological variables and PAL scores, we calculated hazard
   ratios (HR) to investigate their potential associations with COAD
   prognosis. As depicted in Fig. [81]5A–C, across the three independent
   PAL score cohorts, the PAL score model emerged as a significant risk
   factor associated with adverse prognosis in COAD (HR > 1, P < 0.05),
   indicating that the PAL score system is an independent prognostic
   factor for COAD. Time-dependent ROC curve analysis demonstrated that in
   the complete PAL score model, the area under the curve (AUC) values for
   1-year, 3-year, and 5-year survival were 0.692, 0.731, and 0.723,
   respectively. In the training set PAL score model, the corresponding
   AUCs were 0.701, 0.751, and 0.764, and in the validation set PAL score
   model, they were 0.671, 0.686, and 0.616 (Fig. [82]5D–F).

Fig. 5.

   [83]Fig. 5
   [84]Open in a new tab

   Assessment of the Independent Prognostic Value of PAL Score Models in
   Various Independent Cohorts. A–C Evaluation of the independence
   prediction of clinicopathological features and PAL score models across
   different independent cohorts. D–F Time-dependent ROC curve evaluation
   across three independent cohorts

Development of a nomogram diagnostic model based on the PAL scoring model

   Given the independent prognostic value of the PAL scoring system and
   clinicopathological features in COAD prognosis, we developed a nomogram
   diagnostic model based on various clinical baseline variables and PAL
   scores to assess clinical survival probabilities at 1 year, 3 years,
   and 5 years (Fig. [85]6A). ROC curve analysis revealed that the
   nomogram diagnostic model achieved an AUC of 0.822, significantly
   outperforming standalone PAL scores and other pathological features,
   underscoring its superior predictive capability compared to
   conventional clinical indicators (Fig. [86]6B). Additionally,
   calibration curve analysis showed a good correlation between the
   predicted and observed survival probabilities across various time
   points (Fig. [87]6C). The concordance index curve analysis demonstrated
   that the PAL scoring system had superior prognostic accuracy compared
   to other pathological features (Fig. [88]6D). Decision curve analysis
   revealed that the nomogram diagnostic model offered more accurate
   survival predictions for COAD than both the PAL scores and other
   clinical-pathological features (Fig. [89]6E). These findings suggest
   that the nomogram diagnostic model, incorporating PAL scores and
   clinicopathological features, effectively evaluates survival
   probabilities at different time points in COAD, highlighting its
   substantial potential in prognosticating COAD survival outcomes.

Fig. 6.

   [90]Fig. 6
   [91]Open in a new tab

   Development of the nomogram diagnostic model based on the PAL scoring
   system and clinicopathological features. A Construction of the nomogram
   diagnostic model. B Assessment of diagnostic performance of the
   nomogram diagnostic model, PAL scoring system, and pathological
   features. C Calibration curve analysis. D Concordance index analysis
   between PAL scores and clinicopathological features. E Decision curve
   analysis

Analysis of immune infiltration characteristics in PAL score subgroups

   We further assessed the immune infiltration profiles across PAL score
   subgroups to explore potential mechanisms underlying the clinical
   survival differences. ESTIMATE analysis revealed that, in the high PAL
   score subgroup, immune score, stromal score, and ESTIMATE score were
   significantly higher, while tumor purity was notably lower
   (Fig. [92]7A–D). These findings further suggest a close association
   between heightened immune activation and poor prognosis in COAD.
   Additionally, quantitative results from ssGSEA demonstrated significant
   upregulation of immune cell populations in the high PAL score subgroup,
   including MDSCs, macrophages, regulatory T cells, among others
   (Fig. [93]7E). Using Pearson correlation analysis, we evaluated the
   relationships between PAL prognostic signatures and immune infiltrating
   cells, revealing significant positive correlations of MYOSLID and
   ATP2B1-AS1 with 23 immune infiltrating cells, whereas [94]AL137782.1,
   [95]AC019118.1, FLJ21408, and [96]AC092944.1 showed significant
   negative correlations with immune infiltration status (Fig. [97]7F). In
   summary, based on these results, we elucidated the immune infiltration
   characteristics among PAL score subgroups, providing preliminary
   insights into the association between adverse clinical outcomes in COAD
   and immune activation status.

