Lipid Metabolism Related Gene ACSL3 as a Biomarker for Predicting Immunotherapy Outcomes in Lung Adenocarcinoma

Taiping He; Jinhan Hu; Haoyue Guo; Meng Diao; Yuanyuan Wang; Yuhan Wu; Lei Cheng; Chao Zhao; Xuefei Li; Caicun Zhou

doi:10.4143/crt.2024.1119

Articles

Page Path: HOME > Cancer Res Treat > Volume 57(4); 2025 > Article

Original Article Lung and Thoracic cancer Lipid Metabolism Related Gene ACSL3 as a Biomarker for Predicting Immunotherapy Outcomes in Lung Adenocarcinoma: Taiping He¹, Jinhan Hu², Haoyue Guo¹, Meng Diao¹, Yuanyuan Wang³, Yuhan Wu⁴, Lei Cheng⁵, Chao Zhao⁵, Xuefei Li⁵, Caicun Zhou³; Cancer Research and Treatment : Official Journal of Korean Cancer Association 2025;57(4):1000-1018.
DOI: https://doi.org/10.4143/crt.2024.1119
Published online: January 20, 2025

¹Department of Medical Oncology, Shanghai Pulmonary Hospital, Tongji University School of Medicine, Shanghai, China

²Department of Gastroenterology, Shanghai East Hospital, Tongji University School of Medicine, Shanghai, China

³Department of Oncology, Shanghai East Hospital, Tongji University School of Medicine, Shanghai, China

⁴Bengbu Medical University, School of Nursing Bengbu, Bengbu, China

⁵Department of Lung Cancer and Immunology, Shanghai Pulmonary Hospital, Shanghai, China

Correspondence: Xuefei Li, Department of Lung Cancer and Immunology, Shanghai Pulmonary Hospital, 507 Zhengmin Road, Yangpu District, Shanghai 200433, China
Tel: 86-21-65115006 E-mail: bug_lily2003@163.com

Co-correspondence: Caicun Zhou, Department of Oncology, Shanghai East Hospital, Tongji University School of Medicine, 500 Zhennan Road, Putuo District, Shanghai 200331, China
Tel: 86-21-38804518 E-mail: doctorcaicunzhou@163.com

^*Taiping He, Jinhan Hu, and Haoyue Guo contributed equally to this work.

• Received: November 22, 2024 • Accepted: January 16, 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.

4,568 Views
190 Download
5 Web of Science
6 Crossref
4 Scopus

prev next

Full Article

Download PDF

Abstract

Purpose
Investigate the role of lipid metabolism in the tumor immune microenvironment (TIME) of lung adenocarcinoma (LUAD) and identify vital lipid metabolism-related genes (LMRGs) that contribute to immunotherapy outcomes.
Materials and Methods
One thousand one hundred thirty LUAD patients were acquired utilizing public databases. Multiple algorithms were used to analyze the contribution of lipid metabolism in TIME. Importantly, cell lines, clinical samples (52 patients in surgery cohort and 36 in immunotherapy cohort), animal models, RNA sequencing (RNA-seq), experiments in protein and mRNA levels were conducted for identifying and validating key biomarker in LUAD immunotherapy.
Results
A prognostic signature comprising 33 LMRGs was developed and validated as an effective predictor of prognosis and TIME, with a C-index of 0.766 (95% confidence interval, 0.729 to 0.804). Additionally, we identified acyl-CoA synthetase long-chain family member 3 (ACSL3) as a potential biomarker for immunotherapy prognosis. The expression of ACSL3 was verified in 88 clinical tissues from LUAD patients, which indicated that elevated ACSL3 expression was correlated with worse progression-free survival (p < 0.001) and overall survival (p=0.008). Subsequent experiments revealed that knockdown of ACSL3 in vivo enhanced the efficacy of immunotherapy, potentially through increasing interferon-α secretion, as indicated by bulk RNA-seq and enzyme-linked immunosorbent assay analysis, thereby promoting the infiltration of antitumor immune cells.
Conclusion
The study established a model that accurately predicts immunotherapy response, prognosis, and TIME dynamics in LUAD patients. Notably, the pivotal role of ACSL3 in driving tumor progression and immune evasion was uncovered, offering novel insights into the optimization of immunotherapy strategies for LUAD.
Key words: Lung adenocarcinoma, ACSL3, Tumor microenvironment, Immunotherapy, Lipid metabolism

Introduction

Lung cancer, particularly lung adenocarcinoma (LUAD), is a malignant disease characterized by high morbidity and mortality [1]. In recent years, despite the significant success of immune checkpoint inhibitors (ICIs), several factors still hinder patient responsiveness, posing challenges for achieving optimal therapeutic outcomes. Hence, it is crucial to discover reliable biomarkers that can predict the effectiveness of immunotherapy and provide new intervention targets for LUAD.

In the ongoing quest to understand malignant tumors, the reprogramming of lipid metabolism has emerged as a promising new marker. Enhanced lipid metabolism, including absorption, storage, and adipogenesis, is a hallmark of rapidly growing tumor cells [2]. Lipid metabolism-related genes (LMRGs) have been closely associated with tumor development and hold strong prognostic potential [3,4]. Intriguingly, there is a growing connection between lipid metabolism and LUAD, suggesting that LMRGs may have a crucial role in the progression, prognosis, prevention, and treatment in lung cancer [5]. Accumulating evidence indicates that lipid metabolism may also impact the cancer immunological microenvironment. In the harsh conditions of tumor immune microenvironment (TIME), such as hypoxia, acidity, and nutrient deficiency, both tumor and immune cells rely heavily on lipids for energy [6]. Consequently, dysregulated lipid metabolism in TIME can significantly influence tumorigenesis, progression, and metastasis, paradoxically supporting both antitumor and protumor immune responses. While the relationship between TIME and lipid metabolism has been explored in various cancers [4], its specific role in LUAD remains largely unexplored.

To bridge this gap, we performed an exhaustive analysis of LMRGs expression patterns in LUAD, aiming to develop an innovative model that can predict survival outcomes and characterize the TIME in LUAD patients. Specifically, we focused on acyl-CoA synthetase long-chain family member 3 (ACSL3), a pivotal regulator in lipid metabolism, which transforms free long-chain fatty acids to fatty acyl-CoA. Previous research has shown that ACSL3 is implicated in the advancement of numerous cancer types, suggesting a poorer prognosis for patients. Furthermore, ACSL3 plays a crucial role in the process of ferroptosis, enhancing cancer cells’ resistance to this type of cell death. However, there has been limited exploration regarding the role of ACSL3 in the context of immunotherapy. By investigating the function of ACSL3 in LUAD development and elucidating its underlying mechanisms that modulate immunotherapy efficacy, we hope to offer fresh perspectives on the function of LMRGs in LUAD development and identify novel biomarkers for immunotherapy and tailored treatment plans.

Materials and Methods

1. Data collection

The Cancer Genome Atlas (TCGA) database and the Gene Expression Omnibus (GEO) database (GSE13213, GSE26939 and GSE72094) have collated patient data and RNA sequencing (RNA-seq) records. We used the method by Johnson et al. [7] to remove batch effects. Inclusion criteria for clinical information were (1) LUAD samples; (2) samples which have matched expression patterns of genes and patient data; (3) samples with full clinical details, including sex, survival time, and life status. The exclusion criteria are listed as follows: (1) sample of normal tissues, (2) samples with ≥ 10 unexpressed genes, (3) samples with biased expression.

