ELSEVIER
Contents lists available at ScienceDirect
Molecular and Cellular Endocrinology
journal homepage: www.elsevier.com/locate/mce
-
Molecular and Cellular Endocrinology
.
Identifying prognostic hub genes and key pathways in pediatric adrenocortical tumors through RNA sequencing and Co-expression analysis
Check for updates
Luciana Chain Veroneza, Alcides Euzebio Tavares Xavier b, Luiz Fernando Nagano b, Carolina Alves Pereira Correa b, Kleiton Silva Borges , Paula Santos , Mirella Baroni b, Rosane de Paula Silva Queirozª, Sonir Roberto Rauber Antonini ª, José Andres Yunese, Silvia Regina Brandalise e, Carlos Augusto Fernandes Molinaf, Emilia Modolo Pinto &, Elvis Terci Valera ª, Luiz Gonzaga Tone a,b, Carlos Alberto Scrideli a,b,h,”
ª Departments of Pediatrics and Ribeirão Preto Medical School, University of São Paulo, 14049-900, Ribeirão Preto, SP, Brazil
b Departments of Genetics, Ribeirão Preto Medical School, University of São Paulo, 14049-900, Ribeirão Preto, SP, Brazil
” Department of Pediatrics, Harvard Medical School, Boston, USA
d Department of Psychology, Ribeirão Preto Faculty of Philosophy, Sciences and Letters, 14049-900, Ribeirão Preto, SP, Brazil
e Boldrini Children’s Center, 13083-210, Campinas, SP, Brazil
Department of Surgery and Anatomy, Ribeirão Preto Medical School, University of Sao Paulo, São Paulo, 14049-900, Ribeirão Preto, SP, Brazil
% Department of Pathology, St. Jude Children’s Research Hospital, Memphis, TN, 38105, USA
h National Science and Technology Institute for Children’s Cancer Biology and Pediatric Oncology, INCT BioOncoPed, Brazil
ARTICLE INFO
Handling Editor: Carolyn M. Klinge
Keywords: pediatric adrenocortical tumor Prognosis hub genes
ABSTRACT
Pediatric adrenocortical tumors (ACTs), rare conditions with uncertain prognoses, have high incidence in southern and southeastern Brazil. Pediatric ACTs are highly heterogeneous, so establishing prognostic markers for these tumors is challenging. We have conducted transcriptomic analysis on 14 pediatric ACT samples and compared cases with favorable and unfavorable clinical outcomes to identify prognostically significant genes. This comparison showed 1257 differentially expressed genes in favorable and unfavorable cases. Among these genes, 15 out of 60 hub genes were significantly associated with five-year event-free survival (EFS), and 10 had significant diagnostic value for predicting ACT outcomes in an independent microarray dataset of pediatric adrenocortical carcinomas (GSE76019). Overexpression of N4BP2, HSPB6, JUN, APBB1IP, STK17B, CSNK1D, and KDM3A was associated with poorer EFS, whereas lower expression of ISCU, PTPR, PRKAB2, CD48, PRF1, ITGAL, KLK15, and HIST1H3J was associated with worse outcomes. Collectively, these findings underscore the prog- nostic significance of these hub genes and suggest that they play a potential role in pediatric ACT progression and are useful predictors of clinical outcomes.
1. Introduction
Pediatric adrenocortical tumors (ACTs) represent rare neoplasms with worldwide incidence of 0.2-0.3 cases per million children annu- ally. However, in the southern and southeastern regions of Brazil, this incidence surges to 10-15 times the global average, which has been attributed to the heightened prevalence of the TP53 p.R337H founder variant (Sandrini et al., 2005; Latronico et al., 2001; Figueiredo et al., 2006). Depending on disease presentation, the five-year survival of children with ACT ranges between 54 and 77% (Gupta et al., 2018;
Michalkiewicz et al., 2004). Metastatic cases have notably poor out- comes, with an estimated five-year survival rate of less than 20% (Riedmeier et al., 2021). Presently, for children with ACT in early stages, curing this rare disease depends on completely resecting the tumor. In turn, children with advanced stages of this disease typically undergo adjuvant therapy with mitotane, cisplatin, etoposide, and doxorubicin, but approximately 30-50% of these patients will experience local or distant recurrences (Rodriguez-Galindo et al., 2021a; Pinto et al., 2020; Brondani and Fragoso, 2020).
Treatment protocol ARAR0332 proposed by the Children Oncology
* Corresponding author. Department of Pediatrics, Ribeirão Preto Medical School, University of São Paulo, Avenida Bandeirantes 3900, Ribeirão Preto, SP, 14091- 900, Brazil.
E-mail address: scrideli@fmrp.usp.br (C.A. Scrideli).
https://doi.org/10.1016/j.mce.2024.112383
Group (COG) and currently used to treat children with ACT is based on the stage of the disease-stages I and II indicate the best prognoses, while stages III and IV point to the worst prognoses (Rodriguez-Galindo et al., 2021b). However, pediatric ACTs are highly heterogeneous, and tumor evolution between stages remains partially uncertain (Brondani and Fragoso, 2020; Mete et al., 2022; Ilanchezhian et al., 2022).
Despite recent research indicating dysregulated IGF (Lira et al., 2016), WNT (Leal et al., 2011), and Shh pathways (Gomes et al., 2014) and altered microRNA (Veronez et al., 2022) in pediatric ACTs, the biological characteristics and clinical behavior of these tumors are not completely understood. The nature of pediatric ACTs varies, so prog- nostication of this disease is challenging. Therefore, deeper molecular understanding of these tumors is essential for identifying potential therapeutic targets, prognostic markers, and treatment approaches to improve patient survival (Rodriguez-Galindo et al., 2021a; Mete et al., 2022). The present study aims to identify differences in the RNA expression profiles of pediatric ACT cases with favorable and unfavor- able prognoses, to unveil potential therapeutic targets and prognostic biomarkers. By doing so, we hope to enhance our understanding of this rare disease and to improve clinical outcomes.
2. Methods
2.1. Patients and tumor specimens
This study was conducted in accordance with the Declaration of Helsinki, and it was approved by the Ethics Committees of the Ribeirão Preto Medical School, University of São Paulo, Brazil (protocol number 15509/2016) and Boldrini Children’s Center, Campinas, state of São Paulo, Brazil (protocol number 1.75-050809). All the participants or their legal guardians provided signed informed consent allowing spec- imens to be collected. Tumor samples were obtained from pediatric patients (aged from 0 to 18 years) diagnosed with ACT and assisted at two Brazilian treatment centers (Clinics Hospital of Ribeirão Preto, University of São Paulo, and Boldrini Children’s Center, Campinas). Specifically, 14 samples were collected from 14 untreated patients (one sample was collected from each patient) after surgical resection and immediately snap-frozen in liquid nitrogen for later RNA sequencing analysis. The samples were divided into two groups: samples collected from patients with favorable clinical outcomes (n = 9, favorable group), and samples collected from patients with unfavorable clinical outcomes (n = 5, unfavorable group). Recurrence or death due to the disease was considered an unfavorable outcome. Unfavorable events occurred be- tween 7 and 59 months (median of 16 months) after sample collection. The follow-up time was 51-176 months (median of 84 months) for pa- tients with favorable outcomes. Of the patients analyzed, 78.57% were girls, 28.57% were metastatic at diagnosis. The TP53 p.R337H mutation was evaluated by direct DNA sequencing and was detected in 85.71% of patients analyzed. The analysis of the entire coding and boundary re- gions of the TP53 gene revealed the absence of other mutations. Sup- plementary Table 1 lists selected demographic and clinical features of these 14 patients. All the patients in the unfavorable group were treated according to the COG or COG-like protocols based on mitotane, cisplatin, anthracyclines, and etoposide. The patients in the favorable group only underwent surgery. A microarray dataset comprising 34 cases of pediatric adrenocortical carcinomas (GSE76019) was down- loaded from the Gene Expression Omnibus (GEO) database and used to validate the prognostic and diagnostic significance of the hub genes identified through RNA sequencing analysis. Of note, 10 out of the 34 cases in the GSE76019 dataset carried the TP53 p.R337H variant, and all the 34 cases were treated according to the COG protocol (Pinto et al., 2016).
