Abstract
Background:
Mitochondria have always been considered too be closely related to the occurrence and development of malignant tumors. However, the bioinformatic analysis of mitochondria in lung adenocarcinoma (LUAD) has not been reported yet.
Methods:
In the present study, we constructed a novel and reliable algorithm, comprising a consensus cluster analysis and risk assessment model, to predict the survival outcomes and tumor immunity for patients with terminal LUAD.
Results:
Patients with LUAD were classified into three clusters, and patients in cluster 1 exhibited the best survival outcomes. The patients in cluster 3 had the highest expression of PDL1 (encoding programmed cell death 1 ligand 11) and HAVCR2 (encoding Hepatitis A virus cellular receptor 2), and the highest tumor mutation burden (TMB). In the risk assessment model, patients in the low-risk group tended to have a significantly better survival outcome. Furthermore, the risk score combined with stage could act as a reliable independent prognostic indicator for patients with LUAD. The prognostic signature is a novel and effective biomarker to select anti-tumor drugs. Low-risk patients tended to have a higher expression of CTLA4 (encoding cytotoxic T-lymphocyte associated protein 4) and HAVCR2. Moreover, patients in the high-risk group were more sensitive to Cisplatin, Docetaxel, Erlotinib, Gemcitabine, and Paclitaxel, while low-risk patients would probably benefit more from Gefitinib.
Conclusion:
We constructed a novel and reliable algorithm comprising a consensus cluster analysis and risk assessment model to predict survival outcomes, which functions as a reliable guideline for anti-tumor drug treatment for patients with terminal LUAD.
1 Introduction
As the most common malignant tumor worldwide, lung cancer is famous for its high mortality and high heterogeneity among malignant tumors (). Lung cancer has shown the highest estimated incidence and mortality in the United States for years (; ). Similarly, lung cancer has the highest incidence and mortality among patients in China (Xia et al., 2022). As the most common pathological classification of lung cancer, lung adenocarcinoma (LUAD) accounts for approximately 60% of lung cancer, and is considered to be closely related to heredity and gene mutations (Warth et al., 2012). Under appropriate conditions, surgical treatment, especially a video-assisted thoracic surgery, remains the gold standard for the treatment for LUAD, which could dramatically prolong the overall survival (OS) of patients (). Although the emergence of other therapeutic methods (e.g., molecular targeted therapy and immunotherapy) have improved the life quality of patients with terminal stage lung cancer (; ), the 5-year survival rate of patients with distant metastasis is only 7% (). Therefore, exploring the complex pathogenesis of LUAD and seeking novel and reliable biomarkers are important.
The mitochondrion is a membrane-enclosed structure that produces energy for fundamental cell activities (Tan et al., 2017), which is also involves in hepatic lipid metabolism and oxidative stress (; ). Recently, increasing evidence has demonstrated the crucial role of mitochondria in the occurrence and development of malignant tumors, and mitochondria might be an effective target for patients with cancer (Zhao Y. et al., 2013; ; Ubah and Wallace, 2014). For instance, Zhao J. et al. (2013b) proposed that mitochondria could promote the invasion and migration of malignancy by providing a large amount of adenosine triphosphate for pseudopodia. Furthermore, Villa et al. (2017) reported that mitochondria mediated the sensitivity of LUAD cells to chemotherapy drugs by regulating the autophagy signaling pathway. Moreover, Chang et al. revealed that dihydroergotamine tartrate, a drug used to treat migraine, acted on mitochondria in LUAD cells, thereby promoting apoptosis and mitochondrial autophagy (). Thus, mitochondria are involved in the biological behavior of malignant tumor cells, especially LUAD cells.
Mitochondria can not only meet the energy demand of fundamental cellular activities, but also effectively regulate immune activities (). Porporato et al. proposed that cancer cells could modify the tumor immune microenvironment (TIME) and the immune response of the host by releasing dangerous signals and altering the metabolism of mitochondria (). Moreover, reported that mitochondrial oxidative phosphorylation inhibitors targeted cancer-related immune cells in the TIME, and played a crucial part in immune evasion in the occurrence and progression of cancer. Furthermore, introduced the detailed role of mitochondria as sensors and mediators of innate immune receptor signaling. The precise coordination of oxidative stress between intracellular mitochondria and other organelles is crucial for cell survival. The dynamic balance of oxidative stress can not only coordinate complex cellular signaling events in cancer cells, but also affect other components of the tumor immune microenvironment (TIME). Immune cells, such as M2 macrophages, dendritic cells, and T cells, are the main components of immunosuppressive TMIE induced by oxidative stress (). Therefore, mitochondria are closely related to the immune activities of cells.
