RESEARCH
High expression of GMNN predicts malignant progression and poor prognosis in ACC
Check for updates
Xinzhao Zhao1+, Xuezhou Zhang1+, Shixiu Shao1, Qingbo Yang1, Chengquan Shen1, Xuecheng Yang1, Wei Jiao1, Jing Liu2* and Yonghua Wang1*
Abstract
Background: Adrenocortical carcinoma (ACC) is a rare endocrine neoplasm, which is characterized by poor prog- nosis and high recurrence rate. Novel and reliable prognostic and metastatic biomarkers are lacking for ACC patients. This study aims at screening potential prognostic biomarkers and therapeutic targets of ACC through bioinformatic methods and immunohistochemical (IHC) analysis.
Methods: In the present study, by using the Gene Expression Omnibus (GEO) database we identified differentially expressed genes (DEGs) in ACC and validated these DEGs in The Cancer Genome Atlas (TCGA) ACC cohort. A DEGs- based signature was additionally constructed and we assessed its prognosis and prescient worth for ACC by survival analysis and nomogram. Immunohistochemistry (IHC) was used to verify the relationship between hub gene-GMNN expressions and clinicopathologic outcomes in ACC patients.
Results: A total of 24 DEGs correlated with the prognosis of ACC were screened from the TCGA and GEO databases. Five DEGs were subsequently selected in a signature which was closely related to the survival rates of ACC patients and GMNN was identified as the core gene in this signature. Univariate and multivariate Cox regression showed that the GMNN was an independent prognostic factor for ACC patients (P <0.05). Meanwhile, GMNN was closely related to the OS and PFI of ACC patients treated with mitotane (P <0.001). IHC confirmed that GMNN protein was over- expressed in ACC tissues compared with normal adrenal tissues and significantly correlated with stage (P=0.011), metastasis (P=0.028) and Ki-67 index (P=0.014).
Conclusions: GMNN is a novel tumor marker for predicting the malignant progression, metastasis and prognosis of ACC, and may be a potential therapeutic target for ACC.
Keywords: Adrenocortical carcinoma, GMNN, Signature, Prognosis, TCGA
+Xinzhao Zhao and Xuezhou Zhang are contributed equally to this work
*Correspondence: drliujing8712@126.com; wangyonghua@qdu.edu.cn
1 Department of Urology, The Affiliated Hospital of Qingdao University, Qingdao, Shandong, China
2 Department of Research Management and International Cooperation, The Affiliated Hospital of Qingdao University, Qingdao, Shandong, China
Introduction
Adrenocortical carcinoma (ACC) is an orphan endo- crine illness with an occurrence of between 0.5 and 2 for each million populace each year [1]. The majority of ACC patients had metastasis at the time of their initial diagnosis due to the challenges in early diagnosis, which leaves them with few options for treatment and very poor prognosis [2]. Patients with localized and regionalized ACC have a potential for cure with complete surgical
☒ BMC
@ The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
resection, but a mass of patients still develop recurrent or metastatic disease [3, 4]. Although mitotane is approved in the treatment of metastatic ACC, the therapeutic effectiveness and toxic side effects still limit the progno- sis of ACC [5]. Thus, it is very important to obtain reli- able biomarkers that can predict the prognosis and drug efficacy of ACC.
In recent years, bioinformatic analyses and high- throughput microarray technology have been broadly applied to distinguish genetic changes at the genome level, which can overcome the sample size issue and assist us to seek reliable tumor biomarkers. However, at present, there is still lack of effective, reliable and clini- cally applicable biomarkers, which can well predict the prognosis and therapeutic effect of ACC. In the present study, we screened differentially expressed genes (DEGs) related to the prognosis of ACC by bioinformatics analy- sis from TCGA and GEO databases and constructed an effective prognostic prediction signature. Moreover, we further identified the pivotal gene in the prognostic sig- nature and confirmed its potential value on predicting malignant progression, prognosis and mitotane efficacy in ACC.
Materials and methods
Data collection
RNA sequencing expression profiles of ACC patients were obtained from the TCGA database (https://portal. gdc.cancer.gov/, 92 ACC samples) and the GEO database (https://www.ncbi.nlm.nih.gov/geo/, 58 ACC samples and 5 normal samples). The concrete clinical informa- tion of TCGA ACC patients was presented in Additional file 2: Table S1.
Furthermore, the Affiliated Hospital of Qingdao Uni- versity (AHQU)-ACC cohort including 29 ACC tissues and 10 normal adrenal tissues were obtained from the Department of Urology, AHQU, P. R. China (January 2000 to October 2018). Informed consent was acquired from all patients and the research protocol was approved by the Ethics Committee of AHQU (No. QYFY WZLL 26,516). All the patients were received retroperitoneal laparoscopic resection of adrenal tumor and did not receive chemotherapy or radiotherapy before surgery. The clinical information of AHQU-ACC cohort was pre- sented in Table 1.