2. Molecular subgroup classification

Univariate Cox regression analysis was performed to identify LMRGs which affected both lipid metabolism and survival prognosis. ConsensusClusterPlus [8] was used to conduct cluster analysis.

3. Protocol for the immunoassay

During the analysis process, the Sangerbox (http://www.sangerbox.com/tool) provided valuable help. Three correlation scores were computed to estimate stromal and immunological cells using the ESTIMATE [9] algorithm utilizing the R package, IOBR [10]. ‘immunedeconv’ was used to estimate the abundance scores for 35 immune cell types, This analysis was performed by xCell [11] and Single Sample Gene Set Enrichment Analysis (ssGSEA). The online platform provided Tumor Immune Dysfunction and Exclusion (TIDE) grades [12-14].

4. Analysis of the functional aspects

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were used to functionally analyze genes that were differently expressed. We utilized the R software to obtain the enrichment score for each sample in the gene set by Gene Set Variation Analysis (GSVA) ver. 1.40.1.

5. Constructing and evaluating the risk model

Firstly, survival time, status, and gene activity were obtained. Then, using the lasso-Cox approach to perform a regression analysis. To obtain the ideal model, we also conducted 10-fold cross-validation. A total of 33 genes were collated in the model with a lambda of 0.043. The specific risk score calculation formula is shown in Supplementary Methods. The Cox technique was used to create a nomogram and obtained the C-index. Receiver operating characteristics (ROC) and nomograms were used in the final stage of evaluating the model’s performance.

6. Clinical specimens and immunohistochemistry

The frozen LUAD samples, originating from patients were provided by the Shanghai Pulmonary Hospital. Additionally, the tissue microarray (TMA) was supplied by Shanghai Weiao Biotechnology Company. All samples were provided upon patient consent (ethics approval No. K20-288). And the tumor tissues from mice were obtained for immunohistochemical (IHC) staining. All antibodies used are listed: ACSL3 (1:200, BS74664, Bioworld), neuronal PAS domain-containing protein 2 (NPAS2; 1:100, PHR3777, Abmart), CD8 (1:100, GB15068, Servicebio), CD4 (1:100, AB183685, Servicebio), CD163 (1:100, AB87099, Servicebio), iNOS (inducible nitric oxide synthase; 1:100, GB-11119-50, Servicebio). The percentage of IHC scores, positive cells and mean density were calculated for each patient sample, allowing for the assessment of ACSL3, NPAS2, CD8, CD4, CD163, iNOS protein expression levels.

7. Cell culture and shRNA transfection

The human LUAD cell lines, encompassing PC-9, H1975, 95D, HCC827, LLC, CMT-167, KRAS^G12DTP53^−/− (KP) along with the normal lung epithelial cell line BEAS-2b were provided from The Shanghai Pulmonary Hospital. These cells were propagated in RPMI-1640 medium, fortified with 10% fetal bovine serum sourced from Gibco, and nurtured within an incubator set at 37°C and supplemented with 5% CO₂ for optimal growth conditions. The shRNA oligonucleotides (negative control shRNA and ACSL3 shRNA) were synthesized by Transheep. Subsequently, the transfection of these shRNA constructs into the cells was facilitated using Lipofectamine 3000 (Invitrogen). The specific sequences utilized of shRNAs are detailed in S1 Table.

8. Quantitative real-time PCR

Employing RNA extraction kit (Fastagen) to extract total cellular RNA, and subsequently converted into cDNA utilizing reverse transcription reagents (TaKaRa). Subsequently, quantitative real-time PCR (qRT-PCR) was performed. The specific primer sequences are detailed in S2 Table.

9. Western blot

The Western blot procedure was performed following a previously established protocol [15]. All antibodies used are listed: ACSL3 (1:1,000, BS74664, Bioworld); β-tubulin (1:5,000, M20005S, Abmart); secondary antibodies (1:10,000, M212133S, Abmart).

10. Animal experiments

Using the specified sequences of sh-NC and sh-ACSL3#1, we constructed shRNA lentiviral vectors and efficiently transduced them into LLC cells with the help of polybrene. We then established a xenograft model in 6-week-old male C57BL/6 mice by subcutaneously injecting 2×10⁶ sh-NC and sh-ACSL3 LLC cells into the right flank of each mouse. The tumor volumes were computed using the formula: tumor volume (mm³)=0.5×(length×width²). Upon reaching a tumor volume of 100-200 mm³, we classified the mice into groups and subjected to treatment with programmed cell death protein 1 antagonists (anti–PD-1) (S0B0594-50mg, STARTER, 10 mg/kg) administered via intraperitoneal injection, or left untreated. At the conclusion of the experiment, the mice were euthanized, and the xenograft tumors were excised, fixed, prepared for subsequent IHC analyses and bulk RNA-seq. Additionally, the mice’s plasma was harvested for further examination.

11. Enzyme-linked immunosorbent assay

The enzyme-linked immunosorbent assay (ELISA) was conducted with ELISA kit (Jianglaibio, interferon-α [IFN-α]). Samples were diluted 2-fold for loading and triplicate standard curves were used, and the optical density value for each well is then determined at a wavelength set to 450 nm.

12. Bulk RNA-seq

Using a previously established protocol, Sinotech Genomics sequenced the libraries on an Illumina NovaSeq 6000 platform. Gene abundance was quantified as fragments per kilobase of exon model per million mapped reads. Fragment counts within each gene were tallied using Stringtie software, followed by normalization with the TMM algorithm. Differential expression analysis, as well as GO, KEGG, and Gene Set Enrichment Analysis (GSEA) analyses, were conducted as previously described.

13. Analytical statistics

Survival curves were plotted using the Kaplan-Meier method. The ideal cutoff value of the risk score was calculated specifically using the R software program, max stat (ver. 4.0.4, R Core Team). The survey function of the R package, survival, was used to assess the prognostic differences across groups. The log-rank test approach was utilized to determine the significance of the prognostic differences between samples across groups. To calculate area under curve (AUC), a ROC analysis was performed for the predictive ability of the risk model. Patients were grouped according to sex and stage, and a subgroup analysis was conducted. During the data processing, we got valuable help in Sangerbox [16]. In this study, we leveraged the X-tile software (ver. 3.6.1, Yale University) to ascertain the precise cutoff points for an array of continuous and multiple categorical variables, which were subsequently incorporated into Cox regression models and Kaplan-Meier survival curves. The URLs of all the websites that were employed in this study can be found in Supplementary Methods.

Results

1. Characterization of the two subtypes identified using LMRGs

S3 Fig. exhibited the workflow of our study. Firstly, RNA-seq and clinical data of 1,130 patients were obtained from TCGA and GEO databases. S4 Table provides a comprehensive overview of the clinical features of two cohorts. Through univariate Cox analysis, 147 LMRGs related to patient survival in LUAD were identified (S5 Table). Then, we utilized the RNA-sequencing data to classify patients into two distinct groups (K=2): cluster 1 comprised 246 patients, while cluster 2 included 254 patients (Fig. 1A and B). Analysis of the clinical data demonstrated that cluster 1 exhibited a markedly improved overall survival (OS) than cluster 2 (p < 0.001) (Fig. 1C). Furthermore, immunoassays indicated that cluster 1 had significantly higher immune score (p < 0.001) (Fig. 1D). Additionally, the TIDE algorithm revealed that cluster 1 owned a lower score, which suggests superior efficacy of ICIs and an associated longer survival (Fig. 1E). Moreover, xCell Statistical Analysis indicated that cluster 1 was characterized by a more robust immune status, with significant differences observed across 24 distinct cell types. Notably, cluster 2 was enriched in CD4+ memory T cells and CD8+ naïve T cells, while cluster 1 showed a predominance of B cells and CD4+ naïve T cells (Fig. 1F). These results highlight the crucial impact of LMRG expression on the TIME, immune status, and the effectiveness of ICIs in LUAD patients.