2.2. RNA sequencing and data processing
Total RNA was extracted from the 14 pediatric ACT samples by using
the AllPrep DNA/RNA/Protein Mini kit (Qiagen, Hilden, Germany). RNA integrity was assessed with the Agilent Bioanalyzer 2100 (Agilent Technologies, Santa Clara, USA); the manufacturer’s specifications were followed. The library was constructed, and RNA was sequenced with TruSeq® Stranded Total RNA Human (Illumina). Mitochondrial and cytoplasmic RNA were depleted by employing Ribo-Zero Gold magnetic beads. Fragment sizes were evaluated by using the D1000 kit on the TapeStation (Agilent Technologies). Sequencings were performed on the HiSeq® 2500 equipment (Illumina).
Reads were trimmed by using FlexBar, and quality control was conducted with FastQC and MultiQC (Ewels et al., 2016). Processed reads were aligned with HISAT2, which employs a hierarchical indexing scheme based on the Burrows-Wheeler transformation and Ferragina-Manzini index (Kim et al., 2015). Transcripts were assembled and quantified by using Stringtie; the String Merge tool and Features Count were employed for concatenation and read counting, respectively (Liao et al., 2014). Differential expression was analyzed by using DESeq2 (v1.26.0) (Love et al., 2014). Raw and normalized data are available at the GEO database under the accession number GSE182022.
2.3. Identification of differentially expressed genes
Differentially expressed genes (DEGs) were identified by comparing gene expression in pediatric ACT cases with favorable and unfavorable prognoses; recurrence, mortality, or both were considered unfavorable events. In our RNA-seq dataset, DEGs were detected and visualized by using the DESeq2 package in R (Leal et al., 2011) and defined by a |log2 (fold change)| > 1 and an adjusted p-value <0.05. Unsupervised hier- archical clustering and heatmaps were created by using the Complex- Heatmap package (Gu et al., 2016) with Pearson’s distance metric and the WardD2 method for linkage analysis. A volcano plot of the DEGs was generated by using the ‘ggplot2’ package in R.
2.4. Analysis of co-expression modules and hub genes
Gene modules were identified through co-expression analysis by using the CEMiTool R package (Russo et al., 2018) with default pa- rameters, including an unbiased selection of genes based on a variance-based filter (p-value <0.05) and Pearson’s correlation coeffi- cient. DeSeq2 normalized counts of coding genes were employed as input to construct the network in CEMiTool, where the DEGs were grouped into modules via average linkage hierarchical clustering ac- cording to their expression patterns. Additionally, hub genes in each module were identified as the genes with the highest number of con- nections, interactions, or both.
2.5. Enrichment analysis
To understand the predicted functions of the DEGs better, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis was accomplished by using the “clusterProfiler” package in R. GO terms and KEGG pathways with an adjusted p-value <0.05 were considered statistically significant.
2.6. Statistical analysis
The prognostic value of the hub genes for predicting pediatric ACT outcomes was evaluated by using Kaplan-Meier curves and long-rank tests to analyze event-free survival (EFS); the Statistical Package for the Social Sciences 20.0 for Windows (SPSS, Chicago, IL, USA) and the GSE76019 dataset were employed. The pediatric ACT cases were divided into two groups (higher versus lower expression) on the basis of the median expression values of the hub genes. EFS time was defined as the time elapsed between pediatric ACT diagnosis and relapse, death, or both, considered unfavorable events. The association between the levels of expression of the analyzed genes and the clinical and biological
variables gender, age, tumor size, hormonal production, TP53 status, metastasis, and death was determined by the Mann-Whitney test; the median values of expression were employed as cut-off point. A p-value <0.05 was considered significant.
2.7. Diagnostic significance analysis
The diagnostic significance of the hub genes for distinguishing be- tween favorable and unfavorable (relapse or death) pediatric ACT cases was assessed by using receiver operating characteristic (ROC) curves and the area under the curve (AUC) values based on the expression data of the GSE76019 dataset. An AUC value closer to 1 indicates superior diagnostic performance. Significance was set at p-value <0.05.
2.8. Identification of candidate drugs for hub genes
Drugs targeting the hub genes overexpressed in unfavorable pedi- atric ACT cases and which can significantly impact patient survival were identified by using the DrugBank database and DGIdb (http://www. dgidb.org). The DrugBank database is a comprehensive, free-to-access bioinformatics and cheminformatics resource containing information about over 7800 drugs and their targets (Wishart et al., 2018), while DGIdb searches multiple drug databases and uses text-mining to identify drugs matching input proteins (Freshour et al., 2021). Results were visualized by using the “ggplot2 (3.2.1)” and “ggalluvial (0.11.1)” packages in R, with an alluvial diagram illustrating the drug-gene interaction network.
3. Results
3.1. Identifying DEGs in pediatric ACTs with favorable and unfavorable prognoses
We identified 1257 DEGs in ACT samples obtained from pediatric patients with favorable (n = 9) and unfavorable (n = 5) clinical out- comes in our in-house cohort. Among these DEGs, 485 were down- regulated and 772 were upregulated in patients with unfavorable prognoses (Fig. 1A and B, Supplementary Table 2). To explore the po- tential pathways and biological processes involved in pediatric ACT prognoses, we performed KEGG and GO enrichment analysis of the down- and upregulated DEGs.
The downregulated DEGs were primarily associated with pathways involved in cell adhesion molecules and diverse infections (Fig. 1C) as well as GO terms mediating T cell and leukocyte adhesion, activation, and proliferation (Fig. 1D). Conversely, the upregulated DEGs were enriched in pathways such as MAPK signaling, lipid production and atherosclerosis, some hormone signaling pathways, and the Hedgehog signaling pathway (Fig. 1E) and were mainly associated with GO terms mediating cell-cell adhesion, axonogenesis, homophilic cell adhesion, cilium organization, cilium assembly, neuron projection guidance, centrosome cycle, microtubule organizing center organization, ciliary basal body-plasma membrane docking, and centriole replication (Fig. 1F).
3.2. Identifying co-expression modules and hub genes
To characterize how DEGs interact and to identify hub genes, we applied the CEMiTool modular co-expression network analysis. Overall,
A
B
C
subgroups
Cell adhesion molecules
10
Good Prognosis
subgroups
Poor Prognosis
Staphylococcus aureus
infection
00
Tuberculosis
- log10 P adj
Hematopoietic cell lineage
6
zscore 2
Phagosome
p.adjust
1
19-08
20-04
V
0
1
Influenza A
2
Coronavirus disease - COVID-19
N
Systemic lupus erythematosus
O
Epstein-Barr virus infection
-10
-5
0
5
10
Human T-cell leukemia virus
1 infection
Log2 fold change
0
5
10
15
20
25
D
E
F
T cell activation
cell-cell adhesion via plasma-membrane adhesion molecules
MAPK signaling pathway
leukocyte cell-cell adhesion
Lipid and atherosclerosis
axonogenesis
lymphocyte proliferation
homophilic cell adhesion via
Oxytocin signaling pathway
plasma membrane adhesion molecules
p.adjust
mononuclear cell proliferation
fe-14
GnRH signaling pathway
cilium organization
Count
20-14
10
leukocyte proliferation
de-14
20
46-14
p.adjust
Se-14
Estrogen signaling pathway
001
cilium assembly
30
regulation of lymphocyte
0.02
Count
p.adjust
proliferation
30
Circadian entrainment
0.00
neuron projection guidance
40-04
regulation of mononuclear
40
cell proliferation
45
Inflammatory mediator
regulation of TRP channels
centrosome cycle
regulation of leukocyte proliferation
Parathyroid hormone
synthesis, secretion and-
microtubule organizing center organization
action
regulation of T cell proliferation
Hedgehog signaling pathway
ciliary basal body-plasma membrane docking
T cell proliferation
Lysine degradation
centriole replication
0.08
0.09
0.10
0.11
GeneRatio
0
5
10
15
20
0.02
0.04
0.06
GeneRatio
we identified five co-expression modules (M1, M2, M3, M4, and M5) and 60 hub genes, with approximately 12-14 hub genes per module (Fig. 2). The smallest module (M5) contained 39 co-expressed genes, while the largest module (M1) presented 743 genes.