Identification, validation and consensus clustering analysis of DEGs in ACC
The DEGs between ACC and normal samples in the GEO cohort were screened using GEO2R. LogFC (fold change) >|1| and adj. P values <0.05 were considered to be statistically significant. Subsequently, we put over- lapped DEGs into the STRING (Search Tool for the
| Clinicopathological characteristics | Variable | Total (29) | GMNN expression | P value | |
|---|---|---|---|---|---|
| Low | High | ||||
| Age at diagnosis | <60 | 12 | 2 | 10 | |
| ≥60 | 17 | 8 | 9 | 0.291 | |
| Gender | Male | 18 | 7 | 11 | |
| Female | 11 | 3 | 8 | 0.548 | |
| Stage | I | 4 | 3 | 1 | |
| II | 2 | 1 | 1 | ||
| III | 5 | 3 | 2 | ||
| IV | 18 | 3 | 15 | 0.011 | |
| T | T1 | 4 | 3 | 1 | |
| T2 | 8 | 2 | 6 | ||
| T3 | 10 | 3 | 7 | ||
| T4 | 7 | 2 | 5 | 0.327 | |
| N | N0 | 18 | 8 | 10 | |
| N1 | 11 | 2 | 9 | 0.164 | |
| M | M0 | 12 | 7 | 5 | |
| M1 | 17 | 3 | 14 | 0.028 | |
| Ki-67 | <30 | 6 | 5 | 1 | |
| ≥30 | 8 | 1 | 7 | 0.014 | |
Retrieval of Interacting Gene) to formulate a PPI network [6]. Cytoscape (version 3.4.2) was used to draw a PPI net- work and identify the hub gene in the PPI network [7].
The expression levels of DEGs were further analyzed in TCGA cohort. To validate the correlation between the prognosis of ACC patients and the expression levels of DEGs, ACC patients were divided into various clusters by consensus clustering analysis with “ConsensusCluster- Plus”, and the Kaplan-Meier algorithm was performed to explore the survival results among various clusters.
Construction of a risk signature predicting prognosis for ACC patients
We performed the Univariate Cox regression analysis by using the R package “survival” and selection opera- tor (LASSO) Cox regression to establish the optimal DEGs-based signature for predicting the prognosis in ACC patients with the R package “glmnet”. The DEGs- based signature risk score⇒(Bi x Expi) (i=the number of genes). Kaplan-Meier curves was used to evaluate the survival differences between the high and low risk groups. The survival ROC package was used to reveal the receiver operating characteristic curve (ROC) in order to evaluate the signature’s prediction accuracy. Univariate and multivariate Cox regression analyses were adopted to evaluate the prognostic significance of this signature.
Besides, we constructed a prognostic nomogram to evaluate the probability of 1-, 2-, and 3-, 5-year overall survival (OS) for ACC patients with the R package. Cali- bration curves were used to examine the concordance between the predicted and the actual survival rates.
Validation of the hub gene associated with malignant progression and prognosis in ACC
The expression levels of hub gene-GMNN between nor- mal samples and ACC samples in the TCGA ACC cohort were analyzed by Wilcoxon test. Kaplan-Meier curves and Cox regression analyses were also performed to evaluate the correlation between GMNN expression and clinicopathological features and prognosis in the TCGA ACC cohort.
Immunohistochemistry (IHC) was used to detect the expression levels of GMNN protein in the AHQU-ACC cohort. After being deparaffinized and boiled in sodium citrate buffer (0.01 mol/L, PH=6.0) in the microwave for ten minutes as the use of antigen retrieval, tissue sec- tions incubated with anti-GMNN antibody [ab104306] (Abcam, USA, 1:100) overnight at 4 ℃. Biotinylated sec- ondary antibody (goat-anti-mouse-IgG, dilution 1:300) was then added for 30 min. Sections were stained by the avidin-biotin-peroxidase method and using 3-3’-diam- inobenzidine as chromogen. Normal skin and lymph node tissues were used as positive controls for GMNN antibodies. Negative controls were performed using non-immunized mouse serum, omitting the primary antibodies.
Five fields of view were randomly selected by patholo- gist independently. The pathologist estimated the score according to the degree of staining (0: weakly, 1: moder- ately, 2: strongly) and the percentage of staining (0: 0-5%, 1: 5-50%, 2: 50-100%), and further calculated the mean score of the five fields respectively [8]. The final immu- nostaining score was the sum of staining intensity score and staining percentage score. The final immunostaining score > 2 was considered high expression of GMNN.
