HDAC4, a prognostic and chromosomal instability marker, refines the predictive value of MGMT promoter methylation

Wen Cheng 0 1 2 3 4 5 Mingyang Li 0 1 2 3 4 5 Jinquan Cai 0 1 2 3 4 5 Kuanyu Wang 0 1 2 3 4 5 Chuanbao Zhang 0 1 2 3 4 5 Zhaoshi Bao 0 1 2 3 4 5 Yanwei Liu 0 1 2 3 4 5 Anhua Wu 0 1 2 3 4 5 0 M. Li C. Zhang Y. Liu Beijing Neurosurgical Institute, Capital Medical University , Beijing , China 1 W. Cheng M. Li J. Cai K. Wang C. Zhang Z. Bao Y. Liu A. Wu Chinese Glioma Cooperative Group (CGCG) , Beijing , China 2 W. Cheng A. Wu (&) Department of Neurosurgery, The First Hospital of China Medical University , Nanjing Street 155, Heping District, Shenyang 110001 , China 3 K. Wang Department of Neurosurgery, The First Affiliated Hospital of Dalian Medical University , Dalian , China 4 J. Cai Department of Neurosurgery, The Second Affiliated Hospital of Harbin Medical University , Harbin , China 5 M. Li Z. Bao Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University , Beijing , China Chromosomal instability is a hallmark of human cancers and is closely linked to tumorigenesis. The prognostic value of molecular signatures of chromosomal instability (CIN) has been validated in various cancers. However, few studies have examined the relationship between CIN and glioma. Histone deacetylases (HDACs) regulate chromosome structure and are linked to the loss of genomic integrity in cancer cells. In this study, the prognostic value of HDAC4 expression and its association with markers of CIN were investigated by analyzing data from our own and four other large sample databases. The results showed that HDAC4 expression is downregulated in high- as compared to lowgrade glioma and is associated with a favorable clinical outcome. HDAC4 expression and CIN were closely related in glioma from both functional and statistical standpoints. Moreover, the predictive value of the O-6-methylguanineDNA methyltransferase (MGMT) promoter methylation status-a widely used glioma marker-was refined by HDAC4 Wen Cheng and Mingyang Li have contributed equally to this work. - Glioma is the most common type of primary central nervous system (CNS) tumor and a leading cause of tumorrelated mortality. Despite major advances in therapy over the past decades, the clinical outcome for most patients remains poor. This is especially true for glioblastoma (GBM), the most malignant grade of glioma, which has a median survival of 14.6 months and a 2-year survival rate of 510 % even after aggressive therapy [1]. As a major form of genomic instability, chromosomal instability (CIN) is a critical event in early stages of tumorigenesis and, when compounded, leads to the transformation of normal cells into cancer cells [2]. Various types of CIN have been detected in glioma, including mutations, loss of heterozygosity, and copy number aberrations [36]. Several studies have reported that CIN can affect sensitivity to chemotherapy and consequently the prognosis of glioma patients [7, 8]. A close association between CIN and histone acetylation has been demonstrated [911]. Central to histone acetylation are histone deacetylases (HDACs), which maintain genomic integrity by targeting histone and non-histone proteins and thereby regulating DNA repair mechanisms [12]. A total of 18 human HDACs, classified into four groups, have been identified. As a member of group II HDACs, HDAC4 is closely linked to many disease processesincluding cancer, leukemia, diabetes, infection, and cardiac disease [1318] and is also highly expressed in the brain where it plays an important role in brain functioning [1922]. Epigenetic silencing of the O-6-methylguanine-DNA methyltransferase (MGMT) gene by promoter methylation is associated with prolonged survival and sensitivity to chemotherapeutic alkylating agents in GBM patients undergoing standard treatment [23, 24]. The beneficial effects of combined radiochemotherapy vary significantly between GBM patients, even for those with a methylated MGMT promoter [25]. This suggests that while important, MGMT promoter methylation is not the sole factor determining clinical outcome, and highlights the need for evaluating patients based on other factors; for instance, CIN combined with MGMT promoter methylation status may provide more accurate information for predicting disease outcome. CIN is defined as the gain or loss of whole or fractions of chromosomes, and is associated with tumorigenesis, disease prognosis, and acquisition of multi-drug resistance in various cancers, including breast cancer, melanoma, and lymphoma [2632]. High throughput gene expression profiling approaches have established a reasonable link between the expression of specific genes and the degree of CIN in multiple cancers. Carter et al. developed computational methods to measure the CIN score for 10,151 genes, which indicates the correlation between each gene and the CIN degree in tumor samples [30]. Based on the CIN score, the top ranked genes are chosen for forming the CIN signature, which was represented by CIN25 score (Further backgrounds of CIN signature and CIN25 score are shown in the Supplementary Text) [30]. The CIN signature, comprising a specific set of genes that are critical for maintaining genomic integrity, is significantly higher in metastatic foci, and stratifies patients according to clinical outcome in various cancers, suggesting that these genes are responsible for a more aggressive cancer phenotype [30, 33]. However, as it consists of multiple genes, the CIN signature is too complex to be suitable for routine clinical application. The present study investigated whether HDAC4 expression can serve as an alternative marker for assessing the degree of CIN and, in combination with MGMT promoter status, predict the outcome of patients. Materials and methods Patients and samples A total of 539 glioma specimens from the Chinese Glioma Genome Atlas (CGGA) that were contiguously collected at multiple centers were used in this study. Tumor tissue samples were obtained by surgical resection prior to radio and/or chemotherapy, flash-frozen in liquid nitrogen, and stored at -80 C until nucleic acid extraction. The study protocol was approved by the ethics committees of participating hospitals. Each sample was diagnosed and independently confirmed histopathologically at the Department of Pathology according to the 2007 WHO classification system of CNS tumors by two experienced neuropathologists. Clinical data, including age, sex, preoperative KPS score, adjuvant radiation and chemotherapy, and the recorded date of disease progression or death were obtained from medical records. Data on mRNA expression were obtained by whole transcriptome sequencing (N = 325) and whole-genome mRNA expression microarray (N = 299) from the CGGA, and the following four datasets were used for validation: the Cancer Genome Atlas (http://cancergenome.nih.gov); Repository for Molecular Brain Neoplasis Data (REMBRANDT, http://caintegrator.nci.nih.gov/rembrandt); GSE16011 (http:// www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE1601 (...truncated)