3.3. Prognostic value of the hub genes
To investigate whether expression of the 60 hub genes could impact patient survival, we used clinical and expression data from the GSE76019 dataset. This dataset used gene expression to associate the diagnostic tumor expression of these genes with clinical outcomes in 34 ACT pediatric patients treated with the COG ARAR0332 protocol. We identified 15 hub genes that significantly impacted EFS (Fig. 3 and Supplementary Table 3). Overexpression of N4BP2, HSPB6, JUN, APB- B1IP, STK17B, CSNK1D, and KDM3A was associated with poorer EFS, whereas lower expression of ISCU, PTPRC, PRKAB2, CD48, PRF1, ITGAL, KLK15, and HIST1H3J was associated with worse EFS (Fig. 3).
Girls with pediatric ACT had higher expression of ISCU (p = 0.020). Pediatric ACT patients aged less than 3 years presented lower expression of KDM3A (p = 0.020). Pediatric ACT patients with functional tumors presented lower expression of KDM3A (p = 0.043) and higher expression of ISCU (p = 0.006) and PRF1 (p = 0.047). Pediatric ACT patients with TP53 mutation presented lower expression of CSNK1D (p = 0.048) and higher expression of PRKAB2 (p = 0.001).
Pediatric ACT patients with tumor weighing more than 200 g had higher expression of N4BP2 (p = 0.006) and lower expression of KDM3A (p = 0.049), ISCU (p= 0.026), PTPRC (p=0.002), CD48 (p=0.002), PRF1 (p= 0.005), and KLK15 (p = 0.047). The presence of metastasis at diagnosis was associated with lower expression of PRKAB2 (p = 0.024) and HIST1H3J (p = 0.016).
Death was associated with higher expression of N4BP2 (p = 0.019), STK17B (p=0.033), CSNK1D (p=0.007), and KDM3A (p=0.008) and lower expression of ISCU (p = 0.009), PTPRC (p = 0.018), PRKAB2 (p =
0.011), CD48 (p=0.006), PRF1 (p=0.012), and KLK15 (p=0.023).
To assess the diagnostic significance of the hub genes in discrimi- nating between favorable and unfavorable ACT cases, we used the GSE76019 dataset to plot ROC curves and determined the AUC at the optimum threshold. This analysis included only the hub genes that significantly impacted patient survival. Among the 15 identified hub genes, 10 demonstrated significant diagnostic value with accuracies ranging from 0.650 to 0.788. The hub genes with the highest accuracy were CD48 (AUC: 0.788, 95% CI: 0.609-0.966), CSNK1D (AUC: 0.782, 95% CI: 0.609-0.955), KDM3A (AUC: 0.778, 95% CI: 0.610-0.947), and ISCU (AUC: 0.777, 95% CI: 0.610-0.943) (Table 1).
3.4. Identifying druggable hub genes associated with unfavorable prognoses in pediatric ACTs
As shown in Fig. 4 and Supplementary Table 4, by using the DGIdb and DrugBank databases, we identified 88 potential drugs for four out of the seven hub genes whose higher expression was associated with poor survival in pediatric ACTs (HSPB6, JUN, KDM3A, and CSNK1D). Spe- cifically, JUN, CSNK1D, KDM3A, and HSPB6 were targeted by 52, 32, 3, and 1 drug, respectively. No drugs targeting the other three hub genes were predicted in the drug databases.
4. Discussion
Herein, we performed RNA sequencing on 14 samples of untreated pediatric ACTs to understand the molecular characteristics that may distinguish between cases with favorable and unfavorable clinical out- comes. In our analysis, we compared tumors associated with poor out- comes to tumors associated with good outcomes and identified 1257 DEGs. Furthermore, on the basis of interaction network evaluation, we identified 15 hub genes with prognostic value for pediatric ACTs, many of which are dysregulated across several types of cancer.
M1
ST13
M2
PRF1
M3
BTBD3
SKI
HIST 1H3J
MATK
MLLT6
APBB1IP
VASN
SYNGAP1
Hub
STRADA
PRKAB2
THRB
Hub
KDM3A
CD163
LST1
Co-expression
NOG
Degree
PTP4A1
a Co-expression
Interaction
CHORDC1
100
Interaction
200
MLLT1
FZD4
CCL5]
UBE2L6
ITGAL
Degree
Degree
KLC2
DNAJA4
BAG3
300
CSNK1D
CBFA2T2
200
CD53
PTP4A1
200
CSRP1
400
400
Hub
600
Co-expression
JUN
600
RUSC1
CD48
SCIMP
800
3 Interaction
PTPRC
N4BP2
SHC2
MAP3K1
IQGAP2
M4
STK17B
M5
ITPR3
SLC7A7
CYP27A1
GRK1
Degree
Degree
ERBB2
KLK15
100
200
RIPK4
20
DDB1
TM4SF4
CEP72
40
300
60
400
ECM1
80
ERBB3
SUCLG1
500
TMEM54
Hub
ISCU
Hub
ARHGEF2
a Co-expression
Co-expression
Co-expression + Interaction
EPHA7
Interaction
DCST2
HSPB6
a Interaction
MUC6
EXOC3
BRD9
Lower expression
Higher expression
1.0
I
CD48
1.0
MATK
1,0
PFR1
1.0
JUN
1,0
HIST1H3J
0,8
0,8-
0,8
0,8
0,8
Survival probability
Survival probability
Survival propability
Survival probability
Survival probability
0,6-
0,6-
0,6
0,6-
0,6-
0,4
0,4
0,4
0,4
0,4
0,2
0,2-
0,2
0,2
0,2-
0.0
P=0.016
0,0
P<0.001
0,0-
P=0.036
0.0
P=0.032
0,0
P=0.048
,00
20,00
40,00
60,00
80,00
100,00
00
20,00
40,00
60,00
80,00
100,00
,00
20,00
40,00
60,00
80,00
100,00
00
20,00
40,00
60,00
80,00
100,00
00
20,00
40.00
60,00
80,00
100,00
Time (months)
Time (months)
Time (months)
Time (months)
Time (months)
1.0
KDM3A
1,0
PRKAB2
1,0
CSNK1D
1.0
PTPRC
1,0
APBB1IP
0,0
0,8
0,8
0,8
0,8
Survival probability
Survival propability
Survival probability
Survival probability
Survival probability
0,6
0,6-
0,6
0,6-
0,6-
0,4
0,4-
0.4
0,4-
0.4”
0,2
0,2-
0,2
02-
0.2
P=0.016
0,0
P=0.040
0.0
P=0.037
0,0
0,0
P=0.012
0,0
P=0.038
00
20,00
40.00
60,00
80.00
100,00
00
20,00
40,00
60,00
80,00
100,00
00
20,00
40,00
60,00
00,00
100,00
00
20,00
40,00
60,00
80,00
100,00
00
20,00
40,00
60,00
80,00
100,00
Time (months)
Time (months)
EFS
Time (months)
Time (months)
1.0
I
HSPB6
1.0
I
N4BP2
1.0
ISCU
1.0
I
LK15
1,0
I
STK17B
0,8
0,8
0.8
0,8
0,8
Survival probability
Survival probability
Survival probability
Survival probability
Survival probability
0,8
0,6-
0.6
0,6
0,6
0,4
0.4
0.4
0,4
0,4
0,2
0,2
0,2
0,2
0,2
0.0
P=0.031
O.D
P=0.024
D.D
P=0.009
0.0
P=0.045
0,0
P=0.040
,00
20,00
40,00
60,00
80,00
100,00
00
20,00
40,00
60,00
80,00
100,00
00
20,00
40,00
60,00
00,00
100,00
00
20,00
40,00
60,00
80,00
100,00
,00
20,00
40,00
60,00
80,00
100,00
Time (months)
Time (months)
EFS
Time (months)
Time (months)
| Hub gene | AUC | p-value | 95% CI |
|---|---|---|---|
| ISCU | 0.777 | 0.009 | 0.61-0.943 |
| PTPRC | 0.748 | 0.018 | 0.559-0.938 |
| KDM3A | 0.778 | 0.008 | 0.61-0.947 |
| CD48 | 0.788 | 0.006 | 0.609-0.966 |
| N4BP2 | 0.746 | 0.019 | 0.581-0.912 |
| HSPB6 | 0.633 | 0.207 | 0.417-0.848 |
| JUN | 0.705 | 0.052 | 0.503-0.906 |
| PFR1 | 0.765 | 0.012 | 0.588-0.942 |
| CSNK1D | 0.782 | 0.007 | 0.609-0.955 |
| APBB1IP | 0.672 | 0.101 | 0.482-0.863 |
| PRKAB2 | 0.769 | 0.011 | 0.608-0.93 |
| STK17B | 0.723 | 0.033 | 0.539-0.908 |
| ITGAL | 0.65 | 0.155 | 0.456-0.843 |
| KLK15 | 0.739 | 0.023 | 0.563-0.914 |
| HIST1H3J | 0.597 | 0.358 | 0.374-0.819 |
AUC: Area under the curve; CI: Confidence interval.
