PLCG1 regulates pyroptosis to inhibit the progression of non-small cell lung cancer

Abstract

Pyroptosis, an inflammatory form of programmed cell death, has been implicated in non-small cell lung cancer (NSCLC) progression. However, the specific pyroptosis-related genes causally linked to NSCLC and their upstream regulatory mechanisms remain poorly understood. We performed a two-sample Mendelian randomization (MR) study to assess causal relationships between 33 pyroptosis-related genes and NSCLC risk using genetic instruments from the deCODE database and outcome data from the FinnGen consortium (6446 NSCLC cases, 378,749 controls). Inverse-variance weighting was the primary analysis method, with MR-Egger, weighted median, and MR-PRESSO as sensitivity analyses. To identify upstream regulators, we integrated pQTL data from the The UK Biobank Pharma Proteomics Project and performed two-step MR mediation analysis. Identified genes were further validated using mRNA expression data from The Cancer Genome Atlas (TCGA) and genotype-tissue expression (GTEx) databases. MR analysis identified 2 pyroptosis-related genes significantly associated with NSCLC: IL1B (odds ratio [OR] = 1.087, 95% confidence interval [CI] = 1.009–1.171, P = .028) and PLCG1 (OR = 0.874, 95% CI = 0.766–0.997, P = .044). Upstream analysis revealed that AMBP and CCNE1 were causally associated with PLCG1 expression. Mediation analysis demonstrated that PLCG1 mediated 11.7% of the total effect of AMBP on NSCLC risk (mediation effect: OR = 1.02, 95% CI = 1.01–1.03, P = .008). mRNA expression validation confirmed that AMBP and PLCG1 were significantly upregulated in NSCLC tissues compared to normal tissues (P < .001). This study establishes causal links between pyroptosis-related genes and NSCLC, identifying IL1B as a risk factor and PLCG1 as a protective factor. We further demonstrate that AMBP may promote NSCLC progression partly through upregulating PLCG1 expression, with PLCG1 mediating 11.7% of this effect. These findings provide novel insights into pyroptosis regulation in NSCLC and highlight potential therapeutic targets.

1. Introduction

Pyroptosis is a programmed cell death form triggered by inflammasomes and mediated by gasdermin proteins, characterized by cell swelling, pore formation, and the release of cellular contents.^[1,2] As an inflammatory cell death, pyroptosis activates antitumor immunity, offering new potential for cancer treatment. This form of lytic cell death is primarily executed through 2 major pathways: the canonical inflammasome pathway and the noncanonical pathway. In the canonical pathway, sensor proteins such as NOD-like receptor family pyrin domain containing 3 (NLRP3) recognize pathogenic or danger signals, leading to the assembly of inflammasomes and activation of caspase-1.^[3] The noncanonical pathway directly involves human caspase-4/5 (or murine caspase-11), which binds to intracellular lipopolysaccharide.^[4] Both activated caspase-1 and caspase-4/5/11 subsequently cleave gasdermin D at a specific site, releasing its N-terminal pore-forming domain. This domain translocates to the plasma membrane, oligomerizes, and forms lytic pores, resulting in cell swelling, membrane rupture, and the release of pro-inflammatory cytokines such as IL-1B and IL-18.^[5] Most signaling pathways involving gasdermin E, gasdermin D, gasdermin C, or gasdermin B as executors trigger pyroptosis in cancer.^[6,7]

Pyroptosis exhibits both pro-tumor and antitumor effects. Chronic pyroptosis in a harmful tumor microenvironment promotes cancer progression, while acute activation enhances immune cell infiltration and induces cancer cell death, inhibiting tumor growth.^[8] Lung cancer is the second most common cancer and the leading cause of cancer mortality worldwide, with 2.2 million new cases and 1.79 million deaths annually.^[9] Specifically, non-small cell lung cancer (NSCLC) accounts for approximately 85% of all lung cancer cases.^[9] In the context of lung cancer, this immunogenic form of cell death has garnered significant attention, with certain anticancer therapies inducing gasdermin-mediated pyroptosis. This process promotes immune infiltration and reduces the immunosuppressive tumor microenvironment, making pyroptosis induction a promising therapeutic strategy.^[10] However, despite advancements in understanding the mechanisms of pyroptosis, its specific role in NSCLC remains unclear. This ambiguity may be due to variations in cancer stage, gasdermin expression, and the composition of the TME.^[11] The effects of different gasdermin proteins on tumor growth, metastasis, and prognosis in NSCLC are not fully understood, necessitating further investigation.^[12–15]

