A P-body-related risk score predicts prognosis and immune microenvironment in lung adenocarcinoma

Introduction

P-bodies, or processing bodies, are dynamic cytoplasmic granules that play a crucial role in the post-transcriptional regulation of gene expression (1). These granules are involved in mRNA decay, storage, and translational repression (2). P-bodies are composed of various enzymes and proteins that interact with mRNAs, including decapping enzymes, exonucleases, and components of the RNA-induced silencing complex (RISC) (3,4). Their primary function is to regulate mRNA turnover and translation, thereby maintaining cellular homeostasis and enabling adaptive responses to stress conditions. By sequestering mRNAs, P-bodies inhibit their translation and promote their degradation when required. This regulatory mechanism is essential for cells to rapidly adjust to environmental changes and stress, ensuring the proper balance of protein synthesis and degradation (5,6).

Recent studies have further elucidated the critical role of P-bodies in various biological processes, including stress response (7), viral infection (8), and development (9,10). In the context of cancer, P-bodies have been identified as pivotal players in tumorigenesis and cancer progression (11-14). The dysregulation of mRNA metabolism mediated by P-bodies can result in abnormal gene expression profiles, driving the malignant transformation and proliferation of cancer cells (11). By regulating the stability and translation of specific mRNAs, P-bodies influence various oncogenic pathways. For example, the disruption of P-body function has been linked to the stabilization of oncogenic mRNAs and the downregulation of tumor suppressor mRNAs, thereby promoting cancer cell survival and growth (11). P-bodies are also thought to be involved in the regulation of key mRNAs that control the cell cycle, migration, and tumor growth (5,15,16). Abnormalities in P-body components or their regulatory functions are closely associated with increased tumor aggressiveness and poor patient prognosis (17-19). Additionally, studies have shown that the number and composition of P-bodies can change in response to oncogenic stress, indicating a direct role in tumor biology (11,20,21). Although the precise mechanisms by which P-bodies contribute to lung cancer pathogenesis are still under investigation, they likely involve the intricate control of mRNA stability and translation under oncogenic signaling conditions (16,22). These findings underscore the potential of targeting P-bodies or their associated pathways as a therapeutic strategy in lung cancer treatment.

Despite their recognized importance of P-bodies in cancer biology, comprehensive studies on their roles in lung cancer remain scarce. Most existing research has focused on individual P-body components or specific mRNAs, leaving a gap in understanding their broader impact on lung cancer pathogenesis. In this study, we systematically evaluated 24 core P-body-associated genes in lung adenocarcinoma (LUAD), analyzing their genomic alterations, expression patterns, and association with clinical outcomes. Using least absolute shrinkage and selection operator (LASSO)-Cox regression, we developed a “P-body-gene expression related risk score” (hereafter referred to as the P-body-related risk score), which is calculated based on the weighted expression of 5 genes associated with core P-body functions. This score serves as a transcriptomic-level proxy to reflect the transcriptional activity of P-body-related genes in tumor cells. By exploring the relationships between this risk score and tumor biological processes, including RNA processing, genomic instability, and immune microenvironment characteristics, we aimed to explore potential explanation for the prognostic role of P-body-associated gene expression in LUAD and to generate hypotheses for future functional studies. We present this article in accordance with the TRIPOD reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2736/rc).

Methods

Acquisition and processing of data

Data from The Cancer Genome Atlas (TCGA), accessible via the portal at https://portal.gdc.cancer.gov/repository, were obtained to gather genomic, transcriptomic, and proteomic data of primary tumors, along with clinicopathological data for 513 patients. Additionally, RNA expression data and clinical details for the datasets GSE42127, GSE31210, GSE37745, GSE41271, GSE13213, were retrieved from the Gene Expression Omnibus (GEO) database, which can be found at https://www.ncbi.nlm.nih.gov/geo. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Development of a risk model to analyze the P-body-related risk score

A total of 24 core P-body-related genes were curated based on P-body databases (P-body database, RNAgranuleDB, http://rnagranuledb.lunenfeld.ca) (23), including all well-characterized structural components, regulatory factors, and functional effectors of P-bodies (e.g., AGO2, DDX6, DCP1A, LSM14A). Univariable Cox proportional hazards regression analysis (uniCOX) was performed to identify prognostic genes [overall survival (OS)] within the 24 core P-body-related genes. Genes exhibiting a uniCOX P value less than 0.05 were selected for further analysis (8 genes retained), serving as a dimensionality reduction pre-screening step to reduce computational noise.

