Introduction

Hepatocellular carcinoma (HCC) is the most common adult primary liver cancer, and it is one of the highest cancers in the world (4.7%). HCC has a mortality rate of 8.3%, ranking fourth among cancer-related deaths worldwide. It is characterized by late detection and poor prognosis, posing a significant threat to human health [1].

According to statistical data on morbidity and mortality of liver cancer, there is a significant gender difference, in that males have the fifth highest incidence rate (6.3%) and the second highest mortality rate (10.5%), which is much higher than that of females [1]. Epidemiological studies have also shown that the incidence and mortality of male hepatocellular carcinoma are higher than females, and the differences between the two genders are significant [2-5]. Sex hormones are believed to be one of the important reasons why liver cancer incidence and mortality are different between two genders, and they also play different roles in the process of liver cancer development. It is known that androgen stimulates liver cancer production, while estrogen protects human beings from liver cancer onset [6-8]. However, the development of hepatocellular carcinoma is a complicated process, this research further explores how gender affects the pathways of hepatocellular carcinoma development.

Materials and Methods

Tissue Sample

There were 60 individuals whose tumor tissues were utilized for proteomic analysis, including 48 male cases and 12 female cases. It involved 27 patients whose tumor tissues were used for metabolomics analysis, including 22 males and 5 females. These cancer tissues were from Changhai Hospital, the affiliated hospital of Naval Medical University in Shanghai. This study has been ratified by the Clinical Research Ethics Committee of Changhai Hospital and the Third People's Hospital of Sichuan. The differences in clinical information between male and female tumor samples were compared by rank sum test using SPSS software (version 26.0). P-value < 0.05 was considered statistically significant (Supplementary Tables S1 and S2).

LC-ESI-MS/MS Analysis

The results of proteomics and metabolomics were detected by isobaric tags for relative and absolute quantitation (iTRAQ) and liquid chromatography-electrospray ionization mass spectrometry/mass spectrometry (LC-ESI-MS/MS), respectively. The results of iTRAQ and LC-ESI-MS/MS were obtained through BGI-Shenzhen, China(https://www.bgi.com/).

Determination of the Differentially Expressed Genes

The R package DESeq2(version 1.40.2) [9] was applied to screen the differentially expressed proteins (DEPs) between male and female tumor samples. The ratio of expression levels of proteins in males and females was represented as a fold change (FC) value. A statistically significant difference was defined as FDR < 0.05 and |log2FC| > 1.5. Subsequently, the related differentially expressed genes (DEGs) were then obtained by comparing the DEP's access in the UniProt data base(https://www.uniprot.org). Finally, the DEPs were visualized as a spot diagram by using R package ggplot2 (version 3.4.3) [10].

Functional and Pathway Enrichment Analyses

Biological process (BP), cellular component (CC), and molecular function (MF) are instances of Gene Ontology (GO) terms [11]. Apart from the gene function annotation, Kyoto Encyclopedia of Genes and Genomes (KEGG) contributes to clarifying the signaling transduction pathway associated with genes [12]. In order to figure out the key modules of hub genes, GO and KEGG pathway enrichment analyses were performed using the DAVID website(https://david.ncifcrf.gov/) [13]. P-value < 0.05 was considered statistically significant. The R package ggplot2 was used to depict the top 10 discoveries of GO and KEGG results.

Protein-protein Interaction Networks

The online STRING data base (version 12.0, https://cn.string-db.org/) was used to analyze the functional association of protein-protein interactions (PPIs) [14]. A combined score > 0.4 was established as the cutoff criterion, and then the PPI network was visualized by applying the Cytoscape software (Version 3.10.0) [15]. CytoHubba (version 0.1) plug-in in Cytoscape was applied to locate the hub genes by determining the intersection of the top 30 genes ranked by degree score [16]. The size of a node was determined by the degree value.

Survival Analysis of the Hub Genes

Based on the expression and clinical data from the Kaplan-Meier plotter data bases (http://www.kmplot.com), survival analysis was carried out to confirm the effect of hub genes on prognosis.  P < 0.05 was considered to be statistically significant [17-19].

