A novel cuproptosis-related prognostic gene signature in adrenocortical carcinoma
Wenjun Gao1 1 Xiaoyan He2 2 İD Qi Huangfu1 Yanqi Xie1 1 Keliang Chen 3
Chengfang Sun1 1 Jingchao Wei1 Bohan Wang 1
1Department of Urology, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, Zhejiang, China
2Department of Health Education, HangZhou Center for Disease Control and Prevention, Hangzhou, China
3Department of Urology, 4th Affiliated Hospital, Zhejiang University School of Medicine, Yiwu, Zhejiang, China
Correspondence
Jingchao Wei and Bohan Wang, Department of Urology, The Second Affiliated Hospital, School of Medicine, Zhejiang University, Hangzhou, Zhejiang, China.
Email: weijingchao@zju.edu.cn and wangbohan@zju.edu.cn
Funding information
National Natural Science Foundation of China, Grant/Award Number: 81970601 and 82200850
Abstract
Background: Adrenocortical carcinoma (ACC) is an aggressive and rare malignant tumor associated with poor outcomes. Cuproptosis, a new pattern of cell death, relies on mitochondrial respiration and is associated with protein lipoylation. Increasing evi- dence has demonstrated the potential roles of cuproptosis in several tumor entities. However, the relationship between cuproptosis and ACC remains unclear.
Methods: In total, 10 cuproptosis-related genes (CRGs) of patients with ACC were obtained from the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases and differential expression analysis of CRGs was analyzed. Functional enrichment of the CRGs was performed and protein-protein interaction analysis was utilized to explore the association between the CRGs. Cuproptosis- related risk score (CRRS) was constructed by Lasso Cox regression and validated.
Results: In the current study, the alteration and expression patterns of 10 CRGs in TCGA-ACC datasets were analyzed. We identified different expression patterns of CRGs in ACCs, discovered strong associations between CRGs and ACCs, and found that the CRGs were associated with immune infiltration in ACCs. A CRRS was created thereafter to predict overall survival (OS). CRRS=(0.083103718) *FDX1+(-0.278423862) *LIAS+(0.090985682) *DLAT+(-0.018784047)
*PDHA1+(0.297218951) *MTF1+(0.310197964) *CDKN2A. Patients were divided into high- and low-risk groups based on their CRRS, and independent prognostic fac- tors were investigated. Finally, CDKN2A and FDX1 were found to be independent prognostic predictors of patients with ACC.
Conclusions: CDKN2A and FDX1 are independent prognostic predictors of patients with ACC. Cuproptosis may play a role in the development of ACC, providing a new per- spective on therapeutic strategies related to CRGs for cancer prevention and treatment.
KEYWORDS
adrenocortical carcinoma (ACC), CDKN2A, Cuproptosis, FDX1, risk score
Wenjun Gao and Xiaoyan He contributed equally to this work and share the first authorship.
Jingchao Wei and Bohan Wang contributed equally to this work and share the last authorship.
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
@ 2023 The Authors. Journal of Clinical Laboratory Analysis published by Wiley Periodicals LLC.
Adrenocortical carcinoma (ACC) is an aggressive and rare malignant tumor with an incidence of 0.7-2.0 cases/million/year.1,2 ACC de- rives from the adrenal cortex and currently lacks effective systemic treatment options. The median age of diagnosis is in the fifth to the sixth decade with a bimodal distribution, predominantly affecting adults in their 40-60s and children of 1-5 years.3 About one-third of patients with ACC may experience resurgence even after mar- gin-negative surgical resection.4 Despite the development of sur- gery and mitotane-based and platinum-based chemotherapy within recent years, the prognosis for patients with ACC remains poor.3 The median overall survival (OS) is 3-4 years, and the 5-year survival rate for patients with distant metastases is 0%-28%.5,6 Validated prog- nostic factors include tumor stage, resection status, Ki-67 index (or mitotic count), autonomous cortisol secretion, and the patient’s gen- eral condition,3 all of which possess limited performance. Therefore, given the substantial risk of recurrence and poor prognosis of ACC, there is an imperative need to develop more effective and efficient prognostic models.