Fig. 7.

   [98]Fig. 7
   [99]Open in a new tab

   Association Analysis of PAL Score Subgroups with Immune Infiltration
   Status. A–D Assessment of immune infiltration status based on ESTIMATE
   algorithm. E Quantitative analysis of 23 immune infiltrating cells
   using ssGSEA algorithm. F Correlation analysis between PAL prognostic
   signatures and immune infiltration features

Characterization of potential regulatory mechanisms and somatic mutations in
PAL scoring subgroups

   We assessed gene expression differences between PAL score subgroups
   using a differential threshold of |FC|> 2 and p < 0.05, exploring
   potential regulatory mechanisms (Fig. [100]8A). GO analysis of
   differentially expressed genes revealed biological functions related to
   regulation of cellular response to growth factor stimulus,
   collagen-containing extracellular matrix, and extracellular matrix
   structural constituent (Fig. [101]8B). KEGG enrichment analysis
   identified several signaling pathways, including cytoskeleton
   regulation in muscle cells, malaria, dilated cardiomyopathy, and
   vascular smooth muscle contraction, as potential key regulatory
   mechanisms in COAD (Fig. [102]8C). Somatic mutation analysis revealed
   significantly higher mutation frequencies in many somatic genes in the
   high PAL score subgroup, such as APC, TP53, and TTN (Fig. [103]8D, E).
   IPS immune response evaluation suggested that the low PAL score
   subgroup showed better responses to CTLA4 or PD1 immune therapies
   (Fig. [104]8F). Furthermore, our predictions identified potential
   beneficial anti-tumor drugs for PAL score subgroups; IC50 results
   indicated significantly lower scores for Bortezomib, CP466722,
   Crizotinib, and MG-132 in the high PAL score subgroup, suggesting
   potential sensitivity to these treatments (Fig. [105]8G). In summary,
   our findings provide preliminary insights into the potential regulatory
   mechanisms among PAL score subgroups, predict responses to immune and
   targeted therapies, and offer new perspectives for future personalized
   precision medicine approaches.

Fig. 8.

   [106]Fig. 8
   [107]Open in a new tab

   Analysis of Potential Regulatory Mechanisms and Immune Therapy Response
   in PAL Score Subgroups. A Differential expression analysis of genes
   between PAL score subgroups. Selection criteria: |fold change|> 2,
   p < 0.05. Red indicates upregulation, blue indicates downregulation. B,
   C GO and KEGG enrichment analysis of differentially expressed genes. D,
   E Analysis of somatic mutation features in PAL score subgroups. F
   Prediction of immune therapy response in PAL score subgroups. G
   Prediction of small molecule targeted drugs potentially responsive in
   PAL score subgroups

In vitro qRT-PCR validation of prognostic PAL signatures

   We further explored the expression profiles of PAL prognostic features
   in vitro cell experiments. As depicted in Fig. [108]9A–F, PCR results
   indicate significantly decreased expression of [109]AL137782.1 and
   markedly increased expression of MYOSLID, [110]AC019118.1, ATP2B1-AS1,
   FLJ21408, and [111]AC092944.1 in the HCT116 cell line.

Fig. 9.

   [112]Fig. 9
   [113]Open in a new tab

   Expression analysis of PAL prognostic features in NCM460 and HCT116
   cell lines. A mRNA expression analysis of [114]AL137782.1, B FLJ21408,
   C ATP2B1-AS1, D MYOSLID, E [115]AC019118.1, and F [116]AC092944.1. Data
   are presented as mean ± SD (n = 3), *p < 0.05, **p < 0.01, ***p < 0.001

Discussion

   In this study, we established a prognostic model and verified its
   reliability by analyzing PAL, which is associated with COAD prognosis.
   Our results provide new evidence for the role of PANoptosis in COAD. In
   fact, it has been reported that the key protein of PANoptosis ASC is a
   pro-apoptotic gene responsible for activating the procaspase-1 pathway
   in COAD [[117]17]. In addition, COAD prognosis is influenced by the
   methylation of ASC [[118]18]. The effect of ASC on p53-mediated
   chemical sensitivity has also been reported in patients with COAD
   [[119]19]. Together with our results, these evidences support the role
   of PANoptosis in COAD tumorigenesis.