Mendelian randomization (MR) is a method that uses genetic variations associated with specific exposures as instrumental variables (IVs) to estimate causal relationships between the exposure of interest and the desired outcome.^[16,17] This study uses MR to investigate the causal role of the pyroptosis pathway in lung cancer risk. By elucidating pyroptosis mechanisms, the study aims to clarify its role in tumorigenesis, progression, and treatment response, contributing to novel therapeutic strategies for lung cancer.

2. Materials and methods

2.1. Study design

The study follows STROBE-MR guidelines (Table S1, Supplemental Digital Content 1)^[18,19] and includes 3 main components: analyzing the causal relationship between pyroptosis genes and NSCLC (Fig. 1B, b); identifying potential upstream mechanisms (Fig. 1B, c); and mediation analysis of upstream molecules’ effects on NSCLC via pyroptosis (Fig. 1B). Single-nucleotide polymorphisms (SNPs) serve as IVs, based on 3 MR analysis assumptions.

Figure 1.

Mediation analysis model of the UKB-PPP pQTL, pyroptosis gene expression, and NSCLC risk. (A) The 3 core MR assumptions: instrumental variables (SNPs) must have a strong association with the exposure (relevance), no association with confounders (e.g., smoking, age, BMI) (independence), and affect the outcome solely through the exposure (exclusion restriction). (B) The 3 analytical components: (1) effect of pyroptosis gene expression (exposure) on NSCLC risk (outcome); (2) effect of UKB-PPP pQTL (exposure) on PLCG1 expression (outcome); (3) total effect of UKB-PPP pQTL (exposure) on NSCLC risk (outcome), decomposed into a direct effect and an indirect effect mediated by PLCG1. BMI = body mass index, MR = Mendelian randomization, NSCLC = non-small cell lung cancer, SNP = single-nucleotide polymorphisms, UKB-PPP = The UK Biobank Pharma Proteomics Project.

MR relies on 3 core assumptions (Fig. 1A): the IVs are strongly associated with the exposure (relevance); the IVs are independent of any confounders of the exposure-outcome relationship (independence); and the IVs affect the outcome only through the exposure and not via alternative pathways (exclusion restriction). In this study, these assumptions were applied across 3 analytical components: IVs were selected at genome-wide significance (P < 5 × 10⁻⁸) with F-statistics > 10 to ensure relevance (assumption 1). PhenoScanner was used to exclude SNPs associated with known lung cancer confounders (assumption 2). MR-Egger and weighted median methods were performed to detect and correct for horizontal pleiotropy (assumption 3).

2.2. Data sources

2.2.1. Pyroptosis-related gene data sources

This study identified 33 pyroptosis-related genes (Table S2, Supplemental Digital Content 2) from prior research.^[20,21] Raw pQTL data were obtained from the deCODE database^[22] (https://www.decode.com/summarydata/), with SNPs selected based on criteria: P < 5 × 10⁻⁸, clumping window 10,000 kb, and r² < 0.1. The final dataset, intersecting with our genes, included 11 genes’ pQTL data for analysis.

2.2.2. UKB-PPP pQTL data sources

The study retrieved 2940 pQTL files from the UK Biobank Pharma Proteomics Project (UKB-PPP; https://www.synapse.org/Synapse:syn51364943/wiki/622119),^[23] all from the European population. After applying criteria (P < 5 × 10⁻⁸, clumping window 10,000 kb, r² < 0.1), 2619 files remained and were used for the final analysis.

