Society for Endocrinology
Lamin B1: a novel biomarker in adult and pediatric adrenocortical carcinoma
Yihao Chen1,2,*, Jiahong Chen2,*, Yongcheng Shi2,*, Xiaohui Ling3, Shumin Fang4, Chuanfan Zhong5, Fengping Liu6, Weide Zhong6,7,8, Xuecheng Bi9, Zhong Dong2 and Jianming Lu®1,7
1Department of Andrology, Guangzhou First People’s Hospital, Guangzhou Medical University, Guangzhou, China
2Department of Urology, Huizhou Municipal Central Hospital, Huizhou, Guangdong, China
3Reproductive Medicine Center, Huizhou Municipal Central Hospital, Huizhou, Guangdong, China
4Science Research Center, Huizhou Municipal Central Hospital, Huizhou, Guangdong, China
5Department of Urology, Zhujiang Hospital, Southern Medical University, Guangzhou, Guangdong, China
6State Key Laboratory of Quality Research in Chinese Medicine, Macau University of Science and Technology, Taipa, Macau, China
7Department of Urology, Guangdong Key Laboratory of Clinical Molecular Medicine and Diagnostics, Guangzhou First People’s Hospital, Guangzhou Medical University, Guangzhou, China
8Department of Urology, Minimally Invasive Surgery Center, The First Affiliated Hospital of Guangzhou Medical University, Guangdong Key Laboratory of Urology, Guangzhou Institute of Urology, Guangzhou, Guangdong
9Department of Urology, Guangdong Provincial People’s Hospital (Guangdong Academy of Medical Sciences), Southern Medical University, Guangzhou, China
Correspondence should be addressed to Z Dong or J Lu: hzdongzhong@126.com or louiscfc8@gmail.com
*(Y Chen, J Chen and Y Shi contributed equally to this work)
Abstract
Adrenocortical carcinoma (ACC) is a malignancy with a poor prognosis and high mortality rate. A high tumor mutational burden (TMB) has been found to be associated with poor prognosis in ACC. Thus, exploring ACC biomarkers based on TMB holds significant importance for patient risk stratification. In our research, we utilized weighted gene coexpression network analysis and an assay for transposase-accessible chromatin with high- throughput sequencing to identify genes associated with TMB. Through the comprehensive analysis of various public datasets, Lamin B1 (LMNB1) was identified as a biomarker associated with a high TMB and low chromatin accessibility. Immunohistochemical staining demonstrated high expression of LMNB1 in ACC compared to noncancerous tissues. Functional enrichment analyses revealed that the function of LMNB1 is associated with cell proliferation and division. Furthermore, cell assays suggested that LMNB1 promotes tumor proliferation and invasion. In addition, mutation analysis suggested that the high expression of LMNB1 is associated with TP53 mutations. Additionally, LMNB1 was highly expressed in the vast majority of solid tumors across cancers. In our immune analysis, we discovered that the high expression of LMNB1 might suppress the infiltration of CD8+ T cells in the ACC microenvironment. In summary, LMNB1 is a predictive factor for the poor prognosis of adult and pediatric ACC. Its high expression in ACC is positively associated with high TMB and lower chromatin accessibility, and it promotes ACC cell proliferation and invasion. Therefore, LMNB1 holds promise as a novel biomarker and potential therapeutic target for ACC.
Keywords: adrenocortical carcinoma; ATAC-seq; TMB; LMNB1; prognosis
Introduction
Adrenocortical carcinoma (ACC) is a rare tumor, with global annual incidence rates of 1-2 cases per million adults and 0.2 to 0.3 cases per million children (Else et al. 2014, Paschoalin et al. 2022). The 5-year survival rate is a mere 35%. Despite the low incidence of ACC, the prognosis for most patients remains poor, with many experiencing local recurrence and metastasis even after undergoing radical surgery. Furthermore, besides radical surgery, few treatment options are available. Mitotane is currently the only first-line drug treatment for ACC, offering a modest efficacy rate of 30% and strong side effects, thereby limiting its overall effectiveness (Terzolo et al. 2013). Due to the heterogeneity of ACC, aside from tumor staging, there are no superior prognostic indicators (Else et al. 2014). Given the limitations of staging-based prognosis, there is a pressing need to identify novel biomarkers for risk stratification of patients and the discovery of potential therapeutic targets.
It is well recognized that genomic instability and mutations are hallmarks of cancer (Hanahan 2022). In this context, tumor mutational burden (TMB) - defined as the sum of mutations per million base pairs in a cell - serves as a valuable measure of the quantity of mutations in a cancer, thereby aiding in our understanding of ACC’s heterogeneity (Jardim et al. 2021). Elevated TMB often presages poor prognosis in ACC, though the mechanism remains unclear (Landwehr et al. 2021, Luo et al. 2022, Xu et al. 2022).
From the perspective of chromatin accessibility, mutations in tumor genes can generate new transcription factor binding sites, leading to increased accessibility of the respective chromatin region. This, in turn, allows transcription factors to bind to these sites and initiate transcription, resulting in increased gene expression (Corces et al. 2018). Assay for Transposase-Accessible Chromatin using sequencing (ATAC-seq) is a known method for analyzing the genome-wide accessibility of chromatin (Yan et al. 2020). Hence, using ATAC- seq, we can investigate which coding genes in ACC are affected by the accessibility of noncoding regulatory elements and thereby identify differentially expressed genes under specific TMB conditions. We aim to use TMB and ATAC-seq to aid in the identification of novel biomarkers for ACC.
Materials and methods
Flowchart of the present study
As depicted in Fig. 1, this study combines The Cancer Genome Atlas (TCGA) ACC data with weighted gene coexpression network analysis (WGCNA) and ATAC- seq to screen for TMB-related differential peak genes. Subsequently, using receiver operating characteristic
(ROC) curves and Cox regression, we aimed to identify biomarkers contributing to poor prognosis in ACC. In subsequent validation within four GEO cohorts, we selected and confirmed that LMNB1, which is associated with high TMB, plays a prognostic role in ACC patients. Further exploration of its role in ACC was conducted through gene set enrichment analysis (GSEA), immune infiltration, mutation analysis, pancancer analysis, immunohistochemistry, and in vitro experiments.
Data collection and processing
The outset of our study involved acquiring RNA-seq data, corresponding mutation data, and pertinent clinical details of the TCGA-ACC cohort via the UCSC Xena platform (https://xenabrowser.net/datapages/) (Goldman et al. 2020). With respect to the ATAC-seq data, we harnessed the raw count matrix, the normalized count matrix, and the bigWig files, all sourced from the TCGA. From within the TCGA-ACC-ATAC-seq (Corces et al. 2018) we identified and selected 18 samples, all of which had RNA-seq data and clinical details matched. Based on the median TMB values of 79 samples in TCGA- ACC, we divided the 18 samples into TMB high and low groups for differential accessible peak (DAP) analysis. To serve as validation sets, we retrieved four ACC datasets from the GEO database (refer to Supplementary Table 1, see the section on supplementary materials given at the end of this article): GSE19750 (Demeure et al. 2013), GSE10927 (Giordano et al. 2009),GSE76019 (Pinto et al. 2016), and GSE76021 (Pinto et al. 2016).