To eliminate multicollinearity among these 8 candidate genes and construct a robust prognostic model, LASSO-Cox regression analysis with ten-fold cross-validation was performed using the glmnet package in R. The optimal regularization parameter (λ.min) was selected to minimize the cross-validation error, and genes with non-zero regression coefficients were retained for the final model (5 genes retained: MOV10, PCBP1, PCBP2, YBX1, YWHAG). The P-body-related risk score was constructed using the following formula based on the regression coefficients from LASSO-Cox analysis: P-body-related risk score = (0.03311412) × expMOV10 + (0.02487767) × expPCBP1 + (0.20140845) × expPCBP2 + (0.06831339) × expYBX1 + (0.33630468) × expYWHAG, where “exp” refers to the normalized expression levels of the corresponding genes. The LUAD samples were categorized into high and low P-body-related risk score groups according to the 0.5 quantile cutoff of the score in the TCGA cohort.

Copy number variation (CNV) and methylation profile of P-body related genes in LUAD based on GSCA

Gene set cancer analysis (GSCA) platform is a web server that integrates multiomics data based on the TCGA database (http://bioinfo.life.hust.edu.cn/web/GSCA/) (GSCALite: a web server for GSCA). CNV, methylation, are among the studies offered by GSCA. Correlation between mRNA expression quantity of different genes and CNV in LUAD was analyzed by GSCA website. Also, the correlation between the amount of mRNA expression and the degree of methylation of genes in LUAD was analyzed by GSCA website.

Gene set variation analysis (GSVA)

To explore the biological pathways associated with the P-body-related risk score, we performed GSVA using the GSVA package in R (version 4.2.2). The analysis was applied to the transcriptomic data of the TCGA LUAD cohort. We calculated enrichment scores for 50 hallmark gene sets obtained from the Molecular Signatures Database (MSigDB, https://www.gsea-msigdb.org/gsea/msigdb). These hallmark gene sets represent well-defined biological processes and pathways.

Immune infiltration analysis

Single-sample gene set enrichment analysis (ssGSEA) algorithm was used to calculate and compare the scores for 28 types of immune cells. To enhance the rigor of immune infiltration analysis, xCell was further used for independent validation of the key immune cell infiltration results (e.g., activated CD8⁺ T cells, dendritic cells).

Single-gene mutation analysis of the two groups

Following the acquisition and processing of the LUAD mutation data from TCGA, we conducted a comparative analysis of the tumor mutational burden (TMB), tumor neoantigen burden (TNB), CNV fraction, loss of heterozygosity (LOH), ploidy score and homologous recombination deficiency (HRD) score between the high and low score groups.

Statistical analysis

The statistical analyses were conducted employing R software, version 4.2.2. Count data are delineated in terms of numerical count and corresponding rate (percentage), with inter-group comparisons executed utilizing the Fisher’s exact test. Wilcox test was applied for continuous data. The Kaplan-Meier (KM) method coupled with log-rank tests, was implemented to evaluate survival rates among patients stratified into high- and low-P-body-related risk score groups. Multivariable Cox proportional hazards regression analysis was used to assess the independent prognostic value of the risk score after adjusting for clinical factors (tumor stage, gender, age); 1,000-time bootstrap resampling was performed to validate the model stability, and the optimism-corrected C-index was calculated to evaluate prognostic accuracy. For all fold change calculations (e.g., gene expression, pathway activity, protein levels), the value was computed as low group/high group. Statistical significance was established at a P value less than 0.05 for all analyses conducted, and a P value between 0.05 and 0.1 was considered a marginal statistical trend.

Results

Genomic landscape of P-body gene set in LUAD

To establish the biological relevance of P-body-related genes in LUAD, we first characterized their genomic alterations and expression patterns in the TCGA LUAD cohort. Analysis of 24 core P-body-related genes revealed frequent genetic alterations, including point mutations, structural variants, and CNVs (Figure 1A). The top five mutated genes were AGO2 (9%), FMR1 (4%), LSM14B (4%), XRN1 (4%), and HTT (4%), with amplifications being the predominant alteration type. Based on these alterations, we stratified patients into a “wild-type” group (no alterations in any of the 24 genes) and an “altered” group (at least one alteration). Survival analysis showed that the presence of these alterations was significantly associated with disease-free survival (DFS), but not with OS or progression-free survival (PFS), suggesting their potential utility as prognostic biomarkers (Figure 1B).