2.2.3. Outcome data sources

FinnGen is a large-scale genomics project collecting data from 500,000 Finnish biobank participants to study the genetic basis of diseases. The NSCLC genome-wide association study summary data, from the 12th release (R12) (https://www.finngen.fi/en), include 6446 patients and 378,749 controls of European descent.^[22]

2.3. MR analysis

2.3.1. Preliminary analysis

To explore the relationship between pyroptosis and NSCLC, we applied 5 methodologies: inverse-variance weighting (IVW), MR-Egger, weighted model, weighted median, and simple model. IVW uses meta-analysis to combine causal effects from each SNP’s Wald ratio. MR-Egger and weighted median supplement IVW, offering alternative MR estimates under different validity assumptions. IVW assumes all SNPs are valid IVs, providing accurate estimates. MR-Egger evaluates directional pleiotropy, while the weighted median method offers greater precision and consistent estimates, even with 50% invalid instruments. Genes associated with pyroptosis and UKB-PPP pQTLs (P < .05) were analyzed.^[24] The weighted median method provides higher precision compared to the MR-Egger.^[25]

2.3.2. Mediation analysis

We used a two-step MR design for mediation analysis to investigate the mechanisms linking pyroptosis and NSCLC (Fig. 1B). The total effect was decomposed into direct and indirect (mediated) effects. First, we estimated the effect of UKB-PPP pQTLs on NSCLC (βc), pyroptosis pQTLs (βa), and their combined effect on NSCLC (βb). The mediation effect (βc′) was calculated as βa × βb, and the mediated proportion was (βc′/βc) × 100%. Pathways with a mediation proportion >10% and P < .05 were further analyzed.^[26]

2.3.3. Sensitivity analysis

Cochran’s Q test assessed SNP heterogeneity, and scatter plots visualized SNP-exposure and SNP-outcome associations. Leave-one-out analysis tested the impact of each SNP by sequential exclusion and IVW analysis of the remaining SNPs. MR-PRESSO and MR-Egger regression were used to detect horizontal pleiotropy, with MR-PRESSO identifying significant outliers, which were excluded to correct for pleiotropy.^[27]

2.3.4. Statistical analysis and software

All statistical analyses were performed using R software (version 4.2.1; R Foundation for Statistical Computing). The following R packages were used: “TwoSampleMR” (version 0.5.6) for MR analyses, “MR_PRESSO” (version 1.0) for detection of horizontal pleiotropy, and “MendelianRandomization” (version 0.6.0) for additional sensitivity analyses. For the mediation analysis, we used the “mediate” function from the “mediation” package (version 4.5.0).

The significance threshold for instrument selection was set at P < 5 × 10⁻⁸. For the primary MR analyses, statistical significance was defined as P < .05. Due to the hypothesis-driven nature of the pyroptosis gene analysis (33 genes), no multiple testing correction was applied. For the upstream protein analysis involving 2619 proteins from UKB-PPP, we applied Bonferroni correction to account for multiple comparisons, with significance set at P < 1.91 × 10⁻⁵ (0.05/2619). The mediation effect was considered significant if the 95% confidence interval (CI) did not include the null value.

2.4. Validation of gene expression

For validation, mRNA expression levels were obtained from TCGA for both NSCLC and matched normal tissues, while data from the GTEx database provided an independent set of normal tissue controls.^[28,29] The RNA sequencing data in fragments per kilobase of transcript per million mapped reads were normalized to transcripts per million using the Toil pipeline and subsequently log2-transformed.^[30] The expression levels of these genes were visualized as boxplots using the Xiantao tools (https://www.xiantaozi.com/), and the Wilcoxon rank-sum test was employed to assess the differential expression of these genes between NSCLC and normal tissues. To eliminate the confounding effects of age, gender, and smoking history, we employed the MatchIt package (v4.7.2) in R (v4.4.2) to perform propensity score matching, using gender, age, and smoking history as matching variables, with a ratio set to normal: tumor = 1:2, employing the nearest method and a caliper of 0.2. Subsequently, the matched data were analyzed using SPSS 23 to conduct binary logistic regression.