Construction of WGCNA
WGCNA is an R package designed for weighted correlation network analysis (Langfelder & Horvath 2008). To identify high TMB related gene modules in ACC, we divided the TCGA-ACC queue into four groups based on TMB values, defined ‘TMB-Top25%’ as the TMB high group, and divided the remaining three groups ‘TMB_25%~50%’, ‘TMB_50%~75%’, ‘TMB_Bottom 25%’ are defined as the TMB low group. To define potential clusters of highly coexpressed genes, we utilized the ‘WGCNA’ (version 1.71) package to construct a gene coexpression network from the gene expression matrix of the TCGA-ACC cohort. To define potential clusters of highly coexpressed genes, we employed the WGCNA (1.71) package to construct a gene coexpression network. This was achieved by utilizing the mRNA expression (n=19,563) of TCGA-ACC, applying the ‘one-step’ method for network construction. In order to satisfy the condition of scale-free networks, we determined the optimal soft-thresholding power (B=12) and transformed the adjacency matrix into a topological overlap matrix (TOM). Furthermore, we calculated the corresponding dissimilarity (1-TOM) and identified TMB-related modules using the dynamic tree cutting method.
WGCNA
:
=
(Survival information 2 RNA expression data 3ATAC-seq data
:
:
TCGA Adrenocortical Carcinoma Cohort
TN5
=
=
Tumor mutation burden -related genes
ATAC-seq
0.4 ROC’s AUC COX Regression
02
1.0
GEO Gene Expression Omnibus
GSE10927 GSE19750 GSE76019 GSE76021
Prognostic Value
Validation
Gene Set Enrichment
LMNB1
-
03
p=2.7e-9
Lamin B1 (LMNB1)
HR=8.31,95C1%(3.70,18.65)
Number at risk
L
S
35
1
14
28
8
1
0
TG OS.time(Months)
=1
36
114
152
log2(Hazard Ratio(95%CI
Kaplan-Meier Univariate Cox regression Multivariate Cox regression
Rank in Ordered Dataset
1 Colony formation assay 2 Transwell assay Cell experiments
1 Mutation landscape 2 CNV landscape Mutation analysis
Immunohistochemistry
Immune landscape
Prognostic gene evaluation
To identify potential prognostic genes, we employed the ‘timeROC’ (0.4) (Blanche et al. 2013) and ‘survival’ (3.5-5) (https://CRAN.R-project.org/package=survival) R packages to ascertain the area under the receiver operating characteristic (ROC) curve (AUC), concurrently performing a univariate Cox regression and Kaplan- Meier (KM) analysis.
Functional enrichment
We assessed the correlation between LMNB1 expression and all genes using the Spearman correlation method. The ‘clusterProfiler’ (4.2.2) (Wu et al. 2021) R package was used for GSEA to identify significantly enriched terms related to Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways, Reactome, and WikiPathways. We prioritized the top ten terms for visualization based on their adjusted P-values (adj. P), ensuring a focused and statistically significant representation.
Multi-omics analysis
The GISTIC 2.0 analysis (www.genepattern.org/) was used to identify genomic alterations (areas of recurrent
amplification and deletion). The ‘maftools’ R package (2.10.05) (Mayakonda et al. 2018) was utilized to calculate the Tumor Mutation Burden (TMB). The ‘IOBR’ R package (0.99.9) (Zeng et al. 2021) was employed to compute the scores for immune cell infiltration. Visualization was achieved using the ‘ComplexHeatmap’ R package (2.10.0) (Gu et al. 2016).
Immunohistochemistry
The ACC (n=6) and adrenal adenoma (n=5) tissue samples used in this study were procured from the Huizhou Municipal Central Hospital and had ethical approval from the Ethical Committee of Huizhou Municipal Central Hospital. Prior to paraffin embedding, specimens were fixed in 4% paraformaldehyde. Each tissue block was cut into 4 um thick sections, treated with a solution of 1% H2O2, and then blocked with nonimmune goat serum. The sections were incubated overnight with primary antibodies at 4℃, followed by a 30-min incubation at room temperature with biotinylated secondary antibodies to bind to the primary antibodies. The specific IHC staining procedure was consistent with our previous research (Zhong et al. 2023). The final score was obtained by adding the percentage of positively stained cells to the staining intensity score. The scoring for the percentage of
immunoreactive cells was defined as follows: 0 (0%), 1 (1-10%), 2 (11-50%), and 3 (>50%). Visual scoring and stratification of staining intensity were defined as follows: 0 (negative), 1 (weak), 2 (moderate), and 3 (strong) (Lu et al. 2018). The utilized antibody was anti-LMNB1 (YM3036; ImmunoWay, Plano, TX, USA).
Cell transient transfection
In the present study, two human ACC cell lines, SW13 and H295R, were employed. These ACC cells were incubated in DMEM/F12 (Bio-Channel, Nanjing, China) supplemented with 10% fetal bovine serum (ThermoFisher), maintained in a humidified incubator with 5% CO2 at 37°C. Negative control (NC) and LMNB1 siRNA (Genepharma, Suzhou, China) were transfected into ACC cells according to the manufacturer’s instructions using GP-transfect-Mate (Genepharma, Suzhou, China). After placing the plate in the incubator for 48 h, total protein was extracted for Western blot. The siRNA sequences are presented in Supplementary Table 2.
Western blot
For the Western blot, we harvested cells and subjected them to lysis in a RIPA buffer supplemented with protease inhibitors. The extracted protein samples were then separated through SDS-PAGE, transferred to PVDF membranes, and subsequently blocked with a 5% solution of nonfat milk. We then incubated the membranes with an ImmunoWay-supplied primary anti-LMNB1 antibody (at a 1:1000 dilution), and later with a secondary antibody (ImmunoWay, 1:5000 dilution). Following three PBST washes, each lasting 10 min, we exposed the membranes. As a normalization control, B-ACTIN (1:5000; ImmunoWay) was utilized. Using Image J software, we quantified the intensities of the final protein bands.
Transwell assay
We cultured an approximate 4 x 104 count of transiently transfected cells in a serum-free medium within the upper chamber, while placing complete medium into the lower chamber. The cells were then incubated at 37℃ and 5% CO2 for a duration of 48 h. After a wash with PBS and fixation with paraformaldehyde, cells were stained using a 0.1% crystal violet solution. The final step involved capturing images of the cells under a microscope and conducting a count of the stained cells.
Colony formation assay
Each cell line was seeded at 1000 cells/well in six- well plates and incubated for 2 weeks in a 37℃, 5% CO2 incubator. At the endpoint of 2 weeks, wells were washed twice with cold PBS, fixed for 15 min with 4% paraformaldehyde, and stained at room
temperature for 20 min with a 1% crystal violet solution. Visible colonies were counted. This experiment was performed in triplicate.