Figure 1 Genomic alterations of P-body genes and survival analysis in TCGA cohort. (A) Oncoprint showing the genetic alteration profiles of 24 P-body-related genes in LUAD patients. (B) Kaplan-Meier survival curves analyzing the correlation between P-body-related gene alteration status (altered vs. WT) and OS, PFS, and DFS in LUAD patients. (C) Box plots comparing the mRNA expression levels of representative P-body-related genes between normal and tumor tissues. (D) Integrative analysis of the correlation between gene expression and DNA methylation (top) and CNV alterations (bottom) for P-body-related genes. Blue and orange bubbles represent positive/negative correlation. Circle sizes represent the abs [correlation coefficient (R)] and black circle indicating statistical significance. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001; ns, not significant. CNV, copy number variation; DFS, disease-free survival; LUAD, lung adenocarcinoma; OS, overall survival; PFS, progression-free survival; TCGA, The Cancer Genome Atlas; WT, wild-type.

We analyzed the RNA expression levels of these 24 gene in paired tumor and normal tissues. Among these genes, 9 genes were significantly upregulated in tumor tissues compared to normal controls, with varying degrees of statistical significance: EDC3, GEMIN5, LSM14B, LSM2, MOV10, PATL1, RALY, UPF3B, YWHAG; while 3 genes were significantly decreased in tumor tissues: CASC3, DCP1A, HTT (Figure 1C). To further understand the regulatory mechanisms underlying these expression changes, we examined the relationships between DNA methylation, CNVs, and gene expression. Our data demonstrated that while DNA methylation was inversely correlated with gene expression for some genes, CNVs significantly impacted others, highlighting a complex regulatory network governing P-body gene expression in LUAD (Figure 1D).

Gene expression patterns and clinical outcomes in LUAD

Having established that P-body-related genes are frequently dysregulated in LUAD, we next investigated whether their expression levels could predict patient prognosis. Univariate Cox regression analysis identified eight P-body-related genes significantly associated with OS in the TCGA LUAD cohort (P<0.05): EDC3, PATL1, PCBP1, MOV10, LSM14A, PCBP2, YBX1, and YWHAG (Figure 2A). To address multicollinearity among these candidate genes and construct a robust prognostic model, we performed LASSO-Cox regression analysis with ten-fold cross-validation. This procedure effectively eliminated multicollinearity and ultimately retained five genes (MOV10, PCBP1, PCBP2, YBX1, YWHAG) (Figure 2B,2C).

Figure 2 Analysis of P-body-related risk score based on the RNA expression and its association with survival outcomes in TCGA LUAD cohort. (A) Forest plot showing the association between P-body genes and OS. (B) Cross-validation error profile of LASSO-Cox regression. (C) LASSO coefficient profile. The vertical red dashed line marks λ.min =0.0207, retaining 5 genes (YWHAG, PCBP2, MOV10, PCBP1, YBX1) with non-zero coefficients. (D) Kaplan-Meier survival curves for OS, PFI, and DSS, comparing groups with high and low P-body-related risk score. (E) Multivariable cox regression analysis of the impact of P-body-related risk score and clinical factors (tumor stage, gender, age) on OS, PFI, and DSS. *, P<0.05; **, P<0.01; ***, P<0.001. DSS, disease-specific survival; LASSO, least absolute shrinkage and selection operator; LUAD, lung adenocarcinoma; OS, overall survival; PFI, progression-free interval; TCGA, The Cancer Genome Atlas.

Using the regression coefficients from the LASSO model, we calculated a P-body-related risk score for each patient (see “Methods” for formula). To assess model stability and overfitting, we performed 1,000-replicate bootstrap resampling, which yielded an optimism-corrected C-index of 0.614 [95% confidence interval (CI): 0.566–0.666], indicating good model stability without significant overfitting.

KM survival analysis revealed that a high P-body-related risk score was significantly associated with poorer OS, progression-free interval (PFI), and disease-specific survival (DSS) (log-rank P<0.05 for all, Figure 2D). Furthermore, multivariable Cox regression analysis, adjusting for tumor stage, gender, and age, confirmed that the P-body-related risk score was an independent prognostic factor for OS and DSS in LUAD (Figure 2E).

