Genomic Profiling Identified a Mutational Signature Related to Female in Nonsmoking Lung Adenocarcinoma

Jinman Zhuang; Yongsheng Yang; Xuezhen Chen; Guanying Zheng; Xiaopei Shen; Fei He; Baosong Xie

doi:10.4143/crt.2025.489

Articles

Page Path: HOME > J Korean Cancer Assoc > Ahead-of print articles > Article

Original Article Genomic Profiling Identified a Mutational Signature Related to Female in Nonsmoking Lung Adenocarcinoma: Jinman Zhuang¹, Yongsheng Yang¹, Xuezhen Chen¹, Guanying Zheng², Xiaopei Shen³, Fei He^1,4,5,6, Baosong Xie²; DOI: https://doi.org/10.4143/crt.2025.489
Published online: September 11, 2025

¹Department of Epidemiology and Health Statistics, School of Public Health, Fujian Medical University, Fuzhou, China

²Department of Pulmonary and Critical Care Medicine, Fujian Provincial Hospital, Fujian Shengli Medical College, Fujian Medical University, Fuzhou, China

³Department of Bioinformatics, Fujian Key Laboratory of Medical Bioinformatics, School of Medical Technology and Engineering, Fujian Medical University, Fuzhou, China

⁴Fujian Provincial Key Laboratory of Tumor Microbiology, Fujian Medical University, Fuzhou, China

⁵Fujian Digital Tumor Data Research Center, Fuzhou, China

⁶Department of Pulmonary and Critical Care Medicine, Respirology Medicine Centre of Fujian Province, The Second Afliated Hospital of Fujian Medical University, Quanzhou, China

Correspondence: Fei He, Department of Epidemiology and Health Statistics, Fujian Provincial Key Laboratory of Environment Factors and Cancer, School of Public Health, Fujian Medical University, Fuzhou 350122, Fujian Province, China
Tel: 86-13400549537 E-mail: i.fei.he@fjmu.edu.cn

Co-correspondence: Baosong Xie, Department of Pulmonary and Critical Care Medicine, Fujian Provincial Hospital, Fujian Shengli Medical College, Fujian Medical University, Fuzhou 350001, Fujian Province, China
Tel: 86-13763861681 E-mail: xbaosong@126.com

• Received: May 7, 2025 • Accepted: September 10, 2025

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

1,998 Views
94 Download

Full Article

Download PDF

Abstract

Purpose
This study aimed to illustrate mutational characteristics in nonsmoking lung adenocarcinoma (LUAD) and to explore its relationship with clinical factors.
Materials and Methods
We included 86 nonsmokers with LUAD and downloaded the clinical information, whole exome sequencing, and RNA sequencing data from cBioPortal and The Cancer Genome Atlas (TCGA) website. NMF algorithm, “SomaticSignatures,” and “DeconstructSigs” were used to reconstruct and infer new mutational signature. The similarity between Catalogue of Somatic Mutations in Cancer (COSMIC) and reconstructed mutational signature was measured by cosine similarity. The enrichment status of signaling pathways was derived by Gene Set Enrichment Analysis. “pRRophetic” package was used to predict the sensitivity of adjuvant drug for cancer treatments.
Results
The most frequent driver genes in nonsmoking LUAD were epidermal growth factor receptor (EGFR) (59% in cBioPortal cohort, 68% in Fujian nonsmoking LUAD cohort). We identified three new mutation signatures of LUAD in nonsmokers and identified S2 as the most enriched signature in these nonsmokers with LUAD. In cBioPortal and Fujian nonsmoking LUAD cohort, the proportion of S2 was higher in females (p < 0.05) and patients with tumor mutational burden (TMB) (p < 0.01). Similar results were observed in nonsmoking TCGA-LUAD cohort (p_female=0.0013, p_TMB < 0.001). OXIDATIVE_PHOSPHORYLATION signaling pathway was most enriched in S2-enriched group (NES=1.63). S2-enriched group had higher mutation rate of EGFR (p=0.003) and more drug sensitivity to metformin (p=0.035).
Conclusion
We identified a new mutational signature (S2) which is most enriched in LUAD in nonsmokers and related to female and low-TMB. S2 mutational signature may help reveal female-related mutagenesis mechanisms and develop strategies for therapeutic of never-smoker lung adenocarcinoma.
Key words: Mutational signature, Non-smokers, Female, Oxidative phosphorylation, Metformin

Introduction

Cigarette smoke is a well-documented cause of cancer [1]. Deaths attributed to smoking-related cancers are projected to increase by 28.7% for men and 35.8% for women by 2044 [2]. The International Agency for Research on Cancer (IARC) has classified lung, oral cavity, gastric, and bladder cancers as smoking-related cancers [3]; among these, lung cancer is the leading cause of cancer-related mortality [4]. However, nonsmokers also account for a relatively large proportion of cases of lung cancer, especially females [5], and most of these cases have lung adenocarcinoma (LUAD) [6]. In East Asia, there has been a significant increase in the incidence of LUAD in nonsmokers [7]; however, the carcinogenic mechanisms in nonsmokers remain poorly understood [8].

Both environmental and genetic factors have been considered as risk factors for LUAD in nonsmokers [9]. Mutation signature analysis is a nascent field that can unravel the environmental and endogenous sources of mutagenesis in the context of specific tumors [10]. All cancers are caused by somatic mutations caused by DNA damage (incorrect repair or unrepaired) or replication errors [11]; these processes of mutations reveal a pattern of tumor mutation signature. Smoking-related mutation signature was found most enriched in lung cancer, especially in lung squamous cell carcinomas (LSCC), and this mutation signature was similar to that of other tobacco-induced cancers [12]. In contrast to LSCC, LUAD is characterized by distinctly different genomic profiles [13,14], and only a fraction of mutations can be explained [15]. Compared to smokers, LUAD in nonsmokers is characterized by early-onset [16] and high prevalence of epidermal growth factor receptor (EGFR) mutations which have been identified as a poor prognostic factor [17]. Previous studies have explored the mutation profiles in LUAD of nonsmokers, mostly focusing on the effect of driver mutations; these studies constructed a new molecular classification to reveal new potential therapeutic targets for LUAD [18,19]. However, few studies have assessed the correlation of mutational signatures with clinical information in the context of LUAD in nonsmokers to explore the possible etiologies. Therefore, better characterization of the unique mutation signature in LUAD in nonsmokers is a key imperative.