Subsequently, A functional analysis of the differentially expressed genes (DEG) was carried out. We identified 1,178 DEGs in cluster 2 compared to cluster 1, including 721 upregulated and 457 downregulated genes (S6A Fig.). Enrichment analyses showed that these DEGs were primarily involved cell cycle regulation, humoral immunity, and antigen processing, as indicated by GO and KEGG analyses (S6B and S6C Fig.). GSVA and GSEA were employed to assess pathway expression differences. Cluster 2 exhibited reduced lung interstitial formation and alveolar development, along with increased activity in actomyosin assembly and cell division regulation. GSEA analysis additionally showed that cluster 1 exhibited an enrichment of pathways associated with cell cycle progression, spliceosome functionality, and mismatch repair mechanisms. These findings suggest that dysregulation of immune pathways, alterations in the cell cycle, and changes in lung tissue architecture are linked to LMRG expression, potentially contributing to poor prognosis in LUAD patients (S6D and S6E Fig.).

2. Building and validating an LMRG-based risk signature model

For accessing the prognosis and predictive utility of LMRGs in LUAD, we developed a risk signature model using the lasso-Cox approach. With the lambda value adjusted at 0.043, 33 genes were finally retrieved, which were listed in S5 Table. Based on the median value, the created risk model was able to effectively divide LUAD patients into two groups: high-risk and low-risk. We constructed survival curves that illustrated a statistically meaningful distinction between the two patient groups (p < 0.001) (Fig. 2A). Subgroup and regression analyses revealed that risk scores varied by sex (p < 0.05) and stage (p < 0.001) (S6F and S6G Fig.). Furthermore, the TIME was assessed using ESTIMATE algorithms [9], which showed that the high-risk group had significantly higher immune (p < 0.01) score compared to the low-risk group (Fig. 2B). In addition, AUCs of 0.80 (95% confidence interval [CI], 0.86 to 0.73), 0.79 (95% CI, 0.84 to 0.73), and 0.79 (95% CI, 0.85 to 0.72) for 1-year, 3-year, and 5-year ROC curves, respectively, were obtained. Our findings indicated that the established risk model could predict prognosis and was connected to TIME in LUAD patients accurately (Fig. 2C). In addition, the constructed risk model was shown to be a stand-alone predictor of the prognosis of LUAD patients by the univariate and multivariate Cox regression analysis (S7 and S8 Tables).

After successfully removing the batch effects (S6H and S6I Fig.), the validation cohort of LUAD patients confirmed the model’s ability for predicting prognostic outcomes and TIME (Fig. 2D-F). The low-risk individuals showed significantly superior outcomes (p < 0.001) (Fig. 2D). Moreover, the low-risk group had higher immune score compared to the high-risk group, consistent with the training cohort (Fig. 2E).

To improve prognosis prediction, we constructed a nomogram that integrates risk scores and clinical information. In the training cohort, this nomogram achieved an OS C-index of 0.766 (95% CI, 0.729 to 0.804). Validation yielded a C-index of 0.685 (95% CI, 0.648 to 0.722) (Fig. 2G and H). Predictions for 3- and 5-year survival closely matched actual outcomes in both cohorts (Fig. 2I and J). Our findings indicate that the nomogram may accurately predict prognosis in LUAD patients.

3. ACSL3 predicts poor prognosis in surgery cohort of LUAD patients

To identify key biomarkers within LMRGs that are linked to prognosis and immunotherapy response in LUAD, we screened genes among the 33 LMRGs (S5 Table.) included in our established risk model using the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn).We identified two genes: ACSL3 and NPAS2, as having the capacity to independently forecast prognosis. We uncovered a correlation between elevated expression levels of both ACSL3 and NPAS2 and diminished survival outcomes (Fig. 3A and B). Next, we examined the mRNA expression profiles in human normal lung epithelial cells (BEAS-2b) and four LUAD cell lines (PC9, H1975, 95D, and HCC827) through qRT-PCR. Our analysis revealed a marked upregulation of ACSL3 and NPAS2 expression in all LUAD cell lines compared to BEAS-2b cells (Fig. 3C and D). Subsequently, we utilized a TMA which comprises 52 tumor specimens from LUAD patients under surgery treatment. The characteristics of these patients was summarized on S9 Table. The OS does not exhibit a statistically significant difference based on M status, with a p-value of 0.062. This may be attributed to the constraint of sample size, as there are only three patients categorized under M1 status. We performed IHC staining for ACSL3 and NPAS2. The IHC quantitative results were analyzed with H-score and percentage of positive cells. Representative images were showed in Fig. 3E. Our findings indicated that heightened expression of ACSL3 was indicative of poorer OS (H-score: p=0.008; positive cells %: p=0.004) (Fig. 3F and G), whereas no statistically significant correlation was observed between NPAS2 expression level and OS (H-score: p=0.363; positive cells %: p=0.295) (S10A and S10B Fig.).

4. ACSL3 predicts poor efficacy in immunotherapy cohort of LUAD patients

To delve deeper into the relation of ACSL3 with immunotherapy response in LUAD, we selected cell lines which have different response to immunotherapy. KRAS^G12DTP53^−/− (KP) [17] and CMT-167 are immune sensitive cell lines, LLC is immune resistant cell line [18]. We selected the three cell lines for mRNA levels quantification via qRT-PCR, indicating that immune-resistant cells exhibited higher expression levels both ACSL3 and NPAS2 (Fig. 3H and I) Additionally, we leveraged own clinical samples and employed IHC staining to validate. There are a total of 32 patients, we classified patients into responders, including those with complete response and partial response (PR), and non-responders, which encompassed those with stable disease and progressive disease [19]. Therefore, 12 patients are responders and the remaining 20 are non-responders. The characteristics of immunotherapy cohort are summarized in Table 1. We found that the LUAD patients who did not respond to immunotherapy displayed a higher expression of ACSL3, according to the H-score (p=0.036) and mean density (p=0.018) (Fig. 3J). It was worth noting that there were three patients, all of whom responded to immunotherapy, who underwent surgery or radiotherapy after achieving PR from immunotherapy. This influenced the calculation of the progression-free survival (PFS) for immunotherapy. Therefore, when drawing the survival curves and conducting the Cox regression analysis for PFS in the cohort receiving immunotherapy, we excluded these three patients.

Moreover, high expression of ACSL3 correlates with a shorter PFS (H-score: p < 0.001; mean density: p=0.048) (Fig. 3K and L), and its expression was significantly higher in non-responders than in responders, as shown by both H-score and mean density (Fig. 3M and N). Nevertheless, no statistically significant difference in NPAS2 expression was found between groups, and there was no correlation with PFS (H-score: p=0.409; mean density: p=0.159) (S10C-S10F Fig.). This confirmation emphasizes the potential importance of ACSL3 in forecasting immunotherapy outcomes in LUAD.