Analysis of the Differentially Expressed Metabolites

The differentially expressed metabolites (DEMs) between male and female tumor tissues were obtained from the online MetaboAnalyst data base (version 6.0, https://www.metaboanalyst.ca) [20-22].

Principal Component Analysis (PCA) is an unsupervised analysis technique that is utilized to observe the overall distribution trend between male and female tumor tissues and find potential discrete points. Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA), on the other hand, is a supervised approach that is employed to screen the differentially expressed metabolites in tumor tissues between male and female. When all key parameters (OPLS-DA, R2Y and Q2) are greater than 0.5, it indicates that the model is stable and dependable. The permutation test with 100 iterations was employed to assess the statistical significance. P<0.05 demonstrates the model is not overfitting.

In order to identify potential key metabolites, the Variable Importance in Projection (VIP) was used to evaluate the importance of the DEMs [23, 24].

In this study, a statistically significant difference was defined as FDR < 0.05, |log2FC| > 1.5 and VIP > 1. Finally, the classification of the DEMs was obtained by using the PubChem data base (https://pubchem.ncbi. nlm.nih.gov/) [25].

Results

Identification of the DEGs

We analyzed the DEPs in male and female tumor samples following the cutoff criteria | log2FC | > 1.5 and P-value < 0.05.  We identified a total of 307 DEPs with 154 upregulated and 153 downregulated proteins (Figure 1). Subsequently, we obtained 233 DEGs from the UniProt data base based on the accession of DEPs, of which 102 were upregulated and 131 were downregulated. A full list of the DEGs was reported in Supplementary Table S3.

Figure 1: Identification of differentially expressed proteins (DEPs). Spot diagram was utilized to illustrate the DEPs between male and female tumor samples. Each triangle denoted one protein detected in both tissues. Green triangle represented downregulated proteins with log2FC < -1.5 and P-value < 0.05. Red triangle represented the upregulated proteins with log2FC > 1.5 and P-value < 0.05.

Functional Enrichment Analyses

To further investigate the biological function and mechanism of the DEGs, GO and KEGG were carried out using the DAVID data base and R package ggplot2. Changes in BP of the upregulated DEGs were significantly enriched in cell adhesion, response to xenobiotic stimulus, cytoplasmic translation, cell-matrix adhesion, positive regulation of angiogenesis, cell migration, extracellular matrix organization, cell-cell adhesion, translation and positive regulation of transcription from RNA polymerase II promoter (Figure 2A). Furthermore, CC analysis showed that the upregulated DEGs were mainly associated with extracellular exosome, basement membrane, cytosolic large ribosomal subunit, extracellular space, extracellular region, cytosol, cytoplasm, ribonucleoprotein complex, endoplasmic reticulum lumen and perinuclear region of cytoplasm (Figure 2B). Additionally, MF terms prompted that the upregulated DEGs were dramatically related to RNA binding, extracellular matrix structural constituent, integrin binding, cadherin binding, collagen binding, macromolecular complex binding, protein binding, enzyme binding, structural constituent of ribosome and protein homodimerization activity (Figure 2C). The three primary pathways  significantly enhanced by the upregulated DEGs were focal adhesion, actin cytoskeleton regulation and ECM-receptor interaction (Figure 2D).

Figure 2: Enrichment analysis of the upregulated DEGs revealed top 10 genesets in BPs (A), CCs (B), MFs (C) and KEGG pathways (D) by the DAVID data base using the ggplot2 package in R language for visualization. The cutoff is established at P-value<0.05.