Therefore, in this study, we performed whole exome sequencing (WES) and RNA sequencing (RNA-seq) in 86 nonsmoking patients with LUAD. We also downloaded the WES and RNA-seq data of nonsmoking LUAD from cBioPortal, identified the enriched mutation signatures and explored the relationship of mutation signature with clinical information and gene expression. Finally, these associations were also validated in The Cancer Genome Atlas (TCGA) dataset. Our findings revealed a new signature that was enriched in LUAD in nonsmokers and associated with females. This new mutational signature might serve as a predictor of female-related mutagenesis mechanisms and provided the sensitivity of metformin as a concomitant drug in the EGFR tyrosine-kinase inhibitors (TKIs) of nonsmoking LUAD.

Materials and Methods

1. Study population recruitment, tissue sample sequencing, and public data acquisition

Nonsmoking patients with LUAD were recruited from the Affiliated Union Hospital of Fujian Medical University and First Affiliated Hospital of Fujian Medical University. Tumor and adjacent normal tissues (3-5 cm) were harvested by pathologists following surgical resection. All tumor tissues were reviewed by pathologists to confirm that the tumor cells accounted for 70% of the tissue submitted for WES and RNA-seq. Adjacent normal lung tissue was used as a matched normal control. DNA was extracted by Magnetic Universal Genomic DNA Kit; RNA was extracted using Trizol. For WES, library building and capture were performed using Agilent SureSelect Human All Exon V6 (Agilent Technologies). The Illumina Novaseq PE150 test was performed according to the effective concentration and data volume requirements of the library sequence. In order to obtain high-quality clean reads, Fastp was used to filter the raw data using the following criteria for removal: (1) reads containing Adapter; (2) reads in which N (N indicates that the base information cannot be determined) in a single-ended read has > 10% of the total number of bases in the read; (3) reads with low quality (quality value < 5) that have more than 50% of the read length ratio in a single-ended read. For RNA-seq, total RNA was used as input material for the RNA sample preparations. Sequencing libraries were generated using NEBNext Ultra RNA Library Prep Kit for Illumina (NEB); mRNA was purified from total RNA using poly-T oligo-attached magnetic beads.

We utilized WES, RNA-seq, and clinical dataset from two cohorts. Data pertaining to a total of 187 nonsmoking LUAD patients were downloaded from cBioPortal (https://www.cbioportal.org/); data source was the genomic landscape of LUAD study by Chen et al. [20]. While 494 patients with LUAD (70 nonsmokers and 424 smokers) and some smoking-related cancers (including lung squamous cell carcinoma [LUSC], bladder urothelial carcinoma [BLCA], cervical squamous cell carcinoma and endocervical adenocarcinoma [CESC], esophageal carcinoma [ESCA], head and neck squamous cell carcinoma [HNSC], kidney chromophobe [KICH], kidney renal clear cell carcinoma [KIRC], kidney renal papillary cell carcinoma [KIRP], and pancreatic adenocarcinoma [PAAD]) with smoking condition were found in TCGA; data were downloaded from the Genomic Data Commons (GDC) data portal (https://portal.gdc.cancer.gov/).

2. Identification of somatic mutations

Burrows-Wheeler Aligner [21] was used to identify somatic single-nucleotide variants and the sequence reads passing the quality assessment were mapped to the human reference genome GRCh38 (hg 38). SAMtools was applied to transform sequence alignment and map (SAM) files into binary alignment and map (BAM) files and polymerase chain reaction duplicates were first removed and subsequently realigned and recalibrated by GATK (MarkDuplicatesSpark). Using GATK (BaseRecalibrator) for base quality score recalibration; finally, BAM with adjusted base quality was obtained by GATK (ApplyBQSR). The sequencing coverage and average sequencing depth of BAM files were calculated by BAMDST software; Mutect2 was used to obtain the mutation gene of every sample and variant call format (vcf) files were generated. Finally, vcf2maf was used to transfer vcf files to mutation annotation format (maf) files. For all maf files, mutation landscape was evaluated by R package “maftools” and presented as a waterfall figure.

3. Transcriptome analysis

The gene expression profiles for the 43 nonsmoking patients with LUAD was obtained using the RNA-seq technique. Firstly, fastqc was applied to evaluate the quality of raw data and fastp was used to filter the raw data. Secondly, according to human reference genome GRCh38 (hg 38), fastq files were annotated by hisat2 software and transformed into BAM files. Finally, BAM documents were quantified using featureCounts and the expression matrix of counts was obtained.

To render all expression files of nonsmoking LUAD cohort in the same format, the raw read counts for 86 nonsmoking cohort and nonsmoking TCGA-LUAD were normalized using DESeq2 (v.1.16.1) followed by log transformation (adding one pseudo-count and log2 transformation) consistent with that of the cBioportal dataset. Finally, ComBat function in “sva” R package was used to batch correction in cohorts with different experimental conditions.

4. Non-negative matrix clustering and reconstructed signatures

Based on the number of mutations in each of the 96 possible trinucleotide contexts, “NMF” package [22] can be used to deconvolute single-base substitution signatures according to Catalogue of Somatic Mutations in Cancer (COSMIC, https://cancer.sanger.ac.uk). The estimated rank (expected mutational signature numbers) was determined by the residual sum of squares, which reaches a lower plateau. Then we constructed mutational signatures according to the rank by NMF algorithm and output of the reconstructed signature was generated as a Rata using R package “SomaticSignatures” [23].

5. Scale inference based on the reconstructed mutational signature

After uploading the reconstructed mutational signatures Rata, “DeconstructSigs” [24] R package was used to infer the proportion of new mutational signatures in nonsmoking LUAD cohort of 86 patients and all TCGA somatic mutation dataset.

6. Comprehensive analysis of molecular characteristic and drug sensitivity prediction in S2 subgroups

Gene Set Enrichment Analysis (GSEA) and differential expression analysis were performed on all genes between the different S2 subgroup all samples. Signaling pathways were determined based on the hallmark gene sets and Kyoto Encyclopedia of Genes and Genomes (KEGG). pRRophetic package [25] was used to predict drug sensitivity in metformin.

7. Statistic analysis

Statistical analysis was performed using R 4.2.0 (R Foundation for Statistical Computing). The normality of continuous variables was assessed using Shapiro-Wilk test. For non-normally distributed data, Kruskal-Wallis test was used for three-group comparisons and the Wilcoxon rank-sum test for two-group comparisons. Multiple testing correction was applied using the Benjamini-Hochberg method where appropriate. Effect sizes and 95% confidence intervals were calculated for all comparisons.