In recent years, immunotherapy has gradually become an effective tumor treatment strategy as an emerging tumor treatment strategy (). Unlike traditional radiotherapy and chemotherapy, immunotherapy is a treatment strategy that utilizes the human immune system to attack and eliminate cancer cells. It does not directly destroy tumor cells, but rather activates, enhances, or repairs the patient’s own immune system to recognize and kill tumor cells. The most common immune checkpoint molecules are Programmed Death Ligand 1 (PD-L1) and Hepatitis A Virus Cellular Receiver 2 (HAVCR-2) (). PD-L1 is an immune checkpoint protein that plays an important role in the immunotherapy of malignant tumors. It mainly inhibits T cell activity by binding to the Programmed Cell Death Protein 1 (PD-1) receptor, thereby reducing immune response and helping tumor cells evade immune surveillance. Therefore, inhibiting the interaction between PD-L1 or PD-1 can restore the activity of T cells and enhance the immune killing effect on tumor cells (; ). HAVCR-2, also known as T-cell Immunoglobulin and Mucin Domain 3 (TIM-3), is another important immune checkpoint molecule. It plays a crucial role in regulating the immune response process, especially in inhibiting T cell function. HAVCR-2 negatively regulates T cell activity by binding to its ligand, such as Galectin-9, and participates in regulating T cell depletion and immune tolerance phenomena. In immunotherapy, inhibition of HAVCR-2 is believed to enhance the anti-tumor effect of T cells, especially in patients who have failed treatment with PD-1/PD-L1 inhibitors and may play an important role. Therefore, HAVCR-2, as a potential target, is actively being studied and developed to expand and enhance the effectiveness of immunotherapy (; ; ).
The potential relationships among mitochondria, tumor immunity, and LUAD have been reported. For example, found that mitochondrial topoisomerase I was closely related to immune cells and the expression of immune checkpoint inhibitors (ICIs) in patients with LUAD. However, there has been no bioinformatic study of consensus cluster analysis combined with prognostic signature for patients with LUAD. In addition, mitochondria also participate in the expression of PD-L1 and HAVCR-2 in tumor cells, demonstrating an undeniable role in tumor occurrence and development. The latest research indicates that mitochondria are involved in the localization regulation of PD-L1 protein on the outer membrane and mitochondria. Enhancing mitochondrial autophagy helps to degrade mitochondrial localization PD-L1, thereby overcoming the resistance of TNBC to chemotherapy and immunotherapy (). The research results in this area are of great significance for enhancing the efficacy of targeted PD-1/PD-L1 therapy. In addition, studies have found that mitochondrial autophagy can enhance the therapeutic effect of ICI combined with paclitaxel by degrading mitochondrial distribution PD-L1, which can be inhibited by ATAD3A protein (Xie et al., 2023). Besides, in human colorectal cancer cancer cells, mitochondrial dysfunction inhibits the expression of HAVCR-2, thereby affecting the immune escape of tumor cells (). Thus, fully understanding the role of mitochondria in the development of LUAD might provide theoretical guidance and new strategies for future mitochondrial targeted therapy.
In the present study, we established a novel and reliable algorithm compromising molecular subtypes and a risk assessment model to predict prognosis and select sensitive anti-tumor drugs for patients with LUAD.
2 Materials and methods
2.1 Data download
The gene expression at the transcriptome level and corresponding clinical information were downloaded from LUAD project of The Cancer Genome Atlas database (https://portal.gdc.cancer.gov/). Subsequently, the mRNAs and long noncoding RNAs (lncRNAs) were annotated using gene transfer format (GTF) files obtained from Ensembl. A list of mitochondria-related genes (mrgenes) were downloaded from The Gene Ontology Resource (GO, http://geneontology.org/) (The Gene Ontology Consortium, 2019). The mitochondria-related lncRNAs (mrlncRNAs) were identified by performing a Spearman correlation analysis between genes related to mitochondria and lncRNAs (|cor| > 0.4, P < 0.001). Differentially expressed mrlncRNAs (DEmrlncRNAs) were filtered using differential expression analysis (|log FC| > 1, false discovery rate <0.001), and DEmrlncRNAs closely related to survival were screened using a univariate Cox analysis (P < 0.01), which were visualized using a volcano map and a forest map. We obtained mrlncRNAs that were closely related to the occurrence of LUAD and the OS of patients with LUAD, which were the foundation for the subsequent construction of the consensus cluster analysis and risk assessment model.
2.2 Molecular subtypes according to DEmrlncRNAs
The patients with LUAD were classified into different molecular subtypes based on the expression of DEmrlncRNAs by running the ConsensusClusterPlus package (Wilkerson and Hayes, 2010). Then, the survival outcomes of the patients with different molecular subtypes were explored by performing a Kaplan–Meier survival analysis. According to the National Comprehensive Cancer Network guidelines, the expression of common ICIs could reflect the reactivity of patients with LUAD to immunotherapy approximately, which could benefit a large number of patients with a terminal stage tumor (). The expression levels of common ICIs (e.g., PDL1 (encoding programmed cell death 1 ligand 1) and HAVCR2 [encoding Hepatitis A virus cellular receptor 2)] were compared between patients from different clusters, and a series of boxplots were generated for visualization, which were marked as: ***P < 0.001; **P < 0.01; and *P < 0.05. To better comprehend the relative abundance of stromal cells and immune cells in the TIME, the StromalScore, ImmuneScore, and ESTIMATEScore were calculated using the estimate package, which were subsequently compared between different clusters. The tumor mutation burden (TMB) represents the total number of mutations per million bases, which is used as a rough indicator of the efficacy of immunotherapy (; ; ). Several Wilcoxon rank-sum tests were performed to investigate whether there was a statistical difference in the TMB between patients from different clusters.
