Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment features

Fan Rong; Bin Cheng; Ling Guo; Shaobo Zeng; Xunliang Xu; Zhongji Meng

doi:10.1371/journal.pone.0325033

. 2025 Jun 2;20(6):e0325033. doi: 10.1371/journal.pone.0325033

Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment features

Fan Rong ^1,^#, Bin Cheng ^1,^#, Ling Guo ¹, Shaobo Zeng ², Xunliang Xu ³, Zhongji Meng ^1,^*

Editor: Matthew Cserhati⁴

PMCID: PMC12129190 PMID: 40455796

Abstract

Background

Mitochondrial DNA (mtDNA) is an important genetic material in eukaryotic cells. Mitochondrial DNA maintenance-related gene (mtDNA MRG) variants contribute to mitochondrial dysfunction in cancer progression and are associated with cancer prognosis. However, the mechanism of mtDNA MRGs in the tumor microenvironment (TME) of hepatocellular carcinoma (HCC) remains unclear.

Methods

Data for a total of 487 HCC samples were collected from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO). The mitochondrial regulatory pathway gene set was downloaded, and 22 mtDNA MRGs were identified by screening. Based on these 22 genes, the HCC samples were grouped by unsupervised clustering based on a machine learning model. Principal component analysis (PCA) was used to construct the mtDNA score model, and the relationships between the mtDNA score and clinicopathological features, tumor mutation burden (TMB), TME cell infiltration and biological processes were analyzed.

Results

The expression of 22 mtDNA MRGs significantly different in HCC samples vs. normal controls. In this study, HCC samples were divided into three molecular subtypes based on the expression of mtDNA MRGs. The three subtypes exhibit different clinical characteristics and immune infiltration profiles, and the cell infiltration profiles corresponded to the immune rejection, immune inflammation, and immune-desert phenotypes, respectively. A total of 740 core genes were obtained from different molecular subtypes, and these genes were divided into three gene subtypes. The mtDNA score model, which can be used to assess tumor immune cell invasion, clinicopathological features, genetic variation, and prognosis, was subsequently constructed. A high mtDNA score was associated with a high mutation burden, high clinical stage and poor prognosis.

Conclusions

mtDNA MRGs play important roles in HCC TMB, prognosis, clinicopathological features and the immune microenvironment. The mtDNA score can be used to evaluate HCC prognosis, TMB and the immune microenvironment, thereby providing guidance for treatment decision making and prognosis prediction in HCC patients.

Introduction

Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer, accounting for approximately 90% of all primary liver cancers [1]. It is also one of the top four causes of cancer-related death worldwide; 700,000 new cases are diagnosed worldwide each year, half of which are in China [2]. Liver transplantation and early surgical resection can improve the prognosis of HCC [3]. However, because the early symptoms of HCC patients are not typical, most HCC patients are often diagnosed at a late stage, and the mortality rate of HCC patients is still unusually high [4]. Therefore, there is an urgent need to elucidate the pathogenesis of HCC, identify potential biomarkers for early diagnosis, and develop novel therapies for HCC patients.

Mitochondria are ubiquitous in almost all eukaryotic cells and are mainly involved in oxidative phosphorylation (OXPHOS), reactive oxygen species (ROS) production, the tricarboxylic acid (TCA) cycle, heme synthesis, amino acid metabolism, apoptosis regulation, innate immunity, calcium homeostasis and other important processes [5,6]. mtDNA is a double-stranded closed circular DNA molecule composed of 16,569 base pairs that encodes 13 peptides in humans, as well as 22 tRNAs and 2 rRNAs required for mitochondrial protein synthesis [7,8]. mtDNA replication, repair, modification and stability maintenance play important roles in the maintenance of mtDNA homeostasis, aiming to maintain the integrity of mitochondrial structure and function [9]. Genes involved in mtDNA homeostasis maintenance are collectively referred to as mtDNA maintenance-related genes (mtDNA MRGs). Recent studies have shown that mtDNA MRG variants cause mitochondrial defects and dysfunction and lead to various mitochondrial diseases [10]. Mitochondrial dysfunction leads to OXPHOS disruption and increases ROS production, which is involved in HCC progression [11]. Moreover, mitochondrial dysfunction is associated with tumor chemotherapy resistance [12].

The tumor microenvironment (TME) also plays a considerable role in tumor cell progression [13], regulating tumor cell proliferation, invasion, immune escape and the response to immunotherapy through various signaling pathways [14]. The characteristics of cell infiltration in the TME can predict the response to immune checkpoint blockade (ICB) therapy [15]. A comprehensive analysis of the heterogeneity and complexity of the TME landscape may reveal different tumor immunophenotypes, provide an important basis for guiding treatment and predicting the immunotherapy response, and help to identify new therapeutic targets.

mtDNA MRGs have been identified as participants in cancer occurrence and progression [16]. For example, mitochondrial ND3, ND4, and ND5 mutation-mediated ROS elevation is a key factor in cancer pathogenesis and the generation of cancer-promoting phenotypes [17,18]. Increasing evidence shows that mitochondrial dysfunction is closely related to the occurrence and progression of HCC. High expression of APEX1 is associated with resistance to sorafenib and anti-programmed death 1 (PD-1) therapy in HCC patients [19]. The expression level of OGG1 in HCC patients is higher than that in healthy people. OGG1 promotes the occurrence of liver cancer by promoting the expression of cell cycle-related proteins in tumor cells and the repair of oxidative DNA damage [20]. SSBP1 is a key protein in mtDNA replication and is involved in tissue-specific mtDNA wasting diseases. Most of these studies focused on a single mtDNA gene [21]. However, the complex interactions among mtDNA MRGs in the development and progression of HCC are unclear.

In this study, we focused on exploring the important role of mtDNA MRGs in the HCC TME, constructed a mtDNA score model, and evaluated its value in predicting HCC prognosis, TMB, immune microenvironment features, and immunotherapy response. This study may provide more information on the molecular mechanism and prognostic biomarkers of HCC.

Materials and methods

Data acquisition and preprocessing

The HCC RNA transcriptome sequencing data and clinical information were obtained from The Cancer Genome Atlas (TCGA) (http://cancergenome.nih.gov/) and Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/). Data were obtained for 374 HCC samples and 50 normal samples from the TCGA on June 17, 2023, and 115 primary HCC tissues and 52 adjacent tissues from the GEO (GSE76427) cohort [22]. TCGA RNA sequencing data (FPKM format) were downloaded from the Genomic Data Commons (GDC, https://portal.gdc.cancer.gov/) and then used the fpkm function of the “limma” package in R to convert the FPKM value of the RNA data to the TPM value. For the GSE76427 data from the Illumina platform, the normalized matrix file was directly downloaded [23]. Patients without survival information were excluded. The list of mtDNA MRGs was obtained from the MitoCarta3.0database(https://www.broadinstitute.org/mitocarta/mitocarta30-inventory-mammalian-mitochondrial-proteins-and-pathways), and then, the mtDNA MRGs were obtained for integration and batch correction, the performance of oPLS-DA model was assessed by R²Y (goodness of fit) and Q² (goodness of prediction) through cross-validation, along with permutation testing to evaluate robustness. Data were downloaded from the publicly available database hence it was not applicable for additional ethical approval.

Unsupervised clustering of mtDNA MRGs

A total of 22 mtDNA MRGs were identified in this study and are listed as follows: mtDNA replication (DNA2, EXOG, LIG3, MGME1, POLG, POLG2, POLRMT, RNASEH1, SSBP1, TFAM, TFB2M, TOP1MT, TOP3A, TWNK), mtDNA repair (APEX1, EXOG, LIG3, OGG1, POLB, POLG, PRIMPOL, RECQL4, UNG), mtDNA modifications (METTL4), mtDNA stability and decay (ENDOG, EXOG, MGME1). Based on the mtDNA MRGs expression levels, HCC samples were classified for further analysis by using the “ConsensusClusterPlus” package in R for unsupervised cluster analysis, and 50 repetitions were performed with pltem = 0.8, pFeature = 1 to verify the stability of the subtypes. The number of cluster k-values were increased from 2 to 9, the k-values with better clustering stability were selected according to the clustering effect [24].Kaplan–Meier curves were used to assess the overall time to survival (OS) of patients with different clusters.

Gene set variation analysis (GSVA) and functional annotation

GSVA is a non-parametric, unsupervised method that is mainly used to estimate pathway changes in experimental datasets and differences in activity during biological processes [25]. To further explore the biological functions and signaling mechanisms between mtDNA MRG clusters, “GSVA” in the R package was used for GSVA enrichment analysis. “c2.cp.kegg.v7.2.symbols.gmt” was downloaded from the MSigDB database (http://www.gsea-msigdb.org/gsea/msigdb). The p value was adjusted according to the false discovery rate (FDR), with adjusted P < 0.05 as the cut-off criterion.

Analysis of immune cell differences in the TME

To elucidate the intricate mechanisms underlying tumor-immune cell interactions, Pornpimol Charoentong, et al characterized the intratumoral immune landscapes and the cancer antigenomes from 20 solid cancers and created The Cancer Immunome Atlas (https://tcia.at/). Using machine learning, the determinants of tumor immunogenicity were identified and a quantitative scoring scheme called immunophenotypic scoring was developed. The gene set of TME-infiltrating immune cells obtained from Pornpimol Charoentong’s study included activated B cells, activated CD4 T cells, natural killer T cells and regulatory T cells [26]. The single sample gene set enrichment analysis (ssGSEA) algorithm was used to quantify the relative abundance of individual infiltrating cells in the HCC TME. The relative abundance of each immune cell type was represented by an enrichment score in ssGSEA analysis and normalized to unity distribution from 0 to 1 [27]. The degree of infiltration of immune cells in each sample was indicated by the enrichment fraction calculated via ssGSEA.

Identification of mtDNA MRG gene clusters

According to the consensus clustering algorithm, 22 mtDNA MRGs were divided into three different subgroups. Based on the above three subgroups of mtDNA MRGs, the R package “limma” was used to screen differentially expressed genes (DEGs), and an adjusted p value less than 0.001 was used as the screening criterion. The Venn diagram package was used to identify overlapping genes between different subgroups, and KEGG enrichment analysis was subsequently performed to explore the potential related biological functions and biological pathways.

Consensus ClusterPlus cluster analysis in R was used to identify three gene clusters, the Kaplan‒Meier method was used to analyze the prognosis for groups enriched in the three gene clusters, and the log-rank test was used to evaluate the statistical significance. The “pheatmap” package was used to create a heatmap showing the relationships between gene clusters and clinicopathological features.

mtDNA score

To quantify the characteristics of mtDNA MRGs in individual HCC patients, a scoring system called the mtDNA score was constructed using principal component analysis (PCA). Combined with previous studies, core genes were extracted from DEGs, and a univariate Cox regression model was used to assess prognosis according to each gene in the feature model. Then, PCA was performed to extract principal components, and principal components 1 and 2 were selected to construct the signature score model. This approach focuses on the scores of the set with the largest number of well-correlated (or inversely correlated) gene blocks in the set, while eliminating as much as possible the contributions of genes that do not track other members [27]. We adopted a formula similar to previous studies to define the mtDNA score [28]:

\[m t D N A s c o r e = \sum (P C 1 i + P C 2 i) \]

i represents the expression of mtDNA phenotype-associated genes. Patients were divided into high and low mtDNA score groups based on maximum choice ranking statistics.

Correlation of the mtDNA score with the TMB, clinicopathological features, and immunotherapy response status

According to the mtDNA score, the samples were divided into a high mtDNA score group and a low mtDNA score group. A waterfall map was drawn using the “maftools” package to show the difference in mutation rate between the high-mtDNA score group and low-mtDNA score group. Differences in the mtDNA score according to pathological stage and survival status were analyzed. The percentage chart was drawn using the “plyr” package, and the box plot was drawn using the “ggpub” package. The “limma” package was used to compare mtDNA score differences between different mtDNA MRG cluster subgroups and gene subgroups. The immune checkpoint inhibitor (ICI) immunophenotype score (IPS) was downloaded from the cancer immune group database, and differences in the response to CTLA-4 and PD-1 blockade immunotherapies between groups were analyzed.

Statistical analysis

Spearman’s and distance correlation analyses were used to determine the correlation coefficient between TME infiltration of immune cells and mtDNA MRG expression. One-way ANOVA and the Kruskal‒Wallis test were used to analyze differences between three or more groups. Survival curves for prognostic analyses were generated using the Kaplan‒Meier method, and logarithmic rank tests were used to determine the significance of differences. All the statistical analyses were performed using R version 4.1.0. P < 0.05 was considered to indicate statistical significance.

Results

Genetic variation and transcriptional overview of mtDNA MRGs in HCC

In this study, to ensure data reliability, we first performed batch correction on the dataset (S1A Figand B). The orthogonal partial least squares-discriminant analysis (oPLS-DA) model was then applied to assess residual batch effects (S1C Fig, D). The oPLS-DA model demonstrated high efficacy, with an R²Y of 0.767, indicating that 76.7% of the sample classification variance could be explained by the model. The cross-validation predictive ability (Q²Y = 0.672, threshold >0.5) confirmed the model’s robustness (P < 0.05, permutation test), excluding the possibility of random classification. These results suggest that the model is both statistically reliable (high R²Y) and predictively valid (high Q²Y). To further validate the model’s performance, we conducted a permutation test (100 random label shuffles) (S1D Fig). The real model’s R²Y/Q²Y (red line) significantly outperformed the random distribution (gray bars, P = 0.05), confirming that the observed classification was not due to chance. This stringent validation ensures that the identified differentially expressed genes and pathways are biologically meaningful rather than artifacts.

The levels of transcripts and genetic variants of 22 mtDNA MRGs were investigated, and the results revealed that the expression of 22 mtDNA MRGs was upregulated in HCC tissues compared with healthy tissues (Fig 1A). To further understand the genetic variation of mtDNA MRGs in HCC, the copy number variation (CNV) and somatic cell mutation rate of mtDNA MRGs in HCC samples were summarized. Among the 371 HCC samples, 28 (7.55%) had mutations in mtDNA MRGs. The major mutated genes were RECQL4 (2%), APEX1 (1%), DNA2 (1%), LIG3 (1%), POLG (1%), POLG2 (1%), POLRMT (1%), and TOP1MT (1%) (Fig 1B). Among them, 18 genes were significantly related to prognosis (Fig 1C-T). According to the above analysis, the genetic mutations and expression changes of mtDNA MRGs were highly heterogeneous between normal samples and HCC samples, suggesting that the imbalance in the expression of mtDNA MRGs plays an important role in the occurrence and development of HCC.

