Deep structural lipidomic profiling reveals C=C positional isomers as potential biomarkers in lung adenocarcinoma tissue

Introduction

Lung cancer remains a leading cause of cancer-related mortality worldwide, posing a significant threat to human health (1,2). Among lung cancer cases, non-small cell lung cancer (NSCLC) accounts for approximately 85% of all cases, with lung adenocarcinoma (LUAD) being the most prevalent histological subtype (3,4). Despite advances in early detection and treatment strategies, the prognosis for LUAD patients remains poor, with a 5-year survival rate below 20% (5,6). Currently, imaging techniques such as high-resolution computed tomography (CT) and low-dose computed tomography (LDCT) have substantially enhanced the detection of early-stage lung cancer (7,8). However, their utility is limited by disadvantages such as false positives and radiation exposure (9-11). Additionally, biomarkers such as exosomes, proteins, and microRNAs are available for early detection and treatment of LUAD, their diagnostic and therapeutic efficacy remains unsatisfactory (12). Therefore, there is an urgent need to identify novel biomarkers and therapeutic targets to improve early detection and patient outcomes.

Lipids are a structurally diverse class of hydrophobic or amphipathic molecules that play essential roles in energy storage, membrane transport, and metabolic regulation (13-15). Recent studies have shown that lipid metabolic reprogramming is a key feature of cancer initiation and progression, enabling tumor cells to adapt to the immune microenvironment and providing favorable conditions for their survival and development (16,17). Lung cancer is one of the most common malignant tumors globally and is closely associated with lipid metabolism (18). Hartmann et al. reported a negative correlation between cholesterol and total cholesterol levels and the overall survival of LUAD patients (19). Single-cell RNA sequencing analysis further revealed widespread dysregulation of lipid metabolic pathways, particularly glycerophospholipid and glycerolipid metabolism, across multiple cell types in early-stage lung cancer tissues (20). The latest lipidomic studies have identified a variety of specific lipid signatures with potential diagnostic value that could serve as biomarkers for the detection of early-stage lung cancer (21).

Lipids exhibit remarkable structural diversity, with over 40,000 distinct molecular species identified to date. Subtle differences in lipid structures significantly influence their biological functions (22). For example, the number of carbon-carbon double bond (C=C) in fatty acid chains determines the degree of unsaturation of lipid molecules, affecting their fluidity, stability, and intermolecular interactions. The presence of C=C in unsaturated fatty acids causes them to bend within the lipid bilayer, enhancing membrane fluidity and consequently influencing cellular membrane elasticity and permeability (23). Moreover, the instability of C=C in lipid molecules makes them highly susceptible to oxidation, leading to the production of harmful substances such as free radicals, which induce oxidative stress and inflammation, processes closely associated with cancer progression (24). Recent studies have reported alterations in the ratios of lipid C=C isomers in metabolic diseases and cancers, suggesting that these structural changes may serve as potential biomarkers for disease diagnosis and progression (25,26). Despite these advances, the role of lipid fine structures, particularly C=C isomerism, in LUAD remains underexplored. To date, no comprehensive analysis of deep structural lipidomics and relative quantification of lipid C=C isomers has been conducted in LUAD.

In this study, we performed deep structural lipidomic analysis on LUAD tissues and normal lung tissues using the Ω Analyzer, a precision structural reactor developed based on photochemical reaction principles (26), in combination with liquid chromatography-mass spectrometry (LC-MS). By focusing on the relative quantification of lipid C=C isomers, we aimed to uncover the metabolic adaptations of cancer cells and identify potential biomarkers for LUAD. Our findings provide new insights into the role of lipid fine structures in cancer biology and highlight the potential of lipidomics in advancing cancer diagnosis and therapy. We present this article in accordance with the STREGA reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-717/rc).

Methods

Patients

This retrospective study collected tissue samples from 17 patients in the Department of Thoracic Surgery, Beijing Chaoyang Hospital. All cases were pathologically confirmed as LUAD. For each patient, tissue samples were obtained from two regions of the surgically resected lung tissue: LUAD tissue (tumor tissue) and non-tumorous normal lung tissue located at least 5 cm away from the tumor margin. Immediately after collection, the samples were snap-frozen in liquid nitrogen and aliquoted for storage at −80 ℃ until analysis. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Beijing Chaoyang Hospital, Capital Medical University (approval No. 2022-ke-502) and individual consent for this retrospective analysis was waived.

Serum sample processing

The serum samples were retrieved from a −80 ℃ freezer and thawed on ice. A 100 mg tissue sample was weighed and diluted with 1 mL of water, followed by the addition of 1 mL of methanol and 2 mL of chloroform. The mixture was vortexed thoroughly for 10 minutes and centrifuged at 12,000 rpm for 12 minutes to collect the lower phase. The upper phase was re-extracted once using the same procedure, and the lower phases obtained from the two extractions were combined. The combined extract was dried under nitrogen gas and reconstituted with 1 mL of methanol. The reconstituted solution was filtered through a 0.22 µm membrane to obtain the lipid extract stock solution.