2.3 Risk assessment model based on DEmrlncRNAs
To better verify the predictive capability of the constructed risk assessment model, the patients with LUAD were equally divided into a training group and a test group randomly. To prevent over-fitting of the constructed model, the Least absolute shrinkage and selection operator regression combined with a multivariate Cox regression analysis were conducted on the DEmrlncRNAs of patients in the training group to construct a novel prognostic signature related to mitochondria. Then, the risk scores of patients in the training group were calculated using the following formula:
Where E(i) and Coef (i) are the expression and the regression coefficients of the DEmrlncRNAs, respectively. The median value of the risk score for patients in the training group was used as the cut-off point to classify patients into high- and low-risk groups. Subsequently, the regression coefficients and cut-off point of patients in the test group were determined to be in full accordance with those of the patients in train group. Kaplan–Meier survival analyses were conducted to exhibit the survival outcomes of patients in the different risk groups. To evaluate the predictive ability of the risk assessment model, the receiver operating characteristic curves were plotted and the area under the curve was calculated, respectively. To explore the potential relationship between the risk score and survival status, four scatter plots were plotted for visualization. Univariate and multivariate Cox regression analyses were conducted to investigate whether the risk assessment model could function as a reliable independent prognostic indicator for patients with LUAD, which was visualized using four forest maps. Furthermore, to filter the patients with LUAD whose prognosis could be predicted accurately using the risk assessment model, a series of Kaplan–Meier survival analyses were performed for validation. A clinical heatmap was used to exhibit the expression levels of 15 DEmrlncRNAs included in the modeling process, which revealed the potential relationships between the risk group and common clinicopathological characteristics [e.g., node (N), metastasis (M), tumor (T), stage, gender, age, ImmuneScore, and Cluster], in which the clinicopathological characteristics closely related to the risk groups were discussed in detail. Furthermore, to study the response of patients with LUAD to immunotherapy, the TMB and the expression levels of CTLA4 (encoding cytotoxic T-lymphocyte associated protein 4) and HAVCR2 in patients in the different risk groups were compared using Wilcoxon rank-sum tests. Moreover, we used the immunophenoscore (IPS), which represents gene expression levels in immune cells closely related to the tumor, including lymphocytes and macrophages, which has been utilized to assess the response to immunotherapy targeting PD-L1 and CTLA-4 (). The IPS of each patient was downloaded from The Cancer Immunome Atlas (https://tcia.at/) (), which were compared between different risk groups. The statistical differences in StromalScore, ImmuneScore, and ESTIMATEScore between the different risk groups were explored using Wilcoxon rank-sum tests. Single-sample gene-set enrichment analysis was used to quantify the relative abundance of common immune cells and the relative activity of common immune-related signaling pathways, which were compared between different risk groups. To explore the functions and signaling pathways closely related to the risk groups, two bar-plots were plotted for GO and Kyoto Encyclopedia of Genes and Genomes functional enrichment analyses for visualization. A Sankey diagram was generated to visualize the relationship between the molecular subtypes and the risk assessment model. The half maximal inhibitory concentration (IC50) represents the concentration of an anti-tumor drugs that inhibits half of the tumors cells, which could effectively measure the reaction of patients with LUAD to anti-tumor drugs (). In the present study, the pRRophetic package was run to evaluate the IC50 of common anti-tumor drugs including chemotherapy (e.g., Cisplatin, Docetaxel, Gemcitabine, and Paclitaxel) and molecular targeted therapy (e.g., Gefitinib and Erlotinib). A nomogram was plotted to display the calculation process of the risk scores for clinical patients, the predictive capability of which was evaluated using one-, three-, and five-year correction curves. Finally, a survival curve exhibited the survival outcomes of patients from different molecular subtypes and risk groups.
3 Results
3.1 Identification of DEmrlncRNAs
As shown in Figure 1, a multi-step approach was carried out according to the flowchart. We downloaded a total of 551 samples (497 LUAD tissues and 54 normal tissues) from The Cancer Genome Atlas database. Then, we downloaded 1,838 mrgenes from the GO knowledgebase, and obtained 14,087 lncRNAs. After annotation, 3,546 lncRNAs were identified as mrlncRNAs by performing correlation analyses. Next, we identified 1,724 DEmrlncRNAs by differential expression analysis, of which 276 were downregulated, and 1,448 were upregulated in patients with LUADs (Figure 2A). Then, 76 DEmrlncRNAs were identified as DEmrlncRNAs closely related to prognosis using a univariate Cox analysis, 15 of which were included in the modeling process (Figure 2B).
FIGURE 1
FIGURE 2
3.2 The molecular subtype is a reliable indicator for tumor immunity
When the k value was three, the variation of the cumulative distribution function (CDF) was the smallest, and the relative change in the area under CDF curve was the highest (Figures 2C–E). Therefore, the patients with LUAD were classified into three clusters, and the patients in cluster 1 exhibited the best survival outcomes among the clusters, with statistical significance (Figure 2F). According to the expression of ICIs, the patients in cluster 3 had the highest PDL1 (Figure 2G) and HAVCR2 (Figure 2H) expression, followed by the patients in cluster 1; the patients in cluster 2 had the lowest expression, which were all statistically significant. Furthermore, the patients in cluster 3 had the highest TMB (Figure 3A), which indicated that they might be most sensitive to immunotherapy targeting PD-L1 and HAVCR-2. The patients in cluster 2 had the lowest StromalScore, ImmuneScore, and ESTIMATEScore, which suggested that they had the lowest abundance of stromal cells and immune cells, while there was no statistical significance between the remaining two groups (Figures 2I–K).