Fig 1 — A. Expression of mtDNA MRGs in healthy tissue and HCC tissue. B. mtDNA MRG mutation frequencies in 371 HCC patients. The histogram above shows the total TMB. The numbers on the right indicate the frequency of mutations in each gene. The bar chart on the right shows the proportion of each mutation. The stacked bar chart below shows the converted scores. C-T. Survival analysis of HCC patients according to mtDNA MRG expression (P<0.05). APEX1, Apurinic/apyrimidinic endonuclease 1; DNA2, ATP-dependent helicase/nuclease 2; EXOG, Endo/exonuclease G-like; EXDOG, endonuclease G; LIG3, ligase III; METTL4, Methyltransferase-like 4; POLG2, Polymerase-gamma2; UNG, Uracil DNA glycosylase; TOP3A, Topoisomerase IIIalpha; OGG1, 8-oxoguanine DNA glycosylase; TFB2M, Mitochondrial transcription factor B2; TOP1MT, Mitochondrial topoisomerase I; RNASEH1, Ribonuclease H1; SSBP1, Single-stranded DNA-binding protein 1; TFAM, Mtochondrial transcription factor A; RECQL4, RecQ protein-like 4; POLRMT, Mitochondrial RNA polymerase; POLG, Polymerase gamma; POLB, Polymerase beta; TWNK, Twinkle mitochondrial DNA helicase; PRIMPOL, Primase-polymerase; MGME1, Mitochondrial genome maintenance exonuclease 1.

Identification of molecular subtypes of mtDNA MRGs

The interactions and correlations among mtDNA MRGs were demonstrated through a network interaction diagram, as shown in (Fig 2A). POLG and ENDOG were found to be protective factors in terms of the prognosis of HCC. K = 3 was the best cluster value for grouping HCC patients according to the expression of the mtDNA MRGs; the clusters were named as follows: mtDNA MRG cluster A (n = 140), mtDNA MRG cluster B (n = 214), and mtDNA MRG cluster C (n = 131) (Fig 2B-D). Prognosis analysis revealed that patients in clusters A and C had better prognoses than did those in cluster B (Fig 2F). In addition, there were significant differences in clinicopathological features (such as TNM stage, sex, age, and survival status) among the three mtDNA MRG clusters (Fig 2E), and most mtDNA MRGs were highly expressed in mtDNA MRG cluster B.

Biological functions and TME immune cell infiltration features of different mtDNA MRG clusters

GSVA enrichment analysis revealed differences in a variety of pathways among mtDNA MRG clusters A, B and C. mtDNA MRG cluster B was enriched mainly in matrix and carcinogenic activation pathways, such as the mTOR signaling pathway, cell cycle and RNA degradation (Fig 3A), while mtDNA MRG cluster A and mtDNA MRG cluster C were significantly enriched in immune regulation and metabolic pathways, such as the metabolism of fatty acids, arachidonic acid and lysine (Fig 3B-C). Subsequently, the relationships between the mtDNA MRG clusters and immune infiltrating cells were further analyzed, and significant differences in TME cell infiltration were detected among the three mtDNA MRG clusters (Fig 3D). Cluster A had the most infiltrating immune cells and presented enrichment of innate immune cells, including eosinophils, macrophages, and mast cells, while cluster B had the lowest infiltration level. The results of the PCA showed differences between the transcriptome profiles of the three mtDNA MRG clusters (Fig 3E).

Fig 3 — Heatmaps showing different biological processes, with red representing activation pathways and blue representing inhibition pathways. A. mtDNA MRG cluster A and mtDNA MRG cluster B. B. mtDNA MRG cluster A and mtDNA MRG cluster C. mtDNA MRG cluster B and mtDNA MRG cluster C. D. Abundance of TME-infiltrating cells in samples enriched in the three mtDNA MRG clusters. E. Transcriptome-level differences between mtDNA MRG clusters based on PCA.

Identification and biological functions of gene clusters

The differences in the mtDNA MRG clusters were analyzed using the limma package. As shown in the figure, there were 8293 DEGs between clusters A-B, 2043 DEGs between clusters A-C, and 9064 DEGs between clusters B-C. The common genes in the intersection of DEGs among the three mtDNA MRG clusters were identified, and 740 common DEGs were considered core genes (Fig 4A). KEGG analysis revealed that the 740 core genes were enriched in carcinogenesis and RNA modification pathways, mainly in the cell cycle, RNA degradation, and DNA base repair (Fig 4B).

Fig 4 — A. DEGs among mtDNA MRG clusters. The red part represents the DEGs between clusters B-A, the green part represents the DEGs between clusters C-A differential gene, and the gray part represents the DEGs between clusters C-B. B. KEGG enrichment analysis of the core genes.

Based on 740 mtDNA-related genes, HCC patients were divided into three gene clusters by unsupervised cluster analysis: geneCluster A (n = 120), geneCluster B (n = 153), and geneCluster C (n = 212) (Fig 5A-C). There were significant differences in the expression levels of the three mtDNA MRG gene clusters, HCC patients with different gene clusters displayed different clinicopathological characteristics, and most patients with enrichment of geneCluster B showed better survival status or had clinical stage I-II disease (Fig 5D). Kaplan–Meier survival analysis revealed that HCC patients in geneCluster A had a poor prognosis, while those in geneCluster B had a better prognosis (Fig 5E). Most of the mtDNA MRGs were downregulated in gene cluster B but upregulated in geneCluster A (Fig 5F).

Fig 5 — A-C. Consensus clustering of genecluster for k = 3. D. Survival analysis of three gene clusters. E. Heatmaps showing the gene profiles of the three gene clusters and the relationships between the gene clusters and clinicopathological features. F. Expression levels of mtDNA MRGs in three gene clusters. APEX1, Apurinic/apyrimidinic endonuclease 1; DNA2, ATP-dependent helicase/nuclease 2; EXOG, Endo/exonuclease G-like; EXDOG, endonuclease G; LIG3, ligase III; METTL4, Methyltransferase-like 4; POLG2, Polymerase-gamma2; UNG, Uracil DNA glycosylase; TOP3A, Topoisomerase IIIalpha; OGG1, 8-oxoguanine DNA glycosylase; TFB2M, Mitochondrial transcription factor B2; TOP1MT, Mitochondrial topoisomerase I; RNASEH1, Ribonuclease H1; SSBP1, Single-stranded DNA-binding protein 1; TFAM, Mtochondrial transcription factor A; RECQL4, RecQ protein-like 4; POLRMT, Mitochondrial RNA polymerase; POLG, Polymerase gamma; POLB, Polymerase beta.

Construction of the mtDNA score and functional annotation

Due to the individual heterogeneity and complexity of HCC patients, the mtDNA score was established to quantify the mtDNA MRG patterns of individual patients. HCC patients were divided into a high mtDNA score group and a low mtDNA score group, and the “surminer” program was used to obtain the best truncation value (0.05). A Sankey chart was generated to show mtDNA score, survival status, mtDNA MRG clusters and gene cluster correlations (Fig 6A). Immune correlation analysis revealed that the mtDNA score was negatively correlated with the activation of CD8 + T cells, eosinophils and mast cells but was positively correlated with the activation of CD4 + T cells (Fig 6B). The Kruskal–Wallis test was conducted for the mtDNA MRG cluster, and mtDNA MRG cluster B had the highest score. mtDNA MRG cluster C had the lowest score (Fig 6C); for gene clusters, geneCluster A had the highest score, and geneCluster B had the lowest score (Fig 6D). Survival analysis revealed a significant survival advantage in HCC patients with a low mtDNA score (Fig 6E).

Fig 6 — A. Alluvial maps showing changes in the mtDNA MRG cluster, gene cluster, mtDNA score and survival status. B. Correlation between the DNA score and immune cell infiltration. C. mtDNA score differences among the three mtDNA MRG clusters. D. mtDNA score differences among the three mtDNA MRG gene clusters. E. Survival analysis of patients with high mtDNA scores and low mtDNA scores.

mtDNA score and TMB

HCC patients were divided into high-TMB and low-TMB groups based on the TMB. Survival analysis revealed that patients with low TMB had better survival than those with high TMB (Fig 7A). When the mtDNA score was combined with the TMB to predict the prognosis of patients with HCC, patients with both high TMB and high mtDNA score had the worst prognosis (Fig 7B). In addition, the waterfall plots show the landscape of tumor somatic mutations between the high- and low-mtDNA-score groups. The most common somatic mutations were in TP53 (58%) in the high mtDNA score group (Fig 7C) and in CTNNB1 (28%) in the low mtDNA score group (Fig 7D).

Fig 7 — A. Survival analysis of patients with high and low TMB. B. Survival analysis of patients stratified by TMB combined with the mtDNA MRG score. C. Waterfall plot for the high mtDNA MRG score group. D. Waterfall map for the low mtDNA MRG score group. TMB, tumor mutation burden.

Correlation analysis of the mtDNA score with clinicopathological features and immunotherapy efficacy

According to the clinicopathological feature correlation analysis, younger age (< 65 years), stage III-IV, and death at the clinical endpoint were significantly associated with a higher mtDNA score (Fig 8A-C). Stratified analysis revealed that patients in the low mtDNA score group had a better prognosis than patients in the high mtDNA score group (Fig 8D-F). These results further supported that the mtDNA score might be used to predict the prognosis of HCC.

Fig 8 — A-C. Correlation analysis between clinical features and the mtDNA MRG score. D-F. Associations of the mtDNA MRG score with death outcome (D), age (E), and stage **(F)**.

To further explore the correlation between the mtDNA score and immunotherapy efficacy, the therapeutic response to ICIs, such as CTLA-4/PD-1 inhibitors, was analyzed. Fig 9A shows no difference in PD-L1 expression between the high mtDNA score group and the low mtDNA score group (Fig 9A). The anti-CTLA-4 and anti-PD-1 response rates were different between the high mtDNA score group and the low mtDNA score group (p = 0.018) (Fig 9B). However, there was no difference between the high mtDNA score group and the low mtDNA score group in terms of the response to anti-PD-1, anti-CTLA-4, or combined anti-CTLA-4/PD-1 immunotherapy (Fig 9C-E).

Fig 9 — A. Expression of PD-L1 in the low-mtDNA score group and high-mtDNA score group. B. Analysis of differences in the response to CTLA-4 negative and PD-1 negative therapy between the low mtDNA score group and high mtDNA score group. C. Analysis of the difference in anti-PD-1 immunotherapy efficacy between the low mtDNA score group and the high mtDNA score group. D. Analysis of the difference in anti-CTLA-4 immunotherapy efficacy between the low mtDNA score group and the high mtDNA score group. E. Differences in the efficacy of anti-PD-1/anti-CTLA-4 combined immunotherapy between the low-mtDNA MRG score group and the high-mtDNA MRG score group. PD-1, Programmed cell death-1; PD-L1, Programmed death ligand 1; CTLA-4, Cytotoxic T-lymphocyte antigen 4.

Discussion

In this study, 22 mtDNA MRGs and three mtDNA MRG clusters were identified. HCC patients with enrichment of the three mtDNA MRG clusters had different prognoses and TME immune cell infiltration characteristics. mtDNA MRG cluster A is characterized by enhanced tumor matrix activity and abundant innate immune cell infiltration, corresponding to the immune rejection phenotype; mtDNA MRG cluster B is characterized by immunosuppression, corresponding to the immune-desert phenotype; and mtDNA MRG cluster C is characterized by adaptive immune cell infiltration and immune activation, corresponding to the immunoinflammatory phenotype [29]. The immune rejection and immune-desert phenotypes are considered to indicate noninflammatory tumors, primarily tumors that lack immune cell invasion in the parenchyma and stroma, rarely express PD-L1, are located at the opposite end of the tumor immune continuum, and histologically lack immune invasion and antigen presentation (low MHC class I), while exhibiting high tumor cell proliferation. The immunoinflammatory phenotype indicates inflammatory tumors, in which the TME has a high degree of infiltration of immune cells, such as T cells, CD8 + T cells producing IFN-γ, and PD-1-positive immune cells [30]. There were significant differences in the TME features among samples enriched in the three mtDNA MRG clusters. The GSVA results showed that mtDNA MRG clusters A and C were mainly enriched in pathways related to tumor metabolism, leading to relative inhibition of tumor growth and good prognosis [31], while mtDNA MRG cluster B was mainly enriched in the cell cycle, RNA degradation, DNA replication, base repair and other pathways, which are closely related to cancer progression [32–34]. Therefore, HCC patients with enrichment of mtDNA MRG cluster B have a poor survival prognosis.

Analysis of mtDNA MRG clusters is helpful for understanding the TME infiltration characteristics. Due to individual heterogeneity and specificity, a scoring model was constructed to evaluate the mtDNA MRG patterns in individual HCC patients. The mtDNA score was higher for the immune-desert phenotype and lower for the immune inflammatory phenotype. A high mtDNA score is associated with increased immune and stromal cell infiltration [14]. This result suggested that the mtDNA score is a useful tool for comprehensively evaluating the enrichment of mtDNA MRGs in individual tumors.

Mitochondria are important organelles, and mtDNA variants (with mutations/single nucleotide polymorphisms) and disorders of mitochondrial coding genes are associated with cancer progression. The genes with the highest mutation frequencies in HCC patients with high and low mtDNA scores were TP53 and CTNNB1, respectively. TP53 encodes the P53 protein known as a canonical tumor suppressor, and protein isoforms derived from mutated p53 are usually overexpressed in cancer; these isoforms lose their tumor suppressor function and promote tumorigenesis [35]. p53 promotes mitochondrial respiration in a variety of ways. Mutation of p53 can affect the morphology and structure of mitochondria, resulting in loss of mitochondrial function. Mutated p53 affects mitochondrial energy production by regulating mtDNA replication and maintenance, assembly of the mitochondrial respiratory chain complex and regulation of the mitochondrial membrane potential [36]. HCC is a highly heterogeneous tumor, and mutations in the TP53 gene are involved in the development of intratumor heterogeneity and may contribute to treatment failure and drug resistance in many HCC patients [37,38]. In the present study, the poor prognosis of HCC in the high mtDNA score group was associated with excessive TP53 mutation [39]. The probability of CTNNB1 mutation was greater in HCC patients with low mtDNA scores, and CTNNB1 mutation in HCC patients was associated with specific well-differentiated HCC subtypes, in contrast to the findings in TP53-mutated HCC patients, in whom histologically poorly differentiated, highly proliferative, large trabecular clumps developed [40]. HCC patients with high mtDNA scores had a significantly greater probability of having genetic mutations than did those with low mtDNA scores. TMB, the total number of mutations in each coding region of the tumor genome, is a new indicator for predicting immune responses in various cancers [41]. The finding that patients with high TMB and mtDNA scores had the worst prognosis may be related to genomic instability caused by a high frequency of TP53 mutations and poor HCC differentiation [42].