Xenobiotic metabolic process, cellular glucuronidation, glutathione metabolic process, cellular respiration, lipid metabolic process, mitochondrial electron transport, cytochrome c to oxygen, response to ethanol, fatty acid beta-oxidation, fatty acid biosynthetic process and electron transport chain were substantially enriched in BP analysis of the downregulated DEGs (Figure 3A). Moreover, CC analysis revealed that mitochondrion, mitochondrial matrix, extracellular exosome, cytosol, mitochondrial inner membrane, peroxisome, azurophil granule lumen, endoplasmic reticulum membrane, mitochondrial outer membrane and cytoplasm were sensibly involved in the downregulated DEGs (Figure 3B). Besides, MF terms indicated a substantial association within the downregulated DEGs. These functions were involved in  oxidoreductase activity, NADP binding, catalytic activity, electron carrier activity, glucuronosyltransferase activity, UDP-glycosyltransferase activity, heme binding, protein homodimerization activity, identical protein binding and GTP binding. (Figure 3C). The three critical pathways significantly related to the downregulated DEGs were metabolic pathways, metabolism of xenobiotics by cytochrome P450 and chemical carcinogenesis - reactive oxygen species (Figure 3D).

Figure 3: Enrichment analysis of the downregulated DEGs showed top 10 genesets in BPs (A), CCs (B), MFs (C) and KEGG pathways (D) according to the DAVID data base using the ggplot2 package in R language for visualization. The threshold is set at P-value<0.05.

Construction of the PPI Network

The PPI network of the upregulated genes of DEGs was shown in Figure 4A, and the top 30 upregulated hub genes were shown in Figure 4B and Figure 4C. Moreover, the PPI network of the downregulated genes of DEGs was shown in Figure 4D, and the top 30 downregulated hub genes were shown in Figure 4E. The detailed descriptions of the hub genes were shown in Supplementary Table S3.

Figure 4: PPI network of the hub genes. PPI network of the upregulatedDEGs (A). The top 30 genes  selected from the upregulated PPI network (B, C). PPI network of the downregulatedDEGs (D). The top 30 genes from the downregulated PPI network (E).

Analysis of the survival and prognostic value of hub genes

 To identify the overall survival rates of the hub genes, we utilized the online Kaplan Meier plotter data base. We obtained 10 genes with survival and prognostic value in upregulated hub genes(Figure 5A), namely CPSF6, HDAC2, ITGA2, LAMB1, NAP1L1, RPL17, RPL23A, SNRPD1, XRCC5 and XRCC6. Furthermore, we revealed 17 genes with survival and prognostic value in  downregulated hub genes (Figure 5B), namely ADHFE1, ALDH7A1, BCKDK, COX5A, COX5B, CYP4A11, DBT, EHHADH, GRHPR, H6PD, HIBCH, NDUFAB1, SDHA, SHMT1, SOD1, TUFM and UQCRC2.

Figure 5: Prognostic value of hub genes in HCC. Prognostic value of the upregulated hub genes (A). Prognostic value of the downregulated hub genes (B).  P < 0.05 was considered to be statistically significant.

Metabolomics Investigation

We observed that metabolic pathways individually accounted for more than 50% based on the enrichment analysis of downregulated hub genes (Figure 3D). Therefore, we further explored the DEMs on male and female tumor tissues. The general distribution trend could be observed by the PCA analysis performed over all tissues. The PCA scores plot was presented in Figure 6. The first two principal components in the negative-ion mode (ESI-) accounted for 32.8% of  overall variance (Figure 6A). Meanwhile, in the positive-ion mode (ESI+), the first two primary components represented a total variation of 31.3% (Figure 6B). Nevertheless, the metabolic distribution trend of male and female tumor tissues was not well distinguished. Thus, considering that the PCA model is unsupervised, OPLS-DA was performed to further differentiate the DEMs between groups. A clear separation was shown in the OPLS-DA scores plot (ESI- (Figure 6C), ESI+ (Figure 6D)). The permutation tests of the OPLS-DA model based on the ESI- and ESI+ were presented in Figure 6E and Figure 6F, respectively. R2Y and Q2, the key parameters of the OPLS-DA model in the ESI- (R2Y=0.986 and Q2=0.729, Figure 6E) and ESI+ (R2Y=0.922 and Q2=0.616, Figure 6F), were greater than 0.5. Meanwhile, combined with P-value < 0.05, it indicated that the OPLS-DA model was not overfitting and the whole analysis system was stable and reliable.