Statistical analysis
During data processing, R 4.1.0 was utilized for statistical analysis and plotting, with some data visualization performed using the Sanger Box bioinformatics analysis online tool (Shen et al. 2022). Immunohistochemistry experiments and cell experiment data statistical analysis and plotting were carried out using GraphPad Prism 8.0. The Pearson or Spearman correlation coefficients were used to evaluate the correlation between gene expression and immune infiltrating cells. The Wilcoxon rank- sum test and Kruskal-Wallis test were employed to examine differences between two groups and multiple groups, respectively. P-values were two-sided, with statistical significance indicated as *P < 0.05, ** P < 0.01, and *** P < 0.001.
Results
Identifying TMB-related gene modules in ACC
To identify high TMB-related gene modules in ACC, we clustered 79 ACC samples from TCGA that were quartile-divided based on TMB values and removed 1 outlier sample (Supplementary Fig. 1). Subsequently, we selected 0.25 and 30 as the thresholds for module merging and the minimum module size, respectively, and performed WGCNA, resulting in eight different gene modules (Fig. 2A). Next, we explored the relationship between the modules and TMB using a heatmap. After excluding genes with no statistically significant correlations, we identified the yellow gene module as the module most strongly correlated with TMB (r=0.54, P=3e-07) (Fig. 2B and Supplementary Table 3). Additionally, in the gene significance vs module membership plot, yellow module genes displayed consistent results (r=0.51, P=1.7e-26) (Fig. 2C). To investigate the potential mechanisms of the aforementioned module, we performed KEGG enrichment analysis and GO enrichment analysis on the 373 genes in these modules. The results demonstrated significant enrichment of 335 GO terms and 16 KEGG pathways in the module genes, including pathways notably associated with cell proliferation and apoptosis (DNA replication, chromosome segregation, cell cycle, and the p53 signaling pathway) (Fig. 2D, E and Supplementary Tables 4 and 5).
Exploring TMB-related differential peak genes using ATAC-seq
Research has suggested that the open state of chromatin in cancer origin cells is associated with the
A
Cluster Dendrogram
1.0
D
0.9
GO
Height
0.8
organelle fission
5
nuclear division
chromosome segregation
0.6
nuclear chromosome segregation
0.5
Count
mitotic nuclear division
40
DNA replication
Module colors
50
mitotic cell cycle phase transition
60
Merged colors
70
regulation of mitotic cell cycle
80
sister chromatid segregation
B
Module-TMB relationships
90
regulation of cell cycle phase transition
mitotic sister chromatid segregation
MEblue
-0.071(0.5)
-0.16(0.2)
0.0084(0.9)
0.23(0.04)
0.21(0.07)
1
p.adjust
1.50-68
negative regulation of cell cycle process
8.4e-31
MEturquoise
DNA-dependent DNA replication
0.31(0.006)
-0.093(0.4)
-0.13(0.3)
-0.086(0.5)
-0.24(0.04)
1.60-30
regulation of mitotic
2.5e-30
cell cycle phase transition
MEred
0.3(0.008)
0.005(1)
-0.0032(1)
-0.31(0.006)
-0.36(0.001)
0.5
3.3e-30
meiotic cell cycle
cell cycle checkpoint signaling
MEyellow
0.54(3e-07)
-0.052(0.7)
-0.17(0.1)
-0.33(0.003)
-0.53(5e-07)
0
0.25 0.20 0.15 GeneRatio
MEblack
-0.053(0.6)
-0.1(0.4)
-0.11(0.3)
0.28(0.01)
0.19(0.1)
MEbrown
-0.097(0.4)
-0.081(0.5)
-0.11(0.3)
0.3(0.008)
0.22(0.05)
-0.5
E
MEgreen
0.04(0.7)
-0.025(0.8)
-0.24(0.04)
0.23(0.04)
0.065(0.6)
KEGG
MEgrey
0.16(0.2)
-0.23(0.05)
-0.14(0.2)
0.22(0.06)
0.047(0.7)
-1
Cell cycle
TMB_Top25%
TMB_25%~50%
TMB_50%~75%
TMB_Bottom25%
TMB
Oocyte meiosis
DNA replication
MicroRNAs in cancer
Cellular senescence
C
Module membership vs. gene significance
Progesterone-mediated
p.adjust
oocyte maturation
0.6
Fanconi anemia pathway
0.01
Human T-cell leukemia
0.02
virus 1 infection
0.03
Homologous recombination
Gene significance
0.5
0.04
Viral carcinogenesis
0.4
Count
p53 signaling pathway
10
Nucleotide excision repair
0.3
20
Hepatitis B
30
Small cell lung cancer
0.2
cor=0.51,
p=1.7e-26
Mismatch repair
Base excision repair
0.4
0.5
0.6
0.7
0.8
0.9
0.2
0.1
Module Membership in Yellow module
GeneRatio
heterogeneous distribution of tumor mutations (Corces et al. 2016). Therefore, to explore chromatin accessibility differences under different TMB conditions in ACC, we analyzed differential peak genes using ATAC-seq data from TCGA-ACC. The TMB values of the aforementioned 79 samples were grouped according to the median value for ATAC-seq sample annotation,
with 8 in the TMB-Low group and 10 in the TMB-High group. The differential analysis yielded 1950 differential accessibility peaks (DAPs) (Fig. 3A, adj. P < 0.05, log2 FC > 2). Then, we annotated these DAPs and ALL peaks. The results showed that the percentage of distal elements defined as nonpromoter components was higher in DAPs than in ALL peaks, suggesting a
stronger specific response to the TMB state in distal elements (Fig. 3B, C, and D).
Selection and validation of TMB-related biomarkers
Initially, we performed Cox regression analyses on all genes in TCGA-ACC and calculated the AUC with the median overall survival (OS) time and median progression-free interval (PFI), selecting genes with AUC ≥0.7 and HR >1. Intersecting these genes with the previously established WGCNA yellow module genes
and ATAC-seq differential peak genes resulted in six candidates (RCC2, SPC24, NCAGP2, LMNB1, ORC6, and TRIP13) (Fig. 4A). We then computed the AUC and performed univariate Cox regression analyses on these six genes in two pediatric ACC cohorts (GSE76019, GSE76021) and two adult ACC cohorts (GSE19750, GSE10927) from GEO using OS or event-free survival (EFS) as the prognostic endpoints. The results highlighted the robust prognostic prediction capability of LMNB1 across multiple datasets (Fig. 4B and C). Further univariate and multivariate Cox regression analyses in TCGA and GEO datasets revealed that LMNB1 serves as an independent prognostic predictor (Fig. 4D and E).