Previous studies have shown that mitochondrial dysfunction is a key factor in cancer pathogenesis [43]. Dysregulation of mitochondria plays an important role in oncogenesis and progression of HCC [44,45]. However, there are few reports on the relationship between mitochondrial genes and the development of HCC. In our data, HCC patients with high expression of RECQL4 and APEX1 have a worse prognosis. A retrospective study has shown that the expression of RECQL4 mRNA in HCC tissues is significantly higher than that in adjacent normal liver tissues. Overexpression of RECQL4 may give hepatocellular carcinoma cells unlimited proliferation potential and promote hepatocellular tumor occurrence [46]. In addition, APEX1 was significantly upregulated and predicted poor clinical overall survival in HCC patients. Silencing APEX1 inhibited the proliferation of HCC cells in vivo and in vitro, and it repressed invasion and migration [47]. So targeted mitochondrial gene therapy is an attractive strategy to inhibit HCC progression.

Analysis of the correlation between the mtDNA score and clinical data revealed that patients aged < 65 years who died at the clinical endpoint and with stage III-IV disease had relatively high mtDNA scores and poor clinical outcomes. Many studies have shown that the copy number and function of mtDNA decrease with age [48]. An increase in mtDNA copy number leads to mitochondrial dysfunction, thus causing disease. Patients with malignant tumors such as lung cancer, pancreatic cancer and kidney cancer have increased mtDNA copy numbers [49]. Therefore, patients aged < 65 years have higher mtDNA scores than patients aged > 65 years.

The mtDNA score can also be used to evaluate certain clinical characteristics of patients. The effects of anti-CTLA4 and anti-PD-1 treatments differed between the high mtDNA score group and the low mtDNA score group. The value of the mtDNA score in predicting the outcome of PD-L1/CTLA-4 immunotherapy was subsequently further analyzed; however, the mtDNA score did not significantly differ between responders and nonresponders receiving PD-L1, CTLA-4, or combination therapy. These results suggest that an imbalance in mtDNA MRGs affects immunotherapy resistance in HCC patients. Drug resistance is a major obstacle to cancer treatment [50]. Mitochondrial metabolic plasticity contributes to resistance in most types of anticancer therapy [51]. In melanoma, alterations in mitochondrial function may affect the energy metabolism, apoptosis, and immunotherapy response of melanoma cells, thus affecting the progression of melanoma and resistance to PD-1 inhibitors [52]. Therefore, the results of this study can guide future studies on the mechanisms related to immunotherapy resistance and mtDNA MRGs in HCC.

Although we identified 22 mtDNA MRGs, using retrospective data to construct mtDNA scores. In clinical practice, mtDNA scores may be useful in predicting clinical outcomes as well as evaluating the corresponding TME cell infiltration characteristics and TMB in individual HCC patients. However, several limitations should be acknowledged. First, to ensure data robustness, we excluded samples located far from the main cluster, as these may represent biological outliers deviating from the predominant group trend. Given that all specimens were derived from human tissues, inter-individual variability is expected; nevertheless, the overall differences remained within an acceptable range. Moreover, the TCGA and GEO datasets employed in this study have been extensively validated in prior research, supporting their reliability. To minimize technical artifacts, we performed rigorous data correction, ensuring that subsequent findings reflect true biological characteristics rather than batch effects. Second, all clinical data in this study came from public databases, and the mtDNA MRG score established on this basis may have potential biases. In addition, further cell and animal experiments are needed to explore the functional role of mtDNA MRGs in HCC, which could help provide stronger clues to guide clinical applications.

Conclusion

In conclusion, a comprehensive and systematic analysis of mtDNA MRGs was conducted in this study. The mtDNA score established based on mtDNA MRGs can enhance the understanding of the cell infiltration characteristics of the TME and provide innovative ideas regarding potential therapeutic targets and more effective immunotherapy strategies to provide theoretical guidance for treatment decision making and prognosis prediction in HCC patients.

Supporting information

S1 File. The performance of oPLS-DA model was assessed by R²Y (goodness of fit) and Q² (goodness of prediction) through cross-validation, along with permutation testing to evaluate robustness.

(DOCX)

pone.0325033.s001.docx^{(612.6KB, docx)}

Data Availability

The HCC RNA transcriptome sequencing data and clinical information were obtained from The Cancer Genome Atlas (TCGA) (http://cancergenome.nih.gov/) and Gene Expression Omnibus (GEO)(https://www.ncbi.nlm.nih.gov/geo/). TCGA and GEO databases are public databases where relevant data can be downloaded directly.We have already upload figures to figshare, DOI: 10.6084/m9.figshare.27089542.

Funding Statement

This work was supported by the Hubei Provincial Technology Innovation Project (2023BCB129); the Project of Creative Research Groups of Hubei Province (2023AFA023); Special Fund for Artificial Liver of the Beijing Hepatobiliary Charity Foundation (RGGJJ-2021-026); Ligan project of the Beijing Hepatobiliary Charity Foundation (iGandanF-1082024-LG001).

References

1.Zabransky DJ, Danilova L, Leatherman JM, Lopez-Vidal TY, Sanchez J, Charmsaz S, et al. Profiling of syngeneic mouse HCC tumor models as a framework to understand anti-PD-1 sensitive tumor microenvironments. Hepatology. 2023;77(5):1566–79. doi: 10.1002/hep.32707 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.He Q, Yang J, Jin Y. Immune infiltration and clinical significance analyses of the coagulation-related genes in hepatocellular carcinoma. Brief Bioinform. 2022;23(4):bbac291. doi: 10.1093/bib/bbac291 [DOI] [PubMed] [Google Scholar]
3.Ruff SM, Shannon AH, Pawlik TM. Advances in Targeted Immunotherapy for Hepatobiliary Cancers. Int J Mol Sci. 2022;23(22):13961. doi: 10.3390/ijms232213961 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Nevola R, Ruocco R, Criscuolo L, Villani A, Alfano M, Beccia D, et al. Predictors of early and late hepatocellular carcinoma recurrence. World J Gastroenterol. 2023;29(8):1243–60. doi: 10.3748/wjg.v29.i8.1243 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Dong J, Wong L-J, Mims MP. Mitochondrial inheritance and cancer. Transl Res. 2018;202:24–34. doi: 10.1016/j.trsl.2018.06.004 [DOI] [PubMed] [Google Scholar]
6.Zong W-X, Rabinowitz JD, White E. Mitochondria and Cancer. Mol Cell. 2016;61(5):667–76. doi: 10.1016/j.molcel.2016.02.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Wheelhouse NM, Lai PBS, Wigmore SJ, Ross JA, Harrison DJ. Mitochondrial D-loop mutations and deletion profiles of cancerous and noncancerous liver tissue in hepatitis B virus-infected liver. Br J Cancer. 2005;92(7):1268–72. doi: 10.1038/sj.bjc.6602496 [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Picca A, Guerra F, Calvani R, Coelho-Júnior HJ, Leeuwenburgh C, Bucci C, et al. The contribution of mitochondrial DNA alterations to aging, cancer, and neurodegeneration. Exp Gerontol. 2023;178:112203. doi: 10.1016/j.exger.2023.112203 [DOI] [PubMed] [Google Scholar]
9.El-Hattab AW, Craigen WJ, Scaglia F. Mitochondrial DNA maintenance defects. Biochim Biophys Acta Mol Basis Dis. 2017;1863(6):1539–55. doi: 10.1016/j.bbadis.2017.02.017 [DOI] [PubMed] [Google Scholar]
10.Alimardani M, Moghbeli M, Rastgar-Moghadam A, Shandiz FH, Abbaszadegan MR. Single Nucleotide Polymorphisms as the Efficient Prognostic Markers in Breast Cancer. Curr Cancer Drug Targets. 2021;21(9):768–93. doi: 10.2174/1568009621666210525151846 [DOI] [PubMed] [Google Scholar]
11.Lee H-Y, Nga HT, Tian J, Yi H-S. Mitochondrial Metabolic Signatures in Hepatocellular Carcinoma. Cells. 2021;10(8):1901. doi: 10.3390/cells10081901 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Rodrigues T, Ferraz LS. Therapeutic potential of targeting mitochondrial dynamics in cancer. Biochem Pharmacol. 2020;182:114282. doi: 10.1016/j.bcp.2020.114282 [DOI] [PubMed] [Google Scholar]
13.Quail DF, Joyce JA. The Microenvironmental Landscape of Brain Tumors. Cancer Cell. 2017;31(3):326–41. doi: 10.1016/j.ccell.2017.02.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Ma S, Zhu J, Wang M, Zhu J, Wang W, Xiong Y, et al. Comprehensive analysis of m7G modification patterns based on potential m7G regulators and tumor microenvironment infiltration characterization in lung adenocarcinoma. Front Genet. 2022;13:996950. doi: 10.3389/fgene.2022.996950 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Ali HR, Chlon L, Pharoah PDP, Markowetz F, Caldas C. Patterns of Immune Infiltration in Breast Cancer and Their Clinical Implications: A Gene-Expression-Based Retrospective Study. PLoS Med. 2016;13(12):e1002194. doi: 10.1371/journal.pmed.1002194 [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Lin Y-H, Lim S-N, Chen C-Y, Chi H-C, Yeh C-T, Lin W-R, et al. Functional Role of Mitochondrial DNA in Cancer Progression. Int J Mol Sci. 2022;23(3):1659. doi: 10.3390/ijms23031659 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Singh RK, Srivastava A, Kalaiarasan P, Manvati S, Chopra R, Bamezai RNK. mtDNA germ line variation mediated ROS generates retrograde signaling and induces pro-cancerous metabolic features. Sci Rep. 2014;4:6571. doi: 10.1038/srep06571 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Singh RK, Saini S, Verma D, Kalaiarasan P, Bamezai RNK. Mitochondrial ND5 mutation mediated elevated ROS regulates apoptotic pathway epigenetically in a P53 dependent manner for generating pro-cancerous phenotypes. Mitochondrion. 2017;35:35–43. doi: 10.1016/j.mito.2017.05.001 [DOI] [PubMed] [Google Scholar]
19.Cao L, Cheng H, Jiang Q, Li H, Wu Z. APEX1 is a novel diagnostic and prognostic biomarker for hepatocellular carcinoma. Aging (Albany NY). 2020;12(5):4573–91. doi: 10.18632/aging.102913 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Zhang H, Jiang P-J, Lv M-Y, Zhao Y-H, Cui J, Chen J, et al. OGG1 contributes to hepatocellular carcinoma by promoting cell cycle-related protein expression and enhancing DNA oxidative damage repair in tumor cells. J Clin Lab Anal. 2022;36(7):e24561. doi: 10.1002/jcla.24561 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Piro-Mégy C, Sarzi E, Tarrés-Solé A, Péquignot M, Hensen F, Quilès M, et al. Dominant mutations in mtDNA maintenance gene SSBP1 cause optic atrophy and foveopathy. J Clin Invest. 2020;130(1):143–56. doi: 10.1172/JCI128513 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Yan C, Niu Y, Ma L, Tian L, Ma J. System analysis based on the cuproptosis-related genes identifies LIPT1 as a novel therapy target for liver hepatocellular carcinoma. J Transl Med. 2022;20(1):452. doi: 10.1186/s12967-022-03630-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Liu L, Liu B, Yu J, Zhang D, Shi J, Liang P. Development of a Toll-Like Receptor-Based Gene Signature That Can Predict Prognosis, Tumor Microenvironment, and Chemotherapy Response for Hepatocellular Carcinoma. Front Mol Biosci. 2021;8:729789. doi: 10.3389/fmolb.2021.729789 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Wilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26(12):1572–3. doi: 10.1093/bioinformatics/btq170 [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50. doi: 10.1073/pnas.0506580102 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Charoentong P, Finotello F, Angelova M, Mayer C, Efremova M, Rieder D, et al. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell Rep. 2017;18(1):248–62. doi: 10.1016/j.celrep.2016.12.019 [DOI] [PubMed] [Google Scholar]
27.Chong W, Shang L, Liu J, Fang Z, Du F, Wu H, et al. m6A regulator-based methylation modification patterns characterized by distinct tumor microenvironment immune profiles in colon cancer. Theranostics. 2021;11(5):2201–17. doi: 10.7150/thno.52717 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Sotiriou C, Wirapati P, Loi S, Harris A, Fox S, Smeds J, et al. Gene expression profiling in breast cancer: understanding the molecular basis of histologic grade to improve prognosis. J Natl Cancer Inst. 2006;98(4):262–72. doi: 10.1093/jnci/djj052 [DOI] [PubMed] [Google Scholar]
29.Hegde PS, Chen DS. Top 10 Challenges in Cancer Immunotherapy. Immunity. 2020;52(1):17–35. doi: 10.1016/j.immuni.2019.12.011 [DOI] [PubMed] [Google Scholar]
30.Hegde PS, Karanikas V, Evers S. The Where, the When, and the How of Immune Monitoring for Cancer Immunotherapies in the Era of Checkpoint Inhibition. Clin Cancer Res. 2016;22(8):1865–74. doi: 10.1158/1078-0432.CCR-15-1507 [DOI] [PubMed] [Google Scholar]
31.Huang X, Qiu Z, Li L, Chen B, Huang P. m6A regulator-mediated methylation modification patterns and tumor microenvironment infiltration characterization in hepatocellular carcinoma. Aging (Albany NY). 2021;13(16):20698–715. doi: 10.18632/aging.203456 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Liu J, Peng Y, Wei W. Cell cycle on the crossroad of tumorigenesis and cancer therapy. Trends Cell Biol. 2022;32(1):30–44. doi: 10.1016/j.tcb.2021.07.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Li J, Xie H, Ying Y, Chen H, Yan H, He L, et al. YTHDF2 mediates the mRNA degradation of the tumor suppressors to induce AKT phosphorylation in N6-methyladenosine-dependent way in prostate cancer. Mol Cancer. 2020;19(1):152. doi: 10.1186/s12943-020-01267-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Mughal MJ, Mahadevappa R, Kwok HF. DNA replication licensing proteins: Saints and sinners in cancer. Semin Cancer Biol. 2019;58:11–21. doi: 10.1016/j.semcancer.2018.11.009 [DOI] [PubMed] [Google Scholar]
35.Tang Q, Su Z, Gu W, Rustgi AK. Mutant p53 on the Path to Metastasis. Trends Cancer. 2020;6(1):62–73. doi: 10.1016/j.trecan.2019.11.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Kamp WM, Wang P-Y, Hwang PM. TP53 mutation, mitochondria and cancer. Curr Opin Genet Dev. 2016;38:16–22. doi: 10.1016/j.gde.2016.02.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Calderaro J, Ziol M, Paradis V, Zucman-Rossi J. Molecular and histological correlations in liver cancer. J Hepatol. 2019;71(3):616–30. doi: 10.1016/j.jhep.2019.06.001 [DOI] [PubMed] [Google Scholar]
38.Shi M, Wang Y, Tang W, Cui X, Wu H, Tang Y, et al. Identification of TP53 mutation associated-immunotype and prediction of survival in patients with hepatocellular carcinoma. Ann Transl Med. 2020;8(6):321. doi: 10.21037/atm.2020.02.98 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Murai H, Kodama T, Maesaka K, Tange S, Motooka D, Suzuki Y, et al. Multiomics identifies the link between intratumor steatosis and the exhausted tumor immune microenvironment in hepatocellular carcinoma. Hepatology. 2023;77(1):77–91. doi: 10.1002/hep.32573 [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Calderaro J, Couchy G, Imbeaud S, Amaddeo G, Letouzé E, Blanc J-F, et al. Histological subtypes of hepatocellular carcinoma are related to gene mutations and molecular tumour classification. J Hepatol. 2017;67(4):727–38. doi: 10.1016/j.jhep.2017.05.014 [DOI] [PubMed] [Google Scholar]
41.Gabbia D, De Martin S. Tumor Mutational Burden for Predicting Prognosis and Therapy Outcome of Hepatocellular Carcinoma. Int J Mol Sci. 2023;24(4):3441. doi: 10.3390/ijms24043441 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Rao CV, Asch AS, Yamada HY. Frequently mutated genes/pathways and genomic instability as prevention targets in liver cancer. Carcinogenesis. 2017;38(1):2–11. doi: 10.1093/carcin/bgw118 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Luo Y, Ma J, Lu W. The Significance of Mitochondrial Dysfunction in Cancer. Int J Mol Sci. 2020;21(16):5598. doi: 10.3390/ijms21165598 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Yao J, Wang J, Xu Y, Guo Q, Sun Y, Liu J, et al. CDK9 inhibition blocks the initiation of PINK1-PRKN-mediated mitophagy by regulating the SIRT1-FOXO3-BNIP3 axis and enhances the therapeutic effects involving mitochondrial dysfunction in hepatocellular carcinoma. Autophagy. 2022;18(8):1879–97. doi: 10.1080/15548627.2021.2007027 [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Hsu C-C, Lee H-C, Wei Y-H. Mitochondrial DNA alterations and mitochondrial dysfunction in the progression of hepatocellular carcinoma. World J Gastroenterol. 2013;19(47):8880–6. doi: 10.3748/wjg.v19.i47.8880 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Li J, Jin J, Liao M, Dang W, Chen X, Wu Y, et al. Upregulation of RECQL4 expression predicts poor prognosis in hepatocellular carcinoma. Oncol Lett. 2018;15(4):4248–54. doi: 10.3892/ol.2018.7860 [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Sun Z, Chen G, Wang L, Sang Q, Xu G, Zhang N. APEX1 promotes the oncogenicity of hepatocellular carcinoma via regulation of MAP2K6. Aging (Albany NY). 2022;14(19):7959–71. doi: 10.18632/aging.204325 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Chiang JL, Shukla P, Pagidas K, Ahmed NS, Karri S, Gunn DD, et al. Mitochondria in Ovarian Aging and Reproductive Longevity. Ageing Res Rev. 2020;63:101168. doi: 10.1016/j.arr.2020.101168 [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Filograna R, Mennuni M, Alsina D, Larsson N-G. Mitochondrial DNA copy number in human disease: the more the better?. FEBS Lett. 2021;595(8):976–1002. doi: 10.1002/1873-3468.14021 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Zhang C, Liu X, Jin S, Chen Y, Guo R. Ferroptosis in cancer therapy: a novel approach to reversing drug resistance. Mol Cancer. 2022;21(1):47. doi: 10.1186/s12943-022-01530-y [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Jin P, Jiang J, Zhou L, Huang Z, Nice EC, Huang C, et al. Mitochondrial adaptation in cancer drug resistance: prevalence, mechanisms, and management. J Hematol Oncol. 2022;15(1):97. doi: 10.1186/s13045-022-01313-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Du F, Yang L-H, Liu J, Wang J, Fan L, Duangmano S, et al. The role of mitochondria in the resistance of melanoma to PD-1 inhibitors. J Transl Med. 2023;21(1):345. doi: 10.1186/s12967-023-04200-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