Figure 6: The distribution trend of the DEMs between male and female tumor samples. PCA analysis in ESI- (A) and ESI+ (B). OPLS-DA analysis in ESI- (C) and ESI+ (D).OPLS-DA permutation test in ESI-(E) and ESI+ (F).

Identification of the DEMs

The VIP value of the OPLS-DA model was used to filter the important DEMs with the greatest contribution across groups. The DEMs were visualized based on the cut-off value of |log2FC| >1.5 and FDR < 0.05 in ESI- (Figure 7A) and ESI+ (Figure 7B). Moreover, ESI- (Figure 7C) and ESI+ (Figure 7D) displayed the top 15 DEMs with the highest VIP score. Then, the hub DEMs were selected combined with the criteria: |log2FC| >1.5, FDR < 0.05 and VIP > 1. We therefore respectively acquired 23 hub DEMs in ESI- and 10 hub DEMs in ESI+. Finally, we gathered relevant data on the hub DEMs from the PubChem data base. Subsequently, a full list of the key DEMs was reported in Supplementary Table S4. According to the classification of the hub DEMs, we found that the lipid metabolism, especially the metabolism of glycerophospholipids and fatty acids, in male tumor tissues is significantly downregulated compared to female. (Supplementary Table S4).

Figure 7: Identification of the DEMs. Volcano plot was utilized to illustrate the DEMs in ESI- (A) and ESI+ (B) between male and female tumor samples. The top 15 DEMs with the highest VIP score in ESI- (C) and ESI+ (D).

Discussion

Hepatocellular carcinoma (HCC) poses a persistent threat to worldwide human health, and its significant difference in incidence between the two genders warrants further investigation. Genomics [26-28] and epigenetics, including DNA methylation change [29, 30] have been thoroughly explored in the past few years. Sex-dependent DNA methylation events have been proven to exist in HCC [2]. Proteomics and metabolomics have become hotspots for studying the pathogenesis of hepatocellular carcinoma, especially the effect of gender. In recent years, mass spectrometry has emerged as an excellent method for the differential expression analysis of proteins and metabolites. The proteomic and metabolomic data presented herein, along with their comparative analysis, provide vital insights into elucidating the mechanisms that underlie the male higher prevalence of HCC [31, 32].

In this study, we found 233 DEGs between male and female tumor samples. In terms of GO functional enrichment analysis, most of them take part in cell adhesion, cytosol and protein binding. Pathway enrichment revealed that they are mainly involved in metabolic pathways and metabolism of xenobiotics by cytochrome P450, indicating that those DEGs may affect occurrence, metastasis, or invasion of HCC by regulating energy metabolism and metabolism of essential substances such as retinoic acid.