Instruments and methods

The Ω Analyzer (PURSPEC) was coupled with the Agilent 6546 LC/Q-TOF (Agilent) and 1290 Infinity II UHPLC (Agilent) to establish a structural lipidomics platform. Ω Analyzer Software (OA 4.0) (PURSPEC) was integrated into the MS control system for automated data acquisition and structural identification at multiple levels, ensuring comprehensive lipidomic profiling.

Chromatographic conditions

Sample separation was performed using an Agilent 1290 Infinity II Ultra-High-Performance Liquid Chromatography (UHPLC) system coupled with an ACQUITY UPLC BEH HILIC column (2.1×100 mm, 1.7 µm). The column temperature was maintained at 30 ℃ with a flow rate of 0.35 mL/min. The mobile phase system consisted of mobile phase A (10 mmol ammonium acetate aqueous solution containing 0.2% acetic acid) and mobile phase B (acetonitrile/acetone/isopropanol, 50/48/2, v/v/v). The gradient elution program was set as follows: 90% to 85% B from 0 to 2.4 min, 85% to 80% B from 2.4 to 3.2 min, 80% B from 3.2 to 5.0 min, 80% to 70% B from 5.0 to 5.1 min, 70% B from 5.1 to 6.0 min, 70% to 90% B from 6.0 to 6.1 min, and 90% B from 6.1 to 10.0 min. Throughout the analysis, samples were maintained at 6 ℃ in an autosampler. Ammonium acetate and acetic acid were purchased from Anaqua™ Chemicals (Cleveland, USA). Chloroform was purchased from Yonghua Chemicals (Suzhou, China). HPLC-grade, acetone, acetonitrile, isopropanol, and methanol were purchased from Fisher Scientific (Waltham, Massachusetts, USA).

MS conditions

Electrospray ionization (ESI) was employed in both positive and negative ion modes. The samples were separated using an Agilent 1290 Infinity II UHPLC system and analyzed with an Agilent 6546 Q-TOF mass spectrometer. The ESI source parameters were set as follows: gas temperature, 200 ℃; gas flow, 7 L/min; nebulizer, 30 psi; sheath gas temperature, 350 ℃; V Cap, 3.5 kV (positive), 2.5 kV (negative); MS1 scan ranges, 450–1,000; MS/MS scan ranges: 100–1,000.

Quality control (QC) analysis

The project utilized bovine liver polar lipid extract as the QC sample. Bovine liver polar lipid extracts were purchased from Avanti Polar Lipids, Inc. (Alabaster, AL, USA). A 1 mg/mL stock solution was diluted 10-fold to prepare a 100 mg/L working solution. Instrument stability, experimental reproducibility, and data reliability were evaluated by comparing the total ion chromatograms (TICs) and the relative ratios of lipid C=C position isomers. For UHPLC-MS analysis, one QC sample (100 mg/L) was injected after every five test samples from the N group (normal lung tissue) and T group (LUAD tissue). For UHPLC-Paternò-Büchi (PB)-MS/MS analysis, one QC sample was injected after every 12 test samples. These QC measures ensured the reliability and reproducibility of the analytical process.

Statistical analysis

Data analysis

Data analysis was conducted using Ω software, an automated lipid structure analysis platform developed by PURSPEC Technologies (Suzhou, China). The raw MS data in .d format were converted to a standard format (.mzML) using ProteoWizard (version 3.0). The converted data were processed in Ω software with integrated databases from LIPIDMAPS and in-house predicted lipid databases to facilitate comprehensive lipid identification. Lipid identification at the species level was performed by matching precise m/z and retention time. Fatty acyl fragment ions in the MS/MS spectra were recognized for lipid identification at the molecular species level from LC-MS/MS data. The lipid C=C location information was determined by matching C=C diagnostic ion fragments generated from PB reactions at C=C in the LC-PB-MS/MS data. Quantification analysis at the species level was based on the integrated intensity of chromatographic peaks obtained from LC-MS data. Quantification of lipids at the C=C location level was performed by calculating the relative intensities of diagnostic ions corresponding to C=C isomers in LC-PB-MS/MS data, which were used to determine the proportions of C=C isomers in different samples. Statistical analyses were performed using R (version 4.3.1), and data visualization was conducted with R (version 4.3.1) and Origin (version 2023b).

Fold change (FC) analysis

FC analysis was performed to evaluate the differential expression between LUAD tissues and adjacent normal lung tissues. The results were presented in the form of a volcano plot. The x-axis of the plot represents the log₂-transformed fold change (log₂FC), while the y-axis represents the −log₁₀-transformed P value (−log₁₀P). Each point in the plot corresponds to a lipid molecule. Differential lipids were identified based on the thresholds of FC >1.5 or FC <0.67 with a P value <0.05, and they are highlighted in different colors on the plot. Lipids with FC >1.5 and P value <0.05 were considered upregulated lipids and are marked in red, while lipids with FC <0.67 and P value <0.05 were considered downregulated lipids and are marked in blue. Lipids that did not meet the thresholds for significance are shown in gray, indicating no significant difference.