PLoS One. 2025 Jun 2;20(6):e0325033. doi: 10.1371/journal.pone.0325033.r001

Author response to Decision Letter 0

5 Jun 2024

PLoS One. doi: 10.1371/journal.pone.0325033.r002

Decision Letter 0

Matthew Cserhati

5 Aug 2024

PONE-D-24-22922Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment featuresPLOS ONE

Dear Dr. Rong,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Review of PONE-D-24-22922

Please define the mtDNA score, TMB and TME (lines 25-17).
I suggest you use the keyword mitochondrion and not mitochondrial (line 44).
“own superhelical double-linked 61 ring of genetic material” is unnecessary (line 60).
Move gene list from lines 115-118 to a separate table.
Describe ssGSEA in more detail in the Methods section.
What is this data set from Charoenteng? Add a reference to it.
Why was an adjusted p-value of 0.001 used? (line 143)
Lines 166-176: indentation is problematic
You mention 8293 DEGs between clusters A and B and 2043 DEGs between clusters A and C, and 9064 between clusters B and C. Why are there so many? What was the fold change cutoff limit? What would happen if you increase this cutoff limit?
Starting from line 317 you describe the role of TP53 in cancer. This is well known. What is the novelty of your findings? As you write, HCC is very heterogeneous, what other genes play a role in HCC that you found?

Please submit your revised manuscript by Sep 19 2024 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org . When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols . Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols .

We look forward to receiving your revised manuscript.

Kind regards,

Matthew Cserhati, Ph.D

Academic Editor

PLOS ONE

Journal Requirements:

1. When submitting your revision, we need you to address these additional requirements.

Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

3. Thank you for uploading your study's underlying data set. Unfortunately, the repository you have noted in your Data Availability statement does not qualify as an acceptable data repository according to PLOS's standards.

At this time, please upload the minimal data set necessary to replicate your study's findings to a stable, public repository (such as figshare or Dryad) and provide us with the relevant URLs, DOIs, or accession numbers that may be used to access these data. For a list of recommended repositories and additional information on PLOS standards for data deposition, please see https://journals.plos.org/plosone/s/recommended-repositories.

4. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager. Please see the following video for instructions on linking an ORCID iD to your Editorial Manager account: https://www.youtube.com/watch?v=_xcclfuvtxQ

5. Please amend either the abstract on the online submission form (via Edit Submission) or the abstract in the manuscript so that they are identical.

Additional Editor Comments:

Review of PONE-D-24-22922

1. Please define the mtDNA score, TMB and TME (lines 25-17).

2. I suggest you use the keyword mitochondrion and not mitochondrial (line 44).

3. “own superhelical double-linked 61 ring of genetic material” is unnecessary (line 60).

4. Move gene list from lines 115-118 to a separate table.

5. Describe ssGSEA in more detail in the Methods section.

6. What is this data set from Charoenteng? Add a reference to it.

7. Why was an adjusted p-value of 0.001 used? (line 143)

8. Lines 166-176: indentation is problematic

9. You mention 8293 DEGs between clusters A and B and 2043 DEGs between clusters A and C, and 9064 between clusters B and C. Why are there so many? What was the fold change cutoff limit? What would happen if you increase this cutoff limit?

10. Starting from line 317 you describe the role of TP53 in cancer. This is well known. What is the novelty of your findings? As you write, HCC is very heterogeneous, what other genes play a role in HCC that you found?

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The manuscript titled "Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden, and tumor microenvironment features" presents a comprehensive analysis of the relationship between mitochondrial DNA (mtDNA) maintenance-related genes (mtDNA MRGs) and the prognosis of hepatocellular carcinoma (HCC), tumor mutation burden (TMB), and tumor microenvironment (TME) features. The study utilized a dataset of 487 HCC samples from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) to identify 22 mtDNA MRGs through screening and then applied unsupervised clustering and principal component analysis (PCA) to construct an mtDNA score model. This model was used to examine associations with clinicopathological characteristics, TMB, TME cell infiltration, and biological processes.

Methodological Considerations: The study design appears sound, utilizing a large sample size and robust bioinformatic tools. However, there are several areas that could be improved upon for greater clarity and rigor:

1. Sample Selection and Control Group: While the use of TCGA and GEO datasets is commendable, the inclusion criteria for selecting HCC samples should be more clearly described. Moreover, the control group's characteristics should also be provided for comparison purposes.

2. Screening of mtDNA MRGs: The process of identifying the 22 mtDNA MRGs requires further detail. How were these genes prioritized over others? What criteria were used for screening?

3. Unsupervised Clustering: Details about the machine learning model employed for unsupervised clustering need to be elaborated upon, including parameters and validation methods.

4. PCA Model Construction: The construction of the mtDNA score model using PCA should be detailed, including how the principal components were selected and validated.

Statistical Analysis: The statistical methods used appear appropriate for the data type and research question. Nevertheless, the following points should be addressed:

1. Statistical Power: It would be beneficial to include a discussion on the statistical power of the study given the sample size and the number of variables being analyzed.

2. Multiple Testing Correction: With multiple comparisons being performed, a correction method such as Bonferroni or False Discovery Rate (FDR) should be applied and reported.

3. Association Strength and Significance: The strength and significance of correlations should be reported comprehensively, including effect sizes and confidence intervals.

Results Interpretation: The findings of significant differences in mtDNA MRG expression between HCC samples and normal controls, as well as the identification of three molecular subtypes with distinct clinical and immune profiles, are intriguing. However, the interpretation of results should be cautious and conservative:

1. Biological Relevance: The biological implications of the molecular subtypes and their correlation with TME features need to be discussed in the context of existing literature.

2. Causal Inference: The observational nature of the study limits the ability to infer causality. Any discussion of causative mechanisms should be framed within this limitation.

3. External Validation: Although not feasible in this study, future directions should include external validation of the mtDNA score model in independent cohorts.

Discussion: The discussion should address the limitations of the study and consider the broader implications of the findings. It is important to discuss how the results fit into the current understanding of mtDNA MRGs in HCC and what further research is needed.

In conclusion, while the manuscript provides valuable insights into the role of mtDNA MRGs in HCC, several methodological and analytical improvements are necessary for publication. The authors should address these concerns thoroughly in a major revision before the manuscript can be considered for acceptance.

Reviewer #2: Title: Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment features

This paper presents a study to focus on exploring the important role of Mitochondrial DNA maintenance-related genes (mtDNA MRGs) in the hepatocellular carcinoma (HCC). Authors proposed mtDNA score model which can be used to assess tumor immune cell invasion, clinicopathological features, genetic variation, and immunotherapy response. However, there are questions that limit my enthusiasm for the paper, as outlined below.

- Data section needs to be improved

o Add more details regarding the batch correction method used.

o Include a figure showing data before and after batch correction (as supplementary material).

o Specify the type of RNA data used (RNA-seq or microarray).

o Clarify the type of normalization applied to the data.

- The Method section needs to be improved. Authors mentioned the packages used without providing details about the methods. It should not be assumed that all readers are familiar with or can recall the methods implemented in each package (e.g., limma, ConsensusClusterPlus).

- The clustering step using samples and genes is not clear and easy to follow. Please add details of the method, including the type of clustering (e.g., hierarchical clustering) and the distance metric used.

- Additionally, add dendrograms to the heatmaps (e.g., Figures 2E and 5D) to better illustrate the clusters found and align with the findings.

- The mtDNA score model is not clear. Please specify on which subset of genes the PCA analysis was applied. Additionally, Figure 3 related to score modeling and TME was not shared and is missing.

- What does the y-axis label show in Figures 9B-E? Is it the IPS signature score? Additionally, there are more ICB signatures or biomarkers to consider, including PDCD1, CD274, PDCD1LG2, CTLA4, TIM3, LAG3, TIGIT, and resources like bhklab/SignatureSets: Compendium of published molecular signatures (github.com) and/or https://pubmed.ncbi.nlm.nih.gov/36055464/

- A GitHub repository or another container to access all the code is needed.

- Validation is highly recommended.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean? ). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy .

Reviewer #1: No

Reviewer #2: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/ . PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org . Please note that Supporting Information files do not need this step.

PLoS One. 2025 Jun 2;20(6):e0325033. doi: 10.1371/journal.pone.0325033.r003

Author response to Decision Letter 1

12 Oct 2024

Response to Editor

Thanks for your kind suggestions. We have read your suggestions carefully and try our best to give the following response:

Q1. Please define the mtDNA score, TMB and TME (lines 25-17).

Response: Thank you for your suggestions. The tumor microenvironment (TME) was defined (lines 73-75 in the revised manuscript). mtDNA score was defined (lines 185-186 in the revised manuscript). Tumor mutational burden (TMB) reflects cancer mutation quantity (doi:10.1016/j.ccell.2020.10.001), now TMB is emerging as a potential biomarker (doi:10.1093/annonc/mdy495). As a common name, in many articles, TMB is directly cited (doi: 10.1016/j.ccell.2023.09.006; doi:10.7150/thno.52717).

Q2. I suggest you use the keyword mitochondrion and not mitochondrial (line 44).

Response: Thank you for your suggestion. the key word “mitochondrial” has been replaced with “mitochondrion”.

Q3.“own superhelical double-linked 61 ring of genetic material” is unnecessary (line 60)

Response: Thank you for your suggestion. The sentence “A unique feature of mitochondria is that they have their own superhelical double-linked ring of genetic material, called mitochondrial DNA (mtDNA)” has been deleted.

Q4. Move gene list from lines 115-118 to a separate table.