Copper is an indispensable mineral nutrient that takes part in almost every organism and participates in various biological processes, including mitochondrial respiration, antioxidant/de- toxification process, and iron uptake.7 Copper homeostasis reg- ulation is crucial for cellular metabolism and survival.8 Aberrant copper accumulation could result in cell malignant transfor- mation.9 Recently, a novel pattern of copper-induced cell death termed “cuproptosis” was reported, distinct from apoptosis, necroptosis, pyroptosis, and ferroptosis.10 As a result of copper exceeding the homeostasis threshold, cuproptosis relies on mi- tochondrial respiration.10 Cuproptosis is associated with protein lipoylation as copper can bind directly to the lipid components of the tricarboxylic acid (TCA) cycle for enzymatic function, lead- ing to a loss of iron-sulfur (Fe/S) cluster containing proteins and induction of heat shock protein 70 (HSP70), ultimately result- ing in proteotoxic stress and cell death.10,11 Tsvetkov et al. per- formed a whole-genome CRISPR-Cas9 loss-of-function screen in order to validate the genes involved in the progress and identi- fied 10 cuproptosis-related genes (CRGs) and found seven CRGs involved in the positive regulation of cuproptosis, including di- hydrolipoamide S-acetyltransferase (DLAT), dihydrolipoamide dehydrogenase (DLD), ferredoxin 1 (FDX1), lipoyl synthase (LIAS), lipolytransferase 1 (LIPT1), pyruvate dehydrogenase E1 subunit alpha 1 (PDHA1), and pyruvate dehydrogenase E1 subunit beta (PDHB).10 Meanwhile, three CRGs were identified to negatively regulate cuproptosis, namely, cyclin-dependent kinase inhibitor 2A (CDKN2A), glutaminase (GLS), and metal regulatory transcrip- tion factor 1 (MTF1).10 In addition, a study conducted by Bilcikova et al. explored the effects of copper sulfate (CuSO4.5H2O) on cy- totoxicity using a human ACC (NCI-H295R) cell line and found that cell viability was significantly decreased in all experimental groups (3.90-1000 mM of CuSO4.5H2O) compared with the control group (medium without CuSO4.5H2O).12 Therefore, the 10 CRGs may
facilitate the mechanism exploration of copper toxicity, and cu- proptosis may potentially be a part of an effective way to treat ACC.
In this study, we sought to elucidate the potential effects of the 10 CRGs in ACC. The association between the 10 CRGs and the clin- icopathological features of patients with ACC was investigated by analysis of The Cancer Genome Atlas (TCGA) database. An innova- tive risk score was constructed relying on six CRGs, and the ability of the CRGs to predict the prognosis of patients with ACC was as- sessed. Moreover, the correlation between the CRGs and immune infiltration of patients with ACC was also analyzed, among which two independent CRGs were identified as independent cupropto- sis-related prognostic predictors of patients with ACC. In conclu- sion, our findings may contribute to the prognostic prediction of ACC and lay a foundation for the personalized treatment of patients with ACC.
2 MATERIALS AND METHODS |
2.1 Public data acquisition and processing |
The Gene Expression Omnibus (GEO) and TCGA were used to obtain gene expression and relevant prognosis data for ACC. UCSC Xena (https://xena.ucsc.edu/) was leveraged to compare information of 77 ACC tumor samples from TCGA to 128 normal tissue samples from Genotype-Tissue Expression (GTEx, http://commonfund.nih. gov/GTEx). A total of three datasets were obtained from the GEO (GSE10927, GSE19750, and GSE90713), among which the GSE10927 dataset contains 33 cancer samples and 10 normal samples, the GSE19750 dataset contains 44 cancer samples and 4 normal sam- ples, while the GSE90713 dataset contains 57 ACC metastases from 42 patients and five normal samples. All data analysis was executed by R (version 4.3.0) and R Bioconductor.
2.2 Differential expression analysis of CRG |
We investigated the expression level of the CRGs between tumor and normal tissue using the aforementioned three datasets. The limma package in R was used to analyze the different expression levels of CRGs between tumor and normal tissue. It was consid- ered statistically significant if the difference met the criterion of p<0.05.
2.3 Enrichment and interaction analysis of CRGs |
Biology process (BP) of gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were per- formed to illustrate the potential processes and pathways highly associated with CRGs function. R language (version 4.3.0) was uti- lized for statistical analysis and visualization. Enrichment analysis
was performed using mainly the clusterProfiler R package (version 3.14.3), and org.Hs.eg.db R package (version 3.10.0) was utilized for ID conversion. Protein-protein interaction (PPI) was used to explore the CRGs-related proteins involved in ACC. PPI was com- pleted by STRING (https://cn.string-db.org/), and Cytoscape bio- informatics software (https://cytoscape.org/) was used to adjust and beautify it.
2.4 Immune infiltration analysis |
Tumor Immune Estimation Resource (TIMER; cistrome.shinyapps. io/timer)13 was leveraged to investigate the relationship between the expression of CRGs and six types of immune cells (B cells, CD4+ T cells, CD8+ T cells, neutrophils, macrophages, and den- dritic cells).
2.5 Development of cuproptosis-related |
risk score
Using the glmnet package in R software, the 10 CRGs were put into the least absolute shrinkage and selection operator (Lasso) Cox regression analysis to minimize the risk of overfitting. The optimal and minimum criteria for the penalty (1) using 10 times cross-vali- dation were selected. In total, six critical genes were generated and were further undergoing multiple stepwise COX regression analy- ses. Afterward, the formula of the risk scoring model was estab- lished, namely risk score=(0.083103718) *FDX1+(-0.278423862) *LIAS+(0.090985682) *DLAT+(-0.018784047) *PDHA1+ (0.297218951) *MTF1+(0.310197964) *CDKN2A. Patients were given risk scores that were formed based on the Lasso prognostic model and categorized into low-risk or high-risk groups accordingly after comparing their risk scores with the median risk score.
2.6 Evaluation of CRRS independence |
The Kaplan-Meier method was performed using the survminer R package (version 0.4.9) in conjunction with the survival R package
FIGURE 1 Expression and genetic alteration of CRGs in ACC: the expression of 10 CRGs in ACC and normal tissue (tumor in red and normal in blue-black). The upper and lower ends of the boxes represent the interquartile range of values. The lines in the boxes represent the median value. * p<0.05, ** p<0.01, *** p<0.001.