Results

1. Dataset of nonsmoking LUAD patients

In total, our study included 86 nonsmoking LUAD patients from Fujian province and 187 nonsmokers with LUAD from the cBioPortal dataset. Most tumors were at early stage (89.5% of Fujian nonsmoking LUAD cohort and 46.5% of nonsmoking LUAD patients from cBioPortal were at stage I), 87.2% of Fujian nonsmoking LUAD cohort and 70.1% of cBioPortal nonsmoking LUAD patients were female (Table 1). In order to explore the differences of mutational signatures between different smoke status and validate the observed relationship between clinical features and mutational signature contribution, we also downloaded the gene mutation and expression data of TCGA-LUAD including 70 nonsmokers and 424 smokers. To assess the contribution of our reconstructed mutational signatures in other cancers not related to smoking, we also downloaded the dataset of smoking-related cancers (including LUSC, BLCA, CESC, ESCA, HNSC, KICH, KIRC, KIRP, and PAAD) for which the information about smoking status (nonsmoker) was obtained from the TCGA website.

2. Genomic landscape of nonsmoking LUAD

In the cBioPortal nonsmoking LUAD cohort, somatic mutation of 20 top genes were identified in all patients. The most frequent driver genes were EGFR (59%), TP53 (28%), and TTN (15%). Missense_Mutation accounted for the majority of somatic mutations, and the most single-base substitution was C>T (Fig. 1A).

Similar to the cBioPortal dataset, mutations of EGFR (68%), TP53 (18%), and RBM10 (11%) genes were most frequently found in Fujian nonsmoking LUAD cohort. Missense_Mutation and C>T single-base substitution accounted for the majority of mutations (Fig. 1B).

3. Mutation signatures of nonsmoking LUAD

We investigated COSMIC mutation signatures in nonsmokers with LUAD and found that the COSMIC signature 1 (clock-like signature) was the most common in the two nonsmoking LUAD cohorts of our study (Fig. 2A and B).

To better understand the mutation signatures of nonsmoking LUAD, we used NMF algorithm (NMF package of R software) to reconstructed the mutational signatures. Mutational signatures were first discovered using the cBioPortal cohort as the discovery dataset. The discovered signatures were then independently validated in the Fujian cohort and TCGA cohort using DeconstructSigs package. Option “brunet” was selected in our study and 50 iterations were performed to identify the optimal cluster numbers; the number of mutation signature clusters were set from 2 to 10. Using the parameter “cophenetic”, “dispersion” and “silhouette”, the optimal cluster number was determined to be 3 (as the magnitude of the cophenetic and dispersion begin to fall in k=3) (S1 Fig.). After determining the optimal cluster number, we used “DeconstructSigs” R package to construct three new mutational signatures (S1, S2, and S3) (Fig. 3). Subsequently, “cos_sim_matrix” algorithm from “MutationalPatterns” R package was used to explore the cosine similarity between S1, S2, S3 and the mutational signatures in COSMIC and exogenous exposures. The results illustrate that signature S1 was most similar to COSMIC-2 (cosine similarity [CS]=0.877), signature S2 was most similar to COSMIC-1 (CS=0.892), while signature S3 was most similar to COSMIC-4 (CS=0.869) and BaP..0.39.uM....S9 (CS=0.868) (Table 2, Fig. 4A and B). The proportions of somatic mutations in the 3 mutation signatures in individual patients in the cBioPortal nonsmoking LUAD cohort and Fujian nonsmoking LUAD cohort showed that S2 accounted for the largest proportion (Fig. 4C).

4. Enrichment of three new mutation signatures in TCGA-LUAD cohort and other cancers

We explored the mutation landscape of both smoking-TCGA-LUAD cohort and nonsmoking-TCGA-LUAD cohort. The results showed that the most significant mutation in the nonsmoking TCGA-LUAD cohort was EGFR, while the most significant mutation of smoking TCGA-LUAD cohort was TTN (Fig. 5A and B). Next, we calculated the enrichment of these three new mutation signatures in the TCGA-LUAD cohort. S1 was found to be most enriched in smoking TCGA-LUAD cohort, while S2 was most enriched in nonsmoking TCGA-LUAD cohort (Fig. 5C and D).

Next, we also explored the mutation landscape of other cancers in nonsmokers based on data downloaded from the TCGA website. Mutation landscape of nonsmoking LUSC, BLCA, CESC, ESCA, HNSC, KICH, KIRC, KIRP, and PAAD are shown in S2 Fig. Then, we calculated the enrichment of 3 mutational signatures in these nonsmoking cancers. Almost all nonsmoking cancers showed the largest amount of S2 enrichment, except BLCA and CESC (S3A and S3B Fig.).

5. Correlation of clinical characteristics and mutational signatures

Next, we assessed the correlation between clinical characteristics and proportion of different mutational signatures. We combined our Fujian nonsmoking LUAD cohort and cBioPortal nonsmoking LUAD cohort and explored the correlations between mutational signature and clinical characteristics. The proportion of S2 was significantly higher in females (p < 0.05). A potential relationship was observed between stage and the proportion of S1 and S2 (p < 0.001). Patients with tumor mutational burden (TMB) ≥ 1.6 had a significantly higher proportion of S1 (p < 0.001), while patients with TMB < 1.6 had a higher proportion of S2 (p < 0.01) (Fig. 6).

Next, we verified the correlation between clinical characteristics and mutational signatures in the nonsmoking TCGA-LUAD cohort. S2 was found to have a higher proportion in nonsmoking LUAD female patients (p=0.001) and patients with lower TMB (p < 0.001) (Fig. 7).

As the S2 mutational signature was most enriched in nonsmoking LUAD cohort, it is critical to further explore the relationship between S2 signature and gender. Here, we used multivariate logistic regression analysis to further determine the relationship between the S2 mutational signature and gender. In the logistic regression analysis, cases with a relative contribution of S2 ≤ 0.5 were defined as having no S2 mutational signature and were coded as 0. For cases with a relative contribution of S2 > 0.5, they were defined as having S2 mutational signature and were coded as 1. In this study, the age, stage, and TMB were included as adjustment variables in the multivariate logistic regression analysis. The results showed that S2 still related to female (odds ratio, 1.881; 95% confidence interval, 1.013 to 3.494; p=0.045) (S4 Table).

6. Molecular characteristics and drug sensitivity of different S2-subgroups

In order to explored the differences of potential mechanism in S2-subgroups, we combined all transcriptome matrix of 3 cohort. Principal component analysis plots before and after transcriptome matrix batch correction were shown in S5 Fig.