Response: Thank you for your suggestion. If we move gene list to a separate table, there's not much in this table. So, we categorized these genes according to their functions. Maybe it's more categorical and concise. mtDNA replication (DNA2, EXOG, LIG3, POLG, POLG2, POLRMT, RNASEH1, SSBP1, TFAM, TFB2M, TOP1MT, TOP3A, TWNK)�mtDNA repair (APEX1, EXOG, LIG3, OGG1, POLG, PRIMPOL, RECQL4, UNG), mtDNA modifications (METTL4), mtDNA stability and decay (ENDOG, EXOG, MGME1) (lines 125-128 in the revised manuscript).

Q5. Describe ssGSEA in more detail in the Methods section.

Response: Thank you for your comments. The following sentence has been added the Methods section to describe ssGSEA in more detail. ssGSEA was used to calculate the standardized enrichment score via the GSVA package50.The relative abundance of each immune cell type was represented by an enrichment score in ssGSEA analysis and normalized to unity distribution from 0 to 1(doi: 10.7150/thno.52717)(lines 167 to 169).

Q6. What is this data set from Charoenteng? Add a reference to it.

Response: Thank you for your suggestion. Pornpimol Charoentong, et al analyzed tumor-immune cell interactions in 20 solid cancers, revealing the relationship between genotype and immunophenotype and created The Cancer Immunome Atlas (https://tcia.at/). Using machine learning, the determinants of tumor immunogenicity were identified and a quantitative scoring scheme called immunophenotypic scoring was developed. The according reference (doi: 10.1016/j.celrep.2016.12.019) has been added (line 156).

Q7. Why was an adjusted p-value of 0.001 used? (line 143)

Response: We referred to many similar literatures and found that using an adjusted p-value of 0.001 is more reasonable (doi: 10.1186/s12943-020-01170-0; doi: 10.3389/fcell.2022.842220).

Q8. Lines 166-176: indentation is problematic

Response: Thank you for your comment. All the indentation has been deleted in the revised manuscript.

Q9. you mention 8293 DEGs between clusters A and B and 2043 DEGs between clusters A and C, and 9064 between clusters B and C. Why are there so many? What was the fold change cutoff limit? What would happen if you increase this cutoff limit?

Response: Although the consensus clustering algorithm based on mtDNA MRGs expression classified HCC patients into three mtDNA MRG phenotypes, the underlying genetic alterations and expression perturbations within these phenotypes were not well known, so there are so many related genes. P-values were adjusted according to the false discovery rate (FDR), and the P-value less than 0.001 was taken as the screening criteria. If we increase this cutoff limit, more DEGs related genes will be excluded.

Q10. Starting from line 317 you describe the role of TP53 in cancer. This is well known. What is the novelty of your findings? As you write, HCC is very heterogeneous, what other genes play a role in HCC that you found?

Response: The novelty of our findings lies in the following aspects. 1) In the high mtDNA score group, TP53 mutation frequency is the highest, and the poor prognosis of HCC patients in the High mtDNA score group may be related to the high TP53 mutation. 2) Among HCC patients with lower mtDNA scores, HCC patients with TP53 mutations had worse clinicopathological features compared to CTNNB1-mutated HCC patients.

Besides TP53, two genes RECQL4 and APEX1 have been found play a role in HCC, HCC patients with higher expression of RECQL4 and APEX1 have a worse prognosis according with the previous studies (doi:10.3892/ol.2018.78600; doi:10.1016/j.immuni.2019.12.011). In our data, RECQL4 and APEX1 are highly expressed in HCC samples, and the higher expression in RECQL4 and APEX1, the worse prognosis of the HCC patients, which is consistent with the results of other studies. So, our research is feasible.

Response to Reviewer 1

Reviewer 1

Q1. Sample Selection and Control Group: While the use of TCGA and GEO datasets is commendable, the inclusion criteria for selecting HCC samples should be more clearly described. Moreover, the control group's characteristics should also be provided for comparison purposes.

Response: Dear reviewer, thanks for your kind suggestions. HCC samples in TCGA were collected from patients pathologically diagnosed with hepatocellular carcinoma (HCC) after surgical resection and who had not previously received treatment (ablation, chemotherapy, or radiotherapy) specific to their disease. Cases were staged according to the American Joint Committee on Cancer (AJCC). Each frozen primary tumor specimen had a companion normal tissue specimen (blood or blood components, including DNA extracted at the tissue source site). Adjacent tissue was submitted for some cases. Hematoxylin and eosin (H&E) stained sections from each sample were subjected to independent pathology review to confirm that the tumor specimen was histologically consistent with the allowable hepatocellular carcinomas and the adjacent tissue specimen contained no tumor cells. Adjacent tissue with cirrhotic changes was not acceptable as a germline control, but was characterized if accompanied by DNA from a patient matched blood specimen. The percent tumor nuclei, percent necrosis, and other pathology annotations were also assessed. Tumor samples with ≥ 60% tumor nuclei and ≤ 20% or less necrosis were submitted for nucleic acid extraction (doi:10.1016/j.cell.2017.05.046). GSE76427 cohort with 115 HCC samples was downloaded from the GEO database as a validation cohort after the removal of normal tissue samples and tumor samples without follow-up and outcome status information.115 primary HCC tissues and 52 adjacent tissues (Dataset GSE76427, Eastern population) from GEO were enrolled in this study. Percentage of HCC patients with HBV infection and cirrhosis were 46% and 54%, respectively (doi:10.1016/j.heliyon.2023.e14460;doi:10.1002/1878-0261.12153). All HCC samples were downloaded from TCGA and GEO. The inclusion criteria for selecting HCC samples and control group's were HCC patients with survival information in the databases.

Q2. Screening of mtDNA MRGs: The process of identifying the 22 mtDNA MRGs requires further detail. How were these genes prioritized over others? What criteria were used for screening?

Response: Thank you for your suggestions. The process of identifying the 22 mtDNA MRGs was described in more detail as “The list of mtDNA MRGs was obtained from the MitoCarta3.0 database, including 27 mtDNA MRGs for mtDNA replication (DNA2, EXOG, LIG3, MGME1, POLG, POLG2, POLRMT, RNASEH1, SSBP1, TFAM, TFB2M, TOP1MT, TOP3A, TWNK)�mtDNA repair (APEX1, EXOG, LIG3, OGG1, POLB, POLG, PRIMPOL, RECQL4, UNG), mtDNA modifications (METTL4), mtDNA stability and decay (ENDOG, EXOG, MGME1), respectively. The expression profile of these genes was systematically extracted and analyzed in normal and tumor samples. Finally, 22 mtDNA MRGs with statistical significance were identified, with the function of not only packaging mitochondrial DNA, but also providing a stable environment for mitochondrial DNA replication, transcription and repair, so that they are superior to other mtDNA MRGs (doi:10.1016/j.biochi.2013.09.017).

Q3. Unsupervised Clustering: Details about the machine learning model employed for unsupervised clustering need to be elaborated upon, including parameters and validation methods.

Response: Thank you for your suggestions. Unsupervised Clustering was performed using the R package“Consensus Cluster Plus” and 1000 cycles were undertaken to ensure the stability of the classification; the number of cluster k values were increased from 2 to 9. The k = 3 with better clustering stability were selected according to the clustering effect (doi: 10.1093/bioinformatics/btq170), the detailed information has been added to the revised manuscript.

Q4. PCA Model Construction: The construction of the mtDNA score model using PCA should be detailed, including how the principal components were selected and validated.

Response: Thank you for your suggestions. In this study, the mtDNA score model was constructed using PCA refer to previous studies, the overlapping DEGs identified from different mtDNA clusters were selected and employed to perform prognostic analysis for each gene using a univariate Cox regression model. The genes with a significant prognostic impact were extracted for further feature selection by using recursive feature elimination (RFE) with random forest and the 10-fold cross -validation method in the‘caret’package. Then the expression profile of the final determined genes was curated to perform PCA analysis, and principal components 1 and 2 were extracted and served as the signature score. This method mainly focuses on the score on the set with the largest block of well correlated (or inverse-correlated) genes in the set, while downweighting contributions from genes that do not track with other set members. And finally, mtDNA score was obtained by inputting core genes according to the corresponding code (doi:10.7150/thno.52717;doi: 10.1186/s12943-020-01170-0).

Statistical Analysis: The statistical methods used appear appropriate for the data type and research question. Nevertheless, the following points should be addressed:

Q1. Statistical Power: It would be beneficial to include a discussion on the statistical power of the study given the sample size and the number of variables being analyzed.

Response: Thank you for your suggestion. In the method, the statistical parameters are annotated.

Q2. Multiple Testing Correction: With multiple comparisons being performed, a correction method such as Bonferroni or False Discovery Rate (FDR) should be applied and reported.

Response: Thank you for your suggestion. In this study, FDR has been applied and reported (lines167 to 169).

Q3. Association Strength and Significance: The strength and significance of correlations should be reported comprehensively, including effect sizes and confidence intervals.

Response: Thank you for your suggestions. The strength and significance of correlations were reported comprehensively, and the effect sizes and confidence intervals have been added in the revised manuscript.

Q1.Biological Relevance: The biological implications of the molecular subtypes and their correlation with TME features need to be discussed in the context of existing literature.

Response: Thank you for your suggestions. The three molecular subtypes correspond to different classifications of the tumor microenvironment and have different biological processes and prognosis (lines 326 to 346).

Q2. Causal Inference: The observational nature of the study limits the ability to infer causality. Any discussion of causative mechanisms should be framed within this limitation.

Response: Thank you for your suggestions. All discussions of causative mechanisms were based on objective analysis of the expression of 22 mtDNA MRGs in HCC samples in public databases supported by existing articles (doi:10.3389/fcell.2022.785058;doi:10.1016/j.compbiomed.2023.106831;doi:10.1016/j.celrep.2016.12.019).

Q3. External Validation: Although not feasible in this study, future directions should include external validation of the mtDNA score model in independent cohorts.

Response: The lack of external validation is a shortcoming of this study. In the future, the existing results will be analyzed and some genes will be extracted for in vitro and in vivo experiments.

Response to Reviewer 2

Reviewer 2

Data section needs to be improved:1.Add more details regarding the batch correction method used.2. Include a figure showing data before and after batch correction (as supplementa-ry material).3.Specify the type of RNA data used (RNA-seq or microarray).4. Clarify the type of normalization applied to the data.

Response: Thank you for your suggestions.

1. Details regarding the batch correction method used in this study has been added, and changes in the revised manuscript were marked as red font.

2. A supplementary material has been provided, which including a figure showing data before and after batch correction (As shown in the following).

3. TCGA RNA sequencing data (FPKM format) were downloaded from the Genomic Data Commons. For the GSE76427 data from the Illumina platform, the normalized matrix file was directly downloaded.

4. The type of normalization that applies to the data has been clarified in the revised manuscript.

Supplementary Fig. 1 Data set of HCC samples before and after batch correction. A. Batch of HCC samples before correction. B. Batch of HCC samples after correction

The Method section needs to be improved. Authors mentioned the packages used without providing details about the methods. It should not be assumed that all readers are familiar with or can recall the methods implemented in each package (e.g., limma, ConsensusClusterPlus).

Response: Thank you for your suggestions. Details of the method have been added in the revised manuscript. If necessary, we will provide the corresponding code as a reference.

The clustering step using samples and genes is not clear and easy to follow. Please add details of the method, including the type of clustering (e.g., hierarchical clustering) and the distance metric used.

Response: Thank you for your suggestions. Details of the method have been added(lines 110 to 116, lines 139 to 143).

Additionally, add dendrograms to the heatmaps (e.g., Figures 2E and 5D) to better illustrate the clusters found and align with the findings.

Response: Thank you for your suggestion. Dendrograms reflect the similarities or correlations between members within a cluster. Figures 2E and 5D mainly showed the relationship between mtDNA MRGs molecular subtypes, genotypes and clinicopathological stages. Thus, there is no need to add dendrograms.

What does the y-axis label show in Figures 9B-E? Is it the IPS signature score? Additionally, there are more ICB signatures or biomarkers to consider, including PDCD1, CD274, PDCD1LG2, CTLA4, TIM3, LAG3, TIGIT, and resources like bhklab/SignatureSets: Compendium of published molecular signatures (github.com) and/or.

Response: In figures 9B-E, y-axis labels show the IPS signature scores of anti-PD-1 (ctla4-neg-pd1-pos), anti-CTLA-4(ctla4-pos-pd1-neg), or combined anti-CTLA-4/PD-1 (ctla4-pos-pd1-pos) immunotherapy. To promote understanding of tumor-immune cell interactions, Pornpimol Charoentong, et al characterized the intratumoral immune landscapes and the cancer

Attachment

Submitted filename: Response to the reviewers.docx

pone.0325033.s002.docx^{(475.7KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0325033.r004

Decision Letter 1

Matthew Cserhati

27 Nov 2024

PONE-D-24-22922R1Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment featuresPLOS ONE

Dear Dr. Meng,

Please submit your revised manuscript by Jan 11 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org . When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Matthew Cserhati, Ph.D

Academic Editor

PLOS ONE

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #3: (No Response)

Reviewer #4: (No Response)

Reviewer #5: (No Response)

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #3: Partly

Reviewer #4: Yes

Reviewer #5: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: No

Reviewer #4: No

Reviewer #5: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

Reviewer #3: No

Reviewer #4: Yes

Reviewer #5: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

Reviewer #3: No

Reviewer #4: Yes

Reviewer #5: Yes

**********

6. Review Comments to the Author

Reviewer #3: Some issues mentioned by the reviewers have not been addressed, appropriately and comprehensively. For example, Q1 and Q4 from Reviewer 1.

Reviewer #4: (No Response)

Reviewer #5: Thank you for inviting me to review the paper. In this article, the authors explored correlation between mitochondrial DNA maintenance-related genes and HCC tumor features as well as clinical features utilizing multiple bioinformatic methods. Authors have already done edits and improved paper quality as per previous review history. However, closer scrunity reveals there are several drawbacks which I would like authors to revise.

1. There are several flaws in figures and figure legends.

1.1 Please rotate Figure 5A to correct the orientation.

1.2 Figure 7E should be Figure 7D.

1.3 In Figure 8, letter G of Figure 8G is in Figure 8C.

2. Authors should specify how they identified 22 mtDNA MRGs out of 1136 mitochondrial genes in MitoCarta3.0 (https://personal.broadinstitute.org/scalvo/MitoCarta3.0/human.mitocarta3.0.html) since this is the very foundation of this article. Though they have answered this question to Reviewer 1 in previous review round, it seems that their response was not logically sound enough to convince me. Could authors inform me why mtDNA genes related to mtDNA replication, repair, modifications were selected as representative mtDNA MRGs, instead of those involved in other pathways, e.g. immune response?