FIGURE 3
3.3 The prognostic signature acts as a reliable biomarker for patients with LUAD
The risk score of patients with LUAD were calculated with the formula, and the regression coefficients of the DEmrlncRNAs were listed in Table 1. Patients (n = 236) were classified into the training group, while 232 patients were classified into the test group (Figures 3B, C), in which patients in the low-risk group tended to have a significantly better survival outcome (Figures 3D, E). The area under the curve values of the training group and the test group were 0.830 (Figure 3F) and 0.708 (Figure 3G), respectively, which suggested that the prognostic signature had a relatively better predictive capability for patients with LUAD. Furthermore, with increasing risk score, the number of patients who died increased significantly (Figures 3H, I). Moreover, the risk score (hazard ratio = 1.614 [confidence interval 1.383–1.884], P < 0.001) and stage (hazard ratio = 1.803 [confidence interval 1.456–2.231], P < 0.001), could act as reliable independent prognostic indicators for patients with LUAD based on univariate and multivariate Cox regression (Figures 4A–D). According to a series of survival analyses, the prognostic signature exhibited the best predictive ability among patients without distant metastasis (Figure 5D), regardless of age (Figure 4E), sex (Figure 4F), stage (Figure 5A), T (Figure 5B), and N (Figure 5C). The clinical heatmap (Figure 6A) revealed that the prognostic signature was closely related to N (Figure 6B), T (Figure 6C), stage (Figure 6D), ImmuneScore (Figure 6E), and molecular subtypes (Figure 6F). This indicated that high-risk patients tended to have a later stage of LUAD and always had relatively poor survival outcomes.
TABLE 1
| Gene | Coef |
|---|---|
| AC090559.1 | −0.0808422085604288 |
| AL353152.1 | −0.0128420568313749 |
| MIR223HG | −0.0183021082479998 |
| AC018647.1 | −1.20900113082152 |
| AC099850.4 | 0.011776662913192 |
| AC021755.3 | −0.0600844102402713 |
| LINC01116 | 0.030971078922924 |
| AC010999.2 | −0.948844016068309 |
| AL365181.2 | 0.0130238569092484 |
| AL049836.1 | 0.0103501622344718 |
| HIF1A-AS1 | 0.0823378698246388 |
| LINC02323 | 0.124240788617215 |
| TARID | 0.454314559680914 |
| LINC00578 | −0.0348236259612791 |
| AC027031.2 | 0.0131456798726987 |
The regression coefficients of mrlnRNAs included in the lasso regression.
FIGURE 4
FIGURE 5
FIGURE 6
The prognostic signature is a novel and effectively biomarker to select anti-tumor drugs. The low-risk patients tended to have a higher expression of CTLA4 (Figure 6G) and HAVCR2 (Figure 6H) with a lower TMB (Figure 6I), suggesting that they might be more sensitive to immunotherapy targeting CTLA-4 and HAVCR-2. According to the IPS value, low-risk patients would always benefit from anti-PD-L1 therapy (Figure 7A), anti-CTLA-4 therapy (Figure 7B), and their combination (Figure 7C), with statistical significance. The low-risk patients possessed a higher abundance of stromal cells and immune cells, based on the estimate algorithm (Figures 7D–F). Furthermore, the low-risk patients had a higher content of common immune cells (Figure 7G) and more active immune-related signaling pathways (Figure 7H). Therefore, compared with the high-risk patients, the patients in low-risk group tended to have stronger tumor immunity. According to the GO functional enrichment analysis, the mitochondria-related signature was closely related to the process of mitosis, including mitotic sister chromatid segregation, mitotic nuclear division, chromosome segregation, and sister chromatid segregation (Figure 7I). Similarly, the mitosis-related signaling pathways were enriched in the risk assessment model, such as mitotic sister chromatid segregation, mitotic nuclear division, chromosome segregation, nuclear division, and organelle fission (Figure 8A). The majority of high-risk patients were from cluster 2, the most of low-risk patients were from cluster 3, while 80% of the patients from cluster 1 were low-risk (Figure 8B).
FIGURE 7
FIGURE 8
The risk assessment model could function as a robust guideline for clinical medication using common chemotherapies and molecular targeted therapy. For example, patients in the high-risk group were more sensitive to Cisplatin (Figure 8C), Docetaxel (Figure 8D), Erlotinib (Figure 8E), Gemcitabine (Figure 8G), and Paclitaxel (Figure 8H), while low-risk patients would probably benefit more from Gefitinib (Figure 8F). The nomogram simplified the calculation process of the risk score, and provided the corresponding approximate one-, three-, and five-year survival rates based on the calculated risk score (Figure 9A), in which the nomogram exhibited the best predictive capability for 1-year survival (Figure 9B). According to the multi-survival curve, the high-risk patients in cluster 2 and cluster 3 exhibited poor survival outcomes, while the low-risk patients in cluster 1 had a relative survival advantage (Figure 9C).