2.5. Ethics approval and consent to participate

This MR study utilized publicly available anonymous genetic and phenotypic data. Each database had completed ethical approval before data collection. Since no personal-level data were collected and no direct involvement of human subjects was involved, no ethical approval was required.

3. Results

3.1. Identification of IVs

We identified 251 SNPs linked to 11 pyroptosis-related genes with P < 5 × 10⁻⁸, which were used as IVs for these genes (Table S2, Supplemental Digital Content 2). In addition, 129,236 SNPs from the UKB-PPP pQTL dataset, associated with 2619 upstream genes, were also selected as IVs (Table S3, Supplemental Digital Content 3).

3.2. MR analysis of the relationship between pyroptosis and NSCLC

A two-sample MR analysis revealed significant associations between 2 genes, IL1B (OR = 1.087, 95% CI = 1.009–1.171, P = .028) and PLCG1 (OR = 0.874, 95% CI = 0.766–0.997, P = .044), and NSCLC. These genes were further analyzed to explore their upstream mechanisms (Fig. 2).

Figure 2.

Causal effect of pyroptosis genes expression on NSCLC risk. CI = confidence interval, MR = Mendelian randomization, NSCLC = non-small cell lung cancer, SNP = single-nucleotide polymorphisms.

3.3. MR analysis of the relationship between UKB-PPP and NSCLC

A two-sample MR analysis was conducted between the 2619 potential upstream gene IVs identified from the UKB-PPP pQTL dataset and NSCLC outcomes. The analysis revealed that 386 potential upstream molecules were significantly associated with NSCLC, with 144 molecules exhibiting a positive correlation and 242 molecules showing a negative correlation with NSCLC (Table S4, Supplemental Digital Content 4). Among these, 26 genes showed an adjusted P-value of <.05 (Fig. 3). Fifteen genes were positively associated with the occurrence of NSCLC, including B2M (OR = 1.19, 95% CI = 1.12−1.26, P = 1.15 × 10⁻⁹), BTN2A1 (OR = 1.20, 95% CI = 1.15−1.26, P = 2.87 × 10⁻¹⁶), C2 (OR = 1.13, 95% CI = 1.07−1.18, P = 1.88 × 10⁻⁷), CDSN (OR = 1.09, 95% CI = 1.07−1.12, P = 1.33 × 10⁻¹⁵), HLA-A (OR = 1.05, 95% CI = 1.03−1.06, P = 5.88 × 10⁻⁸), HLA-E (OR = 1.11, 95% CI = 1.08−1.15, P = 9.48 × 10⁻¹³), MICB (OR = 1.05, 95% CI = 1.04−1.07, P = 3.43 × 10⁻¹⁰), PDIA5 (OR = 1.22, 95% CI = 1.15−1.29, P = 5.47 × 10⁻¹²). Eleven genes were negatively associated with the occurrence of NSCLC, including AMBP (OR = 0.78, 95% CI = 0.72−0.85, P = 1.45 × 10⁻⁸), BTN3A2 (OR = 0.91, 95% CI = 0.89−0.93, P = 1.82 × 10⁻¹⁶), CCL19 (OR = 0.77, 95% CI = 0.72−0.83, P = 1.97 × 10⁻¹³), CCNE1 (OR = 0.72, 95% CI = 0.63−0.83, P = 5.14 × 10⁻⁶), DAPK2 (OR = 0.85, 95% CI = 0.81−0.90, P = 3.72 × 10⁻⁸), IL18 (OR = 0.88, 95% CI = 0.83−0.93, P = 6.69 × 10⁻⁶), and SYT1 (OR = 0.28, 95% CI = 0.18−0.43, P = 2.77 × 10⁻⁹). To maximize the identification of potential upstream molecules, all molecules with a P value <.05 were incorporated into the subsequent mediation analysis.

Figure 3.

Causal effect of UKB-PPP pQTL on NSCLC risk. CI = confidence interval, NSCLC = non-small cell lung cancer, SNP = single-nucleotide polymorphisms, UKB-PPP = The UK Biobank Pharma Proteomics Project.