A
TMB High vs TMB Low
B
Feature Distribution
40
DAPs
30
Up
All_peak-
Feature
NS
Promoter ( ⇐ 1kb)
Down
Promoter (1-2kb)
-log10 (adj.P.Val)
Promoter (2-3kb)
5’ UTR
3’ UTR
20
LDB2
1st Exon
Other Exon
1st Intron
Other Intron
PRDM9
LDB2
DAPs
Downstream ( ⇐ 300)
GRIP1
Distal Intergenic
10
CRTAM
NLRP11
PTK7O
ZNF622
PPP1R3C
BRD9
0
25
Percentage(%)
50
75
100
0
-4
-3
-2
-1
0
1
2
3
4
5
log2 (fold change)
C
D
All peaks
DAPs
800
Genic
Intergenic
Genic
20000
Intron
Intergenic
Intron
Exon
Upstream
600
Exon
Downstream
Upstream
Distal Intergen
Downstream
Distal Intergen
400
10000
200
0
0
genic
Intron
genic
Intron
Promoter
Intergenic
Exon
distal_intergenic
Intergenic
Promoter
fiveUTR
distal_intergenic
Exon
fiveUTR
threeUTR
threeUTR downstream
downstream
A
RCC2
0
1392
SPC24
7
NCAPG2
1
10
265
LMNB1
ORC6
206
TRIP13
1
22
1
0
0
6
317
0
51
26
4
1437
1150
528
158
WGCNA
DPGs
One-way Cox with OS HR >1
One-way Cox with PFI HR >1
AUC (50%OS) ≥ 0.7
AUC (50%PFI) ≥0.7
D Univariate COX
| Variable | HR | lower 95% CI | upper 95% CI | pvalue | |
|---|---|---|---|---|---|
| TCGA-ACC | · | ||||
| Age | 1.011 | 0.987 | 1.036 | 3.79E-01 | |
| Gender | 1.001 | 0.469 | 2.137 | 9.99E-01 F | - 1 |
| Stage | 2.628 | 1.714 | 4.03 | 9.39E-06 | F 1 |
| clinical_M | 5.3 | 2.364 | 11.883 | 5.14E-05 | I- -1 |
| pathologic_N | 1.742 | 0.679 | 4.473 | 2.48E-01 | F 1 |
| pathologic_T | 3.04 | 1.95 | 4.741 | 9.28E-07 | F 1 |
| LMNB1 | 3.105 | 2.116 | 4.556 | 6.99E-09 | F 1 |
| GSE19750 | |||||
| Age | 1.035 | 0.994 | 1.079 | 9.73E-02 | |
| Gender | 1.262 | 0.463 | 3.441 | 6.50E-01 I- | -1 |
| Stage | 1.027 | 0.739 | 1.428 | 8.73E-01 | 1. -I |
| LMNB1 | 2.186 | 1.119 | 4.272 | 2.21E-02 | I |
| GSE10927 | |||||
| Age | 1.012 | 0.977 | 1.048 | 5.19E-01 | |
| Gender | 1.471 | 0.510 | 4.242 | 4.75E-01 | I 1 |
| Stage | 1.818 | 1.095 | 3.019 | 2.09E-02 | -- I |
| LMNB1 | 10.743 | 1.307 | 88.296 | 2.72E-02 | |
| GSE76019 | |||||
| Age | 1.019 | 1.010 | 1.029 | 7.30E-05 | |
| Gender | 5.684 | 1.697 | 19.033 | 4.83E-03 | : I · I |
| Stage | 3.075 | 1.468 | 6.442 | 2.91E-03 | F 1 |
| LMNB1 | 2.820 | 1.525 | 5.215 | 9.51E-04 | + |
| GSE76021 | |||||
| Age | 1.004 | 0.994 | 1.013 | 4.88E-01 | |
| Gender | 1.294 | 0.378 | 4.436 | 6.81E-01 I | 1 |
| Stage | 1.376 | 0.736 | 2.571 | 3.18E-01 | I- 4 |
| LMNB1 | 3.124 | 1.228 | 7.950 | 1.68E-02 | F 1 |
-1 0
1
2
3
4
5
log2(Hazard Ratio(95%CI
B
LMNB1
Cohort
NCAPG2
GSE10927
GSE19750
GSE76019
ORC6
GSE76021
Type
Risky
Not Significant
TRIP13
RCC2
None
C
LMNB1
0.694
0.899
0.710
0.780
Cohort
NCAPG2
0.542
0.798
0.434
0.890
GSE19750
GSE76019
GSE76021
ORC6
0.618
0.739
0.628
0.795
AUC value
1.0
SPC24
0.420
0.746
0.559
0.595
0.9
0.8
0.7
TRIP13
0.778
0.856
0.441
0.921
0.6
RCC2
0.771
0.674
0.634
None
E
| Variable | HR | lower 95% CI | upper 95% CI | pvalue | ||
|---|---|---|---|---|---|---|
| TCGA-ACC | ||||||
| Stage | 1.355 | 0.363 | 5.054 | 6.51E-01 | + | 1 |
| clinical_M | 0.45 | 0.073 | 2.772 | 3.89E-01 | 1 | |
| pathologic_T | 1.891 | 0.867 | 4.124 | 1.09E-01 | : | |
| LMNB1 | 2.582 | 1.594 | 4.184 | 1.17E-04 | I- -I | |
| GSE19750 | ||||||
| Age | 1.044 | 1.004 | 1.086 | 3.12E-02 | ||
| LMNB1 | 2.615 | 1.285 | 5.324 | 8.02E-03 | F 土 | |
| GSE10927 | ||||||
| Stage | 2.852 | 1.471 | 5.529 | 1.92E-03 | F | |
| LMNB1 | 50.201 | 4.209 | 598.79 | 1.96E-03 | F | |
| GSE76019 | ||||||
| Age | 1.012 | 0.995 | 1.028 | 1.63E-01 | ||
| Gender | 4.671 | 1.18 | 18.497 | 2.81E-02 | I · I | |
| Stage | 1.714 | 0.477 | 6.168 | 4.09E-01 | F | 1 |
| LMNB1 | 2.281 | 1.104 | 4.714 | 2.60E-02 | F + | |
| GSE76021 | ||||||
| Stage | 1.846 | 0.862 | 3.951 | 1.15E-01 | H | -1 |
| LMNB1 | 4.188 | 1.43 | 12.266 | 8.99E-03 |
1
£
-3-2-1 0 1
2
3
4
5
6
7
8
n 9
log2(Hazard Ratio(95%CI))
Figure 4 Screening and verification of TMB prognostic-related genes. (A) Venn diagram of candidate genes with WGCNA module genes, DPGs, univariate COX, and AUC ≥0.7 in TCGA-ACC. (B) Heatmap of univariate COX of the six intersecting genes in GEO datasets. (C) ROC AUC of the six candidate genes in the GEO datasets. (D, E) Univariate and multivariate regression analysis of TCGA and GEO cohorts. A full color version of this figure is available at https://doi. org/10.1530/ERC-23-0178.