Principal component analysis (PCA)

PCA is an unsupervised dimensionality reduction technique widely used in multivariate data analysis. It transforms high-dimensional data into a set of orthogonal variables called principal components, which are ordered by the amount of variance they capture. The first principal component explains the largest variance in the data, while subsequent components capture decreasing levels of variance. PCA does not require class labels and is particularly useful for exploring data structure, visualizing patterns, clustering samples, and detecting outliers. In omics research, PCA is frequently applied to reveal variability within lipidomics, metabolomics, and proteomics datasets, providing insights into overall sample distribution and inherent data patterns.

Sparse partial least squares discriminant analysis (sPLS-DA)

sPLS-DA is a supervised classification method that combines dimensionality reduction with feature selection. Unlike PCA, sPLS-DA incorporates class label information to identify variables that maximize group separation, focusing on biological or experimental relevance. By introducing sparsity constraints, sPLS-DA selects a subset of the most relevant variables, improving model interpretability and reducing overfitting. It constructs latent components by optimizing the covariance between predictors and response variables (class labels), ensuring effective classification of samples into predefined groups. This method is particularly useful in omics research for biomarker discovery, disease classification, and subgroup differentiation in high-dimensional datasets.

Advantages of sPLS-DA over PCA

Compared to PCA, sPLS-DA offers several advantages. Firstly, sPLS-DA is a supervised method that explicitly aims to separate predefined groups, whereas PCA is unsupervised and focuses only on explaining variance, which may not align with class separation. Secondly, sPLS-DA incorporates sparsity constraints to select the most relevant variables, enhancing interpretability and reducing noise, while PCA includes all variables in its analysis, potentially diluting the contribution of biologically significant features. Lastly, sPLS-DA is robust to irrelevant variables and noise, enabling more accurate classification and biological interpretation, making it particularly suited for predictive modeling and biomarker discovery in omics research.

Results

Patient characteristics

A total of 17 patients pathologically diagnosed with LUAD were included in this study. The average age of the patients was 64 years, with 7 (41.2%) male patients and 10 (58.8%) female patients. Lymph node metastasis was observed in 3 (17.6%) patients. According to the tumor-node-metastasis (TNM) staging system, 13 (76.5%) patients were classified as stage I, and 4 (23.5%) patients were classified as stage II. Detailed patient information is presented in Table 1.

Analysis of phospholipids isomers in lung tissue

Ω Analyzer and related software were combined with the UPLC-QTOF system for the development of the structural lipidomics platform. Acetone-containing mobile phase was developed as reagents for online PB reactions and separation of different kinds of phospholipids within 6 min. Under positive ion mode, a single 2 µL injection of lipid extract from normal lung tissue was analyzed by UHPLC-QTOF (Figure 1). The chromatogram clearly displays two well-resolved phospholipid peaks within 6 minutes: phosphatidylethanolamine (PE) eluting at 1.9–2.4 minutes and phosphatidylcholine (PC) eluting at 3.6–4.1 minutes. The TIC and mass spectra (Figure 2A,2B) confirmed the identification of PE and PC at the molecular species level. Subsequent negative ion MS/MS analysis (Figure 3A,3B) further enabled molecular-level identification of these lipids. In the negative ion mode, PE(36:2) (m/z 742.5) was identified to contain C18:0 and C18:2 fatty acyl chains, corresponding to PE(18:0_18:2). Similarly, PC(34:1) (m/z 818.6) was determined to contain C16:0 and C18:1 fatty acyl chains, corresponding to PC(16:0_18:1).

Figure 1 The peak-analysis of PE and PC phospholipids were obviously observed in the lipid structure identification chromatogram. PC, phosphatidylcholine; PE, phosphatidylethanolamine.

Figure 2 MS of phospholipid components PE (A) and PC (B) in lipid extracts from tissue samples. MS, mass spectrometry; PC, phosphatidylcholine; PE, phosphatidylethanolamine.

Figure 3 MS/MS spectra of tissue lipid extracts under negative ion mode. (A) MS/MS spectrum of PE(36:2); (B) MS/MS spectrum of PC(34:1). MS, mass spectrometry; PC, phosphatidylcholine; PE, phosphatidylethanolamine.