5. ACSL3 knockdown improve the efficacy of immunotherapy of LUAD in vivo

After above validation, ACSL3 emerges as a pivotal factor influencing the effectiveness of immunotherapy. Due to the highest expression of ACSL3 in LLC among common murine LUAD cell lines (Fig. 4A), it was chosen for subsequent experiments involving shRNA knockdown. Notably, the protein level of ACSL3 was significantly reduced by sh-ACSL3#1 (Fig. 4A). Next, we constructed xenograft model by administering subcutaneous injections of both sh-NC (negative control) and sh-ACSL3#1 (hereafter abbreviated as sh-ACSL3) LLC cells in C57BL/6 mice. Once tumor volumes attained a range of 100 to 200 mm³, we divided mice into four groups: sh-NC+ phosphate buffered saline (PBS) (group 1); sh-NC+anti–PD-1 (group 2); sh-ACSL3+PBS (group 3); sh-ACSL3+anti–PD-1 (group 4); every 3 days for a duration of 2 weeks, mice received intraperitoneal injections of either PBS or anti–PD-1 (Fig. 4B). Following the conclusion of the experiment, mice were all euthanized, and the tumor sizes among the four different groups were compared. Considering the inherent immune resistance of the LLC cell line, no statistically difference was found between group 1 and group 2 (p=0.580). Conversely, the sh-ACSL3 and anti–PD-1 combination markedly reduced tumor sizes, comparing group 4 and group 3 (p=0.008), comparing group 4 and group 2 (p=0.035), yielding the smallest tumor weights, highlighting the fact that decreasing ACSL3 expression can boost the therapeutic effectiveness of immunotherapy in LUAD (Fig. 4C-E). Due to the significantly smaller tumor size in group 4 compared to group 3, there was a difference in the body weights between the two groups of mice (Fig. 4F).

6. Knockdown of ACSL3 facilitates the formation of an antitumor TIME

To gain further insights into the mechanisms underlying the improved efficacy of immunotherapy following ACSL3 knockdown, bulk RNA-seq was conducted on mouse tumor tissues from group 2 and group 4. We identified a total of 205 significant DEGs, comprising 16 downregulated and 179 upregulated genes, which included multiple IFN-related genes such as igtp, ifi44, ifi203, irf1, ifit1, ifi206, ifi47, as well as genes pertinent to immune system processes like trbv16, trbv5, cd8a, cd8b1, cd3e, and stat1 (Fig. 5A). GO enrichment analysis indicated that the upregulated genes were primarily enriched in pathways involved IFN signaling, cytokine response, positive regulation of immune response, and cytokine-mediated signaling (Fig. 5B). In addition, when LUAD patients in TCGA database were divided into two groups according to the median expression level of ACSL3 and GSEA analysis was performed, it was found that the IFN-α pathway was enriched in the low ACSL3 expression group, aligning with the RNA-seq results (Fig. 5E and F). KEGG analysis further indicated enrichment of JAK-STAT signaling pathway, NOD-like receptor signaling, tumor necrosis factor α, and chemokine signaling pathways among the upregulated genes (Fig. 5C). Moreover, ssGSEA demonstrated an antitumor TIME in group 4, characterized by increased infiltration of antitumor immune cells CD8+ T cells and natural killer cells, tumor-infiltrating lymphocytes, neutrophils, plasmacytoid dendritic cells (pDCs) and M1 macrophages. Additionally, protumor cytokines, angiogenesis, and tumor proliferation rates were reduced, while T cell and antigen-presenting cell co-stimulation were enhanced in this group (Fig. 5D).

Given the multiple associations with IFN-α pathway enrichment, we measured IFN-α levels in mouse plasma. ELISA results suggested a significant promotion in IFN-α levels in group 4 (Fig. 5G). Furthermore, IHC staining was used to assess immune cell infiltration, confirming an increased infiltration of CD8+ T cells in group 4, along with a decrease in CD163 expression, which is the marker of M2 macrophages. However, no statistically notable variation was detected in the quantities of CD4+ T cells or in the manifestation of iNOS, which is the marker of M1 macrophages (Fig. 5H). Consequently, these results demonstrate that knockdown of ACSL3 fosters an antitumor TIME, thereby potentiating the efficacy of immunotherapy.

Discussion

Among all subtypes of lung cancer, LUAD has the highest incidence rate, accounting for approximately 40%. Therefore, studying the pathogenesis of LUAD is of utmost importance for improving the survival of lung cancer patients. Given the significant malignancy and unfavorable prognosis associated with LUAD, with a hallmark of reprogramming of lipid metabolism [5]. There is a pressing need to investigate the contribution of lipid metabolism in TIME of LUAD and its correlation with immunotherapy and prognosis. In our research, we observed that LUAD patients displaying distinct expression profiles of LMAGs exhibited disparities in their prognosis and TIME. Consequently, we devised a risk model aimed at accurately forecasting the response to immunotherapy and TIME in LUAD. Furthermore, we demonstrated that ACSL3 is overexpressed in LUAD, which correlates with a poorer prognosis. Notably, combining ACSL3 knockdown with anti–PD-1 therapy impedes LUAD progression by upregulating the release of IFN-α, thereby bolstering the abundance of CD8+T while inhibiting the infiltration of immunosuppressive cell types M2-like macrophages.

Lipid metabolism disorders represent a prominent metabolic shift observed in cancer, where cancerous cells exploit this process to acquire energy, structural elements for cell membranes, as well as signaling factors essential for growth, viability, aggression, dissemination, and adaptation to the tumor’s surroundings and therapeutic interventions [20]. Tumor cells primarily induce lipid metabolic reprogramming by affecting the uptake, synthesis, and degradation of three major lipid molecules: fatty acids, phospholipids, and cholesterol [21]. ACSL3 plays a vital role in catalyzing the activation of fatty acids through esterification with coenzyme A, which is involved in the process of fatty acid oxidation (FAO) and lipogenesis [22]. ACSL3 is overexpressed in a wide spectrum of cancers [4], encompassing clear cell renal cell carcinoma, pancreatic cancer, prostate cancer, and lung cancer, which is intimately associated with a dire prognosis. Existing literature suggests ACSL3 promotes cancer progression via various pathways: Quan et al. [23] reported that transforming growth factor β1 up-regulated the expression of ACSL3, resulting in increased level of FAO, which was required for metastasis and epithelial-mesenchymal transition in colorectal cancer; Saliakoura et al. [24] showed that ACSL3 promoted the synthesis of prostaglandin, thereby facilitating the growth of non–small cell lung cancers. In this research, we also verified the high expression of ACSL3 in tumor tissues through clinical samples as well as common cell lines, and found that it was negatively associated with the OS of LUAD patients.

When it comes to the relation between ACSL3 and tumor immunity, Rossi Sebastiano et al. [15] demonstrated ACSL3 enhances fibrosis in pancreatic ductal adenocarcinoma by increasing the secretion of plasminogen activator inhibitor-1 in cancer cells, which contributes to an immunosuppressive TIME. However, there is a lack of investigation on the function of ACSL3 in antitumor immunity against LUAD. In our study, after confirming high expression of ACSL3 as a biomarker of worse prognosis, we investigated the role of ACSL3 in LUAD immunity. We found that individuals who exhibited non-responsiveness to immunotherapy displayed significantly elevated expression of ACSL3 compared to those who were responsive by detecting clinical samples. Additionally, the combination of ACSL3 knockdown and anti–PD-1 therapy led to a substantial decrease in tumor size, whereas ACSL3 knockdown alone had no impact on tumor size, suggesting that ACSL3 may serve as a prognostic biomarker for immunotherapy response in LUAD. Regarding NPAS2, numerous literature reports have indicated its overexpression in various types of tumors and its association with poor prognosis [25]. In our ongoing research, an initial bioinformatics analysis has yielded comparable results. Nevertheless, our analysis of clinical samples showed no correlation between NPAS2 levels and prognosis. Therefore, further investigation into this aspect was not pursued.