As the S2 mutational signature was most enriched and related to female, it may be able to indicate the potential causes of LUAD in nonsmoking female. GSEA was performed to explore the differences in signaling pathways between the two S2-subgroups. The gene sets of the S2-enriched subgroup were most enriched in HALLMARK_OXIDATIVE_PHOSPHORYLATION (Normalized Enrichment Score=1.63, p < 0.05) (Fig. 8A). Besides, oxidative phosphorylation also found enriched in KEGG analysis of differentially expressed genes between S2-subgroups (Fig. 8B). Next, we explored the gene mutations to gain further biological insight into S2-subgroups. The EGFR mutation was higher in the S2-enriched group (Fig. 8C).

Finally, we used pRRophetic to predict the drug sensitivity of metformin, which can be used as an adjuvant medication for EGFR-TKIs. The IC₅₀ of metformin was lower in S2-enriched group, indicating that the drug sensitivity of metformin is relatively strong in S2-enriched group (p=0.035) (Fig. 8D).

Discussion

Owing to the relatively large proportion of nonsmokers with LUAD, identification of the potential etiology of cancers in nonsmokers and its association with mutational signature is of much clinical significance. This study provides a comprehensive characterization of nonsmoking LUAD genomes. EGFR was found to be the most significantly mutated gene in nonsmoking LUAD. In addition, new mutational signature clusters of nonsmoking LUAD were exhibited in our present study, including S1, S2, and S3. S2 accounted for the largest proportion of these mutation signatures in nonsmoking LUAD, and our results showed that S2 may be associated with sex and TMB. Furthermore, we also observed that oxidative phosphorylation is enriched in S2-enriched group. In addition, S2-enriched group showed sensitivity to metformin.

EGFR is known as the most frequent activating mutation gene in nonsmoking LUAD, especially, in Asian patients [26]. Our study focused on LUAD in nonsmokers in both Asian and European cohort. The findings showed that in addition to the Asian nonsmoking patients a high mutation in EGFR was also observed in European nonsmokers from the TCGA-LUAD cohort. Similar to previous studies, nonsmoking Asian patients have higher EGFR mutation rate than European patients. Although there are some interracial differences with respect to the mutation landscape of LUAD [27], nonsmokers with LUAD have the same significant mutation genes; TP53 and TTN also showed high mutation rate in nonsmoking LUAD patients in these cohorts.

COSMIC Signature 4 (SBS4, smoking-related mutation signature) is most commonly found in non–small cell lung cancer (NSCLC) because of its association with smoking [28]. While for lung cancer in nonsmokers, previous study showed that COSMIC Signature 1 (SBS1, clock-related mutation signature) is the most enriched signature [18]. In our study, we also compared mutational signature to COSMIC database, and the results were similar to those of previous studies. These explorations exhibited that age is a risk factor for LUAD in nonsmokers. To explore new mutational signature related to the causation of LUAD in nonsmokers, we further identified 3 new mutation clusters in nonsmokers with LUAD. The proportion of S2 signature was the largest in LUAD with nonsmokers in our study, indicating that the increased risk of LUAD in nonsmokers may be related to this mutation signature. S2 signature showed the greatest similarity to SBS1; however, we found that S2 signature may be associated with female sex and low-TMB instead of the patients’ age in our study.

In the past, the difference of lung cancer incidence between males and females was attributed to the higher prevalence of smoking habit in males [29,30]. However, with increasing incidence of lung cancer in nonsmokers, the differential risk of lung cancer between males and females could not be fully explained by smoking [5]. Previous studies have identified that the sexual dimorphism in genes enhances the susceptibility to lung cancer [31,32]. In our current study, the largest mutational signature enrichment was found to be related to female sex; this further indicated that genetic factors play an important part in the causation of LUAD in females. Moreover, low-TMB also showed an association with S2 signature. Pan-cancer study has illustrated that male cancer patients have higher TMB [33], which is also consistent with our findings. The relationship between smoking status and TMB has been explored in NSCLC patients treated with immune checkpoint inhibitor therapy; the results showed a positive correlation of TMB with smoking in LUAD patients [34]. Another study found a similar association of TMB-high with smoking history [35]. These results further confirmed the molecular characteristics related with the S2 signature which we found most enriched in nonsmoking LUAD patients.

Oxidative phosphorylation is the process in the respiratory chain where a proton gradient generated through redox reactions drives the conversion of ADP to ATP [36]. Previous studies have found that oxidative phosphorylation is essential for tumor initiation and progression [37-41]. Regardless of genetic background, oxidative phosphorylation was supportive for lung cancer development in lung cancer [42]. In our study, oxidative phosphorylation was the main characteristic of differences in the S2 mutational signature. What’s more, S2 mutational signature was enriched in females. These results may suggest potential causes for lung cancer in nonsmoking women. For endogenous factors, somatic mitochondrial DNA (mtDNA) mutations can increase oxidative phosphorylation defects, which contributes to accelerated tumorigenesis [43]. The maternal inheritance characteristic of mtDNA might explain some of reasons for oxidative phosphorylation and nonsmoking females with LUAD [44]. In female body, during the metabolism of endogenous hormones such as estrogen, reactive oxygen species may be produced and interfere with the oxidative phosphorylation process [45]. For exogenous environmental exposure factors, include SO₂, NO₂, PM_2.5, and BaP, cause adverse health outcomes by effects the active site complex for oxidative phosphorylation [46]. Exposure to cooking oil fume was found associate with risk of lung cancer [47]. Cooking oil fume is a common source of harmful environmental exposure for female, contains a large number of substances that can trigger oxidative phosphorylation. Our exploration also provided a reference for the study of risk factors for lung cancer in nonsmoking females.

As the high mutation rate of EGFR in nonsmoking LUAD, EGFR-TKIs has applied in the treatment paradigm for patients [48]. According to previous studies, EGFR-TKIs drugs have shown remarkable therapeutic effects on EGFR mutant NSCLC. However, the problem of acquired resistance to EGFR-TKIs is a challenge in the treatment of patients [49,50]. Metformin is derived from goat beans and is a type of biguanide derivative that has the effects of inhibiting hepatic gluconeogenesis and promoting peripheral glucose uptake [51]. Studies found that metformin not only can control the symptoms of diabetes, but also can reduce the incidence of malignant tumors in patients with type 2 diabetes mellitus, and can effectively improve the prognosis of cancer patients [52]. Clinical experimental studies have revealed that the combination therapy of metformin and the EGFR-TKIs for NSCLC shows a synergistic effect, which can significantly increase the median OS of patients [53,54]. What’s more, metformin can be used as oxidative phosphorylation inhibitors to target cancer subtypes in which oxidative phosphorylation is upregulated and to alleviate therapeutically adverse tumor hypoxia [55]. In our study, S2-enriched group was found have higher proportion of EGFR mutations and higher drug sensitivity of metformin. This result indicates that metformin can be used as a medication for treating resistance to EGFR-TKIs in the nonsmoking LUAD with S2-enriched. Besides, the results about upregulation of oxidative phosphorylation in S2-enriched group in our exploration also suggests that metformin may be able to serve as an adjunctive medication for the S2-enriched group.