(version 3.2-10), in order to illustrate the prognostic outcomes of patients with ACC. The receiver operating characteristic (ROC) curve and the area under the curve (AUC) were used to evalu- ate the prognostic capability of the cuproptosis-related risk score (CRRS) by using the survival ROC and time ROC packages in R. In addition, univariate and multivariate Cox regression analyses were used to investigate the independent prognostic factor of ACC. The independent risk factors in CRGs were evaluated by OS and AUC.
2.7 Statistical analysis
Student’s t-test, and logistic regression were used to assess the re- lationships between clinical characteristics and CRG expression lev- els. All statistical analyses were performed using R software (version 4.3.0). p Values were two-sided, and statistical significance was set at p<0.05.
3 RESULTS
|
3.1 Different expression of CRG in ACC |
We curated a catalog of 10 genes (CDKN2A, DLD, DLAT, FDX1, GLS, LIAS, LIPT1, MTF1, PDHA1, and PDHB) that function closely with cuproptosis as CRGs.10 In the comparison of differentially expressed genes (DEGs) between normal tissue and tumor in patients with ACC using data from TCGA, nine genes were differentially expressed be- tween tumor and normal tissue (Figure 1). Mutation of CRGs expres- sions in ACC in the TCGA datasets was investigated and found that CDKN2A has the highest genetic alteration (Figure S1).
3.2 Correlation between different expression of CRG in ACC |
In addition, we investigated the correlation between the expression of the different CRGs in ACC. As a result, several genes had high cor- relation, for example, LIAS was highly and positively correlated with
10
*
*
*
The expression levels Log2 (TPM+1)
8
.
**
6
Normal
**
Tumor
ns
4
2
0
FDX1
LIAS
LIPT1
DLD
DLAT
PDHA1
PDHB
MTF1
GLS CDKN2A
DLD (r=0.69, p=0.16x 10-163; Figure 2), indicating the possibility of the two taking part in the same biological process.
3.3 Validation of CRG expression
The heatmap (Figure 3) of the 10 CRGs was obtained using the GSE10927, GSE19750, and GSE90713 datasets, respectively. The ex- pressions of LIPT1, PDHB, FDX1, and MTF1 were significantly lower in tumors compared to normal tissue of patients with ACC in GSE10927 (Figure 3A). LISA expression was statistically higher in tumors in GSE19750 (Figure 3B), while the expressions of PDHB and GLS were significantly lower in tumors than in normal tissue of ACC patients in GSE19750 (Figure 3B) and GSE90713 (Figure 3C), respectively.
3.4 Functional enrichment and PPI analysis of ACC |
To demonstrate the biological function of the CRGs, relevant path- ways were analyzed using GO and KEGG databases. According to the GO analysis, the biological processes of the 10 CRGs included mainly the sulfur compound biosynthetic process, the nucleoside bisphos- phate biosynthetic process, the acyl-CoA biosynthetic process, the thioester biosynthetic process, the TCA metabolic process, the acetyl- CoA metabolic process, the citrate metabolic process, the TCA cycle, the acetyl-CoA biosynthetic process, and the acetyl-CoA biosynthetic process from pyruvate (Figure 4A). According to the KEGG pathway enrichment analysis, the 10 CRGs were mainly related to the Carbon metabolism, glycolysis/gluconeogenesis, pyruvate metabolism, the citrate cycle (TCA cycle), central carbon metabolism in cancer, the HIF-1 signaling pathway, the glucagon signaling pathway, glyoxylate
and dicarboxylate metabolism, proximal tubule bicarbonate reclama- tion, and arginine biosynthesis (Figure 4B). A PPI analysis was per- formed to explore the interactions of the CRGs, which revealed that DLAT, DLD, PDHA1, and PDHB were hub genes (Figure 4℃).
3.5 Correlation between CRGs and immune cells |
To further illustrate whether CRGs could affect immune cell recruit- ment to influence ACC prognosis, we analyzed the relationship be- tween the 10 CRGs and immune infiltration in ACC using TIMER.13 The correlation plot displayed a strong association between the 10 CRGs and immune cells (Figure 5, Figure S2). FDX1 showed sig- nificant negative correlations with CD4+ T cells, macrophages, and myeloid dendritic cells. LIPT1 and PDHB showed significant posi- tive correlations with B cells. LIPT1 and CDKN2A showed significant positive correlations with neutrophils. DLD, DLAT, and MTF showed significant negative correlations with CD4+ T cells. PDHA1 and GLS showed significant positive correlations with CD8+ T cells and mac- rophages. Taken together, these results revealed the association be- tween the CRGs and immune infiltration in ACC.