Examples of these hub genes include PRKAB2, CSNK1D, and JUN, pivotal regulators of the cell cycle and co-expressed in M1 in our anal- ysis. PRKAB2 plays a central role in regulating cellular metabolism and response to energy stress. Its activation triggers a cascade of events that impact metabolic pathways and gene expression critical for cell prolif- eration. PRKAB2 is dysregulated in various cancers including breast, endometrial, lung, colon, and ovarian cancers. In these tumors, PRKAB2
dysregulation promotes cell survival, uncontrolled proliferation, and resistance to apoptosis. Moreover, AMPK activation induced by met- formin, an AMPK activator, is associated with antitumor effects in numerous types of cancer, including lung, colon, and ovarian cancers, suggesting that PRKAB2 is potentially involved in these contexts (Li et al., 2021; Chen et al., 2021; Sheng et al., 2018; Oliveira et al., 2003; Shi et al., 2022a). In line with this, downregulated PRKAB2 was asso- ciated with lower EFS and overall survival (OS) in both COG and our cohort. Remarkably, our group had already described that PRKAB2 is specific and sensitive for predicting the clinical outcomes of pediatric ACT patients (Xavier et al., 2024).
CSNK1D, a gene encoding a serine/threonine kinase, plays a multi- faceted role in cellular dynamics, including DNA replication and repair. Its high expression is associated with more aggressive tumorigenic fea- tures and lower patient survival in various cancers, including glioblas- toma, hepatocellular carcinoma, and prostrate, breast, and colon cancers (Liu et al., 2022; Qi et al., 2023; Wang et al., 2023a; Bar et al., 2018). Overexpression of CSNK1D is also associated with higher somatic mutation frequencies of TP53 in human cancers (Wang et al., 2023a), which was found in 87% of our pediatric patients. Similarly, here lower expression of CSNK1D was associated with better survival in pediatric ACTs. CSNK1D phosphorylates key proteins such as beta-catenin, thereby affecting cell cycle progression and differentiation pathways, which could explain its association with cancer progression and unfa- vorable clinical outcomes. JUN, a nuclear transcription factor that is part of the AP-1 complex, wields authority over genes governing cell proliferation, differentiation, and apoptosis. In our study, high expres- sion of JUN was correlated with poorer EFS in pediatric ACTs. Consistent
(-)-CAMPHOR
(RS)-ROSCOVITINE
2-MERCAPTOPYRIMIDINE
Adapalene
AEG3482
ALSTERPAULLONE
AMINEPTINE
ANTHRACENE-9-CARBOXYLIC-ACID
ATOMOXETINE-HYDROCHLORIDE
AZD-1152-HQPA
AZELASTINE-HYDROCHLORIDE
AZX-100
BAY-888
BENZENETHIOL
BENZO[B]FLUORANTHENE
BMS-345541
BRUCEANTIN
CSNK1D
BUPROPION-HYDROCHLORIDE
BUTINOLINE
CARBOXYMETHYL-TRIMETHYL-ARSONIUM
CC-90001
CENISERTIB
CHEMBL225519
CHEMBL261454
CHEMBL275260
CHEMBL375293
CHEMBL477052
CINNARIZINE
CIPROFIBRATE
CK1-IN-1
CLOFIBRATE
CLOTRIMAZOLE
COLCHICINE
CRIDANIMOD
HSPB6
CUPRIC-CHLORIDE
CYC-116
DIPHENHYDRAMINE-HYDROCHLORIDE
D4476
DOVITINIB
FENOFIBRATE
GEFITINIB
GEMFIBROZIL
GW441756X
Halicin
HOLACANTHONE
C261
LORASERTIB
IOX1
IQ-3
IRISOLIDONE
SOLIQUIRITIGENIN
JIB-04
JNJ-6204
UNK-Inhibitor-VII
LINIFANIB
LIPOIC-ACID-ALPHA
ongdaysin
MECHLORETHAMINE-HYDROCHLORIDE
METHIMAZOLE
MU1742
JUN
NAFRONYL-OXALATE
NEOCHAMAEJASMIN-A
PATULIN
PF-5006739
PF-670462-dihydrochloride
QUINAPRIL-HYDROCHLORIDE
R-1487
RETINYLRETINOATE
ROTENONE
SANGIVAMYCIN
SB-203580
SB-220025
SB-242235
SERGEOLIDE
SERTRALINE
SODIUM-SELENITE
SP-600125
SP600125
SR-1277
SR3029
T-5224
TA-01
TAK-715
TRIFLUPROMAZINE-HYDROCHLORIDE
TROPISETRON
Vafidemstat
VINBLASTINE-SULFATE
VINORELBINE-TARTRATE
KDM3A
Drug
Gene
with our findings, high expression of this gene is also associated with tumor progression and drug resistance in hepatocellular carcinoma (Xiang et al., 2019) and tumor growth and development in xenograft models of prostate cancer (Niu et al., 2016). Additionally, silencing of this gene in ovarian cancer cell lines reduces the migratory and prolif- erative capacity of these cells (Liu et al., 2020; Xiaohua et al., 2022).
Interestingly, many clinically significant hub genes identified in M2, including PFR1, CD48, PTPRC, and ITGAL, regulate tumor immunity (Stigliani et al., 2015; He et al., 2023; Wang et al., 2021; Park et al., 2022; Li et al., 2022, 2023; Wu et al., 2022; Zhang et al., 2022). On the basis of our analysis, these genes were downregulated in unfavorable pediatric ACTs. Indeed, low expression levels of other immunologic markers or related genes have been reported in both pediatric and adult cases of ACTs, suggesting a poor immunogenic phenotype (Pinto et al., 2016; Leite et al., 2014; Zheng et al., 2016).
PFR1 encodes perforin 1, a protein secreted by cytotoxic T and nat- ural killer cells to mediate elimination of infected and cancerous cells. In accordance with our findings, low expression of PFR1 is associated with lower EFS in adult ACTs, whereas patients with higher expression of this gene exhibited better OS (Roufas et al., 2018).
Regarding CD48, it is dysregulated in several types of cancer, including hepatocellular carcinoma (Wang et al., 2021), lymphomas (Chiba et al., 2022) and cervical (Yue et al., 2024) and non-small cell lung (Park et al., 2022) cancers. In breast cancer and lung adenocarci- noma, the expression of PTPRC is associated with tumor-infiltrating immune cells, prognosis, and sensitivity to drugs (Li et al., 2023; Wei et al., 2021).
In the context of cancers, ITGAL plays a crucial role in modulating the immune microenvironment. In patients with gastric cancer, positive regulation of this gene and its higher expression are consistently
associated with unfavorable outcomes, characterized by more aggres- sive disease progression (Zhang et al., 2022). In addition, increased expression of ITGAL is correlated with poorer outcomes in acute myeloid leukemia (AML) (Li et al., 2022). Conversely, in cases of non-small cell lung cancer, negatively regulated ITGAL is associated with unfavorable prognosis and increased tumor aggressiveness (Wang et al., 2023b), which is aligned with our findings. In our analysis, lower expression of ITGAL was associated with shorter EFS in pediatric ACTs.
APBB1IP and KDM3A emerged as clinically significant hub genes in pediatric ACTs. APBB1IP is notably associated with immune infiltrates. In a pan-cancer analysis, this gene proved a promising prognostic biomarker, and its positive regulation is primarily correlated with in- filtrations of CD8 T cells and NK cells (Ge et al., 2020). However, here, high expression of APBB1IP and shorter EFS were correlated in pediatric ACTs.
Concerning KDM3A expression, this gene is upregulated in various types of cancer, including colorectal carcinoma (where it acts on tumor cell migration and invasion) (Wang et al., 2019; Liu et al., 2019) and pancreatic (Hou et al., 2021) and bladder (Wang et al., 2020) cancers. In our analysis, positively regulated KDM3A was associated with worse ACT prognoses.