Further UPLC-PB-MS/MS experiments in the positive ion mode were conducted to analyze the PB reaction products of unsaturated lipids with Ω Analyzer, providing specific diagnostic ions that reveal the positions of C=C double bonds. As shown in Figure 4A, the PB reaction product peak of polyunsaturated lipid PE(18:0_18:2) at m/z 802.5 produced multiple pairs of diagnostic ions—m/z 495.4 and m/z 521.5, as well as m/z 535.5 and m/z 561.5—indicating that the double bonds are located at Δ9 and Δ12. This lipid was thus identified as PE[18:0_18:2(Δ9,Δ12)]. Similarly, Figure 4B shows the PB reaction product peak of PC(16:0_18:1) at m/z 818.6, which exhibited two pairs of diagnostic ions at m/z 650.5 and m/z 676.5, pointing to double bonds located at Δ9 and Δ11. Therefore, PC(16:0_18:1) was identified as containing two positional C=C isomers: PC[16:0_18:1(Δ9)] and PC[16:0_18:1(Δ11)].

Figure 4 PE(18:0_18:2) (A) and PC(16:0_18:1) (B) were used as examples to analyze the double bonds of lipid C=C in lipid isomers. C=C, carbon-carbon double bond; PC, phosphatidylcholine; PE, phosphatidylethanolamine.

The lipid identification results at three structural levels—lipid class level, lipid molecular species level, and lipid C=C positional isomer level—showed that the number of lipids identified in LUAD tissue was slightly higher than that in normal lung tissue (Figure 5). The identification of phospholipids PE and PC at different molecular levels in the N and T groups revealed that the number of lipids identified in phospholipid PE at each level was higher in the T group compared to the N group (Figure 6A). However, for phospholipid PC, the number of lipids identified at each level showed little difference between the N and T groups (Figure 6B). The majority of lipid features were consistently detected in the T groups and the N group at three structural levels, with only a minor subset being uniquely identified in either the T group or the N group (Figure 7). The majority of these uniquely identified features corresponded to low-abundance lipid species, which can be attributed to the limited detection sensitivity of the analytical methodology.

Figure 5 Comparison of the amount of PC and PE identified in group N and group T samples. N, normal lung tissue; T, lung adenocarcinoma tissue. PC, phosphatidylcholine; PE, phosphatidylethanolamine.

Figure 6 Comparison of the amounts of PE (A) and PC (B) identified in group N and group T samples. N, normal lung tissue; T, lung adenocarcinoma tissue. PC, phosphatidylcholine; PE, phosphatidylethanolamine.

Figure 7 Venn diagram illustrating the overlap of PC and PE identified between group N and group T samples. (A) Venn diagram illustrating the overlap of PC and PE at lipid class level identified between group N and group T samples. (B) Venn diagram illustrating the overlap of PC and PE at lipid molecular species level identified between group N and group T samples. (C) Venn diagram illustrating the overlap of PC and PE at C=C positional isomer level identified between group N and group T samples. N, normal lung tissue; T, lung adenocarcinoma tissue. C=C, carbon-carbon double bond; PC, phosphatidylcholine; PE, phosphatidylethanolamine.

Differential analysis at lipid species level

FC analysis (Figure 8) identified 15 differentially expressed lipid molecules between T and N groups, comprising 12 upregulated and 3 downregulated molecules, visualized through a volcano plot. Figure 9 provides detailed information on 15 lipid molecules that showed significant differences between the two groups. Among them, Figure 9A-9C represents lipid molecules that were downregulated in LUAD tissues, while Figure 9D-9O depicts lipid molecules that were upregulated.

Figure 8 Volcano plot analysis of lipid molecular species at the molecular level in group N and T samples. N, normal lung tissue; T, lung adenocarcinoma tissue.

Figure 9 Details of 15 lipid molecules with significant differences between group N and T samples. (A) PC(34:3). (B) PE(42:1). (C) PE(42:2). (D) PC(40:2). (E) PC(42:6). (F) PC(42:7). (G) PC(42:8). (H) PC(44:8). (I) PE(32:1). (J) PE(36:4). (K) PE(38:5). (L) PE(40:3). (M) PE(40:5). (N) PE(40:6). (O) PE(40:7). N, normal lung tissue; T, lung adenocarcinoma tissue. PC, phosphatidylcholine; PE, phosphatidylethanolamine.

To explore the differences in lipid profiles between N and T groups, we employed both PCA and sPLS-DA. While PCA analysis failed to achieve complete separation between the groups at the lipid species level (Figure 10A), sPLS-DA successfully demonstrated clear discrimination (Figure 10B). Component 1 (primary discriminant axis) was predominantly weighted by PC(42:8), with corresponding lipids showing significant upregulation in group T (Figure 11A). Component 2 (secondary axis) exhibited the highest loadings for PE(40:3) and PC(34:2), both elevated in group T (Figure 11B). Loading values (Loading 1/Loading 2) quantified lipid contributions to each component, identifying PC(42:8), PE(40:3), and PC(34:2) as key discriminators. This dimensionality reduction enabled clear visualization of group separation while preserving biological interpretability through lipid-specific weightings.