3.6 Construction of the prognostic CRGs risk model |
We used Lasso Cox regression to analyze the 10 CRGs and found that six genes, namely, CDKN2A, DLAT, FDX1, LIAS, MTF1, and PDHA1 were highly correlated with the progno- sis of patients with ACC (Figure 6A,B). CRRS was then con- structed. CRRS=(0.083103718) *FDX1+(-0.278423862) *LIAS+(0.090985682) *DLAT+(-0.018784047) *PDHA1+
| LIAS LIPT1 | DLD | DLAT | PDHA1 | PDHB | MTF1 | GLS | CDKN2A | ||
|---|---|---|---|---|---|---|---|---|---|
| FDX1 | 0.29 | 0.32 | 0.48 | 0.54 | -0.15 | 0.53 | 0.3 | 0.15 | 0.1 |
| LIAS | 0.24 | 0.69 | 0.4 | 0.31 | 0.28 | 0.53 | 0.24 | 0.02 | |
| LIPT1 | 0.29 | 0.48 | 0.01 | 0.44 | 0.51 | 0.47 | 0.31 | ||
| DLD | 0.68 | 0.35 | 0.46 | 0.68 | 0.39 | -0.01 | |||
| DLAT | 0.31 | 0.69 | 0.61 | 0.56 | 0.32 | ||||
| Correlation 1.0 | PDHA1 | 0.15 | 0.21 | 0.46 | -0.02 | ||||
| 0.5 0.0 | PDHB | 0.39 | 0.48 | 0.27 | |||||
| -0.5 | MTF1 | 0.33 | 0.31 | ||||||
| -1.0 | GLS 0.01 | ||||||||
FIGURE 2 Correlation analyses between the expression of CRGs in ACC. Data were analyzed using Pearson correlation analysis.
(A)
GSE10927
(B)
GSE19750
group
group
CDKN2A (ns)
GLS (ns)
PDHA1 (ns)
4
PDHA1(ns)
2
LIAS (ns)
2
LIAS (*)
1
LIPTİ (##)
0
FDX1 (ns)
0
PDHB (##)
-2
DLAT (ns)
-1
GLS (ns)
-4
DLD (ns)
-2
DLD (ns)
group
PDHB (#)
group
Adrenocortical Adenoma
Adrenocortical Carcinoma
DLAT (ns)
Normal Adrenal Cortex
MTF1 (ns)
Normal adrenal gland
FDX1 (#)
CDKN2A(ns)
MTF1 (#)
LIPT1(ns)
(C)
GSE90713
group
PDHA1 (ns)
MTF1 (ns)
4
LIAS (ns)
2
GLS (#)
0
i
CDKN2A (ns)
-2
FDX1 (ns)
-4
LIPT1 (ns)
group
PDHB (ns)
Adrenocortical Adenoma
Normal Adrenal Cortex
DLD (ns)
DLAT (ns)
(0.297218951) *MTF1+(0.310197964) *CDKN2A. The distribu- tion of risk scores and the survival status of patients are shown in Figure 6C, along with the expression level of six CRGs involved in the high- and low-risk groups. The survival time shortened and the number of deaths increased as the risk score increased.
3.7 Validation of independent cuproptosis-related prognostic factors
Patients with ACC were classified into the low- and high-risk groups according to their CRRS, and patients in the high-risk group demon- strated considerably shorter survival time compared with those in the low-risk group (Figure 7A). The AUC values of CRRS in predict- ing the OS time were 0.399 at 1 year, 0.787 at 3 years, and 0.833 at 5 years (Figure 7B).
To determine whether the 6 CRGs could serve as an independent prognostic predictor of OS, univariate and multivariate Cox regres- sion analyses were performed to assess the predictive value of the prognostic model. In univariate Cox regression analysis, a significant difference was found in FDX1 and CDKN2A (Figure 7C). The prog- nostic value of the two CRGs for OS was confirmed in multivariate Cox regression analysis.
3.8 Survival analysis curves of CDKN2A and FDX1
CDKN2A and FDX1 expressions were independent prognostic pre- dictors of patients with ACC. As shown in Figure 8A, we found that
patients with low-CDKN2A expression had dramatically higher sur- vival probability, and the AUC curves of CDKN2A in predicting the survival of ACC was as high as 0.799 (Figure 8C), suggesting that pa- tients with ACC with lower CDKN2A expression might have longer survival. Consistently, low-FDX1 expression was associated with higher survival (Figure 8B), and the AUC curves of FDX1 in predict- ing the survival of ACC was 0.607 (Figure 8D), suggesting that lower FDX1 expression in patients with ACC was related to longer survival. Taken together, CDKN2A and FDX1 expressions had higher confi- dence levels in predicting the prognosis of patients with ACC.