We identified another interesting hub gene, STK17B. In M4, posi- tively regulated STK17B was associated with lower EFS in pediatric ACTs. In accordance with our findings, this gene is positively regulated in breast cancer, and its aberrant expression is associated with tumor progression. Besides, attenuated expression of STK17B decreases the tumorigenic capacity in in vitro and in vivo assays (Jiang et al., 2021). Therefore, elevated expression of STK17B may be a potential prognostic indicator, as observed in pediatric ACTs. Additionally, researchers are currently investigating specific inhibitors of STK17B to modulate cellular activity and to inhibit cancer cell proliferation and dissemination.
KLK15, which we identified as a hub gene with prognostic value in M4, belongs to the human kallikrein gene family. Interestingly, many kallikreins are differentially expressed in endocrine-related malig- nancies (Diamandis and Yousef, 2001; Yousef and Diamandis, 2001). In our analysis, high expression of KLK15 was correlated with improved EFS in pediatric ACTs. Furthermore, high expression of KLK15 is asso- ciated with better survival and is an independent favorable prognostic marker of breast cancer (Yousef et al., 2002).
Heat shock proteins, such as HSPB6, comprise a large family of proteins that participate in the regulation of cellular responses to stress, thereby protecting the cell from both extra- and intracellular insults (Wu et al., 2023). In our analysis, positively regulated HSPB6 was correlated with decreased EFS in pediatric ACTs. In ovarian and breast cancers, expression of HSPB6 has been negatively correlated with the degree of malignancy and recurrence-free survival (Qiao et al., 2014; Yang et al., 2022).
Some genes displaying prognostic value in our analysis have not been extensively investigated in cancers. This is mainly true for HIST1H3J, N4BP2, and ISCU. For instance, HIST1H3J is upregulated in high-grade glioma compared to lower-grade gliomas (grades 1, 2, and 3) (Hervás-Corpió et al., 2021). In addition, in laryngeal squamous cell carcinoma, high expression of HIST1H3J is associated with lower OS (Xiang et al., 2021). As for N4BP2, we verified that its high expression was correlated with shorter EFS in pediatric ACTs.
ISCU, which is involved in iron metabolism, is a target of p53, so it regulates the intracellular iron pool. Reduced expression of ISCU is significantly associated with TP53 mutation, as evidenced by its notably lower expression in hepatocellular carcinoma tissues with mutant TP53 as compared to hepatocellular carcinoma tissues with wild type TP53 (Funauchi et al., 2015). In melanoma and colorectal cancer, expression of ISCU is lower in tumors compared to normal samples (Suwei et al., 2022; Shi et al., 2022b). Moreover, expression of ISCU, along with expression of seven other genes, has been used as a prognostic signature in bladder cancer, where its expression is considered a low-risk factor (Yi
et al., 2021). In agreement, our data indicated that patients with lower expression of ISCU had shorter EFS.
To the best of our knowledge, in silico studies on pediatric ACTs have not been conducted, but some in silico studies comparing adult ACTs and normal tissue have been published (Xia et al., 2019; Ma et al., 2021; Yin et al., 2023; Guo et al., 2020). These studies found some common DEGs and hub genes, as well as several different ones. These DEGs and hub genes differed from the genes we found here, suggesting that different DEGs and hub genes could be involved in pediatric ACTs, especially in the TP53 p.R337H series.
By using the DGIdb and DrugBank databases, we found 88 potential drugs for four out of the seven overexpressed hub genes associated with poor survival in pediatric ACTs (HSPB6, JUN, KDM3A, and CSNK1D). These drugs are related to important pathways and genes in cancers and include inhibitors of histone demethylase (IOX1 and JIB-04), aurora kinases (AZD-1152-HQPA, Cenisertib, CYC-116, Ilorasertinib), p38 MAPK (SB220025, SB24223, SB203580), EGFR inhibitor (Gefitinib), VEGF/PDGF (Linifanib, Ilorasertinib, Dovitinib), CDKs (Roscovitine), and casein kinase (SR3029, MU1742, PF-5006739, and IC261). SP600125, which inhibits Jun N-terminal kinase, selectively kills p53- deficient cancer cells (Jema et al., 2012). Some clinical trials are ongoing to test these potential drugs against human cancers: Roscovitine is being tested against solid tumors and lung and breast cancers; AZD- 1522 is being tested against AML, Genfitinib is being tested against lung, head and neck, and breast cancers; Linifanib is being tested against refractory solid tumors and metastatic tumors; Ilorasertib is being tested against solid metastatic, refractory hematological tumors; CYC-116 is being tested against solid tumors; and Dovitinib is being tested against solid refractory and advanced adrenal tumors (https://clinicaltrials. gov/).
However, our study has some limitations. Notwithstanding the fact that pediatric ACTS are rare, the number of patients analyzed and validated in the independent cohort (GSE76019) was relatively small. The high prevalence of the TP53 p.R337H mutation in our cohort may also be a potential bias when generalizing our results in other cohorts without this mutation.
In summary, our results cast light on the potential role of novel genes that might be involved in ACT progression in children. These genes hold promise not only as prognostic markers, but also as potential therapeutic targets for future studies on pediatric ACTs. In this scenario, our tran- scriptomic analysis opens opportunities to improve the prognostic classification of pediatric ACTs and to understand this rare disease.
CRediT authorship contribution statement
Luciana Chain Veronez: Writing - review & editing, Writing - original draft, Methodology, Formal analysis, Conceptualization. Alcides Euzebio Tavares Xavier: Writing - review & editing, Writing - original draft, Methodology, Formal analysis, Conceptualization. Luiz Fernando Nagano: Methodology, Formal analysis. Carolina Alves Pereira Correa: Methodology. Kleiton Silva Borges: Writing - review & editing, Methodology, Formal analysis. Paula Santos: Methodology, Formal analysis. Mirella Baroni: Methodology. Rosane de Paula Silva Queiroz: Methodology, Formal analysis, Data curation. Sonir Roberto Rauber Antonini: Writing - review & editing, Data curation. José Andres Yunes: Writing - review & editing, Data curation. Silvia Regina Brandalise: Writing - review & editing, Data curation. Carlos Augusto Fernandes Molina: Data curation. Emilia Modolo Pinto: Writing - review & editing, Data curation. Elvis Terci Valera: Writing - review & editing, Data curation. Luiz Gonzaga Tone: Writing - review & editing, Funding acquisition, Conceptualization. Carlos Alberto Scrideli: Writing - review & editing, Writing - original draft, Resources, Funding acquisition, Formal analysis, Data curation, Conceptualization.
Funding
This work was supported by the Brazilian Public Research Agencies Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, Grant Number, 2014/20341-0) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, Grant Number 406484/2022-8 - INCT BioOncoPed).
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi. org/10.1016/j.mce.2024.112383.
Data availability
Data will be made available on request.
References
Bar, I., Merhi, A., Larbanoix, L., Constant, M., Haussy, S., Laurent, S., Canon, J.L., Delrée, P., 2018. Silencing of casein kinase 1 delta reduces migration and metastasis of triple negative breast cancer cells. Oncotarget 9, 30821-30836. https://doi.org/ 10.18632/oncotarget.25738.
Brondani, V.B., Fragoso, M.C.B.V., 2020. Pediatric adrenocortical tumor-review and management update. Curr. Opin. Endocrinol. Diabetes Obes. 27, 177-186. https:// doi.org/10.1097/MED.0000000000000540.
Chen, X., Zhang, W., Zhu, H., Lin, F., 2021. Development and validation of a 5-gene autophagy-based prognostic index in endometrial carcinoma. Med. Sci. Monit. 27, 1-13. https://doi.org/10.12659/MSM.928949.
Chiba, M., Shimono, J., Ishio, T., Takei, N., Kasahara, K., Ogasawara, R., Ara, T., Goto, H., Izumiyama, K., Otsuguro, S., et al., 2022. Genome-wide CRISPR screens identify CD48 defining susceptibility to NK cytotoxicity in peripheral T-cell lymphomas. Blood 140, 1951-1963. https://doi.org/10.1182/blood.2022015646.
Diamandis, E.P., Yousef, G.M., 2001. Human tissue kallikrein gene family: a rich source of novel disease biomarkers. Expert Rev. Mol. Diagn. 1, 182-190. https://doi.org/ 10.1586/14737159.1.2.182.