Results
Identification of DEGs and construction of DEGs-based prognostic signature for ACC
457 DEGs were identified in the GEO-ACC cohort including 377 downregulated genes and 80 upregu- lated genes. Then, a total of 28 DEGs were screened in the most significant module by using Cytoscape. Among these 28 DEGs, 24 genes were most differen- tially expressed (Additional file 1: Fig. S1) and related to the OS of ACC patients in TCGA ACC cohort (Fig. 1A). Multivariate Cox and lasso regression analysis were then used to construct a DEGs-based prognostic signature, which contained 5 genes including CCNB2 (cyclin B2),
CDK1 (cyclin dependent kinase 1), DTL (denticleless E3 ubiquitin protein ligase homolog), GMNN (gemi- nin DNA replication inhibitor), and UBE2C (ubiquitin conjugating enzyme E2C). The coefficient of prognos- tic risk score was as follows:(0.0294 x the expression level value of CCNB2) +(0.0369 x the expression level value of CDK1)+(0.0762 x the expression level value of DTL)+(0.843 x the expression level value of GMNN+(0.0099 x the expression level value of UBE2C).
Survival analyses indicated that high-risk scores were associated with poor OS, disease-specific survival (DSS), and progression-free interval (PFI) of ACC (Fig. 1B-D). Furthermore, the risk score was positively correlated with the treatment efficacy of mitotane for ACC patients (P=7.666e-05, Fig. 1E). ROC curve analysis showed that the AUC was 0.849, which indicated excellent predictive accuracy of the prognostic signature (Fig. 1F). Moreover, the heatmap indicated that risk scores of the signature were closely related to tumor stage, T, and M in ACC patients (P<0.05, Fig. 1G).
The DEGs-based signature was an independent prognostic factor for ACC
Univariate and multivariate Cox regression analyses were used to assess the prognostic factors related to ACC including the DEGs-based signature and clinicopatho- logical features. Univariate Cox regression showed that stage (P<0.001), T (P<0.001), M (P<0.001), Weiss score (P<0.05), invasion (P<0.05), and risk scores (P<0.001) were significantly related to the OS of ACC patients (Fig. 1H). Multivariate Cox regression analysis confirmed that the DEG-based signature was an independent prog- nostic factor for ACC patients (P<0.001, Fig. 1I).
High expression of GMNN was correlated with clinicopathologic outcomes and poor prognosis in ACC patients
MCODE analysis indicated that GMNN was a hub gene in the most significant module (Fig. 2A). TCGA-ACC cohort showed that there was a significant difference in the expression of GMNN between ACC samples and normal samples and GMNN was significantly related to OS (P=6.318e-04), DSS (P=1.131e-03), and PFI (P=3.009e-03) for ACC patients (Fig. 2B-E). GMNN was also correlated with the stage, T and M in ACC and univariate and multivariate Cox regression showed that the GMNN was an independent prognostic factor for ACC patients (P<0.05, Fig. 2F-J). More importantly, GMNN was closely related to the OS (P<0.001) and PFI (P<0.001) of ACC patients treated with mitotane (Fig. 2K, L).
We further validated the protein expressions of GMNN in 29 ACC samples and 10 normal adrenal samples from
(See figure on next page.)
Fig. 1 Construction of a DEGs based signature to predict the prognosis of ACC. (A) Univariate Cox regression analysis showed that 23 DEGs are closely associated with the OS of ACC patients. (B-D) Kaplan-Meier curves revealed that the high-risk group had significantly shorter OS, DSS, and PFI compared with the low-risk group. (E) In addition, high-risk scores also indicate poor OS in ACC patients who received mitotane therapy. (F) ROC curves showed that the AUC of the risk score was 0.849. (G) Heatmap showed that a significant difference was found between the high- and low-risk groups for the stage, T, and M. (H) Univariate Cox regression analysis showed that stage, T, M, Weiss score, invasion, and risk score were significantly related to the survival of ACC patients. (I) Multivariate Cox regression analysis demonstrated that the signature could serve as an independent prognostic predictor in ACC. * P < 0.05; ** P < 0.01; *** P < 0.001; DEG differentially expressed gene, ACC adrenocortical carcinoma, OS overall survival, DSS disease-specific survival, PF/ progression-free interval, ROC receiver operating characteristic curve, AUC area under curve
AHQU ACC cohort by IHC. We observed that GMNN was aberrantly expressed in 65.52% of ACC patients and the protein expression level of GMNN was significantly higher in ACC tissues than that in normal adrenal tissues (P<0.05). In addition, in ACC cells, the immunoexpres- sions of GMNN were primarily in the nucleus (Fig. 3A- C). To evaluate the role of GMNN expression in ACC, the correlations between GMNN immunoreactivity and clin- icopathologic features were analyzed and summarized in Table 1. Interestingly, we found that GMNN expres- sion was significantly correlated with stage (P=0.011), M (P=0.028) and Ki67 index (P=0.014), indicating that GMNN may play an important role in the malignant pro- gression and metastasis of ACC.
