Cancers (Basel). 2021;13(15):cancers13153890.

Mass Spectrometry-Based Proteomic Discovery of Prognostic Biomarkers in Adrenal Cortical Carcinoma.

, , , , , ,

제1저자 장한나, 문선준 · 교신저자 김정희, 한도현

PMID 34359790 PMC8345732 DOI 10.3390/cancers13153890 IF 4.8 내분비대사내과

소속 (8)

Department of Internal Medicine, Seoul National University College of Medicine, Seoul 03080, Korea.
Department of Internal Medicine, Seoul National University Hospital, Seoul 03080, Korea.
Division of Endocrinology and Metabolism, Department of Internal Medicine, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul 03080, Korea.
Department of Pathology, Seoul National University College of Medicine, Seoul 03080, Korea.
Department of Translational Medicine, Seoul National University College of Medicine, Seoul 03080, Korea.
Integrated Major in Innovative Medical Science, Seoul National University Graduate School, Seoul 03080, Korea.
Department of Internal Medicine, Seoul Metropolitan Government Seoul National University Boramae Medical Center, Seoul 03080, Korea.
Proteomics Core Facility, Biomedical Research Institute Seoul National University Hospital, Seoul 03080, Korea.

Abstract

Adrenal cortical carcinoma (ACC) is an extremely rare disease with a variable prognosis. Current prognostic markers have limitations in identifying patients with a poor prognosis. Herein, we aimed to investigate the prognostic protein biomarkers of ACC using mass-spectrometry-based proteomics. We performed the liquid chromatography-tandem mass spectrometry (LC-MS/MS) using formalin-fixed paraffin-embedded (FFPE) tissues of 45 adrenal tumors. Then, we selected 117 differentially expressed proteins (DEPs) among tumors with different stages using the machine learning algorithm. Next, we conducted a survival analysis to assess whether the levels of DEPs were related to survival. Among 117 DEPs, HNRNPA1, C8A, CHMP6, LTBP4, SPR, NCEH1, MRPS23, POLDIP2, and WBSCR16 were significantly correlated with the survival of ACC. In age- and stage-adjusted Cox proportional hazard regression models, only HNRNPA1, LTBP4, MRPS23, POLDIP2, and WBSCR16 expression remained significant. These five proteins were also validated in TCGA data as the prognostic biomarkers. In this study, we found that HNRNPA1, LTBP4, MRPS23, POLDIP2, and WBSCR16 were protein biomarkers for predicting the prognosis of ACC.

Keywords

HNRNPA1adrenal cortical carcinomamass spectrometryprognosisproteomics

Figures

**Figure 1** Overall workflow and results of proteomic analysis. (A) Overall workflow for proteomic analysis of adrenal cortical carcinoma FFPE tissues with respect to ENSAT staging is presented. This figure was created with Biorender.com (accessed on 10 June 2021) and exported under a paid subscription; (B) The number of total proteins expressed in benign adrenal adenomas and ACC, stages 1–2 and 3–4 ACCs, and stages 1–3 and 4 ACCs are expressed, respectively. The overlapping part of the circle indicates the number of proteins expressed in both groups; (C) The intensity of the expression of well-known proteins in the ACC. ACC—adrenal cortical carcinoma; FASP—filter aided sample preparation; FFPE—formalin-fixed paraffin-embedded; ENSAT—European Network for the Study of Adrenal Tumors; LS-MS/MS—liquid chromatography–tandem mass spectrometry; PPT—precipitation.

**Figure 2** Volcano plots for differential expression of proteins. Volcano plots showed significantly differentially abundant proteins. The −log₁₀ (p-value) is plotted against log₂ (fold-change). The non-axial vertical lines denote a two-fold difference in expression, while the non-axial horizontal line denotes p = 0.05, which is our significance threshold. The differential expression of proteins in (A) benign adrenal adenomas and ACC, (B) stages 1–2 and 3–4 ACCs, and (C) stages 1–3 and 4 ACCs is expressed, respectively. ACC—adrenal cortical carcinoma.

**Figure 3** Feature selection of proteins for classification (top 20 DEPs). Feature selection was performed using ReliefF, infoGain, and ANOVA for DEP. Using each algorithm, the top 20 DEPs were obtained, and proteins with high rankings in all algorithms are displayed in overlapping circles. Differential expressions of proteins in (A) benign adrenal adenomas and ACC, (B) stages 1–2 and 3–4 ACCs, and (C) stages 1–3 and 4 ACCs are expressed, respectively. Red letters indicate highly expressed proteins in the worse group, while green letters indicate highly expressed proteins in the better group. ACC—adrenal cortical carcinoma; ANOVA—analysis of variance; DEP—differential expression of proteins.