Ewels, P., Magnusson, M., Lundin, S., Käller, M., 2016. MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics 32, 3047-3048. https://doi.org/10.1093/bioinformatics/btw354.
Figueiredo, B.C., Sandrini, R., Zambetti, G.P., Pereira, R.M., Cheng, C., Liu, W., Lacerda, L., Pianovski, M.A., Michalkiewicz, E., Jenkins, J., et al., 2006. Penetrance of adrenocortical tumours associated with the germline TP53 R337H mutation. J. Med. Genet. 43, 91-96. https://doi.org/10.1136/jmg.2004.030551.
Freshour, S.L., Kiwala, S., Cotto, K.C., Coffman, A.C., McMichael, J.F., Song, J.J., Griffith, M., Griffith, O.L., Wagner, A.H., 2021. Integration of the drug-gene interaction database (DGIdb 4.0) with open crowdsource efforts. Nucleic Acids Res. 49, D1144-D1151. https://doi.org/10.1093/nar/gkaa1084.
Funauchi, Y., Tanikawa, C., Yi Lo, P., et al., 2015. Regulation of iron homeostasis by the p53-ISCU pathway. Sci. Rep. 5, 16497. https://doi.org/10.1038/srep16497, 16497.
Ge, Q., Li, G., Chen, J., Song, J., Cai, G., He, Y., Zhang, X., Liang, H., Ding, Z., Zhang, B., 2020. Immunological role and prognostic value of APBB1IP in pan-cancer analysis. J. Cancer 12, 595-610. https://doi.org/10.7150/jca.50785.
Gomes, D.C., Leal, L.F., Mermejo, L.M., Scrideli, C.A., Martinelli, C.E., Fragoso, M.C.B.V., Latronico, A.C., Tone, L.G., Tucci, S., Yunes, J.A., et al., 2014. Sonic Hedgehog signaling is active in human adrenal cortex development and deregulated in adrenocortical tumors. J. Clin. Endocrinol. Metab. 99, 209-1216. https://doi.org/ 10.1210/jc.2013-4098.
Gu, Z., Eils, R., Schlesner, M., 2016. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics 32, 2847-2849. https://doi.org/ 10.1093/bioinformatics/btw313.
Guo, J., Gu, Y., Ma, X., Zhang, L., Li, H., Yan, Z., Han, Y., Xie, L., Guo, X., 2020. Identification of hub genes and pathways in adrenocortical carcinoma by integrated bioinformatic analysis. J. Cell Mol. Med. 24 (8), 4428-4438. https://doi.org/ 10.1111/jcmm.15102. Epub 2020 Mar 8. PMID: 32147961; PMCID: PMC7176852.
Gupta, N., Rivera, M., Novotny, P., Rodriguez, V., Bancos, I., Lteif, A., 2018. Adrenocortical carcinoma in children: a clinicopathological analysis of 41 patients at the mayo clinic from 1950 to 2017. Horm. Res. Paediatr. 90 (1), 8-18. https://doi. org/10.1159/000488855.
He, M., Yu, J., Chen, S., Mi, H.A., 2023. Systematic immune and prognostic analysis of CD48 interaction with tumor microenvironment in pan-cancer. Int. J. Gen. Med. 16, 5255-5269. https://doi.org/10.2147/ijgm.s431696.
Hou, X., Li, Q., Yang, L., Yang, Z., He, J., Li, Q., Li, D., 2021. KDM1A and KDM3A promote tumor growth by upregulating cell cycle-associated genes in pancreatic cancer. Exp. Biol. Med. 246, 1869-1883. https://doi.org/10.1177/ 15353702211023473.
Ilanchezhian, M., Varghese, D.G., Glod, J.W., Reilly, K.M., Widemann, B.C., Pommier, Y., Kaplan, R.N., Del Rivero, J., 2022. Pediatric adrenocortical carcinoma. Front. Endocrinol. 13, 1-9. https://doi.org/10.3389/fendo.2022.961650.
Jiang, L., Zhou, J., Zhao, S., Wang, X., Chen, Y., 2021. STK17B promotes the progression of ovarian cancer. Ann. Transl. Med. 9, 475. https://doi.org/10.21037/atm-21-601, 475.
Kim, D., Langmead, B., Salzberg, S.L., 2015. HISAT: a fast spliced aligner with low memory requirements. Nat. Methods 12, 357-360. https://doi.org/10.1038/ nmeth.3317.
Latronico, A.C., Pinto, E.M., Domenice, S., Fragoso, M.C., Martin, R.M., Zerbini, M.C., Lucon, A.M., Mendonca, B.B., 2001. An inherited mutation outside the highly conserved DNA-binding domain of the P53 tumor suppressor protein in children and adults with sporadic adrenocortical tumors. J. Clin. Endocrinol. Metab. 86 (10), 4970-4973. https://doi.org/10.1210/jcem.86.10.7957.
Leal, L.F., Mermejo, L.M., Ramalho, L.Z., Martinelli, C.E., Yunes, J.A., Seidinger, A.L., Mastellaro, M.J., Cardinalli, I.A., Brandalise, S.R., Moreira, A.C., et al., 2011. Wnt/ B-Catenin pathway deregulation in childhood adrenocortical tumors. J. Clin. Endocrinol. Metab. 96, 3106-3114. https://doi.org/10.1210/jc.2011-0363.
Leite, F.A., Lira, R.C.P., Fedatto, P.F., Antonini, S.R.R., Martinelli, C.E., de Castro, M., Neder, L., Ramalho, L.N.Z., Tucci, S., Mastelaro, M.J., et al., 2014. Low expression of HLA-DRA, HLA-DPA1, and HLA-DPB1 is associated with poor prognosis in pediatric adrenocortical tumors (ACT). Pediatr. Blood Cancer 61, 1940-1948. https://doi.org/ 10.1002/pbc.25118.
Li, X., Cao, Y., Yu, X., Jin, F., Li, Y., 2021. A novel autophagy-related genes prognostic risk model and validation of autophagy-related oncogene VPS35 in breast cancer. Cancer Cell Int. 21, 1-14. https://doi.org/10.1186/s12935-021-01970-4.
Li, R., Wu, X., Xue, K., Li, J., 2022. ITGAL infers adverse prognosis and correlates with immunity in acute myeloid leukemia. Cancer Cell Int. 22, 1-11. https://doi.org/ 10.1186/s12935-022-02684-x.
Li, P., Wang, W., Wang, S., Cao, G., Pan, T., Huang, Y., Wan, H., Zhang, W., Huang, Y., Jin, H., et al., 2023. PTPRC promoted CD8+ T cell mediated tumor immunity and drug sensitivity in breast cancer: based on pan-cancer analysis and artificial intelligence modeling of immunogenic cell death-based drug sensitivity stratification. Front. Immunol. 14, 1-19. https://doi.org/10.3389/ fimmu.2023.1145481.
Liao, Y., Smyth, G.K., Shi, W., 2014. FeatureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 30, 923-930. https://doi.org/10.1093/bioinformatics/btt656.
Lira, R.C.P., Fedatto, P.F., Santos, D., Antonio, M., Leal, L.F., Martinelli, C.E., Castro, M. De, Tucci, S., Neder, L., Ramalho, L., et al., 2016. IGF2 and IGF1R in pediatric adrenocortical tumors : roles in metastasis and steroidogenesis. Endocr. Relat. Cancer 23 (6), 481-493. https://doi.org/10.1530/ERC-15-0426.
Liu, J., Liang, T., Zhangsun, W., 2019. KDM3A is associated with tumor metastasis and modulates colorectal cancer cell migration and invasion. Int. J. Biol. Macromol. 126, 318-325. https://doi.org/10.1016/j.ijbiomac.2018.12.105.
Liu, L., Ning, Y., Yi, J., Yuan, J., Fang, W., Lin, Z., Zeng, Z., 2020. MiR-6089/MYH9/ ß-Catenin/c-Jun negative feedback loop inhibits ovarian cancer carcinogenesis and progression. Biomed. Pharmacother. 125, 109865. https://doi.org/10.1016/j. biopha.2020.109865.