Construction of a nomogram predicting prognosis for ACC To obtain a clinically applicable tool for monitoring the prognosis of ACC patients, we also constructed a nom- ogram containing the DEGs-based signature, GMNN expression levels and clinicopathologic features (includ- ing age, stage, gender, TNM, invasion, mitotane treat- ment, and Weiss score). Our results showed that the nomogram could well predict the 1-, 2-, 3-, 5-year OS for ACC patients (Fig. 4A). The calibration plots also dem- onstrated excellent coincidence between observation and prediction for the 1- and 3-year OS probabilities for ACC patients (Fig. 4B, C).
Discussion
ACC is characterized by low incidence rate, aggres- sive behaviors and poor prognosis. Thus, it is urgent to explore novel and reliable biomarkers which can well predict prognosis and provide potential therapeutic tar- get for ACC [9]. Previous studies have already identi- fied a number of potential hub genes in ACC by using bioinformatic methods. Guo et al. described a prog- nostic signature for ACC containing 9 genes, including CCNB1 (cyclin B1), CDK1, TOP2A (topoisomerase IIa), CCNA2 (cyclin A2), CDKN3 (cyclin-dependent kinase inhibitor 3), MAD2L1 (mitosis arrest-deficient 2-like 1), RACGAP1 (rac GTPase activating protein 1), BUB1
(benzimidazole 1 homolog beta) and CCNB2 [10]. Xu et al. suggested that ZWINT (ZW10 interactor), PRC1 (protein regulator of cytokinesis 1), CDKN3 (cyclin- dependent kinase inhibitor 3), CDK1 and CCNA2 were involved in early recurrence of ACC [2]. In our study, we identified 24 DEGs correlated with prognosis of ACC and constructed a five gene (CCNB2, CDK1, DTL, GMNN, and UBE2C)-based prognostic signature. Different from previous studies, for the first time we found that GMNN was the core gene in this signature, which was closely correlated with clinicopathologic outcomes and progno- sis of ACC.
GMNN is an important gene for cell cycle regulation. GMNN encodes a protein inhibits histone acetyltrans- ferase activity of KAT7/HBO1 in a CDT1-dependent manner, inhibiting DNA replication licensing and his- tone H4 acetylation, enrolling them in cell proliferative control. Previous studies have demonstrated that GMNN expression was correlated with prognosis in oral squa- mous cell carcinoma, squamous cell carcinoma of the tongue, and melanoma [11-13]. Blanchard et al. found that upregulated GMNN acted as an oncogene that pro- motes cytokinesis failure and the progression of breast tumors and thus became an effective therapeutic tar- get for aggressive breast cancer [14]. However, relevant research on GMNN in ACC is still limited. In the pre- sent study, we found that GMNN was significantly over- expressed in ACC tissues and was closely related to the pathological features and prognosis. Previous studies demonstrated that high GMNN/Ki-67 ratio indicated a short G1 phase and a high rate of cell proliferation [15- 18]. Interestingly, we found that GMNN expression was significantly correlated with Ki-67 index in ACC. Fur- thermore, inhibiting GMNN activity could selectively arrest cancer cell proliferation by inducing DNA dam- age-mediated apoptosis and DNA re-replication without affecting the normal cells [19]. Thus, we speculate that GMNN may promote the malignant progression of ACC by regulating the cell cycle and promoting cell prolif- eration. However, the underlying mechanism of GMNN
A