4 DISCUSSION |
Cuproptosis is a copper-dependent form of cell death that has been linked with various cancers.14 Copper, a trace element in the human body, has been closely linked with various signaling pathways and tumor-related biological behaviors.15 An overabun- dance of copper can lead to cell death, and for a long time, the mechanisms and specific forms of copper-induced cell death have remained a mystery.16 Recent studies have proposed that cuprop- tosis is an independent form of cell death, which is believed to be strongly associated with mitochondrial respiration and the lipoic acid (LA) pathway.16
In this study, we investigated the expression of 10 CRGs in ACC tissue and explored their relationship with OS. A novel cupropto- sis-related prognostic score was constructed for the first time. Our study was the first to examine the correlation between CRGs and the development of ACC. Interestingly, most of the 10 CRGs were differentially expressed between tumor and normal tissue, and out
(A)
(B)
enrichment analysis_GO_BP
enrichment analysis_KEGG
sulfur compound biosynthetic process
Carbon metabolism
nucleoside bisphosphate biosynthetic process -
Glycolysis / Gluconeogenesis
acyl-CoA biosynthetic process
p.adjust
Pyruvate metabolism
p.adjust
6e-07
0.08
4e-07
0.06
thioester biosynthetic process
Citrate cycle (TCA cycle)
0.04
tricarboxylic acid metabolic process
2e-07
0.02
Central carbon metabolism in cancer
Counts
HIF-1 signaling pathway
Counts
acetyl-CoA metabolic process
4.0
1
☐ 4.5
☐
2
citrate metabolic process
☐
Glucagon signaling pathway
☐
3
5.0
tricarboxylic acid cycle -
Glyoxylate and dicarboxylate metabolism -
☐
4
acetyl-CoA biosynthetic process
Proximal tubule bicarbonate reclamation -
acetyl-CoA biosynthetic process from pyruvate -
Arginine biosynthesis
0.40
0.42
0.44
0.46 0
0.48
0.50
0.1
0.2
0.3
0.4
0.5
GeneRatio
GeneRatio
(C)
MCM2
PDK2
GSR
CDKN2A
RPIA
SUCLG1
SUCLA2
DLAT
PDHX
PDHA1
LIAS
APRT
ABAT
SDHB
DLST
PDK3
MTF1
FDX1
DLD
PDHB
BOAT1
RACK1
PDHA2
OGDH
LIPT1
AP1M1
OAT
TRAF4
BCKDHA
GLS
of which 6 were significantly associated with OS, suggesting a prob- able role of cuproptosis in the prognosis of ACC and a potential pre- dictive value of this score in the prognosis of ACC.
The prognostic risk model constructed in this study is composed of 6 CRGs (CDKN2A, DLAT, FDX1, LIAS, MTF1, and PDHA1), includ- ing 4 pro-cuproptosis genes (DLTA, FDX1, LISA, and PDHA1) and 2 anti-cuproptosis genes (CDKN2A and MTF1), among which FDX1 and CDKN2A were found to be independent prognostic factors.10
FDX1, a member of the [2Fe-2S] cluster-containing ferredoxin family, encodes a small iron-sulfur protein that is integral to the re- duction of mitochondrial cytochromes and the synthesis of various steroid hormones.17 FDX1 serves as a regulator of protein lipoyla- tion, mediating cuproptosis,10 and is intricately linked to the cellular function, immune response, and disease prognosis of tumors.18,19 It encodes a reductase that is known for reducing Cu2+ to its more
toxic form, Cu1+,10 and exhibits differential expression and modifi- cation levels across various tumors.19 A comprehensive multi-omics pan-cancer study conducted by Xu et al. discovered that FDX1 is highly expressed in 15 types of tumors and exhibits low expression in 11 tumor types.19 Interestingly, FDX1 expression was found to be downregulated in most types of cancers, with higher expression indicating improved OS and death-specific survival rates.2º For in- stance, a pan-cancer analysis revealed that FDX1 acts as a protec- tive gene in several types of cancers including kidney renal clear cell carcinoma, cervical squamous cell carcinoma and endocervical ade- nocarcinoma, liver hepatocellular carcinoma, kidney renal papillary cell carcinoma, mesothelioma, and thyroid carcinoma.21 However, it’s important to note that high-FDX1 expression has also been as- sociated with poor prognosis in glioma.22 These findings suggest that FDX1 could potentially serve as a valuable biomarker for these
FDX1
*
*
**
LIAS
* p < 0.05
LIPT1
*
*
** p < 0.01
DLD
**
Correlation
DLAT
1.0
*
PDHA1
0.5
*
**
PDHB
*
0.0
MTF1
*
-0.5
GLS
**
**
-1.0
CDKN2A
*
B cell
T cell CD4+
T cell CD8+
Neutrophil
Macrophage
Myeloid dendritic cell
cancers. Nevertheless, further research is required to fully compre- hend the role of FDX1 in these and other types of cancer.