GSE10927
SPC24
Therefore, we conjecture that LMNB1 might serve as a novel biomarker for ACC.
High expression of LMNB1 indicates poor prognosis in ACC patients and displays abnormal chromatin accessibility
To investigate the relationship between LMNB1 expression and ACC patient prognosis, we conducted KM survival analysis across multiple cohorts, revealing a clear trend of poor prognosis in ACC patients with high LMNB1 expression levels (Fig. 5A, B, C, D, E and F). Moreover, we observed a significant increase in LMNB1 expression levels in the TMB-High group (P=0.00029) (Fig. 5G). In the ATAC-seq data, a change was observed in the LMNB1 peak in the TMB-High group. IGV display showed a significant decrease in accessibility in the exonic and intronic component regions of LMNB1 corresponding to ACC-25728 and ACC-25731 (Fig. 5H and I). These findings suggest that high LMNB1 expression levels indicate poor prognosis in multiple adult and pediatric ACC cohorts and may be associated with a high TMB and reduced chromatin accessibility.
Exploring the biological characteristics of LMNB1
To reveal the biological characteristics of LMNB1, we performed functional enrichment analysis based on the correlations of LMNB1 with other genes. Figure 6A and B display the significantly enriched GO terms and KEGG pathways for LMNB1 (Fig. 6A, B, Supplementary Tables 6 and 7). The GO terms encompassed gene sets highly correlated with LMNB1, including activation of terms related to cell proliferation, such as ‘attachment of spindle microtubules to kinetochore’, ‘cell cycle checkpoint signaling’, and ‘cell cycle DNA replication’, and inhibition of immune processes, such as ‘antigen processing and presentation’ (Fig. 6A). KEGG pathways included the activation of cell proliferation-related pathways, such as ‘cell cycle’ and ‘DNA replication’, and inhibition of oxidative phosphorylation-related pathways, such as ‘metabolism of xenobiotics by cytochrome P450’ and ‘oxidative phosphorylation’ (Fig. 6B). Furthermore, in our GSEA, we employed the Reactome and WikiPathways (Supplementary Tables 8 and 9). The outcomes of these analyses were consistent with the results obtained from the GO and KEGG enrichment analyses, as illustrated in Supplementary Fig. 2. Based on these findings, LMNB1 may participate in activating tumoral cell proliferation and growth signaling pathways while inhibiting cell immunity and oxidative phosphorylation.
Experimental validation of LMNB1
To ascertain the expression patterns of LMNB1, we collected clinical samples of ACC for immunohistochemical staining. Figure 7A displays
representative images of LMNB1 immunohistochemical staining results. LMNB1 is expressed in the nuclear envelope of adrenal cells, and its protein expression level in the ACC group was higher than that in the noncancer group (P=0.04). Subsequently, to verify the function of LMNB1 in ACC cells, we transfected specific LMNB1 siRNAs into SW13 and H295R cell lines and selected si-LMNB1#1/2 as the most efficient siRNA fragment based on Western blotting. We then conducted Transwell and plate colony formation experiments, which suggested that knocking down LMNB1 can significantly inhibit the invasion and colony formation abilities of SW13 and H295R cancer cell lines (Fig. 7C and D). In addition, upon the overexpression of LMNB1 in SW13 and H295R cell lines (Supplementary Fig. 3A), there was a marked enhancement in the cells’ colony formation and invasion abilities (Fig. 3B). Therefore, the high expression levels of LMNB1 in ACC and their oncogenic effect provide evidence for the potential role of LMNB1 as a biomarker.
Multidimensional features of LMNB1
Next, we sought to investigate the role of LMNB1 in ACC from a multiomics perspective. At the gene mutation level, we plotted mutation and CNV landscape diagrams of the top 20 frequently mutated genes (FMGs) in ACC patients (Fig. 8A). Notably, TP53, a tumor suppressor gene with a high propensity for mutation in cancer, exhibited a mutation frequency of 29% in the high LMNB1 expression group, compared to a mere 3% mutation frequency in the low LMNB1 expression group. The difference in mutation frequencies between these two groups was statistically significant (P < 0.01) (Supplementary Fig. 4). Additionally, in CNVs, we observed that deletion mutations were more common in the high LMNB1 group (Fig. 8A, B and Supplementary Fig. 4). Furthermore, we assessed the expression of LMNB1 in the remaining 31 solid tumors in TCGA. Overall, LMNB1 was highly expressed in most cancer tissues (Supplementary Fig. 5A), and it was a poor prognostic factor in multiple types of tumors, including BLCA, BRCA, KICH, and SARC (Supplementary Fig. 5B and C). Finally, we analyzed the relationship between LMNB1 and the ACCimmune microenvironment (Supplementary Fig. 6A) and conducted a correlation analysis between LMNB1 expression and major immune cell infiltration (Fig. 6B). The immune infiltration scores of CD8+ T cells were negatively correlated with LMNB1 expression in all six algorithms, suggesting that high expression levels of LMNB1 might lead to poor prognosis by inhibiting CD8+ T-cell infiltration in the ACC microenvironment (Frankiw & Li 2022, Wang et al. 2022).