P-value
Hazard ratio
B
OS
C
DSS
ANLN
<0.001
1.133(1.074-1.194)
BIRC5
<0.001
CCNB1
1.181(1.123-1.241)
<0.001
1.068(1.046-1.091)
Risk
High risk
Low risk
Risk
High risk
Low risk
CCNB2
<0.001
1.185(1.124-1.250)
1.00
1.00-
CDK1
<0.001
1.143(1.099-1.188)
CDKN3
<0.001
1.145(1.088-1.205)
Survival probability
Survival probability
CENPU
<0.001
1.151(1.093-1.213)
0.75
0.75
DEPDC1B
0.117
1.082(0.981-1.194)
DTL
<0.001
1.686(1.439-1.975)
GINS1
<0.001
1.457(1.251-1.697)
0.50
0.50
GMNN
<0.001
1.087(1.047-1.129)
HMMR
<0.001
KIF11
1.210(1.088-1.346)
<0.001
1.221(1.145-1.302)
MAD2L1
0.25
0.25
<0.001
1.294(1.153-1.453)
p=8.798e-08
p=2.113e-07
NCAPG
<0.001
1.384(1.242-1.542)
PARPBP
<0.001
3.320(2.214-4.979)
PRC1
<0.001
0.00
0.00
RACGAP1
1.118(1.068-1.170)
<0.001
1.182(1.121-1.247)
0
1
2
3
4
5
6
7
8
9
10
11
12
0
1
2
3
4
5
6
7
8
9
10
11
12
RAD51AP1
<0.001
1,481(1.304-1.683)
Time(years)
Time(years)
RFC4
<0.001
1.225(1.129-1.328)
TOP2A
<0.001
UBE2C
1.091(1.058-1.125)
<0.001
1.042(1.029-1.055)
UBE2T
<0.001
1.059(1.039-1.080)
Risk
High risk -
40
36
21
14
8
6
3
2
2
UN
0
NO
1.103(1.072-1.135)
Low risk
40
40
37
30
22
18
13
a
O
4
NO
Risk
High risk
38
34
20
37
13
7
30
6
3
Low risk
40
40
18
13
9
2
O
5
0
4
0
2
0 2
ZWINT
<0.001
2
22
1.0
2.0
3.0
0
1
2
3
4
5
6
7
8
9
10
11
12
0
1
2
3
4
5
6
7
8
9
10
11
12
Hazard ratio
Time(years)
Time(years)
D
PFI
E
OS (mitotane)
F
ROC curve ( AUC = 0.849 )
Risk
High risk
Low risk
1.00-
Risk
High risk
Low risk
1.00
1.0
Survival probability
0.75
Survival probability
0.75
0.8
0.50
0.50
True positive rate
0.6
0.25
p=5.026e-06
0.25
p=7.666e-05
0.4
0.00
0.00
0
1
2
3
4
5
6
7
8
9
10
11
12
0
1
2
3
4
5
6
7
8
9
10
11
12
Time(years)
Time(years)
0.2
Risk
High risk Low risk
40 20
10
5
4
3
1
1
5
1
WO
NO
NO
Risk
High risk-
24
13
9
5
22
11
4
Low risk
12
4
1
0
O
O
40
36
31
16
15
10
0
10
23
21
15
8
4
2
1
0.0
0
1
2
3
4
5
6
7
8
9
10
11
12
0
1
2
3
4
5
6
7
8
9
10
11
12
Time(years)
Time(years)
0.0
0.2
0.4
0.6
0.8
1.0
False positive rate
G
Risk
6
Risk
T ***
high
stage **
N
4
low
gender
T ***
M*
2
T1
age
T2
0
T3
T4
UBE2C
-2
stage **
-4
Stage I
Stage II
-6
Stage III
Stage IV
GMNN
N
NO
N1
gender
CCNB2
FEMALE
MALE
M*
MO
M1
CDK1
age
<49
>=49
DTL
1
I
P-vlaue
Hazard ratio
I
age
0.073
1.035(0.997-1.074)
gender
0.606
1.369(0.416-4.507)
stage
0.981
1.021(0.172-6.078)
T
0.137
2.566(0.740-8.890)
N
0.074
4.502(0.864-23.457)
1
M
0.632
0.577(0.061-5.488)
I
weiss score
0.635
1.081(0.783-1.495)
invasion
0.783
1.249(0.257-6.076)
mitotane therapy
0.544
0.676(0.191-2.394)
risk score
<0.001
1.175(1.076-1.284)
0
5
10
15
0
5
10
15
20
Hazard ratio
Hazard ratio
Fig. 1 (See legend on previous page.)
| H | P-value | Hazard ratio |
|---|---|---|
| age | 0.257 | 1.014(0.990-1.040) |
| gender | 0.057 | 2.295(0.976-5.394) |
| stage | <0.001 | 2.726(1.670-4.451) |
| T | <0.001 | 3.245(1.920-5.483) |
| N | 0.264 | 1.857(0.626-5.508) |
| M | <0.001 | 6.770(2.651-17.291) |
| weiss score | 0.007 | 1.350(1.084-1.682) |
| invasion | 0.016 | 3.331(1.246-8.905) |
| mitotane therapy | 0.151 | 2.091(0.765-5.719) |
| risk score | <0.001 | 1.192(1.118-1.271) |
A
RACGAP
AURKA
MADZLI
COKN3
PARPEP
B
8
p=0.012
RFCA
TOPZA
CENPU
CDK1
NCAPG
GMNN expression
8
MCM7
BIRCS
DEPDC18
ZWINT
GMNN
4
UBEZT
UBEZC
CONB1
DTL
2
ANUN
RADS1AP
GINS1
CCNB2
0
CKS2
KIAA0101
PRC1
KIF11
HMMR
Normal
Tumor
C
OS
DSS
E
PFI
expression
High
Low
expression
High
*
1.00-
1.00
Low
1.00-
expression
High
Low
Survival probability
0.75
Survival probability