3. In paragraph 2 (line 56-72) of the Introduction, since authors have already mentioned mitochondria, mtDNA and mtDNA MRGs, why was paragraph 4 (line 81-96) created to describe mtDNA MRGs again?

4. How would authors explain mtDNA MRG clusters A and C had better survival compared with cluster B, while A and B were identified as immune rejection phenotype and immunosuppression phenotype, respectively, which is associated with lack of response and/or resistance to immunotherapy? What made mtDNA MRG cluster A possess better survival than B? I don't think the argument on this point in the first paragraph (line 286-308) of Discussion was strong enough to clarify the difference.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean? ). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy .

Reviewer #3: No

Reviewer #4: No

Reviewer #5: No

**********

Attachment

Submitted filename: comments to the authers PONE-D-24-22922.docx

pone.0325033.s003.docx^{(14.8KB, docx)}

PLoS One. 2025 Jun 2;20(6):e0325033. doi: 10.1371/journal.pone.0325033.r005

Author response to Decision Letter 2

6 Feb 2025

Dear Editor and Reviewers:

Thank you very much for your efforts to review our manuscript entitled “Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment features”. Those comments are all valuable and very helpful for revising and improving our paper. According to the reviewers’comments, we have included a point-by-point response to each comment.

We appreciate the suggestions and comments by the Editor and the Reviewers. We would like to express our great appreciation to you and reviewers for comments on our paper, Looking forward to hearing from you.

Yours sincerely,

Zhongji Meng

Department of Infectious Disease, Taihe Hospital, Hubei University of Medicine

No. 32. South Renmin Road, Shiyan 442009, China.

E-mail: zhongji.meng@163.com

Reviewer #3

Some issues mentioned by the reviewers have not been addressed, appropriately and comprehensively. For example, Q1 and Q4 from Reviewer #1.

Q1 from Reviewer #1. Sample Selection and Control Group: While the use of TCGA and GEO datasets is commendable, the inclusion criteria for selecting HCC samples should be more clearly described. Moreover, the control group's characteristics should also be provided for comparison purposes.

Response: Thank you for your valuable feedback regarding the sample selection and control group in our study. Why did we choose TCGA and GEO datasets? Because we needed to do prognostic analysis, so no prognostic data set was excluded. The larger the sample size, the smaller the error. Finally, we choosed TCGA and GEO Inclusion Criteria for HCC Samples: HCC samples in TCGA were collected from patients pathologically diagnosed with hepatocellular carcinoma (HCC) after surgical resection and who had not previously received treatment (ablation, chemotherapy, or radiotherapy) specific to their disease. Cases were staged according to the American Joint Committee on Cancer (AJCC) (doi:10.1016/j.cell.2017.05.046).Data were obtained for 374 HCC samples (Western population) and 50 normal samples from the TCGA on June 17, 2023, and 115 primary HCC tissues (Eastern population) and 52 adjacent tissues from the GEO (GSE76427) cohort(doi:10.1186/s12967-022-03630-1).Clinical Data Availability: Only patients with complete clinical information, including age, gender, tumor stage, and survival data, were included.Exclusion Criteria: Samples with incomplete data, non-primary tumors, or patients who received preoperative treatments were excluded to ensure data consistency and reliability.

Control Group Characteristics: Control samples were derived from adjacent non-tumorous liver tissues available within the TCGA. Clinical Parameters: Detailed clinical characteristics, including liver function tests (ALT, AST levels), presence of underlying liver diseases (e.g., hepatitis B/C, cirrhosis), and lifestyle factors (e.g., alcohol consumption), have been provided to facilitate comprehensive comparisons with the HCC samples. Matching Criteria: Controls were matched to HCC cases based on age and gender to minimize potential confounding factors(doi:10.1016/j.heliyon.2023.e14460;doi:10.1002/1878-0261.12153).

The details are as follows�

Characteristic TCGA Group (N = 352) GEO Group (N = 114) P-value

Age (years) 0.07

Mean (SD) 61.5 (13.1) 63.5 (12.7)

Gender 0.069

Male (%) 192 (55%) 60 (52.6%)

Female (%) 160 (45%) 54 (47.4%)

Stage 0.32

Stage_I (%) 175 (49.7%) 55 (48.2%)

Stage_II (%) 86 (24.4%) 35 (30.7%)

Stage_III (%) 86 (24.4%) 21 (18.4%)

Stage_IV (%) 5 (1.5%) 3 (3.7%)

Q4 from Reviewer 1. PCA Model Construction: The construction of the mtDNA score model using PCA should be detailed, including how the principal components were selected and validated.

Response: Thank you for your suggestions. To quantify the characteristics of mtDNA MRGs in individual HCC patients, a scoring system called the mtDNA score was constructed using principal component analysis (PCA). First of all, based on the mtDNA MRGs expression levels, HCC samples were classified for further analysis by using the “ConsensusClusterPlus” package in R for unsupervised cluster analysis, and 1000 repetitions were performed with pltem=0.8. K=3 was the best cluster value. To verify the stability of the subtypes, Cumulative distribution function (CDF) and Delta plot were used.CDF display consensus distributions for each k. When k takes any value, CDF reaches an approximate maximum, and the cluster analysis results are the most reliable at this time. That is, the k value with small CDF descent slope is considered. The Delta plot shows that the inflection point of change in cluster stability with an increase in k value usually represents the optimal number of clusters(doi:10.1093/bioinformatics/btq170).Then, through the above analysis , the overlapping DEGs identified from different mtDNA clusters were selected and employed to perform prognostic analysis for each gene using a univariate Cox regression model. The genes with a significant prognostic impact were extracted for further feature selection by using recursive feature elimination (RFE) with random forest and the 10-fold cross -validation method in the‘caret’package. Then the expression profile of the final determined genes was curated to perform PCA analysis, and principal components 1 and 2 were extracted and served as the signature score(1 representatived mtDNAGeneExp. 2 representatived mtDNACluster). This method mainly focuses on the score on the set with the largest block of well correlated (or inverse-correlated) genes in the set, while downweighting contributions from genes that do not track with other set members. And finally, mtDNA score was obtained by inputting core genes according to the corresponding code (doi:10.7150/thno.52717;doi: 10.1186/s12943-020-01170-0).

Statistical Analysis: The statistical methods used appear appropriate for the data type and research question. Nevertheless, the following points should be addressed:

Q1. Statistical Power: It would be beneficial to include a discussion on the statistical power of the study given the sample size and the number of variables being analyzed.

Response: Thank you for your suggestion. In the method, the statistical parameters are annotated.

Q2. Multiple Testing Correction: With multiple comparisons being performed, a correction method such as Bonferroni or False Discovery Rate (FDR) should be applied and reported.

Response: Thank you for your suggestion. When making multiple comparisons, we agree to use methods such as Bonferroni or FDR (False Discovery Rate) for verification. The FDR approach is more powerful than methods like the Bonferroni procedure that control false positive rates. Controlling the false discovery rate in a study that arguably consisted of scientifically driven hypotheses found nearly as many significant results as without any adjustment, whereas the Bonferroni procedure found no significant results(doi: 10.1016/j.jclinepi.2014.03.012).False discovery rate(FDR) control has become an increasingly standard practice in genomic research and analysis of microarray data for extensive testing(doi: 10.1016/j.jclinepi.2007.04.017). We have applied the FDR to adjust for multiple testing and have included this information in the revised manuscript(lines143 to 145, line163).

Q3. Association Strength and Significance: The strength and significance of correlations should be reported comprehensively, including effect sizes and confidence intervals.

Q1.Biological Relevance: The biological implications of the molecular subtypes and their correlation with TME features need to be discussed in the context of existing literature.

Response: Thank you for your suggestions. HCC patients with enrichment of the three mtDNA MRG clusters had different prognoses and TME immune cell infiltration characteristics. mtDNA MRG cluster A is characterized by enhanced tumor matrix activity and abundant innate immune cell infiltration, corresponding to the immune rejection phenotype; mtDNA MRG cluster B is characterized by immunosuppression, corresponding to the immune-desert phenotype; and mtDNA MRG cluster C is characterized by adaptive immune cell infiltration and immune activation, corresponding to the immunoinflammatory phenotype(doi:10.1016/j.immuni.2019.12.011). The immune rejection and immune-desert phenotypes are considered to indicate noninflammatory tumors, primarily tumors that lack immune cell invasion in the parenchyma and stroma, rarely express PD-L1, are located at the opposite end of the tumor immune continuum, and histologically lack immune invasion and antigen presentation (low MHC class I), while exhibiting high tumor cell proliferation. The immunoinflammatory phenotype indicates inflammatory tumors, in which the TME has a high degree of infiltration of immune cells, such as T cells, CD8+ T cells producing IFN-γ,and PD-1-positive immune cells(doi:10.1158/1078-0432.CCR-15-1507; 10.18632/aging.203456). There were significant differences in the TME features among samples enriched in the three mtDNA MRG clusters.

Q2. Causal Inference: The observational nature of the study limits the ability to infer causality. Any discussion of causative mechanisms should be framed within this limitation.

Response: Thank you for your suggestions. All discussions of causative mechanisms were based on objective analysis of the expression of 22 mtDNA MRGs in HCC samples in public databases supported by existingarticles(doi:10.3389/fcell.2022.785058;doi:10.1016/j.compbiomed.2023.106831;doi:10.1016/j.celrep.2016.12.019). In the revised manuscript, we have made changes to some of the results(lines 363 to 364; lines 404 to 405; lines 410 to 411).

Q3. External Validation: Although not feasible in this study, future directions should include external validation of the mtDNA score model in independent cohorts.

Response: Thank you for your suggestions.The lack of external validation is a shortcoming of this study. Plan for future external validation: In our future research, we will conduct external validation of the mtDNA score model using independent cohorts. We will collect data from various regions and populations to assess the model's effectiveness and stability in different settings. We plan to select at least two independent cohorts for validation, coming from different geographic regions and demographic backgrounds, to assess the generalizability and reliability of the mtDNA score model in various settings. Additionally, we will conduct stratified analysis (based on age, gender, ethnicity, etc) to assess how the model performs in different subgroups. We will also incorporate cross-validation techniques to increase the reliability and accuracy of the evaluation. Based on previous research, We will select representative genes (such as SSBP1, OGG1, etc.) for further verification in the cohort to verify the stability of mtDNA score(doi:10.1002/jcla.24561;10.1172/JCI128513).

Response to Reviewer #5

Reviewer 5

1. There are several flaws in figures and figure legends.

1.1 Please rotate Figure 5A to correct the orientation.

Response: Thank you for your suggestions. We have rotated Figure 5A to the normal direction

1.2 Figure 7E should be Figure 7D.

Response: Thank you for reminding me. We apologize for this careless mistake. We have corrected Figure7E to 7D.

1.3 In Figure 8, letter G of Figure 8G is in Figure 8C.

Response: Sorry for our carelessness. We have removed the letter G in picture 8C.

Response: We downloaded 1136 mitochondrial genes from the MitoCarta3.0database. Data for all 1136 human genes with high confidence of mitochondrial localization (based on integrated proteomics, computation, and microscopy). We focused on MitoPathways, Hierarchy of biological pathways and list of MitoCarta3.0 genes assigned to each pathway.Although mitochondrial repair mechanisms are highly efficient, the mitochondrial genome is highly sensitive to oxidative damage and other exogenous and endogenous induced DNA damage due to the lack of protective histones and their proximity to major reactive oxygen species (ROS) production sites. Mutations in mitochondrial replication ,repair, modifications and and stability genes can lead to depletion and deletion of mtDNA, which in turn leads to instability of the mitochondrial genome. The combination of mutations and deletions can lead to impaired mitochondrial genome maintenance and trigger various mitochondrialdiseases(doi:10.1016/j.bbabio.2022.148554;10.1161/CIRCULATIONAHA.123.068358).The expression profile of these genes was systematically extracted and analyzed in normal and tumor samples. Finally, 22 mtDNA MRGs with statistical significance were identified. The chosen genes have been extensively documented in existing literature to be associated with mtDNA maintenance, supported by substantial experimental evidence(doi: 10.1016/j.bbamcr.2021.119167; 10.1038/s41582-018-0101-0).

Although mitochondrial genes also play roles in other pathways, such as immune response, the focus of our study is on the maintenance mechanisms of mtDNA. Therefore, we concentrated on genes directly involved in mtDNA replication, repair, and modifications.

3. In paragraph 2 (line 56-72) of the Introduction, since authors have already mentioned mitochondria, mtDNA and mtDNA MRGs, why was paragraph 4 (line 81-96) created to describe mtDNA MRGs again?

Response: Thank you for your question. line 56-72 focuses on the function of mitochondria and the importance of mtDNA MRGs. Lines 81-96 describe mtDNA MRGs again, mainly because we selected representative genes with research basis for their role in HCC, which is consistent with our subsequent studies. At the same time, we have removed duplicate descriptions.

Response: Thank you for your question. Most human solid tumours exhibit one of three distinct immunological phenotypes:immune inflamed, immune excluded, or immune desert(doi: 10.1038/nature14011). Immunomonitoring can help researchers and clinicians better understand the tumor's response to the immune system, including the presence of tumor-infiltrating lymphocytes (TILs), the activity of CD8+ T cells, and the expression of PD-L1, which can evaluate the effectiveness of immunotherapy an

PLoS One. doi: 10.1371/journal.pone.0325033.r006

Decision Letter 2

Matthew Cserhati

25 Feb 2025

PONE-D-24-22922R2Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment features

PLOS ONE

Dear Dr. Meng,

The authors addressed an interesting topic used appropriate methodology, highlighted the strengths of their study, and presented some interesting results and conclusions.Overall, this is a clear, concise, and well-written manuscript. This article covers lots of information. On the one hand, many general statements are made without proper references. Please try to remove some of the general statements or speculations but add some details for some key refs for better illustration. I have some concerns and remarks that I hope the authors can address to improve the paper The title and abstract cover the central aspect of the workBut is better to have a graphical abstract to simplify the idea of the studied methodology -The introduction is relevant, and theory based. But: Sufficient information about the previous study findings should be presented for readers to follow the present study rationale and procedures. - The methods are generally appropriate.- Please add the Data analysis and machine learning models- I cant reach how These analyses were done and what was the used software version ?- How many machine learning models were used?-how the performance of the oPLS-DA model was assessed ? The results are clear, , the current study has innovations and advantages. However,- the quality of the figures is poor and should be rectified ( very very poor)- please mention all abbreviations under the figures-Points located far from the main cluster for their group may be potential outliers, indicating samples that do not follow the group trend. These samples should be investigated further, as they may represent unique biological conditions or data quality issues.- Parameters for evaluation of each machine learning model discriminating pairs of groups should be clarified The performed study and results could have implication for research and clinical practice in the future. Therefore, the present article could be accepted for publication after revision and correction of the figures============2nd set of questions: 1. “Pornpimol Charoentong, et al analyzed tumor-immune cell interactions in 20 solid cancers, revealing the relationship between genotype and immunophenotype and created The Cancer Immunome Atlas (https://tcia.at/). Using machine learning, the determinants of tumor immunogenicity were identified and a quantitative scoring scheme called immunophenotypic scoring was developed.” Please add this explanation to the section titled “Analysis of cell differences in TME”2. Lines 176-177 of revised manuscript: did you use the Draw Equation tool? Please put ‘i’ into subscript. Also, does the sigma have any initial and terminal values, i.e. i=0 .. n? Please start line 177 with “Where i represents the expression of mtDNA…”3. In response to question #9, I do apologize, I stated my question incorrectly. I wanted to ask, how many genes do you increase the fold-change cutoff? If the fold change is higher, you should get fewer genes.