Discussion
ACC is a rare malignancy with a poor prognosis; it has an annual incidence rate of approximately 1-2
A
TCGA-ACC
B
TCGA-ACC
C
GSE10927
G
Survival probability
1.0
LMNB1
1.0
LMNB1
1.0
LMNB1
Low
Survival probability
Low
Survival probability
Low
High
High
15.0
0.8
0.8
0.8
High
Wilcoxon, p = 0.00029
0.5
0.5
0.5
12.5
Expression
0.3
0.3
0.3
10.0
0.0
log-rank test p=2.7e-9
0.0
log-rank test p=1.2e-7
0.0
log-rank test p=0.09
Number at risk
Number at risk
Number at risk
7.5
Low
53
35
14
6
2
Low
38
21
10
4
1
Low
7
2
1
1
1
High
26
6
1
1
1
High
41
3
1
1
1
High
17
3
1
1
1
5.0
0
38
76
114
76
74
OS.time(Months)
152
0
38
PFI.time(Months)
114
152
0
37
OS.time(Months)
111
148
Low
High
TMB
D
E
F
H
GSE76019
GSE76021
GSE19750
Survival probability
1.0
LMNB1,
1.0
LMNB1
1.0
LMNB1
Low
Survival probability
Low
Survival probability
Low
10.5-
High
Wilcoxon
0.8
0.8
High
0.8
High
Wilcoxon
p=0.021
p=0.0044
10.0
0.5
0.5
0.5
Peak_vst
9.5
0.3
0.3
0.3
9.0
0.0
log-rank test p=2.0e-5
0.0
log-rank test p=2.1e-3
0.0
log-rank test p=0.04
1
Number at risk
Number at risk
Number at risk
8.5
Low 17
14
10
3
1
Low
13
6
5
1
1
Low 5
5
2
2
1
ACC_25728
ACC_25731
High
17
7
4
1
1
High
6
1
1
1
1
High
16
4
3
2
1
0
21
EFS.time(Months)
42
63
84
0
48
EFS.time(Months)
96
144
192
0
54
108
162
216
Low
High
Low
High
OS.time(Months)
TMB
126,820 kb
126,840 kb
126,860 kb
126,880 kb
chr5
ACC-25728
ACC-25731
TCGA-OR-A5J3-01A
L
TCGA-OR-A5JZ-01A
TCGA-PA-A5YG-01A
TCGA-PK-A5H8-01A
TCGA-OR-A5J2-01A
TCGA-OR-A5J6-01A
TCGA-OR-A5J9-01A
TCGA-OR-A5K8-01A
TCGA-OR-A5KX-01A
TMB-Low group
TMB-High group
LMNB1
MARCHF3
A
B
GOBP ANTIGEN PROCESSING AND PRESENTATION
KEGG DRUG METABOLISM CYTOCHROME P450
GOBP ANTIGEN PROCESSING AND PRESENTATION OF EXOGENOUS ANTIGEN
KEGG OXIDATIVE PHOSPHORYLATION-
GOBP CELL CYCLE G1 S PHASE TRANSITION
KEGG METABOLISM OF XENOBIOTICS BY CYTOCHROME P450-
GOBP CELLULAR RESPONSE TO DNA DAMAGE STIMULUS
KEGG PARKINSONS DISEASE
NES
NES
GOBP CELL CYCLE G2 M PHASE TRANSITION
2
KEGG GRAFT VERSUS HOST DISEASE
0
-2
4
GOBP CELL CYCLE PHASE TRANSITION
KEGG HEMATOPOIETIC CELLLINEAGE
GOBP ATTACHMENT OF SPINDLE MICROTUBULES TO KINETOCHORE
KEGG ALLOGRAFT REJECTION
GOBP CELL DIVISION
KEGG RETINOL METABOLISM-
GOBP CELL CYCLE DNA REPLICATION
KEGG DNA REPLICATION
GOBP CELL CYCLE CHECKPOINT SIGNALING
KEGG CELL CYCLE
-0.5
0.0
0.5
1.0
-0.5
0.0
0.5
1.0
Enrichment Scores
Enrichment Scores
2
0
-2
4
per million, accounting for approximately 0.12% of all malignant tumors (Else et al. 2014). In this study, we sought to identify new biomarkers for ACC using TMB as an entry point. To identify gene modules with high TMB in ACC, we employed WGCNA to assess TMB- associated gene modules. Subsequently, combined with the results of ATAC-seq, these results indicate that the proportion of distal elements in differentially accessible peaks (DAPs) of ACC samples with different TMB statuses was higher than that of all peaks, indicating that distal elements are more sensitive to specific TMB states. Finally, using several adult and pediatric ACC cohorts, we discovered that LMNB1 is a potential ACC biomarker.
ATAC-seq is recognized as a method to study chromatin accessibility at the epigenomic level; it can identify mutations in regulatory genes that drive cancer initiation and progression (Corces et al. 2018). Given that differences in chromatin accessibility play a critical role in tumor regulatory gene mutations, ATAC- seq can assist in identifying biomarkers associated with TMB. Furthermore, ATAC-seq peaks may form part of regulatory units or enhancers, and these peaks have strong connections with genes expected to be regulated by these peaks (Corces et al. 2018). Therefore, we can explore which genes are affected by variations in tumor mutation load through differential accessibility analysis. In our study, we conducted a differential analysis between the TMB-Low group and the TMB-High group in ACC patients to identify DAPs, and using peak-to-gene links, we identified DPGs under different TMB states. Furthermore, we observed
that the proportion of distal elements in DAPs from ACC samples with different TMB statuses was higher than that in ALL peaks, indicating that distal elements are more sensitive to specific TMB statuses. By combining the results of this analysis with those of the ATAC-seq analysis, we identified LMNB1, a gene with significant chromatin accessibility differences in the high TMB gene module.
LMNB1 encodes one of the two B-type lamins, constituting part of the nuclear lamina. Its main function is to maintain the integrity of the nuclear skeleton, and it participates in cell proliferation and aging by affecting chromosome distribution, gene expression, and DNA damage repair (Shumaker et al. 2008, Dittmer & Misteli 2011). LMNB1 has been reported to exhibit oncogenic effects in a variety of cancers. For example, in the context of cervical cancer, Yang et al. discovered that LMNB1 could increase cervical cancer progression and nuclear autophagy (Yang et al. 2019). Li and colleagues found that knockdown of LMNB1 inhibits the proliferation of lung adenocarcinoma cells by inducing DNA damage and cell senescence. Additionally, the absence of LMNB1 leads to a decrease in H3k9me3 protein expression, thereby increasing chromosome accessibility (Li et al. 2022). In cases of liver cancer, Sun et al. found that LMNB1 was overexpressed in both the tumors and plasma proteome of patients, and it was linked to tumor size, tumor stage, and the presence of tumor nodules (Sun et al. 2010). Furthermore, research conducted by Izdebska and colleagues revealed that high expression levels of LMNB1 in human colon cancer cells resulted
A
10X
20X
Non-cancer (n=6)
Immunoreactive Score of LMNB1
10%
15
*
10
ACC (n=5)
5
0
B
Non-cancer
ACC
NC
Si-1
Si-2
Si-3
SW13
NCI-H295R
SW13
LMNB1
70kd
Relative LMNB1 expression
1.5
Relative LMNB1 expression
1.5
B-ACTIN
43kd
1.0
1.0
NC
Si-1
Si-2
Si-3
H295R
LMNB1
70kd
0.5
0.5
B-ACTIN
43kd
0.0
0.0
C
NC
Si-1
Si-2
Si-3
NC
Si-1
Si-2
Si-3
NC
Si-1
Si-2
SW13
NCI-H295R
200
250
SW13
cell numbers
150
cell numbers
200
150
100
*
*
100
H295R
50
50
0
0
NC
Si-1
Si-2
NC
Si-1
Si-2
D
NC
Si-1
Si-2
SW13
NCI-H295R
SW13
400
100
colony numbers
300
colony numbers
80
60
200
40
H295R
100
20
0
0
NC
Si-1
Si-2
NC
Si-1
Si-2
A