0.75
Survival probability
0.75
0.50
0.50
0.50
0.25
p=6.318e-04
0.25
p=1.131e-03
0.25
p=3.009e-03
0.00
0.00
0.00
0
1
2
3
4
5
6
7
8
9
10
11
12
0
1
2
3
4
5
6
7
8
9
10
11
12
0
1
2
3
4
5
6
7
8
9
10
11
12
Expression
Time(years)
Time(years)
Time(years)
High
39 40
35
22
18
12 18
Expression
Expression
11
6
3
8
UU
3
A
1
High Low
37 40
33
21
66
17
11
11
6
Low
3
High
39
10
8
7
4
40
36
26
13
10
40
36
26
18
13
10
3
che
1
40
34
A
Low
1
4
28
12
11
8
4
2
0
1
2
3
4
5
6
7
8
9
10
1
12
0
1
2
3
4
5
6
7
8
9
10
11
12
0
1
2
3
4
5
6
7
8
9
10
11
Time(years)
12
Time(years)
Time(years)
F
P-value
Hazard ratio
G
age
0.257
1.014(0.990-1.040)
P-value
Hazard ratio
gender
0.057
2.295(0.976-5.394)
oBe
0.133
1.028(0.992-1.067)
stage
<0.001
2.726(1.670-4.451)
gender
0.032
3.998(1.124-14.213)
T
<0.001
3.245(1.920-5.483)
stage
0.887
0.880(0.152-5.105)
H
1.857(0.626-5.508)
H
T
0.264
0.096
3.084(0.819-11.610)
N
M
<0.001
6.770(2.651-17.291)
N
0.159
3.003(0.650-13.881)
M
1.350(1.084-1.682)
0.874
0.850(0.115-6.298)
weiss score
0.007
weiss score
0.087
0.738(0.522-1.045)
invasion
0.016
3.331(1.246-8.905)
invasion
0.055
4.973(0.969-25.535)
mitotane therapy
0.151
2.091(0.765-5.719)
mitotane thearpy
0.327
0.530(0.149-1.887)
H
GMNN
<0.001
1.077(1.033-1.122)
GMNN
0.002
1.129(1.047-1.218)
0
5
10
15
0
5
10
15
20
25
Hazard ratio
Hazard ratio
H
80
p=1.76e-04
8
p=0.001
J
8
p=0.007
GMNN expression
80
GMNN expression
80
GMNN expression
80
4
4
:
4
..
..
20
20
20
&
7
0
0
0
stage I-II
stage III-IV
T1-2
T3-4
MO
M1
K
1.00-
L
OS(mitotane)
1.00
PFI(mitotane)
GMNN
GMNN
Low
Low
Survival probability
0.75
High
Survival probability
0.75
High
0.50
0.50
0.25
p<0.001
0.25
<0
901
0.00
0.00
0
1
2
3
4
5
6
7
8
9
10
11
12
0
1
2
3
4
5
6
1
8
9
10
11
12
Time(years)
Time(years)
Number at risk
Number at risk
GMNN
Low
34
34 28
19
14
10
5
2
1
1
1
1
1
GMNN
Low
30
24
18
9
8
7
4
2
1
1
1
1
14
11
5
4
2
2
1 0
0
0
0
0
0
0
0
High
18
7
2
2
1
1
0
0
0
Q
0
0
0
1
2
3
4
5
6
7
8
9
10
11
Time(years)
12
0
1
2
3
4
5
6
7
8
9
10
11
12
Time(years)
A
10x
20x
Normal tissue
B
10x
1
20x
C
Non-metastatic ACC tissue
10x
.
20x
Metastatic ACC tissue
A
0
10
20
30
40
50
60
70
80
90
100
Points
age
10
20
30
40
50
60
70
80
MALE
gender
FEMALE
stage
Stage III-IV
Stage I-II
T3-4
T
T1-2
N1
N
NO
MO
M
M1
mitotane_therapy
NO
YES
Present
Invasion
Absent
GMNN
0
5
10
20
30
40
8
weiss_score
MY
9
risk score
0
5
10
15
20
25
30
35
Total Points
0
20
40
60
80
100
120
140
160
180
200
220
1-year survival
0.99
0.9
0.8 0.70.60.50.40.30.20.10.05
2-year survival
0.99
0.9
0.8 0.70.60.50.40.30.20.10.05
3-year survival
0.99
0.9
0.8 0.70.60.50.40.30.20.10.05
5-year survival
0.99
0.9 0.8 0.70.60.50.40.30.20.1
B
1.0
C
1.0
0.8
0.8
Actual 1-Year Survival
0.6
Actual 3-Year Survival
0.6
0.4
0.4
0.2
0.2
0.0
0.0
0.0
0.2
0.4
0.6
0.8
1.0
0.0
0.2
0.4
0.6
0.8
1.0
Nomogram-Predicted Probability of 1-Year Survival
Nomogram-Predicted Probability of 3-Year Survival
Fig. 4 A prognostic nomogram with clinical features, GMNN expression levels, and DEGs-based signature risk score was established for ACC. (A) The nomogram could superiorly predict 1-, 2-, and 3-, 5-year OS of ACC patients. (B-C) The calibration plots also demonstrate excellent agreement between prediction and observation for the 1- and 3-year OS probabilities of the TCGA ACC patients. DEG differentially expressed gene, ACC adrenocortical carcinoma, OS overall survival
affecting the prognosis of ACC still needs to be further studied.