Liu, Y., He, W., Guo, Y., Qu, S., Yao, F., Liu, J., Chai, J., Yang, Y., Xu, T., Liu, Y., et al., 2022. CSNK1D is associated with stemness and invasiveness in glioblastoma. Pathol. Res. Pract. 240, 154187. https://doi.org/10.1016/j.prp.2022.154187.
Love, M.I., Huber, W., Anders, S., 2014. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, 1-21. https://doi.org/ 10.1186/s13059-014-0550-8.
Ma, C., Xiong, J., Su, H., Li, H., 2021. The underlying molecular mechanism and drugs for treatment in adrenal cortical carcinoma. Int. J. Med. Sci. 18 (13), 3026-3038. https://doi.org/10.7150/ijms.60261. PMID: 34220331; PMCID: PMC8241782.
Mete, O., Erickson, L.A., Juhlin, C.C., de Krijger, R.R., Sasano, H., Volante, M., Papotti, M.G., 2022. Overview of the 2022 WHO classification of adrenal cortical tumors. Endocr. Pathol. 33 (1), 155-196. https://doi.org/10.1007/s12022-022- 09710-8.
Michalkiewicz, E., Sandrini, R., Figueiredo, B., Miranda, E.C., Caran, E., Oliveira- Filho, A.G., Marques, R., Pianovski, M.A., Lacerda, L., Cristofani, L.M., et al., 2004. Clinical and outcome characteristics of children with adrenocortical tumors: a report from the International Pediatric Adrenocortical Tumor Registry. J. Clin. Oncol. 22 (5), 838-845. https://doi.org/10.1200/JCO.2004.08.085.
Niu, Y.N., Wang, K., Jin, S., Fan, D.D., Wang, M.S., Xing, N.Z., Xia, S.J., 2016. The intriguing role of fibroblasts and C-Jun in the chemopreventive and therapeutic effect of finasteride on xenograft models of prostate cancer. Asian J. Androl. 18, 913-919. https://doi.org/10.4103/1008-682X.167714.
Oliveira, S.M.J., Ehtisham, J., Redwood, C.S., Ostman-Smith, I., Blair, E.M., Watkins, H., 2003. Mutation analysis of AMP-activated protein kinase subunits in inherited cardiomyopathies: implications for kinase function and disease pathogenesis. J. Mol. Cell. Cardiol. 35, 1251-1255. https://doi.org/10.1016/S0022-2828(03)00237-2.
Park, E.J., Jun, H.W., Na, I.H., Lee, H.K., Yun, J., Kim, H.S., Kim, Y., Hong, J.T., Han, S. B., 2022. CD48-Expressing non-small-cell lung cancer cells are susceptible to natural killer cell-mediated cytotoxicity. Arch Pharm. Res. (Seoul) 45, 1-10. https://doi. org/10.1007/s12272-021-01365-z.
Pinto, E.M., Rodriguez-Galindo, C., Choi, J.K., Pounds, S., Liu, Z., Neale, G., Finkelstein, D., Hicks, J.M., Pappo, A.S., Figueiredo, B.C., et al., 2016. Prognostic significance of major histocompatibility complex class II expression in pediatric
adrenocortical tumors: a st. Jude and Children’s Oncology group study. Clin. Cancer Res. 22, 6247-6255. https://doi.org/10.1158/1078-0432.CCR-15-2738.
Pinto, E.M., Zambetti, G.P., Rodriguez-Galindo, C., 2020. Pediatric adrenocortical tumours. Best Pract. Res. Clin. Endocrinol. Metab 34, 101448. https://doi.org/ 10.1016/j.beem.2020.101448.
Qi, F., Shen, W., Wei, X., Cheng, Y., Xu, F., Zheng, Y., Li, L., Qin, C., Li, X., 2023. CSNK1D-Mediated phosphorylation of HNRNPA2B1 induces MiR-25-3p/MiR-93-5p maturation to promote prostate cancer cell proliferation and migration through M6A-dependent manner. Cell. Mol. Life Sci. 80, 1-18. https://doi.org/10.1007/ s00018-023-04798-5.
Qiao, N., Zhu, Y., Li, H., Qu, Z., Xiao, Z., 2014. Expression of heat shock protein 20 inversely correlated with tumor progression in patients with ovarian cancer. Eur. J. Gynaecol. Oncol. 35, 576-579. https://doi.org/10.12892/ejgot25602014.
Riedmeier, M., Decarolis, B., Haubitz, I., Müller, S., Uttinger, K., Börner, K., Reibetanz, J., Wiegering, A., Härtel, C., Schlegel, P.G., et al., 2021. Adrenocortical carcinoma in childhood: a systematic review. Cancers 13, 1-23. https://doi.org/10.3390/ cancers13215266.
Rodriguez-Galindo, C., Krailo, M.D., Pinto, E.M., Pashankar, F., Weldon, C.B., Huang, L., Caran, E.M., Hicks, J., McCarville, M.B., Malkin, D., et al., 2021a. Treatment of pediatric adrenocortical carcinoma with surgery, retroperitoneal lymph node dissection, and chemotherapy: the Children’s Oncology group ARAR0332 protocol. J. Clin. Oncol. 39, 2463-2473. https://doi.org/10.1200/JCO.20.02871.
Rodriguez-Galindo, C., Krailo, M.D., Pinto, E.M., Pashankar, F., Weldon, C.B., Huang, L., Caran, E.M., Hicks, J., McCarville, M.B., Malkin, D., et al., 2021b. Treatment of pediatric adrenocortical carcinoma with surgery, retroperitoneal lymph node dissection, and chemotherapy: the children’s Oncology group ARAR0332 protocol. J. Clin. Oncol. 39 (22), 2463-2473. https://doi.org/10.1200/JCO.20.02871.
Roufas, C., Chasiotis, D., Makris, A., Efstathiades, C., Dimopoulos, C., Zaravinos, A., 2018. The expression and prognostic impact of immune cytolytic activity-related markers in human malignancies: a comprehensive meta-analysis. Front. Oncol. 8, 1-18. https://doi.org/10.3389/fonc.2018.00027.
Russo, P.S.T., Ferreira, G.R., Cardozo, L.E., Bürger, M.C., Arias-Carrasco, R., Maruyama, S.R., Hirata, T.D.C., Lima, D.S., Passos, F.M., Fukutani, K.F., et al., 2018. CEMiTool: a bioconductor package for performing comprehensive modular Co- expression analyses. BMC Bioinf. 19, 1-13. https://doi.org/10.1186/s12859-018- 2053-1.
Sandrini, F., Villani, D.P., Tucci, S., Moreira, A.C., de Castro, M., Elias, L.L., 2005. Inheritance of R337H p53 gene mutation in children with sporadic adrenocortical tumor. Horm. Metab. Res. 37 (4), 231-235. https://doi.org/10.1055/s-2005- 861373.
Sheng, M., Dong, Z., Xie, Y., 2018. Identification of tumor-educated platelet biomarkers of non-small-cell lung cancer. OncoTargets Ther. 11, 8143-8151. https://doi.org/ 10.2147/OTT.S177384.
Shi, F., Tang, Z., Jiang, S., Xiong, Z., Zhang, W., Li, Y., Lin, H., Luo, Z., Ying, Y., 2022a. AMP-activated protein kinase § 1 or § 2 deletion enhances colon cancer cell growth and tumorigenesis. Acta Biochim. Biophys. Sin. 54, 1140-1147. https://doi.org/ 10.3724/abbs.2022086.
Shi, W.K., Liu, Y.X., Qiu, X.Y., Zhou, J.Y., Zhou, J.L., Lin, G., 2022b. Le construction and validation of a novel ferroptosis-related gene signature predictive model in rectal cancer. BMC Genom. 23, 1-18. https://doi.org/10.1186/s12864-022-08996-6.
Stigliani, S., Croce, M., Morandi, F., Scaruffi, P., Rigo, V., Carlini, B., Manzitti, C., Gigliotti, A.R., Tonini, G.P., Pistoia, V., et al., 2015. Expression of FOXP3, CD14, and ARG1 in neuroblastoma tumor tissue from high-risk patients predicts event-free and overall survival. BioMed Res. Int. 2015, 1-10. https://doi.org/10.1155/2015/ 347867.