Validation of P-body-related risk score prognostic value in GEO datasets

To validate the robustness and generalizability of the P-body-related risk score, we assessed its prognostic value using 5 independent GEO cohorts (GSE42127, GSE31210 GSE37745, GSE41271, GSE13213). Consistent with our primary analysis in the TCGA cohort, KM analyses demonstrated that a high P-body-related risk score was consistently associated with poorer survival outcomes across all cohorts (Figure 3). The prognostic association reached statistical significance (log-rank P<0.05) in four cohorts (GSE42127, GSE41271, GSE31210, GSE13213) and showed consistent adverse trends in the remaining cohort (GSE37745, P=0.07). Despite some variability in statistical significance, the consistent direction of effect across all 5 independent cohorts supports the overall robustness and generalizability of the P-body-related risk score as a prognostic biomarker for LUAD.

Figure 3 Validation of the prognostic value of P-body-related risk score in LUAD cohort from GEO datasets. (A-E) Kaplan-Meier survival curves comparing survival outcomes between high and low P-body-related risk score groups for 5 different cohorts from GEO databases. GEO, Gene Expression Omnibus; LUAD, lung adenocarcinoma.

RNA functional regulation associated with P-body-related risk score

Given the central role of P-bodies in post-transcriptional regulation, we next explored whether the P-body-related risk score reflects differences in RNA processing activities. GSVA revealed significant associations between the risk score and various RNA-mediated processes, including those involving mRNA, non-coding RNA, and regulatory RNA. Specifically, a low score was significantly correlated with decreased activity of RNA polymerase-related processes, including transcription, elongation, and termination (Figure 4A), suggesting reduced overall RNA turnover.

Figure 4 Differential gene expression and functional enrichment analysis. (A) Bar plot showing the GSVA score in RNA processing between high and low P-body-related risk score groups. (B-D) Fold change analysis of GSVA score of pathways related to mRNA, tRNA and rRNA procession between high and low P-body-related risk score groups. Orange represents upregulated pathways in high score group. Blue represents downregulated pathways in high score group. *, P<0.05; **, P<0.01; ****, P<0.0001. GSVA, gene set variation analysis.

Focusing specifically on mRNA metabolism which is the primary function of P-bodies, we observed distinct patterns associated with the risk score. High score group exhibited enriched expression of genes involved in mRNA decay, capping, deadenylation, and splicing, whereas a low score correlated with enhanced mRNA editing activity (Figure 4B). Additionally, reduced tRNA and rRNA functional activity was observed in the low score group (Figure 4C,4D). Collectively, these findings reveal a significant association between the transcriptional activity of P-body-related risk score and the regulation of RNA processing, particularly mRNA dynamics.

Correlation between P-body-related risk score and genomic alterations

Recent studies have linked RNA granules to DNA damage response and genomic integrity. We therefore examined whether the P-body-related risk score is associated with features of genomic instability. Compared to the low score group, patients in the high score group exhibited significantly elevated levels of multiple genomic instability markers, including TMB, TNB, CNV score, HRD score, LOH fraction, and ploidy (Figure 5A, P<0.05 for all).

Figure 5 Association of P-body-related risk score with genomic instability of tumors in LUAD. (A) Box plots showing the association between P-body-related risk score and genomic instability characteristics: TMB, TNB, CNV score, HRD score, LOH, and Ploidy. (B) Volcano plot depicting the differentially mutated genes between high and low P-body-related risk score groups. (C) Forest plot showing the OR with 95% CI for various mutated pathways with high and low P-body-related risk score groups. Orange: pathways with statistically significant differences in mutation frequency between the high- and low-risk score groups; blue: pathways without statistically significant differences in mutation frequency between the high- and low-risk score groups. *, P<0.05; **, P<0.01; ****, P<0.0001. CI, confidence interval; HRD, homologous recombination deficiency; LOH, loss of heterozygosity; LUAD, lung adenocarcinoma; OR, odds ratio; TMB, tumor mutational burden; TNB, tumor neoantigen burden.

To explore the potential mechanisms underlying these associations, we compared gene mutation profiles between the high score and low score groups. Significantly higher frequency of mutation in driver genes including TP53, TTN, CSMD3, PTPRD was observed in high score group (Fisher’s exact test, Figure 5B). Pathway enrichment analysis of mutated genes revealed high score group had significant higher proportions of patients with mutations in pathways of p53, Hippo, Notch, HAT, HMT, DDR, cell cycle, TGF-beta, RTK-RAS, and SWI/SNF (Fisher’s exact test, Figure 5C). These findings provide a potential mechanistic link between high P-body-related risk score and the observed genomic instability.