Mitotane is the only available drug approved for the treatment of advanced ACC. Considering the toxicity and side effects, it is very important to screen ACC patients who can benefit from mitotane [20, 21]. However, there is still a lack of biomarkers for predicting the efficacy of mitotane in ACC. Another interesting finding in our study was that GMNN showed close relation with the OS and PFI of ACC patients treated with mitotane. Moreo- ver, a nomogram containing GMNN expression levels and mitotane treatment was constructed and showed a relatively high accuracy of prediction. These findings indicated that GMNN could be used as a potential bio- marker for predicting mitotane efficacy in ACC.
There are still some limitations in this study. First, as a retrospective study, it is inevitable to have possible selec- tion bias and low statistical ability; secondly, as this study mainly aimed to explore the potential clinical values of selected hub genes in the diagnosis and therapy of ACC, the details of their mechanisms were not comprehen- sively explored, especially GMNN with very limited num- ber of previous studies of its effect on ACC. In addition, bioinformatics research based on public database needs further experimental research to explore its potential role and deep-seated molecular mechanism.
In conclusion, we constructed a five DEGs (CCNB2, CDK1, DTL, GMNN, and UBE2C)-based prognostic signature for ACC and first identified GMNN as a novel tumor biomarker for predicting the malignant progres- sion, mitotane efficacy, and prognosis of ACC.
Abbreviations
ACC: Adrenocortical carcinoma; IHC: Immunohistochemical; GMNN: Geminin DNA replication inhibitor; DEGs: Differentially expressed genes; GEO: Gene expression omnibus; TCGA: The cancer genome atlas; PPI: Protein-protein interactions; OS: Overall survival; DSS: Disease-specific survival; PFI: Progres- sion-free interval; CCNB2: Cyclin B2; CDK1: Cyclin dependent kinase 1; DTL: Denticleless E3 ubiquitin protein ligase homolog; UBE2C: Ubiquitin conjugat- ing enzyme E2 C; AUC: Area under the curve; STRING: Search tool for the retrieval of interacting gene; K-M: Kaplan-Meier; ENSAT: European network for the study of adrenal tumors; LASSO: Least absolute shrinkage and selection operator; ROC: Receiver operating characteristic curve; CCNB1: Cyclin B1; TOP2A: Topoisomerase Ila; CCNA2: Cyclin A2; CDKN3: Cyclin-dependent kinase inhibitor 3; MAD2L1: Mitosis arrest-deficient 2-like 1; RACGAP1: Rac GTPase activating protein 1; BUB1: Benzimidazole 1 homolog beta; ZWINT: ZW10 interactor; PRC1: Protein regulator of cytokinesis 1; CDKN3: Cyclin-dependent kinase inhibitor 3.
Supplementary Information
The online version contains supplementary material available at https://doi. org/10.1186/s40001-022-00950-2.
Additional file 1. Differential clinical outcomes of ACC patients in the two different clusters. Heatmap was generated to validate the expression levels of DEGs between ACC and normal samples in the TCGA cohort. ACC adrenocortical carcinoma.
Additional file 2: Table S1. TCGA-adrenocortical carcinoma (ACC) patient characteristics.
Author contributions
XZZ and XZZ: conceptualization, methodology, writing-original draft. SXS: data curation. QBY and CQS: validation, visualization. WJ and XCY: software. JL: supervision. YHW: writing-review and editing. All authors read and approved the final version of the manuscript.
Funding
This work was supported by the National Natural Science Foundation of China (Nos. 81972378, 81101932).
Availability of data and materials
The data used to support the findings of this study is included within the arti- cle, and the data are available from the corresponding author upon request.
Declarations
Ethics approval and consent to participate
The studies involving human participants were reviewed and approved by Ethical review committees of The Affiliated Hospital of Qingdao University (No. QYFY WZLL 26516). The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.
Competing interests
The authors declare that they have no conflict of interest.
Received: 13 September 2022 Accepted: 14 December 2022 Published online: 20 December 2022
References
1. Calissendorff J, Calissendorff F, Falhammar H. Adrenocortical cancer: mortality, hormone secretion, proliferation and urine steroids-experi- ence from a single centre spanning three decades. BMC Endocr Disord. 2016;16:15.
2. Xu WH, Wu J, Wang J, Wan FN, Wang HK, Cao DL, et al. Screening and identification of potential prognostic biomarkers in adrenocortical carci- noma. Front Genet. 2019;10:821.