FIGURE 9
4 Discussion
Changes in mitochondrial homeostasis are closely related to many human diseases, such as cancer, neurodegenerative diseases [Parkinson’s disease (), Alzheimer’s disease (Swerdlow, 2018), and Huntington’s disease ()], and myopathy (; Xie et al., 2020; ). Recently, the bioinformatics-based construction of prognostic signatures to predict the prognosis and guide treatment for patients with the terminal stage of malignant tumors has become a research hotspot (; ; Zhang et al., 2024). Zhuo et al. (2021) established a mitophagy-related signature to explore the survival outcomes, tumor immunity, mutation, and chemotherapy response in pancreatic cancer. In addition, Zhang et al. (2021) constructed a mitochondria-related signature to explore the TIME, infiltration of immune cells, and immunotherapy of patients with hepatocellular carcinoma.
In the present study, we identified DEmrlncRNAs related to prognosis, which were utilized for subsequent consensus cluster analysis and prognostic signature construction. The patients with LUAD were classified into three clusters, and patients in cluster 1 exhibited the best survival outcomes. The patients in cluster 3 had the highest expression of PDL1 and HAVCR2, and the highest TMB, which indicated that they might be the most sensitive to immunotherapy. The patients in cluster 2 had the lowest StromalScore, ImmuneScore, and ESTIMATEScore. According to the risk assessment model, patients in the low-risk group tended to have a significantly better survival outcome than the patients in the other groups. Furthermore, the risk score and stage could act as reliable independent prognostic indicators for patients with LUAD. The prognostic signature exhibited an excellent predictive ability among patients without distant metastasis, regardless of age, sex, stage, T, and N. The clinical heatmap revealed that the prognostic signature was closely related to N, T, stage, ImmuneScore, and molecular subtypes.
The prognostic signature is a novel and effective biomarker to select anti-tumor drugs. The low-risk patients tended to have a higher expression of CTLA4 and HAVCR2, with a lower TMB. According to the IPS value, the low-risk patients would always benefit from anti-PD-L1 therapy, anti-CTLA4 therapy, and their combination, which was significantly different. Furthermore, compared with the high-risk patients, the patients in the low-risk group tended to have a stronger tumor immunity. Furthermore, the risk assessment model could function as a robust guideline to select clinical medication comprising common chemotherapy and molecular targeted therapy. For example, patients in the high-risk group were more sensitive to Cisplatin, Docetaxel, Erlotinib, Gemcitabine, and Paclitaxel, while the low-risk patients would probably benefit more from Gefitinib. The high-risk patients in cluster 2 and cluster 3 exhibited poor survival outcomes, while the low-risk patients in cluster 1 showed a relative survival advantage.
The present study was the first to carry out bioinformatic analyses closely related to mitochondria and patients with LUAD. Furthermore, compared with traditional modeling process, we established a novel mitochondria-related algorithm containing a consensus cluster analyses and risk assessment model, in which patients with LUAD were classified into six groups that received different treatment strategies based on the corresponding groups. Moreover, least absolute shrinkage and selection operator regression analysis was carried out together with a multivariate Cox regression analysis to avoid overfitting of the model.
Although the algorithm might function as a guideline for clinical medication, there are also several limitations. Firstly, we conducted the internal verification using The Cancer Genome Atlas database, rather than other databases (e.g., GEO datasets). Secondly, all analyses were confined to bioinformatic analyses, and the study lacks validation of clinical specimens and molecular biological experiments, which are necessary for the clinical application of this algorithm.
Finally, based on the results of this manuscript, we have provided prospects for future potential research directions in this field. In recent years, the additional molecular pathways and incorporating multi-omics data integrations may play an increasingly important role in exploring the potential mechanisms of LUAD progression and treatment response. Nowadays, more and more research focuses on transcriptome sequencing. For example, studies based on transcriptome sequencing have demonstrated through in vitro and in vivo functional and mechanistic experiments that B4GALT1 promotes immune escape at both transcriptional and post transcriptional levels, thereby promoting the progression of LUAD (). In addition, there are also studies based on single-cell RNA sequencing analysis data, which calculate the immunogenic cell death value of cells to construct a set of prognostic models that can predict the prognosis of LUAD patients and immunotherapy, and to some extent guide the clinical treatment of LUAD patients (Zhang et al., 2023a). More interestingly, researchers explored the role of Treg cells in Esophageal squamous cell carcinoma by combining single-cell RNA sequencing and bulk RNA-seq analysis, in order to predict patient prognosis and immune therapy responsiveness as a prognostic model (Zhang et al., 2023b).
In the present study, we constructed a novel and reliable algorithm comprising a consensus cluster analysis and risk assessment model to predict the survival outcomes, which function as a reliable guideline to select anti-tumor drugs to treat patients with terminal LUAD, which might provide a theoretical foundation for customized individualized treatment.
Statements
Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.
Author contributions
XW: Validation, Writing–review and editing. HC: Formal Analysis, Software, Writing–original draft, Writing–review and editing. ZG: Software, Writing–original draft. BL: Software, Writing–original draft. HP: Validation, Writing–original draft. YS: Validation, Writing–original draft. ZX: Validation, Writing–original draft. CZ: Conceptualization, Writing–original draft, Funding acquisition.
Funding
The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. The study was funded by Basic public welfare project of Ningbo (Grant No. 2022S042).