To investigate potential mechanisms, we conducted bulk RNA-seq. We identified that the IFN signaling was enriched in group 4 according to GO, ssGSEA and GSEA analysis. ELISA further confirmed an increased level of IFN-α. IFNs including type I, type II, and type III play important role in tumor progression and antitumor immunity [26]. IFN-α, which belongs to type I IFNs (IFN-Is), is the most used cytokine in patients [27]. IFN-α can be produced by various cells in the body, with pDCs are major IFN-α–producing cells [28]. In this study, there were more pDCs infiltration in group 4, which contributed a higher level of IFN-α. Additionally, pDCs are primarily activated through the Toll-like receptor (TLR) pathway (especially TLR7 and TLR9), leading to the production of IFN-Is and an increase in the expression of costimulatory molecules. In this study, the TLR pathway was activated in the group 4 (Fig. 5C), suggesting that knockdown of ACSL3 may promote the activation of pDCs by activating the TLR pathway, thereby enhancing the secretion of IFN-Is by pDCs to exert antitumor effects.

IFN-Is can modulate the functions of nearly all immune cells, encompassing T cells, B cells, dendritic cells, and macrophages [28]. In addition, IFN has been demonstrated to promote the polarization of macrophages into M1 type than immunosuppressive M2 macrophages [29]. By binding to IFNAR (IFN-α/β receptor), IFN-Is activate the JAK-STAT pathway, playing a role in anti-proliferation, immune regulation, and other functions [28]. In our study, KEGG analysis revealed a pronounced enrichment of the JAK-STAT pathway in group 4 when compared to the group 2. Additionally, group 4 displayed heightened infiltration of CD8+ T cells, along with a reduction in the infiltration of immunosuppressive M2 macrophages, in comparison to the group 2 according to IHC staining. Therefore, knockdown of ACSL3 improves the efficacy of immunotherapy by increasing the infiltration of pDCs, enhancing the secretion of IFN-α, thereby facilitating the infiltration of CD8+ T cells, decreasing the immunosuppressive M2 macrophages, ultimately promoting the formation of an antitumor TIME. Furthermore, a recent study reported that combination of IFN-α and PD-1 inhibitors can significantly improve the response to immunotherapy in patients with liver cancer [30]. According to our investigation, the synergistic application of IFN-α, anti–PD-1, and ACSL3 inhibitors holds the potential to enhance immunotherapeutic efficacy, thereby presenting a novel therapeutic avenue of hope for individuals suffering from LUAD.

Overall, this study developed an accurate prognostic risk model utilizing LMRGs, and identified the potential role of ACSL3 in LUAD progression and immunotherapy, thus offering novel perspective for enhancing immunotherapy strategies. However, our study has certain limitations. First, since the three chosen databases comprised different expression profiles, some risk genes were not obtained in the validation cohort, leading to deviations in the calculation of risk scores and a reduction in the efficiency of the test. Second, the combination of IFN-α, anti–PD-1, and ACSL3 inhibitors in vivo can improve the efficacy of immunotherapy have not been explored. Further research is required to validate the aforementioned results.

Electronic Supplementary Material

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

crt-2024-1119_S1_Table.pdf

crt-2024-1119_S2_Table.pdf

crt-2024-1119_S3_Fig.pdf

crt-2024-1119_S4_Table.pdf

crt-2024-1119_S5_Table.pdf

crt-2024-1119_S6_Fig.pdf

crt-2024-1119_S7_Table.pdf

crt-2024-1119_S8_Table.pdf

crt-2024-1119_S9_Table.pdf

crt-2024-1119_S10_Fig.pdf

crt-2024-1119_Supplementary_Methods.pdf

NOTES

Ethical Statement

All procedures performed in studies involving human participants were ethically approved by the Shanghai Pulmonary Hospital’s Medical Ethics Committee (Ethics approval No. K20-288). All samples were provided upon patient consent.

Author Contributions

Conceived and designed the analysis: He T, Hu J, Guo H, Wang Y, Wu Y, Cheng L, Zhao C, Li X, Zhou C.

Collected the data: He T, Hu J, Guo H, Wang Y, Cheng L.

Contributed data or analysis tools: He T, Hu J, Guo H, Diao M, Wu Y, Zhao C.

Performed the analysis: He T, Hu J, Guo H.

Wrote the paper: He T, Hu J, Diao M.

Administrated the study: Li X, Zhou C.

Conflict of Interest

Conflict of interest relevant to this article was not reported.

Funding

This study was granted by Key projects of the National Natural Science Foundation of China (82141101), Collaborative innovation project of Shanghai Municipal Health Commission (2020CXJQ02), National Natural Science Foundation of China (82102766).

Fig. 1.

Characterization of two subtypes identified by lipid metabolism-related genes (LMRGs). (A) Consensus clustering of patients based on the 147 prognostic genes generated, the optimal number of clusters K=2 was determined based on (B) relative change in area under cumulative distribution function (CDF) curve. (C) Survival curves of two patient subgroups. (D) Immune score of two patient subgroups. (E) TIDE (Tumor Immune Dysfunction and Exclusion) score statistics, higher scores are associated with greater probability of immune escape. CI, confidence interval; HR, hazard ratio. F) The xCell Statistical Analysis results of two subgroups. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^**** p < 0.0001.

Fig. 2.

Construction and verification of a risk signature model based on lipid metabolism-related genes. From the training cohort, survival curves of lung adenocarcinoma patients with different risk groups (A), immune score by ESTIMATE algorithm in different risk groups from the training cohort (B), receiver operating characteristic (ROC) curves for risk models constructed in the training cohort (C). On Verification Queue, survival curves for different risk groups (D), immune score calculated by ESTIMATE algorithm in both high and low risk (E). (F) ROC curves for risk models constructed in the validation cohort. AUC, area under the curve; CI, confidence interval; HR, hazard ratio. (G) Nomogram of combined clinical characteristics and risk scores in the training cohort. (H) Nomogram of combined clinical characteristics and risk scores in the validation cohort. (I) Nomogram calibration at 1, 3, and 5 years in the training cohort. (J) Nomogram calibration at 1, 3, and 5 years in the validation cohort. ^*p < 0.05, ^****p < 0.0001.

Fig. 3.

