EnM ENDOCRINOLOGY AND METABOLISM
Original Article
Check for updates
Protein Expression of Cyclin B1, Transferrin Receptor, and Fibronectin Is Correlated with the Prognosis of Adrenal Cortical Carcinoma
Sun Joon Moon1, Jung Hee Kim1,2, Sung Hye Kong’, Chan Soo Shin1
1Department of Internal Medicine, Seoul National University Hospital, Seoul National University College of Medicine;
2Department of Molecular Medicine and Biopharmaceutical Sciences, Graduate School of Convergence Science and Technology, Seoul National University College of Medicine, Seoul, Korea
Background: Adrenal cortical carcinoma (ACC) is a rare cancer with a variable prognosis. Several prognostic factors of ACC have been previously reported, but a proteomic analysis has not yet been performed. This study aimed to investigate prognostic biomark- ers for ACC using a proteomic approach.
Methods: We used reverse-phase protein array data from The Cancer Proteome Atlas, and identified differentially expressed pro- teins in metastatic ACCs. Multivariate Cox regression analysis adjusted by age and staging was used for survival analysis, and the C- index and category-free net reclassification improvement (cfNRI) were utilized to evaluate additive prognostic value.
Results: In 46 patients with ACC, cyclin B1, transferrin receptor (TfR1), and fibronectin were significantly overexpressed in pa- tients with distant metastasis. In multivariate models, high expression of cyclin B1 and TfR1 was significantly associated with mor- tality (hazard ratio [HR], 6.13; 95% confidence interval [CI], 1.02 to 36.7; and HR, 6.59; 95% CI, 1.14 to 38.2; respectively), where- as high fibronectin expression was not (HR, 3.92; 95% CI, 0.75 to 20.4). Combinations of high cyclin B1/high TfR1, high cyclin B1/ high fibronectin, and high TfR1/high fibronectin were strongly associated with mortality ([HR, 13.72; 95% CI, 1.89 to 99.66], [HR, 9.22; 95% CI, 1.34 to 63.55], and [HR, 18.59; 95% CI, 2.54 to 135.88], respectively). In reclassification analyses, cyclin B1, TfR1, fibronectin, and combinations thereof improved the prognostic performance (C-index, 0.78 to 0.82-0.86; cfNRI, all P values <0.05). Conclusion: In ACC patients, the overexpression of cyclin B1, TfR1, and fibronectin and combinations thereof were associated with poor prognosis.
Keywords: Adrenocortical carcinoma; Protein array analysis; Proteomics; Prognosis
| INTRODUCTION | prognosis, with a 5-year overall survival rate of less than 40% for all cancers and 10% for metastatic cancers [2]. However, |
| Adrenal cortical carcinoma (ACC) is a rare cancer, with an an- | even in metastatic ACC, the prognosis is quite variable, with re- |
| nual incidence of about 1-2 per million [1]. ACC has a poor | ported survival ranging from a few months to more than 10 |
| Received: 7 October 2019, Revised: 25 November 2019, | Copyright @ 2020 Korean Endocrine Society |
| Accepted: 13 January 2020 | This is an Open Access article distributed under the terms of the Creative Com- |
| Corresponding author: Jung Hee Kim | mons Attribution Non-Commercial License (https://creativecommons.org/ |
| Department of Internal Medicine, Seoul National University Hospital, Seoul | licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribu- |
| National University College of Medicine, 101 Daehak-ro, Jongno-gu, Seoul 03080, Korea | tion, and reproduction in any medium, provided the original work is properly cited. |
| Tel: +82-2-2072-4839, Fax: +82-2-764-2199, E-mail: jungheeendo@gmail.com |
EnM
years, suggesting the heterogeneity of these tumors [3]. Tumor, node, metastasis (TNM) staging has been used as a predictor of survival and has been modified to improve its prognostic power; currently, the eighth edition of TNM staging and European Net- work for the Study of Adrenal Tumors (ENSAT) staging sys- tems are used [4-6]. Nevertheless, a significant proportion of patients with localized disease according to TNM staging expe- rience recurrence after surgery (up to 70% within 3 years), which is the only curative treatment to date [2]. Tumor grade, as assessed by mitotic count and proliferation indices such as the Ki-67 index, has shown good predictive value for localized ACC [7]. However, tumor grade still has the limitation of poor reproducibility [8]. Molecular studies have focused on several potential driver genes, but their independent prognostic values are controversial [9,10]. Recently, distinct molecular subgroups have been identified via pan-genomic approaches using DNA copy number, mRNA expression, miRNA expression, and DNA methylation, and these subgroups have shown associations with different survival outcomes [11,12]. Moreover, this classifica- tion demonstrated high prognostic value when combining with clinical staging and tumor grade [13]. However, the pan-genom- ic approach is quite complex and not easy to use in clinical practice. In addition, pan-genomic biomarkers had limited value in prognostic performance for metastatic ACC [13].