The relationship between mutational signature and sex is not well characterized in the contemporary literature. Although we identified a mutational signature that was most enriched in nonsmoking LUAD and which may be related to female, some limitations of our study should be acknowledged. First, most of the patients-related data were downloaded from a public database; detailed information about the risk factors for LUAD in nonsmokers (secondhand smoke exposure, occupational exposure, oil fumes exposure, menopausal status, hormone therapy use, and tumor purity) was missing in these datasets. Therefore, the interaction of the mutational signature with these exposures could not be explored. However, our study provided an insight that S2 signature may provide references for the endogenous and exogenous factors contributing to lung cancer in nonsmoking female. Secondly, the new mutational signature was developed from the cBioPortal dataset, which exclusively comprised of Asian patients; therefore, we could not include other races to re-cluster the mutational signature in nonsmoking LUAD patients and generalizability to broader populations. However, the new mutation signatures were also suitable for patients in the TCGA cohort and similar results were found in our findings. Third, the study lacks functional validation of the proposed mechanisms linking S2 signature to oxidative phosphorylation in female patients. Fourth, we cannot completely rule out the influence of unmeasured confounders such as environmental exposures or genetic background on the observed associations.

In summary, we re-clustered the mutational landscape in nonsmokers with LUAD and identified a mutation signature that may be related to female and low-TMB. This new mutational signature indicated that oxidative phosphorylation may contributing to lung cancer in nonsmoking female and may serves as a marker of treatment management in nonsmoking LUAD. Our study provides insights into the potential relation of molecular characteristics in the context of LUAD in nonsmokers.

Electronic Supplementary Material

Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).

crt-2025-489_S1_Fig.pdf

crt-2025-489_S2_Fig.pdf

crt-2025-489_S3_Fig.pdf

crt-2025-489_S4_Table.pdf

crt-2025-489_S5_Fig.pdf

NOTES

Ethical Statement

This study was approved by the Ethics Committee of the Fujian Medical University. All participants were informed and signed informed consent forms. The study was conducted in accordance with the Declaration of Helsinki.

Author Contributions

Conceived and designed the analysis: Shen X, He F, Xie B.

Collected the data: Yang Y, Chen X, Zheng G, Xie B.

Contributed data or analysis tools: Zhuang J, Shen X, He F.

Performed the analysis: Zhuang J.

Wrote the paper: Zhuang J, He F.

Conflicts of Interest

Conflict of interest relevant to this article was not reported.

Funding

This work was supported by National Natural Science Foundation of Fujian province [grant number: 2023J01093], Innovation of science and technology of Fujian province [grant number: 2020Y9028] and Health Commission’s Program for Cultivating Middle-aged and Young Backbone Talents of Fujian province [grant number: 2023GGA002].

Fig. 1.

The mutation landscape of nonsmoking lung adenocarcinoma (LUAD) patients (A) cBioPortal dataset (B) Fujian nonsmoking LUAD cohort dataset. DEL, deletion; DNP, dinucleotide polymorphism; IGR, intergenic region; INS, insertion; SNP, single nucleotide polymorphism; SNV, single nucleotide variant; UTR, untranslated region.

Fig. 2.

The enrichment of COSMIC mutational signature in nonsmoking lung adenocarcinoma (LUAD) patients. (A) cBioPortal dataset. (B) Fujian nonsmoking LUAD cohort dataset.

Fig. 3.

Three de novo mutation signatures detected in nonsmoking lung adenocarcinoma of cBioPortal dataset.

Fig. 4.

Associate of three de novo mutation signatures with COSMIC signature and enrichment of 3 mutation signatures in nonsmoking lung adenocarcinoma (LUAD) patients. (A) Cosine similarity between three de novo mutation signatures and COSMIC signature. (B) Cosine similarity between three de novo mutation signatures and environmental agents signature. (C) The enrichment of three mutational signature in cBioPortal dataset and Fujian nonsmoking LUAD cohort.

Fig. 5.

Mutational files in The Cancer Genome Atlas (TCGA) cohort. (A) The mutational landscape of smoking lung adenocarcinoma (LUAD) patients in TCGA dataset. (B) The mutational landscape of nonsmoking LUAD patients in TCGA dataset. (C) The enrichment of three mutational signature in smoking LUAD patients of TCGA dataset. (D) The enrichment of three mutational signature in nonsmoking LUAD patients of TCGA dataset.

Fig. 6.

Relationship between age (A), sex (B), stage (C), tumor mutational burden (TMB) (D) and the enrichment of three de novo mutation signatures in cBioPortal dataset combine Fujian nonsmoking lung adenocarcinoma cohort.

Fig. 7.

Relationship between age, sex, stage, tumor mutational burden (TMB) and the enrichment of three de novo mutation signatures S1 (A), S2 (B), and S3 (C) in nonsmoking The Cancer Genome Atlas lung adenocarcinoma (TCGA-LUAD) dataset.

Fig. 8.

Molecular characteristics and drug sensitivity of different S2-subgroups. (A) Gene Set Enrichment Analysis analysis of S2-enriched subgroup. (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of different expression gene set between S2-subgroups. (C) Relationship between EGFR mutation and S2-subgroups. (D) Prediction of drug sensitivity on metformin of S2-subgroups.

Table 1.

Clinical characteristic of nonsmoking patients with lung adenocarcinoma in our study

Clinical parameter	BGI+GSK cohort	Fujian nonsmoking cohort
Age (yr)
< 62	81 (43.3)	43 (50.0)
≥ 62	106 (56.7)	43 (50.0)
Sex
Male	56 (29.9)	11 (12.8)
Female	131 (70.1)	75 (87.2)
Stage
I	87 (46.5)	77 (89.5)
II	36 (19.3)	5 (5.8)
III	54 (28.9)	4 (4.7)
IV	10 (5.3)	0
TMB
< 1.6	93 (49.7)	45 (52.3)
≥ 1.6	94 (50.3)	41 (47.7)
EGFR
Wild	74 (39.6)	27 (31.4)
Mutation	113 (60.4)	59 (68.6)

Values are presented as number (%). EGFR, epidermal growth factor receptor; TMB, tumor mutational burden.