Differential pathway enrichment and phosphoproteomics analysis in high vs. low P-body-related risk score groups

To further elucidate the biological processes underlying the prognostic differences, we performed pathway-level analyses comparing high score and low score groups. GSVA identified significant enrichment of hallmark cancer pathways in the high score group, including mitotic spindle, G2/M checkpoint, PI3K-AKT-mTOR signaling, and E2F targets (Figure 6A). All core pathways associated with cell proliferation and tumor progression.

Figure 6 Differential pathway enrichment and phosphor-proteomics analysis in high vs. low P-body-related risk score groups. (A) Bar plots showing the comparison of GSVA score of the 50 pathways of cancer hall markers. Orange: pathways significantly higher enriched in high-risk score group; blue: pathways significantly higher enriched in low-risk score group. (B) Volcano plot displaying differentially expressed proteins between high and low P-body-related risk score groups. Orange dots represent proteins with significantly higher expression in the high score group (log₂FC <0); blue dots represent proteins with significantly higher expression in the low score group (log₂FC >0). (C) Lollipop plot illustrating the log₂FC of differentially phosphorated proteins, with genes upregulated in high P-body-related risk score group shown in orange and downregulated in blue. FC, fold change; GSVA, gene set variation analysis.

These findings were corroborated at the protein level using reverse-phase protein array (RPPA) data. The high score group showed upregulation of proteins involved in cell proliferation, metabolism, and DNA repair, including PAR, EIF4E/G, CYCLINB1, Aurora A/B, CDC25C, and MMP14 (Figure 6B). Furthermore, phosphoproteomic analysis revealed increased activation of DNA repair molecules (e.g., p-Wee1, p-P27) and pro-proliferative signaling molecules (e.g., p-mTOR, p-MEK1, p-S6) in the high score group (Figure 6C). Together, these multi-omics analyses consistently indicate that a high P-body-related risk score is associated with enhanced proliferative capacity, metabolic activity, and DNA repair signaling.

Immune microenvironment differences

Given the emerging role of RNA granules in immune regulation, we finally characterized the tumor immune microenvironment in correlation to the P-body-related risk score. Analysis of immune subtypes revealed that the low score group had a significantly higher proportion of C3 (inflammatory) and IE/F subtypes, indicating a more inflammatory immune microenvironment (Figure 7A,7B). Consistently, the low score group exhibited significantly higher T cell receptor (TCR) and B cell receptor (BCR) diversity, as measured by Shannon and richness indices (Figure 7C, P<0.05), suggesting a more diverse immune repertoire.

Figure 7 Tumor immune microenvironment comparison between patients with high and low P-body-related risk score. (A) Sankey diagram illustrating the immune subtype distribution of high and low P-body-related risk score groups. (B) Proportion of immune subtypes in high and low P-body-related risk score groups. (C) Box plots showing the distribution of BCR/TCR richness, evenness, and Shannon index between high and low P-body-related risk score groups. (D) Box plots comparing the expression levels of 28 immune cell types between high and low P-body-related risk score group. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001; ns, not significant. BCR, B cell receptor; D, immune-depleted; E, immune-enriched; F, fibrotic; TCR, T cell receptor.

To ensure robustness, we assessed immune cell infiltration using two independent algorithms. ssGSEA analysis showed that the low score group had significantly higher infiltration of key anti-tumor immune cell populations, including activated CD8⁺ T cells, CD4⁺ T cells, B cells, and dendritic cells (Figure 7D, P<0.05 for all). These findings were validated using the xCELL algorithm, which confirmed significantly higher levels of CD8⁺ T cells, B cells, dendritic cells, and overall immune score in the low score group (Table S1, P<0.05 for all).

At the molecular level, the low score group exhibited upregulation of antigen presentation genes (e.g., HLA-DMA, HLA-DPB1, HLA-DOB) and downregulation of chemokines (e.g., CXCL9, CXCL10, CXCL11, CCL5) (Figure 8A). In contrast, the high score group showed elevated expression of immune checkpoint molecules, including CD276, PDCD1LG2, and CD274 [programmed death-ligand 1 (PD-L1)] (Figure 8A), suggesting a potential association with immunosuppressive status. These observations were further supported by RPPA analysis, which revealed higher protein levels of immune components such as CD45, CD4 , and lower protein level of B7-H3, PDL1 in the low score group (Figure 8B).Taken together, these multi-dimensional immune analyses demonstrate that the P-body-related risk score stratifies LUAD patients into distinct immune phenotypes: a low score is related with an immunologically active, “hot” tumor microenvironment, while a high score correlates with an immunosuppressive, “cold” phenotype.