3.4. Mediation analysis of UKB-PPP – pyroptosis – NSCLC relationship

We conducted a two-step MR analysis to explore the upstream mechanisms linking NSCLC-related pyroptosis genes with 386 potential upstream molecules (Table S5, Supplemental Digital Content 5). Significant pathways were identified with mediation proportions exceeding 10% and P < .05. We found that 2 upstream genes, AMBP (OR = 1.24, 95% CI = 1.15–1.34, P = 5.83 × 10⁻⁸) and CCNE1 (OR = 1.47, 95% CI = 1.34–1.62, P = 2.68 × 10⁻¹⁶), significantly influenced PLCG1 expression, suggesting a positive correlation. Using the IVW method, we estimated the MR effects of these genes on NSCLC: AMBP (OR = 0.78, 95% CI = 0.72–0.85, P = 1.45 × 10⁻⁸) and CCNE1 (OR = 0.72, 95% CI = 0.63–0.83, P = 5.14 × 10⁻⁶). In the second step, we assessed PLCG1 as a mediator of NSCLC risk (Fig. 1B) and calculated the total, mediated, and direct effects (Table 1). The mediating effects of PLCG1 contributed 11.7% and 16.1% to the total effect of AMBP and CCNE1 on NSCLC, respectively.

Table 1.

Mediator analysis results of UKB-PPP pQTL, PLCG1 gene expression, and NSCLC risk.

UKB-PPP pQTL	Total effect (βc)	Path a (βa)	Path b (βb)	Indirect effect (βc′)	Direct effect (βc − βc′)	Proportion mediated (%)	P value
AMBP	−0.248	0.216	−0.135	−0.029	−0.219	11.70%	.020
CCNE1	−0.324	0.388	−0.135	−0.052	−0.272	16.12%	.011

NSCLC = non-small cell lung cancer, UKB-PPP = The UK Biobank Pharma Proteomics Project.

3.5. Sensitivity analysis

The sensitivity analysis revealed no evidence of heterogeneity or horizontal pleiotropy in the Cochran’s Q test and MR-Egger regression analysis. The MR-PRESSO test identified no outliers, and the leave-one-out analysis confirmed that our results were robust, further reinforcing the reliability of our MR analysis findings (Table S6, Supplemental Digital Content 6 and Fig. S1, Supplemental Digital Content 7).

3.6. Validation of gene expression

To validate the expression levels of AMBP, CCNE1, and PLCG1 in NSCLC and normal tissues, mRNA expression levels of these genes were obtained from the TCGA and GTEx databases. RNA sequencing data in fragments per kilobase of transcript per million mapped reads were normalized to transcripts per million using the Toil pipeline and log2-transformed (Table S7, Supplemental Digital Content 8). The results were visualized as boxplots (Fig. 4), and the Wilcoxon rank-sum test was used to evaluate the expression differences of these genes between NSCLC and normal tissues. The results revealed that the expression levels of AMBP and PLCG1 were significantly higher in NSCLC tissues compared to normal tissues, and similarly, the expression of CCNE1 was also considerably higher in NSCLC tissues. After propensity score matching, a matched cohort of 103 pairs (307 samples) was successfully constructed. Subsequent binary logistic regression analysis revealed that age, sex, and smoking history were not significantly associated with NSCLC in the matched data. In contrast, the association between PLCG1 and NSCLC remained significant (Table S8, Supplemental Digital Content 9).

Figure 4.

The expression of AMBP, CCNE1, and PLCG1 in NSCLC and normal tissues. NSCLC = non-small cell lung cancer. *P < .05, **P < .01, ***P < .001 vs. control group.