Table 2.

Characteristics of reconstructed mutational signatures (S1-S3)

Mutational signature	Characteristic	Reference signature	Cosine similarity	Description
S1	High proportion in C>G and C>T;only in T[C>G]A,T[C>G]T,T[C>T]A,T[C>T]T	COSMIC-2	0.877	Activity of APOBEC family of cytidine deaminases
S1		SSR 1.25 J	0.489	Radiation-like
S2	High proportion in C>T	COSMIC-1	0.892	Spontaneous deamination of 5-methylcytosine
S2		DMH 11.6 mM S9	0.524	Alkylating Agents-like
S3	High proportion in C>A	COSMIC-4	0.869	Tobacco smoking
S3		BaP 0.39 μM S9	0.868	PAHs

REFERENCES

1. Khariwala SS, Ma B, Ruszczak C, Carmella SG, Lindgren B, Hatsukami DK, et al. High level of tobacco carcinogen-derived DNA damage in oral cells is an independent predictor of oral/head and neck cancer risk in smokers. Cancer Prev Res (Phila). 2017;10:507–13. Article PubMed PMC PDF
2. Luo Q, Steinberg J, Yu XQ, Weber M, Caruana M, Yap S, et al. Projections of smoking-related cancer mortality in Australia to 2044. J Epidemiol Community Health. 2022;76:792–9. Article PubMed PMC
3. Suwazono Y, Nogawa K, Uetani M, Miura K, Sakata K, Okayama A, et al. Application of hybrid approach for estimating the benchmark dose of urinary cadmium for adverse renal effects in the general population of Japan. J Appl Toxicol. 2011;31:89–93. Article PubMed
4. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68:394–424. Article PubMed PDF
5. Jemal A, Miller KD, Ma J, Siegel RL, Fedewa SA, Islami F, et al. Higher lung cancer incidence in young women than young men in the United States. N Engl J Med. 2018;378:1999–2009. Article PubMed PMC
6. Luo W, Tian P, Wang Y, Xu H, Chen L, Tang C, et al. Characteristics of genomic alterations of lung adenocarcinoma in young never-smokers. Int J Cancer. 2018;143:1696–705. Article PubMed PMC PDF
7. Zang R, Shi JF, Lerut TE, Wang L, Liu CC, Brunelli A, et al. Ten-year trends of clinicopathologic features and surgical treatment of lung cancer in China. Ann Thorac Surg. 2020;109:389–95. Article PubMed
8. Gazdar AF, Zhou C. Lung cancer in never-smokers: a different disease. In: Pass HI, Ball D, Scagliotti GV, editors. IASLC thoracic oncology. 2nd ed. Elsevier; 2018. p. 23–9.
9. Choi JR, Park SY, Noh OK, Koh YW, Kang DR. Gene mutation discovery research of non-smoking lung cancer patients due to indoor radon exposure. Ann Occup Environ Med. 2016;28:13.Article PubMed PMC
10. Koh G, Degasperi A, Zou X, Momen S, Nik-Zainal S. Mutational signatures: emerging concepts, caveats and clinical applications. Nat Rev Cancer. 2021;21:619–37. Article PubMed PDF
11. Martincorena I, Campbell PJ. Somatic mutation in cancer and normal cells. Science. 2015;349:1483–9. Article PubMed
12. Alexandrov LB, Nik-Zainal S, Wedge DC, Aparicio SA, Behjati S, Biankin AV, et al. Signatures of mutational processes in human cancer. Nature. 2013;500:415–21. PubMed PMC
13. Cai X, Chen Z, Deng M, Li Z, Wu Q, Wei J, et al. Unique genomic features and prognostic value of COSMIC mutational signature 4 in lung adenocarcinoma and lung squamous cell carcinoma. Ann Transl Med. 2020;8:1176.Article PubMed PMC
14. Campbell JD, Alexandrov A, Kim J, Wala J, Berger AH, Pedamallu CS, et al. Distinct patterns of somatic genome alterations in lung adenocarcinomas and squamous cell carcinomas. Nat Genet. 2016;48:607–16. Article PubMed PMC PDF
15. Cancer Genome Atlas Research Network. Comprehensive molecular profiling of lung adenocarcinoma. Nature. 2014;511:543–50. Article PubMed PMC PDF
16. Kawaguchi T, Matsumura A, Fukai S, Tamura A, Saito R, Zell JA, et al. Japanese ethnicity compared with Caucasian ethnicity and never-smoking status are independent favorable prognostic factors for overall survival in non-small cell lung cancer: a collaborative epidemiologic study of the National Hospital Organization Study Group for Lung Cancer (NHSGLC) in Japan and a Southern California Regional Cancer Registry databases. J Thorac Oncol. 2010;5:1001–10. Article PubMed
17. Kelly WJ, Shah NJ, Subramaniam DS. Management of brain metastases in epidermal growth factor receptor mutant non-small-cell lung cancer. Front Oncol. 2018;8:208.Article PubMed PMC
18. Devarakonda S, Li Y, Martins Rodrigues F, Sankararaman S, Kadara H, Goparaju C, et al. Genomic profiling of lung adenocarcinoma in never-smokers. J Clin Oncol. 2021;39:3747–58. Article PubMed PMC
19. Chen YJ, Roumeliotis TI, Chang YH, Chen CT, Han CL, Lin MH, et al. Proteogenomics of non-smoking lung cancer in East Asia delineates molecular signatures of pathogenesis and progression. Cell. 2020;182:226–44. Article PubMed
20. Chen J, Yang H, Teo AS, Amer LB, Sherbaf FG, Tan CQ, et al. Genomic landscape of lung adenocarcinoma in East Asians. Nat Genet. 2020;52:177–86. PubMed
21. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25:1754–60. Article PubMed PMC PDF
22. Gaujoux R, Seoighe C. A flexible R package for nonnegative matrix factorization. BMC Bioinformatics. 2010;11:367.Article PubMed PMC PDF
23. Gehring JS, Fischer B, Lawrence M, Huber W. SomaticSignatures: inferring mutational signatures from single-nucleotide variants. Bioinformatics. 2015;31:3673–5. Article PubMed PMC PDF
24. Rosenthal R, McGranahan N, Herrero J, Taylor BS, Swanton C. DeconstructSigs: delineating mutational processes in single tumors distinguishes DNA repair deficiencies and patterns of carcinoma evolution. Genome Biol. 2016;17:31.Article PubMed PMC PDF
25. Geeleher P, Cox N, Huang RS. pRRophetic: an R package for prediction of clinical chemotherapeutic response from tumor gene expression levels. PLoS One. 2014;9:e107468Article PubMed PMC
26. Shi Y, Au JS, Thongprasert S, Srinivasan S, Tsai CM, Khoa MT, et al. A prospective, molecular epidemiology study of EGFR mutations in Asian patients with advanced non-small-cell lung cancer of adenocarcinoma histology (PIONEER). J Thorac Oncol. 2014;9:154–62. Article PubMed PMC
27. Shigematsu H, Lin L, Takahashi T, Nomura M, Suzuki M, Wistuba II, et al. Clinical and biological features associated with epidermal growth factor receptor gene mutations in lung cancers. J Natl Cancer Inst. 2005;97:339–46. PubMed