Figure 8 P-body-related risk score stratification reveals distinct expression patterns of immune-related molecules in lung adenocarcinoma. (A) Volcano plots showing differentially expressed immune genes (HLA, chemokines, immune checkpoint inhibitors and stimulators) in high and low P-body-related risk score groups. Orange dots represent genes with significantly higher expression in the high score group (log₂FC <0); blue dots represent genes with significantly higher expression in the low score group (log₂FC >0). (B) Box plots displaying the protein levels of specific immune marker proteins or checkpoint molecules in high and low P-body-related risk score groups. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001. FC, fold change; HLA, human leukocyte antigen.

Discussion

In this study, we developed a P-body-related risk score using RNA expression levels of five genes that encode core P-body components. We then systematically examined its relationships with clinical outcomes, molecular characteristics, and immune microenvironment status in LUAD. Our results indicate that the risk score is significantly associated with patient survival, tumor proliferation, genomic instability, and immune contexture. These findings underline the clinical relevance of P-body-associated genes in LUAD.

The observation that a high P-body-related risk score predicts poorer overall and DFS is consistent with recent studies linking P-body components to cancer progression (16,24,25). P-bodies are dynamic cytoplasmic granules that control mRNA stability, storage, and translational repression (3). Disruption of these processes can change expression of oncogenes and tumor suppressors, which contributes to malignant transformation (25). The five genes in our risk score (MOV10, PCBP1, PCBP2, YBX1, and YWHAG) are well-characterized P-body components with established roles in mRNA metabolism, translational regulation, and oncogenic signaling (26-29). The prognostic value of our risk score may therefore reflect the combined effects of these genes on mRNA metabolism and oncogenic signaling.

High P-body-related risk scores were associated with enriched cell cycle and DNA repair pathways, along with several markers of genomic instability such as increased TMB, HRD score, and copy number alterations. These results agree with studies that connect RNA granules to replication stress responses and genomic integrity (30,31). For example, P-body proteins are involved in cellular responses to replication stress (30), and their dysregulation may promote genomic alterations. The increased frequency of DNA damage response pathway mutations (e.g., p53, DDR) in high score tumors further supported this connection. Although these findings are correlative, it raises the possibility that deserves further investigation: high score patients may benefit from targeted therapies such as PARP inhibitors or cell cycle checkpoint inhibitors (28,32).

One of the most notable findings in this study is the strong relationship between the P-body-related risk score and the tumor immune microenvironment. Low score tumors showed features of an immunologically active, “hot” microenvironment, including greater infiltration of CD8⁺ T cells, B cells, and dendritic cells, as well as higher expression of antigen presentation genes and chemokines that support immune recruitment (e.g., CXCL9, CXCL10). By contrast, high score tumors displayed a “cold” phenotype marked by reduced effector immune cell infiltration.

These results are consistent with recent reports showing that RNA granules contribute to immune regulation. For instance, stress granules and P-bodies can modulate T cell activation by controlling the translation of immune-related mRNAs (33,34). Franchini et al. showed that stress granules enhance translation of inhibitory immune checkpoint mRNAs (e.g., PD-1, CTLA4) in T cells, which in turn shapes antitumor immune responses (35). While our data do not prove causality, they suggest that expression of P-body-associated genes may influence immune modulation in LUAD. These findings support the idea that the risk score could help stratify patients for immunotherapy. Low score patients with “hot” tumors may be more likely to benefit from immune checkpoint inhibitors, whereas high score patients with “cold” tumors may need combination strategies to reverse immunosuppression. These hypotheses, however, still require prospective validation.

Limitations

Several limitations should be acknowledged. First, all cohorts included are retrospective and heterogeneous in terms of patient demographics, treatment regimens, and follow-up times. Although validation across multiple independent cohorts supports the robustness of our results, prospective studies with standardized protocols are needed to confirm clinical utility before routine application. Second, the five genes in our risk score are not exclusive to P-bodies. MOV10, PCBP1, PCBP2, and YBX1 are multifunctional RNA-binding proteins with roles beyond P-bodies (e.g., nuclear RNA metabolism, stress granule dynamics), and YWHAG is primarily known as a signaling adaptor, though some 14-3-3 family members have been reported to bind RNA. Moreover, because our score is based on bulk mRNA expression, it cannot capture P-body formation, a process involving liquid-liquid phase separation and post-translational modifications invisible to transcriptomics (4,6). Whether this signature reflects true P-body biology or broader cellular processes remains to be determined by future IHC/IF studies directly quantifying P-body granules (e.g., using DDX6 or EDC4) and correlating them with our score. Third, our immune analyses rely on bulk transcriptomic deconvolution; spatial transcriptomics and multiplex immunohistochemistry would provide valuable validation. Finally, all findings are observational and do not establish causality. Functional experiments are required to elucidate the mechanistic roles of individual P-body components in LUAD biology.