CDKN2A, located on chromosome 9p21, is among the most commonly altered tumor suppressor genes in human cancers.23 It encodes two distinct proteins, p14ARF and p16INK4a, each of which exerts unique regulatory effects on the cell cycle.23-25 Specifically, p14ARF interacts with MDM2 and promotes its degra- dation, thereby preventing p53 inactivation through ubiquitin-me- diated proteolysis or transcriptional silencing. On the contrary, p16INK4a binds to cyclin-dependent kinase (CDK) 4/6, effectively blocking cell division during the G1/S phase of the cell cycle.23-25 The occurrence of loss-of-function mutations or homozygous de- letions of CDKN2A leads to the loss of both proteins, resulting in uninhibited cell proliferation and subsequent tumor progres- sion.26 A review of literature databases reveals that CDKN2A has been extensively researched in the context of various tumor types including melanoma and pancreatic cancer.24 However, the
(A)
(B)
10
9
8
6
1
10
) 10 10
10 9 9
9
9
9
9
8
8
8
8
7
6
5
5 441
0
MTF1
Partial Likelihood Deviance
11.0
0.5-
DLAT
CDKN2A
10.5
Coefficients
FDX1
0.0-
PDHA1
10.0
9.5
-0.5
9.0
LIAS
-6
-5
-4
-3
-2
-6
-5
Log(x)
-4
-3
-2
(C)
Log(x)
Risk score
3
Risk group
2
· Low
1
· High
Survival time
4000
3000
Status
2000
Alive
Dead
1000
FDX1
LIAS
2
DLAT
PDHA1
0
MTF1
-2
CDKN2A
(A)
1.0
L
Risk Score
Low
High
Survival probability
0.8
0.6
0.4
0.2
HR = 5.08 (2.40-10.77)
0.0
Log-rank P < 0.001
0
1000
2000
3000
4000
Time(days)
(B)
1.0
0.8
Sensitivity (TPR)
0.6
0.4
0.2
1-Year (AUC = 0.399)
3-Year (AUC = 0.787)
5-Year (AUC = 0.833)
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1-Specificity (FPR)
(C)
| Gene | HR(95% CI) (UniCox) | P (UniCox) | HR(95% CI) (MultiCox) | P (MultiCox) |
|---|---|---|---|---|
| FDX1 | 2.307 (1.041-5.113) | 0.040 | 2.302 (1.039-5.098) | 0.040 |
| LIAS | 0.664 (0.314-1.405) | 0.284 | ||
| DLAT | 1.666 (0.780-3.560) | 0.188 | ||
| PDHA1 | 0.634 (0.297-1.356) | 0.240 | ||
| MTF1 | 1.749 (0.819-3.738) | 0.149 | ||
| CDKN2A | 3.365 (1.474-7.686) | 0.004 | 3.363 (1.472-7.681) | 0.004 |
0
2
4
6
8
(A)
(B)
Adrenocortical Carcinoma
Adrenocortical Carcinoma
1.0
CDKN2A
1.0
FDX1
Low
Low
High
High
Survival probability
0.8
Survival probability
0.8
0.6
0.6
0.4
0.4
0.2
Overall Survival
0.2
Overall Survival
HR = 3.37 (1.47-7.69)
HR = 2.31 (1.04-5.11)
0.0
P = 0.004
0.0
P = 0.04
0
50
100
150
0
50
100
150
Time (months)
Time (months)
(C)
(D)
Adrenocortical Carcinoma
Adrenocortical Carcinoma
1.0
1.0
0.8
0.8
Sensitivity (TPR)
Sensitivity (TPR)
0.6
0.6
0.4
0.4
0.2
CDKN2A
0.2
FDX1
AUC: 0.799
AUC: 0.607
CI: 0.522-0.692
0.0
CI: 0.725-0.874
0.0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
1-Specificity (FPR)
1-Specificity (FPR)
FIGURE 8 Survival analysis curves of CDKN2A and FDX1 expression. The Kaplan-Meier plot for the expression of (A) CDKN2A and (B) FDX1 and overall survival. The AUC curves of CDKN2A (C) and FDX1 (D) in predicting the survival of ACC.
findings from these studies remain controversial.24,27 In this study, CDKN2A was also confirmed to have a relatively high percentage of nonsense-mutation in the 10 CRGs. The survival rate of patients
with ACC with high-CDKN2A expression was significantly worse than those with low-CDKN2A expression. Two possible explana- tions include: (1) CDKN2A mutation results in loss of the encoding
of cycle regulators p14ARF and p16INK4a, leading to uncontrolled cell proliferation and tumor formation.28 (2) CDKN2A may not be critical for tumorigenesis despite its high-single nucleotide variation.29
This study features scientific value as follows: Firstly, this is cur- rently the first prognostic CRG risk model targeting ACC. Cell death, as a basis for cancer origin and development, has been an intense and hot area in tumor research.3º Cuproptosis, as a novel way of cop- per-dependent cell death, is distinct from all other known cell death mechanisms.10 Moreover, traditional prognostic factors, such as tumor stage and resection status, have limited effects in predicting the prog- nosis.3 Due to patient scarcity, poor outcomes, and high-resource de- mand, there are only a few randomized controlled ACC trials. Studies targeting cuproptosis may offer a new way to make a better prognos- tic evaluation and potential treatment for patients with ACC.
In terms of ongoing or future studies that explore the thera- peutic targeting of cuproptosis in ACC, there indeed exist studies that investigate cuproptosis as a mechanism and potential inter- vention target in various diseases.31-33 For instance, the use of copper chelators, inhibitors of copper chaperone proteins, and copper ionophores could potentially serve as therapeutics tar- geting cuproptosis in cardiovascular diseases.31 Moreover, tetra- thiomolybdate (TTM), a copper-chelating agent known for its high affinity for copper, is currently utilized in the treatment of Wilson’s disease, a condition characterized by the excessive ac- cumulation of copper in the liver.34 However, studies specifically focusing on ACC are currently limited. We are of the belief that further research in this area could potentially pave the way for new therapeutic interventions.