28. van den Heuvel GR, Kroeze LI, Ligtenberg MJ, Grunberg K, Jansen EA, von Rhein D, et al. Mutational signature analysis in non-small cell lung cancer patients with a high tumor mutational burden. Respir Res. 2021;22:302.PubMed PMC
29. Hammond EC. Smoking in relation to the death rates of one million men and women. Natl Cancer Inst Monogr. 1966;19:127–204. PubMed
30. Doll R, Gray R, Hafner B, Peto R. Mortality in relation to smoking: 22 years’ observations on female British doctors. Br Med J. 1980;280:967–71. Article PubMed PMC
31. Lan Q, Hsiung CA, Matsuo K, Hong YC, Seow A, Wang Z, et al. Genome-wide association analysis identifies new lung cancer susceptibility loci in never-smoking women in Asia. Nat Genet. 2012;44:1330–5. PubMed PMC
32. Zang EA, Wynder EL. Differences in lung cancer risk between men and women: examination of the evidence. J Natl Cancer Inst. 1996;88:183–92. Article PubMed
33. Han J, Yang Y, Li X, Wu J, Sheng Y, Qiu J, et al. Pan-cancer analysis reveals sex-specific signatures in the tumor microenvironment. Mol Oncol. 2022;16:2153–73. Article PubMed PMC PDF
34. Sun LY, Cen WJ, Tang WT, Long YK, Yang XH, Ji XM, et al. Smoking status combined with tumor mutational burden as a prognosis predictor for combination immune checkpoint inhibitor therapy in non-small cell lung cancer. Cancer Med. 2021;10:6610–7. Article PubMed PMC PDF
35. Nagahashi M, Sato S, Yuza K, Shimada Y, Ichikawa H, Watanabe S, et al. Common driver mutations and smoking history affect tumor mutation burden in lung adenocarcinoma. J Surg Res. 2018;230:181–5. Article PubMed PMC
36. Wilson DF. Oxidative phosphorylation: regulation and role in cellular and tissue metabolism. J Physiol. 2017;595:7023–38. Article PubMed PMC PDF
37. Frederick M, Skinner HD, Kazi SA, Sikora AG, Sandulache VC. High expression of oxidative phosphorylation genes predicts improved survival in squamous cell carcinomas of the head and neck and lung. Sci Rep. 2020;10:6380.Article PubMed PMC PDF
38. Becherini P, Caffa I, Piacente F, Damonte P, Vellone VG, Passalacqua M, et al. SIRT6 enhances oxidative phosphorylation in breast cancer and promotes mammary tumorigenesis in mice. Cancer Metab. 2021;9:6.Article PubMed PMC PDF
39. Valle S, Alcala S, Martin-Hijano L, Cabezas-Sainz P, Navarro D, Munoz ER, et al. Exploiting oxidative phosphorylation to promote the stem and immunoevasive properties of pancreatic cancer stem cells. Nat Commun. 2020;11:5265.Article PubMed PMC PDF
40. Ponton-Almodovar A, Udumula MP, Khullar V, Rashid F, Rattan R, Bernard JJ, et al. GPT2 mediates glutamine metabolism-driven metabolic alterations in platinum-resistant ovarian cancer cells. Sci Rep. 2025;15:30528.Article PubMed PMC PDF
41. Bonnay F, Veloso A, Steinmann V, Kocher T, Abdusselamoglu MD, Bajaj S, et al. Oxidative metabolism drives immortalization of neural stem cells during tumorigenesis. Cell. 2020;182:1490–507. Article PubMed
42. Rao S, Mondragon L, Pranjic B, Hanada T, Stoll G, Kocher T, et al. AIF-regulated oxidative phosphorylation supports lung cancer development. Cell Res. 2019;29:579–91. Article PubMed PMC PDF
43. Smith AL, Whitehall JC, Bradshaw C, Gay D, Robertson F, Blain AP, et al. Age-associated mitochondrial DNA mutations cause metabolic remodelling that contributes to accelerated intestinal tumorigenesis. Nat Cancer. 2020;1:976–89. Article PubMed PMC PDF
44. Arnadottir ER, Moore KH, Guethmundsdottir VB, Ebenesersdottir SS, Guity K, Jonsson H, et al. The rate and nature of mitochondrial DNA mutations in human pedigrees. Cell. 2024;187:3904–18. Article PubMed
45. Kim HN, Ponte F, Nookaew I, Ucer Ozgurel S, Marques-Carvalho A, Iyer S, et al. Estrogens decrease osteoclast number by attenuating mitochondria oxidative phosphorylation and ATP production in early osteoclast precursors. Sci Rep. 2020;10:11933.Article PubMed PMC PDF
46. McCully KS. Environmental pollution, oxidative stress and thioretinaco ozonide: effects of glyphosate, fluoride and electromagnetic fields on mitochondrial dysfunction in carcinogenesis, atherogenesis and aging. Ann Clin Lab Sci. 2020;50:408–11. PubMed
47. Chen TY, Fang YH, Chen HL, Chang CH, Huang H, Chen YS, et al. Impact of cooking oil fume exposure and fume extractor use on lung cancer risk in non-smoking Han Chinese women. Sci Rep. 2020;10:6774.Article PubMed PMC PDF
48. Zhou F, Guo H, Xia Y, Le X, Tan DS, Ramalingam SS, et al. The changing treatment landscape of EGFR-mutant non-small-cell lung cancer. Nat Rev Clin Oncol. 2025;22:95–116. Article PubMed PDF
49. Laface C, Maselli FM, Santoro AN, Iaia ML, Ambrogio F, Laterza M, et al. The resistance to EGFR-TKIs in non-small cell lung cancer: from molecular mechanisms to clinical application of new therapeutic strategies. Pharmaceutics. 2023;15:1604.Article PubMed PMC
50. Li Y, Mao T, Wang J, Zheng H, Hu Z, Cao P, et al. Toward the next generation EGFR inhibitors: an overview of osimertinib resistance mediated by EGFR mutations in non-small cell lung cancer. Cell Commun Signal. 2023;21:71.Article PubMed PMC PDF
51. Wang G, Lin F, Wan Q, Wu J, Luo M. Mechanisms of action of metformin and its regulatory effect on microRNAs related to angiogenesis. Pharmacol Res. 2021;164:105390.Article PubMed
52. Lv Z, Guo Y. Metformin and its benefits for various diseases. Front Endocrinol (Lausanne). 2020;11:191.Article PubMed PMC
53. Wang H, Yang X, Sun Y, Li Y, Dong Y, Shan D, et al. Synergistic effect of metformin and EGFR-TKI in the treatment of non-small cell lung cancer. Transl Cancer Res. 2020;9:372–81. Article PubMed PMC
54. Arrieta O, Barron F, Padilla MS, Aviles-Salas A, Ramirez-Tirado LA, Arguelles Jimenez MJ, et al. Effect of metformin plus tyrosine kinase inhibitors compared with tyrosine kinase inhibitors alone in patients with epidermal growth factor receptor-mutated lung adenocarcinoma: a phase 2 randomized clinical trial. JAMA Oncol. 2019;5:e192553Article PubMed PMC
55. Ashton TM, McKenna WG, Kunz-Schughart LA, Higgins GS. Oxidative phosphorylation as an emerging target in cancer therapy. Clin Cancer Res. 2018;24:2482–90. Article PubMed PDF

Figure & Data

REFERENCES

Citations

Citations to this article as recorded by

PubReader
ePub Link

Cite

CITE: export Copy Download Format; Close

Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:

RIS — For EndNote, ProCite, RefWorks, and most other reference management software
BibTeX — For JabRef, BibDesk, and other BibTeX-specific software

Include:

Citation for the content below
Citation and abstract for the content below

Genomic Profiling Identified a Mutational Signature Related to Female in Nonsmoking Lung Adenocarcinoma

DOI: https://doi.org/10.4143/crt.2025.489

XML Download

Abstract
Introduction
Materials and Methods
Results
Discussion
Electronic Supplementary Material
NOTES
REFERENCES

Genomic Profiling Identified a Mutational Signature Related to Female in Nonsmoking Lung Adenocarcinoma

Fig. 1. The mutation landscape of nonsmoking lung adenocarcinoma (LUAD) patients (A) cBioPortal dataset (B) Fujian nonsmoking LUAD cohort dataset. DEL, deletion; DNP, dinucleotide polymorphism; IGR, intergenic region; INS, insertion; SNP, single nucleotide polymorphism; SNV, single nucleotide variant; UTR, untranslated region.

Fig. 2. The enrichment of COSMIC mutational signature in nonsmoking lung adenocarcinoma (LUAD) patients. (A) cBioPortal dataset. (B) Fujian nonsmoking LUAD cohort dataset.

Fig. 3. Three de novo mutation signatures detected in nonsmoking lung adenocarcinoma of cBioPortal dataset.

Fig. 4. Associate of three de novo mutation signatures with COSMIC signature and enrichment of 3 mutation signatures in nonsmoking lung adenocarcinoma (LUAD) patients. (A) Cosine similarity between three de novo mutation signatures and COSMIC signature. (B) Cosine similarity between three de novo mutation signatures and environmental agents signature. (C) The enrichment of three mutational signature in cBioPortal dataset and Fujian nonsmoking LUAD cohort.

Fig. 5. Mutational files in The Cancer Genome Atlas (TCGA) cohort. (A) The mutational landscape of smoking lung adenocarcinoma (LUAD) patients in TCGA dataset. (B) The mutational landscape of nonsmoking LUAD patients in TCGA dataset. (C) The enrichment of three mutational signature in smoking LUAD patients of TCGA dataset. (D) The enrichment of three mutational signature in nonsmoking LUAD patients of TCGA dataset.

Fig. 6. Relationship between age (A), sex (B), stage (C), tumor mutational burden (TMB) (D) and the enrichment of three de novo mutation signatures in cBioPortal dataset combine Fujian nonsmoking lung adenocarcinoma cohort.

Fig. 7. Relationship between age, sex, stage, tumor mutational burden (TMB) and the enrichment of three de novo mutation signatures S1 (A), S2 (B), and S3 (C) in nonsmoking The Cancer Genome Atlas lung adenocarcinoma (TCGA-LUAD) dataset.

Fig. 8. Molecular characteristics and drug sensitivity of different S2-subgroups. (A) Gene Set Enrichment Analysis analysis of S2-enriched subgroup. (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of different expression gene set between S2-subgroups. (C) Relationship between EGFR mutation and S2-subgroups. (D) Prediction of drug sensitivity on metformin of S2-subgroups.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Genomic Profiling Identified a Mutational Signature Related to Female in Nonsmoking Lung Adenocarcinoma

Clinical parameter	BGI+GSK cohort	Fujian nonsmoking cohort
Age (yr)
< 62	81 (43.3)	43 (50.0)
≥ 62	106 (56.7)	43 (50.0)
Sex
Male	56 (29.9)	11 (12.8)
Female	131 (70.1)	75 (87.2)
Stage
I	87 (46.5)	77 (89.5)
II	36 (19.3)	5 (5.8)
III	54 (28.9)	4 (4.7)
IV	10 (5.3)	0
TMB
< 1.6	93 (49.7)	45 (52.3)
≥ 1.6	94 (50.3)	41 (47.7)
EGFR
Wild	74 (39.6)	27 (31.4)
Mutation	113 (60.4)	59 (68.6)

Mutational signature	Characteristic	Reference signature	Cosine similarity	Description
S1	High proportion in C>G and C>T;only in T[C>G]A,T[C>G]T,T[C>T]A,T[C>T]T	COSMIC-2	0.877	Activity of APOBEC family of cytidine deaminases
S1		SSR 1.25 J	0.489	Radiation-like
S2	High proportion in C>T	COSMIC-1	0.892	Spontaneous deamination of 5-methylcytosine
S2		DMH 11.6 mM S9	0.524	Alkylating Agents-like
S3	High proportion in C>A	COSMIC-4	0.869	Tobacco smoking
S3		BaP 0.39 μM S9	0.868	PAHs

Table 1. Clinical characteristic of nonsmoking patients with lung adenocarcinoma in our study

Values are presented as number (%). EGFR, epidermal growth factor receptor; TMB, tumor mutational burden.

Table 2. Characteristics of reconstructed mutational signatures (S1-S3)