Acknowledgments
We thank the TCGA database for generously sharing a large amount of data.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Abbreviations
LUAD, Lung Adenocarcinoma; PDL1, Drogrammed cell Death Ligand 1; HAVCR2, Hepatitis A Virus Cellular Receptor 2; TMB, tumor mutation burden; CTLA4, Cytotoxic T-Lymphocyte Associated protein 4; OS, overall survival; TIME, Tumor Immune Microenvironment; ICIs, Immune Checkpoint Inhibitors; lncRNAs, Long Noncoding RNAs; GTF, Gene Transfer Format; GO, Gene Ontology; mrRNA, Mitochondria-Related lncRNAs; DEmrlncRNAs, Differentially Expressed mrlncRNAs; IPS, Immunophenoscore; IC50, half maximal Inhibitory Concentration; CDF, Cumulative Distribution Function.
References
1
American Cancer Society (2024). Non-small cell lung cancer survival rates by stage[OL]. Available at: www.cancer.org/cancer/non-small-cell-lung-cancer/detection-diagnosis-staging/survival-rates.html (Date last updated: 04/20/21).
2
AnagnostouV.BruhmD.NiknafsN.WhiteJ.ShaoX.SidhomJ.et al (2020). Integrative tumor and immune cell multi-omic analyses predict response to immune checkpoint blockade in melanoma. Cell Rep. Med.1 (8), 100139. 10.1016/j.xcrm.2020.100139
3
AndersonA.JollerN.KuchrooV. (2016). Lag-3, tim-3, and TIGIT: Co-inhibitory receptors with specialized functions in immune regulation. Immunity44 (5), 989–1004. 10.1016/j.immuni.2016.05.001
4
AndréJ. (1994). Mitochondria. Biol. Cell80, 103–106. 10.1111/j.1768-322x.1994.tb00915.x
5
AndrieuxP.ChevillardC.Cunha-NetoE.NunesJ. (2021). Mitochondria as a cellular hub in infection and inflammation. Int. J. Mol. Sci.22 (21), 11338. 10.3390/ijms222111338
6
BanothB.CasselS. (2018). Mitochondria in innate immune signaling. Transl. Res.202, 52–68. 10.1016/j.trsl.2018.07.014
7
BolandM.ChourasiaA.MacleodK. (2013). Mitochondrial dysfunction in cancer. Front. Oncol.3, 292. 10.3389/fonc.2013.00292
8
ChangS.LeeA.YuK.ParkJ.KimK.ChoM. (2016). Dihydroergotamine tartrate induces lung cancer cell death through apoptosis and mitophagy. Chemotherapy61 (6), 304–312. 10.1159/000445044
9
CharoentongP.FinotelloF.AngelovaM.MayerC.EfremovaM.RiederD.et al (2020). Pan-cancer immunogenomic analyses reveal genotype-immunophenotype relationships and predictors of response to checkpoint blockade. Cell Rep.18 (1), 248–262. 10.1016/j.celrep.2016.12.019
10
ChenH.HuZ.SangM.NiS.LinY.WuC.et al (2021b). Identification of an autophagy-related lncRNA prognostic signature and related tumor immunity research in lung adenocarcinoma. Front. Genet.12, 767694. 10.3389/fgene.2021.767694
11
ChenH.ShenW.NiS.SangM.WuS.MuY.et al (2021a). Construction of an immune-related lncRNA signature as a novel prognosis biomarker for LUAD. Aging (Albany NY)13 (16), 20684–20697. 10.18632/aging.203455
12
CloonanS.ChoiA. (2013). Mitochondria: sensors and mediators of innate immune receptor signaling. Curr. Opin. Microbiol.16 (3), 327–338. 10.1016/j.mib.2013.05.005
13
CuiY.LiJ.ZhangP.YinD.WangZ.DaiJ.et al (2023). B4GALT1 promotes immune escape by regulating the expression of PD-L1 at multiple levels in lung adenocarcinoma. J. Exp. Clin. Cancer Res.42, 146. 10.1186/s13046-023-02711-3
14
DasM.ZhuC.KuchrooV. (2017). Tim-3 and its role in regulating anti-tumor immunity. Immunol. Rev.276 (1), 97–111. 10.1111/imr.12520
15
DongJ.LiB.LinD.ZhouQ.HuangD. (2019). Advances in targeted therapy and immunotherapy for non-small cell lung cancer based on accurate molecular typing. Front. Pharmacol.10, 230. 10.3389/fphar.2019.00230
16
EttingerD.WoodD.AisnerD.AkerleyW.BaumanJ.BharatA.et al (2021). NCCN guidelines insights: non-small cell lung cancer, version 2.2021. J. Natl. Compr. Canc Netw.19 (3), 254–266. 10.6004/jnccn.2021.0013
17
FeiL.LuZ.XuY.HouG. (2022). A comprehensive pan-cancer analysis of the expression characteristics, prognostic value, and immune characteristics of TOP1MT. Front. Genet.13, 920897. 10.3389/fgene.2022.920897
18
FordeP.ChaftJ.SmithK.AnagnostouV.CottrellT.HellmannM.et al (2018). Neoadjuvant PD-1 blockade in resectable lung cancer. N. Engl. J. Med.378 (21), 1976–1986. 10.1056/NEJMoa1716078
19
HerbstR.SoriaJ.KowanetzM.FineG.HamidO.GordonM.et al (2014). Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature515 (7528), 563–567. 10.1038/nature14011