4. Discussion

This study employs MR to investigate the causal relationship between pyroptosis and NSCLC, identifying IL1B and PLCG1 as key pyroptosis-related genes. IL1B encodes interleukin-1β (IL-1β), a central effector of pyroptosis: it is activated by inflammasome-associated caspase-1 and secreted via gasdermin D (GSDMD) pores, exerting dual pro- and antitumor effects within the tumor microenvironment.^[31,32] PLCG1 encodes phospholipase C gamma 1, which hydrolyzes PIP₂ to generate IP₃ and DAG; IP₃ triggers Ca²⁺ release from the endoplasmic reticulum, a critical signal for NLRP3 inflammasome activation, thereby modulating the pyroptosis threshold.^[33,34] In addition to these genes, our MR analysis implicated upstream regulators including AMBP and CCNE1. While CCNE1 showed a positive correlation with PLCG1 in MR, its elevated expression in NSCLC tumor tissues (TCGA/GTEx) contradicted a protective role, leading us to focus on AMBP. We propose that AMBP may indirectly promote NSCLC progression by modulating PLCG1 expression, with PLCG1 mediating 11.7% of AMBP’s total effect.

Our results identified a significant association between IL1B and NSCLC risk (OR = 1.087). This aligns with IL-1β’s dual role in inflammation and pyroptosis. Mechanistically, IL-1β is synthesized as pro-IL-1β and matures via inflammasome-activated caspase-1.^[35] Its release depends on GSDMD pore formation, linking it directly to pyroptosis execution.^[36] In NSCLC, this pathway has complex effects. Secreted IL-1β can promote a pro-tumorigenic TME, enhancing angiogenesis, metastasis, and immunosuppression.^[37,38] Conversely, robust pyroptosis via this axis can induce immunogenic cell death and activate antitumor immunity.^[38] The modest MR effect size suggests IL-1β acts as a critical modulator within this network, not a sole driver. Therapeutically, the CANTOS trial showed that inhibiting IL-1β with canakinumab reduces lung cancer incidence, supporting this axis’s relevance.^[39]

Our analysis revealed a modest protective effect for PLCG1 (OR = 0.874). PLCG1 is a signaling enzyme that hydrolyzes PIP₂ to generate IP₃ and DAG. The IP₃ then binds to its receptor on the endoplasmic reticulum, triggering Ca²⁺ release into the cytosol.^[34,40] This elevation in intracellular Ca²⁺ is a well-established “signal 2” for the activation of the NLRP3 inflammasome, a crucial sensor upstream of pyroptosis.^[34] Therefore, we hypothesize that PLCG1, through fine-tuning the spatiotemporal dynamics of Ca²⁺ signaling, may act as a critical modulator that sets the activation threshold for the NLRP3 inflammasome and subsequent pyroptosis. This indirect, calibrating role is consistent with the modest genetic effect observed. Concurrently, the DAG produced activates protein kinase C, which can phosphorylate and regulate components of inflammatory and cell death pathways.^[41] This dual output (Ca²⁺ and protein kinase C) positions PLCG1 as a hub linking extracellular signals to pyroptosis. Notably, PLCG1 also promotes NSCLC tumor survival and metastasis,^[42] contrasting our protective association. This paradox may be explained by context-dependent functions: in the specific setting of an inflamed TME, PLCG1-mediated sensitization to pyroptosis (a tumor-suppressive process) might outweigh its direct pro-tumorigenic signaling. This aligns with therapeutic strategies disrupting calcium homeostasis to induce pyroptosis.^[43]

AMBP (α-1-microglobulin-bikunin precursor gene) encodes 2 proteins: α-1-microglobulin (A1M) and bikunin. A1M belongs to the lipocalin family and performs physiological functions as an antioxidant and tissue-cleaning agent through its chemical reductase activity, free radical scavenging, and heme-binding mechanisms.^[44] Bikunin, on the other hand, is a structural component of the extracellular matrix and a Kunitz-type plasma protease inhibitor.^[45,46]

Our study reveals that PLCG1 partially mediates (11.7%) the effect of AMBP on NSCLC risk. To interpret this statistical association, we propose a biologically plausible link: Bikunin, as an uPA inhibitor, may indirectly influence signaling pathways downstream of uPA.^[47] Notably, uPA/uPAR signaling is a classic upstream activator of the phospholipase C pathway, including the PLCG1 isoform.^[48,49] Therefore, by inhibiting uPA, bikunin theoretically possesses the potential to perturb downstream PLCG1-related signaling. This provides a mechanistic context for our genetic findings. Consequently, we refine our hypothesis: AMBP, likely through bikunin, may exert part of its effect by inhibiting uPA, thereby indirectly modulating the activity of intracellular signaling networks centered on PLCG1 in the tumor microenvironment. This modulation could affect downstream processes such as pyroptosis, in which PLCG1 is implicated.

