Advertisement

Nuclear Exclusion of TET1 Is Associated with Loss of 5-Hydroxymethylcytosine in IDH1 Wild-Type Gliomas

      The recent identification of isocitrate dehydrogenase 1 (IDH1) gene mutations in gliomas stimulated various studies to explore the molecular consequences and the clinical implications of such alterations. The Cancer Genome Atlas Research Network showed evidence for a CpG island methylator phenotype in glioblastomas that was associated with IDH1 mutations. These alterations were associated with the production of the oncometabolite, 2-hydroxyglutarate, that inhibits oxygenases [ie, ten-eleven translocation (TET) enzymes involved in the oxidation of 5-methylcytosine to 5-hydroxymethylcytosine (5hmC)]. We investigated 60 gliomas for 5hmC presence, 5-methylcytosine content, TET1 expression, and IDH1 mutation to gain insight into their relationships on a histological level. Of gliomas, 61% revealed no immunoreactivity for 5hmC, and no correlation was observed between IDH1 mutations and loss of 5hmC. Interestingly, expression of TET1 showed remarkable differences regarding overall protein levels and subcellular localization. We found a highly significant (P = 0.0007) correlation between IDH1 mutations and nuclear accumulation of TET1, but not with loss of 5hmC. Of 5hmC-negative gliomas, 70% showed either exclusive or dominant cytoplasmic expression, or no detectable TET1 protein (P = 0.0122). Our data suggest that the loss of 5hmC is a frequent event in gliomas, independent of IDH1 mutation, and may be influenced by the nuclear exclusion of TET1 from the nuclei of glioma cells.
      Glioblastoma (World Health Organization grade IV) is the most frequent and most malignant type of glioma. Because of poor response to radiation therapy and most forms of chemotherapy, glioblastomas have a dismal prognosis, with a median survival time of <12 months.
      • Kleihues P.
      • Louis D.N.
      • Scheithauer B.W.
      • Rorke L.B.
      • Reifenberger G.
      • Burger P.C.
      • Cavenee W.K.
      The WHO classification of tumors of the nervous system.
      Primary or de novo glioblastomas are characterized by multiple genetic alterations, commonly including loss of heterozygosity on chromosome 10, EGFR amplification and overexpression, CDKN2A deletion, and PTEN mutation.
      • Ohgaki H.
      • Kleihues P.
      Genetic pathways to primary and secondary glioblastoma.
      Only 5% of glioblastomas, the so-called secondary glioblastomas, develop through progression from pre-existing astrocytomas of WHO grade 2 or 3. They typically lack EGFR amplification but frequently carry TP53 mutations.
      Recently, somatic mutations of the isocitrate dehydrogenase (IDH1/IDH2) genes have been reported in most secondary glioblastomas and diffuse astrocytic and oligodendroglial tumors as well. These mutations, however, are rare in primary glioblastomas.
      • Parsons D.W.
      • Jones S.
      • Zhang X.
      • Lin J.C.-H.
      • Leary R.J.
      • Angenendt P.
      • Mankoo P.
      • Carter H.
      • Siu I.-M.
      • Gallia G.L.
      • Olivi A.
      • McLendon R.
      • Rasheed B.A.
      • Keir S.
      • Nikolskaya T.
      • Nikolsky Y.
      • Busam D.A.
      • Tekleab H.
      • Diaz L.A.
      • Hartigan J.
      • Smith D.R.
      • Strausberg R.L.
      • Marie S.K.N.
      • Shinjo S.M.O.
      • Yan H.
      • Riggins G.J.
      • Bigner D.D.
      • Karchin R.
      • Papadopoulos N.
      • Parmigiani G.
      • Vogelstein B.
      • Velculescu V.E.
      • Kinzler K.W.
      An integrated genomic analysis of human glioblastoma multiforme.
      • Yan H.
      • Parsons D.W.
      • Jin G.
      • McLendon R.
      • Rasheed B.A.
      • Yuan W.
      • Kos I.
      • Batinic-Haberle I.
      • Jones S.