Conclusions

In summary, we developed a P-body-related risk score that independently predicts prognosis in LUAD and correlates with distinct tumor biological features and immune phenotypes. While the score serves as a transcriptomic proxy rather than a direct measure of P-body formation, it offers a clinically useful tool for risk stratification and generates testable hypotheses. Prospective validation and functional studies are needed to establish its clinical value.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2736/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2736/prf

Funding: None.

Conflicts of Interest: Both authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2025-1-2736/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Senatore E, Rinaldi L, Chiuso F, et al. Reshaping cancer cell plasticity by P-body dynamics and protein translation. Commun Biol 2026;9:160. [Crossref] [PubMed]
2. Feng MW, Mossanen-Parsi A, Stonys V, et al. P-bodies act as dynamic control hubs for RNA processing and storage. J Biol Chem 2026;302:111273. [Crossref] [PubMed]
3. Standart N, Weil D. P-Bodies: Cytosolic Droplets for Coordinated mRNA Storage. Trends Genet 2018;34:612-26. [Crossref] [PubMed]
4. Hubstenberger A, Courel M, Bénard M, et al. P-Body Purification Reveals the Condensation of Repressed mRNA Regulons. Mol Cell 2017;68:144-157.e5. [Crossref] [PubMed]
5. Decker CJ, Parker R. P-bodies and stress granules: possible roles in the control of translation and mRNA degradation. Cold Spring Harb Perspect Biol 2012;4:a012286. [Crossref] [PubMed]
6. Luo Y, Na Z, Slavoff SA. P-Bodies: Composition, Properties, and Functions. Biochemistry 2018;57:2424-31. [Crossref] [PubMed]
7. Tosiano MA, Lanni F, Mitchell AP, et al. Roles of P-body factors in Candida albicans filamentation and stress response. PLoS Genet 2025;21:e1011632. [Crossref] [PubMed]
8. Hatfield D, Rodriguez W, Mehrmann T, et al. The anti-viral protein Shiftless blocks p-body formation during KSHV infection. J Gen Virol 2026;107:002229. [Crossref] [PubMed]
9. Araoyinbo A, Salat-Canela C, Vještica A. Gamete fusion triggers cytosolic functions and P-body recruitment of the RNA-binding protein Mei2 to drive fission yeast zygotic development. Genes Dev 2026;40:328-50. [Crossref] [PubMed]
10. Ru W, Cheng J, Gao Y, et al. METTL3-mediated m(6)A modification regulates muscle development by promoting TM4SF1 mRNA degradation in P-body via YTHDF2. Int J Biol Macromol 2025;295:139576. [Crossref] [PubMed]
11. Kodali S, Proietti L, Valcarcel G, et al. RNA sequestration in P-bodies sustains myeloid leukaemia. Nat Cell Biol 2024;26:1745-58. [Crossref] [PubMed]
12. Garg M, Lamba S, Rajyaguru PI. Emerging trends in the disassembly of stress granules and P-bodies. Cell Rep 2025;44:116340. [Crossref] [PubMed]
13. Wang H, Peng H, Zhang Z, et al. FTO-mediated m(6)A demethylation regulates IGFBP3 expression and AKT activation through IMP3-dependent P-body re-localisation in lung cancer. Clin Transl Med 2025;15:e70392. [Crossref] [PubMed]
14. Senatore E, Avolio R, Rinaldi L, et al. Praja2 controls P-body assembly and translation in glioblastoma by non-proteolytic ubiquitylation of DDX6. EMBO Rep 2025;26:2347-77. [Crossref] [PubMed]
15. Minagawa K, Katayama Y, Nishikawa S, et al. Inhibition of G(1) to S phase progression by a novel zinc finger protein P58(TFL) at P-bodies. Mol Cancer Res 2009;7:880-9. [Crossref] [PubMed]
16. Shen X, Peng X, Guo Y, et al. YAP/TAZ enhances P-body formation to promote tumorigenesis. Elife 2024;12:RP88573. [Crossref] [PubMed]