The observed “partial mediation” (11.7%) aligns with the pleiotropic nature of AMBP function.^[50] Its impact on NSCLC likely arises from a synergistic network of multiple pathways. Beyond the potential uPA-PLCG1 axis, the well-established direct immunomodulatory functions of A1M^[51] and the direct anti-invasive functions of bikunin^[52] are also critically important. Thus, the PLCG1 pathway represents one component within AMBP’s broader functional network. This enhanced understanding of AMBP’s multifaceted role may inform new strategies for lung cancer research.

This study is the first to explore the relationship between pyroptosis and NSCLC using mediation MR analysis. We identified putative causal links between pyroptosis-related genes (IL1B and PLCG1) and NSCLC risk and explored their upstream mechanisms using UKB-PPP pQTL data. However, several limitations should be noted. First, findings require experimental validation. Second, the genetic data from UKB-PPP and TCGA used in this study are predominantly derived from individuals of European ancestry. Although the core biology of pyroptosis and phospholipase C signaling is evolutionarily conserved,^[53,54] differences in genetic architecture, allele frequencies, and environmental exposures across populations may influence the generalizability of our specific causal estimates and mediation proportions (e.g., for the AMBP-PLCG1 axis) to non-European populations.^[55] Future multiethnic MR studies are needed to verify the transferability of these instruments and findings.^[56] Future research should incorporate larger and more diverse datasets (e.g., multi-ancestry pQTL resources), additional cell death-related genes, and experimental models to further elucidate the role of pyroptosis in NSCLC progression, which could ultimately inform patient treatment strategies.

5. Conclusions

In conclusion, our study provides novel insights into the role of pyroptosis in NSCLC progression. Through MR and bioinformatic analyses, we identified 2 key pyroptosis-related genes, IL1B and PLCG1, that are causally linked to NSCLC. Further investigation revealed that AMBP may promote NSCLC development by modulating PLCG1 expression, with PLCG1 serving as a key mediator in this process. These findings suggest that targeting pyroptosis-related genes, particularly PLCG1, could offer new therapeutic avenues for NSCLC management and warrant further exploration in clinical settings.

Acknowledgments

The deCODE database, UKB-PPP, and FinnGen database provide pQTL data for pyroptosis genes, and GWAS data for NSCLC, respectively. We gratefully acknowledge all the researchers for sharing the genome-wide summary statistics.

Author contributions

Conceptualization: Yihui Zhang, Guihui Wang.

Data curation: Yihui Zhang, Guihui Wang, Ziyan Mo.

Formal analysis: Yihui Zhang, Guihui Wang, Ziwen Qin.

Investigation: Yihui Zhang, Guihui Wang.

Methodology: Yihui Zhang, Guihui Wang, Jixian Liu.

Project administration: Yihui Zhang, Guihui Wang.

Resources: Yihui Zhang, Guihui Wang.

Software: Yihui Zhang, Guihui Wang, Ziwen Qin.

Supervision: Yihui Zhang, Guihui Wang, Jixian Liu.

Validation: Yihui Zhang, Guihui Wang, Ziyan Mo.

Visualization: Yihui Zhang, Guihui Wang, Ziwen Qin.

Funding acquisition: Jixian Liu.

Writing – original draft: Yihui Zhang, Guihui Wang, Ziwen Qin, Ziyan Mo, Jixian Liu.

Writing – review & editing: Yihui Zhang, Guihui Wang, Ziwen Qin, Ziyan Mo, Jixian Liu.