Furthermore, the top 10 DEGs with the highest degree in PPIs are H6PD, SDHA, NDUFAB1, G6PD, DBT, ACSS2, CYCS, HIBADH, EHHADH and ALDH7A1. Specifically, DBT, SDHA, EHHADH, ALDH7A1, H6PD and NDUFAB1 demonstrate prognostic value in HCC. Reduced NADPH is generated in the endoplasmic reticulum cavity during the H6PD-catalyzed conversion of glucose 6-phosphate to 6-phosphogluconic acid. Notably, NADPH, a cofactor for numerous reductases, needs to be synthesized within the endoplasmic reticulum cavity due to the inability of NADPH to penetrate biological membranes. Consequently, H6PD is pivotal in preserving homeostasis within the endoplasmic reticulum cavity [33, 34]. SDHA is a subunit of the SDHx located within the mitochondrial inner membrane. It is essential for the transmission of electrons from succinic acid to CoQ in the mitochondrial electron transfer chain. Remarkably, SDH mutations are prevalent in malignancies, with the most common mutation being SDHA [35, 36]. Research has revealed that, in comparison to normal liver cells, SDHx is significantly reduced in HCC. This decrease constantly affects the TCA cycle, which results in an accumulation of succinates within the body [37-39]. DBT, a subunit of the ranched-chain alpha-keto acid dehydrogenase complex (BCKD), plays a pivotal role in the catabolism of branched-chain amino acids such as leucine and valine within mitochondria [40]. Studies have substantiated that the expression of DBT is significantly reduced in a multitude of malignant tumors. Furthermore, it has been observed that the level of DBT expression is intimately associated with both the progression and prognosis of these tumors [41-43]. EHHADH-encoded protein is a bifunctional enzyme with the ability to take part in lipid metabolism [44, 45]. Studies have demonstrated that the expression of EHHADH is markedly downregulated in HCC. Furthermore, there is a pronounced negative correlation between EHHADH and progression as well as the prognosis of the tumor, particularly within HCC [46-48]. ALDH7A1, a member of the ALDH superfamily, has been implicated in tumor progression, prognosis and DNA methylation[ 49-51], etc. However, its expression varies across different tumors. Notably, ALDH7A1 is upregulated in colon cancer and PDAC [49, 52], while it is downregulated in LSCC, HCC and oral cancer [51, 53, 54]. Further researches are required to elucidate the tumor-related effects of ALDH7A1. NDUFAB1 as an acyl carrier protein is crucial for the energy transfer within the mitochondrial respiratory chain and involved in lipid metabolism [55]. Furthermore, it plays a significant role in tumor progression, such as proliferation, migration, and immune infiltration [56, 57].

Notably, the top 10 DEGs wereall implicated in metabolic pathways, implying that metabolism could be the pivotal factor contributing to the gender disparity in the incidence of HCC. Consequently, we investigated the expression of metabolism. Interestingly, our findings indicated a significant downregulation of lipid metabolism, with a particular emphasis on glycerophospholipid and fatty acid metabolism, in male tumor tissues relative to female counterparts, suggesting that investigating the glycerophospholipid and fatty acid metabolism could yield valuable insights into the gender disparity observed in HCC. Glycerophospholipids, ubiquitously present in the body, are pivotal in determining the structure and functionality of cell membranes. This study revealed a significant downregulation in the expression of glycerophosphoserines (PS) and glycerophosphoglycerols (PG) within glycerophospholipids. This could potentially be associated with decreased hepatocyte function or an alteration in lipid homeostasis [58, 59]. Fatty acids, the fundamental constituents of complex lipids, significantly influence cancer progression through abnormal metabolism. The reduction in fatty acid has been reported in several previous investigations [59, 60].

In conclusion, our proteomics and metabolomics analyses of male and female HCC patients revealed 27 hub genes with differential expression that possess prognostic and diagnostic value. Notably, the top 10 of these genes were predominantly associated with metabolic processes. Furthermore, we observed that lipid metabolism was significantly downregulated in male patients compared to females. These findings hold potential implications for  screening and prognosis of male tumor patients.