B Low LMNB1 High LMNB1
3210
log10(TMB+2)
LMNB1
chr1
chr1
Loss
Gain
17%
TP53
16%
CTNNB1
16%
MUC16
chr2
chr2
11%
TTN
8%
CNTNAP5
8%
HMCN1
chr3
chr3
8%
PKHD1
7%
APOB
chr4
chr4
7%
KMT2B
7%
NF1
7%
PRKAR1A
chr5
chr5
7%
SVEP1
7%
TUT7
chr6
chr6
5%
ASXL3
5%
MEN1
chr7
chr7
5%
CMYA5
5%
FRAS1
chr8
chr8
5%
LRP1
5%
STAB1
chr9
chr9
4%
ZNRF3
chr10
76%
12q14.1-Amp
chr10
76%
12q14.3-Amp
chr11
chr11
76%
12q15-Amp
75%
5p15.33-Amp
chr12
chr12
69%
5q35.3-Amp
63%
5p15.31-Amp
chr13
chr13
56%
22q12.1-Del
chr14
chr14
43%
1p36.23-Del
chr15:
chr15:
29%
4q35.1-Del
29%
4q34.3-Del
chr16
chr16
28%
9p21.3-Del
chr17
25%
3q13.31-Del
chr18
chr17
chr19
chr18
80%
CDK4
0
0.5
dhr19
-
chr20
chr20 chr22
80%
OS9
PCT
chr21
chr2
80%
AGAP2
56%
ZNRF3-AS1
LMNB1
Alterations
CNA (arm-level)
CNA (gene-level)
☒ Low
☒ Mutated
Gain
Gain
High
☐ Loss
High_balanced_gain
Loss
High_balanced_loss
40
20
0
20
40
60
80
40
20
0
20
40
80
Frequency
Frequency
in the induction of mitotic catastrophe following 5-FU treatment, the enhancement of intercellular connections, and a restriction on cell migration (Izdebska et al. 2018). Our research serves to fill the existing research void concerning the role of LMNB1 in ACC. In our study, high expression levels of LMNB1 in ACC indicated OS, PFI, and EFS. The GSEA results suggested that LMNB1 is closely related to tumor development and progression pathways, such as the proliferation and growth of tumor cells. Our experimental results indicated that the LMNB1 could promote the proliferation and invasion of ACC cells. Within the mutation analysis, we identified that TP53 harbors the highest mutation frequency in ACC, with a higher incidence of TP53 mutations observed in the high LMNB1 expression group. TP53 is regarded as one of the most pivotal cancer suppressor genes, and mutations in TP53 play a critical role in the onset and progression of ACC (Olivier et al. 2010, Curylova et al. 2022). These findings further underscore the positive correlation between LMNB1 and elevated TMB. However, some limitations exist in our study. First, although we validated our results using public databases, our own sample size of ACC is still limited. Second, we only validated our results at the cellular experimental level. In the future, we aim to further validate our results through animal experiments and other assays.
Conclusion
Our study is the first to report that LMNB1 is associated with a high TMB and exhibits abnormal chromatin accessibility in ACC. High expression levels of LMNB1 in ACC, in both adults and children, indicate a poor prognosis. Furthermore, LMNB1 was found to be overexpressed in ACC samples. Through in vitro assays, LMNB1 was found to promote the invasiveness and proliferation of ACC cells. In summary, LMNB1 is a promising biomarker for the evaluation and potential therapeutic targeting of ACC.
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/ ERC-23-0178.
Declaration of interest
The authors declare that they do not have any conflicts of interest.
Funding
This research was supported by grants from the National Natural Science Foundation of China (No.82003271), the Science and Technology Development Fund, Macau SAR (File no. 0116/2023/RIA2,0090/2022/A), the Guangzhou Planned Project of Science and Technology (2023A04J1269, 202102080089), Medical Science and Technology Research Fund of Guangdong Province (grant no. B2021317), and Huizhou High Level Hospital Construction Science and Technology Special Project (grant No.2022CZ010004).
Data availability statement
The public data used in this study have been described in the Materials and Methods section.
Author contribution statement
JML, JHC, and ZD played instrumental roles in the conceptualization and design of the study. Bioinformatics analysis was conducted by JML, CYH, YCS, and FPL The collection of clinical samples was undertaken by JHC and CYH. Supervision of the research was carried out by JML, ZD, XCB and WDZ. Manuscript preparation was done by JHC, YHC, and XHL, while experimental validation was achieved by CFZ, SMF, and YHC. All authors made substantial contributions to the manuscript and granted approval for its submission.
Acknowledgement
We would like to express our gratitude to all the contributors of the public datasets.
References
Blanche P, Dartigues JF & Jacqmin-Gadda H 2013 Estimating and comparing time-dependent areas under receiver operating characteristic curves for censored event times with competing risks. Statistics in Medicine 32 5381-5397. (https://doi.org/10.1002/sim.5958)
Corces MR, Buenrostro JD, Wu B, Greenside PG, Chan SM, Koenig JL, Snyder MP, Pritchard JK, Kundaje A, Greenleaf WJ, et al. 2016 Lineage- specific and single-cell chromatin accessibility charts human hematopoiesis and leukemia evolution. Nature Genetics 48 1193-1203. (https://doi.org/10.1038/ng.3646)
Corces MR, Granja JM, Shams S, Louie BH, Seoane JA, Zhou W, Silva TC, Groeneveld C, Wong CK, Cho SW, et al. 2018 The chromatin accessibility landscape of primary human cancers. Science 362. (https://doi. org/10.1126/science.aav1898)
Curylova L, Ramos H, Saraiva L & Skoda J 2022 Noncanonical roles of p53 in cancer stemness and their implications in sarcomas. Cancer Letters 525 131-145. (https://doi.org/10.1016/j.canlet.2021.10.037)
Demeure MJ, Coan KE, Grant CS, Komorowski RA, Stephan E, Sinari S, Mount D & Bussey KJ 2013 PTTG1 overexpression in adrenocortical cancer is associated with poor survival and represents a potential therapeutic target. Surgery 154 1405-1416. (https://doi.org/10.1016/j. surg.2013.06.058)
Dittmer TA & Misteli T 2011 The lamin protein family. Genome Biology 12 222. (https://doi.org/10.1186/gb-2011-12-5-222)
Else T, Kim AC, Sabolch A, Raymond VM, Kandathil A, Caoili EM, Jolly S, Miller BS, Giordano TJ & Hammer GD 2014 Adrenocortical carcinoma. Endocrine Reviews 35 282-326. (https://doi.org/10.1210/er.2013-1029)
Frankiw L & Li G 2022 The tumor microenvironment’s role in malignant progression and treatment response. Cancer Letters 548 215888. (https:// doi.org/10.1016/j.canlet.2022.215888)
Giordano TJ, Kuick R, Else T, Gauger PG, Vinco M, Bauersfeld J, Sanders D, Thomas DG, Doherty G & Hammer G 2009 Molecular classification and prognostication of adrenocortical tumors by transcriptome profiling. Clinical Cancer Research 15 668-676. (https://doi. org/10.1158/1078-0432.CCR-08-1067)