   Given the large number of lncRNAs and the limited research available,
   only two of the six lncRNAs we identified have been reported to
   potentially influence the mechanisms of COAD. ATP2B1-AS1 has been found
   to regulate the NF-kappaB signaling pathway [[120]20, [121]21].
   Combined with the effect of NF-kappaB signaling pathway on COAD growth,
   angiogenesis, and tumor metastasis, ATP2B1-AS1 may at least partially
   affect COAD progression through this pathway [[122]22–[123]24].
   MYOSLID, as a competitive endogenous RNA, regulates the expression of
   MCL-1 [[124]25]. MCL-1 is associated with drug resistance, growth, and
   apoptosis in colon cancer [[125]26–[126]28]. In addition, MYOSLID
   regulates the epithelial-mesenchymal transition (EMT) sequence of
   tumors to promote invasion and metastasis [[127]29]. Therefore, there
   is a theoretical basis for MYOSLID to influence COAD.

   Our pathway enrichment experiments showed that Cytoskeleton-related
   pathway enrichment was the most significant. Cytoskeletal remodeling is
   one of the important preconditions for tumor cell metastasis [[128]30].
   LncRNAs play a significant role in the remodeling of the cytoskeleton
   in cancer. They can affect cellular functions by directly interacting
   with actin and its associated proteins, as well as by modulating
   cytoskeleton-related regulatory pathways, including Rho GTPase and its
   downstream effector proteins [[129]31]. Multiple lncRNAs participate in
   malignant behavior of tumors through cytoskeletal remodeling [[130]32].
   The cytoskeletal network in colon cancer is also involved in regulating
   cell polarity, differentiation, proliferation, migration, and invasion
   [[131]33]. The causality and mechanisms underlying the prognostic
   differences among various PAL groups and their association with
   cytoskeleton-related pathways warrant further investigation.

   In the development of colorectal cancer, a sustained and sustained
   inflammatory response is essential. Long-term accumulation of colon
   inflammation significantly increases the risk of colorectal cancer
   [[132]34]. Myeloid Derived Suppressor Cells (MDSCs) play a key role in
   the progression of colorectal cancer associated with colitis [[133]35].
   MDSCs contribute to tumor progression by enhancing angiogenesis,
   promoting chronic inflammation, and creating a tumor microenvironment
   that suppresses immune system activity [[134]36]. MDSC may also promote
   colitis-associated colorectal cancer progression by promoting the
   stemness of colon cancer cells [[135]37]. Reducing MDSC can effectively
   slow tumor progression and produce anti-tumor effects [[136]38]. In
   addition, the reduction of MDSCs content contributes to the
   effectiveness of immunotherapy [[137]39]. Our results also showed
   higher levels of MDSCs in the TME of COAD patients with poor prognosis.
   As a result, several regimens targeting MDSCs are currently underway in
   COAD patients [[138]39, [139]40].

   In conclusion, we successfully developed the PAL scoring model and
   preliminarily validated its effectiveness. It is important to note that
   most of the results in this study were derived from public databases,
   which may introduce geographical and ethnic biases. Moreover, the
   bioinformatics analysis primarily involves correlation, with a lack of
   causal discussion. Therefore, further multi-center, large-scale
   studies, along with in vitro experiments, are needed to explore the
   molecular mechanisms more thoroughly. These additional investigations
   will help confirm the conclusions of this study and enhance our
   understanding of the role of PAL in COAD.

Supplementary Information

   [140]Supplementary Material 1.^ (9.2KB, xlsx)
   [141]Supplementary Material 2.^ (9KB, xlsx)

Author contributions

   YW conceived and designed the study. SZ contributed the data collection
   and data analysis. BZ conceived the original ideas and composed this
   manuscript. SD and LS designed and completed the experimental section.
   MX contributed the table and figures of this manuscript. YW and SZ
   contributed equally to this manuscript. All authors contributed to the
   article and approved the submitted version.

Data availability

   The data used to support the fndings of this study are available from
   the corresponding author upon request.

Declarations

Competing interests

   The authors declare no competing interests.

Footnotes

   Publisher's Note

   Springer Nature remains neutral with regard to jurisdictional claims in
   published maps and institutional affiliations.

   Yangyang Wang and Shihui Zhao contributed equally to this work.

References