References

1. Sung H, Ferlay J, Siegel RL (2021) Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 71:209-249.
2. Ye W, Siwko S, Tsai R (2021) Sex and Race-Related DNA Methylation Changes in Hepatocellular Carcinoma. Int J Mol Sci 22:3820.
3. Zhang T, Xu J, Ye L (2021) Age, Gender and Geographic Differences in Global Health Burden of Cirrhosis and Liver Cancer due to Nonalcoholic Steatohepatitis. J Cancer 12:2855-2865.
4. Ajayi F, Jan J, Singal AG, Rich NE (2020) Racial and Sex Disparities in Hepatocellular Carcinoma in the USA. Curr Hepatol Rep 19:462-469.
5. Kim SY, Song HK, Lee SK (2020) Sex-Biased Molecular Signature for Overall Survival of Liver Cancer Patients. Biomol Ther (Seoul) 28:491-502.
6. Zheng B, Zhu YJ, Wang HY, Chen L (2017) Gender disparity in hepatocellular carcinoma (HCC): multiple underlying mechanisms. Sci China Life Sci 60:575-584.
7. Li Y, Xu A, Jia S, Huang J (2019) Recent advances in the molecular mechanism of sex disparity in hepatocellular carcinoma. Oncol Lett 17:4222-4228.
8. Wu MH, Ma WL, Hsu CL (2010) Androgen receptor promotes hepatitis B virus-induced hepatocarcinogenesis through modulation of hepatitis B virus RNA transcription. Sci Transl Med 2: 32ra35.
9. Love MI, Huber W, Anders S (2014) Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 15:550.
10. Wickham H (2011) ggplot2. WIREs Computational Statistics 3:180-185.
11. Harris MA, Clark J, Ireland A (2004) The Gene Ontology (GO) data base and informatics resource. Nucleic Acids Res 32:D258-261.
12. Ogata H, Goto S, Fujibuchi W, Kanehisa M (1998) Computation with the KEGG pathway data base. Biosystems 47:119-128.
13. Dennis G, Sherman BT, Hosack DA (2003) DAVID: Data base for Annotation, Visualization, and Integrated Discovery. Genome Biology 4:P3.
14. Szklarczyk D, Kirsch R, Koutrouli M (2023) The STRING data base in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res 51:D638-D646.
15. Doncheva NT, Morris JH, Gorodkin J, Jensen LJ (2019) Cytoscape StringApp: Network Analysis and Visualization of Proteomics Data. J Proteome Res 18:623-632.
16. Chin CH, Chen SH, Wu HH, Ho CW, Ko MT, et al. (2014) cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC Syst Biol 8:S11.
17. Mu X, Huang O, Jiang M, Xie Z, Chen D, et al. (2019) Prognostic value of ephrin B receptors in breast cancer: An online survival analysis using the microarray data of 3,554 patients. Oncol Lett 18:742-750.
18. Zheng S, Yang L, Dai Y (2019) Screening and Survival Analysis of Hub Genes in Gastric Cancer Based on Bioinformatics. J Comput Biol 26:1316-1325.
19. Wang L, Xu D, Cai L, Dai J, Li Y, et al. (2021) Expression and survival analysis of the STAT gene family in diffuse gliomas using integrated bioinformatics. Curr Res Transl Med 69:103274.
20. Pang Z, Chong J, Zhou G (2021) MetaboAnalyst 5.0: narrowing the gap between raw spectra and functional insights. Nucleic Acids Res 49:W388-W396.
21. Oliveira MS, Santo RCE, Silva JMS,  Alabarse PVG, Brenol CV, et al. (2023) Urinary metabolomic biomarker candidates for skeletal muscle wasting in patients with rheumatoid arthritis. J Cachexia Sarcopenia Muscle 14:1657-1669.
22. Xu Z, Sheng Y, Zeng G, Zeng Z, Li B, et al. (2021) Metabonomic Study on the Plasma of High-Fat Diet-Induced Dyslipidemia Rats Treated with Ge Gen Qin Lian Decoction by Ultrahigh-Performance Liquid Chromatography-Mass Spectrometry. Evid Based Complement Alternat Med 6692456.
23. Liang L, Liu G, Yu G, Zhang F, Linhardt RJ, et al. (2020) Urinary metabolomics analysis reveals the anti-diabetic effect of stachyose in high-fat diet/streptozotocin-induced type 2 diabetic rats. Carbohydr Polym 229:115534.