Goldman MJ, Craft B, Hastie M, Repečka K, McDade F, Kamath A, Banerjee A, Luo Y, Rogers D, Brooks AN, et al. 2020 Visualizing and interpreting cancer genomics data via the Xena platform. Nature Biotechnology 38 675-678. (https://doi.org/10.1038/s41587-020-0546-8)
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)
Hanahan D 2022 Hallmarks of cancer: new dimensions. Cancer Discovery 12 31-46. (https://doi.org/10.1158/2159-8290.CD-21-1059)
Izdebska M, Gagat M & Grzanka A 2018 Overexpression of lamin B1 induces mitotic catastrophe in colon cancer LoVo cells and is associated with worse clinical outcomes. International Journal of Oncology 52 89-102. (https://doi.org/10.3892/ijo.2017.4182)
Jardim DL, Goodman A, de Melo Gagliato D & Kurzrock R 2021 The challenges of tumor mutational burden as an immunotherapy biomarker. Cancer Cell 39 154-173. (https://doi.org/10.1016/j. ccell.2020.10.001)
Landwehr LS, Schreiner J, Appenzeller S, Kircher S, Herterich S, Sbiera S, Fassnacht M, Kroiss M & Weigand I 2021 A novel patient-derived cell line of adrenocortical carcinoma shows a pathogenic role of germline MUTYH mutation and high tumour mutational burden. European Journal of Endocrinology 184 823-835. (https://doi.org/10.1530/EJE-20-1423)
Langfelder P & Horvath S 2008 WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9 559. (https://doi. org/10.1186/1471-2105-9-559)
Li J, Sun Z, Cui Y, Qin L, Wu F, Li Y, Du N & Li X 2022 Knockdown of LMNB1 inhibits the proliferation of lung adenocarcinoma cells by inducing DNA damage and cell senescence. Frontiers in Oncology 12 913740. (https://doi.org/10.3389/fonc.2022.913740)
Lu J, Dong W, He H, Han Z, Zhuo Y, Mo R, Liang Y, Zhu J, Li R, Qu H, et al. 2018 Autophagy induced by overexpression of DCTPP1 promotes tumor progression and predicts poor clinical outcome in prostate cancer. International Journal of Biological Macromolecules 118 599-609. (https:// doi.org/10.1016/j.ijbiomac.2018.06.005)
Luo Y, Chen Q & Lin J 2022 Identification and validation of a tumor mutation burden-related signature combined with immune microenvironment infiltration in adrenocortical carcinoma. Mathematical Biosciences and Engineering 19 7055-7075. (https://doi.org/10.3934/ mbe.2022333)
Mayakonda A, Lin DC, Assenov Y, Plass C & Koeffler HP 2018 Maftools: efficient and comprehensive analysis of somatic variants in cancer. Genome Research 28 1747-1756. (https://doi.org/10.1101/gr.239244.118)
Olivier M, Hollstein M & Hainaut P 2010 TP53 mutations in human cancers: origins, consequences, and clinical use. Cold Spring Harbor Perspectives in Biology 2 a001008. (https://doi.org/10.1101/cshperspect. a001008)
Paschoalin VP, Tucci Júnior S, Estevanato AB, Tiraboschi R, Antonini SR, Muglia VF, Chahud F, Mermejo LM, de Bessa Júnior J & Fernandes Molina CA 2022 Histopathological criteria for paediatric adrenocortical carcinoma. Hormone Research in Paediatrics 95 347-353. (https://doi. org/10.1159/000524892)
Pinto EM, Rodriguez-Galindo C, Choi JK, Pounds S, Liu Z, Neale G, Finkelstein D, Hicks JM, Pappo AS, Figueiredo BC, et al. 2016 Prognostic significance of major histocompatibility complex Class II expression in pediatric adrenocortical tumors: ASt. Jude and Children’s Oncology
Group study. Clinical Cancer Research 22 6247-6255. (https://doi. org/10.1158/1078-0432.CCR-15-2738)
Shen W, Song Z, Zhong X, Huang M, Shen D, Gao P, Qian X, Wang M, He X, Wang T, et al. 2022 Sangerbox: a comprehensive, interaction- friendly clinical bioinformatics analysis platform. iMeta 1 e36. (https:// doi.org/10.1002/imt2.36)
Shumaker DK, Solimando L, Sengupta K, Shimi T, Adam SA, Grunwald A, Strelkov SV, Aebi U, Cardoso MC & Goldman RD 2008 The highly conserved nuclear lamin Ig-fold binds to PCNA: its role in DNA replication. Journal of Cell Biology 181 269-280. (https://doi.org/10.1083/ jcb.200708155)
Sun S, Xu MZ, Poon RT, Day PJ & Luk JM 2010 Circulating Lamin B1 (LMNB1) biomarker detects early stages of liver cancer in patients. Journal of Proteome Research 9 70-78. (https://doi.org/10.1021/pr9002118)
Terzolo M, Baudin AE, Ardito A, Kroiss M, Leboulleux S, Daffara F, Perotti P, Feelders RA, deVries JH, Zaggia B, et al. 2013 Mitotane levels predict the outcome of patients with adrenocortical carcinoma treated adjuvantly following radical resection. European Journal of Endocrinology 169 263-270. (https://doi.org/10.1530/EJE-13-0242)
Wang J, Mi S, Ding M, Li X & Yuan S 2022 Metabolism and polarization regulation of macrophages in the tumor microenvironment. Cancer Letters 543 215766. (https://doi.org/10.1016/j.canlet.2022.215766)
Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, Feng T, Zhou L, Tang W, Zhan L, et al. 2021 clusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innovation 2 100141. (https://doi.org/10.1016/j. xinn.2021.100141)
Xu F, Guan Y, Zhang P, Xue L, Ma Y, Gao M, Chong T & Ren BC 2022 Tumor mutational burden presents limiting effects on predicting the efficacy of immune checkpoint inhibitors and prognostic assessment in adrenocortical carcinoma. BMC Endocrine Disorders 22 130. (https://doi. org/10.1186/s12902-022-01017-3)
Yan F, Powell DR, Curtis DJ & Wong NC 2020 From reads to insight: a hitchhiker’s guide to ATAC-seq data analysis. Genome Biology 21 22. (https://doi.org/10.1186/s13059-020-1929-3)
Yang Z, Sun Q, Guo J, Wang S, Song G, Liu W, Liu M & Tang H 2019 GRSF1-mediated MIR-G-1 promotes malignant behavior and nuclear autophagy by directly upregulating TMED5 and LMNB1 in cervical cancer cells. Autophagy 15 668-685. (https://doi.org/10.1080/15548627.2 018.1539590)
Zeng D, Ye Z, Shen R, Yu G, Wu J, Xiong Y, Zhou R, Qiu W, Huang N, Sun L, et al. 2021 IOBR: multi-omics immuno-oncology biological research to decode tumor microenvironment and signatures. Frontiers in Immunology 12 687975. (https://doi.org/10.3389/fimmu.2021.687975)
Zhong C, Long Z, Yang T, Wang S, Zhong W, Hu F, Teoh JY, Lu J & Mao X 2023 M6A-modified circRBM33 promotes prostate cancer progression via PDHA1-mediated mitochondrial respiration regulation and presents a potential target for ARSI therapy. International Journal of Biological Sciences 19 1543-1563. (https://doi.org/10.7150/ijbs.77133)