17. Ghashghaei M, Le CT, Shaalan H, et al. miR-148a-3p and DDX6 functional link promotes survival of myeloid leukemia cells. Blood Adv 2023;7:3846-61. [Crossref] [PubMed]
18. Li J, Chen H, Bai L, et al. RNF115 upregulation by YBX1-dependent m5C modification promotes hepatocellular carcinoma progression. NPJ Precis Oncol 2025;9:375. [Crossref] [PubMed]
19. Khozyainova AA, Menyailo ME, Subrakova VG, et al. Spatial Dynamics of Tumor Cell Plasticity in Lung Adenocarcinoma Revealed by Region-Specific Single Cell Transcriptomics. Genes Chromosomes Cancer 2025;64:e70101. [Crossref] [PubMed]
20. Legrand N, Dixon DA, Sobolewski C. AU-rich element-binding proteins in colorectal cancer. World J Gastrointest Oncol 2019;11:71-90. [Crossref] [PubMed]
21. Fang L, Zhang L, Wang M, et al. Pooled CRISPR Screening Identifies P-Bodies as Repressors of Cancer Epithelial-Mesenchymal Transition. Cancer Res 2024;84:659-74. [Crossref] [PubMed]
22. Kim EO, Kang SE, Choi M, et al. CCR4 NOT transcription complex subunit 2 regulates TRAIL sensitivity in non small cell lung cancer cells via the STAT3 pathway. Int J Mol Med 2020;45:324-32. [Crossref] [PubMed]
23. Millar SR, Huang JQ, Schreiber KJ, et al. A New Phase of Networking: The Molecular Composition and Regulatory Dynamics of Mammalian Stress Granules. Chem Rev 2023;123:9036-64. [Crossref] [PubMed]
24. Nsengimana B, Khan FA, Ngowi EE, et al. Processing body (P-body) and its mediators in cancer. Mol Cell Biochem 2022;477:1217-38. [Crossref] [PubMed]
25. Kodali S, Proietti L, Valcarcel G, et al. RNA sequestration in P-bodies sustains myeloid leukaemia. Nat Cell Biol 2024;26:1745-58. [Crossref] [PubMed]
26. Wang F, Ruan L, Yang W, et al. MOV10, a novel immunotherapy and prognostic biomarker, contributes to glioma development by regulating autophagy. Sci Rep 2026;16:9501. [Crossref] [PubMed]
27. Fréreux C, Karam JAQ, Howley BV, et al. PCBP1 binding to single-stranded poly-cytosine motifs enhances cGAS sensing and impairs breast cancer development. Commun Biol 2026;9:179. [Crossref] [PubMed]
28. Qi Z, He L, Xu Z, et al. PCBP2 Mediates Olaparib Resistance in Breast Cancer by Inhibiting m6A Methylation to Stabilize PARP1 mRNA. Cancer Res 2025;85:3949-65. [Crossref] [PubMed]
29. Li C, Yuan Y, Jiang X, et al. Unveiling the immunological functions of the RNA m(5)C reader YBX1 in cancer. Front Immunol 2025;16:1703697. [Crossref] [PubMed]
30. Loll-Krippleber R, Brown GW. P-body proteins regulate transcriptional rewiring to promote DNA replication stress resistance. Nat Commun 2017;8:558. [Crossref] [PubMed]
31. Emam A, Wu X, Xu S, et al. Stalled replication fork protection limits cGAS-STING and P-body-dependent innate immune signalling. Nat Cell Biol 2022;24:1154-64. [Crossref] [PubMed]
32. Hernández G, Ramírez MJ, Minguillón J, et al. Decapping protein EDC4 regulates DNA repair and phenocopies BRCA1. Nat Commun 2018;9:967. [Crossref] [PubMed]
33. Curdy N, Lanvin O, Cerapio JP, et al. The proteome and transcriptome of stress granules and P bodies during human T lymphocyte activation. Cell Rep 2025;44:116442. [Crossref] [PubMed]
34. Zhang Y, Ren H, Ma C, et al. A PSAT1 buff of YBX1 transcriptionally sustains HLA-E-mediated evasion of NK immunity. Proc Natl Acad Sci U S A 2025;122:e2505658122. [Crossref] [PubMed]
35. Franchini DM, Lanvin O, Tosolini M, et al. Microtubule-Driven Stress Granule Dynamics Regulate Inhibitory Immune Checkpoint Expression in T Cells. Cell Rep 2019;26:94-107.e7. [Crossref] [PubMed]