Few studies have investigated the use of a proteomic ap- proach to predict the prognosis of ACC. Moreover, protein markers can be useful in clinical practice since they can be straightforwardly analyzed using immunohistochemistry or Western blots. Therefore, we aimed to investigate prognostic biomarkers for ACCs using a proteomic approach.
METHODS
Study subjects and data source
In this study, ACC datasets from The Cancer Proteome Atlas (TCPA) database were extracted from the cBioPortal website (http://www.cbioportal.org/) using The Cancer Genome Atlas (TCGA) provisional dataset [14]. We obtained data from re- verse-phase protein arrays (RPPA), an antibody-based quantita- tive method assessing multiple protein markers in a cost-effec- tive, sensitive, and high-throughput manner [14]. RPPA data that had been normalized using the z-score, as provided in cBio- Portal, were used. Among the 91 ACC patients in the TCGA da- tabase, we included 46 patients with RPPA data. Clinical data, including staging (ENSAT), treatment, and survival informa- tion, were also obtained from the same source [5].
Differential protein expression analysis
Differential protein expression analysis of the RPPA data was performed using the Perseus software (http://www.perseus- framework.org). Differentially expressed proteins (DEPs) in metastatic and non-metastatic ACCs were identified, and the false-discovery rate (FDR) was controlled using the Benjamini- Hochberg method. The list of DEPs was limited to proteins showing a fold-change of equal to or greater than +2, or equal to or less than -2, and an FDR lower than 0.05. Protein expression above the median value was defined as high expression.
Statistical analysis
Kaplan-Meier analysis and the log-rank test were used to con- struct the survival curve. Cox proportional-hazards regression analyses were used to calculate the hazard ratios (HRs) of DEPs for all-cause mortality. Age and staging were adjusted in the multivariable Cox regression analyses. To determine the incre- mental discriminative value of DEPs additive to age and staging for prognosis, we calculated C-statistics and the category-free net reclassification improvement (cfNRI). In addition, we per- formed a subgroup analysis of non-metastatic patients. All P values were two-sided, and P values of lower than 0.05 were considered to indicate statistical significance in all analyses. The statistical analyses were performed using STATA software ver- sion 13 (StataCorp, College Station, TX, USA; https://www.sta- ta.com) and R version 3.6.1 (R Foundation, Vienna, Austria; https://www.r-project.org).
Ethical statement
Written informed consent by subjects was waived due to the public database. All data were anonymized, and this study re- ceived Institutional Review Board (IRB) approval at Seoul Na- tional University Hospital (IRB No. E-1910-005-1067).
RESULTS
Baseline characteristics and clinical outcomes
The baseline characteristics of the participants are shown in Ta- ble 1. Their mean age was 47.7±14.4 years, and 18 (39.1%) were male. Thirty-five (76.1%) were white, two (4.4%) were non-white, and the others’ racial information was unknown. The number of patients with ENSAT stage I, II, III, and IV disease was two (4.3%), 26 (56.5%), 10 (21.7%), and eight (17.4%), re- spectively. Complete resection was performed in 33 (71.7%) patients, of whom 11 experienced recurrence. Thirty-two pa- tients (69.6%) received adjuvant chemotherapy after surgery.
EnM
| Table 1. Baseline Characteristics of Adrenal Cortical Carcinoma Patients (n=46) in the TCPA Database | ||||
|---|---|---|---|---|
| Characteristic | No metastasis (n=38) | Metastasis (n=8) | Total (n=46) | P valueª |
| Age, yr | 47.6±14.8 | 48.0±13.1 | 47.7±14.4 | 0.948 |
| Male sex | 18 (47.4) | 0 | 18 (39.1) | 0.015 |
| Race | 0.440 | |||
| White | 27 (71.1) | 8 (100) | 35 (76.1) | |
| Black | 1 (2.6) | 0 | 1 (2.2) | |
| Asian | 1 (2.6) | 0 | 1 (2.2) | |
| Unknown | 9 (23.7) | 0 | 9 (19.6) | |
| Survival (alive) | 30 (78.9) | 2 (25.0) | 32 (69.6) | 0.006 |
| Follow-up, yr | 3.3 (1.3-5.3) | 2.2 (1.2-3.7) | 3.1 (1.3-5.3) | 0.041 |
| Staging (ENSAT) | <0.001 | |||
| I | 2 (5.3) | 0 | 2 (4.3) | |
| II | 26 (68.4) | 0 | 26 (56.5) | |
| III | 10 (26.3) | 0 | 10 (21.7) | |
| IV | 0 | 8 (100) | 8 (17.4) | |
| Complete resection | 32 (84.2) | 1 (12.5) | 33 (71.7) | <0.001 |
| Chemotherapy | 25 (65.8) | 7 (87.5) | 32 (69.6) | 0.403 |
| Recurrence after complete resectionb | 11 (34.4) | 0 | 11 (33.3) | 1.000 |
| Weiss score℃ | 5.8±1.7 | 6.1±3.0 | 5.9±2.0 | 0.503f |
| Mitotic countd | ||||
| Count/HPF | 9.8±8.2 | 33.0±39.9 | 13.6±18.8 | 0.589f |
| No. of >20/HPF | 2 (7.7) | 2 (28.6) | 4 (12.1) | 0.190 |
| Cortisol secretione | 15 (42.9) | 5 (62.5) | 20 (46.5) | 0.440 |
| Differentially expressed proteins | ||||
| High cyclin B1 | 16 (42.1) | 7 (87.5) | 23 (50.0) | 0.047 |
| High TfR1 | 17 (44.7) | 6 (75.0) | 23 (50.0) | 0.243 |
| High fibronectin | 15 (39.5) | 8 (100) | 23 (50.0) | 0.004 |
| High cyclin B1+TfR1 | 11 (28.9) | 6 (75.0) | 17 (37.0) | 0.038 |
| High cyclin B1+fibronectin | 7 (18.4) | 7 (87.5) | 14 (30.4) | <0.001 |
| High TfR1+fibronectin | 8 (21.1) | 6 (75.0) | 14 (30.4) | 0.006 |
| High cyclin B1+TfR1+fibronectin | 7 (18.4) | 6 (75.0) | 13 (28.3) | 0.004 |
Values are expressed as mean ± standard deviation, number (%), or median (interquartile range).