Acyl-CoA synthetase long-chain family member 3 (ACSL3) is overexpressed and associated with poor prognosis in lung adenocarcinoma (LUAD). (A, B) Kaplan-Meier curves of LUAD patients in The Cancer Genome Atlas (TCGA)–LUAD datasets utilizing Gene Expression Profiling Interactive Analysis website based on the expression of ACSL3 and neuronal PAS domain-containing protein 2 (NPAS2) respectively. (C, D) Expression of ACSL3 and NPAS2 in BEAS-2b, PC-9, H1975, 95D, and HCC827 cells. (E) Representative images of immunohistochemical (IHC) staining of ACSL3 and NPAS2 protein in the surgery cohort. Scale bars=25 μm. (F, G) Quantitative results of IHC staining of ACSL3 expression in tissue microarray (TMA) cohort, the survival curves are based on H-score, percentage of ACSL3-positive cells, and overall survival. (H, I) Expression of ACSL3 and NPAS2 in LLC, KP, CMT-167 cells. (J) Representative images of IHC staining of ACSL3 and NPAS2 protein in the immunotherapy cohort. Scale bars=25 μm. (K, L) Results of IHC staining of ACSL3 expression in the cohort of immunotherapy patients, the survival curves are based on H-score, mean density of ACSL3 and progression-free survival. (M, N) Quantitative presentation of IHC staining for ACSL3 between responder and non-responder group. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001.

Fig. 4.

Acyl-CoA synthetase long-chain family member 3 (ACSL3) knockdown improve the efficacy of immunotherapy of lung adenocarcinoma (LUAD). (A) The expression of ACSL3 protein in LLC, KL, KP, CMT-167 cells, and efficacy of shRNAs in suppressing expression of ACSL3 in LLC cells. (B) Treatment scheme of mice experiment. (C) Gross appearance of tumor xenografts in sh-NC and sh-ACSL3 groups (n=3). (D) The tumor weight on day 19 of each group. (E) Tumor growth curve of four groups. (F) The mice body weight curve of four groups. ^*p < 0.05, ^**p < 0.01; ns, not significant. Group 1: sh-NC+phosphate buffered saline (PBS); group 2: sh-NC+anti–programmed cell death protein 1 (PD-1); group 3: sh-ACSL3+PBS; group 4: sh-ACSL3+anti–PD-1.

Fig. 5.

The underlying mechanisms of knocking down acyl-CoA synthetase long-chain family member 3 (ACSL3) improves the efficacy of immunotherapy of lung adenocarcinoma (LUAD). (A) The Volcano map of group 4 compared to group 2. (B) The results of Gene Ontology (GO) Enrichment analysis between group 4 and group 2, the size of the circles represents gene count, the shades of color represent the magnitude of the p-values. (C) The results of Kyoto Encyclopedia of Genes and Genomes Enrichment analysis between group 4 and group 2, the size of the circles represents gene count, the shades of color represent the magnitude of the p-values. (D) The immune cell infiltration of group 2 and group 4 using the Single Sample Gene Set Enrichment Analysis (ssGSEA). (E) GSEA plot of IFN-α response based on the group 4 versus group 2 gene expression profiles. (F) GSEA plot of interferon alpha response based on the lung adenocarcinoma patients in The Cancer Genome Atlas database. ACSL3-low group versus ACSL3-high group. (G) Concentration of IFN-α in mouse plasma of group 2 and group 4, detected by enzyme-linked immunosorbent assay. (H) The representative images and quantitative presentation of immunohistochemical staining for CD4, CD8, CD163, and inducible nitric oxide synthase (iNOS), in group 2 and group 4. APC, antigen-presenting cell; DC, dendritic cell; EMT, epithelial–mesenchymal transition; FDR, false discovery rate; IFN, interferon; IL-17, interleukin 17; NES, normalized enrichment score; NK, natural killer; PD-1, programmed cell death protein 1; TCGA, The Cancer Genome Atlas; Th1 cell, T helper type 1 cell; Th2 cell, T helper type 2 cell; TIL, tumor-infiltrating lymphocyte; TNF, tumor necrosis factor; Tregs, regulatory T cells. Scale bars=50 μm. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001; ns, not significant.

Table 1.

Univariable Cox regression for PFS in the cohort of immunotherapy patients

Factor	No.	HR (95% CI)	p-value
Sex
Female	7	1.443 (0.481-4.328)	0.512
Male	22
Age (yr)
≤ 70	21	0.809 (0.293-2.237)	0.683
> 70	8
T category
1-2	9	1.320 (0.502-3.473)	0.574
3-4	20
N category
0-2	16	6.364 (2.212-18.309)	< 0.001
3	13
M category
0	3	2.077 (0.481-8.077)	0.327
1	26
Stage
1-3	4	2.077 (0.481-8.977)	0.327
4	25
Response to immunotherapy
Responders	9	8.190 (2.381-138.968)	0.005
Non-responders	20
Expression of ACSL3
Positive cells (%)
Low (≤ 98.18)	11	1.927 (0.69-5.382)	0.069
High (> 98.18)	18
Mean density
Low (≤ 0.1724)	11	2.688 (1.093-6.608)	0.048
High (> 0.1724)	18
H-score
Low (≤ 273.28)	23	5.025 (0.825-30.6)	< 0.001
High (> 273.28)	6
Expression of NPAS2
Positive cells (%)
Low (≤ 99.2)	18	0.644 (0.255-1.628)	0.352
High (> 99.2)	11
Mean density
Low (≤ 0.15)	15	1.892 (0.7688-4.655)	0.160
High (> 0.15)	14
H-score
Low (≤ 286.45)	19	1.444 (0.557-3.744)	0.409
High (> 286.45)	10

All HRs are based on the first value of each group as reference. ACSL3, long-chain acyl-coenzyme A synthase 3; CI, confidence interval; HR, hazard ratio; NPAS2, neuronal PAS domain-containing protein 2; PFS, progression-free survival.