Suwei, D., Zhen, L., Zhimin, L., Mei, L., Jianping, K., Zhuohui, P., Yanbin, X., Xiang, M., 2022. Hypoxia modulates melanoma cells proliferation and apoptosis via MiRNA- 210/ISCU/ROS signaling. Bull. Exp. Biol. Med. 173, 645-650. https://doi.org/ 10.1007/s10517-022-05605-0.
Veronez, L.C., Fedatto, P.F., Correa, C.A.P., Lira, R.C.P., Baroni, M., da Silva, K.R., Santos, P., Antonio, D.S.M., Queiroz, R.P.S., Antonini, S.R.R., Tucci Jr., S., et al., 2022. MicroRNA expression profile predicts prognosis of pediatric adrenocortical tumors. Pediatr. Blood Cancer 69 (7), e29553. https://doi.org/10.1002/pbc.29553.
Wang, Z., Yang, X., Liu, C., Li, X., Zhang, B., Wang, B., Zhang, Y., Song, C., Zhang, T., Liu, M., et al., 2019. Acetylation of PHF5A modulates stress responses and colorectal carcinogenesis through alternative splicing-mediated upregulation of KDM3A. Mol. Cell 74, 1250-1263.e6. https://doi.org/10.1016/j.molcel.2019.04.009.
Wang, T., Luo, R., Li, W., Yan, H., Xie, S., Xiao, W., Wang, Y., Chen, B., Bai, P., Xing, J., 2020. Dihydroartemisinin suppresses bladder cancer cell invasion and migration by regulating KDM3A and P21. J. Cancer 11, 1115-1124. https://doi.org/10.7150/ jca.36174.
Wang, Z., He, L., Li, W., Xu, C., Zhang, J., Wang, D., Dou, K., Zhuang, R., Jin, B., Zhang, W., et al., 2021. GDF15 induces immunosuppression via CD48 on regulatory
T cells in hepatocellular carcinoma. J. Immunother. Cancer 9, 1-16. https://doi.org/ 10.1136/jitc-2021-002787.
Wang, J., Hu, B., Wang, W., 2023a. Prognostic value and immunological role of CSNK1D in human cancers. Aging (Albany. NY) 15, 8948-8975. https://doi.org/10.18632/ aging.205009.
Wang, Q., Xiao, G.J., Li, N., Jiang, X., Li, C., 2023b. LncRNA PCBP1-AS1 mediated downregulation of ITGAL as a prognostic biomarker in lung adenocarcinoma. Aging (Albany. NY) 15, 4510-4523. https://doi.org/10.18632/aging.204756.
Wei, J., Fang, D., Zhou, W., 2021. CCR2 and PTPRC are regulators of tumor microenvironment and potential prognostic biomarkers of lung adenocarcinoma. Ann. Transl. Med. 9, 1419. https://doi.org/10.21037/atm-21-3301, 1419.
Wishart, D.S., Feunang, Y.D., Marcu, A., Guo, A.C., Liang, K., Vázquez-Fresno, R., Sajed, T., Johnson, D., Li, C., Karu, N., et al., 2018. Hmdb 4.0: the human metabolome database for 2018. Nucleic Acids Res. 46, D608-D617. https://doi.org/ 10.1093/nar/gkx1089.
Wu, Z., Zhou, J., Xiao, Y., Ming, J., Zhou, J., Dong, F., Zhou, X., Xu, Z., Zhao, X., Lei, P., et al., 2022. CD20+CD22+ADAM28+ B cells in tertiary lymphoid structures promote immunotherapy response. Front. Immunol. 13, 1-18. https://doi.org/ 10.3389/fimmu.2022.865596.
Wu, Y., Zhao, J., Tian, Y., Jin, H., 2023. Cellular functions of heat shock protein 20 (HSPB6) in cancer: a review. Cell. Signal. 112, 110928. https://doi.org/10.1016/j. cellsig.2023.110928.
Xavier, A.E.T., Veronez, L.C., Nagano, L.F.P., Correa, C.A.P., Baroni, M., Ramos, M.S., Queiroz, R.G.P., Fernandes Molina, C.A., Yunes, J.A., Brandalise, S.R., et al., 2024. Low PRKAB2 expression is associated with poor outcomes in pediatric adrenocortical tumors, and treatment with rottlerin increases the PRKAB2 level and inhibits tumorigenic aspects in the NCI-H295R adrenocortical cancer cell line. Cancers 16, 1094. https://doi.org/10.3390/cancers16061094, 1094.
Xia, W.X., Yu, Q., Li, G.H., Liu, Y.W., Xiao, F.H., Yang, L.Q., Rahman, Z.U., Wang, H.T., Kong, Q.P., 2019. Identification of four hub genes associated with adrenocortical carcinoma progression by WGCNA. PeerJ 7, e6555. https://doi.org/10.7717/ peerj.6555. PMID: 30886771; PMCID: PMC6421058.
Xiang, D.M., Sun, W., Zhou, T., Zhang, C., Cheng, Z., Li, S.C., Jiang, W., Wang, R., Fu, G., Cui, X., et al., 2019. Oncofetal HLF transactivates C-Jun to promote hepatocellular carcinoma development and sorafenib resistance. Gut 68, 1858-1871. https://doi. org/10.1136/gutjnl-2018-317440.
Xiang, F., Li, L., Lin, J., Li, S., Peng, G., 2021. Network pharmacology and bioinformatics methods reveal the mechanism of zao-jiao-ci in the treatment of LSCC. J. Oncol. 2021, 1-15. https://doi.org/10.1155/2021/8862821.
Xiaohua, Zhu, Xie, Y.Y., Huang, W., Chen, Z., Guo, S.Q., 2022. NAP1L1 promotes tumor proliferation through HDGF/C-JUN signaling in ovarian cancer. BMC Cancer 22, 1-10. https://doi.org/10.1186/s12885-022-09356-z.
Yang, Y., Wu, Y., Hou, L., Ge, X., Song, G., Jin, H., 2022. Heat shock protein 20 suppresses breast carcinogenesis by inhibiting the MAPK and AKT signaling pathways. Oncol. Lett. 24, 1-11. https://doi.org/10.3892/ol.2022.13582.
Yi, K., Liu, J.C., Rong, Y., Wang, C., Tang, X., Zhang, X.P., Xiong, Y., Wang, F., 2021. Biological functions and prognostic value of ferroptosis-related genes in bladder cancer. Front. Mol. Biosci. 8, 1-12. https://doi.org/10.3389/fmolb.2021.631152.
Yin, M., Wang, Y., Ren, X., Han, M., Li, S., Liang, R., Wang, G., Gang, X., 2023. Identification of key genes and pathways in adrenocortical carcinoma: evidence from bioinformatic analysis. Front. Endocrinol. 14, 1250033. https://doi.org/10.3389/ fendo.2023.1250033. PMID: 38053725; PMCID: PMC10694291.
Yousef, G.M., Diamandis, E.P., 2001. The new human tissue kallikrein gene family: structure, function, and association to disease. Endocr. Rev. 22, 184-204. https:// doi.org/10.1210/er.22.2.184.
Yousef, G.M., Scorilas, A., Magklara, A., Memari, N., Ponzone, R., Sismondi, P., Biglia, N., Abd Ellatif, M., Diamandis, E.P., 2002. The androgen-regulated gene human kallikrein 15 (KLK15) is an independent and favourable prognostic marker for breast cancer. Br. J. Cancer 87, 1294-1300. https://doi.org/10.1038/sj. bjc.6600590.
Yue, M., Zhuo, Y., Jing, L., Danbo, W., 2024. CD48 suppresses proliferation and migration as an immune-related prognostic signature in the cervical cancer immune microenvironment. Carcinogenesis 45, 57-68. https://doi.org/10.1093/carcin/ bgad039.
Zhang, J., Wang, H., Yuan, C., Wu, J., Xu, J., Chen, S., Zhang, C., He, Y., 2022. ITGAL as a prognostic biomarker correlated with immune infiltrates in gastric cancer. Front. Cell Dev. Biol. 10, 1-15. https://doi.org/10.3389/fcell.2022.808212.
Zheng, S., Cherniack, A.D., Dewal, N., Moffitt, R.A., Danilova, L., Murray, B.A., Lerario, A.M., Else, T., Knijnenburg, T.A., Ciriello, G., et al., 2016. Comprehensive pan-genomic characterization of adrenocortical carcinoma. Cancer Cell 29723-29736. https://doi.org/10.1016/j.ccell.2016.04.002.