TCPA, The Cancer Proteome Atlas; ENSAT, European Network for the Study of Adrenal Tumors; HPF, high power field; TfR1, transferrin receptor. aP value for a comparison between patients with metastasis and without metastasis; bThe denominator was patients with complete resection; “The avail- able number was 32 because of missing values; dThe available number was 33 because of missing values; “The available number was 43 because of miss- ing values; “The Mann-Whitney test was used, because the Weiss score and mitotic count did not follow normal distributions.
The patients’ mean Weiss score was 5.9±2.0.
Differential expression of proteins
The expression levels of 198 proteins in the RPPA data are shown in Supplemental Table S1. In patients with distant metas- tasis, three proteins were found to be upregulated (fold-change ≥2)- CCNB1 (cyclin B1), TFRC (transferrin receptor [TfR1]),
and FN1 (fibronectin)-and none was found to be downregulat- ed (fold-change ≤-2) A volcano plot of DEPs according to me- tastasis is presented in Fig. 1.
Survival analyses
During the median follow-up period of 3.1 years (interquartile range, 1.3 to 5.3), mortality occurred in 14 patients (31.4%).
Cyclin B1
3.5
TfR1
· Fibronectin
3.0
2.5
-Log10 (P value)
2.0
1.5
1.0
0.5
0
-1
-0.5
0
0.5
1
Log2 (fold change)
Kaplan-Meier curves for each of the above-mentioned DEPs and combinations thereof are shown in Fig. 2. The survival rate was significantly lower in patients with high expression levels of each of these three proteins than in those with low expression levels (overall log-rank P value <0.001 for all). In addition, sig- nificantly lower survival rates were seen for all combinations of these proteins (high cyclin B1/high TfR1, high cyclin B1/high fibronectin, high TfR1/high fibronectin, high cyclin B1/high TfR1/high fibronectin; overall log-rank P value <0.001 for all).
In the unadjusted Cox regression analyses, high expression of each protein and all combinations thereof were significantly as- sociated with mortality (Table 2). In multivariate Cox regression analyses adjusted for age and staging, high expression levels of cyclin B1 and TfR1 were significantly associated with high mortality. However, the expression level of fibronectin failed to predict mortality in the multivariate Cox regression model. All combinations of each protein expression signature (high cyclin B1/high TfR1, high cyclin B1/high fibronectin, high TfR1/high fibronectin, and high cyclin B1/high TfR1/high fibronectin) sig- nificantly predicted mortality. The HRs for mortality of combi-
nations involving each DEP were higher than those of each pro- tein individually.
Reclassification analyses of individual DEPs and combinations thereof for overall survival
C-statistics and the cfNRI were used to evaluate the prognostic values of the DEPs additive to age and staging (Table 3). The C- index of the reference model (age and staging) was 0.78, and it increased when each protein (high cyclin B1, high TfR1, and high fibronectin) was added, to 0.82, 0.83, and 0.82, respective- ly. The C-index increased further when combinations of the pro- teins were considered (high cyclin B1/high TfR1, high cyclin B1/high fibronectin, high TfR1/high fibronectin, and high cyclin B1/high TfR1/high fibronectin), which yielded C-index values of 0.84, 0.86, 0.86, and 0.86, respectively. In addition, the cfNRI values were also significant for each protein, individually and in combination with others (P value <0.05 for all). The event and nonevent cfNRIs were both positive in all models.
Subgroup analysis of non-metastatic ACC patients
We performed a subgroup analysis of non-metastatic patients (Table 4). Some combinations of DEPs (high cyclin B/high TfR1 and high TfR1/high fibronectin) showed significant asso- ciations with mortality, while other combinations (high cyclin B1/high fibronectin and high cyclin B1/high TfR1/high fibro- nectin) showed near-significant associations with mortality.
