Accepted Manuscript
Epigenetic dysregulation in adrenocortical carcinoma, a systematic review of the literature
P.K.C. Jonker, V.M. Meyer, S. Kruijff
| PII: | S0303-7207(17)30431-8 |
| DOI: | 10.1016/j.mce.2017.08.009 |
| Reference: | MCE 10046 |
| To appear in: | Molecular and Cellular Endocrinology |
| Received Date: | 31 March 2017 |
| Revised Date: | 17 August 2017 |
| Accepted Date: | 17 August 2017 |
195H 0000-720P
Molecular and Cellular Endocrinology
Please cite this article as: Jonker, P.K.C., Meyer, V.M., Kruijff, S., Epigenetic dysregulation in adrenocortical carcinoma, a systematic review of the literature, Molecular and Cellular Endocrinology (2017), doi: 10.1016/j.mce.2017.08.009.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
TITLE Epigenetic Dysregulation in Adrenocortical Carcinoma, a Systematic Review of the Literature
P.K.C. Jonker, V.M. Meyer, S. Kruijff
HIGHLIGHTS
- Methylation levels of promotor regions in adrenocortical carcinoma (ACC) are inversely correlated with overall survival.
- ACCs can be categorized based on CpG-island methylation phenotypes (CIMPs) with a different overall survival.
- DNA methylation levels might be useful to differentiate adrenocortical carcinomas from adrenocortical adenomas and normal adrenocortical tissue.
- 14 hypermethylated genes with low mRNA expression in ACC are reported in two or more independent studies.
- Pyrosequencing and Methylation Specific Multiplex Ligation Dependent Probe Amplification (MS-MLPA) are potential platforms for future application of DNA methylation as diagnostic and prognostic tool.
ABSTRACT
Adrenocortical carcinoma (ACC) is a rare and aggressive endocrine malignancy with a poor prognosis. Diagnosis and treatment of this tumor remains challenging. The Weiss score, the current gold standard for the histopathological diagnosis of ACC, lacks diagnostic accuracy of borderline tumors (Weiss score 2 or 3) and is subject to inter observer variability. Furthermore, adjuvant and palliative systemic therapy have limited effect and no proven overall survival benefit. A better insight in the molecular background of ACC might identify markers that improve diagnostic accuracy, predict treatment response or even provide novel therapeutic targets. This systematic review of the literature aims to provide an overview of alterations in DNA methylation, histone modifications and their potential clinical relevance in ACC.
BACKGROUND
Adrenocortical carcinoma (ACC) is an endocrine malignancy with a prevalence of 0.7-2 per million inhabitants.(1) The overall prognosis of ACC is poor with reported 5-year overall survival rates ranging between 16% and 44%.(2-8) At initial presentation, 48% of the tumors is localized within the adrenal gland (stadium I-II), 16-27% has locoregional invasion or lymph node metastasis (stadium III) and 11-36% of the patients present with distant metastasis (stage IV a-c).(9-11) Five-year survival of patients with ACC is stage dependent. From stage I to IV it declines from 82%, towards 61%, 50% and less than 13 % respectively.(1) Surgical resection of locoregional disease remains the sole curative treatment option for ACC.(1) Despite novel insights in the carcinogenesis of this rare and aggressive tumor, accurate diagnosis of borderline adrenocortical tumors remains difficult. More importantly, there is no proven effective adjuvant therapy for ACC that improves overall survival.
Accurate diagnosis of a subcategory of “borderline” adrenocortical tumors with a Weiss score of 2 or 3 may be challenging.(12-14) The Weiss score is accepted as the gold standard for histopathological diagnosis of adrenocortical tumors due to its accuracy, simplicity and reliability.(15,16) However, the score has a low accuracy for accurate diagnosis of borderline adrenocortical tumors.(13,14) Some patients initially diagnosed with benign borderline tumors develop local recurrence during follow-up, sometimes with distant metastasis.(12,17,18) Current understanding about carcinogenesis of borderline tumors is hampered because of a low number of reported cases, limited follow-up and heterogeneity of endpoints.(13) The identification of molecular diagnostic markers that differentiate
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49
ACCs from adrenocortical adenoma (ACA) might be beneficial for the diagnostic accuracy of this subcategory of adrenocortical tumors.(12)
Currently, the only FDA approved adjuvant treatment of ACC is the adrenocorticolitic mitotane (Lysodren®). Adjuvant treatment with mitotane may increase recurrence-free survival but does not affect overall survival.(19,20) On top of this, treatment with mitotane comes with substantial, mainly gastrointestinal and neurological side effects.(21). According to consensus agreements and guidelines, adjuvant treatment with mitotane is indicated for patients with locoregional, high risk disease following surgical resection or for patients with metastatic disease.(22,23) The KI67 index is one of the main parameters used for patient selection for adjuvant therapy. However, its predictive capacity for treatment response has not been assessed. Furthermore, the scoring of Ki67 in tumor cells is subject to inter observer variability, which greatly affects the estimated survival.(24) The 62 identification of additional molecular markers may improve the selection of patients that could benefit from adjuvant treatment.
Over the last decades, multiple molecular alterations have emerged as factors in ACC carcinogenesis. Whole genome gene expression analysis of ACC identified two transcriptomic clusters with distinct gene expression patterns and survival. These clusters, denominated as C1a and C1b, corresponded with a five-year overall survival of 20% and 91% respectively.(25,26) Altered gene expression levels can be caused by many factors, both genetic and non-genetic. Mutations and copy number alterations may be one of these factors and are frequently reported in ACC. Mutations in genes involved in Wnt/B-catenin signaling such as CTNNB1, ZNRF3, APC and MEN1 can be found in up to 40% of ACCs, while the p53-Rb pathway is the second most affected signaling route (33% of ACCs) with alterations like inactivating mutations or homozygous deletions of TP53.(27)The proportion of TP53 mutations in benign adrenocortical tissues is low, whereas 6-catenin mutations can be found in both ACCs and ACAs.(28-31)
77 Epigenetic mechanisms are another factor involved in the regulation of gene expression. DNA methylation is one of these epigenetic regulatory mechanisms. In normal cells, DNA methylation is responsible for selective regulation of gene expression. The process of DNA methylation is mediated by DNA methyl transferases (DNMTs) and may occur throughout the genome in C-phosphate-G (CpG) dinucleotides. The genome is characterized by an overall low CpG content, except for areas in gene promoter regions. The CpG dinucleotides cluster in gene promotor regions, so called CpG islands.(32) Methylation of CpG islands in these regions mechanically blocks transcription by creating a physical barrier.(33) Methyl-CpG binding proteins (MBD proteins) adhere to hypermethylated DNA regions and induce chromatin remodeling by recruiting histone deacetylases (HDACs) and chromatin remodeling complexes.(34,35) The opposite occurs in hypomethylated CpG areas. Methylation of CpG islands may affect gene expression, for example by silencing oncoregulatory genes. As such, an increased prevalence of CpG island methylation has been proposed to be involved in carcinogenesis of several tumors. The exact molecular basis of this phenomenon remains to be explored.(43) 90 Denominated as CpG-island methylation phenotype (CIMP), this methylation pattern was identified for the first time in colon carcinoma.(36) Subsequently, CIMPs were identified in multiple other malignancies.(37-42) Although some similarities exist between the CIMP patterns in different tumors, it is likely that the molecular basis of the CIMP phenotype is tumor specific. Nevertheless an exact definition of (tumor specific) CIMP phenotypes is lacking.(43)
A second epigenetic regulatory mechanism involved in the regulation of chromatin and gene expression are histone modifications. Histones form the spool on which double stranded DNA is wound to form nucleosomes. A nucleosome consists of 146 basepairs wrapped around 8 histones. The eight histones are formed from 2 copies of H2A, H2b, H3 and H4. The amino acids forming the histone core and tails can be modified by various mechanisms such as methylation, acetylation,
50 51 52 53 54 55 56 57 58 59 60 61
63 64 65 66 67 68 69 70 71 72 73 74 75 76 78 79 80 81 82 83 84 85 86 87 88 89 91 92 93 94 95 96 97 98 99 100
phosphorylation, ubiquitination, sumoylation, citrullination and ADP-ribosylation.(44) These mechanisms may affect the order of amino acids, ultimately forming a “histone code” which can attract enzymes that in turn could influence the chromatin structure. Acetylation and deacetylation of histone tails is performed by histone modifying proteins such as histone acetyltransferases (HATs) and HDACS. (45,46) Ultimately, the level of chromatin condensation influences gene expression. If relaxed due to a mechanism such as hypomethylation, chromatin allows for transcription and gene expression is promoted. Hypermethylation may cause chromatin to condense, thereby blocking transcriptional factors and silencing gene expression.(47-49) Epigenetic changes are defined as heritable chromosomal modifications that affect gene expression without changing the coding DNA sequence itself. These alterations might attribute to carcinogenesis.(50-53) Epigenetic dysregulation may affect gene expression either directly by regulating gene transcription or indirectly by affecting translation through factors such as non-coding RNAs. The role of non-coding RNAs in ACC has been extensively reviewed previously.(54,55) Therefore, this review will focus on DNA methylation, histone modifications and their potential clinical relevance in ACC. These epigenetic mechanisms may help to differentiate ACC from ACA. Furthermore, alterations of these mechanisms in ACC might help in the selection of patients that can benefit from adjuvant therapy.
METHODS
Search strategy
Data collection and analysis were performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.(56) The Intelligent Gateway to Biomedical & Pharmacological Information (EMBASE) and the National Library of Medicine (MEDLINE/PubMed) were systematically searched. The search strategy is provided in supplementary table 1. All records published until 28th February 2017 were searched. Abstracts were independently screened by 2 authors (PJ & VM) and non-relevant articles were excluded based on in- and exclusion criteria. References of articles included for full-text review were checked for relevant publications. Full-text review of remaining records was performed by both authors (PJ & VM) independently. Non- relevant articles were excluded according to in- and exclusion criteria. In case of disagreement, consensus was reached through discussion with all authors.