Figure 10 Comparative analysis of lipid molecules in group N and group T samples using PCA and sPLS-DA methods. (A) PCA analysis method could not distinguish the lipid molecules of group N and group T samples at the species level. (B) sPLS-DA analysis method can effectively distinguish the lipid molecules of group N and group T samples at the species level. N, normal lung tissue; T, lung adenocarcinoma tissue. PCA, principal component analysis; sPLS-DA, sparse partial least squares discriminant analysis.

Figure 11 Loadings plot of the weights of major lipid molecules. (A) Component 1: dominant contribution of PC(42:8) with upregulated expression in group T. (B) Component 2: high-weight lipids PE(40:3) and PC(34:2) showing elevated levels in group T. N, normal lung tissue; T, lung adenocarcinoma tissue. PC, phosphatidylcholine; PE, phosphatidylethanolamine.

Through univariate and multivariate statistical analyses of lipids at lipid species levels, this study identified differential lipids meeting both criteria of P value <0.01 in volcano plot analysis and variable importance in projection (VIP) >1 (Figure S1) in sPLS-DA analysis. These differential lipids were visualized with hierarchical clustering analysis, and diagnostic model evaluation was performed to investigate interrelationships among these lipids and assess their potential for disease diagnosis. Hierarchical clustering analysis using a heatmap (Figure 12) demonstrated effective clustering trends between N and T groups based on 23 lipid molecules at the species level. The predominance of red regions in the T group indicated that the differentiating lipid molecules were primarily upregulated species.

Figure 12 The molecular types of lipid in group N and group T were analyzed by clustering heat map. N, normal lung tissue; T, lung adenocarcinoma tissue.

In conclusion, at the species level, polyunsaturated lipid molecules showed overall elevated expression in the T group (LUAD tissue), and these differential lipid features effectively distinguished between N and T group samples. These findings provide crucial insights into the mechanisms of altered lipid metabolism associated with LUAD.

Differential analysis of C=C positional isomers

FC analysis of C=C positional isomer ratio lipid molecules between groups N and T (Figure 13) identified five differentially expressed lipid molecules, with two downregulated and three upregulated. Figure 14 provides detailed information on these lipids. The upregulated lipids (Figure 14A-14C) included PC{16:0_18:1[Δ8/(Δ8+Δ9+Δ11)]}, PC[18:1(Δ9/Δ11)_20:0], PC(16:1(Δ6/(Δ6+Δ7+Δ9))_18:0). These lipids showed increased isomer ratios in Group T, suggesting potential alterations in desaturation or elongation pathways. Conversely, the downregulated lipids (Figure 14D,14E) included PE{18:1[Δ8/(Δ8+Δ9+Δ11)]_20:2(Δ8,Δ11)} and PE{18:1[Δ8/(Δ8+Δ9+Δ11)]_20:2(Δ11,Δ14)}. These lipids exhibited reduced isomer ratios in Group T, indicating potential shifts in lipid metabolism or structural remodeling.

Figure 13 Volcano plot analysis of C=C positional isomers relative ratios in lipid molecules in group N and group T samples. C=C, carbon-carbon double bond.

Figure 14 Details of C=C positional isomers with significant differences between group N and T samples. (A) PC{16:0_18:1[Δ8/(Δ8+Δ9+Δ11)]}. (B) PC{16:1[Δ6/(Δ6+Δ7+Δ9)]_18:0}. (C) PC[18:1(Δ9/Δ11)_20:0]. (D) PE{18:1[Δ8/(Δ8+Δ9+Δ11)]_20:2(Δ11, Δ14)}. (E) PE{18:1[Δ8/(Δ8+Δ9+Δ11)]_20:2(Δ8, Δ11)}. *, P<0.01; **, P<0.005; ***, P<0.001. N, normal lung tissue; T, lung adenocarcinoma tissue. C=C, carbon-carbon double bond; FC, fold change; PC, phosphatidylcholine; PE, phosphatidylethanolamine.

PCA (Figure 15A) showed that C=C positional isomer ratio lipid molecules could not completely distinguish between groups N and T. However, sPLS-DA (Figure 15B) revealed a clear separation trend between the two groups at the C=C positional level. Further loading analysis (Figure 16) indicated that in Component 1, PC[18:1(Δ9/Δ11)_20:0] exhibited significantly higher loading than other lipid molecules and showed the highest content in group T. In Component 2, PC{18:1[Δ9/(Δ8+Δ9+Δ11)]_22:1(Δ13)} showed the highest proportion in group N, while other high-loading molecules were primarily PE[18:1(Δ9/Δ11)_20:3(Δ11,Δ14,Δ17)] and PC{18:1[Δ11/(Δ8+Δ9+Δ11)]_22:1(Δ13)}.