DISCUSSION
In the present study, we identified three proteins (cyclin B1, TfR1, and fibronectin) that were highly expressed in metastatic ACC in the TCPA database. With a median follow-up of 3.1 years, high expression of each of these three proteins was asso- ciated with a poor survival rate. Subjects with high expression levels of a combination of two DEPs were at a higher risk for mortality than those with a high expression levels of only one DEP. Cyclin B1, TfR1, and all combinations of the three DEPs showed meaningful prognostic performance independent of age and staging. Moreover, among non-metastatic patients, combi- nations of these three DEPs showed significant or near-signifi- cant associations with mortality. The reason for the non-signifi- cance of fibronectin alone needs to be elucidated, but the small number of patients may have contributed to this finding. In ad- dition, the C-index and cfNRI values of high cyclin B1/high TfR1/high fibronectin were the same as those of high cyclin B1/ high fibronectin. This phenomenon may be attributed to the
1.00
Log-rank P<0.001
Overall survival
0.75
0.50
0.25
Cyclin B1 low
Cyclin B1 high
0
-
0
2
4
6
8
10
12
14
1.00
Log-rank P<0.001
Overall survival
0.75
0.50
0.25
Fibronectin low
Fibronectin high
0
-
0
2
4
6
8
10
12
14
Number at risk
Years
Fibronectin low
23
16
13
6
2
1
0
0
Fibronectin high 23
13
6
2
1
1
1
0
C
1.00
Log-rank P<0.001
Overall survival
0.75
0.50
0.25
Other
High cyclin B1/High TfR1
0
-
0
2
4
6
8
10
12
14
Number at risk
Years
Cyclin B1/TfR1 high
17
7
3
0
0
0
0
0
Other
29
22
16
8
3
2
1
0
E
1.00
Overall survival
0.75
0.50
Log-rank P<0.001
Other
0.25
High cyclin B1/High TfR1/
High fibronectin
0
H
0
2
4
6
8
10
12
14
Number at risk
Years
Cyclin B1/TfR1/Fibronectin high 13
6
2
0
0
0
0
0
Other
33
23
17
8
3
2
1
0
G
1.00
Log-rank P<0.001
Overall survival
0.75
0.50
0.25
TfR1 low
TfR1 high
0
1
0
2
4
6
8
10
12
14
Years
Number at risk
TfR1 low
23
17
11
6
1
1
1
0
TfR1 high
23
12
8
2
2
1
0
0
B
1.00
Log-rank P<0.001
Overall survival
0.75
0.50
0.25
Other
High cyclin B1/High fibronectin
0
H
0
2
4
6
8
10
12
14
Years
Number at risk
Cyclin B1/Fibronectin high
14
7
3
0
0
0
0
0
D
Other 32
22
16
8
3
2
1
0
1.00
Log-rank P<0.001
Overall survival
0.75
0.50
0.25
Other
High TfR1/High fibronectin
0
H
0
2
4
6
8
10
12
14
Number at risk
Years
TfR1/Fibronectin high 14
7
3
0
0
0
0
0
Other 32
22
16
8
3
2
1
0
F
| Years | ||||||||
|---|---|---|---|---|---|---|---|---|
| Number at risk | ||||||||
| Cyclin B1 low 23 | 18 | 13 | 7 | 3 | 2 | 1 | 0 | |
| Cyclin B1 high 23 | 11 | 6 | 1 | 0 | 0 | 0 | 0 | A |
EnM
| Variable | Univariate | Multivariateª | ||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Age (per 10-year increment) | 1.35 (0.91-2.02) | 0.140 | - | - |
| Staging (ENSAT)b | 11.82 (2.49-56.09) | 0.002 | - | - |
| High cyclin B1 | 11.40 (2.47-52.59) | 0.002 | 6.13 (1.02-36.65) | 0.047 |
| High TfR1 | 8.76 (1.94-39.54) | 0.005 | 6.59 (1.14-38.23) | 0.036 |
| High fibronectin | 8.71 (1.93-39.35) | 0.005 | 3.92 (0.75-20.43) | 0.105 |
| High cyclin B1+high TfR1℃ | 22.23 (4.59-107.61) | <0.001 | 13.72 (1.89-99.66) | 0.010 |
| High cyclin B1+high fibronectine | 19.44 (4.22-89.57) | <0.001 | 9.22 (1.34-63.55) | 0.024 |
| High TfR1+high fibronectin“ | 21.48 (4.57-100.95) | <0.001 | 18.59 (2.54-135.88) | 0.004 |
| High cyclin B1+high TfR1+high fibronectine | 15.43 (4.05-58.83) | <0.001 | 8.41 (1.54-45.90) | 0.014 |
HR, hazard ratio; CI, confidence interval; ENSAT, European Network for the Study of Adrenal Tumors; TfR1, transferrin receptor.