3. Fassnacht M, Johanssen S, Quinkler M, Bucsky P, Willenberg HS, Beuschlein F, et al. Limited prognostic value of the 2004 international union against cancer staging classification for adrenocortical carcinoma: proposal for a revised TNM classification. Cancer. 2009;115(2):243-50.
4. Schulick RD, Brennan MF. Long-term survival after complete resection and repeat resection in patients with adrenocortical carcinoma. Ann Surg Oncol. 1999;6(8):719-26.
5. Bedrose S, Daher M, Altameemi L, Habra MA. Adjuvant therapy in adrenocortical carcinoma: reflections and future directions. Cancers. 2020;12(2):508.
6. Szklarczyk D, Morris JH, Cook H, Kuhn M, Wyder S, Simonovic M, et al. The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res. 2017;45(D1):D362-8.
7. Smoot ME, Ono K, Ruscheinski J, Wang PL, Ideker T. Cytoscape 28: new features for data integration and network visualization. Bioinformatics. 2011;27(3):431-2.
8. Taheri D, Zahavi DJ, Del Carmen RM, Meliti A, Rezaee N, Yonescu R, et al. For staining of ALK protein, the novel D5F3 antibody demonstrates superior overall performance in terms of intensity and extent of staining in comparison to the currently used ALK1 antibody. Virchows Arch. 2016;469(3):345-50.
9. Assié G, Letouzé E, Fassnacht M, Jouinot A, Luscap W, Barreau O, et al. Integrated genomic characterization of adrenocortical carcinoma. Nat Genet. 2014;46(6):607-12.
10. Guo J, Gu Y, Ma X, Zhang L, Li H, Yan Z, et al. Identification of hub genes and pathways in adrenocortical carcinoma by integrated bioinformatic analysis. J Cell Mol Med. 2020;24(8):4428-38.
11. Jiang Q, Yu YC, Ding XJ, Luo Y, Ruan H. Bioinformatics analysis reveals sig- nificant genes and pathways to target for oral squamous cell carcinoma. Asian Pacific J Cancer Prev. 2014;15(5):2273-8.
12. Quan L, Shi J, Tian Y, Zhang Q, Zhang Y, Zhang Y, et al. Identification of potential therapeutic targets for melanoma using gene expression analy- sis. Neoplasma. 2015;62(5):733-9.
13. Zhou RS, Zhang EX, Sun QF, Ye ZJ, Liu JW, Zhou DH, et al. Integrated analysis of lncRNA-miRNA-mRNA ceRNA network in squamous cell carci - noma of tongue. BMC Cancer. 2019;19(1):779.
14. Blanchard Z, Malik R, Mullins N, Maric C, Luk H, Horio D, et al. Geminin overexpression induces mammary tumors via suppressing cytokinesis. Oncotarget. 2011;2(12):1011-27.
15. Cates JM, Memoli VA, Gonzalez RS. Cell cycle and apoptosis regulatory proteins, proliferative markers, cell signaling molecules, CD209, and decorin immunoreactivity in low-grade myxofibrosarcoma and myxoma. Virchows Arch. 2015;467(2):211-6.
16. de Andrade BA, León JE, Carlos R, Delgado-Azañero W, Mosqueda- Taylor A, de Almeida OP. Expression of minichromosome maintenance 2, Ki-67, and geminin in oral nevi and melanoma. Ann Diagn Pathol. 2013;17(1):32-6.
17. Kushwaha PP, Rapalli KC, Kumar S. Geminin a multi task protein involved in cancer pathophysiology and developmental process: a review. Bio- chimie. 2016;131:115-27.
18. Wharton SB, Hibberd S, Eward KL, Crimmins D, Jellinek DA, Levy D, et al. DNA replication licensing and cell cycle kinetics of oligodendroglial tumours. Br J Cancer. 2004;91(2):262-9.
19. Zhu W, Depamphilis ML. Selective killing of cancer cells by suppression of geminin activity. Can Res. 2009;69(11):4870-7.
20. Shariq OA, Mckenzie TJ. Adrenocortical carcinoma: current state of the art, ongoing controversies, and future directions in diagnosis and treat- ment. Ther Adv Chronic Dis. 2021;12:20406223211033104.
21. Terzolo M, Fassnacht M. Endocrine tumours: our experience with the management of patients with non-metastatic adrenocortical carcinoma. Eur J Endocrinol. 2022;187(3):R27-r40.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations.
Ready to submit your research? Choose BMC and benefit from:
· fast, convenient online submission
· thorough peer review by experienced researchers in your field
· rapid publication on acceptance
· support for research data, including large and complex data types
· gold Open Access which fosters wider collaboration and increased citations
· maximum visibility for your research: over 100M website views per year
At BMC, research is always in progress.
BMC
Learn more biomedcentral.com/submissions