Selection of studies
Inclusion criteria were English written studies reporting on methylation or histone modification in ACC in humans. Reviews, conference abstracts, case reports, non-English articles, in vitro studies, animal studies and studies in patients < 18 year were excluded.
RESULTS
The initial search of EMBASE and MEDLINE identified 376 records, of which 127 were duplicates. Abstract analysis excluded 221 records. One additional record was identified via references of records included for full-text review. Following full text-review of 30 articles, 6 articles were removed based on in- and exclusion criteria. Finally, 24 articles were included for this review (Figure 1, supplementary table 2). This systematic literature review aimed to provide an extensive overview of all relevant articles regarding DNA methylation and histone modifications in adult patients with ACC. The applied search strategy aimed to maximize sensitivity and not specificity for articles regarding these subjects. The downside of the method applied for this review lies in the fact that studies were not assessed for quality of their study design. The majority of studies included for this review focused on alterations in DNA methylation and their potential relevance for future clinical applications. Dysregulated DNA methylation patterns might be highly relevant for the differentiation of ACCs from benign adrenal tissues or to predict overall survival for individual patients.
Genome wide DNA methylation patterns and their association with survival
101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151
Six studies applied whole genome methylation profiling to characterize adrenocortical tumors based on methylation levels.(27,57-61) When relevant, we will describe the methods used for analysis of methylation data and hierarchical clustering to illustrate possible biases that may explain inter study variability in reported results. Each of these studies quantified methylation levels using either ß- value or M-value metrics. The ß-value has a more intuitive biological representation. It is preferable to use the M-value for methylation analysis due to its superior statistical validity, expressed by a better detection rate (DR) and true positive rate (TPR) for both highly methylated and unmethylated CpG sites.(62)
Three methylation profiling studies compared methylation levels between ACC, ACA and normal adrenocortical tissue (NAC).(57,60,61) The first whole genome methylation profiling study identified a global hypomethylation signature (adjusted P ≤ 0.01) by comparing 8 ACC, 12 metastatic ACC, 48 ACAs and 19 NAC samples.(57) Furthermore, it was found that hierarchical clustering (unknown metric) of methylation profiling data could separate ACC from benign adrenal tissue (ACA and NAC) with the exception of one ACC and one benign adrenal sample that consistently clustered with samples from the opposite histotype.(57) Rechache analyzed whole genome methylation levels of CpG islands, shores (0-2 kb from CpG islands), shelves (2-4 kb from CpG islands) and open sea loci (isolated loci without destination) using the Infinium HumanMethylation (HM)450 Beadchip (Illumina Inc, San Diego, CA). This assay includes 485,000 CpG methylation sites, covering 99% of RefSeq genes, 96% of CpG islands and 92% of CpG shores. Methylation data was quantified with the B-value method (cutoff values P ≤ 0.01 and 46 ≤ -0.20 or 40 ≥0.20). Legendre quantified the difference of methylation levels between 17 ACCs, 1 metastatic ACC and 6 NAC samples. This study confirmed a significant (P < 0.0001) level of hypomethylation in ACC compared to normal adrenocortical tissue (cutoff values P ≤ 0.01 and 40 ≤ -0.20 or 4₿ ≥ 0.20).(60) Furthermore, it was found that unsupervised cluster analysis (Euclidian distance) could separate ACCs from NACs based on altered methylation levels.(60) The third study that applied the HM450 array to assess methylation levels between ACC and normal tissue analyzed 10 ACCs, 75 ACAs and 2 NACs.(61) It was seen that methylation differences between ACC and NAC were higher compared to ACC and ACA. However, no statistical information was provided to illustrate these findings. 70-77% of the hypomethylated areas that differentiate ACC from ACA and normal adrenocortical tissue are situated in open sea regions, whereas hypermethylated areas are mainly situated in CpG island promoter regions (40-50%).(57,60)
The level of hypermethylation of CpG promoter regions in ACC is inversely correlated with overall survival.(27,59,63,64) The first study that described this correlation also proposed to categorize ACCs based on methylation level in CpG-island methylation phenotype (CIMP) groups.(58) This study used the Infinium HM27 Beadchip array (Illumina Inc, San Diego, CA) to analyze methylation levels of 51 ACCs and 84 ACAs (supplementary table 2). The HM27 analyses methylation levels of 27,578 CpG sites in proximal promoter regions of 14,475 genes. Methylation data was quantified using the M- value method, followed by unsupervised hierarchical clustering of the top 2824 CpGs with the highest methylation variability (SD > 1.5) in ACC (Manhattan metric, Ward method). First, two distinct methylation clusters were identified. These were denominated as non-CIMP (significantly hypermethylated compared to ACA) and CIMP (significantly hypermethylated compared to ACA and non-CIMP). The CIMP cluster was further subdivided in a CIMP-high (highest methylation level) and CIMP-low cluster (intermediate methylation level). No information about the significance of methylation levels between CIMP-high and CIMP-low clusters was provided, which may suggest that this was not significant. ACCs assigned to a CIMP and non-CIMP cluster were associated with 5-year overall survival of +35% and +62% respectively (log rank P = 0.04). The prognosis of CIMP-high compared to non-CIMP ACCs was even poorer, with 5-year overall survival of 0% and ±62% respectively (P = 0.006). More recently, the hypothesis on the existence of three methylation clusters in ACC (CIMP-high, CIMP-intermediate and CIMP-low) was confirmed by a study from The
152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202
Cancer Genome Atlas (TCGA) network.(59) The authors also concluded that assignment to each of these methylation clusters is associated with a different overall survival (log-rank p-value 0.0001). The patients in the CIMP-high and CIMP-low clusters had a three-year overall survival of less than 50% or 96% respectively. Methylation levels of 79 ACCs included in the TCGA study were acquired using Illumina HM450 arrays (supplementary table 2) and quantified with ß-value metrics and subsequent unsupervised clustering of the most variable 1% of CpG island promoter probes (Euclidean metrics, partitioning around medoids). Although the method used for data analysis is different from Barreau et al., Zheng concluded that the identified methylation clusters in both studies are similar. Interestingly, another study identified four distinct methylation clusters in ACC; CIMP-high, CIMP-low and two clusters within the non-CIMP group.(27) The first non-CIMP cluster was characterized by hypomethylation of CpG sites outside CpG islands, while the second non-CIMP cluster had low number of differentially methylated CpG sites. The reported 5-year overall survival for CIMP-high, CIMP-low and the two non-CIMP clusters was 0%, +35%, 50% and 90% respectively. The authors applied the same methodology as Barreau et al. to analyze the methylation levels of 51 ACCs and 30 ACAs with the Illumina Infinium HM27 assay. However, they applied a recursively partitioned mixture model (RPMM) using the 10% most variant CpG sites for cluster analysis, which is different from methods used by both Barreau and Zheng.(65)
We may conclude from these studies that methylation levels can be used to distinguish ACCs from ACAs and normal tissues. For the highest proportions of ACC, the level of methylation of DNA promoter regions is inversely correlated with overall survival. Furthermore, ACCs can be categorized in clusters with distinct methylation patterns and overall survival. The three key studies that defined methylation clusters in ACC used different methods to assess methylation levels and define robust methylation clusters. (27,58,59) The reported CIMP-phenotypes therefore might represent categories with different biological behavior. A possible explanation for conflicting results regarding the number of methylation clusters might be the proportion of most variant CpG sites used for cluster analysis. Inclusion of a higher proportion of aberrant methylated CpG sites might provide the statistical power to define a larger number of distinct methylation clusters. Furthermore, a lack of a definition of the ACC “CIMP-phenotype” might have attributed to different methodologies. This may be another explanation why one study identified four CIMP-phenotypes in ACC, while two other studies observed three CIMPs. Future studies should define the number of methylation clusters based on consensus. As described below, some studies already validated ACC methylation clusters in separate cohorts. Methylation levels might be useful as tool to differentiate ACCs from ACAs or to predict overall survival. CIMP and non-CIMP phenotypes may help to identify patients with poor prognosis. Prospective studies are warranted to define consensus based methylation clusters and to establish the diagnostic and therapeutic relevance of aberrant methylation levels in ACC. Interestingly, each of key studies uploaded their data in the public domain. This comes with the possibility to combine raw methylation assay data from each study and re-analyze this dataset with a consensus based methodology. Subsequently, a methylation probe set that accurately describes the predefined CIMP-phenotype can be designed and validated in either FFPE tissue or a prospective cohort. Ultimately, this might attribute to the establishment of a well-defined, reproducible and clinical relevant CIMP-phenotype in ACC.(43)
DNA methylation and gene expression
The identification of CpG islands in promoter regions of individual genes that are specifically hypermethylated in ACC might help to differentiate ACC from ACA or NAC. Since the level of DNA methylation is inversely correlated with overall survival, identification of hypermethylated genes with low mRNA expression might also be relevant to predict prognosis. 16 studies included for this review reported both methylation and mRNA expression levels for individual genes in ACC and ACA and/or NAC.(57-60,66-77) Two of these studies found low methylation levels in promoter regions of TP53 and MGMT, hypothesizing that other mechanisms are involved in the regulation of these genes
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253
in ACC.(71,77) The remaining 14 studies described a relation between one or more hypermethylated promoter regions and subsequent downregulated expression of the associated gene (supplementary table 4). Interestingly, a poor overlap exists between hypermethylated genes with low mRNA expression identified in these studies. Only 14 genes were reported by two or more studies (table 1). A possible explanation for the poor overlap of the reported hypermethylated and downregulated genes between studies could be found in differences in analyzed tissue types, number of samples analyzed and methodologies used for the analysis itself (supplementary table 2). Another cause of the inter-study heterogeneity might be the variation in definitions used for hypermethylated, downregulated genes (supplementary table 3). The H19 gene was the most frequently reported individual gene that is both hypermethylated and subsequently has low mRNA expression (supplementary table 4). Already in 1994, the first alteration in DNA methylation in ACC was reported at the 11p15 locus, encoding H19 and IGF2.(78) The maternal copy of 11p15 in normal adrenal tissue is unmethylated at the Imprinting Center Region (ICR) 1. This region is associated with expression of the non-coding tumor suppressor H19 in normal adrenocortical tissue, while the ICR1 region of the paternal copy is methylated and expresses IGF2. ACCs have a significant higher methylation levels at H19 promoter regions compared to ACAs or normal tissue, which causes low H19 and high IGF2 mRNA expression levels.(66,72,79) This is likely to be caused by paternal uniparental disomy at the 11p15 locus, a late genetic events in progression of benign adrenal adenomas towards ACC.(79,80) LOH of the 11p15 locus can be found in nearly 95% of ACCs.(80) Interestingly, some ACCs with similar LOH and tumor phenotypes have normal IGF2 and moderate H19 expression levels, which might be caused by a decreased ICR1 methylation levels.(81) Altogether, differentially methylated control regions at the 11p15 locus are involved in ACC carcinogenesis and may differentiate ACCs from benign tissues. More recently, mRNA expression of RARRES2 was found to be downregulated as a consequence of promoter hypermethylation.(57,76) RARRES2 overexpression in ACC induced reduced cell invasion, cell proliferation and tumorigenicity in vitro and in vivo. These observations might be caused by reduced phosphorylation and degradation of ß-catenin or phosphorylation of p38 mitogen-activated protein kinase (MAPK). These mechanism were found to be immune-independent, because it does not depend on the recruitment of CMKLR1-expressing immune cells in vitro.(76) The mechanisms behind hypermethylation and subsequent low mRNA expression in ACC are not reported for the remaining 12 genes (CD74, HSD38B2, S100A6, CYP7B1, GIPC2, GYPC, GOS2, TM7SF2, RAB34, MLH3, SPINT2, SLC7A4).(57-60)
Three studies aimed to provide insight in the correlation between methylation clusters and transcriptomic clusters in ACC.(27,58,59) Transcriptomic clusters categorize ACCs based on robust clusters that represent patterns of differentially expressed mRNA levels. The characteristics of a robust cluster are dependent on the methods and cutoff values used to define the cluster. As discussed previously, variation in both methods and cutoff values might lead to different outcomes. This may result in difficulties when comparing results from hierarchical clustering between different studies. Analysis of the methods used to define the transcriptomic clusters of the three aforementioned studies is beyond the scope of this review. With this potential bias in mind, we will describe the relation between methylation transcriptomic clusters as reported by these studies. There is evidence that transcriptomic clusters in ACC are strongly correlated with subgroups based on DNA methylation (P = 6.3 x 10-5, x2 test).(27) All studies reported data from an integrated analysis of multiple -omic domains, including transcriptomic and methylation data, for individual patients. Each of the three studies identified at least two transcriptomic clusters, consistently denominated as C1a (poor overall survival) and C1b (better overall survival) as described by de Reyniès.(26) 90-100% of the ACCs with CIMP-high phenotypes identified in these three studies were associated with a C1a transcriptomic phenotype.(27,58,59) Assie et al. divided ACCs with a C1a phenotype in three subgroups based on their association with DNA methylation clusters. Besides the CIMP-high and CIMP-low phenotypes, a third methylation cluster denominated as “hypomethylated” was associated with the poor prognosis C1a transcriptomic cluster.(27) Overall, 24-31 % of the ACCs with
254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304
the lowest methylation level phenotypes (denominated as non-CIMP(27,58) or CIMP-low(59)) reported in the three studies were associated with the C1a phenotype. This could be interesting, because the reported 5-year overall of these low methylation level ACC phenotypes ranges between 65-96%, while it ranges between 20-35 % for ACCs associated with the C1a transcriptomic cluster.(27,58,59) This observation might suggest the existence of a subcategory of hypomethylated ACCs that is associated with a poor prognosis transcriptomic phenotype.
Altogether, 14 genes reported in 2 or more studies have both promoter hypermethylation and lower mRNA expression in ACC compared to benign adrenal tissue. Methylation levels of CpG promoter regions of these genes may be useful to differentiate ACC from ACA or NAC. Furthermore, 24-31 % of ACCs with a hypomethylation phenotype (mainly associated with a good overall survival) were associated with a C1a transcriptomic phenotype (associated with a poor prognosis). This observation might affect the power of methylation levels to predict prognosis.
Histone modifications in ACC
This systematic literature review identified one study that aimed to identify altered mRNA expression levels of genes involved in histone modifications, while another study briefly mentioned altered expression levels of histone modification genes in ACC. The authors of the first study screened mRNA levels of histone methyltransferases, demethylases and associated factors. EZH2 was identified as the most significantly upregulated histone modifying factor in ACCs. EZH2 is part of the Polycomb Repressor Complex 2 (PRC2) and functions as catalytic core protein. It catalyzes the trimethylation of histone H3 lysine 27 (H3K27me3), thereby attributing to the silencing of target genes. EZH2 overexpression is correlated with advanced stages of cancer and poor prognosis.(82) High EZH2 expression levels are related to poor prognosis due to significant expression in the previously described poor prognosis C1A transcriptomic cluster. This may suggest an involvement of this gene in aggressive ACC subtypes. Both gene set enrichment analyses indicated that EZH2 is associated with proliferation. Zeng briefly reported that mRNA expression levels of histone modification genes (MLL, MLL2, and MLL4) and chromatin remodeling genes (ATRX and DAXX) were dysregulated in 22% of the analyzed samples.(59) Future studies should further clarify the clinical relevancy of the altered mRNA expression of histone modifying genes in ACC.
The clinical application of epigenetic signatures characterizing ACC
Before the epigenetic signatures that characterize ACC can be clinically applied, development and subsequent validation of tools that accurately quantify these signatures are warranted. This requires a well-defined and validated methylation probe set that either accurately differentiates ACC from ACA or identifies distinct CIMP-phenotypes in ACC.
Multiple studies reported that ACC might be differentiated from benign tissues based on methylation levels.(57,58,60,72) The only study that investigated the accuracy of DNA methylation levels for diagnostic purposes focused on methylation levels of H19/IGF2 regulatory regions.(72) In this study a discovery cohort of 24 ACCs, 14 ACAs and 11 NACs was established to determine the combination of differentially methylated areas that yielded the highest diagnostic accuracy. Mean methylation levels for ACCs and ACAs were quantified as absolute standard deviation scores (| SDS|) compared to NAC methylation levels. The most optimal sensitivity of 96% and specificity of 100% (cutoff value mean |SDS | 2.617) was observed when methylation levels of H19 promoter region, 2nd - 4th CpG in DMR2 and 5th - 7th CpG in CTCF 3 were used to predict prognosis. A sensitivity of 89 % and specificity of 92 % (cutoff value mean | SDS| 2,617) was calculated from an independent validation cohort of 9 ACCs and 13 ACAs. Based on DNA methylation levels of the selected regions with the highest diagnostic accuracy, “borderline” tumors excluded for study participation were predicted to be ACCs with an | SDS | of 4.29 and 4.20 respectively. While follow-up is ongoing, it is interesting to note that these patients did not develop metastasis during a follow-up period of 27
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355
and 8 months respectively. Therefore, H19 promoter methylation status might be helpful to improve diagnostic accuracy of adrenocortical tumors. However, the majority of methylation assays used in clinical practice to diagnose colorectal, prostate or bladder cancer use a combination of two or more molecular markers to obtain an adequate diagnostic power.(83-87) It is therefore plausible that additional molecular markers are needed to acquire a sufficient accuracy for a methylation based diagnostic test in ACC that accurately diagnoses borderline tumors.
Zheng aimed to identify a probe set that could accurately predict the CIMP-phenotypes as defined in their study.(59) The probe set was established by applying a random forest classifier together with the Boruta method (no cut-off values provided), using the 1% most variable probes commonly expressed from the Illumina Infinium HM450 arrays of two methylation profiling cohorts.(27,59) The number of ACC samples used for this analysis was not reported. Finally, a set of 68 methylation probes was selected and categorized ACCs to their predefined CIMP phenotypes with 92 % accuracy. To validate the prognostic value of these probes, a second cohort of 83 patients was classified in CIMP-phenotypes. Patients assigned to the CIMP-high category were associated with a poor 3-year overall survival of less than 50%, while patients in the CIMP-low category had a three-year overall survival of 96%. These results are similar to the overall survival of matching CIMP phenotypes in the discovery cohort, suggesting that this probe set accurately may predict CIMP phenotypes and prognosis.(59)
Barreau et al. reported that multiplex ligation-dependent probe amplification (MS-MLPA) can be used to identify CIMP-phenotypes in ACC and predict overall survival and disease free survival based on methylation levels.(58,64) Based on these observations, a recent study identified 4 methylation probes (GSTP1, PYCARD, PAX6 and PAX5) as a simplified molecular tool to predict prognosis using the MS-MLPA platform.(64) The probes were selected from a set of 27 methylation probes, based on their correlation with CpG island methylation status as defined by the Illumina Infinium HM27 array (58) (Pearson correlation, P < 0.05), association with disease free survival and overall survival (Cox regression, P < 0.05) in a retrospective training cohort of 50 ACCs. Subsequently, the 4 probes were tested in a retrospective validation cohort of 203 ACCs. Methylation levels measured by MS-MLPA were a significant prognostic factor of disease free survival (P < 0.0001) and overall survival (P < 0.0001). Multivariate analysis showed that the prognostic value of methylation is independent from the gold standard KI67 index and tumor stage for recurrence (HR 1.012, P = 0.0005) and death (HR 1,014, P = 0.0006) per 1% increase of methylation levels. The authors combined methylation, KI67 index and ENSAT stage in a prognostic score that robustly predicted disease free survival (log-rank P < 10-16) and overall survival (log-rank P < 10-16).
These first steps towards the translation of altered DNA methylation levels in ACC towards clinical application are promising. The studies indicate that altered DNA methylation patterns might be useful to differentiate ACC from benign adrenal tumors and to predict survival. Future, prospective studies in larger cohorts should be executed to validate these results and quantify the accuracy of pyrosequencing and MS-MLPA for diagnostic and prognostic purposes in ACC.
CONCLUSION
Epigenetic regulatory mechanisms have emerged as a potential clinically relevant factor in ACC. The most important clinical benefits of epigenetic dysregulation in ACC might be found in the improvement of diagnostic tools and prediction of the benefit of individual patients for adjuvant treatment. The main epigenetic markers that could be relevant for these purposes are DNA methylation levels and CIMP-phenotypes. The therapeutic potential of epigenetic alterations in ACC is unclear and needs further research. Future studies should also focus on the feasibility of diagnostic tests that predict diagnosis of adrenocortical tumors based on altered methylation levels.
356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
ACCEPTED MANUSCRIPT
Furthermore, a consensus based CIMP-phenotype needs to be established and requires subsequent prospective validation in independent cohorts.
REFERENTIES
1. Fassnacht M, Libé R, Kroiss M, Allolio B. Adrenocortical carcinoma: a clinician’s update. Nat Rev Endocrinol. 2011 Jun;7(6):323-35.
2. Søreide JA, Brabrand K, Thoresen SO. Adrenal cortical carcinoma in Norway, 1970-1984. World J Surg. 1992 Jul;16(4):663-7-discussion668.
3. Abiven G, Coste J, Groussin L, Anract P, Tissier F, Legmann P, et al. Clinical and biological features in the prognosis of adrenocortical cancer: poor outcome of cortisol-secreting tumors in a series of 202 consecutive patients. The Journal of Clinical Endocrinology & Metabolism. Endocrine Society; 2006 Jul;91(7):2650-5.
4. Luton JP, Cerdas S, Billaud L, Thomas G, Guilhaume B, Bertagna X, et al. Clinical features of adrenocortical carcinoma, prognostic factors, and the effect of mitotane therapy. N Engl J Med. Massachusetts Medical Society; 1990 Apr 26;322(17):1195-201.
5. Pommier RF, Brennan MF. An eleven-year experience with adrenocortical carcinoma. Surgery. 1992 Dec;112(6):963-70-discussion970-1.
6. Icard P, Goudet P, Charpenay C, Andreassian B, Carnaille B, Chapuis Y, et al. Adrenocortical carcinomas: surgical trends and results of a 253-patient series from the French Association of Endocrine Surgeons study group. World J Surg. 2001 Jul;25(7):891-7.
7. Haak HR, Hermans J, van de Velde CJ, Lentjes EG, Goslings BM, Fleuren GJ, et al. Optimal treatment of adrenocortical carcinoma with mitotane: results in a consecutive series of 96 patients. Br J Cancer. Nature Publishing Group; 1994 May;69(5):947-51.
8. Bellantone R, Ferrante A, Boscherini M, Lombardi CP, Crucitti P, Crucitti F, et al. Role of reoperation in recurrence of adrenal cortical carcinoma: results from 188 cases collected in the Italian National Registry for Adrenal Cortical Carcinoma. Surgery. 1997 Dec;122(6):1212- 8.
9. 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. Wiley Subscription Services, Inc., A Wiley Company; 2009 Jan 15;115(2):243-50.
10. Libé R, Borget I, Ronchi CL, Zaggia B, Kroiss M, Kerkhofs T, et al. Prognostic factors in stage III- IV adrenocortical carcinomas (ACC): an European Network for the Study of Adrenal Tumor (ENSAT) study. Ann Oncol. 2015 Oct;26(10):2119-25.
11. Bilimoria KY, Shen WT, Elaraj D, Bentrem DJ, Winchester DJ, Kebebew E, et al. Adrenocortical carcinoma in the United States: treatment utilization and prognostic factors. Cancer. 2008 Dec 1;113(11):3130-6.
12. Giordano TJ. Adrenocortical tumors: an integrated clinical, pathologic, and molecular approach at the University of Michigan. 2010 Oct;134(10):1440-3.
13. de Krijger RR, Papathomas TG. Adrenocortical neoplasia: evolving concepts in tumorigenesis
407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447
with an emphasis on adrenal cortical carcinoma variants. Virchows Arch. 2012 Jan;460(1):9- 18.
14. Lau SK, Weiss LM. The Weiss system for evaluating adrenocortical neoplasms: 25 years later. Human Pathology. 2009 Jun;40(6):757-68.
15. Weiss LM, Medeiros LJ, Vickery AL. Pathologic features of prognostic significance in adrenocortical carcinoma. Am J Surg Pathol. 1989 Mar;13(3):202-6.
16. Aubert S, Wacrenier A, Leroy X, Devos P, Carnaille B, Proye C, et al. Weiss system revisited: a clinicopathologic and immunohistochemical study of 49 adrenocortical tumors. Am J Surg Pathol. 2002 Dec;26(12):1612-9.
17. Klibanski A, Stephen AE, Greene MF, Blake MA, Wu C-L. Case records of the Massachusetts General Hospital. Case 36-2006. A 35-year-old pregnant woman with new hypertension. N Engl J Med. 2006 Nov 23;355(21):2237-45.
18. Pohlink C, Tannapfe A, Eichfeld U, Schmidt F, Führer D, Paschke R, et al. Does tumor heterogeneity limit the use of the Weiss criteria in the evaluation of adrenocortical tumors? J Endocrinol Invest. 2004 Jun;27(6):565-9.
19. Terzolo M, Angeli A, Fassnacht M, Daffara F, Tauchmanova L, Conton PA, et al. Adjuvant mitotane treatment for adrenocortical carcinoma. N Engl J Med. 2007 Jun 7;356(23):2372-80.
20. Postlewait LM, Ethun CG, Tran TB, Prescott JD, Pawlik TM, Wang TS, et al. Outcomes of Adjuvant Mitotane after Resection of Adrenocortical Carcinoma: A 13-Institution Study by the US Adrenocortical Carcinoma Group. J Am Coll Surg. 2016 Apr;222(4):480-90.
21. Hoff AO, Berruti A. 5th International ACC Symposium: Future and Current Therapeutic Trials in Adrenocortical Carcinoma. Horm Cancer. 2016 Feb;7(1):29-35.
22. Berruti A, Fassnacht M, Baudin E, Hammer G, Haak H, Leboulleux S, et al. Adjuvant therapy in patients with adrenocortical carcinoma: a position of an international panel. J Clin Oncol. 2010 Aug 10;28(23):e401-2-authorreplye403.
23. Kulke MH, Shah MH, Benson AB, Bergsland E, Berlin JD, Blaszkowsky LS, et al. Neuroendocrine tumors, version 1.2015. Vol. 13, Journal of the National Comprehensive Cancer Network : JNCCN. 2015. pp. 78-108.
24. 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 Apr;40(4):569- 76.
25. Assié G, Guillaud-Bataille M, Ragazzon B, Bertagna X, Bertherat J, Clauser E. The pathophysiology, diagnosis and prognosis of adrenocortical tumors revisited by transcriptome analyses. Trends Endocrinol Metab. 2010 May;21(5):325-34.
26. de Reyniès A, Assié G, Rickman DS, Tissier F, Groussin L, René-Corail F, et al. Gene expression profiling reveals a new classification of adrenocortical tumors and identifies molecular predictors of malignancy and survival. J Clin Oncol. American Society of Clinical Oncology; 2009 Mar 1;27(7):1108-15.
27. Assié G, Letouze E, Fassnacht M, Jouinot A, Luscap W, Barreau O, et al. Integrated genomic
448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486
| ACCEPTED MANUSCRIPT | ||
|---|---|---|
| 487 | characterization of adrenocortical carcinoma. Nat Genet. 2014 Jun;46(6):607-12. | |
| 488 489 | 28. | Ohgaki H, Kleihues P, Heitz PU. p53 mutations in sporadic adrenocortical tumors. Int J Cancer. 1993 May 28;54(3):408-10. |
| 490 | 29. | Reincke M, Karl M, Travis WH, Mastorakos G, Allolio B, Linehan HM, et al. p53 mutations in |
| 491 | human adrenocortical neoplasms: immunohistochemical and molecular studies. The Journal | |
| 492 | of Clinical Endocrinology & Metabolism. 1994 Mar;78(3):790-4. | |
| 493 | 30. | Libé R, Groussin L, Tissier F, Elie C, René-Corail F, Fratticci A, et al. Somatic TP53 mutations are |
| 494 | relatively rare among adrenocortical cancers with the frequent 17p13 loss of heterozygosity. | |
| 495 | Clin Cancer Res. American Association for Cancer Research; 2007 Feb 1;13(3):844-50. | |
| 496 | 31. | Tissier F, Cavard C, Groussin L, Perlemoine K, Fumey G, Hagnéré A-M, et al. Mutations of |
| 497 | beta-catenin in adrenocortical tumors: activation of the Wnt signaling pathway is a frequent | |
| 498 | event in both benign and malignant adrenocortical tumors. Cancer Res. 2005 Sep | |
| 499 | 1;65(17):7622-7. | |
| 500 | 32. | Takai D, Jones PA. Comprehensive analysis of CpG islands in human chromosomes 21 and 22. |
| 501 | Proc Natl Acad Sci USA. National Acad Sciences; 2002 Mar 19;99(6):3740-5. | |
| 502 | 33. | Rodriguez C, Borgel J, Court F, Cathala G, Forné T, Piette J. CTCF is a DNA methylation- |
| 503 | sensitive positive regulator of the INK/ARF locus. Biochem Biophys Res Commun. 2010 Feb | |
| 504 | 5;392(2):129-34. | |
| 505 | 34. | Ballestar E, Wolffe AP. Methyl-CpG-binding proteins. Targeting specific gene repression. Eur J |
| 506 | Biochem. 2001 Jan;268(1):1-6. | |
| 507 | 35. | Prokhortchouk E, Hendrich B. Methyl-CpG binding proteins and cancer: are MeCpGs more |
| 508 | important than MBDs? Oncogene. 2002 Aug 12;21(35):5394-9. | |
| 509 | 36. | Toyota M, Ahuja N, Ohe-Toyota M, Herman JG, Baylin SB, Issa JP. CpG island methylator |
| 510 | phenotype in colorectal cancer. Proc Natl Acad Sci USA. 1999 Jul 20;96(15):8681-6. | |
| 511 | 37. | Liu Z, Zhao J, Chen X-F, Li W, Liu R, Lei Z, et al. CpG island methylator phenotype involving |
| 512 | tumor suppressor genes located on chromosome 3p in non-small cell lung cancer. Lung | |
| 513 | Cancer. 2008 Oct;62(1):15-22. | |
| 514 | 38. | Strathdee G, Appleton K, Illand M, Millan DW, Sargent J, Paul J, et al. Primary ovarian |
| 515 | carcinomas display multiple methylator phenotypes involving known tumor suppressor | |
| 516 | genes. Am J Pathol. 2001 Mar;158(3):1121-7. | |
| 517 | 39. | Ueki T, Toyota M, Sohn T, Yeo CJ, Issa JP, Hruban RH, et al. Hypermethylation of multiple |
| 518 | genes in pancreatic adenocarcinoma. Cancer Res. 2000 Apr 1;60(7):1835-9. | |
| 519 | 40. | Arai E, Chiku S, Mori T, Gotoh M, Nakagawa T, Fujimoto H, et al. Single-CpG-resolution |
| 520 | methylome analysis identifies clinicopathologically aggressive CpG island methylator | |
| 521 | phenotype clear cell renal cell carcinomas. Carcinogenesis. Oxford University Press; 2012 | |
| 522 | Aug;33(8):1487-93. | |
| 523 | 41. | Abe M, Ohira M, Kaneda A, Yagi Y, Yamamoto S, Kitano Y, et al. CpG island methylator |
| 524 | phenotype is a strong determinant of poor prognosis in neuroblastomas. Cancer Res. 2005 | |
| 525 | Feb 1;65(3):828-34. |
42. Tanemura A, Terando AM, Sim M-S, van Hoesel AQ, de Maat MFG, Morton DL, et al. CpG island methylator phenotype predicts progression of malignant melanoma. Clin Cancer Res. American Association for Cancer Research; 2009 Mar 1;15(5):1801-7.
43. Hughes LAE, Melotte V, de Schrijver J, de Maat M, Smit VTHBM, Bovée JVMG, et al. The CpG island methylator phenotype: what’s in a name? Cancer Res. 2013 Oct;73(19):5858-68.
44. Bannister AJ, Kouzarides T. Regulation of chromatin by histone modifications. Cell Res. Nature Publishing Group; 2011 Mar;21(3):381-95.
45. Bird AP, Wolffe AP. Methylation-induced repression — belts, braces, and chromatin. CELL. 1999 Nov 24;99(5):451-4.
46. Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nat Rev Genet. 2002 Jun;3(6):415-28.
47. Verdone L, Agricola E, Caserta M, Di Mauro E. Histone acetylation in gene regulation. Brief Funct Genomic Proteomic. 2006 Sep;5(3):209-21.
48. Kuo MH, Allis CD. Roles of histone acetyltransferases and deacetylases in gene regulation. Bioessays. 1998 Aug;20(8):615-26.
49. de Ruijter AJM, van Gennip AH, Caron HN, Kemp S, van Kuilenburg ABP. Histone deacetylases (HDACs): characterization of the classical HDAC family. Biochem J. 2003 Mar 15;370(Pt 3):737-49.
50. Kulis M, Esteller M. DNA methylation and cancer. Adv Genet. Elsevier; 2010;70:27-56.
51. Clark SJ, Melki J. DNA methylation and gene silencing in cancer: which is the guilty party? Oncogene. Nature Publishing Group; 2002 Aug 12;21(35):5380-7.
52. Mukamel Z, Tanay A. Hypomethylation marks enhancers within transposable elements. Nat Genet. 2013 Jul;45(7):717-8.
53. Paz MF, Fraga MF, Avila S, Guo M, Pollan M, Herman JG, et al. A systematic profile of DNA methylation in human cancer cell lines. Cancer Res. 2003 Mar 1;63(5):1114-21.
54. Hassan N, Zhao JT, Sidhu SB. The role of microRNAs in the pathophysiology of adrenal tumors. Molecular and Cellular Endocrinology. Elsevier Ireland Ltd; 2017 Jan 4 ;: 1-8.
55. Igaz P, Igaz I, Nagy Z, Nyírő G, Szabó PM, Falus A, et al. MicroRNAs in adrenal tumors: relevance for pathogenesis, diagnosis, and therapy. Cell Mol Life Sci. 2015 Feb;72(3):417-28.
56. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015 Jan 1;4(1):1.
57. Rechache NS, Wang Y, Stevenson HS, Killian JK, Edelman DC, Merino M, et al. DNA methylation profiling identifies global methylation differences and markers of adrenocortical tumors. The Journal of Clinical Endocrinology & Metabolism. 2012 Jun;97(6):E1004-13.
58. Barreau O, Assié G, Wilmot-Roussel H, Ragazzon B, Baudry C, Perlemoine K, et al. Identification of a CpG island methylator phenotype in adrenocortical carcinomas. The
526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562
Journal of Clinical Endocrinology & Metabolism. 2013 Jan;98(1):E174-84.
59. Zheng S, Cherniack AD, Dewal N, Moffitt RA, Danilova L, Murray BA, et al. Comprehensive Pan-Genomic Characterization of Adrenocortical Carcinoma. Cancer Cell. 2016 May 9;29(5):723-36.
60. Legendre CR, Demeure MJ, Whitsett TG, Gooden GC, Bussey KJ, Jung S, et al. Pathway Implications of Aberrant Global Methylation in Adrenocortical Cancer. PLOS One. 2016;11(3):e0150629.
61. Gara SK, Wang Y, Patel D, Liu-Chittenden Y, Jain M, Boufragech M, et al. Integrated genome- wide analysis of genomic changes and gene regulation in human adrenocortical tissue samples. Nucleic Acids Res. 2015 Oct;43(19):9327-39.
62. Du P, Zhang X, Huang C-C, Jafari N, Kibbe WA, Hou L, et al. Comparison of Beta-value and M- value methods for quantifying methylation levels by microarray analysis. BMC Bioinformatics. BioMed Central; 2010 Nov 30;11(1):587.
63. Barreau O, de Reyniès A, Wilmot-Roussel H, Guillaud-Bataille M, Auzan C, René-Corail F, et al. Clinical and Pathophysiological Implications of Chromosomal Alterations in Adrenocortical Tumors: An Integrated Genomic Approach. The Journal of Clinical Endocrinology & Metabolism. Endocrine Society Chevy Chase, MD; 2011 Nov 23;97(2):E301-11.
64. Jouinot A, Assié G, Libé R, Fassnacht M, Papathomas T, Barreau O, et al. DNA methylation is an independent prognostic marker of survival in adrenocortical cancer. J Clin Endocrinol Metab. 2016 Dec 14 ;: jc20163205.
65. Houseman EA, Christensen BC, Yeh R-F, Marsit CJ, Karagas MR, Wrensch M, et al. Model- based clustering of DNA methylation array data: a recursive-partitioning algorithm for high- dimensional data arising as a mixture of beta distributions. BMC Bioinformatics. BioMed Central; 2008 Sep 9;9(1):365.
66. Gao Z-H, Suppola S, Liu J, Heikkila P, Jänne J, Voutilainen R. Association of H19 promoter methylation with the expression of H19 and IGF-II genes in adrenocortical tumors. The Journal of Clinical Endocrinology & Metabolism. 2002 Mar;87(3):1170-6.
67. Simi L, Malentacchi F, Luciani P, Gelmini S, Deledda C, Arvia R, et al. Seladin-1 expression is regulated by promoter methylation in adrenal cancer. BMC Cancer. 2010 May;10:201.
68. Korah R, Healy JM, Kunstman JW, Fonseca AL, Ameri AH, Prasad ML, et al. Epigenetic silencing of RASSF1A deregulates cytoskeleton and promotes malignant behavior of adrenocortical carcinoma. Mol Cancer. 2013 Aug 5;12(1):87.
69. Mitsui Y, Yasumoto H, Nagami T, Hiraki M, Arichi N, Ishikawa N, et al. Extracellular activation of Wnt signaling through epigenetic dysregulation of Wnt inhibitory factor-1 (Wif-1) is associated with pathogenesis of adrenocortical tumor. Oncotarget. 2014 Apr;5(8):2198-207.
70. Pilon C, Rebellato A, Urbanet R, Guzzardo V, Cappellesso R, Sasano H, et al. Methylation Status of Vitamin D Receptor Gene Promoter in Benign and Malignant Adrenal Tumors. Int J Endocrinol. 2015;2015(1):375349-7.
71. Creemers SG, Van Koetsveld PM, Van Den Dungen ESR, Korpershoek E, van Kemenade FJ, Franssen GJH, et al. Inhibition of Human Adrenocortical Cancer Cell Growth by Temozolomide
563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602
in Vitro and the Role of the MGMT Gene. J Clin Endocrinol Metab. 2016 Dec;101(12):4574- 84.
72. Creemers SG, Van Koetsveld PM, van Kemenade FJ, Papathomas TG, Franssen GJH, Dogan F, et al. Methylation of IGF2 regulatory regions to diagnose adrenocortical carcinomas. Endocr Relat Cancer. BioScientifica; 2016 Sep;23(9):727-37.
73. Fonseca AL, Kugelberg J, Starker LF, Scholl U, Choi M, Hellman P, et al. Comprehensive DNA methylation analysis of benign and malignant adrenocortical tumors. Genes Chromosomes Cancer. 2012 Oct;51(10):949-60.
74. Cheng JY, Brown TC, Murtha TD, Stenman A, Juhlin CC, Larsson C, et al. A novel FOX01- mediated dedifferentiation blocking role for DKK3 in adrenocortical carcinogenesis. BMC Cancer. 2017 Mar;17(1):164.
75. Hofland J, Steenbergen J, Voorsluijs JM, Verbiest MMPJ, de Krijger RR, Hofland LJ, et al. Inhibin alpha-subunit (INHA) expression in adrenocortical cancer is linked to genetic and epigenetic INHA promoter variation. PLOS One. 2014;9(8):e104944.
76. Liu-Chittenden Y, Jain M, Gaskins K, Wang S, Merino MJ, Kotian S, et al. RARRES2 functions as a tumor suppressor by promoting ß-catenin phosphorylation/degradation and inhibiting p38 phosphorylation in adrenocortical carcinoma. Oncogene. Nature Publishing Group; 2017 Jan 23;23:viii6.
77. Sidhu S, Martin E, Gicquel C, Melki J, Clark SJ, Campbell P, et al. Mutation and methylation analysis of TP53 in adrenal carcinogenesis. Eur J Surg Oncol. 2005 Jun;31(5):549-54.
78. Gicquel C, Bertagna X, Schneid H, Francillard-Leblond M, Luton JP, Girard F, et al. Rearrangements at the 11p15 locus and overexpression of insulin-like growth factor-Il gene in sporadic adrenocortical tumors. The Journal of Clinical Endocrinology & Metabolism. 1994 Jun; 78(6):1444-53.
79. Nielsen HM, How-Kit A, Guerin C, Castinetti F, Vollan HKM, De Micco C, et al. Copy number variations alter methylation and parallel IGF2 overexpression in adrenal tumors. Endocr Relat Cancer. 2015 Dec;22(6):953-67.
80. Gicquel C, Raffin-Sanson ML, Gaston V, Bertagna X, Plouin PF, Schlumberger M, et al. Structural and functional abnormalities at 11p15 are associated with the malignant phenotype in sporadic adrenocortical tumors: study on a series of 82 tumors. The Journal of Clinical Endocrinology & Metabolism. 1997 Aug;82(8):2559-65.
81. Guillaud-Bataille M, Ragazzon B, de Reynies A, Chevalier C, Francillard I, Barreau O, et al. IGF2 promotes growth of adrenocortical carcinoma cells, but its overexpression does not modify phenotypic and molecular features of adrenocortical carcinoma. PLOS One. 2014;9(8):e103744.
82. Drelon C, Berthon A, Mathieu M, Ragazzon B, Kuick R, Tabbal H, et al. EZH2 is overexpressed in adrenocortical carcinoma and is associated with disease progression. Hum Mol Genet. 2016 Jul;25(13):2789-800.
83. Imperiale TF, Ransohoff DF, Itzkowitz SH, Levin TR, Lavin P, Lidgard GP, et al. Multitarget stool DNA testing for colorectal-cancer screening. N Engl J Med. Massachusetts Medical Society; 2014 Apr 3;370(14):1287-97.
603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643
84. Church TR, Wandell M, Lofton-Day C, Mongin SJ, Burger M, Payne SR, et al. Prospective evaluation of methylated SEPT9 in plasma for detection of asymptomatic colorectal cancer. Gut. 2014 Feb;63(2):317-25.
85. Stewart GD, Van Neste L, Delvenne P, Delrée P, Delga A, McNeill SA, et al. Clinical utility of an epigenetic assay to detect occult prostate cancer in histopathologically negative biopsies: results of the MATLOC study. J Urol. 2013 Mar;189(3):1110-6.
86. Partin AW, Van Neste L, Klein EA, Marks LS, Gee JR, Troyer DA, et al. Clinical validation of an epigenetic assay to predict negative histopathological results in repeat prostate biopsies. J Urol. 2014 Oct;192(4):1081-7.
P
87. Renard I, Joniau S, van Cleynenbreugel B, Collette C, Naômé C, Vlassenbroeck I, et al. Identification and validation of the methylated TWIST1 and NID2 genes through real-time methylation-specific polymerase chain reaction assays for the noninvasive detection of primary bladder cancer in urine samples. Eur Urol. 2010 Jul;58(1):96-104.
ACCEPTED MANUS a.o.comand
644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664
Records identified after EMBASE search: (n = 217)
Records identified after MEDLINE search: (n = 159)
Total records identified (n = 376)
Duplicates removed (n = 127)
Records screened for title and abstract (n =249)
Records excluded (n = 221)
Articles included for full-text review (n =29)
Articles included via references (n =1)
Full-text articles assessed for eligibility (n =30)
Records excluded Other topic (n = 3) Conference abstracts (n = 2) Case report (n = 1)
Articles included for review (n = 24)
ACCEPTE
666 667 668 669
670 671 672
| Gene | Location | Function | Reference |
|---|---|---|---|
| CD74 | 5q32 | MHC class II antigen processing | (57,59) |
| HSD3B2 | 1p12 | Mineralobiosynthesis of all classes of hormonal steroids | (57,58) |
| S100A6 | 1q21 | Cellular calcium signalling | (57,59) |
| CYP7B1 | 8q21.3 | Primary bile acids synthesis | (57,59) |
| GIPC2 | 1p31.1 | Unknown | (57,59) |
| GYPC | 2q14.3 | Regulating red cell stability | (57,58) |
| H19 | 11p15.5 | Non-protein coding tumor suppressor | |
| G0S2 | 1q32.2 | HIGHER Promoting apoptosis, preventing BCL2-BAX heterodimer formation | (58,59) |
| TM7SF2 | 11q13.1 | Lanosterol to cholesterol conversion | (58,59) |
| RAB34 | 17q11.2 | Protein transport. | (57,58) |
| MLH3 | 14q24.3 | DNA mismatch repair | (57,59) |
| SPINT2 | 19q13.2 | Serine protease inhibitor | (57,60) |
| SLC7A4 | 22q11.21 | Amino-acid transporter | (57,60) |
| RARRES2 | 7q36.1 | ACCEPTED MAL PHERA Regulation of adipogenesis, metabolism and inflammation | (57,76) |
673 674
Supplementary table 1 Systematic literature search details
| Database | Search strategy | Date of search |
|---|---|---|
| MEDLINE | ("adrenocortical carcinoma"[MeSH Terms] OR "adrenal cortex neoplasms"[MeSH Terms] OR ("adrenocortical"[Title/Abstract] AND ("cancer"[Title/Abstract] OR "tumor"[Title/Abstract] OR "carcinoma"[Title/Abstract] OR "neoplasm")) OR ("adrenal"[Title/Abstract] AND ("cancer"[Title/Abstract] OR "tumor"[Title/Abstract] OR "carcinoma"[Title/Abstract] OR "neoplasm")) OR ("adrenal cortex"[Title/Abstract]) AND ("cancer"[Title/Abstract] OR "tumor"[Title/Abstract] OR "carcinoma"[Title/Abstract] OR "neoplasm")) AND (((("methylation"[MeSH Terms] OR "methylation"[Title/Abstract]) OR ("histones"[MeSH Terms] OR "histones"[Title/Abstract] OR "histone"[Title/Abstract])) OR ("acetylation"[MeSH Terms] OR "acetylation"[Title/Abstract])) OR ("epigenomics"[MeSH Terms] OR "epigenomics"[Title/Abstract] OR "epigenetic"[Title/Abstract])) CRAT | 28-02-2017 |
| EMBASE | ACCEPTEDegen ('adrenal cortex carcinoma'/exp OR ('adrenal gland'/exp AND 'neoplasm'/exp) OR ('adrenocortical':ab,ti AND ('cancer':ab,ti OR 'tumor':ab,ti OR 'carcinoma':ab,ti OR 'neoplasm':ab,ti)) OR ('adrenal':ab,ti AND ('cancer':ab,ti OR 'tumor':ab,ti OR 'carcinoma':ab,ti OR 'neoplasm':ab,ti)) OR 'adrenal cortex':ab,ti AND ('cancer':ab,ti OR 'tumor':ab,ti OR 'carcinoma':ab,ti OR 'neoplasm':ab,ti)) AND (('methylation'/exp OR 'methylation':ti,ab) OR ('histone'/exp OR 'histone':ti,ab OR 'histones':ti,ab) OR ('acetylation'/exp OR 'acetylation':ti,ab) OR ('epigenetics'/exp OR 'epigenomics':ti,ab OR 'epigenetic':ti,ab)) | 28-02-2017 |
Legend Detailed description of the applied search strategy
677 678 679
Supplementary table 2 Overview of studies included for the systematic review
| Authors | Tissues | Methods used for analysis | Methylation related results |
|---|---|---|---|
| Gicquel et al. (78) | 6 ACC, 17 ACA - | - Southern Blott - Dot and Northern Blot# | - IGF2 promoter demethylation is associated with increased mRNA in ACC compared to ACA |
| Gicquel et al.(80) | 29 ACC, 35 ACA" - - 18 suspect tumors NOST.# | - Allele specific methylation (AVAII HPAII), Southern blott - Dot and Northern blot, RT-PCR# | - AGF2 mRNA overexpression and LOH is associated with 11p15 abnormalities in 93% of the ACC and 8.6% of the benign tumors Low H19 mRNA expression correlated with LOH or - pathological imprinting in ACC |
| Gao et al.(66) | - 10 ACC, 16 ACA, 16 NACT,* | - Bisulfite sequencing and PCR' - Northern Blot# | - H19 promotor hypermethylation is involved reduced H19 and increased IGF2 mRNA expression in ACC |
| Sidhu et al. (77) | - 20 ACC, 5 NAC' | - Bisulfite sequencing and nested PCR | - Altered TP53 promoter methylation is not involved in TP53 expression in ACC |
| Simi et al. (67) | 9 ACC, 5 NAC - - H295R and SW13 cell line | - Bisulfite sequencing and MS- & RT-qPCR" - Pyrosequencing | - Seladin-1 promoter hypermethylation is involved in |
| reduced seladin-1 mRNA expression in ACC. | |||
| - Decitabine induced seladin-1 upregulation in ACC in vitro | |||
| Fonseca et al. (73) | - 15 ACCs, 27 ACA, 6 NACT - 6 ACC, 10 ACA, 6 NAC* - H295R cell line | - Bisulfite treatment" - Illumina Infinium HM27+ - RT-PCR (selected genes)* | - The identification of 50 hypermethylated genes with low mRNA expression in ACC - Decitabine reverses hypermethylation and increases gene expression in ACC in vitro |
| Rechache et al.(57) | 8 ACC, 12 MACC, 48 ACA, 19 NAC+ - 5 ACC, 74 ACB* | - Bisulfite treatment' | - Distinct methylation patterns could differentiate ACC, |
| - Illumina Infinium HM450* | MACC, ACA and NAC | ||
| - Affymetrix HG U133 plus 2.0* | - ACC are overall hypomethylated compared to ACA and | ||
| NAC | |||
| - Identification of 52 hypermethylated genes with | |||
| downregulated mRNA expression in ACC |
ACCEPTED MANUSCRIPT
| Barreau et al. (58) | 51 ACC, 84 ACA' - - 34 ACC, 53 ACA* | - Bisulfite treatment' - Illumina Infinium HM27' - MS-MLPA' - Affymetrix HG U133 plus 2.0* | - CpG island methylation levels in promoter regions are higher in ACC compared to ACA - The identification of 3 distinct CpG Island Methylation |
|---|---|---|---|
| Phenotypes (CIMP) in ACC - Methylation levels are inversely correlated with overall survival in ACC | |||
| - It is feasible to use MS-MLPA to identify CIMPs in ACC | |||
| Korah et al. (68) | - 7 ACC, 8 ACA, 6 NACt,# - SW-13 cell line | - MS- & RT-qPCR+,# - Immunohistochemistry - Western blot | - RASSF1A promotor hypermethylation is involved reduced RASF1A mRNA and protein expression in ACC |
| - Increased RASSF1A mRNA expression is associated with reduced tumor growth in vitro | |||
| - RASSF1A might be involved in modulation of microtubule dynamics in the adrenal cortex in ACC | |||
| Guillaud-Bataille et al.(81) | - 33 ACCT - 29 ACC, 47 ACA* - H295R cell line | - ASMM RT-qPCR' - Affymetrix HG U133 plus 2.0+ - RT-qPCR+ - Western blot TED MED. | - H19 promotor hypermethylation is involved reduced H19 and increased IGF2 mRNA expression in ACC - Low IGF2 expression in ACC can be explained by altered methylation at the 11p15 locus - There are no phenotypical or transcriptomic differences between IGF2 high and IGF2 low ACC - IGF2 knock down impairs growth and promotes apoptosis in ACC in vitro |
| Hofland et al.(75) | - 12 ACC, 3 NAC+ *** - 25 ACC, 11 ACA, 20 ACH, 10 NAC# *** - 37 ACC *** | - Bisulfite treatment' - RT-qPCR++ | - INHAA promotor hypermethylation is involved reduced INHA mRNA but not with serum inhibin pro-aC expression in a subgroup of ACC |
| Assie et al. (27) | 51 ACC, 30 ACA' - - 47 ACC* | - Illumina Infinium HM27+ - Affymetrix HG U133 plus 2.0 S+ - Multiplex PCR | - The identification of 4 CIMP methylation clusters in ACC associated with distinct overall survival - The identification of 2 distinct transcriptomic clusters in ACC: C1a (poor overall survival) and C1b (better overall survival) |
| - Recurrent alterations in known driver genes (CTNNB1, |
ACCEPTED MANUSCRIPT
| TP53, CDKN2A, RB1 and MEN1) - Alterations in not previously reported genes (ZNRF3, DAXX, TERT and MED12) | |||
|---|---|---|---|
| Mitsui et al. (69) | - 3 ACC, 39 ACA'' | - Bisulfite sequencingt - MS-, US- and RT-qPCR ** | - WIF1 promotor hypermethylation is involved in reduced WIF1 mRNA expression in ACC - WIF1 mRNA expression is inversely correlated with intracellular ß-catenin accumulation or ß-catenin mRNA transcription in ACC |
| - WIF1 mRNA expression is inversely correlated with cyclinD1 expression in ACC | |||
| Pilon et al. (70) | - 8 ACC, 15 ACA | - Bisulfite sequencing - RT-qPCR" .* - Western blot - Immunohistochemistry ANU | - VDR promotor hypermethylation is involved in reduced VDR mRNA expression in ACC |
| Gara et al. (61) | 2 NAC, 75ACA, 10 ACC - - 17 NAC, 10 ACA, 10 ACC™ (validation cohort) 10 ACC, 26 ACA, 21 NAC* - - NIC-H295R and BD140A cell lines | - Illumina Infinium HM450' - Affymetrix HG U133 plus 2.0$ - Illumina HumanCytoSNP-12 v2.1 - Exiqon miRCURY LNA miRNA array - Caspase-Glo® 3/7 apoptosis assay - Western blot | - DNA methylation levels in ACC are more dysregulated compared to ACA and NAC - The majority of dysregulated genes in ACC are downregulated |
| Nielsen et al. (79) | - 20 ACC, 32 ACA, 10 PCCt,# | - Bisulfite treatment' - RT-PCR - Pyrosequencing - IlluminaOmni2.5M array¥ 0 | - H19 promotor hypermethylation is involved reduced H19 and increased IGF2 mRNA expression in ACC - Hypermethylation of the H19 ICR is caused by loss of the maternal allele of the 11p15 locus and correlates with IGF2 overexpression |
| Legendre et al. (60) | 17 ACCs, 1MACC, 6 NAC - - 14 ACCs, 6 NAC+ | - Bisulfite treatment" | - ACC are overall hypomethylated compared to ACA and |
| - Illumina Infinium HM450+ | NAC - Epigenetic alterations may affect mRNA expression of genes in WNT, TP53 and IGF pathways in ACC | ||
| - Affymetrix HG U133 plus 2.0* | |||
| Zheng et al. (59) | - 79 ACC, 120 NTS | - Illumina Infinium HM450' | - Three molecular distinct ACC subtypes are captured by a DNA-methylation signature |
ACCEPTED MANUSCRIPT
| - The identification of 3 different CIMP phenotypes with distinct overall survival - A methylation signature of 68 probes accurately predicts survival in an independent cohort | |||
| Creemers et al. (72) | 24 ACC, 14 ACA, 11 NAC - - 9 ACC, 11 ACA' | - Bisulfite treatment' - Pyrosequencing' - RT-PCR* | - H19 promotor hypermethylation reduces mRNA |
| expression of H19 and increases IGF2 mRNA expression - Pyrosequencing of IGF2 regulatory regions can differentiate ACC and ACA with a sensitivity of 89% and specificity of 92% | |||
| Drelon et al. (82) | - 7 ACC, 7 ACA - 47 ACC, 41 ACA and 4 NAC+ (Cochin cohort) - 3 ACC, 22 ACA and 10 NAC+ (Michigan cohort) - 79 ACC (TGCA Cohort) - H295R, SW13 cells+ | - RT-qPCR* - Western blot+ - | Histone methyltransferase EZH2 mRNA is overexpressed - in ACC as result of dysregulation of the P53/RB/E2F pathway - EZH2 mRNA overexpression in ACC is associated with increased proliferation and poorer prognosis - Inhibition of EZH2 with DZNep reduces proliferation and induces apoptosis in vitro - The effect of DZNep is increased with additional administration of Mitotane |
| Chittenden et al. (76) | - 26 ACC, 68 ACA, 21 NAC+ - 8 ACC, 42 benign NOS " (same data as Rechache et al 2012) - H295-R, SW13 cells | - Bisulfite treatment" - qPCR* - pyrosequencing' - | - RARRES2 promotor hypermethylation reduces mRNA expression of RARRES2 in ACC - RARRES2 has a tumor suppressor function independent from CMKLR1- expressing immune cell recruitment - RARRESS2 mRNA overexpression supress tumor growth in vivo |
| Jouinot et al. (64) | - 203 ACCT | - Bisulfite treatment" -Infinium HumanMethylation27 Beadchip (Illumina)" MS-MLPA' - | - MS-MLPA is a methylation assay that can be used in prognostification for ACC survival and recurrence. |
| Creemers et al. (71) | - 29 ACC, 16 ACA, 11 NAC' | - PCR' - Immunohistochemistry | - MGMT promotor hypermethylation reduces mRNA expression of MGMT in ACC |
| - Temozolomide inhibits ACC growth in vitro |
Cheng et al. (74)
38 ACC, 14 ACA H-295R, SW13 cells*
- RT-qPCR*
- MS-RT-qPCR’
- immunofluorescense*
- DKK3 promotor hypermethylation and copy number alterations reduce mRNA expression of DKK3 in ACC
- DKK3 might have a FOXO1-mediated differentiation- promoting role in ACC
Legend Overview of all included studies for the systematic review, ranked according to publication date. Per study, analyzed tissues, methods and outcome variables are shown. The methods applied to assess methylation and gene expression were marked for each study (see below). Normally fresh frozen tissue, except when marked with two asterisks.
* = Normal tissue samples were randomly selected from 12 TCGA projects. t = methylation analysis, # = gene expression analysis, ¥ = mutational analysis, ** formalin-fixed paraffin embedded tissues, *** combination of formalin-fixed paraffin embedded and frozen tissues, ” performed with cells (SW 13 or H295-R) only
Abbreviations: ASMM = allele-specific methylated multiplex real time quantitative PCR, ACC = adrenocortical carcinoma, MACC = metastasis of adrenocortical carcinoma, ACA = adrenocortical adenoma, ACH = adrenocortical hyperplasia, NAC = normal adrenocortical tissue, NTS = normal tissue samples, PCC = pheochromocytoma, MS-MLPA = methylation-specific multiplex ligation-dependent probe amplifications, NOS = not otherwise specified
| Study | Performed statistics | Definition hypermethylated CpG site | Definition downregulated gene |
|---|---|---|---|
| Gao et al.(66) | Kruskal-Wallis Mann-Whitney U test Univariate regression analysis | A target locus methylation level (p<0.05) | A target gene mRNA expression / 1 housekeeping gene mRNA (p<0.05) |
| Simi et al.(67) | Kolmogorov-Smirnov | A target locus methylation level (p<0.05) | A target gene mRNA expression / 1 housekeeping gene mRNA expression (p<0.05) |
| Korah et al.(68) | Shapiro-Wilk test Levene's test T-test ANOVA | A target locus methylation level (p<0.05) | A target gene mRNA expression / 2 housekeeping genes mRNA expression (p<0.05) |
| Mitsui et al.(69) | Mann-Whitney U test Chi-square test Spearman's rank correlation | A target locus methylation level (p<0.05) | A target gene mRNA expression / 2 housekeeping genes mRNA expression (p<0.05) |
| Hofland et al.(75) | Chi-square test One-way analyses of variance Tukey's multiple comparison T-test Spearman's rank correlation | A target locus methylation level (p<0.05) | A target gene mRNA expression / 2 housekeeping genes mRNA expression (p<0.05) |
| Pilon et al.(70) | Mann-Whitney U test Spearman's rank correlation | A target locus methylation level (p<0.05) | A target gene mRNA expression / 1 housekeeping gene mRNA expression (p<0.05) |
| Creemers et al.(72) | Mann-Whitney U test Spearman's rank correlation | A target locus methylation level (p<0.05) | A target gene mRNA expression / 1 housekeeping gene mRNA expression (p<0.05) |
| Rechache et al.(57) | ANOVA Benjamini-Hochberg method | 4ß ≤-0.20 or 4ß ≥ 0.20 of the M-value per locus (FDR adjusted p ≤ 0.01) | 4 ≥ 2 fold change in RMA normalized mRNA expression (FDR adjusted p ≤ 0.05) |
| Fonseca et al.(73) | Wilcoxon Rank test Unpaired Student's t-test Fisher exact test | A ß-value per locus (p<0.05) | A target gene mRNA expression / 3 housekeeping genes RNA expression (p<0.05) |
| Barreau et al. (58) | ANOVA P Limma moderated t-test Benjami-Hochberg method Pearson correlation test | & M-value per locus (FDR adjusted p < 0.05) | A RMA normalized mRNA expression data (FDR adjusted p < 0.05) |
Legendre et al.(60)
Wilcox Rank
Δ Δβ - 0.20 or Δβ > 0.20 of the M-value per
A RMA normalized mRNA expression data (FDR
| T-test Benjamini-Hochberg method Discretization method | locus (FDR adjusted p ≤ 0.05) | adjusted p < 0.05) | |
|---|---|---|---|
| Zheng et al.(59) | Descriptive statistics T-test | ≥ 25% of promotor CPG sites has ß <= 0.1 in ≥ 90% of normal samples and B >= 0.3 in ≥5% ACC samples. | 4 RMA normalized mRNA expression data ACC (B >= 0.3) vs normal tissue (ß < 0.1) (FDR adjusted p < 0.01) |
| Liu-Chittenden et al.(76) | T-test | A target locus methylation level (p<0.05) | A target gene mRNA expression / 1 housekeeping gene mRNA expression (p<0.05) |
| Cheng(74) | D' Agostino/Pearson omnibus test Two-tailed t-test Mann-Whitney U test Kruskal-Wallis test | A target locus methylation level (p<0.05) | A target gene mRNA expression / 1 housekeeping gene mRNA expression (p<0.05) |
ACCEPTED MATEN NU SOLOS)
| Study | Identified hypermethylated and downregulated genes |
|---|---|
| Gao et al.(66) | H19 |
| Simi et al.(67) | DHCR24 (SELADIN1) |
| Korah et al.(68) | RASSF1A |
| Mitsui et al.(69) | WIF1 |
| Hofland et al.(75) | INHA |
| Pilon et al.(70) | VDR |
| Creemers et al.(72) | H19 |
| Rechache et al.(57) | ABCA1, CD74, COL4A3, GOS2, GATA6, HSD3B2, KCNQ1, MAP3K5, NCOA7, RAPGEF4, RARRES2, S100A6, SPTBN1, TNFSF13, TNS1, ADCK3, ALDH3B1, CSDC2, CYP7B1, GIPC2, HOOK1, MEIS1, MLH3, MRPL33, NME5, RGNEF, TCIRG1, AMPD3, B4GALT6, CAB39L, CD55, GYPC, NDRG4, RAB34, RBPMS, SEMA6A, TNFS1F2-TNFSF13, SLC16A9, PHF11 |
| Fonseca et al.(73) | CDKN2A, GATA4, DLEC1, HDAC10, PYCARD, SCGB3A1/HIN1 |
| Barreau et al. | H19, GSTM1, GSTP1, GOS2, GSTT1, RAB34, GYPC, GIPC2, PLAGL1, LY6D, PCOLCE, NDN, AMT, LGALS3BP, APOC1, TM7SF2, PPAPDC3, PTPN7, SCNN1A, HSD3B2, ACAA2, CTSZ, PYGM, KRT8, NDRG2 |
| Legendre et al.(60) ** | ERBB3, TFAP2A, DENND1B, ESRP2, C19orf33, PTHLH, SDCBP2, S100A16, FMO5, RHOF, CAMK2B, ATP8A1, SH2B1, DYRK2, SETD7, SLC2A13, ANXA11, NAV2, ITGB4, NINJ2, TK1, RBM5, SPSB1, C1orf86, NR4A2, ASPHD2, ZNF660, SOS2, ACAT1, GLT25D1, FBP1, ST7, YIF1B, PARP1, VEPH1, ZMYND8, ONECUT1, PCK2, DUSP7, SH3BP2, ADAM17, C4orf41, PDK2, SPINT2, TBC1D1, SMAD6, ITGB2, CLDN11, FGD2, ACO2, TRABD, NT5C, OSTC, ADRBK1, PLEC, HDAC4, ZNF853, TPD52, C8orf44 /// SGK3, AGA, TRAK1, FERMT1, RRP1B, EIF4E2, LOC148413, FABP3, COQ10A, HLA-B, NRP1, TFEB, STIL, CHPF2, CCDC8, SKAP1, USP36, C11orf51, TPK1, SLC37A4, CDC7, MSL2, MAP7, ANKFY1, ST6GALNAC4, ACAD10, RPP21 /// TRIM39 /// TRIM39R, WDR86, CHML, DCTN4, RGL2, ATG7, NDRG1, SMARCD2, PRDM5, MCM10, NKTR, IVNS1ABP, GART, ASPHD1, TTBK1, ZBTB7B, WNT3, SLC16A3, PAX3, GPATCH3, TAPT1, VAX2, RTEL1, LCORL, KIAA1244, CACNA1D, MRPL20, DGCR2, MAPK4, ING3, SPG7, JPH4, CCNH, CTNS, GLT8D2, MAN1C1, NACC1, RGS1, MTF1, SEC22A, ZNF382, ANO7, SLC7A4, ITGB1, PLXNB2, FER, GULP1, ABHD1, EIF2C4, EPC1, MKI67IP, SDK2, EGLN2, LOC641518, ITSN2, OGFOD1, DENR, SUZ12, F2RL2, RPL13, CCHCR1, KIAA1826, TMEM209, AIDA, PRDM15, MAPT |
| Zheng et al.(59) | ACADL, ACTR3C, ADAM15, AGBL2, APOBEC3G, ARSI, AS3MT, ASS1, B3GNT5, C10orf25, C15orf56, C1orf106, C1orf172, C1QTNF1, C21orf88, C2CD4A, C9orf106, C9orf170, C9orf66, CA8, CALCB, CALHM2, CD14, CD74, CHST6, CLCF1, CLDN3, CLVS2, COPZ2, CRIP3, CSF1, CSRP1, CYP27A1, CYP7B1, DDIT4L, DNAJA4, DSC3, ECHDC3, EMB, EMP3, EPN3, ESR2, FADS6, FAM129B, FBXO17, FBXO27, |
| G0S2, GBGT1, GBP1, GIPC2, GLIPR1L1, GPR37, GPX3, GPX7, RAMD1A, GRAMD2, GSDMD, HAAO, HCP5, HIST1H2AM, HIST1H2BO, HIST1H4J, HOXB4, IKBKE, IRAK3, IRF6, ITPKB, JMJD8, KCNJ3, KCNV1, KCTD14, KY, LAD1, LHFPL3, LMOD1, LOC100271722, LOXL1, LRFN2, LRRC34, LRRC61, LRRC8E, MAMDC2, MAPK13, METTL7B, MFSD2A, MLH3, MYD88, NQO1, NXPH2, P2RY2, PCDH10, PLIN5, PLTP, POM121L2, PPP1R9B, PTER, PTGES, PTRF, RAB37, RAB3D, RAP2B, RBP4, RDM1, RGL3, RHOD, RNF208, RNF39, RPS6KL1, S100A11, S100A6, SASH1, SERP2, SERPINE1, SLC13A5, SLC2A10, SLC30A2, SLC40A1, SLC6A15, SLC7A4, SLCO4C1, SMOC1, SPINT2, ST8SIA6, STL, STXBP2, SUSD3, SYT13, TERC, TM7SF2, TMEM106A, TRIM38, TRIP6, TRPA1, TTC22, TUBA1C, TXN, UBXN10, XKR8, ZDHHC1, ZDHHC8P1, ZNF22, ZNF439, ZNF665, ZNF835, ZNF860, ZSCAN16 | |
| Liu-Chittenden et al.(76) | RARRES2 |
| Cheng(74) | DKK3 |
Legend Overview of genes identified in studies reporting on methylation in ACC. For each study the description of the outcome measure between gene methylation and effect on expression is reported. Genes reported in 2 or more studies have bold characters.
* Gene defined as epigenetically silenced if ≥25% promoter CpG sites meet all 3 criteria: 1) ≥ 90% normal samples unmethylated (6 ⇐ 0.1), 2) ≥ 5% ACC methylated ( 6 >= 0.3), 3) t-test comparing expression levels in methylated (6 >= 0.3) and unmethylated tumor samples (6 < 0.1) significant FDR < 0.01.
** Overview of significantly hypermethylated and downregulated in > 7 cases. Extracted from supplementary table 2 of Legendre et al.