ªMultivariate analyses are adjusted by age and staging; bThe categorization of I+II vs. III+IV was applied using the ENSAT system; “The reference group is other subjects, such as those except patients with both high cyclin B1 and high TfR1.
| C-index (95% CI) | cfNRI (95% CI) | P value (cfNRI) | Event cfNRI | Nonevent cfNRI | |
|---|---|---|---|---|---|
| Ref (age+stage)ª | 0.78 (0.62-0.94) | - | - | - | - |
| Ref+high cyclin B1 | 0.82 (0.72-0.92) | 1.03 (0.53-1.52) | <0.001 | 0.71 | 0.31 |
| Ref+high TfR1 | 0.83 (0.71-0.94) | 1.03 (0.53-1.52) | <0.001 | 0.71 | 0.31 |
| Ref+high fibronectin | 0.82 (0.70-0.94) | 0.74 (0.16-1.32) | 0.012 | 0.43 | 0.31 |
| Ref+high cyclin B1+high TfR1 | 0.84 (0.76-0.93) | 0.84 (0.34-1.34) | 0.001 | 0.71 | 0.13 |
| Ref+high cyclin B1+high fibronectin | 0.86 (0.78-0.93) | 0.88 (0.34-1.43) | 0.001 | 0.57 | 0.31 |
| Ref+high TfR1+high fibronectin | 0.86 (0.78-0.95) | 1.03 (0.53-1.52) | <0.001 | 0.71 | 0.31 |
| Ref+high cyclin B1+high TfR1+high fibronectin | 0.86 (0.78-0.93) | 0.88 (0.34-1.43) | 0.001 | 0.57 | 0.31 |
CI, confidence interval; cfNRI, category-free net reclassification improvement; TfR1, transferrin receptor.
aAge was analyzed in terms of 10-year increments; for stage, the categorization of I+II vs. III+IV was applied using the European Network for the Study of Adrenal Tumors (ENSAT) system.
| Variable | Univariate | Multivariateª | ||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Age (per 10-year increment) | 1.52 (0.89-2.62) | 0.126 | - | - |
| Staging (ENSAT)b | 7.85 (1.33-46.04) | 0.022 | - | - |
| High cyclin B1 | 7.75 (1.51-39.86) | 0.014 | 5.46 (0.80-37.27) | 0.083 |
| High TfR1 | 11.27 (1.38-92.35) | 0.024 | 6.70 (0.70-63.77) | 0.098 |
| High fibronectin | 5.91 (1.19-29.43) | 0.030 | 3.60 (0.63-20.66) | 0.150 |
| High cyclin B1+high TfR1℃ | 30.19 (3.45-264.24) | 0.002 | 21.69 (1.58-298.63) | 0.021 |
| High cyclin B1+high fibronectine | 16.20 (3.07-85.37) | 0.001 | 9.25 (0.96-88.87) | 0.054 |
| High TfR1+high fibronectin“ | 29.22 (3.48-245.26) | 0.002 | 84.15 (2.39-2961.35) | 0.015 |
| High cyclin B1+high TfR1+high fibronectine | 16.20 (3.07-85.37) | 0.001 | 9.25 (0.96-88.87) | 0.054 |
HR, hazard ratio; CI, confidence interval; ENSAT, European Network for the Study of Adrenal Tumors; TfR1, transferrin receptor.
aMultivariate analyses were adjusted by age and staging; bThe categorization of I+II vs. III+IV was applied using ENSAT system; “The reference group is other subjects, such as those except patients with both high cyclin B1 and high TfR1.
similar number of subjects with high expression of all three DEPs and with high cyclin B1 and high fibronectin expression.
We conducted a proteomic analysis to identify prognostic markers for ACC. Several studies have explored prognostic markers for ACC, but most studies have reported clinical or pathological prognostic markers [5,7]. Molecular studies re- vealed that mutations in several potential driver genes such as CTNNB1 and TP53 may serve as prognostic markers [10,15]. Nevertheless, the independent prognostic value of these mark- ers is controversial [9,10]. Recently, pan-genomic approaches- encompassing DNA copy number, mRNA expression, miRNA expression, and DNA methylation-classified ACC patients into distinct molecular subgroups, which showed different sur- vival outcomes [11-13]. Despite the good performance of pan- genomic prognostic markers in previous studies, the relevant analyses are quite complex to perform, and their prognostic val- ues were insufficient in advanced ACC (stage IV) [13]. To over- come these limitations, our present study utilized a proteomic approach, which has not been investigated in previous research on ACC. In this study, the expression status of three DEPs had prognostic value in addition to age and staging, as assessed by C-index and cfNRI. In addition, the highest C-index of our pre- diction model was quite high (0.86), which is similar to the re- sult of a previous study [13], given that we included stage IV patients. Therefore, the DEPs that we identified can be applied as relatively simple and powerful prognostic markers in ACC.
The three identified DEPs-cyclin B1, TfR1, and fibronec- tin-have been studied as prognostic markers for other cancers. Cyclin B1 is a regulatory protein that plays an important role in mitosis by forming a complex with Cdk1 [16]. It acts as a switch-like manner in the decision to progress from G2 to the M phase in the cell. In many cancers, such as breast, cervical, colorectal, esophageal, lung, and prostate cancers, high expres- sion levels of cyclin B1 have been found [17-21] and are associ- ated with a poor prognosis [17,21]. In addition, the use of small interfering RNA that targeted cyclin B1 downregulated tumor proliferation and enhanced sensitivity to systemic chemotherapy [22]. Moreover, in a previous study using ACC cell lines (H295R and SW-13 cells), the mechanism of mitotane was found to involve the cyclin B1/Cdk complex [23].