This study also presents several limitations: First, all three datasets used in this article are of limited size (n <50), amplifying the influence of error bias. For instance, the expression of FDX1 in ACC tumor tissue was observed to be lower than that of nor- mal tissue, but prognostic analysis indicated that lower expres- sion of FDX1 was associated with a higher patient survival rate. This anomaly may have been caused by an insufficient number of patients with ACC. The World Health Organization (WHO) has recognized four histological variants of ACC, namely, oncocytic, myxoid, sarcomatoid and pediatric.35,36 Different subtypes may have different prognoses. It is crucial to further increase the bi- ological research of ACC with a larger sample size in the future. Second, although the prognostic model using CRGs expression demonstrated a good performance in predicting ACC survival, other important genes and traditional prognostic factors were not included in this study. A combination of multiple influence factors may result in more accurate prognosis predictions. Third, due to the low incidence of ACC, prospective studies with larger sample sizes in multiple centers are required. Four, 10 CRGs are derived from an article published in Science.1º However, these 10 genes may not comprehensively illuminate the mechanism of cupro- ptosis, suggesting that there could be additional CRGs awaiting discovery.
5 CONCLUSION |
In conclusion, this study shows that cuproptosis may play a role in the development of ACC. Patients with ACC with higher CRRS were associated with shorter survival time. Among the 6 CRGs used to es- tablish CRRS, CDKN2A, and FDX1 are independent prognostic pre- dictors in patients with ACC. Such results provide a new perspective on therapeutic strategies related to CRGs for cancer prevention and treatment.
AUTHOR CONTRIBUTIONS
The study was designed by Jingchao Wei and Bohan Wang. Data ac- quisition was carried out by Jingchao Wei, Wenjun Gao, and Xiaoyan He. Bioinformatics analysis was conducted by Jingchao Wei. Qi Huangfu, Chengfang Sun, and Jingchao Wei provided useful advice for the analyses of the data. The manuscript was drafted by Wenjun Gao and Xiaoyan He, and was revised by all authors before the final version was approved to be published.
ACKNOWLEDGMENTS
We thank Dr. Mushuang Hu for the help in editing and language pol- ishing the manuscript.
FUNDING INFORMATION
This manuscript was supported by the National Natural Science Foundation of China (no. 81970601 to Bohan Wang, no. 82200850 to Jingchao Wei).
CONFLICT OF INTEREST STATEMENT
The authors report no conflicts of interest.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available in the Gene Expression Omnibus (GEO) at https://www.ncbi.nlm. nih.gov/geo/, the Cancer Genome Atlas (TCGA) at https://www.can- cer.gov/ccg/research/genome-sequencing/tcga and the Genotype- Tissue Expression (GTEx) at https://commonfund.nih.gov/GTEx/.
ORCID Wenjun Gao (D https://orcid.org/0000-0003-2272-1813
REFERENCES
1. Kebebew E, Reiff E, Duh QY, Clark OH, McMillan A. Extent of dis- ease at presentation and outcome for adrenocortical carcinoma: have we made progress? World J Surg. 2006;30(5):872-878.
2. Kerkhofs TM, Verhoeven RH, Van Der Zwan JM, et al. Adrenocortical carcinoma: a population-based study on incidence and survival in The Netherlands since 1993. Eur J Cancer. 2013;49(11):2579-2586.
3. Crona J, Beuschlein F. Adrenocortical carcinoma-towards genom- ics guided clinical care. Nat Rev Endocrinol. 2019;15(9):548-560.
4. Fassnacht M, Johanssen S, Fenske W, et al. Improved survival in patients with stage II adrenocortical carcinoma followed up prospectively by specialized centers. Clin Endocrinol Metab. 2010;95(11):4925-4932.
| WILEY
5. Fassnacht M, Johanssen S, Quinkler M, et al. Limited prognostic value of the 2004 International Union against Cancer staging clas- sification for adrenocortical carcinoma: proposal for a revised TNM classification. Cancer. 2009;115(2):243-250.
6. Fassnacht M, Terzolo M, Allolio B, et al. Combination chemo- therapy in advanced adrenocortical carcinoma. N Engl J Med. 2012;366(23):2189-2197.
7. Ruiz LM, Libedinsky A, Elorza AA. Role of copper on mitochondrial function and metabolism. Front Mol Biosci. 2021;8:711227.
8. Ge EJ, Bush AI, Casini A, et al. Connecting copper and cancer: from transition metal signalling to metalloplasia. Nat Rev Cancer. 2022;22(2):102-113.
9. Gunjan D, Shalimar NN, et al. Hepatocellular carcinoma: an unusual complication of longstanding Wilson disease. Clin Exp Hepatol. 2017;7(2):152-154.
10. Tsvetkov P, Coy S, Petrova B, et al. Copper induces cell death by targeting lipoylated TCA cycle proteins. Science. 2022;375(6586):1254-1261.
11. 1. Tardito S, Bassanetti I, Bignardi C, et al. Copper binding agents acting as copper ionophores lead to caspase inhibition and paraptotic cell death in human cancer cells. J Am Chem Soc. 2011;133(16):6235-6242.
12. Bilcikova J, Fialkova V, Duranova H, et al. Copper affects ste- roidogenesis and viability of human adrenocortical carcinoma (NCI-H295R) cell line in vitro. J Environ Sci Health A Tox Hazard Subst Environ Eng. 2020;55(9):1070-1077.