24. Zhang C, Zhou S, Chang H, Zhuang F, Shi Y, et al. (2021) Metabolomic Profiling Identified Serum Metabolite Biomarkers and Related Metabolic Pathways of Colorectal Cancer. Dis Markers 6858809.
25. Kim S (2021) Exploring Chemical Information in PubChem. Curr Protoc 1:e217.
26. Nia A, Dhanasekaran R (2020) Genomic Landscape of HCC. Curr Hepatol Rep 19:448-461.
27. Meier T, Timm M, Montani M, Wilkens L (2021) Gene networks and transcriptional regulators associated with liver cancer development and progression.BMC  Med Genomics 14:41.
28. Yim SY,  Lee  JS (2019) Genomic Perspective on Mouse Liver Cancer Models.Cancers  11:1648.
29. Cerapio JP, Marchio A, Cano L, López I, Fournié JJ, et al. (2021) Global DNA hypermethylation pattern and unique gene expression signature in liver cancer from patients with Indigenous American ancestry. Oncotarget 12:475-492.
30. Toh TB, Lim JJ, Chow EKH (2019) Epigenetics of hepatocellular carcinoma. Clin Transl Med 8:13.
31. Li H, Rong Z, Wang H, Zhang N, Pu C, et al. (2020) Proteomic analysis revealed common, unique and systemic signatures in gender-dependent hepatocarcinogenesis. Biol Sex Differ 11:46.     
32. Rong Z, Fan T, Li H, Li J, Wang K, et al. (2017) Differential Proteomic Analysis of Gender-dependent Hepatic Tumorigenesis in Hras12V Transgenic Mice. Mol Cell Proteomics 16:1475-1490.
33. Tsachaki M, Mladenovic N, Štambergová H, Birk J, Odermatt A (2018) Hexose-6-phosphate dehydrogenase controls cancer cell proliferation and migration through pleiotropic effects on the unfolded-protein response, calcium homeostasis, and redox balance. FASEB J 32:2690-2705.
34. Draper N, Walker EA, Bujalska IJ, Tomlinson JW, Chalder SM, et al. (2003) Mutations in the genes encoding 11beta-hydroxysteroid dehydrogenase type 1 and hexose-6-phosphate dehydrogenase interact to cause cortisone reductase deficiency.  Nat Genet 34:434-9.
35. Nan H, Guo P, Fan J, Zeng W, Hu C, et al. (2023) Comprehensive analysis of the prognosis, tumor microenvironment, and immunotherapy response of SDHs in colon adenocarcinoma. Front Immunol 14:1093974.
36. Ibrahim A, Chopra S (2020) Succinate Dehydrogenase-Deficient Gastrointestinal Stromal Tumors. Arch Pathol Lab Med 144:655-660.
37. Kanungo S, Wells K, Tribett T, El-Gharbawy A  (2018) Glycogen metabolism and glycogen storage disorders. Ann Transl Med 6:474.
38. Li J, Liang N, Long X, Zhao J, Yang J, et al. (2019) SDHC-related deficiency of SDH complex activity promotes growth and metastasis of hepatocellular carcinoma via ROS/NF?B signaling. Cancer Lett 461:44-55.
39. Bai W, Cheng L, Xiong L, Wang M, Liu H, et al. (2022) Protein succinylation associated with the progress of hepatocellular carcinoma.  J  Cell Mol Med 26:5702-5712.
40. Friedrich T, Lambert AM, Masino MA, Downes GB (2012) Mutation of zebrafish dihydrolipoamide branched-chain transacylase E2 results in motor dysfunction and models maple syrup urine disease. Dis Model Mech 5:248-58.
41. Zhang C, Huang G, Yang J, Jiang Y, Huang R, et al. (2023) Overexpression of DBT suppresses the aggressiveness of renal clear cell carcinoma and correlates with immune infiltration. Front Immunol 14:1197011.
42. Miao D, Wang Q, Shi J, Lv Q, Tan D, et al. (2023) N6-methyladenosine-modified DBT alleviates lipid accumulation and inhibits tumor progression in clear cell renal cell carcinoma through the ANXA2/YAP axis-regulated Hippo pathway. Cancer Commun (Lond) 43:480-502.
43. Lai SW, Weng PW, Yadav VK, Pikatan NW, Yeh CT, et al. (2024) Underlying mechanisms of novel cuproptosis-related dihydrolipoamide branched-chain transacylase E2 (DBT) signature in sunitinib-resistant clear-cell renal cell carcinoma. Aging (Albany NY) 16:2679-2701.