TfR1 is a transmembrane glycoprotein that imports iron- bound transferrin into cells by endocytosis [24]. It has also been reported that TfR1 is overexpressed in many cancers, such as breast, prostate, colon, liver, lung, brain, ovarian, and hemato- logic cancers [25-31]. This high expression of TfR1 in cancer is explained by the high iron requirements of cancer cells, as well
as by the regulatory role of TfR1 in anti-apoptotic regulation, reactive oxygen species production, and mitochondrial respira- tion [30,32,33]. In addition, as potential TfR1 inhibitors, cur- cumin, antibody A24 or JST-TFR09, and miR-320 have been studied in several cancers [34-37].
Fibronectin is a glycoprotein of the extracellular matrix that plays a role in cell growth, differentiation, migration, and wound healing [38]. It has also been reported that abnormal ex- pression of fibronectin promotes invasion and migration of can- cers, including breast, ovarian, prostate, lung, and colon cancers [39-43]. In addition, fibronectin has been reported to play a role in reducing the response of cancer to cytotoxic therapy [44,45]. In vitro studies of lung, pancreatic neuroendocrine, and breast cancers have shown that several molecules (e.g., PP2, PF- 04554878, pUR4B, and AdF512v1) that inhibit fibronectin and its upstream or downstream pathways reduced its effects on tu- mor cell proliferation, migration, and angiogenesis [42,46-48].
As discussed above, previous studies have suggested that all three of the proteins (cyclin B1, TfR1, and fibronectin) identi- fied as DEPs in our study may be prognostic markers in several cancers. However, to the best of our knowledge, no study has yet demonstrated an association of these proteins with mortality in ACC patients. Accordingly, these proteins may also be useful novel markers of poor prognosis in ACC patients.
There are several limitations of our study. The number of sub- jects was small, and our results were not validated in other co- horts. Thus, the prognostic power of these three DEPs remains to be confirmed. In addition, the RPPA data of ACC patients from the TCPA, which were used as the data source in our study, contained information on a small number of proteins. Therefore, it is possible that other protein markers might predict survival in ACC patients. Advanced proteomics techniques, such as gas or liquid chromatography tandem mass spectrome- try, might enable the identification of more prognostic protein markers in ACC patients.
Taken together, the overexpression of cyclin B1, TfR1, and fibronectin proteins may have prognostic value for ACC pa- tients. Further studies need to validate these three proteins as prognostic markers in another ACC cohort.
CONFLICTS OF INTEREST
No potential conflict of interest relevant to this article was re- ported.
EnM
ACKNOWLEDGMENTS
This study was supported by the Korean Endocrine Hyperten- sion-Adrenal Study Group from the Korean Endocrinology So- ciety. This work was presented in abstract form at the 2019 En- docrine Society in New Orleans, USA.
AUTHOR CONTRIBUTIONS
Conception or design: S.J.M., J.H.K., S.H.K., C.S.S. Acquisi- tion, analysis, or interpretation of data: S.J.M., J.H.K. Drafting the work or revising: S.J.M., J.H.K. Final approval of the manu- script: S.J.M., J.H.K., S.H.K., C.S.S.
ORCID
Sun Joon Moon https://orcid.org/0000-0002-6286-7254 Jung Hee Kim https://orcid.org/0000-0003-1932-0234
REFERENCES
1. Allolio B, Fassnacht M. Clinical review: adrenocortical car- cinoma: clinical update. J Clin Endocrinol Metab 2006;91: 2027-37.
2. Jouinot A, Bertherat J. Management of endocrine disease: adrenocortical carcinoma: differentiating the good from the poor prognosis tumors. Eur J Endocrinol 2018;178:R215- 30.
3. Else T, Kim AC, Sabolch A, Raymond VM, Kandathil A, Caoili EM, et al. Adrenocortical carcinoma. Endocr Rev 2014; 35:282-326.
4. DeLellis RA, Lloyd RV, Heitz PU, Eng C. Pathology and ge- netics of tumours of endocrine organs. 3rd ed. Lyon: IARC Press; 2004.
5. Fassnacht M, Johanssen S, Quinkler M, Bucsky P, Willen- berg HS, Beuschlein F, et al. Limited prognostic value of the 2004 International Union Against Cancer staging classifica- tion for adrenocortical carcinoma: proposal for a revised TNM classification. Cancer 2009;115:243-50.
6. Amin MB, Edge S, Greene F, Byrd DR, Brookland RK, Washington MK, et al. AJCC cancer staging manual. 8th ed. Chicago: Springer; 2017. Chapter 76, Adrenal cortical carci- noma.