**Figure 4** IPA analyses of the differential expression of proteins (Canonical pathway). IPA analysis was performed in the differentially expressed proteins identified in each group, and canonical pathways with low p-values were presented. (A) Pathways identified in benign adrenal adenomas are indicated in green, while pathways identified in ACC are indicated in red; (B) Pathways identified in stages 1–2 ACC are indicated in green, while pathways identified in stages 3–4 are indicated in red; (C) Pathways identified in stages 1–3 are indicated in green, while pathways identified in stage 4 are indicated in red. The Z-score indicates that the level of protein expression was significantly high, and it was significant when the absolute value was more than 1. ACC—adrenal cortical carcinoma; GP6—Glycoprotein VI; IPA—ingenuity pathway analysis; LXR/RXR—Liver X Receptor-Retinoid X Receptor; TCA—tricarboxylic acid cycle.

**Figure 5** Kaplan–Meier analyses of significant DEPs in SNUH cohort. When Kaplan–Meier analyses were performed on the featured proteins, nine proteins with significant differences were identified. It was analyzed by dividing it into low and high groups based on the median value. When comparing benign adrenal adenomas with ACC, the high expression of HNRNPA1 was associated with poor overall survival. When comparing stage 1–2 and stage 3–4 ACCs, high expressions of C8A, CHMP6, and LTBP4 and low expression of SPR were associated with poor overall survival. In addition, when comparing stages 1–3 and stage 4 ACCs, high expression of NCEH1 and low expressions of MRPS23, POLDIP2, and WBSCR16 were associated with poor overall survival. ACC—adrenal cortical carcinoma; C8A—complement C8 alpha chain; CHMP6—charged multivesicular body protein 6; CI—confidence interval; DEP—differential expression of proteins; HNRNPA1—heterogeneous nuclear ribonucleoprotein A1; LTBP4—latent transforming growth factor beta binding protein 4; MRPS23—mitochondrial ribosomal protein S23; NCEH1—neutral cholesterol ester hydrolase 1; POLDIP2—polymerase delta-interacting protein 2; SNUH—Seoul National University Hospital; SPR—sepiapterin reductase; WBSCR16—Williams-Beuren syndrome chromosome region 16.

**Figure 6** Cox proportional hazard regression models of significant DEPs in SNUH cohort. When Cox proportional hazard regression model analysis was performed on the feature-selected protein, nine proteins with significant differences were identified. ACC—adrenal cortical carcinoma; C8A—complement C8 alpha chain; CHMP6—charged multivesicular body protein 6; CI—confidence interval; DEP—differential expression of proteins; HNRNPA1—heterogeneous nuclear ribonucleoprotein A1; LTBP4—latent transforming growth factor beta binding protein 4; MRPS23—mitochondrial ribosomal protein S23; NCEH1—neutral cholesterol ester hydrolase 1; POLDIP2—polymerase delta-interacting protein 2; SNUH—Seoul National University Hospital; SPR—sepiapterin reductase; WBSCR16—Williams-Beuren syndrome chromosome region 16.

**Figure 7** Kaplan–Meier analyses of significant DEPs in TCGA study. We tried to validate DEPs, which showed a significant difference in survival in the SNUH cohort, by performing Kaplan–Meier analyses in the TCGA study. It was analyzed by dividing it into low and high groups based on the median value. C8A—complement C8 alpha chain; CHMP6—charged multivesicular body protein 6; CI—confidence interval; DEP—differential expression of proteins; HNRNPA1—heterogeneous nuclear ribonucleo-protein A1; LTBP4—latent transforming growth factor beta binding protein 4; MRPS23—mitochondrial ribosomal protein S23; NCEH1—neutral cholesterol ester hydrolase 1; POLDIP2—polymerase delta-interacting protein 2; SPR—sepiapterin reductase; TCGA—The Cancer Genome Atlas; WBSCR16—Williams-Beuren syndrome chromosome region 16.

**Figure 8** Cox proportional hazard regression models of significant DEPs in TCGA study. Cox proportional hazard regression model analysis was performed on DEPs, which showed a significant difference in survival in the TCGA study. C8A—complement C8 alpha chain; CHMP6—charged multivesicular body protein 6; CI—confidence interval; DEP—differential expression of proteins; HNRNPA1—heterogeneous nuclear ribonucleo-protein A1; LTBP4—latent transforming growth factor beta binding protein 4; MRPS23—mitochondrial ribosomal protein S23; NCEH1—neutral cholesterol ester hydrolase 1; POLDIP2—polymerase delta-interacting protein 2; SNUH—Seoul National University Hospital; SPR—sepiapterin reductase; TCGA—The Cancer Genome Atlas; WBSCR16—Williams-Beuren syndrome chromosome region 16.

그림 출처: PubMed Central 공개 논문 (PMC8345732)

Conflict of interest statement

No potential conflict of interest relevant to this article was reported. The funding source had no role in the design, data collection, analysis, interpretation of data, writing of the manuscript, or in the decision to submit the manuscript for publication.

그 밖의 정보