44. Ranea-Robles P, Violante S, Argmann C, Dodatko T, Bhattacharya D, et al. (2021) Murine deficiency of peroxisomal L-bifunctional protein (EHHADH) causes medium-chain 3-hydroxydicarboxylic aciduria and perturbs hepatic cholesterol homeostasis. Cell Mol Life Sci 78:5631-5646.
45. Li N, Li N, Wen S, Li B, Zhang Y, et al. (2021) HSP60 Regulates Lipid Metabolism in Human Ovarian Cancer. Oxid Med Cell Longev 6610529.
46. Ren H, Li W, Liu X, Li S, Guo H, et al. (2021) Identification and Validation of an 6-Metabolism-Related Gene Signature and Its Correlation With Immune Checkpoint in Hepatocellular Carcinoma. Front Oncol 11:783934.
47. Chen P, Wang F, Feng J, Zhou R, Chang Y, et al. (2017) Co-expression network analysis identified six hub genes in association with metastasis risk and prognosis in hepatocellular carcinoma. Oncotarget 8:48948-48958.
48. Pilz JF, Klein M, Neumann-Haefelin E, Ganner A (2024) VHL-dependence of EHHADH Expression in a Human Renal Cell Carcinoma Cell Line. J Kidney Cancer VHL 11:12-18.
49. Lee JS, Lee H, Woo SM, Jang H, Jeon Y, et al. (2021) Overall survival of pancreatic ductal adenocarcinoma is doubled by Aldh7a1 deletion in the KPC mouse. Theranostics 11:3472-3488.
50. Stoll G, Kremer M, Bloy N, Joseph A, Castedo M, et al. (2019) Metabolic enzymes expressed by cancer cells impact the immune infiltrate. Oncoimmunology 8:e1571389.
51. Zhang M, Sun L, Ru Y, Zhang S, Miao J, et al. (2020) A risk score system based on DNA methylation levels and a nomogram survival model for lung squamous cell carcinoma. Int J Mol Med 46:252-264.
52. Liu G, Wu X, Chen J (2022) Identification and validation of a glycolysis-related gene signature for depicting clinical characteristics and its relationship with tumor immunity in patients with colon cancer. Aging (Albany NY) 14:8700-8718.
53. Andrejeva D, Kugler JM, Nguyen HT, Malmendal A, Holm ML, et al. (2018) Metabolic control of PPAR activity by aldehyde dehydrogenase regulates invasive cell behavior and predicts survival in hepatocellular and renal clear cell carcinoma. BMC Cancer 18:1180.
54. Lu HJ, Chuang CY, Chen MK, Su CW, Yang WE, et al. (2022) The impact of ALDH7A1 variants in oral cancer development and prognosis. Aging (Albany NY) 14:4556-4571.
55. Hou T, Zhang R, Jian C, Ding W, Wang Y, et al. (2019) NDUFAB1 confers cardio-protection by enhancing mitochondrial bioenergetics through coordination of respiratory complex and supercomplex assembly. Cell Res 29:754-766.
56. Chen Y, Wu W, Jin C, Cui J, Diao Y, et al. (2023) Integrating Single-Cell RNA-Seq and Bulk RNA-Seq Data to Explore the Key Role of Fatty Acid Metabolism in Breast Cancer. Int J Mol Sci 24:13209.
57. Chang X, Xing P (2022) Identification of a novel lipid metabolism-related gene signature within the tumour immune microenvironment for breast cancer. Lipids Health Dis 21:43.
58. Passos-Castilho AM, Carvalho VM, Cardozo KHM, Kikuchi L, Chagas AL, et al. (2015) Serum lipidomic profiling as a useful tool for screening potential biomarkers of hepatitis B-related hepatocellular carcinoma by ultraperformance liquid chromatography-mass spectrometry. BMC Cancer 15:985.
59. Rashid MM, Varghese RS, Ding Y, Ressom HW (2023) Biomarker Discovery for Hepatocellular Carcinoma in Patients with Liver Cirrhosis Using Untargeted Metabolomics and Lipidomics Studies. Metabolites 13:1047.
60. Qi SA, Wu Q, Chen Z, Zhang W, Zhou Y, et al. (2021) High-resolution metabolomic biomarkers for lung cancer diagnosis and prognosis. Sci Rep 11:11805.