7. Beuschlein F, Weigel J, Saeger W, Kroiss M, Wild V, Daf- fara F, et al. Major prognostic role of Ki67 in localized adre- nocortical carcinoma after complete resection. J Clin Endo-
crinol Metab 2015;100:841-9.
8. Papathomas TG, Pucci E, Giordano TJ, Lu H, Duregon E, Volante M, et al. An international Ki67 reproducibility study in adrenal cortical carcinoma. Am J Surg Pathol 2016;40: 569-76.
9. Heaton JH, Wood MA, Kim AC, Lima LO, Barlaskar FM, Almeida MQ, et al. Progression to adrenocortical tumori- genesis in mice and humans through insulin-like growth factor 2 and B-catenin. Am J Pathol 2012;181:1017-33.
10. Waldmann J, Patsalis N, Fendrich V, Langer P, Saeger W, Chaloupka B, et al. Clinical impact of TP53 alterations in ad- renocortical carcinomas. Langenbecks Arch Surg 2012;397: 209-16.
11. Assie G, Letouze E, Fassnacht M, Jouinot A, Luscap W, Barreau O, et al. Integrated genomic characterization of ad- renocortical carcinoma. Nat Genet 2014;46:607-12.
12. Zheng S, Cherniack AD, Dewal N, Moffitt RA, Danilova L, Murray BA, et al. Comprehensive pan-genomic character- ization of adrenocortical carcinoma. Cancer Cell 2016;29: 723-36.
13. Assie G, Jouinot A, Fassnacht M, Libe R, Garinet S, Jacob L, et al. Value of molecular classification for prognostic assess- ment of adrenocortical carcinoma. JAMA Oncol 2019;5: 1440-7.
14. Li J, Lu Y, Akbani R, Ju Z, Roebuck PL, Liu W, et al. TCPA: a resource for cancer functional proteomics data. Nat Meth- ods 2013;10:1046-7.
15. Kovach AE, Nucera C, Lam QT, Nguyen A, Dias-Santagata D, Sadow PM. Genomic and immunohistochemical analysis in human adrenal cortical neoplasia reveal beta-catenin muta- tions as potential prognostic biomarker. Discoveries (Craio- va) 2015;3:e40.
16. Nurse P. Universal control mechanism regulating onset of M-phase. Nature 1990;344:503-8.
17. Suzuki T, Urano T, Miki Y, Moriya T, Akahira J, Ishida T, et al. Nuclear cyclin B1 in human breast carcinoma as a potent prognostic factor. Cancer Sci 2007;98:644-51.
18. Wang A, Yoshimi N, Ino N, Tanaka T, Mori H. Overexpres- sion of cyclin B1 in human colorectal cancers. J Cancer Res Clin Oncol 1997;123:124-7.
19. Nozoe T, Korenaga D, Kabashima A, Ohga T, Saeki H, Sugi- machi K. Significance of cyclin B1 expression as an inde- pendent prognostic indicator of patients with squamous cell carcinoma of the esophagus. Clin Cancer Res 2002;8:817-22.
20. Zhao M, Kim YT, Yoon BS, Kim SW, Kang MH, Kim SH, et al. Expression profiling of cyclin B1 and D1 in cervical
EnM
Moon SJ, et al.
carcinoma. Exp Oncol 2006;28:44-8.
21. Egloff AM, Weissfeld J, Land SR, Finn OJ. Evaluation of anticyclin B1 serum antibody as a diagnostic and prognostic biomarker for lung cancer. Ann N Y Acad Sci 2005;1062:29- 40.
22. Androic I, Kramer A, Yan R, Rodel F, Gatje R, Kaufmann M, et al. Targeting cyclin B1 inhibits proliferation and sensitizes breast cancer cells to taxol. BMC Cancer 2008;8:391.
23. Cerquetti L, Sampaoli C, Amendola D, Bucci B, Misiti S, Raza G, et al. Mitotane sensitizes adrenocortical cancer cells to ionizing radiations by involvement of the cyclin B1/CDK complex in G2 arrest and mismatch repair enzymes modula- tion. Int J Oncol 2010;37:493-501.
24. Aisen P. Transferrin receptor 1. Int J Biochem Cell Biol 2004;36:2137-43.
25. Rychtarcikova Z, Lettlova S, Tomkova V, Korenkova V, Langerova L, Simonova E, et al. Tumor-initiating cells of breast and prostate origin show alterations in the expression of genes related to iron metabolism. Oncotarget 2017;8:6376- 98.
26. Horniblow RD, Bedford M, Hollingworth R, Evans S, Sut- ton E, Lal N, et al. BRAF mutations are associated with in- creased iron regulatory protein-2 expression in colorectal tu- morigenesis. Cancer Sci 2017;108:1135-43.
27. Kindrat I, Tryndyak V, de Conti A, Shpyleva S, Mudalige TK, Kobets T, et al. MicroRNA-152-mediated dysregulation of hepatic transferrin receptor 1 in liver carcinogenesis. On- cotarget 2016;7:1276-87.