REFERENCES

1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, 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. Article PubMed PDF
2. Cheng C, Geng F, Cheng X, Guo D. Lipid metabolism reprogramming and its potential targets in cancer. Cancer Commun (Lond). 2018;38:27.Article PubMed PMC PDF
3. Bao M, Shi R, Zhang K, Zhao Y, Wang Y, Bao X. Development of a membrane lipid metabolism-based signature to predict overall survival for personalized medicine in ccRCC patients. EPMA J. 2019;10:383–93. Article PubMed PMC PDF
4. Qian H, Lei T, Hu Y, Lei P. Expression of lipid-metabolism genes is correlated with immune microenvironment and predicts prognosis in osteosarcoma. Front Cell Dev Biol. 2021;9:673827.Article PubMed PMC
5. Merino Salvador M, Gomez de Cedron M, Moreno Rubio J, Falagan Martinez S, Sanchez Martinez R, Casado E, et al. Lipid metabolism and lung cancer. Crit Rev Oncol Hematol. 2017;112:31–40. Article PubMed
6. Yu W, Lei Q, Yang L, Qin G, Liu S, Wang D, et al. Contradictory roles of lipid metabolism in immune response within the tumor microenvironment. J Hematol Oncol. 2021;14:187.Article PubMed PMC PDF
7. Johnson WE, Li C, Rabinovic A. Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics. 2007;8:118–27. Article PubMed
8. Wilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26:1572–3. Article PubMed PMC PDF
9. Yoshihara K, Shahmoradgoli M, Martinez E, Vegesna R, Kim H, Torres-Garcia W, et al. Inferring tumour purity and stromal and immune cell admixture from expression data. Nat Commun. 2013;4:2612.Article PubMed PMC PDF
10. Zeng D, Ye Z, Shen R, Yu G, Wu J, Xiong Y, et al. IOBR: Multiomics immuno-oncology biological research to decode tumor microenvironment and signatures. Front Immunol. 2021;12:687975.Article PubMed PMC
11. Aran D, Hu Z, Butte AJ. xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome Biol. 2017;18:220.Article PubMed PMC PDF
12. Jiang P, Gu S, Pan D, Fu J, Sahu A, Hu X, et al. Signatures of T cell dysfunction and exclusion predict cancer immunotherapy response. Nat Med. 2018;24:1550–8. Article PubMed PMC PDF
13. Togashi Y, Shitara K, Nishikawa H. Regulatory T cells in cancer immunosuppression: implications for anticancer therapy. Nat Rev Clin Oncol. 2019;16:356–71. Article PubMed PDF
14. Wang Q, Li M, Yang M, Yang Y, Song F, Zhang W, et al. Analysis of immune-related signatures of lung adenocarcinoma identified two distinct subtypes: implications for immune checkpoint blockade therapy. Aging (Albany NY). 2020;12:3312–39. Article PubMed PMC
15. Rossi Sebastiano M, Pozzato C, Saliakoura M, Yang Z, Peng RW, Galie M, et al. ACSL3-PAI-1 signaling axis mediates tumor-stroma cross-talk promoting pancreatic cancer progression. Sci Adv. 2020;6:eabb9200PubMed PMC
16. Shen W, Song Z, Zhong X, Huang M, Shen D, Gao P, et al. Sangerbox: a comprehensive, interaction-friendly clinical bioinformatics analysis platform. Imeta. 2022;1:e36Article PubMed PMC PDF
17. Skoulidis F, Goldberg ME, Greenawalt DM, Hellmann MD, Awad MM, Gainor JF, et al. STK11/LKB1 mutations and PD-1 inhibitor resistance in KRAS-mutant lung adenocarcinoma. Cancer Discov. 2018;8:822–35. PubMed PMC
18. Li HY, McSharry M, Bullock B, Nguyen TT, Kwak J, Poczobutt JM, et al. The tumor microenvironment regulates sensitivity of murine lung tumors to PD-1/PD-L1 antibody blockade. Cancer Immunol Res. 2017;5:767–77. Article PubMed PMC PDF
19. Lee JS, Nair NU, Dinstag G, Chapman L, Chung Y, Wang K, et al. Synthetic lethality-mediated precision oncology via the tumor transcriptome. Cell. 2021;184:2487–502. Article PubMed PMC
20. Bian X, Liu R, Meng Y, Xing D, Xu D, Lu Z. Lipid metabolism and cancer. J Exp Med. 2021;218:e20201606Article PubMed PMC PDF
21. Fu Y, Zou T, Shen X, Nelson PJ, Li J, Wu C, et al. Lipid metabolism in cancer progression and therapeutic strategies. MedComm (2020). 2021;2:27–59. Article PubMed PMC PDF
22. Grevengoed TJ, Klett EL, Coleman RA. Acyl-CoA metabolism and partitioning. Annu Rev Nutr. 2014;34:1–30. Article PubMed PMC
23. Quan J, Cheng C, Tan Y, Jiang N, Liao C, Liao W, et al. Acyl-CoA synthetase long-chain 3-mediated fatty acid oxidation is required for TGFbeta1-induced epithelial-mesenchymal transition and metastasis of colorectal carcinoma. Int J Biol Sci. 2022;18:2484–96. Article PubMed PMC
24. Saliakoura M, Reynoso-Moreno I, Pozzato C, Rossi Sebastiano M, Galie M, Gertsch J, et al. The ACSL3-LPIAT1 signaling drives prostaglandin synthesis in non-small cell lung cancer. Oncogene. 2020;39:2948–60. Article PubMed PMC PDF
25. Yuan P, Li J, Zhou F, Huang Q, Zhang J, Guo X, et al. NPAS2 promotes cell survival of hepatocellular carcinoma by transactivating CDC25A. Cell Death Dis. 2017;8:e2704Article PubMed PMC PDF
26. Stanifer ML, Pervolaraki K, Boulant S. Differential regulation of type I and type III interferon signaling. Int J Mol Sci. 2019;20:1445.Article PubMed PMC
27. Pfeffer LM, Dinarello CA, Herberman RB, Williams BR, Borden EC, Bordens R, et al. Biological properties of recombinant alpha-interferons: 40th anniversary of the discovery of interferons. Cancer Res. 1998;58:2489–99. PubMed
28. Parker BS, Rautela J, Hertzog PJ. Antitumour actions of interferons: implications for cancer therapy. Nat Rev Cancer. 2016;16:131–44. Article PubMed PDF
29. Duluc D, Corvaisier M, Blanchard S, Catala L, Descamps P, Gamelin E, et al. Interferon-gamma reverses the immunosuppressive and protumoral properties and prevents the generation of human tumor-associated macrophages. Int J Cancer. 2009;125:367–73. PubMed
30. Jacoberger-Foissac C, Blake SJ, Liu J, McDonald E, Triscott H, Nakamura K, et al. Concomitant or delayed anti-TNF differentially impact on immune-related adverse events and antitumor efficacy after anti-CD40 therapy. J Immunother Cancer. 2020;8:e001687Article PubMed PMC

Figure & Data

REFERENCES

Citations

Citations to this article as recorded by

The multifaceted roles of the ACSL family in cancer: Metabolic reprogramming, ferroptosis regulation and tumour immune microenvironment remodelling
Haocai Li, Weijian Wang, Juncheng Zhan, Yuxiang Xiao, Xiaoping Chen, Chen Su, Peng Zhu
Clinical and Translational Medicine.2026;[Epub] CrossRef
Multiomic and functional validation of ACSL3, a regulator of fatty acid metabolism, as a lymph node metastasis-associated gene in lung adenocarcinoma
Yicheng Liang, Linchuan Liang, Minjun Du, Yangyang Lei, Yushun Gao, Shanqing Li
Respiratory Research.2026;[Epub] CrossRef
Vitamin D Mitigates Klebsiella pneumoniae‐Induced Pneumonia by Regulating Macrophage Polarization Through miR‐223/ACSL3 Axis‐Mediated Lipid Metabolism Reprogramming
Jiajia Hu, Yushen Lin, Wanrong Guo, Jingcong Zhang, Benquan Wu, Yanhong Wang
The Journal of Gene Medicine.2026;[Epub] CrossRef
Lipid Metabolism Reprogramming in Cancer: Insights into Tumor Cells and Immune Cells Within the Tumor Microenvironment
Rundong Liu, Chendong Wang, Zhen Tao, Guangyuan Hu
Biomedicines.2025; 13(8): 1895. CrossRef
Independent validation of lung adenocarcinoma prognostic risk scores incorporating cholesterol and estrogen metabolism related transcriptional biomarkers
Qian Zhu, Yuemei Zhang, Jian Ma, Yongjia Li, Hongya Liu, Zhongwen Gong, Ming Du, Xuemei Lian
Scientific Reports.2025;[Epub] CrossRef
Long-chain acyl-CoA synthetases: biological functions, diseases and therapeutic targets
Xiaoliang Deng, Yanqun Luo, Ying Gao, Tao Wu
Molecular Biomedicine.2025;[Epub] CrossRef

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

Lipid Metabolism Related Gene ACSL3 as a Biomarker for Predicting Immunotherapy Outcomes in Lung Adenocarcinoma

Cancer Res Treat. 2025;57(4):1000-1018. Published online January 20, 2025

DOI: https://doi.org/10.4143/crt.2024.1119

XML Download

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

Lipid Metabolism Related Gene ACSL3 as a Biomarker for Predicting Immunotherapy Outcomes in Lung Adenocarcinoma

Fig. 1. Characterization of two subtypes identified by lipid metabolism-related genes (LMRGs). (A) Consensus clustering of patients based on the 147 prognostic genes generated, the optimal number of clusters K=2 was determined based on (B) relative change in area under cumulative distribution function (CDF) curve. (C) Survival curves of two patient subgroups. (D) Immune score of two patient subgroups. (E) TIDE (Tumor Immune Dysfunction and Exclusion) score statistics, higher scores are associated with greater probability of immune escape. CI, confidence interval; HR, hazard ratio. F) The xCell Statistical Analysis results of two subgroups. *p < 0.05, **p < 0.01, ***p < 0.001, **** p < 0.0001.