20
JabbarzadehK.SalimianF.AghapourS.XiangS.ZhaoQ.LiM.et al (2020). Akt-targeted therapy as a promising strategy to overcome drug resistance in breast cancer - a comprehensive review from chemotherapy to immunotherapy. Pharmacol. Res.156, 104806. 10.1016/j.phrs.2020.104806
21
Jodeiri FarshbafM.GhaediK. (2017). Huntington's disease and mitochondria. Neurotox. Res.32 (3), 518–529. 10.1007/s12640-017-9766-1
22
KimJ. (2017). Myopathy, drugs, and mitochondria. J. Korean Med. Sci.32 (11), 1732–1733. 10.3346/jkms.2017.32.11.1732
23
KleinK.HeK.YounesA.BarsoumianH.ChenD.OzgenT.et al (2020). Role of mitochondria in cancer immune evasion and potential therapeutic approaches. Front. Immunol.11, 573326. 10.3389/fimmu.2020.573326
24
KuoC.BabuharisankarA. P.LinY.LienH.LoY.ChouH.et al (2022). Mitochondrial oxidative stress in the tumor microenvironment and cancer immunoescape: foe or friend?J. Biomed. Sci.29 (1), 74. 10.1186/s12929-022-00859-2
25
MacdonaldR.BarnesK.HastingsC.MortiboysH. (2018). Mitochondrial abnormalities in Parkinson's disease and Alzheimer's disease: can mitochondria be targeted therapeutically?Biochem. Soc. Trans.46 (4), 891–909. 10.1042/BST20170501
26
MansouriA.GattolliatC.AsselahT. (2018). Mitochondrial dysfunction and signaling in chronic liver diseases. Gastroenterology155 (3), 629–647. 10.1053/j.gastro.2018.06.083
27
MarabelleA.FakihM.LopezJ.ShahM.Shapira-FrommerR.NakagawaK.et al (2020). Association of tumour mutational burden with outcomes in patients with advanced solid tumours treated with pembrolizumab: prospective biomarker analysis of the multicohort, open-label, phase 2 KEYNOTE-158 study. Lancet Oncol.21 (10), 1353–1365. 10.1016/S1470-2045(20)30445-9
28
MerchantN.McKennaR.OnughaO. (2018). Is there a role for VATS sleeve lobectomy in lung cancer?Surg. Technol. Int.32, 225–229.
29
OliverA. (2022). Lung cancer: epidemiology and screening. Surg. Clin. N. Am.102 (3), 335–344. 10.1016/j.suc.2021.12.001
30
PardollD. (2012). The blockade of immune checkpoints in cancer immunotherapy. Nat. Rev. Cancer12 (4), 252–264. 10.1038/nrc3239
31
PilottoS.Molina-VilaM.KarachaliouN.CarbogninL.ViteriS.González-CaoM.et al (2015). Integrating the molecular background of targeted therapy and immunotherapy in lung cancer: a way to explore the impact of mutational landscape on tumor immunogenicity. Transl. Lung Cancer Res.4 (6), 721–727. 10.3978/j.issn.2218-6751.2015.10.11
32
PorporatoP.FilighedduN.PedroJ.KroemerG.GalluzziL. (2018). Mitochondrial metabolism and cancer. Cell Res.28 (3), 265–280. 10.1038/cr.2017.155
33
PriorF.ClarkK.CommeanP.FreymannJ.JaffeC.KirbyJ.et al (2013). TCIA: an information resource to enable open science. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc.2013, 1282–1285. 10.1109/EMBC.2013.6609742
34
RibasA.WolchokJ. (2018). Cancer immunotherapy using checkpoint blockade. Science359 (6382), 1350–1355. 10.1126/science.aar4060
35
SakhnevychS.YasinskaI.Fasler-KanE.SumbayevV. (2019). Mitochondrial defunctionalization supresses tim-3-galectin-9 secretory pathway in human colorectal cancer cells and thus can possibly affect tumor immune escape. Front. Pharmacol.10, 342. 10.3389/fphar.2019.00342
36
SakuishiK.ApetohL.SullivanJ.BlazarB.KuchrooV.AndersonA. (2010). Targeting Tim-3 and PD-1 pathways to reverse T cell exhaustion and restore anti-tumor immunity. J. Exp. Med.207 (10), 2187–2194. 10.1084/jem.20100643
37
SebaughJ. (2011). Guidelines for accurate EC50/IC50 estimation. Pharm. Stat.10 (2), 128–134. 10.1002/pst.426
38
SiegelR.MillerK.FuchsH.JemalA. (2021). Cancer statistics, 2021. CA-Cancer J. Clin.71 (1), 7–33. 10.3322/caac.21654
39
SiegelR.MillerK.FuchsH.JemalA. (2022). Cancer statistics, 2022. CA-Cancer J. Clin.72 (1), 7–33. 10.3322/caac.21708
40
SunshineJ.TaubeJ. (2105). PD-1/PD-L1 inhibitors. Curr. Opin. Pharmacol.23, 32–38. 10.1016/j.coph.2015.05.011
41
SwerdlowR. (2018). Mitochondria and mitochondrial cascades in Alzheimer's disease. J. Alzheimers Dis.62 (3), 1403–1416. 10.3233/JAD-170585
42