13. Li T, Fan J, Wang B, et al. TIMER: a web server for comprehen- sive analysis of tumor-infiltrating immune cells. Cancer Res. 2017;77(21):e108-e110.
14. Xie J, Yang Y, Gao Y, He J. Cuproptosis: mechanisms and links with cancers. Mol Cancer. 2023;22(1):46.
15. Cheng F, Peng G, Lu Y, et al. Relationship between copper and immunity: the potential role of copper in tumor immunity. Front Oncologia. 2022;12:1019153.
16. Wang Y, Zhang L, Zhou F. Cuproptosis: a new form of programmed cell death. Cell Mol Immunol. 2022;19(8):867-868.
17. . Zhang J, Kong X, Zhang Y, et al. FDXR regulates TP73 tumor sup- pressor via IRP2 to modulate aging and tumor suppression. J Pathol. 2020;251(3):284-296.
8. Lu H, Zhou L, Zhang B, Xie Y, Yang H, Wang Z. Cuproptosis key gene FDX1 is a prognostic biomarker and associated with immune infiltration in glioma. Front Med (Lausanne). 2022;9:939776.
19. Xu J, Hu Z, Cao H, et al. Multi-omics pan-cancer study of cupropto- sis core gene FDX1 and its role in kidney renal clear cell carcinoma. Front Immunol. 2022;13:981764.
20. Xiao C, Yang L, Jin L, et al. Prognostic and immunological role of cuproptosis-related protein FDX1 in pan-cancer. Front Genet. 2022;13:962028.
21. Zhang Q, Liu N, Wu D, et al. Pan-cancer analysis reveals the value of FDX1 as a novel biomarker in the prognosis and immunotherapy in human tumors. 2022.
22. . Zhang G, Shen L, Li Z, Zhao Y. FDX1 serves as a prognostic bio- marker and promotes glioma progression by regulating the immune response. Aging. 2023;15(11):4963-4985.
23. Adib E, Nassar AH, Akl EW, et al. CDKN2A alterations and re- sponse to immunotherapy in solid tumors. Clin Cancer Res. 2021;27(14):4025-4035.
24. Chan SH, Chiang J. CDKN2A germline alterations and the relevance of genotype-phenotype associations in cancer predisposition. Hered Cancer Clin Pract. 2021;19(1):21.
25. Sulong S, Moorman AV, Irving JA, et al. A comprehensive analysis of the CDKN2A gene in childhood acute lymphoblastic leukemia reveals genomic deletion, copy number neutral loss of heterozy- gosity, and association with specific cytogenetic subgroups. Blood. 2009;113(1):100-107.
26. Bihl MP, Foerster A, Lugli A, Zlobec I. Characterization of CDKN2A(p16) methylation and impact in colorectal cancer: sys- tematic analysis using pyrosequencing. J Transl Med. 2012;10:173.
27. Ipenburg NA, El Sharouni MA, Van Doorn R, et al. Lack of associa- tion between CDKN2A germline mutations and survival in patients with melanoma: a retrospective cohort study. J Am Acad Dermatol. 2022;87(2):479-482.
28. . Zhao R, Choi BY, Lee MH, Bode AM, Dong Z. Implications of ge- netic and epigenetic alterations of CDKN2A (p16 INK4a) in cancer. EBioMedicine. 2016;8:30-39.
29. Liu H. Pan-cancer profiles of the cuproptosis gene set. Am J Cancer Res. 2022;12(8):4074-4081.
30. Strasser A, Vaux DL. Cell death in the origin and treatment of can- cer. Mol Cell. 2020;78(6):1045-1054.
31. Yang Y, Feng Q, Luan Y, et al. Exploring cuproptosis as a mechanism and potential intervention target in cardiovascular diseases. Front Pharmacol. 2023;14:1229297.
32. . Tang D, Chen X, Kroemer G. Cuproptosis: a copper-triggered mo- dality of mitochondrial cell death. Cell Res. 2022;32(5):417-418.
33. Li D, Gao Z, Li Q, Liu X, Liu H. Cuproptosis-a potential target for the treatment of osteoporosis. Front Endocrinol (Lausanne). 2023;14:1135181.
34. Borchard S, Raschke S, Zak KM, et al. Bis-choline tetrathiomolyb- date prevents copper-induced blood-brain barrier damage. Life Sci Alliance. 2022;5(3):1164-1174.
35. Lam AK. Adrenocortical carcinoma: updates of clinical and patho- logical features after renewed World Health Organisation classifi- cation and pathology staging. Biomedicines. 2021;9(2):175-199.
36. Minner S, Schreiner J, Saeger W. Adrenal cancer: relevance of different grading systems and subtypes. Clin Transl Oncol. 2021;23(7):1350-1357.
SUPPORTING INFORMATION
Additional supporting information can be found online in the Supporting Information section at the end of this article.
How to cite this article: Gao W, He X, Huangfu Q, et al. A novel cuproptosis-related prognostic gene signature in adrenocortical carcinoma. J Clin Lab Anal. 2023;37:e24981. doi:10.1002/jcla.24981