Fig. 2. Construction and verification of a risk signature model based on lipid metabolism-related genes. From the training cohort, survival curves of lung adenocarcinoma patients with different risk groups (A), immune score by ESTIMATE algorithm in different risk groups from the training cohort (B), receiver operating characteristic (ROC) curves for risk models constructed in the training cohort (C). On Verification Queue, survival curves for different risk groups (D), immune score calculated by ESTIMATE algorithm in both high and low risk (E). (F) ROC curves for risk models constructed in the validation cohort. AUC, area under the curve; CI, confidence interval; HR, hazard ratio. (G) Nomogram of combined clinical characteristics and risk scores in the training cohort. (H) Nomogram of combined clinical characteristics and risk scores in the validation cohort. (I) Nomogram calibration at 1, 3, and 5 years in the training cohort. (J) Nomogram calibration at 1, 3, and 5 years in the validation cohort. *p < 0.05, ****p < 0.0001.

Fig. 3. Acyl-CoA synthetase long-chain family member 3 (ACSL3) is overexpressed and associated with poor prognosis in lung adenocarcinoma (LUAD). (A, B) Kaplan-Meier curves of LUAD patients in The Cancer Genome Atlas (TCGA)–LUAD datasets utilizing Gene Expression Profiling Interactive Analysis website based on the expression of ACSL3 and neuronal PAS domain-containing protein 2 (NPAS2) respectively. (C, D) Expression of ACSL3 and NPAS2 in BEAS-2b, PC-9, H1975, 95D, and HCC827 cells. (E) Representative images of immunohistochemical (IHC) staining of ACSL3 and NPAS2 protein in the surgery cohort. Scale bars=25 μm. (F, G) Quantitative results of IHC staining of ACSL3 expression in tissue microarray (TMA) cohort, the survival curves are based on H-score, percentage of ACSL3-positive cells, and overall survival. (H, I) Expression of ACSL3 and NPAS2 in LLC, KP, CMT-167 cells. (J) Representative images of IHC staining of ACSL3 and NPAS2 protein in the immunotherapy cohort. Scale bars=25 μm. (K, L) Results of IHC staining of ACSL3 expression in the cohort of immunotherapy patients, the survival curves are based on H-score, mean density of ACSL3 and progression-free survival. (M, N) Quantitative presentation of IHC staining for ACSL3 between responder and non-responder group. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fig. 4. Acyl-CoA synthetase long-chain family member 3 (ACSL3) knockdown improve the efficacy of immunotherapy of lung adenocarcinoma (LUAD). (A) The expression of ACSL3 protein in LLC, KL, KP, CMT-167 cells, and efficacy of shRNAs in suppressing expression of ACSL3 in LLC cells. (B) Treatment scheme of mice experiment. (C) Gross appearance of tumor xenografts in sh-NC and sh-ACSL3 groups (n=3). (D) The tumor weight on day 19 of each group. (E) Tumor growth curve of four groups. (F) The mice body weight curve of four groups. *p < 0.05, **p < 0.01; ns, not significant. Group 1: sh-NC+phosphate buffered saline (PBS); group 2: sh-NC+anti–programmed cell death protein 1 (PD-1); group 3: sh-ACSL3+PBS; group 4: sh-ACSL3+anti–PD-1.

Fig. 5. The underlying mechanisms of knocking down acyl-CoA synthetase long-chain family member 3 (ACSL3) improves the efficacy of immunotherapy of lung adenocarcinoma (LUAD). (A) The Volcano map of group 4 compared to group 2. (B) The results of Gene Ontology (GO) Enrichment analysis between group 4 and group 2, the size of the circles represents gene count, the shades of color represent the magnitude of the p-values. (C) The results of Kyoto Encyclopedia of Genes and Genomes Enrichment analysis between group 4 and group 2, the size of the circles represents gene count, the shades of color represent the magnitude of the p-values. (D) The immune cell infiltration of group 2 and group 4 using the Single Sample Gene Set Enrichment Analysis (ssGSEA). (E) GSEA plot of IFN-α response based on the group 4 versus group 2 gene expression profiles. (F) GSEA plot of interferon alpha response based on the lung adenocarcinoma patients in The Cancer Genome Atlas database. ACSL3-low group versus ACSL3-high group. (G) Concentration of IFN-α in mouse plasma of group 2 and group 4, detected by enzyme-linked immunosorbent assay. (H) The representative images and quantitative presentation of immunohistochemical staining for CD4, CD8, CD163, and inducible nitric oxide synthase (iNOS), in group 2 and group 4. APC, antigen-presenting cell; DC, dendritic cell; EMT, epithelial–mesenchymal transition; FDR, false discovery rate; IFN, interferon; IL-17, interleukin 17; NES, normalized enrichment score; NK, natural killer; PD-1, programmed cell death protein 1; TCGA, The Cancer Genome Atlas; Th1 cell, T helper type 1 cell; Th2 cell, T helper type 2 cell; TIL, tumor-infiltrating lymphocyte; TNF, tumor necrosis factor; Tregs, regulatory T cells. Scale bars=50 μm. *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Lipid Metabolism Related Gene ACSL3 as a Biomarker for Predicting Immunotherapy Outcomes in Lung Adenocarcinoma

Factor	No.	HR (95% CI)	p-value
Sex
Female	7	1.443 (0.481-4.328)	0.512
Male	22
Age (yr)
≤ 70	21	0.809 (0.293-2.237)	0.683
> 70	8
T category
1-2	9	1.320 (0.502-3.473)	0.574
3-4	20
N category
0-2	16	6.364 (2.212-18.309)	< 0.001
3	13
M category
0	3	2.077 (0.481-8.077)	0.327
1	26
Stage
1-3	4	2.077 (0.481-8.977)	0.327
4	25
Response to immunotherapy
Responders	9	8.190 (2.381-138.968)	0.005
Non-responders	20
Expression of ACSL3
Positive cells (%)
Low (≤ 98.18)	11	1.927 (0.69-5.382)	0.069
High (> 98.18)	18
Mean density
Low (≤ 0.1724)	11	2.688 (1.093-6.608)	0.048
High (> 0.1724)	18
H-score
Low (≤ 273.28)	23	5.025 (0.825-30.6)	< 0.001
High (> 273.28)	6
Expression of NPAS2
Positive cells (%)
Low (≤ 99.2)	18	0.644 (0.255-1.628)	0.352
High (> 99.2)	11
Mean density
Low (≤ 0.15)	15	1.892 (0.7688-4.655)	0.160
High (> 0.15)	14
H-score
Low (≤ 286.45)	19	1.444 (0.557-3.744)	0.409
High (> 286.45)	10

Table 1. Univariable Cox regression for PFS in the cohort of immunotherapy patients