TanK.LiC.LiY.FeiJ.YangB.FuY.et al (2017). Real-time monitoring atp in mitochondrion of living cells: a specific fluorescent probe for atp by dual recognition sites. Anal. Chem.89 (3), 1749–1756. 10.1021/acs.analchem.6b04020
43
The Gene Ontology Consortium (2019). The gene Ontology resource: 20 years and still GOing strong. Nucleic Acids Res.47, D330–D338. 10.1093/nar/gky1055
44
UbahO.WallaceH. (2014). Cancer therapy: targeting mitochondria and other sub-cellular organelles. Curr. Pharm. Des.20 (2), 201–222. 10.2174/13816128113199990031
45
VillaE.ProïcsE.Rubio-PatiñoC.ObbaS.ZuninoB.BossowskiJ.et al (2017). Parkin-independent mitophagy controls chemotherapeutic response in cancer cells. Cell Rep.20 (12), 2846–2859. 10.1016/j.celrep.2017.08.087
46
WarthA.MuleyT.MeisterM.StenzingerA.ThomasM.SchirmacherP.et al (2012). The novel histologic International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society classification system of lung adenocarcinoma is a stage-independent predictor of survival. J. Clin. Oncol.30 (13), 1438–1446. 10.1200/JCO.2011.37.2185
47
WilkersonM.HayesD. (2010). ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics26 (12), 1572–1573. 10.1093/bioinformatics/btq170
48
XiaC.DongX.LiH.CaoM.SunD.HeS.et al (2022). Cancer statistics in China and United States, 2022: profiles, trends, and determinants. Chin. Med. J-Peking135 (5), 584–590. 10.1097/CM9.0000000000002108
49
XieX.YangY.WangQ.LiuH.FangX.LiC.et al (2023). Targeting ATAD3A-PINK1-mitophagy axis overcomes chemoimmunotherapy resistance by redirecting PD-L1 to mitochondria. Cell Res.33 (3), 215–228. 10.1038/s41422-022-00766-z
50
XieY.LiuJ.KangR.TangD. (2020). Mitophagy receptors in tumor biology. Front. Cell Dev. Biol.8, 594203. 10.3389/fcell.2020.594203
51
ZhangP.DongS.SunW.ZhongW.XiongJ.GongX.et al (2023b). Deciphering Treg cell roles in esophageal squamous cell carcinoma: a comprehensive prognostic and immunotherapeutic analysis. Front. Mol. Biosci.10, 1277530. 10.3389/fmolb.2023.1277530
52
ZhangP.PeiS.ZhouG.ZhangM.ZhangL.ZhangZ. (2024). Purine metabolism in lung adenocarcinoma: a single-cell analysis revealing prognostic and immunotherapeutic insights. J. Cell Mol. Med.28, e18284. 10.1111/jcmm.18284
53
ZhangP.ZhangH.TangJ.RenQ.ZhangJ.ChiH.et al (2023a). The integrated single-cell analysis developed an immunogenic cell death signature to predict lung adenocarcinoma prognosis and immunotherapy. Aging (Albany NY)15, 10305–10329. 10.18632/aging.205077
54
ZhangT.NieY.GuJ.CaiK.ChenX.LiH.et al (2021). Identification of mitochondrial-related prognostic biomarkers associated with primary bile acid biosynthesis and tumor microenvironment of hepatocellular carcinoma. Front. Oncol.11, 587479. 10.3389/fonc.2021.587479
55
ZhaoJ.ZhangJ.YuM.XieY.HuangY.WolffD.et al (2013b). Mitochondrial dynamics regulates migration and invasion of breast cancer cells. Oncogene32 (40), 4814–4824. 10.1038/onc.2012.494
56
ZhaoY.ButlerE.TanM. (2013a). Targeting cellular metabolism to improve cancer therapeutics. Cell Death Dis.4 (3), e532. 10.1038/cddis.2013.60
57
ZhuoZ.LinH.LiangJ.MaP.LiJ.HuangL.et al (2021). Mitophagy-related gene signature for prediction prognosis, immune scenery, mutation, and chemotherapy response in pancreatic cancer. Front. Cell Dev. Biol.9, 802528. 10.3389/fcell.2021.802528
Summary
Keywords
mitochondria, tumor immunity, consensus cluster, risk assessment model, lung adenocarcinoma
Citation
Wu X, Chen H, Ge Z, Luo B, Pan H, Shen Y, Xie Z and Zhou C (2024) A novel mitochondria-related algorithm for predicting the survival outcomes and drug sensitivity of patients with lung adenocarcinoma. Front. Mol. Biosci. 11:1397281. doi: 10.3389/fmolb.2024.1397281
Received
08 March 2024
Accepted
26 July 2024
Published
08 August 2024
Volume
11 - 2024
Edited by
Federico Ávila-Moreno, National Autonomous University of Mexico, Mexico
Reviewed by
Pengpeng Zhang, Nanjing Medical University, China
Zebo Jiang, Zhuhai Hospital of Integrated Traditional Chinese and Western Medicine, China
Updates
Copyright
© 2024 Wu, Chen, Ge, Luo, Pan, Shen, Xie and Zhou.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Chengwei Zhou, fyzhouchengwei@nbu.edu.cn
† These authors have contributed equally to this work
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.