Figure 15 Comparative analysis of C=C positional isomers between group N and group T using PCA and sPLS-DA. (A) PCA fails to differentiate C=C positional isomer profiles between groups at the species level. (B) Sparse partial least squares discriminant analysis effectively discriminates group-specific C=C positional isomer patterns. N, normal lung tissue; T, lung adenocarcinoma tissue. C=C, carbon-carbon double bond; PCA, principal component analysis; sPLS-DA, sparse partial least squares discriminant analysis.

Figure 16 Component loadings of C=C positional isomer ratios in lipid molecules. (A) Loading 1: lipid contributions to the primary discriminant component. (B) Loading 2: weightings for the secondary orthogonal component. C=C, carbon-carbon double bond.

Through univariate and multivariate statistical analyses of lipids at lipid C=C positional levels, this study identified differential lipids meeting both criteria of P value <0.01 in univariate analysis and VIP >1 (Figure S2) in multivariate analysis. These differential lipids were visualized with hierarchical clustering analysis, and diagnostic model evaluation was performed to investigate interrelationships among these lipids and assess their potential for disease diagnosis. Hierarchical clustering analysis of C=C positional lipid molecules between groups N and T is presented in Figure 17. The ratios of diagnostic ion pair abundances for 11 phospholipids containing C18:1 and C16:1 showed good clustering trends for groups N and T. Regarding changes in C=C positional isomer ratio intensities in group T versus N, upregulated lipid molecules predominantly contained C18:1(Δ9) isomers, while downregulated lipid molecules mainly contained C18:1(Δ8) isomers.

Figure 17 The C=C positional isomer ratios in lipid molecules in group N and group T were analyzed by clustering heat map. N, normal lung tissue; T, lung adenocarcinoma tissue. C=C, carbon-carbon double bond.

Discussion

LUAD, a major subtype of NSCLC, remains one of the leading causes of cancer-related mortality globally. Despite advances in early detection and treatment, the prognosis for LUAD patients remains poor. This underscores the need for more refined diagnostic and therapeutic strategies. The emerging field of lipidomics, focusing on the alteration of lipid metabolism in cancer, offers promising insights. Lipid synthesis and metabolism are closely linked to key biological processes such as tumor progression, metastasis, and apoptosis, with changes in lipid types and structures serving as critical hallmarks of tumorigenesis and potential therapeutic targets (27,28). Notably, structural features of lipids, particularly C=C positional isomers, may provide valuable information on the molecular mechanisms driving cancer progression (24). However, detailed research on the fine structure of tumor lipids, including C=C double bonds and positional isomers, remains limited. Cao et al. (29) achieved the identification and quantitation of lipids with C=C double bonds and sn-specificities and classified abnormalities in C=C double bonds. However, this study only performed methodological validation on lung cancer and breast cancer cells as well as a limited number of tissue samples. Addressing this gap, our study conducted an exploratory lipidomic analysis of LUAD tumor tissues, revealing distinct patterns of lipid species, particularly C=C isomers, and achieving promising results. These findings, synthesized with recent literature, underscore the potential of lipidomics for advancing LUAD diagnosis and therapy.

Tumor cells often exhibit heightened lipid biosynthesis, a process essential for cellular membrane formation, energy storage, and signaling (28,30,31). In our study, 5 phospholipid species were identified at the C=C isomer level, with 2 lipid molecules were down-regulated and 3 lipid molecules were up-regulated. Abnormal expression of polyunsaturated lipids was reported as a hallmark of cancer biology that have been implicated in metastatic extravasation, ferroptosis and immunotherapy resistance, which mirrors the lipidomic shifts observed in other cancers (32-34). Studies have shown that two critical fatty acid desaturases, stearoyl-CoA desaturase-1 (SCD1) and acyl-CoA 6-desaturase (FADS2), were aberrantly upregulated, accelerating lipid metabolic activities and tumor aggressiveness of ovarian cancer cells (35). Moreover, genetically increased polyunsaturated fatty acid (PUFA) desaturase activity has been linked to a higher risk of colorectal cancer, esophageal squamous cell carcinoma, lung cancer, and basal cell carcinoma. The PUFA biosynthesis pathway could serve as a potential target for interventions aimed at preventing colorectal cancer and esophageal squamous cell carcinoma (36). Although our findings have not yet established a direct link between abnormal lipid expression and tumorigenesis, the lipidomic profiles of LUAD and normal tissue in our study support the notion that lipid signatures can effectively differentiate between tumor and healthy cells. This provides a foundation for the development of lipid-based diagnostic tools.

In this study, the sPLS-DA data analysis method was used to effectively distinguish tumors from normal tissues at the lipid type level and carbon-carbon double bond level, and five lipid molecules with higher weights were screened out: PC[18:1(Δ9/Δ11)_20:0], PC{18:1[Δ9/(Δ8+Δ9+Δ11)]_22:1(Δ13)}, PC[18:1(Δ9/Δ11)_20:3(Δ11,Δ14,Δ17)], PC{18:1[Δ11/(Δ8+Δ9+Δ11)]_22:1(Δ13)} and PC[18:1(Δ9/Δ11)_22:1(Δ13)]. These differential features basically achieved clustering between groups, indicating that the relative proportion of C=C positional isomers can be used as a marker to distinguish LUAD from normal tissue samples. This provides a basis for our further research to analysis fine structures such as the sn-position, with the latest proposed lipidomics workflows (25).