Please submit your revised manuscript by Apr 11 2025 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org . When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

We look forward to receiving your revised manuscript.

Kind regards,

Matthew Cserhati, Ph.D

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments:

The authors addressed an interesting topic used appropriate methodology, highlighted the strengths of their study, and presented some interesting results and conclusions.

Overall, this is a clear, concise, and well-written manuscript. This article covers lots of information. On the one hand, many general statements are made without proper references. Please try to remove some of the general statements or speculations but add some details for some key refs for better illustration.

I have some concerns and remarks that I hope the authors can address to improve the paper

The title and abstract cover the central aspect of the work

But is better to have a graphical abstract to simplify the idea of the studied methodology

-The introduction is relevant, and theory based. But:

Sufficient information about the previous study findings should be presented for readers to follow the present study rationale and procedures.

- The methods are generally appropriate.

- Please add the Data analysis and machine learning models

- I cant reach how These analyses were done and what was the used software version ?

- How many machine learning models were used?

-how the performance of the oPLS-DA model was assessed ?

The results are clear, , the current study has innovations and advantages. However,

- the quality of the figures is poor and should be rectified ( very very poor)

- please mention all abbreviations under the figures

-Points located far from the main cluster for their group may be potential outliers, indicating samples that do not follow the group trend. These samples should be investigated further, as they may represent unique biological conditions or data quality issues.

- Parameters for evaluation of each machine learning model discriminating pairs of groups should be clarified

The performed study and results could have implication for research and clinical practice in the future. Therefore, the present article could be accepted for publication after revision and correction of the figures

============

2nd set of questions:

1. “Pornpimol Charoentong, et al analyzed tumor-immune cell interactions in 20 solid cancers, revealing the relationship between genotype and immunophenotype and created The Cancer Immunome Atlas (https://tcia.at/). Using machine learning, the determinants of tumor immunogenicity were identified and a quantitative scoring scheme called immunophenotypic scoring was developed.” Please add this explanation to the section titled “Analysis of cell differences in TME”

2. Lines 176-177 of revised manuscript: did you use the Draw Equation tool? Please put ‘i’ into subscript. Also, does the sigma have any initial and terminal values, i.e. i=0 .. n? Please start line 177 with “Where i represents the expression of mtDNA…”

3. In response to question #9, I do apologize, I stated my question incorrectly. I wanted to ask, how many genes do you increase the fold-change cutoff? If the fold change is higher, you should get fewer genes.

[Note: HTML markup is below. Please do not edit.]

PLoS One. 2025 Jun 2;20(6):e0325033. doi: 10.1371/journal.pone.0325033.r007

Author response to Decision Letter 3

10 Apr 2025

Dear Editor:

Thank you very much for your efforts to review our manuscript entitled “Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment features”. We have diligently addressed all the comments, and incorporated appropriate revisions into the revised manuscript. The revisions have significantly strengthened the manuscript's scientific rigor and clarity. Enclosed below please find a point-by-point response document detailing how each suggestion has been implemented. The revised manuscript with all modifications clearly highlighted in red has been resubmitted to the system. We sincerely appreciate the reviewers' insightful feedback and trust that the revised manuscript now meets the journal's publication standards.

Yours sincerely,

Zhongji Meng

Department of Infectious Disease, Taihe Hospital, Hubei University of Medicine

No. 32. South Renmin Road, Shiyan 442009, China.

E-mail: zhongji.meng@163.com

Additional Editor Comments:

I have some concerns and remarks that I hope the authors can address to improve the paper

The title and abstract cover the central aspect of the work

But is better to have a graphical abstract to simplify the idea of the studied methodology

Response: Thank you for your suggestions. The simplified graphical abstract is as follows:

The introduction is relevant, and theory based. But:

Sufficient information about the previous study findings should be presented for readers to follow the present study rationale and procedures.

Response: Thank you for your suggestions. Based on some of the investigators' previous studies (lines 84 to 93), we found that mitochondrial DNA maintenance-related genes (mtDNA MRGs) are closely related to the occurrence and progression of HCC. While the complex interactions among mtDNA MRGs in the development and progression of HCC are unclear, so that it’s necessary to investigate the mechanism of mtDNA MRGs in TME of HCC.

- The methods are generally appropriate,

- - Please add the Data analysis and machine learning models

- I cant reach how These analyses were done and what was the used software version ?

How many machine learning models were used?

Response: All the statistical analyses were performed using R version 4.2.2. In our study, Consensus Clustering and Principal Component Analysis (PCA) of unsupervised machine learning models are mainly used. In the analysis of Consensus Clustering, specific important parameters are set as follows: maxK=9, reps=50, pItem=0.8, pFeature=1.

-how the performance of the oPLS-DA model was assessed ?

Response:

1. Data correction (before/after removal of batch effect)

Problem: Different data sets (such as TCGA and GSE76427) have a "batch effect" due to differences in experimental conditions, resulting in false positive results.

The solution: Before correction: 89.7% explained by Dim1 (dominated by technical differences). After correction: Dim1 decreased to 18.9% (biological differences highlighted).

Significance: To eliminate technical interference and ensure that the differences found later truly reflect biological characteristics.

2. Data integration (oPLS-DA score chart)

Key points: Integrating multiple omics data (gene expression + clinical typing).

Graphic interpretation: Horizontal/vertical axis: principal components t1 (3%) and t2 (16%).

Sample distribution: TCGA (red) and GSE76427 (blue) partially overlap but are generally separable.

Significance: It proves that data from different sources can cooperatively reveal biological laws.

3. Model efficacy (model efficacy -Rplot)

Core indicators:

R2Y=0.767: The model can explain 76.7% of the sample classification differences.

Q2Y=0.672: 67.2% cross-validation prediction ability (>0.5 is valid).

p<0.05: The possibility of random guessing is excluded by substitution test.

Significance: The model is both reliable (high R2Y) and practical (high Q2Y).

4. False positive control (replacement test chart)

Method: Randomly scramble the labels 100 times and compare the real model (red line) with the random result (gray bar).

Result: The real R2Y/Q2Y was significantly higher than the random distribution (p=0.05).

No "virtual high" performance occurred.

Significance: Tat the differential genes and pathways discovered are not accidental.

In summary, our analytical process is rigorous (from data cleansing to model validation), and the biomarkers and typing results found have biological significance and potential clinical application.

The specific flow chart is as follows

A. Batch of HCC samples before correction. B. Batch of HCC samples after correction. C. Data integration (oPLS-DA score Chart). D. Model efficacy (model efficacy -Rplot).

the results are clear, , the current study has innovations and advantages. However,

- the quality of the figures is poor and should be rectified ( very very poor)

- please mention all abbreviations under the figures

Response: Thank you for your suggestions. All the figures have been refined and uploaded in the system. At the same time, all the abbreviations appearing in the figures have been fully expanded in their respective legends to ensure terminological clarity.

Response: Thank you for your suggestions. These samples are all human specimens, individual differences are inevitable, while overall differences are within the acceptable range. On the other hand, TCGA and GEO data have been verified and used in many studies, so the data is relatively reliable. In our study, data correction has been carried out to remove technical interference and ensure that the differences found later reflect the biological characteristics.

- Parameters for evaluation of each machine learning model discriminating pairs of groups should be clarified.

Response: Thank you for your suggestions. In the manuscript, the evaluation parameters for each machine learning model have been annotated (Such as line 125, line 157, etc.)

2nd set of questions:

Response: Thank you for your suggestions. We have added the above sentences “Pornpimol Charoentong, et al analyzed tumor-immune cell interactions in 20 solid cancers, revealing the relationship between genotype and immunophenotype and created The Cancer Immunome Atlas (https://tcia.at/). Using machine learning, the determinants of tumor immunogenicity were identified and a quantitative scoring scheme called immunophenotypic scoring was developed” to the section titled “Analysis of immune cell differences in the TME” (lines 140 to 145).

2. Lines 176-177 of revised manuscript: did you use the Draw Equation tool? Please put‘i’into subscript. Also, does the sigma have any initial and terminal values, i.e. i=0 .. n? Please start line 177 with “Where i represents the expression of mtDNA…”

Response: Thank you for your suggestions. We are very sorry for our carelessness. We have put‘i’ in the bottom right corner. i represents the expression of mtDNA phenotype-associated genes. The minimum value of ‘i’ is 0 and the maximum value of ‘i’ is 740.

3. In response to question #9, I do apologize, I stated my question incorrectly. I wanted to ask, how many genes do you increase the fold-change cutoff ? If the fold change is higher, you should get fewer genes.

Response: In our research, the P-value less than 0.001 was taken as the screening criteria, and the fold change of 1.5 was set as cutoff. If the fold change is higher, fewer genes will be screened out and some important genes with less changes will be filtered out.

Attachment

Submitted filename: Response_to_the_reviewers_auresp_3.docx

pone.0325033.s004.docx^{(512.8KB, docx)}

PLoS One. doi: 10.1371/journal.pone.0325033.r008

Decision Letter 3

Matthew Cserhati

7 May 2025

Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment features

PONE-D-24-22922R3

Dear Dr. Meng,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. If you have any questions relating to publication charges, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Matthew Cserhati, Ph.D

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Please do not forget to upload the final version of your manuscript, along with the supplementary data titled:

Revised Manuscript with TrackChanges-250502.docx

Supplementary materials.docx

Response to the reviewers250501.docx

Reviewers' comments:

PLoS One. doi: 10.1371/journal.pone.0325033.r009

Acceptance letter

Matthew Cserhati

PONE-D-24-22922R3

PLOS ONE

Dear Dr. Meng,

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now being handed over to our production team.

At this stage, our production department will prepare your paper for publication. This includes ensuring the following:

* All references, tables, and figures are properly cited

* All relevant supporting information is included in the manuscript submission,

* There are no issues that prevent the paper from being properly typeset

You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps.

Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

If we can help with anything else, please email us at customercare@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Matthew Cserhati

Academic Editor

PLOS ONE

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 File. The performance of oPLS-DA model was assessed by R²Y (goodness of fit) and Q² (goodness of prediction) through cross-validation, along with permutation testing to evaluate robustness.

(DOCX)

pone.0325033.s001.docx^{(612.6KB, docx)}

Attachment

Submitted filename: Response to the reviewers.docx

pone.0325033.s002.docx^{(475.7KB, docx)}

Attachment

Submitted filename: comments to the authers PONE-D-24-22922.docx

pone.0325033.s003.docx^{(14.8KB, docx)}

Attachment

Submitted filename: Response_to_the_reviewers_auresp_3.docx

pone.0325033.s004.docx^{(512.8KB, docx)}

Data Availability Statement

The HCC RNA transcriptome sequencing data and clinical information were obtained from The Cancer Genome Atlas (TCGA) (http://cancergenome.nih.gov/) and Gene Expression Omnibus (GEO)(https://www.ncbi.nlm.nih.gov/geo/). TCGA and GEO databases are public databases where relevant data can be downloaded directly.We have already upload figures to figshare, DOI: 10.6084/m9.figshare.27089542.