28. Wang B, Zhang J, Song F, Tian M, Shi B, Jiang H, et al. EGFR regulates iron homeostasis to promote cancer growth through redistribution of transferrin receptor 1. Cancer Lett 2016;381:331-40.
29. Rosager AM, Sorensen MD, Dahlrot RH, Hansen S, Schon- berg DL, Rich JN, et al. Transferrin receptor-1 and ferritin heavy and light chains in astrocytic brain tumors: expression and prognostic value. PLoS One 2017;12:e0182954.
30. Basuli D, Tesfay L, Deng Z, Paul B, Yamamoto Y, Ning G, et al. Iron addiction: a novel therapeutic target in ovarian cancer. Oncogene 2017;36:4089-99.
31. Gu Z, Wang H, Xia J, Yang Y, Jin Z, Xu H, et al. Decreased ferroportin promotes myeloma cell growth and osteoclast differentiation. Cancer Res 2015;75:2211-21.
32. Jian J, Yang Q, Huang X. Src regulates Tyr(20) phosphory- lation of transferrin receptor-1 and potentiates breast cancer cell survival. J Biol Chem 2011;286:35708-15.
33. Jeong SM, Hwang S, Seong RH. Transferrin receptor regu-
lates pancreatic cancer growth by modulating mitochondrial respiration and ROS generation. Biochem Biophys Res Commun 2016;471:373-9.
34. Yang C, Ma X, Wang Z, Zeng X, Hu Z, Ye Z, et al. Curcum- in induces apoptosis and protective autophagy in castration- resistant prostate cancer cells through iron chelation. Drug Des Devel Ther 2017;11:431-9.
35. Callens C, Moura IC, Lepelletier Y, Coulon S, Renand A, Dussiot M, et al. Recent advances in adult T-cell leukemia therapy: focus on a new anti-transferrin receptor monoclonal antibody. Leukemia 2008;22:42-8.
36. Shimosaki S, Nakahata S, Ichikawa T, Kitanaka A, Kameda T, Hidaka T, et al. Development of a complete human IgG monoclonal antibody to transferrin receptor 1 targeted for adult T-cell leukemia/lymphoma. Biochem Biophys Res Commun 2017;485:144-51.
37. Schaar DG, Medina DJ, Moore DF, Strair RK, Ting Y. miR- 320 targets transferrin receptor 1 (CD71) and inhibits cell proliferation. Exp Hematol 2009;37:245-55.
38. Pankov R, Yamada KM. Fibronectin at a glance. J Cell Sci 2002;115(Pt 20):3861-3.
39. Nam JM, Onodera Y, Bissell MJ, Park CC. Breast cancer cells in three-dimensional culture display an enhanced ra- dioresponse after coordinate targeting of integrin alpha5be- ta1 and fibronectin. Cancer Res 2010;70:5238-48.
40. Mitra AK, Sawada K, Tiwari P, Mui K, Gwin K, Lengyel E. Ligand-independent activation of c-Met by fibronectin and a(5)ß(1)-integrin regulates ovarian cancer invasion and me- tastasis. Oncogene 2011;30:1566-76.
41. Han Z, Zhou Z, Shi X, Wang J, Wu X, Sun D, et al. EDB fi- bronectin specific peptide for prostate cancer targeting. Bio- conjug Chem 2015;26:830-8.
42. Meng XN, Jin Y, Yu Y, Bai J, Liu GY, Zhu J, et al. Charac- terisation of fibronectin-mediated FAK signalling pathways in lung cancer cell migration and invasion. Br J Cancer 2009; 101:327-34.
43. Wei PL, Kuo LJ, Huang MT, Ting WC, Ho YS, Wang W, et al. Nicotine enhances colon cancer cell migration by induc- tion of fibronectin. Ann Surg Oncol 2011;18:1782-90.
44. Eke I, Storch K, Krause M, Cordes N. Cetuximab attenuates its cytotoxic and radiosensitizing potential by inducing fi- bronectin biosynthesis. Cancer Res 2013;73:5869-79.
45. Pontiggia O, Sampayo R, Raffo D, Motter A, Xu R, Bissell MJ, et al. The tumor microenvironment modulates tamoxi- fen resistance in breast cancer: a role for soluble stromal factors and fibronectin through ß1 integrin. Breast Cancer
EnM
Res Treat 2012;133:459-71.
46. Francois RA, Maeng K, Nawab A, Kaye FJ, Hochwald SN, Zajac-Kaye M. Targeting focal adhesion kinase and resis- tance to mTOR inhibition in pancreatic neuroendocrine tu- mors. J Natl Cancer Inst 2015;107:djv123.
47. Hielscher A, Ellis K, Qiu C, Porterfield J, Gerecht S. Fibro- nectin deposition participates in extracellular matrix assembly
and vascular morphogenesis. PLoS One 2016;11:e0147600.
48. Lopez MV, Rivera AA, Viale DL, Benedetti L, Cuneo N, Kimball KJ, et al. A tumor-stroma targeted oncolytic adeno- virus replicated in human ovary cancer samples and inhibited growth of disseminated solid tumors in mice. Mol Ther 2012;20:2222-33.