Our study highlights the potential of using the relative abundances of C=C positional isomers as biomarkers for LUAD, offering a promising step toward more precise diagnostic tools. Given the challenges of early detection in LUAD, these findings underscore the critical role of lipid reprogramming in cancer progression and the potential of lipidomic biomarkers to improve diagnostics. Specifically, the observed upregulation of polyunsaturated lipids and the differential distribution of C=C isomers provide a molecular fingerprint capable of differentiating tumor tissues from normal tissues. Such advancements could pave the way for the development of non-invasive lipidomic biomarkers, complementing existing diagnostic techniques and enabling more accurate and earlier detection of LUAD.

Despite the strengths of this study, several limitations should be acknowledged. First, the sample size was relatively small, which may limit the generalizability of our findings. Future studies with larger cohorts are needed to validate the diagnostic and prognostic value of the identified lipid signatures. Second, while our study focused on phospholipids and their C=C positional isomers, other lipid classes, such as sphingolipids and sterols, may also play critical roles in LUAD and warrant further investigation. Additionally, the functional implications of the observed alterations in C=C isomer ratios remain unclear. While our findings suggest that these structural changes may influence membrane properties and lipid signaling pathways, further mechanistic studies are needed to elucidate their precise roles in LUAD biology. Finally, the potential impact of tumor heterogeneity and the tumor microenvironment on lipidomic profiles should be explored in future research.

Conclusions

Our deep structural lipidomic of LUAD tissues reveals significant alterations in lipid composition, with particular emphasis on the differential regulation of C=C positional isomers. These findings contribute to the growing body of evidence supporting the role of lipid metabolism in cancer progression and offer novel insights into potential diagnostic and therapeutic strategies for LUAD. The identification of lipid signatures, especially those associated with C=C isomers, could significantly improve early detection and lead to more precise, individualized therapies.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STREGA reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-717/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-717/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-717/prf

Funding: This work was supported by Beijing Clinical Key Specialty Construction Project.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-717/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. The study was approved by the Ethics Committee of Beijing Chaoyang Hospital, Capital Medical University (approval No. 2022-ke-502) and individual consent for this retrospective analysis was waived.

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. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Siegel RL, Miller KD, Wagle NS, et al. Cancer statistics, 2023. CA Cancer J Clin 2023;73:17-48. [Crossref] [PubMed]
3. Gridelli C, Rossi A, Carbone DP, et al. Non-small-cell lung cancer. Nat Rev Dis Primers 2015;1:15009. [Crossref] [PubMed]
4. Brody H. Lung cancer. Nature 2020;587:S7. [Crossref] [PubMed]
5. Hirsch FR, Scagliotti GV, Mulshine JL, et al. Lung cancer: current therapies and new targeted treatments. Lancet 2017;389:299-311. [Crossref] [PubMed]
6. Dalwadi SM, Lewis GD, Bernicker EH, et al. Disparities in the Treatment and Outcome of Stage I Non-Small-Cell Lung Cancer in the 21st Century. Clin Lung Cancer 2019;20:194-200. [Crossref] [PubMed]
7. International Early Lung Cancer Action Program Investigators. Survival of patients with stage I lung cancer detected on CT screening. N Engl J Med 2006;355:1763-71. [Crossref] [PubMed]
8. Hoffman RM, Atallah RP, Struble RD, et al. Lung Cancer Screening with Low-Dose CT: a Meta-Analysis. J Gen Intern Med 2020;35:3015-25. [Crossref] [PubMed]
9. National Lung Screening Trial Research Team. The National Lung Screening Trial: overview and study design. Radiology 2011;258:243-53. [Crossref] [PubMed]
10. Bach PB, Mirkin JN, Oliver TK, et al. Benefits and harms of CT screening for lung cancer: a systematic review. JAMA 2012;307:2418-29. [Crossref] [PubMed]
11. Vansteenkiste J, Dooms C, Mascaux C, et al. Screening and early detection of lung cancer. Ann Oncol 2012;23:x320-7. [Crossref] [PubMed]
12. Yan G, Li J, Gao X, et al. Comprehensive analysis of the diagnostic and therapeutic value, immune infiltration, and drug treatment mechanisms of GTSE1 in lung adenocarcinoma. Front Med (Lausanne) 2024;11:1433601. [Crossref] [PubMed]
13. Fahy E, Subramaniam S, Murphy RC, et al. Update of the LIPID MAPS comprehensive classification system for lipids. J Lipid Res 2009;50:S9-14. [Crossref] [PubMed]
14. Shevchenko A, Simons K. Lipidomics: coming to grips with lipid diversity. Nat Rev Mol Cell Biol 2010;11:593-8. [Crossref] [PubMed]
15. Xu T, Hu C, Xuan Q, et al. Recent advances in analytical strategies for mass spectrometry-based lipidomics. Anal Chim Acta 2020;1137:156-69. [Crossref] [PubMed]
16. Beloribi-Djefaflia S, Vasseur S, Guillaumond F. Lipid metabolic reprogramming in cancer cells. Oncogenesis 2016;5:e189. [Crossref] [PubMed]
17. Wu Y, Pu X, Wang X, et al. Reprogramming of lipid metabolism in the tumor microenvironment: a strategy for tumor immunotherapy. Lipids Health Dis 2024;23:35. [Crossref] [PubMed]
18. Guo S, Wang Y, Zhou D, et al. Significantly increased monounsaturated lipids relative to polyunsaturated lipids in six types of cancer microenvironment are observed by mass spectrometry imaging. Sci Rep 2014;4:5959. [Crossref] [PubMed]
19. Hartmann P, Trufa DI, Hohenberger K, et al. Contribution of serum lipids and cholesterol cellular metabolism in lung cancer development and progression. Sci Rep 2023;13:5662. [Crossref] [PubMed]
20. Wang G, Qiu M, Xing X, et al. Lung cancer scRNA-seq and lipidomics reveal aberrant lipid metabolism for early-stage diagnosis. Sci Transl Med 2022;14:eabk2756. [Crossref] [PubMed]
21. Sun T, Chen J, Yang F, et al. Lipidomics reveals new lipid-based lung adenocarcinoma early diagnosis model. EMBO Mol Med 2024;16:854-69. [Crossref] [PubMed]
22. Zhang W, Jian R, Zhao J, et al. Deep-lipidotyping by mass spectrometry: recent technical advances and applications. J Lipid Res 2022;63:100219. [Crossref] [PubMed]
23. Jump DB. The biochemistry of n-3 polyunsaturated fatty acids. J Biol Chem 2002;277:8755-8. [Crossref] [PubMed]
24. Wenk MR. Lipidomics: new tools and applications. Cell 2010;143:888-95. [Crossref] [PubMed]
25. Zhao X, Zhang W, Zhang D, et al. A lipidomic workflow capable of resolving sn- and C=C location isomers of phosphatidylcholines. Chem Sci 2019;10:10740-8. [Crossref] [PubMed]
26. Zhang W, Zhang D, Chen Q, et al. Online photochemical derivatization enables comprehensive mass spectrometric analysis of unsaturated phospholipid isomers. Nat Commun 2019;10:79. [Crossref] [PubMed]
27. Zheng M, Wang W, Liu J, et al. Lipid Metabolism in Cancer Cells. Adv Exp Med Biol 2021;1316:49-69. [Crossref] [PubMed]
28. Martin-Perez M, Urdiroz-Urricelqui U, Bigas C, et al. The role of lipids in cancer progression and metastasis. Cell Metab 2022;34:1675-99. [Crossref] [PubMed]
29. Cao W, Cheng S, Yang J, et al. Large-scale lipid analysis with C=C location and sn-position isomer resolving power. Nat Commun 2020;11:375. [Crossref] [PubMed]
30. Kim J, DeBerardinis RJ. Mechanisms and Implications of Metabolic Heterogeneity in Cancer. Cell Metab 2019;30:434-46. [Crossref] [PubMed]
31. He M, Xu S, Yan F, et al. Fatty Acid Metabolism: A New Perspective in Breast Cancer Precision Therapy. Front Biosci (Landmark Ed) 2023;28:348. [Crossref] [PubMed]
32. Wang Y, Hu M, Cao J, et al. ACSL4 and polyunsaturated lipids support metastatic extravasation and colonization. Cell 2025;188:412-429.e27. [Crossref] [PubMed]
33. Suda A, Umaru BA, Yamamoto Y, et al. Polyunsaturated fatty acids-induced ferroptosis suppresses pancreatic cancer growth. Sci Rep 2024;14:4409. [Crossref] [PubMed]
34. Xiang W, Lv H, Xing F, et al. Inhibition of ACLY overcomes cancer immunotherapy resistance via polyunsaturated fatty acids peroxidation and cGAS-STING activation. Sci Adv 2023;9:eadi2465. [Crossref] [PubMed]
35. Xuan Y, Wang H, Yung MM, et al. SCD1/FADS2 fatty acid desaturases equipoise lipid metabolic activity and redox-driven ferroptosis in ascites-derived ovarian cancer cells. Theranostics 2022;12:3534-52. [Crossref] [PubMed]
36. Haycock PC, Borges MC, Burrows K, et al. The association between genetically elevated polyunsaturated fatty acids and risk of cancer. EBioMedicine 2023;91:104510. [Crossref] [PubMed]