[pone.0325033.ref001] 1.Zabransky DJ, Danilova L, Leatherman JM, Lopez-Vidal TY, Sanchez J, Charmsaz S, et al. Profiling of syngeneic mouse HCC tumor models as a framework to understand anti-PD-1 sensitive tumor microenvironments. Hepatology. 2023;77(5):1566–79. doi: 10.1002/hep.32707 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref002] 2.He Q, Yang J, Jin Y. Immune infiltration and clinical significance analyses of the coagulation-related genes in hepatocellular carcinoma. Brief Bioinform. 2022;23(4):bbac291. doi: 10.1093/bib/bbac291 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref003] 3.Ruff SM, Shannon AH, Pawlik TM. Advances in Targeted Immunotherapy for Hepatobiliary Cancers. Int J Mol Sci. 2022;23(22):13961. doi: 10.3390/ijms232213961 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref004] 4.Nevola R, Ruocco R, Criscuolo L, Villani A, Alfano M, Beccia D, et al. Predictors of early and late hepatocellular carcinoma recurrence. World J Gastroenterol. 2023;29(8):1243–60. doi: 10.3748/wjg.v29.i8.1243 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref005] 5.Dong J, Wong L-J, Mims MP. Mitochondrial inheritance and cancer. Transl Res. 2018;202:24–34. doi: 10.1016/j.trsl.2018.06.004 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref006] 6.Zong W-X, Rabinowitz JD, White E. Mitochondria and Cancer. Mol Cell. 2016;61(5):667–76. doi: 10.1016/j.molcel.2016.02.011 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref007] 7.Wheelhouse NM, Lai PBS, Wigmore SJ, Ross JA, Harrison DJ. Mitochondrial D-loop mutations and deletion profiles of cancerous and noncancerous liver tissue in hepatitis B virus-infected liver. Br J Cancer. 2005;92(7):1268–72. doi: 10.1038/sj.bjc.6602496 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref008] 8.Picca A, Guerra F, Calvani R, Coelho-Júnior HJ, Leeuwenburgh C, Bucci C, et al. The contribution of mitochondrial DNA alterations to aging, cancer, and neurodegeneration. Exp Gerontol. 2023;178:112203. doi: 10.1016/j.exger.2023.112203 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref009] 9.El-Hattab AW, Craigen WJ, Scaglia F. Mitochondrial DNA maintenance defects. Biochim Biophys Acta Mol Basis Dis. 2017;1863(6):1539–55. doi: 10.1016/j.bbadis.2017.02.017 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref010] 10.Alimardani M, Moghbeli M, Rastgar-Moghadam A, Shandiz FH, Abbaszadegan MR. Single Nucleotide Polymorphisms as the Efficient Prognostic Markers in Breast Cancer. Curr Cancer Drug Targets. 2021;21(9):768–93. doi: 10.2174/1568009621666210525151846 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref011] 11.Lee H-Y, Nga HT, Tian J, Yi H-S. Mitochondrial Metabolic Signatures in Hepatocellular Carcinoma. Cells. 2021;10(8):1901. doi: 10.3390/cells10081901 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref012] 12.Rodrigues T, Ferraz LS. Therapeutic potential of targeting mitochondrial dynamics in cancer. Biochem Pharmacol. 2020;182:114282. doi: 10.1016/j.bcp.2020.114282 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref013] 13.Quail DF, Joyce JA. The Microenvironmental Landscape of Brain Tumors. Cancer Cell. 2017;31(3):326–41. doi: 10.1016/j.ccell.2017.02.009 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref014] 14.Ma S, Zhu J, Wang M, Zhu J, Wang W, Xiong Y, et al. Comprehensive analysis of m7G modification patterns based on potential m7G regulators and tumor microenvironment infiltration characterization in lung adenocarcinoma. Front Genet. 2022;13:996950. doi: 10.3389/fgene.2022.996950 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref015] 15.Ali HR, Chlon L, Pharoah PDP, Markowetz F, Caldas C. Patterns of Immune Infiltration in Breast Cancer and Their Clinical Implications: A Gene-Expression-Based Retrospective Study. PLoS Med. 2016;13(12):e1002194. doi: 10.1371/journal.pmed.1002194 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref016] 16.Lin Y-H, Lim S-N, Chen C-Y, Chi H-C, Yeh C-T, Lin W-R, et al. Functional Role of Mitochondrial DNA in Cancer Progression. Int J Mol Sci. 2022;23(3):1659. doi: 10.3390/ijms23031659 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref017] 17.Singh RK, Srivastava A, Kalaiarasan P, Manvati S, Chopra R, Bamezai RNK. mtDNA germ line variation mediated ROS generates retrograde signaling and induces pro-cancerous metabolic features. Sci Rep. 2014;4:6571. doi: 10.1038/srep06571 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref018] 18.Singh RK, Saini S, Verma D, Kalaiarasan P, Bamezai RNK. Mitochondrial ND5 mutation mediated elevated ROS regulates apoptotic pathway epigenetically in a P53 dependent manner for generating pro-cancerous phenotypes. Mitochondrion. 2017;35:35–43. doi: 10.1016/j.mito.2017.05.001 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref019] 19.Cao L, Cheng H, Jiang Q, Li H, Wu Z. APEX1 is a novel diagnostic and prognostic biomarker for hepatocellular carcinoma. Aging (Albany NY). 2020;12(5):4573–91. doi: 10.18632/aging.102913 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref020] 20.Zhang H, Jiang P-J, Lv M-Y, Zhao Y-H, Cui J, Chen J, et al. OGG1 contributes to hepatocellular carcinoma by promoting cell cycle-related protein expression and enhancing DNA oxidative damage repair in tumor cells. J Clin Lab Anal. 2022;36(7):e24561. doi: 10.1002/jcla.24561 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref021] 21.Piro-Mégy C, Sarzi E, Tarrés-Solé A, Péquignot M, Hensen F, Quilès M, et al. Dominant mutations in mtDNA maintenance gene SSBP1 cause optic atrophy and foveopathy. J Clin Invest. 2020;130(1):143–56. doi: 10.1172/JCI128513 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref022] 22.Yan C, Niu Y, Ma L, Tian L, Ma J. System analysis based on the cuproptosis-related genes identifies LIPT1 as a novel therapy target for liver hepatocellular carcinoma. J Transl Med. 2022;20(1):452. doi: 10.1186/s12967-022-03630-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref023] 23.Liu L, Liu B, Yu J, Zhang D, Shi J, Liang P. Development of a Toll-Like Receptor-Based Gene Signature That Can Predict Prognosis, Tumor Microenvironment, and Chemotherapy Response for Hepatocellular Carcinoma. Front Mol Biosci. 2021;8:729789. doi: 10.3389/fmolb.2021.729789 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref024] 24.Wilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics. 2010;26(12):1572–3. doi: 10.1093/bioinformatics/btq170 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref025] 25.Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50. doi: 10.1073/pnas.0506580102 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref026] 26.Charoentong P, Finotello F, Angelova M, Mayer C, Efremova M, Rieder D, et al. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell Rep. 2017;18(1):248–62. doi: 10.1016/j.celrep.2016.12.019 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref027] 27.Chong W, Shang L, Liu J, Fang Z, Du F, Wu H, et al. m6A regulator-based methylation modification patterns characterized by distinct tumor microenvironment immune profiles in colon cancer. Theranostics. 2021;11(5):2201–17. doi: 10.7150/thno.52717 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref028] 28.Sotiriou C, Wirapati P, Loi S, Harris A, Fox S, Smeds J, et al. Gene expression profiling in breast cancer: understanding the molecular basis of histologic grade to improve prognosis. J Natl Cancer Inst. 2006;98(4):262–72. doi: 10.1093/jnci/djj052 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref029] 29.Hegde PS, Chen DS. Top 10 Challenges in Cancer Immunotherapy. Immunity. 2020;52(1):17–35. doi: 10.1016/j.immuni.2019.12.011 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref030] 30.Hegde PS, Karanikas V, Evers S. The Where, the When, and the How of Immune Monitoring for Cancer Immunotherapies in the Era of Checkpoint Inhibition. Clin Cancer Res. 2016;22(8):1865–74. doi: 10.1158/1078-0432.CCR-15-1507 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref031] 31.Huang X, Qiu Z, Li L, Chen B, Huang P. m6A regulator-mediated methylation modification patterns and tumor microenvironment infiltration characterization in hepatocellular carcinoma. Aging (Albany NY). 2021;13(16):20698–715. doi: 10.18632/aging.203456 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref032] 32.Liu J, Peng Y, Wei W. Cell cycle on the crossroad of tumorigenesis and cancer therapy. Trends Cell Biol. 2022;32(1):30–44. doi: 10.1016/j.tcb.2021.07.001 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref033] 33.Li J, Xie H, Ying Y, Chen H, Yan H, He L, et al. YTHDF2 mediates the mRNA degradation of the tumor suppressors to induce AKT phosphorylation in N6-methyladenosine-dependent way in prostate cancer. Mol Cancer. 2020;19(1):152. doi: 10.1186/s12943-020-01267-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref034] 34.Mughal MJ, Mahadevappa R, Kwok HF. DNA replication licensing proteins: Saints and sinners in cancer. Semin Cancer Biol. 2019;58:11–21. doi: 10.1016/j.semcancer.2018.11.009 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref035] 35.Tang Q, Su Z, Gu W, Rustgi AK. Mutant p53 on the Path to Metastasis. Trends Cancer. 2020;6(1):62–73. doi: 10.1016/j.trecan.2019.11.004 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref036] 36.Kamp WM, Wang P-Y, Hwang PM. TP53 mutation, mitochondria and cancer. Curr Opin Genet Dev. 2016;38:16–22. doi: 10.1016/j.gde.2016.02.007 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref037] 37.Calderaro J, Ziol M, Paradis V, Zucman-Rossi J. Molecular and histological correlations in liver cancer. J Hepatol. 2019;71(3):616–30. doi: 10.1016/j.jhep.2019.06.001 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref038] 38.Shi M, Wang Y, Tang W, Cui X, Wu H, Tang Y, et al. Identification of TP53 mutation associated-immunotype and prediction of survival in patients with hepatocellular carcinoma. Ann Transl Med. 2020;8(6):321. doi: 10.21037/atm.2020.02.98 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref039] 39.Murai H, Kodama T, Maesaka K, Tange S, Motooka D, Suzuki Y, et al. Multiomics identifies the link between intratumor steatosis and the exhausted tumor immune microenvironment in hepatocellular carcinoma. Hepatology. 2023;77(1):77–91. doi: 10.1002/hep.32573 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref040] 40.Calderaro J, Couchy G, Imbeaud S, Amaddeo G, Letouzé E, Blanc J-F, et al. Histological subtypes of hepatocellular carcinoma are related to gene mutations and molecular tumour classification. J Hepatol. 2017;67(4):727–38. doi: 10.1016/j.jhep.2017.05.014 [DOI] [PubMed] [Google Scholar]

[pone.0325033.ref041] 41.Gabbia D, De Martin S. Tumor Mutational Burden for Predicting Prognosis and Therapy Outcome of Hepatocellular Carcinoma. Int J Mol Sci. 2023;24(4):3441. doi: 10.3390/ijms24043441 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref042] 42.Rao CV, Asch AS, Yamada HY. Frequently mutated genes/pathways and genomic instability as prevention targets in liver cancer. Carcinogenesis. 2017;38(1):2–11. doi: 10.1093/carcin/bgw118 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref043] 43.Luo Y, Ma J, Lu W. The Significance of Mitochondrial Dysfunction in Cancer. Int J Mol Sci. 2020;21(16):5598. doi: 10.3390/ijms21165598 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref044] 44.Yao J, Wang J, Xu Y, Guo Q, Sun Y, Liu J, et al. CDK9 inhibition blocks the initiation of PINK1-PRKN-mediated mitophagy by regulating the SIRT1-FOXO3-BNIP3 axis and enhances the therapeutic effects involving mitochondrial dysfunction in hepatocellular carcinoma. Autophagy. 2022;18(8):1879–97. doi: 10.1080/15548627.2021.2007027 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref045] 45.Hsu C-C, Lee H-C, Wei Y-H. Mitochondrial DNA alterations and mitochondrial dysfunction in the progression of hepatocellular carcinoma. World J Gastroenterol. 2013;19(47):8880–6. doi: 10.3748/wjg.v19.i47.8880 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref046] 46.Li J, Jin J, Liao M, Dang W, Chen X, Wu Y, et al. Upregulation of RECQL4 expression predicts poor prognosis in hepatocellular carcinoma. Oncol Lett. 2018;15(4):4248–54. doi: 10.3892/ol.2018.7860 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref047] 47.Sun Z, Chen G, Wang L, Sang Q, Xu G, Zhang N. APEX1 promotes the oncogenicity of hepatocellular carcinoma via regulation of MAP2K6. Aging (Albany NY). 2022;14(19):7959–71. doi: 10.18632/aging.204325 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref048] 48.Chiang JL, Shukla P, Pagidas K, Ahmed NS, Karri S, Gunn DD, et al. Mitochondria in Ovarian Aging and Reproductive Longevity. Ageing Res Rev. 2020;63:101168. doi: 10.1016/j.arr.2020.101168 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref049] 49.Filograna R, Mennuni M, Alsina D, Larsson N-G. Mitochondrial DNA copy number in human disease: the more the better?. FEBS Lett. 2021;595(8):976–1002. doi: 10.1002/1873-3468.14021 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref050] 50.Zhang C, Liu X, Jin S, Chen Y, Guo R. Ferroptosis in cancer therapy: a novel approach to reversing drug resistance. Mol Cancer. 2022;21(1):47. doi: 10.1186/s12943-022-01530-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref051] 51.Jin P, Jiang J, Zhou L, Huang Z, Nice EC, Huang C, et al. Mitochondrial adaptation in cancer drug resistance: prevalence, mechanisms, and management. J Hematol Oncol. 2022;15(1):97. doi: 10.1186/s13045-022-01313-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0325033.ref052] 52.Du F, Yang L-H, Liu J, Wang J, Fan L, Duangmano S, et al. The role of mitochondria in the resistance of melanoma to PD-1 inhibitors. J Transl Med. 2023;21(1):345. doi: 10.1186/s12967-023-04200-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Correlation analysis of mitochondrial DNA maintenance-related genes with HCC prognosis, tumor mutation burden and tumor microenvironment features

Fan Rong

Bin Cheng

Ling Guo

Shaobo Zeng

Xunliang Xu

Zhongji Meng

Roles

Abstract

Background

Methods

Results

Conclusions

Introduction

Materials and methods

Data acquisition and preprocessing

Unsupervised clustering of mtDNA MRGs

Gene set variation analysis (GSVA) and functional annotation

Analysis of immune cell differences in the TME

Identification of mtDNA MRG gene clusters

mtDNA score

Correlation of the mtDNA score with the TMB, clinicopathological features, and immunotherapy response status

Statistical analysis

Results

Genetic variation and transcriptional overview of mtDNA MRGs in HCC

Fig 1. Overview of mtDNA MRG mutations and transcription in HCC.

Identification of molecular subtypes of mtDNA MRGs

Fig 2. Consensus clustering of mtDNA MRGs.

Biological functions and TME immune cell infiltration features of different mtDNA MRG clusters

Fig 3. Biological characteristics and tumor microenvironments of samples enriched in different mtDNA MRG clusters.

Identification and biological functions of gene clusters

Fig 4. Identification and functional analysis of DEGs in the mtDNA MRG clusters.

Fig 5. Three mtDNA MRG gene clusters in HCC patients.

Construction of the mtDNA score and functional annotation

Fig 6. Construction and functional annotations of mtDNA score signatures.

mtDNA score and TMB

Fig 7. mtDNA score and tumor somatic mutations.

Correlation analysis of the mtDNA score with clinicopathological features and immunotherapy efficacy

Fig 8. Analysis of correlation between the mtDNA MRG score and prognosis in different clinical subgroups.

Fig 9. Correlations between the mtDNA score and immunotherapy response.

Discussion

Conclusion

Supporting information

Data Availability

Funding Statement

References

Author response to Decision Letter 0

Decision Letter 0

Matthew Cserhati

Roles

Author response to Decision Letter 1

Decision Letter 1

Matthew Cserhati

Roles

Author response to Decision Letter 2

Decision Letter 2

Matthew Cserhati

Roles

Author response to Decision Letter 3

Decision Letter 3

Matthew Cserhati

Roles

Acceptance letter

Matthew Cserhati

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases