Cancer Medicine. 2020;9:8373–8385. wileyonlinelibrary.com/journal/cam4
|
8373O R I G I N A L R E S E A R C H
Methylome analyses of three glioblastoma cohorts reveal
chemotherapy sensitivity markers within DDR genes
Tobias Kessler
1,2|
Anne Berberich
1,2|
Ahmed Sadik
3|
Felix Sahm
4,5|
Thierry Gorlia
6|
Christoph Meisner
7|
Dirk C. Hoffmann
1,2,8|
Antje Wick
2|
Philipp Kickingereder
9|
Petra Rübmann
1|
Martin Bendszus
9|
Christiane Opitz
3|
Michael Weller
10|
Martin van den Bent
11|
Roger Stupp
12|
Frank Winkler
1,2|
Alba Brandes
13|
Andreas von Deimling
4,5|
Michael Platten
14,15|
Wolfgang Wick
1,2This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
© 2020 The Authors. Cancer Medicine published by John Wiley & Sons Ltd.
1Clinical Cooperation Unit Neurooncology,
German Cancer Consortium (DKTK), German Cancer Research Center (DKFZ), Heidelberg, Germany
2Department of Neurology and
Neurooncology Program of the National Center for Tumor Diseases, Heidelberg University Hospital, Heidelberg, Germany
3Brain Tumor Metabolism, DKTK, DKFZ,
Heidelberg, Germany
4Department of Neuropathology,
Heidelberg University Hospital, Heidelberg, Germany
5Clinical Cooperation Unit Neuropathology,
DKTK, DKFZ, Heidelberg, Germany
6European Organization for Research
and Treatment of Cancer Headquarters, Brussels, Belgium
7Institute for Clinical Epidemiology and
Biometry, Tübingen, Germany
8Faculty of Biosciences, Heidelberg
University, Heidelberg, Germany
9Department of Neuroradiology, Heidelberg
University Hospital, Heidelberg, Germany
10Department of Neurology, University
Hospital and University of Zurich, Zurich, Switzerland
11The Brain Tumor Center, Erasmus
MC Cancer Institute, Rotterdam, The Netherlands
12Feinberg School of Medicine,
Northwestern University, Chicago, IL, USA
Abstract
Background: Gliomas evade current therapies through primary and acquired resist-ance and the effect of temozolomide is mainly restricted to
methylguanin-O6-methyl-transferase promoter (MGMT) promoter hypermethylated tumors. Further resistance
markers are largely unknown and would help for better stratification.
Methods: Clinical data and methylation profiles from the NOA-08 (104, elderly glioblastoma) and the EORTC 26101 (297, glioblastoma) studies and 398 patients with glioblastoma from the Heidelberg Neuro-Oncology center have been analyzed focused on the predictive effect of DNA damage response (DDR) gene methylation. Candidate genes were validated in vitro.
Results: Twenty-eight glioblastoma 5'-cytosine-phosphat-guanine-3' (CpGs) from 17 DDR genes negatively correlated with expression and were used together with telomerase reverse transcriptase (TERT) promoter mutations in further analysis. CpG methylation of DDR genes shows highest association with the mesenchymal (MES) and receptor tyrosine kinase (RTK) II glioblastoma subgroup. MES tumors have lower tumor purity compared to RTK I and II subgroup tumors. CpG hypo-methylation of DDR genes TP73 and PRPF19 correlated with worse patient survival in particular in MGMT promoter unmethylated tumors. TERT promoter mutation is most frequent in RTK I and II subtypes and associated with worse survival. Primary glioma cells show methylation patterns that resemble RTK I and II glioblastoma and long term established glioma cell lines do not match with glioblastoma subtypes. Silencing of selected resistance genes PRPF19 and TERT increase sensitivity to te-mozolomide in vitro.
1
|
INTRODUCTION
Diffuse gliomas regularly evade current therapies and sev-eral mechanisms behind resistance and sensitivity have not only recently been discovered with high-throughput efforts
demonstrating the molecular faith of gliomas at recurrence,1
but also novel targeted approaches proposing a novel radiation
sensitivity biomarker.2 Classically, methylation of the
methyl-guanin-O6-methyltransferase promoter (MGMT) promoter is
known to predict response to alkylating chemotherapy, with at best minimal responses for patients with an unmethylated
MGMT promoter at diagnosis and progression.3,4 The NOA-08
study demonstrated an impressive survival benefit in MGMT promoter methylated elderly patients with temozolomide
treat-ment compared to radiotherapy alone,5 but this may potentially
be restricted to a specific molecular tissue context.6,7
Nevertheless, disease control for more than a few years is achieved in barely 15%-20% of younger glioblastoma pa-tients with tumors with a hypermethylated MGMT promoter only. Therefore, further markers to better stratify patients for treatment response prediction and to decide for chemo- or ra-diotherapy are of great interest. Recent approaches suggested that the DNA damage response (DDR) gene methylome could facilitate to predict resistance vs sensitivity to radio- and che-motherapy in World Health Organization grade II, IDH
mu-tant gliomas.8 These markers have not been validated in an
independent data set yet and an effect of differential DDR
gene methylation in IDH wild-type glioblastomas dependent on the clinically relevant MGMT promoter methylation is unknown. Also, various inhibitors of DDR components are in preclinical and clinical development allowing to further
exploit the concept of synthetic lethality.9,10
Recent studies suggested a benefit from a methylated
MGMT promoter when receiving temozolomide
chemother-apy only in patients with tumors with additional promoter mutation of the DDR gene telomerase reverse transcriptase
(TERT).11,12 We identified a subset of glioblastoma patients
by methylation clustering, who benefitted most from temo-zolomide treatment when the MGMT promoter is methylated.
These patients showed enhanced TERT expression.6 In
ad-dition, TERT promoter mutation was found to be one of the earlier but not earliest events of gliomagenesis and leads to
increased TERT expression rather than TERT methylation.13
Therefore, methylation of DDR genes and mutation status in particular for TERT may alter the sensitivity to standard radiochemotherapy treatment especially in tumors lacking
MGMT promoter methylation. In these tumors prognostic
factors are rarely known and would be important for patient stratification. We here explored the potential of DDR methyl-ation as further prognostic markers. We based or hypothesis
on 450 genes that have been identified as DDR genes14 and
restricted the analysis to genes where methylation negatively impacted expression. Of the DDR genes, TERT is unique in the sense that TERT promoter mutations mainly influence
13Department of Medical Oncology,
Azienda USL-IRCCS Institute of Neurological Sciences, Bologna, Italy
14Clinical Cooperation Unit
Neuroimmunology and Brain Tumor Immunology, DKTK, DKFZ, Heidelberg, Germany
15Department of Neurology, Medical
Faculty Mannheim, MCTN, Heidelberg University, Mannheim, Germany
Correspondence
Wolfgang Wick, Neurology Clinic & Neuroonology Program at the National Center for Tumor Diseases, Im Neuenheimer Feld 400, D-69120 Heidelberg, Germany.
Email: wolfgang.wick@med.uni-heidelberg.de
Funding information
The molecular analyses have been supported by the German Cancer Research Center, the NCT Heidelberg and by a grant from the German Research Foundation (DFG, SFB 1389, TP A03 to TK and WW). The clinical conduct of the study has been supported by F Hoffmann—La Roche Ltd.
Conclusion: Hypomethylation of DDR genes and TERT promoter mutations is asso-ciated with worse tumor prognosis, dependent on the methylation cluster and MGMT promoter methylation status in IDH wild-type glioblastoma.
K E Y W O R D S
DNA damage response (DDR), methylation profiling, methylguanin-O6-methyltransferase promoter (MGMT), telomerase reverse transcriptase (TERT)
expression, and therefore, for TERT not methylation but the promoter mutations status was included. Based on these as-sumptions, we aimed to identify prognostic markers of DDR genes in a combined analysis of three large well-documented
glioblastoma cohorts from the NOA-087 and EORC 2610115
studies as well as a patient cohort from our Heidelberg Neuro-Oncology Center spanning a variety of conditions. Promising candidates are validated in vitro.
2
|
METHODS
2.1
|
Patient cohorts
Two glioma studies and a well-documented cohort of patients treated in Heidelberg with study grade follow up that was used as an exploratory, hypothesis-generating data set were included in the present work. Altogether this study involves 799 patients with diffuse IDH wild-type glioma. Regression analyses were performed either for each study separately or combined for all three studies together with correction for the confounding effect of the study, as indicated.
2.1.1
|
NOA-08
The NOA-08 study compared radiotherapy (RT) with temozo-lomide (TMZ) chemotherapy in elderly patients (age at
diag-nosis >65 years) with anaplastic astrocytoma or glioblastoma.5
The original study population consisted of 373 patients. The investigated biomarker cohort consists of 104 patients (radio-therapy group: 53, temozolomide group: 51).
2.1.2
|
EORTC 26101
The EORTC 26101 study randomized patients with progres-sive glioblastoma between bevacizumab (BEV) with lomustine
(CCNU) and CCNU alone.15 The original study cohort
con-sisted of 596 patients. The investigated biomarker cohort where methylation array data and paraffin tissue for evaluating TERT mutations status was available consisted of 297 patients.
2.1.3
|
Heidelberg cohort
About 398 patients from the Heidelberg Neuro-Oncology center diagnosed with IDH wild-type glioblastoma between 07/2014 and 01/2018 based on histopathological and molecu-lar characteristics. This cohort includes 43 patients from the
NCT Neuro Master Match (N2M2) pilot study extensively
characterized with whole-genome sequencing, whole-exome
sequencing (WES), RNAseq, and methylation analysis.16
Inclusion of patients into this analysis is covered by a local Heidelberg ethics vote (no. S307/2019).
Patients of the two clinical study biomarker cohorts (NOA-08 and EORTC-26101) are comparable to the original study population regarding survival times and treatment (Table 1).
Biomarker cohort Full study cohort
NOA-08
Patient number [n (%)] 104 (28%) 373 (100%)
Overall survival [median (95% CI)]a 11.2 (9.6-13.8) 8.7 (8.0-9.8)
Event-free survival [median (95% CI)]a 4.1 (3.7-4.5) 4.4 (3.6-5.5)
TMZ group [n (%)] 51 (49%) 195 (52%)
RT group [n (%)] 53 (51%) 178 (48%)
EORTC-26101
Patient number [n (%)] 297 (50%) 596 (100%)
Lomustine first group [n (%)] 117 (39%) 231 (39%)
BEV ± Lomustine first group [n (%)] 180 (61%) 365 (61%)
Overall survival [median (95% CI)]a 8.6 (7.9-9.9) 8.9 (8.2-9.6)
Progression-free survival [median (95% CI)]a 3.0 (2.8-3.7) 2.9 (2.8-3.0)
Heidelberg cohort
Patient number [n] 398 NA
Overall survival [median (95% CI)]a 24.9 (19.2-31.0) NA
Progression-free survival [median (95% CI)]a 8.2 (7.2-9.2) NA
Patients with RT + TMZ [n (%)] 252 (63%) NA
Abbreviations: BEV, bevacizumab; RT, radiotherapy; TMZ, temozolomide. aSurvival times are given in months.
TABLE 1 Comparison of Biomarker cohorts in the present analysis with the original study cohorts.
2.2
|
Illumina HumanMethylation450 and
HumanMethylationEPIC arrays
The Illumina Infinium HumanMethylation450 (450k) bead chip and MethylationEPIC kits were used to obtain the DNA methylation status at >450 000 and >850 000 5'-cytosine-phosphat-guanine-3' (CpG) sites, respectively (Illumina), according to the manufacturer's instructions at the Genomics and Proteomics Core Facility of the German Cancer Research Center (DKFZ) in Heidelberg, Germany from fresh frozen and paraffin embedded tissue of study cohorts as well as selected primary patient-derived cul-tures. Samples were analyzed using the R (www.r-proje
ct.org) based methylation pipeline “ChAMP” 2.10.1.17 In
brief, filtering was done for multihit sites, SNPs, and XY chromosome-related CpGs, then, data were normalized with a BMIQ based method and analyzed for batch effects with a singular value decomposition algorithm. Batch ef-fects related to the tissue used (paraffin embedded [FFPE] vs fresh frozen [KRYO]) were corrected using ComBat.
MGMT promoter methylation status was determined by the
algorithm of Bady et al.18 The classifier score and
associa-tion with specific classifier types (RTK I, II, and MES) was performed using the Neuropathology 2.0 tool described in
Capper el al.19 Custom scripts based on the R packages
“minfi” (version 1.26.2) and “conumee” (version 1.14.0) were implemented for copy-number variation profiling and visualization.
2.3
|
Tumor purity estimation
Tumor purity estimation was performed on normalized beta values of methylation data with the R package InfiniumPurity
version 1.3.1.20
2.4
|
Selection of functional DDR gene
methylation CpGs
Potential DDR genes were obtained from a 450 putative gene
list that was published previously.14 CpGs present in the 450k
methylation array in the promoter region were tested for negative association with expression in The Cancer Genome Atlas (TCGA, RRID:SCR_003193) data set obtained from the National Cancer Institute Genomic Commons Data Portal (GDC Portal, portal.gdc.cancer.gov). CpGs with a correlation
r < −.3 and a Benjamini-Hochberg adjusted P < .05 were
re-garded as “functional” in a sense that high methylation cor-relates with low RNA expression. About 28 CpGs from 17 genes met this criterion and were used for further analysis. The thereafter used term “functional DDR CpGs” refers to these 28 CpGs.
2.5
|
Cell culture and in vitro assays
A detailed description of in vitro assays is given in the Methods S1. Generation and maintenance of established glioma cell lines and primary glioma cell cultures was
per-formed using standard methods as described previously.21,22
2.6
|
Data analysis
Statistical analyses were carried out with R 3.5.1 (www.r-proje ct.org, RRID:SCR_001905) and Microsoft Excel 2016 (Microsoft Corporation; RRID:SCR_016137). If not oth-erwise stated, a P < .05 was considered as significant and marked with a “*.” No outliers have been excluded. Correction for testing of multiple CpGs or genes was performed using the Benjamini-Hochberg procedure. The performances of the multivariate proportional hazards models were calculated using a c-index concordance statistic in R. If not otherwise indicated bar charts show mean values and standard devia-tion of at least three independent experiments. For survival analysis, we used custom adaptations of the R packages “sur-vival” (version 2.42-6) and “survminer” (version 0.4.3). A Cox proportional hazards model for univariate and multivari-ate analysis was used to assess the correlation between CpG methylation and survival as implemented in the “coxph” function. Clustering of methylation array samples was done
as described before6 using the ConsensusClusterPlus
pack-age in R. Dimensionality reduction and network analysis were also performed with R and are described in the Methods S1. All figures were produced using R-based packages.
2.7
|
Data availability
Methylation raw and processed data are made accessible via the Gene Expression Omnibus (GEO) database (https:// www.ncbi.nlm.nih.gov/geo/; RRID:SCR_005012) under the GEO accession numbers GSE12 2920 (NOA-08 biomarker cohort), GSE14 3755 (EORTC 26101 biomarker cohort), GSE12 2994 and GSE143842 (both Heidelberg cohort). TCGA data from glioblastoma patients can be accessed via CDC portal (portal.gdc.cancer.gov).
3
|
RESULTS
3.1
|
The functional DDR methylome of IDH
wild-type glioblastoma
Figure S1 shows the workflow of the project and study co-horts used. DDR CpGs were derived from a list of 450
correlated RNA expression of these genes with CpG meth-ylation in the TCGA glioblastoma data set and negatively correlated CpGs (r < −.3, P.adj < .05, see Section 2) were re-garded as functional and used for further analysis. In total, 28 CpGs from 17 genes were identified (Table S1). Additionally,
TERT expression correlates with TERT promoter mutations
in samples from Heidelberg, for which expression and muta-tion data were available (Figure S2). Mutamuta-tion frequencies of DDR genes excluding TERT were rather low in the TCGA glioblastoma data set with only three genes exceeding a pre-defined 3% mutation frequency in the TCGA glioblastoma data set (STAG [5%], TP53 [53%], ATRX [9%]) all without harboring functional CpGs in their promoter. Similar results were obtained in the WES Heidelberg glioblastoma IDH wild-type cohort (n = 43). Figure 1A,B show a heatmap with
methylation values of functional DDR CpGs together with methylation classifier assignments of all 799 IDH wild-type tumors from the EORTC26101, NOA-08, and Heidelberg cohorts. Hierarchical clustering divides the CpGs into four different groups, those with generally low methylation, high methylation, and two groups with a broad range of methyla-tion values (Figure S3).
Principle component analysis of DDR methylation re-veals two main directions. Dimension 1 (24.3% variance) is dominated by XRCC3, CUL4A, and CSK1E and dimension 2 (11.1% variance) by POLE4 and TP73 (Figure 2A). Tumor purity was estimated from methylation data and is strongly associated with dimension 1, highlighting the importance of tumor purity on methylation profiles (Figure 2B). Tumors with mesenchymal (MES) classifier assignment differed
FIGURE 1 Functional DNA damage response 5'-cytosine-phosphat-guanine-3' (CpG) methylation in clinical study cohorts. Heatmaps
of functional CpGs of all three studies (NOA-08, n = 104, EORTC 26101 , n = 297 and Heidelberg, n = 398) combined showing normalized methylation beta values (A) and row scaled values (B). 5ʹ-UTR, 5ʹ untranslated region; chr, chromosome; TSS200, 0-200 base pairs upstream of transcription start site; TSS1500, 200-1500 base pairs upstream of transcription start site; a full list of classifier abbreviations can be found in the Supporting Information TP73 cg13943358 TP73 cg03050669 TP73 cg23163013 PARP3 cg12554573 RPA3 cg19739774 SMC5 cg14289863 PARP4 cg20765408 XRCC3 cg25999604 XRCC3 cg23193616 XRCC3 cg18597188 CUL4A cg26825849 CCND3 cg17945153 PRPF19 cg00778920 EXO1 cg16121177 GEN1 cg11330941 POLR2E cg26528128 EXO1 cg04706276 CHEK1 cg00554702 MVP cg26990835 CSNK1E cg13110239 CSNK1E cg04728789 MGMT cg14194875 MGMT cg12434587 POLE4 cg20919922 POLE4 cg02307033 0 5 10 15 20 25 TSS1500 TSS200 5'UTR 0.2 0.4 0.6 0.8 Heidelberg EORTC 26101 NOA−08 methylated unmethylated APA REACT GBM MES GBM Midline GBM MYCN GBM RTK I GBM RTK II H3 K27M LGG DNT CNS NB FOXR2 INFLAM LGG GG no match chr feature beta value cohort MGMT classifier chr CHEK1 cg00554702 PRPF19 cg00778920 EXO1 cg04706276 EXO1 cg16121177 CSNK1E cg04728789 CSNK1E cg13110239 XRCC3 cg23193616 XRCC3 cg25999604 XRCC3 cg18597188 CUL4A cg26825849 GEN1 cg11330941 POLR2E cg26528128 RPA3 cg19739774 SMC5 cg14289863 MGMT cg12434587 MGMT cg14194875 MVP cg26990835 POLE4 cg02307033 POLE4 cg20919922 PARP3 cg12554573 PARP4 cg20765408 TP73 cg23163013 TP73 cg03050669 CCND3 cg17945153 TP73 cg13943358 0 5 10 15 20 25 TSS1500 TSS200 5'UTR −4 −2 0 2 4 Heidelberg EORTC 26101 NOA−08 methylated unmethylated APA REACT GBM MES GBM Midline GBM MYCN GBM RTK I GBM RTK II H3 K27M LGG DNT CNS NB FOXR2 INFLAM LGG GG no match A B cohort MGMT feature classifier z-score
A B C D E F APA CNS NB FOXR2 GBM MES GBM Midline GBM MYCN GBM RTK I GBM RTK II H3K27M INFLAM LGG DNT LGG GG no match REACT classifier tumor purity [%] 25 50 75 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● 0.2 0.4 0.6 0.8
mean XRCC3 methylation [beta]
tumor purity 0.2 0.4 0.6 25 50 75 G PARP3 cg12554573
PARP4 cg20765408 EXO1 cg04706276EXO1 cg16121177
RPA3 cg19739774 POLE4 cg02307033 POLE4 cg20919922 CUL4A cg26825849 POLR2E cg26528128 MGMT cg12434587 MGMT cg14194875 SMC5 cg14289863 GEN1 cg11330941 XRCC3 cg18597188 XRCC3 cg23193616 XRCC3 cg25999604 CHEK1 cg00554702 CCND3 cg17945153 TP73 cg03050669 TP73 cg23163013 TP73 cg13943358 MVP cg26990835 PRPF19 cg00778920 CSNK1E cg04728789 CSNK1E cg13110239 −1.0 −0.5 0.0 0.5 1.0 −1.0 −0.5 0.0 0.5 1.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● −3 0 3 6 −5 0 5 SMC5 cg14289863 XRCC3 cg23193616 XRCC3 cg25999604 CUL4A cg26825849 XRCC3 cg18597188 EXO1 cg16121177 CSNK1E cg04728789 CSNK1E cg13110239 CHEK1 cg00554702 EXO1 cg04706276 PRPF19 cg00778920 TP73 cg23163013 TP73 cg13943358 RPA3 cg19739774 PARP3 cg12554573 TP73 cg03050669 GEN1 cg11330941 CCND3 cg17945153 POLE4 cg02307033 POLE4 cg20919922 MVP cg26990835 POLR2E cg26528128 MGMT cg12434587 PARP4 cg20765408 MGMT cg14194875 GBM RT K I GBM RT K II GBM ME S GBM RT K I GBM RT K II GBM ME S GBM RT K I GBM RT K II GB MM ES SMC5 cg14289863 XRCC3 cg23193616 XRCC3 cg25999604 XRCC3 cg18597188 CUL4A cg26825849 PRPF19 cg00778920 EXO1 cg04706276 CHEK1 cg00554702 EXO1 cg16121177 CSNK1E cg04728789 CSNK1E cg13110239 TP73 cg23163013 TP73 cg03050669 TP73 cg13943358 GEN1 cg11330941 RPA3 cg19739774 CCND3 cg17945153 PARP3 cg12554573 PARP4 cg20765408 POLR2E cg26528128 POLE4 cg02307033 MVP cg26990835 POLE4 cg20919922 MGMT cg12434587 MGMT cg14194875 RPA3 cg19739774 TP73 cg13943358 PARP3 cg12554573 EXO1 cg04706276 EXO1 cg16121177 TP73 cg03050669 TP73 cg23163013 SMC5 cg14289863 XRCC3 cg23193616 GEN1 cg11330941 CSNK1E cg04728789 CSNK1E cg13110239 XRCC3 cg25999604 XRCC3 cg18597188 CUL4A cg26825849 PRPF19 cg00778920 PARP4 cg20765408 POLR2E cg26528128 CCND3 cg17945153 POLE4 cg02307033 CHEK1 cg00554702 MGMT cg12434587 MVP cg26990835 POLE4 cg20919922 MGMT cg14194875 Heidelberg NOA-08 −3 0 3 6 −5 0 5 −50 0 50 −80 −40 0 40 Individuals − PCA EORTC-26101 chromosome 0 1020 feature TSS1500 TSS200 5'UTR z-score 0 2 -2 GpCs - PCA PCA1 (24.3%) PCA2 (11.1% ) 2.5 5.07.5 cont. PCA1 (24.3%) PCA2 (11.1% ) samples - PCA
av. meth. beta 0.2 0.4 0.6 0.8 PCA1 (24.3%) PCA2 (11.1% ) samples - PCA APA CNS NB FOXR2 GBM MES GBM Midline GBM MYCN GBM RTK I GBM RTK II H3K27M INFLAM LGG DNT LGG GG no match REACT APA CNS NB FOXR2 GBM MES GBM Midline GBM MYCN GBM RTK I GBM RTK II H3K27M INFLAM LGG DNT LGG GG no match REACT PCA1 (15.7%) PCA2 (13.0% ) tumor purity [%]
from the RTK I and RTK II tumors (Figure 2C,D). We an-alyzed the association of the methylation of single DDR CpGs with classifier assignments in the study cohorts, with a cutoff P < .001 (Figure 2E). Distribution of gene methyl-ation was similar in all three investigated glioblastoma co-horts. Methylation of POLE4 and MVP was highest in RKT II tumors, whereas most other genes including XRCC3 and CSNK1E were hypermethylated in mesenchymal classified tumors. Mean MGMT methylation was lowest in mesenchy-mal tumors; however, highest percentage of MGMT unmeth-ylated tumors was found in the RTK I subgroup. Comparison between average DDR methylation and global methylation revealed similar patterns with RTK I tumors having lowest
methylation levels (P < 2.2 × 10−16 for both MES and RTK
II against RTK I, Figure S4A,B).
Tumor purity might influence the prognostic effect of CpG methylation. Mesenchymal tumors had an average lower tumor purity compared to RTK I or RTK II tumors (median:
0.57 vs 0.79 and 0.75, P < 1 × 10−57 for both comparisons,
Figure 2F). This may additionally explain the on average lower MGMT promoter methylation level. Conclusively,
XRCC3, CUL4A, and CSK1E methylation highly correlated
with tumor purity (Figure 2G; Figure 4SC-F).
3.2
|
DDR methylome and interaction
with therapy outcome in glioblastoma
A univariate cox proportional hazard model was applied to identify survival associated functional CpGs. There were five genes identified in the exploratory Heidelberg cohort, which were associated with overall survival (OS). Of these, only MGMT promoter methylation was correlated with better prognosis consistently in all three studies (Data S1).
In a pooled multivariate cox analysis of all three stud-ies seven genes prove to be independent markers in both OS and progression-free survival (PFS) after correcting for
MGMT promoter methylation status and multiple testing
(Data S1). Beforehand analysis revealed that the criterion of proportionality was not violated. Further analysis sep-arated by MGMT promoter methylation revealed that four of the DDR genes (TP73, CSNK1E, EXO1, and PRPF19) showed significant association with OS and PFS only in
MGMT unmethylated tumors (Data S2). Moreover, after
addition of tumor purity into the multivariate model, only
PRPF19 (P = .005-.04, c-index = 0.71-0.73) and TP73
(P = .02-.04, c-index = 0.71-0.72) methylation were sig-nificantly associated with better OS and PFS (Data S2). C-indices for multivariate analyses ranged for significant CpGs between 0.66 ± 0.02 and 0.73 ± 0.02 confirming on average good performance of the models (see respective Supporting Information Data).
The recent NOA-08 long term study analysis showed worse outcome of patients in the RTK I subgroup in our
Heidelberg and the NOA-08 biomarker cohort7 as well as
the best prognosis for patients with MGMT promoter meth-ylated tumors with a RTK II classifier assignment. Though with a global P-value of .058 not significant, there was a trend also in the EORTC 26101 recurrent glioblastoma study toward worse survival of patients with the low methylated RTK I tumors (Figure 5A). Furthermore, recent preliminary data link EGFRvIII to a better prognosis in MGMT
meth-ylated tumors.23 In concordance, we found EGFRvIII and
EGFR amplification both highest in the RTK II subgroup
(Figure 5C,D) as one potential factor driving chemotherapy sensitivity in RTK II tumors.
Focusing on subtype-specific CpG markers, we identified only MGMT promoter methylation specifically correlating with improved survival in RTK II tumors, MVP was prog-nostic in RTK I tumors and no CpG was progprog-nostic in MES subgroup tumors (Data S3).
The NOA-08 study with the chemotherapy vs radiother-apy regimen additionally offers the chance of finding CpGs that are specifically associated with benefit from either treat-ment. MGMT, GEN1, PARP4, and CSNK1E were found to be associated with better survival in the chemotherapy group, whereas TP73 and CCND3 methylation were linked to ra-diotherapy sensitivity in the OS analysis, however, none of these reached significance after correction for multiple test-ing (Data S4). For PFS, only MGMT was prognostic in the chemotherapy group.
3.3
|
The association of TERT
promoter mutation with methylation
profiles and survival
In the two cohorts with available TERT status, TERT pro-moter mutation was found in 377 of 455 (83%) patients. Differences were noted between both cohorts (EORTC 26101
FIGURE 2 DNA damage response (DDR) methylation profiles. A, Principle component analysis (PCA) showing
5'-cytosine-phosphat-guanine-3' (CpG) direction. B, PCA with samples colored by Classifier assignment with the 25 DDR CpGs (C) and the 5000 most variable CpGs (D). E, Methylation of DDR CpGs according to the three most abundant glioblastoma subgroups (MES, RTK I, and RTK II). F, Tumor purity according to tumor subtype. G, Example of correlation between tumor purity and XRCC3 methylation. 5ʹ-UTR, 5ʹ untranslated region; av., average; cont., contribution to a principle component; meth., methylation; TSS200, 0-200 base pairs upstream of transcription start site, TSS1500: 200-1500 base pairs upstream of transcription start site; PCA1, principle component 1; PCA2, principle component 2; a full list of classifier abbreviations can be found in the Supporting Information
++++++++++++ +++++ + + ++ ++ ++ + + +++ + + + + + + + + ++ + +++ + + + + 92 (0) 39 (20) 14 (25) 5 (27) 2 (28) 1 (29) 22 (0) 7 (6) 2 (7) 1 (7) 1 (7) 1 (7) 44 (0) 30 (6) 13 (10) 9 (10) 5 (10) 4 (10) TERT=wild−type TERT=C250T TERT=C228T 0 200 400 600 800 1000 Time [days] Number at risk (number censored)
+ ++ ++++ ++ + + + + + + +++++ + + + 189 (0) 39 (1) 4 (10) 1 (11) 1 (11) 0 (11) 74 (0) 12 (1) 2 (3) 0 (4) 0 (4) 0 (4) 34 (0) 13 (0) 1 (6) 0 (6) 0 (6) 0 (6) TERT=wildtype TERT=C250T TERT=C228T 0 200 400 600 800 1000 Time [days]
Number at risk (number censored) time [days]
0 200 400 600 800 1000 3 3 3 , 0 0 0,667 0,193 0,221 0,586 2 , 0 0,2 0,6 2 , 0 0 0,8 0,062 0,296 0,642 0,097 0,2 0,703 6 8 2 , 0 0 0,714 0 0 1 0 0 1 0,5 0 0,5 0 0 1 0,333 0 0,667 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1819 20 2122 176.98 177 chr2 Mb −1 −0.5 0 0.5 1 sex PFS OS MGMT C228 T mtu T R E T tumor p urity GBM RTK I GBM RTK II GBM MES GBM Midline GBM MYC N MEgrey60 MEtan MEdarkgreen MEdarkred MEpurple MEorange MEblue MEred MEblack MElightgreen MEmidnightblue MElightcyan MEpink MEturquoise MEwhite MEpaleturquoise MEdarkorange MEroyalblue MEviolet MEsaddlebrown MEskyblue MEbrown MEdarkgrey MEsteelblue MElightyellow MEmagenta MEgreen MEcyan MEdarkturquoise MEsalmon MEgreenyellow MEyellow MEgrey −0.052
(0.4) (0.3)0.06 (0.03)0.13 −0.026(0.7) 0.078(0.2) −0.064(0.3) 0.032(0.6) (6e−04)−0.2 (2e−53)−0.73 (7e−12)0.39 (2e−04)0.22 (0.006)−0.16 −0.005(0.9) 0.032
(0.6) −0.11(0.07) (0.001)−0.19 −0.097(0.1) −0.0088(0.9) −0.029(0.6) −0.064(0.3) (0.04)0.12 (0.003)0.18 −0.14(0.02) −0.057(0.3) 0.049(0.4) (3e−04)0.21 −0.14
(0.02) −0.057(0.3) −0.11(0.06) 0.0041(0.9) −0.048(0.4) 0.051(0.4) −0.0017(1) (4e−27)0.57 (6e−06)0.27 (0.002)0.18 (2e−14)−0.43 0.08(0.2) 0.012(0.8) −0.1
(0.08) −0.04(0.5) −0.094(0.1) −0.035(0.6) −0.042(0.5) 0.081(0.2) (0.3)0.06 (1e−21)0.52 (2e−10)0.37 0.05(0.4) (7e−12)−0.39 0.089(0.1) 0.075(0.2) −0.023
(0.7) (9e−04)−0.2 (3e−05)−0.24 −0.13(0.03) −0.09(0.1) 0.024(0.7) (0.04)−0.12 −0.11(0.08) (6e−04)0.2 (1e−09)−0.35 (0.002)0.18 (0.09)0.1 0.032(0.6) −0.025 (0.7) −0.12(0.04) (0.006)−0.16 −0.11(0.07) −0.14(0.02) (0.05)0.12 −0.058(0.3) −0.052(0.4) (2e−05)0.25 (3e−09)−0.34 (0.02)0.13 0.067(0.3) 0.046(0.4) −0.074 (0.2) −0.078(0.2) −0.12(0.04) −0.043(0.5) −0.066(0.3) 0.07(0.2) −0.0013(1) (0.01)0.15 (0.003)0.18 −0.095(0.1) −0.047(0.4) 0.046(0.4) −0.041(0.5) −0.051 (0.4) −0.011(0.9) −0.038(0.5) −0.053(0.4) −0.071(0.2) 0.032(0.6) −0.075(0.2) (0.005)−0.17 −0.09(0.1) −0.13(0.03) (3e−04)0.21 0.024(0.7) −0.02(0.7) −0.028 (0.6) −0.098(0.1) −0.15(0.01) −0.15(0.01) −0.11(0.08) 0.077(0.2) −0.06(0.3) −0.028(0.6) (0.04)0.12 (2e−05)−0.25 (0.03)0.13 0.055(0.4) 0.099(0.1) −0.056 (0.4) −0.14(0.02) (0.002)−0.19 −0.13(0.03) −0.11(0.08) (0.06)0.11 2.2e−05(1) −0.021(0.7) 0.081(0.2) (4e−05)−0.24 (0.001)0.19 −0.0081(0.9) −0.0022(1) −0.031
(0.6) (0.003)−0.18 (1e−06)−0.28 (8e−04)−0.2 −0.13(0.02) (0.01)0.15 0.0067(0.9) 0.057(0.3) (2e−05)0.25 (3e−09)−0.34 (0.03)0.13 0.054(0.4) (0.09)0.1 0.051
(0.4) −0.041(0.5) 0.017(0.8) −0.038(0.5) 0.021(0.7) −0.079(0.2) −0.094(0.1) (0.002)−0.18 −0.023(0.7) −0.058(0.3) 0.079(0.2) 0.042(0.5) −0.023(0.7) 0.15
(0.01) −0.066(0.3) −0.1(0.1) −0.028(0.6) −0.091(0.1) 0.035(0.6) (0.08)−0.1 (6e−23)−0.53 (7e−11)0.37 (7e−43)−0.68 (8e−13)0.41 (0.06)0.11 −0.052(0.4) 0.16
(0.007) −0.0038(0.9) 0.013(0.8) −0.067(0.3) −0.034(0.6) 0.00087(1) −0.059(0.3) (8e−120)−0.88 −0.098(0.1) (2e−27)−0.57 (5e−50)0.71 −0.02(0.7) −0.072(0.2) 0.033
(0.6) −0.033(0.6) −0.034(0.6) 0.036(0.5) −0.0048(0.9) −0.029(0.6) −0.057(0.3) (0.1)0.1 (0.03)0.13 −0.0099(0.9) −0.11(0.06) 0.084(0.2) 0.0072(0.9) −0.089
(0.1) 0.062(0.3) 0.063(0.3) 0.023(0.7) 0.038(0.5) −0.028(0.6) 0.021(0.7) (3e−18)0.48 (0.03)0.13 (2e−05)0.25 (6e−12)−0.39 0.017(0.8) (0.3)0.06 −0.03
(0.6) −0.084(0.2) −0.12(0.04) −0.036(0.5) −0.064(0.3) 0.066(0.3) −0.0049(0.9) (5e−23)0.53 (1e−33)0.62 −0.076(0.2) (2e−18)−0.48 (0.1)0.1 (0.03)0.13 0.1
(0.1) −0.087(0.1) −0.046(0.4) −0.088(0.1) −0.061(0.3) 0.033(0.6) −0.055(0.4) (0.002)−0.18 (3e−17)0.47 (5e−21)−0.51 (0.04)0.12 0.023(0.7) (0.01)0.15 −0.023
(0.7) 0.017(0.8) −0.0089(0.9) 0.0071(0.9) −0.01(0.9) (0.09)0.1 (0.01)0.15 (7e−16)0.45 (7e−05)0.23 (0.001)0.19 (6e−12)−0.39 −0.025(0.7) −0.0062(0.9) −0.075
(0.2) −0.046(0.4) 0.039(0.5) 0.042(0.5) −0.034(0.6) (0.03)0.13 (0.01)0.15 (1e−15)0.45 (1e−22)0.53 −0.02(0.7) (2e−15)−0.44 0.046(0.4) 0.065(0.3) −0.085
(0.2) −0.036(0.6) −0.014(0.8) 0.044(0.5) −0.016(0.8) (0.01)0.15 (2e−04)0.22 (1e−26)0.57 (8e−15)0.44 (0.01)0.15 (1e−22)−0.53 −0.03(0.6) 0.057(0.3) −0.15
(0.01) −0.047(0.4) (0.09)−0.1 0.041(0.5) 0.012(0.8) 0.027(0.7) 0.066(0.3) (2e−179)0.94 (6e−11)0.37 (6e−15)0.44 (4e−79)−0.81 0.088(0.1) 0.074(0.2) −0.14
(0.02) −0.013(0.8) 0.051(0.4) 0.056(0.4) −0.063(0.3) (0.007)0.16 (0.008)0.16 (2e−31)0.61 (3e−09)0.34 (4e−06)0.27 (7e−28)−0.58 (0.03)−0.13 (0.07)0.11 −0.077
(0.2) 0.015(0.8) 0.034(0.6) 0.089(0.1) −0.05(0.4) 0.071(0.2) 0.029(0.6) (9e−22)0.52 (3e−07)0.3 (6e−04)0.2 (7e−19)−0.49 0.073(0.2) 0.037(0.5) 0.00014
(1) −0.013(0.8) −0.027(0.7) −0.018(0.8) −0.025(0.7) 0.063(0.3) (0.3)0.06 (3e−11)0.38 (3e−11)0.38 −0.02(0.7) (1e−08)−0.33 0.077(0.2) 0.095(0.1) 0.019
(0.7) −0.031(0.6) −0.035(0.6) −0.11(0.08) −0.071(0.2) 0.076(0.2) −0.00062(1) 0.031(0.6) (0.002)0.18 (0.008)−0.16 −0.0036(1) 0.018(0.8) 0.087(0.1) 0.049
(0.4) −0.068(0.3) −0.081(0.2) −0.024(0.7) 0.032(0.6) −0.03(0.6) 0.0077(0.9) 0.028(0.6) (5e−04)0.21 −0.11(0.06) −0.074(0.2) 0.051(0.4) (0.4)0.05 −0.0062
(0.9) −0.061(0.3) −0.098(0.1) −0.01(0.9) 0.0081(0.9) 0.027(0.7) (0.3)0.06 (1e−10)0.37 (2e−12)0.4 −0.066(0.3) (8e−08)−0.31 (0.02)0.14 (0.06)0.11 −0.011
(0.9) (0.09)−0.1 −0.13(0.03) 0.0051(0.9) 0.0021(1) 0.031(0.6) (0.4)0.056 (2e−09)0.34 (4e−17)0.47 (0.08)−0.1 (1e−07)−0.31 0.092(0.1) 0.074(0.2) −0.051
(0.4) −0.12(0.05) (0.003)−0.18 −0.06(0.3) −0.11(0.08) (0.02)0.14 (0.5)0.038 (4e−13)0.41 (2e−28)0.58 (0.001)−0.19 (4e−08)−0.32 (0.09)0.1 0.099(0.1) −0.055
(0.4) −0.077(0.2) (0.007)−0.16 −0.046(0.4) 0.013(0.8) 0.0097(0.9) 0.039(0.5) (3e−21)0.51 (2e−11)0.38 −0.0092(0.9) (1e−10)−0.37 (0.02)0.13 (0.005)0.17 −0.092
(0.1) −0.016(0.8) −0.097(0.1) 0.016(0.8) 0.045(0.5) −0.0074(0.9) 0.068(0.3) (1e−39)0.66 (1e−13)0.42 (0.02)0.13 (5e−23)−0.53 (0.02)0.14 0.072(0.2) −0.042
(0.5) −0.12(0.05) (0.002)−0.18 −0.011(0.9) −0.047(0.4) 0.083(0.2) 0.055(0.4) (3e−27)0.57 (1e−42)0.68 −0.099(0.1) (7e−20)−0.5 (0.01)0.15 0.067(0.3)
C250T B C A D E F TERT C228T C250T wildtype classifier
APA GBM MES GBM Midline GBM MYCN GBM RTK I GBM RTK II H3 K27M INFLAM LGG DNT LGG GG no match REACT
numbe
r
0 50 100
primary tumor - HD cohort
overall surviv al [f ra ction] 0 0.25 0.50 0.75 1.00 p = 0.013 time [days] 0 200 400 600 800 1000 C228T C250T wildtype
recurrent tumor - EORTC 26101 cohort
C228T C250T wildtype over all surviv al [f ra ction] 0 0.25 0.50 0.75 1.00 p = 0.26
88%, Heidelberg cohort 72%). TERT promoter mutation was predominantly found in the three main glioblastoma groups (MES, RTK I, and RTK II). For TERT wild-type tumors 21% belong to these groups, as well as 14% of the TERT mutated do (Figure 3A). TERT wild-type status was associated with better PFS in the Heidelberg cohort, but not in the recurrent EORTC 26101 cohort (Figure 3B,C). We restricted the
anal-ysis to RTK I, RTK II, MES, Midline, and MYCN18 tumors
and found 500 differentially methylated CpGs compared to >10 000 differential CpGs in the unrestricted analysis un-derling the effect of the glioma classification over TERT mutation. Differentially methylated regions showed high overlap between C228T and C250T tumors (Figure 3D). On chromosome 2, a cluster of DMRs in the promoter regions of HOXD genes was identified (Figure 3E). A weighted-gene correlation network analysis (WGCNA) identified 30 mod-ules of CpGs within the EORTC 26101 data set (Figure 3F). Most modules correlated strongly with tumor purity and dif-ferentiated between the RTK I, RTK II, and MES subtypes. Highest correlation to TERT promoter mutation status was conclusively found in a module specific for the RTK I pheno-type with CpGs restricted to chromosome 2 in the HOXD12,
HOXD4, HOXD3, and MIR10B promoters. However, no
spe-cific survival associated module was correlated with TERT promoter mutation. Copy-number analysis revealed associa-tion of TERT mutaassocia-tion with amplificaassocia-tion of chromosome 7 and loss of chromosome 10, for example, with a typical glio-blastoma phenotype in differential and WGCNA analysis.
3.4
|
DDR genes and therapy response in
glioblastoma cell lines and primary glioma
cell cultures
TERT status, DDR methylome, MGMT promoter
meth-ylation and methmeth-ylation patterns in primary glioblastoma cell cultures and cell lines are depicted in Figure 4A. Hierarchical clustering separated adherently growing cell lines from primary cells based on DDR methyla-tion. All samples harbor either TERT promoter mutation C228T or C250T, all but one (9/10) are MGMT methylated (Table S2). Methylation is retained in cell culture as all but one (5/6) primary cell lines show a matching profile with one of the main glioblastoma methylation subgroups. Of note, adherent cell lines grown in serum change their
methylation profile in cell culture being most conclusive with a pediatric plexus tumor while copy number variation (CNV) profiles still allow identification as derived from glioblastoma (Table S2; Figure S6). T-SNE analysis of methylation patterns shows clear separation between cell lines and primary cultures as well as IDH wild-type glioma samples (Figure 4B).
PRFP19 methylation was associated with survival in
the combined analysis and in particular in MGMT promoter unmethylated, therefore, mainly temozolomide resistant, tu-mors and retained significance after controlling for tumor purity. Therefore, besides TERT promoter mutation, PRPF19 methylation might be an interesting prognostic marker. We independently confirmed a negative correlation between
PRPF19 methylation and expression (Figure 4C; r = −.39)
and positive correlation of TERT mutation and TERT expres-sion in the subset of the Heidelberg cohort with available ex-pression data (Figure S2). Low methylation/high exex-pression primary glioma cultures were picked for lentiviral gene ex-pression modulation of PRPF19 and TERT. Knockdown of
PRPF19 and TERT was validated via quantitative real-time
PCR (Figure S7) and resulted in an enhanced response to te-mozolomide treatment. Cell cycle analysis revealed a higher proportion of G2 arrested cells after temozolomide treatment in tumor cells deficient of PRFP19 and TERT compared to equally treated tumor cells transfected with respective control vector (Figure 4D,E,H). Similarly, clonogenicity of tumor cells treated with temozolomide was reduced particularly in
PRPF19 and TERT knockdown cells (Figure 4F,G,I). These
effects were not observed for irradiation with both knock-downs (Figure S8). A summary with relevant findings is given in Figure 5A,B.
4
|
DISCUSSION
With this cross-study analysis based on large recent cohorts of glioblastoma patients, we provide evidence for a prognos-tic role of DDR genes including DDR methylome and TERT promoter mutation status. In our view, this holds several im-portant implications:
Besides the well described DDR gene MGMT, we identi-fied DDR genes for which methylation is linked to survival, in particular in patients with MGMT promoter unmethylated tumors. This is of particular interest as these tumors at best
FIGURE 3 Correlation of telomerase reverse transcriptase (TERT) status with glioblastoma subgroups. A, TERT promoter mutation status and
tumor classifier subgroup. Survival analysis of patients with primary glioblastoma (B) Heidelberg cohort and recurrent glioblastoma (C) EORTC 26101 cohort. D, Visualization of genome distribution of differential methylated regions between TERT promoter mutated and wild-type tumors. E, Differential methylated regions on chromosome 2 in C225T (upper row) and C250T (lower row) tumors. F, Heatmap of different 5'-cytosine-phosphat-guanine-3' modules as the result of the WGCNA analysis and relationship to several tumor-specific markers. C225T, C250T, mutation location upstream of the TERT transcription start site; chr2, chromosome 2; PFS, progression-free survival; Mb, megabase; OS, overall survival, a full list of classifier abbreviations can be found in the Supporting Information
● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● r = -0.39 PRPF19 expression
PRPF19 methylation [beta value] 0.05 0.10 0.15 0.20 40 50 60 70 TP73 cg13943358 TP73 cg03050669 RPA3 cg19739774 PARP4 cg20765408 SMC5 cg14289863 XRCC3 cg25999604 CCND3 cg17945153 PARP3 cg12554573 MGMT cg14194875 MGMT cg12434587 MVP cg26990835 POLE4 cg20919922 POLE4 cg02307033 XRCC3 cg18597188 XRCC3 cg23193616 CUL4A cg26825849 CSNK1E cg04728789 CSNK1E cg13110239 EXO1 cg04706276 CHEK1 cg00554702 EXO1 cg16121177 PRPF19 cg00778920 POLR2E cg26528128 GEN1 cg11330941 T1 S24 BG5 L2 T325 P3XX A172 LN229 U8 7 LN308 0 5 10 15 20 25 chr −10 −5 0 5 z-score TSS1500 TSS200 5'UTR feature GB RTK I GB RTK II plexus tumor ped B no match
classifier
cell line primary glioma cell
type TERT C250T C228T A C B cell line D LN308 - clonogenicity temozolomide [µM]0 5 0 100 LN308 - cloogenicity * * 0 5 E * * * * * * * * * * F G H I
primary cultureIDHmut glioma IDHwt glioma PC10 25 -25 -50 PC 2 0 -50 50 LN308 - cell cycle temozolomide [µM] cells in G2 phase [%] 0 5 0 10 20 30 control shPRPF19_1 shPRPF19_2 control shPRPF19_1 shPRPF19_2 control shPRPF19_1 shPRPF19_2 control shPRPF19_1 shPRPF19_2 control shTERT_1 shTERT_2 control shTERT_1 shTERT_2 S24 - cell cycle cells in G2 phase [%] temozolomide [µM] 0 100 0 10 20 30 clonogenicity [%] 0 2 4 6 S24 - clonogenicity clonogenicity [%] 0 20 40 60 80 LN308 - cell cycle temozolomide [µM] temozolomide [µM] cells in G2 phase [%] 0 10 20 30 temozolomide [µM]0 5 clonogenicity [%] 0 2 4 6
have very limited response to standard chemotherapy with temozolomide and since further prognostic molecular factors have remained elusive. In our analysis, we identified methyl-ation of the two DDR genes PRPF19 and TP73 significantly associated with better OS and PFS in particular in patients with MGMT promoter unmethylated tumors in our multivar-iate analysis. PRPF19 was previously reported to be involved
in DDR24 and has a potential role in oncogenesis,25 but
lit-tle is known about its function in glioblastoma. TP73 is a
member of the TP53 gene family and overexpressed in a
vari-ety of cancers.26,27 Its regulation is complex and not fully
un-derstood, especially the association between methylation and
gene expression is a controversy in different cancers,26,28 but
a regulatory role in chemotherapy response and probably
sen-sitivity through DNA methylation has been described.29 We
here describe a negative association between methylation and expression for the prognosis relevant CpG cg13943358 and cg2316013 in glioblastoma. The association in chemotherapy response makes both genes very plausible markers for glioma prognosis in the absence of MGMT methylation as a sensitiv-ity factor against chemotherapy, however, formal testing for a predictive effect of the methylation levels of both genes on chemotherapy response was not in the intention of the study and the analysis comparing different treatment methods in the NOA-08 cohort lacks sufficient sample size for difference detection, and therefore, regarded as exploratory.
Functional evidence for chemosensitivity, however, was validated for PRPF19 by gene silencing in glioblas-toma cells. Depletion of PRPF19 expression resulted in the anticipated sensitization to temozolomide in glioblastoma cell lines and primary cell cultures, and may therefore, be investigated further as a predictive marker in MGMT promoter unmethylated glioblastoma. A limitation to use
PRPF19 methylation as prognostic marker is its overall
rel-atively low methylation, but combination with expression may improve prognostic relevance. The TP73 gene was not functionally analyzed in this study, but represents an at-tractive area for further research. The effect of upfront al-kylating agents vs targeted treatments in MGMT promoter unmethylated glioblastoma on its prognostic impact could be answered by subgroup analysis of our currently
recruit-ing N2M2 clinical study.30
RTK I glioblastomas remain a less understood, poorly performing group that have lower DDR and overall meth-ylation levels. Only MVP methmeth-ylation was prognostic in this group, however, overall promoter methylation of this
FIGURE 4 DNA damage response (DDR) genes and therapy response in glioblastoma cells. A, DDR methylome, telomerase reverse
transcriptase (TERT) status, MGMT promoter methylation and methylation patterns in primary glioblastoma cell cultures and cell lines. B, T-SNE analysis of cell lines, primary cell cultures and IDH wild-type samples. C, Correlation between PRPF19 methylation and expression. D, Cell cycle analysis in LN308 cells, silenced for PRPF19 or transfected with respective control vector, after treatment with 5 µmol/L of temozolomide or dimethyl sulfoxide (DMSO) as control. Two different knockdown constructs were used for analysis. E, Cell cycle analysis in S24 cells, silenced for PRPF19 or transfected with respective control vector, after treatment with 100 µmol/L of temozolomide or DMSO as control. Two different knockdown constructs were used for analysis. F, Clonogenicity of LN-308 cells, silenced for PRPF19 or transfected with respective control vector, after treatment with 5 µmol/L of temozolomide or DMSO as control. Two different knockdown constructs were used for analysis. G, Clonogenicity of S24 cells, silenced for PRPF19 or transfected with respective control vector, after treatment with 100 µmol/L of temozolomide or DMSO as control. Two different knockdown constructs were used for analysis. H, Cell cycle analysis in LN308 cells silenced for TERT or transfected with respective control vector, after treatment with 5 µmol/L of temozolomide or DMSO as control. Two different knockdown constructs were used for analysis. I, Clonogenicity in LN-308 silenced for TERT or transfected with respective control vector, after treatment with 5 µmol/L of temozolomide or DMSO as control. Two different knockdown constructs were used for analysis. For panels D-I the mean value and SD of three independent experiments is shown. *P < .05. 5ʹ-UTR: 5ʹ untranslated region; chr, chromosome; TSS200, 0-200 base pairs upstream of transcription start site; TSS1500, 200-1500 base pairs upstream of transcription start site
FIGURE 5 Summary of relevant findings. A, Methylation
analysis of 450 DNA damage response (DDR) genes revealed 17 functional DDR genes of which in seven genes hypomethylation showed association with reduced survival. Hypomethylation of
PRPF19 and TP73 was associated with worse survival in patients
with MGMT promoter unmethylated tumors with adjustment for tumor purity. TERT promoter mutations were correlated with methylation groups and survival times. B, PRPF19 and TERT k/d-induced sensitivity in glioblastoma cells toward temozolomide but not radiotherapy. k/d, knock down; MGMT, unmethylated MGMT promoter; TMZ, temozolomide; RT, radiotherapy
TERT k/d PRPF19 k/d
TERT promoter mutation methylation analysis (450 genes)
functional methylation (17 genes)
TMZ RT +
DNA damage response
survial (7 genes) + tumor purity MGMT - (PRPF19, TP73) A survival (TERT) B
gene was low challenging its suitability as a robust marker. Further studies might dive deeper into the differences es-pecially between RTK I and II tumors and their differential methylation profiles.
Further limitations of this methylation analysis approach include the heterogeneity of the three well-documented patient cohorts used that might reduce sensitivity for markers poten-tially present only in certain subgroups, but enables to cover a variety of conditions for detection of strong universal markers. Even the NOA-08 study compared radiotherapy vs chemother-apy the methylation analysis was not powered nor intended to detect chemosensitivity of certain CpGs. Furthermore, the co-horts based on the two large studies EORTC 26101 and NOA-08 were subsets of the original study population based on the availability of tissue for methylation array analysis. Therefore a sampling bias that often tends toward a better prognosis in the biomarker cohorts cannot be fully excluded. The approach to include only CpGs with negative correlation of methylation with expression in glioblastoma ensures a higher chance of finding functionally relevant genes from a small defined set, but may not be exhaustive as also a subset of CpGs with pos-itive correlations of methylation and expression or CpGs not captured by the stringent threshold could be robust prognostic factors though complex regulations.
TERT promotor mutation is the main factor facilitating
TERT expression and several studies reported differential out-comes based on TERT mutation and MGMT promoter
methyla-tion,6,11,12 this should be taken with caution as these might have
included patients with nonglioblastoma methylation groups as a potential confounder. Here, we demonstrated that TERT mu-tation is associated with worse survival in well characterized cohorts and silencing of TERT expression in glioma tumor cells was associated with an enhanced response to temozolomide treatment.
This study furthermore holds implications for preclinical models. Primary glioma cultures nicely retain the glioblasto-ma-like methylation state, whereas cell lines change to a meth-ylation profile most consistent with a pediatric plexus tumor. Although we have observed similar results for our functional studies between primary glioma cells and adherent cell lines and both models cluster outside the patient tumor samples, the methylation profiling strongly encourages the use of pri-mary cell lines as an appropriate model glioma for glioma biology as they retain a well-preserved glioma methylation phenotype. Of note, we have not observed a glioblastoma MES primary cell line in our sample. This might be because of the relatively low number of primary cell lines (n = 9), but the lower tumor purity in MES glioblastomas might prevent detection of this phenotype in cell culture.
In summary, low methylation of DDR genes and TERT promoter mutation are associated with worse prognosis in glioblastoma patients and current studies on DDR inhibitors with and without other cytotoxic or immunological therapies
may finally yield benefit especially for the heavily under-served patient population with tumors having an unmethyl-ated MGMT promoter.
ACKNOWLEDGMENTS
We are grateful to F. Hoffmann—La Roche Ltd., the German Cancer Research Center (DKFZ) and the NCT Heidelberg and the German Research Foundation. We thank all the patients and relatives for participating and all physicians, nurses, and stuff for care and thorough docu-mentation. We thank the IT core facility of the DKFZ for providing cluster computing capacities for the methylation analysis. We thank Laura Dörner and Antje Habel from the Department of Neuropathology, Heidelberg University Hospital for support with tissue handling and TERT pro-moter mutation analysis. Open access funding enabled and organized by ProjektDEAL.
CONFLICT OF INTEREST
TK, AB, AS, FS, TG, CM, DH, AW, PH, PR, MB, CO, RS, FW, AB, AD, and MP reported no conflict of interest. Michael Weller: Research grants from Abbvie, Adastra, Bristol Meyer Squibb (BMS), Dracen, Merck, Sharp & Dohme (MSD), Merck (EMD), Novocure, Piqur and Roche, and honoraria for lectures or advisory board partici-pation or consulting from Abbvie, Basilea, Bristol Meyer Squibb (BMS), Celgene, Merck, Sharp & Dohme (MSD), Merck (EMD), Novocure, Orbus, Roche and Tocagen. Martin van den Bent: Consulting for Celgene, BMS, Agios, Boehringer, Abbvie, Bayer, Carthera, Nerviano, Genenta. Wolfgang Wick has received study drug support from Apogenix, Pfizer, and Roche and consulted for MSD and Roche with all financial reimbursement to the University Clinic.
AUTHOR CONTRIBUTIONS
Study concept and design: TK, WW; acquisition of molecu-lar and in vitro data: TK, AB, FS, DH, PR, WW; data acquisi-tion of clinical studies: TK, FS, TG, CM, AW, MW, MB, RS, FW, AB, AD, MP, WW; analysis and interpretation of data: TK, AB, AS, WW; statistical analysis: TK; study supervi-sion and coordination: WW; writing the manuscript: TK, AB, WW; critical revision of manuscript for intellectual content: AS, FS, TG, C.M, DH, AW, PK, PR, MB, CO, MW, MB, RS, FW, AB, AD, MP.
ORCID
Tobias Kessler https://orcid.org/0000-0001-8350-7074
Ahmed Sadik https://orcid.org/0000-0002-0328-1990
Dirk C. Hoffmann https://orcid.org/0000-0003-1370-1933
Alba Brandes https://orcid.org/0000-0002-2503-9089
Michael Platten https://orcid.org/0000-0002-4746-887X
REFERENCES
1. Körber V, Yang J, Barah P, et al. Evolutionary trajectories of IDH(WT) glioblastomas reveal a common path of early tumori-genesis instigated years ahead of initial diagnosis. Cancer Cell. 2019;35:692-704.e12.
2. Ma J, Benitez JA, Li J, et al. Inhibition of nuclear PTEN tyrosine phosphorylation enhances glioma radiation sensitivity through at-tenuated DNA repair. Cancer Cell. 2019;36:690-691.
3. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med. 2005;352:997-1003.
4. Weller M, Tabatabai G, Kastner B, et al. MGMT promoter methyl-ation is a strong prognostic biomarker for benefit from dose-inten-sified temozolomide rechallenge in progressive glioblastoma: the DIRECTOR trial. Clin Cancer Res. 2015;21:2057-2064.
5. Wick W, Platten M, Meisner C, et al. Temozolomide chemother-apy alone versus radiotherchemother-apy alone for malignant astrocytoma in the elderly: the NOA-08 randomised, phase 3 trial. Lancet Oncol. 2012;13:707-715.
6. Kessler T, Sahm F, Sadik A, et al. Molecular differences in IDH wildtype glioblastoma according to MGMT promoter methylation.
Neuro Oncol. 2018;20:367-379.
7. Wick A, Kessler T, Platten M, et al. Superiority of temozolomide over radiotherapy for elderly patients with RTK II methylation class, MGMT promoter methylated malignant astrocytoma. Neuro
Oncol. 2020;22(8):1162-1172.
8. Bady P, Kurscheid S, Delorenzi M, et al. The DNA methylome of DDR genes and benefit from RT or TMZ in IDH mutant low-grade gli-oma treated in EORTC 22033. Acta Neuropathol. 2018;135:601-615. 9. Brown JS, O'Carrigan B, Jackson SP, Yap TA. Targeting DNA
repair in cancer: beyond PARP inhibitors. Cancer Discov. 2017;7:20-37.
10. Lord CJ, Ashworth A. The DNA damage response and cancer ther-apy. Nature. 2012;481:287-294.
11. Arita H, Yamasaki K, Matsushita Y, et al. A combination of TERT promoter mutation and MGMT methylation status predicts clini-cally relevant subgroups of newly diagnosed glioblastomas. Acta
Neuropathol Commun. 2016;4:79.
12. Nguyen HN, Lie A, Li T, et al. Human TERT promoter mutation enables survival advantage from MGMT promoter methylation in IDH1 wild-type primary glioblastoma treated by standard chemo-radiotherapy. Neuro Oncol. 2017;19:394-404.
13. Abou-El-Ardat K, Seifert M, Becker K, et al. Comprehensive molecular characterization of multifocal glioblastoma proves its monoclonal origin and reveals novel insights into clonal evolution and heterogeneity of glioblastomas. Neuro Oncol. 2017;19:546-557.
14. Pearl LH, Schierz AC, Ward SE, Al-Lazikani B, Pearl FM. Therapeutic opportunities within the DNA damage response. Nat
Rev Cancer. 2015;15:166-180.
15. Wick W, Gorlia T, Bendszus M, et al. Lomustine and bevacizumab in progressive glioblastoma. N Engl J Med. 2017;377:1954-1963. 16. Pfaff E, Kessler T, Balasubramanian GP, et al. Feasibility of real-time
molecular profiling for patients with newly diagnosed glioblastoma without MGMT promoter hypermethylation-the NCT Neuro Master Match (N2M2) pilot study. Neuro Oncol. 2018;20:826-837. 17. Duarte RC, Morrison A. Dynamic stability of sequential
stimu-lus representations in adapting neuronal networks. Front Comput
Neurosci. 2014;8:124.
18. Bady P, Sciuscio D, Diserens A-C, et al. MGMT methylation analysis of glioblastoma on the Infinium methylation BeadChip identifies two distinct CpG regions associated with gene silenc-ing and outcome, yieldsilenc-ing a prediction model for comparisons across datasets, tumor grades, and CIMP-status. Acta Neuropathol. 2012;124:547-560.
19. Capper D, Jones DTW, Sill M, et al. DNA methylation-based classification of central nervous system tumours. Nature. 2018;555:469-474.
20. Zheng X, Zhang N, Wu HJ, Wu H. Estimating and accounting for tumor purity in the analysis of DNA methylation data from cancer studies. Genome Biol. 2017;18:17.
21. Lemke D, Weiler M, Blaes J, et al. Primary glioblastoma cultures: can profiling of stem cell markers predict radiotherapy sensitivity?
J Neurochem. 2014;131:251-264.
22. Kessler T, Sahm F, Blaes J, et al. Glioma cell VEGFR-2 confers resistance to chemotherapeutic and antiangiogenic treatments in PTEN-deficient glioblastoma. Oncotarget. 2015;6:31050-31068. 23. Struve N, Brend T, Ott L, et al. EGFRvIII: a predictive marker for
Temozolomide response in O6-methylguanine-DNA methyltrans-ferase negative glioblastoma cells and tumor xenografts. Neuro
Oncol. 2016;18(Suppl 4):iv33.
24. Maréchal A, Li J-M, Ji X, et al. PRP19 transforms into a sensor of RPA-ssDNA after DNA damage and drives ATR activation via a ubiquitin-mediated circuitry. Mol Cell. 2014;53:235-246.
25. Yin J, Zhu JM, Shen XZ. New insights into pre-mRNA pro-cessing factor 19: a multi-faceted protein in humans. Biol Cell. 2012;104:695-705.
26. Yao Z, Di Poto C, Mavodza G, Oliver E, Ressom HW, Sherif ZA. DNA methylation activates TP73 expression in hepatocellular car-cinoma and gastrointestinal cancer. Sci Rep. 2019;9:19367. 27. Stiewe T, Zimmermann S, Frilling A, Esche H, Putzer BM.
Transactivation-deficient DeltaTA-p73 acts as an oncogene.
Cancer Res. 2002;62:3598-3602.
28. García-Martínez A, Sottile J, Sánchez-Tejada L, et al. DNA meth-ylation of tumor suppressor genes in pituitary neuroendocrine tu-mors. J Clin Endocrinol Metab. 2019;104:1272-1282.
29. Ohmura H, Ito M, Uchino K, et al. Methylation of drug resis-tance-related genes in chemotherapy-sensitive Epstein-Barr vi-rus-associated gastric cancer. FEBS Open Bio. 2020;10:147-157. 30. Wick W, Dettmer S, Berberich A, et al. N2M2 (NOA-20) phase I/II
trial of molecularly matched targeted therapies plus radiotherapy in patients with newly diagnosed non-MGMT hypermethylated glio-blastoma. Neuro Oncol. 2019;21:95-105.
SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section.
How to cite this article: Kessler T, Berberich A, Sadik A, et al. Methylome analyses of three
glioblastoma cohorts reveal chemotherapy sensitivity markers within DDR genes. Cancer Med.