      • Riggins G.J.
      • Friedman H.
      • Friedman A.
      • Reardon D.
      • Herndon J.
      • Kinzler K.W.
      • Velculescu V.E.
      • Vogelstein B.
      • Bigner D.D.
      IDH1 and IDH2 mutations in gliomas.
      Interestingly, an IDH1 mutation appears to be associated with increased DNA methylation at 5′-CpG islands of multiple genes, the so-called CpG island methylator phenotype.
      • Noushmehr H.
      • Weisenberger D.J.
      • Diefes K.
      • Phillips H.S.
      • Pujara K.
      • Berman B.P.
      • Pan F.
      • Pelloski C.E.
      • Sulman E.P.
      • Bhat K.P.
      • Verhaak R.G.W.
      • Hoadley K.A.
      • Hayes D.N.
      • Perou C.M.
      • Schmidt H.K.
      • Ding L.
      • Wilson R.K.
      • Van Den Berg D.
      • Shen H.
      • Bengtsson H.
      • Neuvial P.
      • Cope L.M.
      • Buckley J.
      • Herman J.G.
      • Baylin S.B.
      • Laird P.W.
      • Aldape K.
      Cancer Genome Atlas Research Network
      Identification of a CpG island methylator phenotype that defines a distinct subgroup of glioma.
      An IDH1 mutation in cancer cells leads to increased levels of the oncometabolite, 2-hydroxyglutarate (2-HG), and to a decrease of α-ketoglutarate (α-KG),
      • Dang L.
      • White D.W.
      • Gross S.
      • Bennett B.D.
      • Bittinger M.A.
      • Driggers E.M.
      • Fantin V.R.
      • Jang H.G.
      • Jin S.
      • Keenan M.C.
      • Marks K.M.
      • Prins R.M.
      • Ward P.S.
      • Yen K.E.
      • Liau L.M.
      • Rabinowitz J.D.
      • Cantley L.C.
      • Thompson C.B.
      • Vander Heiden M.G.
      • Su S.M.
      Cancer-associated IDH1 mutations produce 2-hydroxyglutarate.
      inhibiting the activity of α-KG–dependent oxygenases, which are involved in demethylation of genomic DNA and histone tails. The 2-oxoglutarate and FeII-dependent dioxygenases TET1 and TET2 (ten-eleven translocation) are responsible for the oxidation of 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC) and require α-KG and oxygen.
      • Tahiliani M.
      • Koh K.P.
      • Shen Y.
      • Pastor W.A.
      • Bandukwala H.
      • Brudno Y.
      • Agarwal S.
      • Iyer L.M.
      • Liu D.R.
      • Aravind L.
      • Rao A.
      Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1.
      TET3 is involved in the oxidation of 5mC to 5hmC in zygotic paternal DNA after fertilization.
      • Gu T.-P.
      • Guo F.
      • Yang H.
      • Wu H.-P.
      • Xu G.-F.
      • Liu W.
      • Xie Z.-G.
      • Shi L.
      • He X.
      • Jin S.-G.
      • Iqbal K.
      • Shi Y.G.
      • Deng Z.
      • Szabó P.E.
      • Pfeifer G.P.
      • Li J.
      • Xu G.-L.
      The role of Tet3 DNA dioxygenase in epigenetic reprogramming by oocytes.
      • Iqbal K.
      • Jin S.-G.
      • Pfeifer G.P.
      • Szabó P.E.
      Reprogramming of the paternal genome upon fertilization involves genome-wide oxidation of 5-methylcytosine.
      The amount of 5hmC may vary between different tissues,
      • Li W.
      • Liu M.
      Distribution of 5-hydroxymethylcytosine in different human tissues.
      whereas the content of 5mC does not show significant differences between different cell types. Interestingly, the highest levels have been found in brain tissue,
      • Kriaucionis S.
      • Heintz N.
      The nuclear DNA base 5-hydroxymethylcytosine is present in Purkinje neurons and the brain.
      • Pastor W.A.
      • Pape U.J.
      • Huang Y.
      • Henderson H.R.
      • Lister R.
      • Ko M.
      • McLoughlin E.M.
      • Brudno Y.
      • Mahapatra S.
      • Kapranov P.
      • Tahiliani M.
      • Daley G.Q.
      • Liu X.S.
      • Ecker J.R.
      • Milos P.M.
      • Agarwal S.
      • Rao A.
      Genome-wide mapping of 5-hydroxymethylcytosine in embryonic stem cells.
      and they have been located predominantly within CpG-rich promoters.
      • Jin S.G.
      • Wu X.
      • Li A.X.
      • Pfeifer G.P.
      Genomic mapping of 5-hydroxymethylcytosine in the human brain.
      Several lines of evidence indicate that 5hmC is an intermediate oxidation product that may ultimately lead to demethylation of the respective region by either an active or a passive mechanism.
      • Tahiliani M.
      • Koh K.P.
      • Shen Y.
      • Pastor W.A.
      • Bandukwala H.
      • Brudno Y.
      • Agarwal S.
      • Iyer L.M.
      • Liu D.R.
      • Aravind L.
      • Rao A.
      Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1.
      Although silencing of cancer-associated genes by CpG hypermethylation at CpG islands is a well-established pathomechanism,
      • Jones P.A.
      • Baylin S.B.
      The epigenomics of cancer.
      the biological and pathophysiological function of 5hmC is still unknown and needs to be elucidated. In a genome-wide study on the distribution of 5hmC, 5mC, and TET1 in embryonic stem cells, Williams et al
      • Williams K.
      • Christensen J.
      • Pedersen M.T.
      • Johansen J.V.
      • Cloos P.A.
      • Rappsilber J.
      • Helin K.
      TET1 and hydroxymethylcytosine in transcription and DNA methylation fidelity.
      observed a predominant localization of TET1 to transcription start sites of CpG–rich promoters. In contrast to 5mC, 5hmC was frequently contained in these regions. Based on these observations, the authors suggested that TET1 opposes aberrant DNA methylation at CpG–rich sequences.
      • Williams K.
      • Christensen J.
      • Pedersen M.T.
      • Johansen J.V.
      • Cloos P.A.
      • Rappsilber J.
      • Helin K.
      TET1 and hydroxymethylcytosine in transcription and DNA methylation fidelity.
      Therefore, it is possible that deregulation of this process may contribute to hypermethylation events of cancer cells or the GCIMP phenotype. Recently published data show decreased levels of 5hmC in several human cancers.
      • Li W.
      • Liu M.
      Distribution of 5-hydroxymethylcytosine in different human tissues.
      • Jin S.-G.
      • Jiang Y.
      • Qiu R.
      • Rauch T.A.
      • Wang Y.
      • Schackert G.
      • Krex D.
      • Lu Q.
      • Pfeifer G.P.
      5-Hydroxymethylcytosine is strongly depleted in human cancers but its levels do not correlate with IDH1 mutations.
      There is a strong correlation between the degree of cellular differentiation and the content of 5hmC, with terminal differentiated cells showing the highest levels of 5hmC.
      • Haffner M.C.
      • Chaux A.
      • Meeker A.K.
      • Esopi D.M.
      • Gerber J.
      • Pellakuru L.G.
      • Toubaji A.
      • Argani P.
      • Iacobuzio-Donahue C.
      • Nelson W.G.
      • Netto G.J.
      • De Marzo A.M.
      • Yegnasubramanian S.
      Global 5-hydroxymethylcytosine content is significantly reduced in tissue stem/progenitor cell compartments and in human cancers.
      In this study, we investigated the distribution of 5hmC, 5mC, and TET1 in primary glioma tissues (anaplastic astrocytomas and secondary and primary glioblastomas) and glioblastoma cell lines. In addition, transcript levels of TET13 genes and the presence of IDH1 and IDH2 mutations were determined in our patient cohort to obtain insight into the interplay between these molecular events. Our results indicate that loss of 5hmC from nuclei is a frequent event in gliomas and that it is independent from IDH1 mutation status. Most interestingly, gliomas with absent or significantly reduced 5hmC content show a remarkable nuclear exclusion or no TET1 protein expression. Nuclear accumulation of TET1, on the other hand, was observed in gliomas with IDH1 mutations.

      Materials and Methods

      Patient Population and Tumor Specimens

      Tumor specimens from 60 patients, including 39 with primary glioblastomas (WHO grade 4), 11 with secondary glioblastomas (WHO grade 4), and 10 with anaplastic astrocytomas (World Health Organization grade 3) were included in the study. The patients were 23 females and 37 males. All tumors were diagnosed according to the WHO classification of tumors of the central nervous system.
      • Louis D.N.
      • Ohgaki H.
      • Wiestler O.D.
      • Cavenee W.K.
      • Burger P.C.
      • Jouvet A.
      • Scheithauer B.W.
      • Kleihues P.
      The 2007 WHO classification of tumours of the central nervous system.

      DNA and RNA Extraction

      Materials were selected for DNA extraction after careful examination of corresponding H&E-stained sections. All samples selected contained at least 80% of vital tumor. Extraction of DNA and RNA was performed as previously described.
      • Ichimura K.
      • Schmidt E.E.
      • Goike H.M.
      • Collins V.P.
      Human glioblastomas with no alterations of the CDKN2A (p16INK4A, MTS1) and CDK4 genes have frequent mutations of the retinoblastoma gene.
      For all tumor specimens, matched formalin-fixed, paraffin-embedded tumor tissues were available for immunostaining. All tissue samples were used in an anonymous manner, as approved by the local ethics committee at the University of Bonn Medical Center, Bonn, Germany.

      Tissue Culture

      Glioma cell lines A172, U373MG, T98G, U178, U87MG, and LN229 were obtained from the Ludwig Institute for Cancer Research (San Diego, CA), cultured as previously described,
      • Waha A.
      • Felsberg J.
      • Hartmann W.
      • von dem Knesebeck A.
      • Mikeska T.
      • Joos S.
      • Wolter M.
      • Koch A.
      • Yan P.S.
      • Endl E.
      • Wiestler O.D.
      • Reifenberger G.
      • Pietsch T.
      • Waha A.
      Epigenetic downregulation of mitogen-activated protein kinase phosphatase MKP-2 relieves its growth suppressive activity in glioma cells.
      and the identity was confirmed by short tandem repeat DNA profiling of 15 loci plus sex-determining marker, amelogenin (Genetica DNA Laboratories, Inc., Cincinnati, OH).

      IHC Data

      Paraffin-embedded tissue specimens, divided into serial sections (4 μm thick), on positively charged slides were routinely processed, air dried overnight at 37°C, deparaffinized in xylene, and rehydrated in a graded alcohol sequence. Antigen retrieval was performed by microwaving the slides in 10 mmol/L citrate buffer (pH 6,0). The slides were incubated in 3% hydrogen peroxide for 5 minutes at room temperature to block endogenous peroxidase activity and incubated in blocking solution (CSA II Kit; Dako, Glostrup, Denmark) for 90 minutes at room temperature. Slides were incubated overnight with primary antibodies [5mC, 1:8000, mouse monoclonal antibody (Eurogentec, Seraing, Belgium); 5hmC, 1:5000, rabbit anti-5hmC polyclonal antibody (Zymo Research, Irvine, CA); and TET1, 1:1000, rabbit anti-TET1 polyclonal antibody (Sigma-Aldrich, Munich, Germany)]. Slides were rinsed with Tris-buffered saline with Tween, and bound antibody was detected using the CSA system embodies technology (Dako CSA II, Biotin-Free Catalyzed Amplification System; Dako, Hamburg, Germany) and visualized by 3,3-diaminobenzidine tetrahydrochloride. The samples were counterstained with hematoxylin. The samples were dehydrated in a graded alcohol sequence and mounted in Richard-Allan Scientific Cytoseal XYL (Thermo Scientific, Waltham, MA). Immunohistochemical (IHC) staining with antibody against R132H mutant form of IDH1 protein (Dianova, Hamburg, Germany) at 1:100 dilutions was performed on a Ventana Automated Staining System (Roche Ventana, Darmstadt, Germany).
      Glioblastoma cell lines A172, U178, LN229, T98G, U373MG, and U87MG were trypsinized; washed with PBS; and fixed in 4% formalin for 24 hours at 4°C. Cell pellets were embedded in paraffin and divided into sections (4 μm thick). Immunostaining with 5mC, 5hmC, and TET1 was performed as previously described.

      IDH1/2 Mutation Analysis by Pyrosequencing

      Somatic sequence alterations of the codon R132 (IDH1) and R172 (IDH2) were investigated by pyrosequencing, as recently described.
      • Setty P.
      • Hammes J.
      • Rothämel T.
      • Vladimirova V.
      • Kramm C.M.
      • Pietsch T.
      • Waha A.
      A pyrosequencing-based assay for the rapid detection of IDH1 mutations in clinical samples.
      The IDH1 PCR amplification primer flanking the R132 mutation hot spot within exon 4 of IDH1 was as follows: 5′-CACCATACGAAATATTCTGG-3′ (forward), which amplifies 135 bp of genomic DNA. Similarly, an 87-bp fragment from exon 4 of IDH2 containing the R172 coding region was amplified using the following primer set: IDH2, 5′-AAACATCCCACGCCTAGTCC-3′ (forward) and 5′-biotin-TCTCCACCCTGGCCTACCT-3′ (reverse). For the pyrosequencing reaction, single-stranded DNA templates were immobilized on streptavidin-coated Sepharose high-performance beads (GE Healthcare, Uppsala, Sweden) using the PSQ Vacuum Prep Tool and Vacuum Prep Worktable (Biotage, Uppsala), according to the manufacturer's instructions, then incubated at 80°C for 2 minutes and allowed to anneal to 0.4 mmol/L sequencing primer (IDH1-Py, 5′-GTGAGTGGATGGGTAAAACC-3′; and IDH2-Py, 5′-AGCCCATCACCATTG-3′) at room temperature. Pyrosequencing was performed using PyroGold Reagents (Biotage) on the Pyromark Q24 instrument (Biotage), according to the manufacturer's instructions. Pyrogram outputs were analyzed by the PyroMark Q24 software version 1.0.10 (Biotage) using the allele quantification software to determine the percentage of mutant versus wild-type alleles, according to percentage relative peak height. A sequential nucleotide dispensation protocol was used, reflecting the expected order of nucleotide incorporation and the potential base change within the first or second positions of the hot spot codons of IDH1 and IDH2. Peak heights are proportional to the number of nucleotides that are incorporated with each dispensation.

      MGMT and DUSP4 Methylation Analysis

      Genomic DNA, 0.5 μg, was treated with sodium bisulfite using the EpiTect Bisulfite kit (Qiagen, Hilden, Germany), according to the manufacturer's recommendations. The methylation status of the MGMT gene was determined by pyrosequencing, as previously described.
      • Mikeska T.
      • Bock C.
      • El-Maarri O.
      • Hübner A.
      • Ehrentraut D.
      • Schramm J.
      • Felsberg J.
      • Kahl P.
      • Büttner R.
      • Pietsch T.
      • Waha A.
      Optimization of quantitative MGMT promoter methylation analysis using pyrosequencing and combined bisulfite restriction analysis.
      For DUSP4/MKP2, a pyrosequencing assay was developed that targets six CpG positions in a CpG island 5′ of the DUSP4/MKP2 gene.
      • Waha A.
      • Felsberg J.
      • Hartmann W.
      • von dem Knesebeck A.
      • Mikeska T.
      • Joos S.
      • Wolter M.
      • Koch A.
      • Yan P.S.
      • Endl E.
      • Wiestler O.D.
      • Reifenberger G.
      • Pietsch T.
      • Waha A.
      Epigenetic downregulation of mitogen-activated protein kinase phosphatase MKP-2 relieves its growth suppressive activity in glioma cells.
      Primer sequences were as follows: DUSP4, 5′-GTTTTAGGGGTTTAGATAGTTTAAGGTTGA-3′ (forward) and 5′-TCCTCCCCCTAAATTTAACTATTAT-3′ (reverse), yielding a PCR product of 348 bp. The reverse primer was biotinylated at the 5′ position. The sequencing primer was DUSP4-p, 5′-GGTTTAGATAGTTTAAGGTTGAT-3′.

      TET1, TET2, and TET3 mRNA Expression Analysis by Real-Time RT-PCR

      The expression of the three human TET genes (TET1, TET2, and TET3) was investigated by real-time RT-PCR using the ABI PRISM 5700 sequence detection system (Applied Biosystems, Darmstadt). cDNA synthesis was performed with 1 μg of total RNA using random hexamers and SuperScript II Reverse Transcriptase (Invitrogen, Darmstadt), according to the instructions of the manufacturer. Transcript levels of TET genes were normalized to the transcript level of ARF1 (ADP-ribosylation factor 1, GenBank accession number M36340). Primer sequences were as follows: TET1-rt, 5′-CAAGTGTTGCTGCTGTCAGG-3′ (forward) and 5′-AATTGGACACCCATGAGAGC-3′ (reverse); TET2-rt, 5′-CCAATAGGACATGATCCAGG-3′ (forward) and 5′-TCTGGATGAGCTCTCTCAGG-3′ (reverse); and TET3-rt, 5′-TTCAGAAGGAGAAGCTGAGC-3′ (forward) and 5′-TGTTCATGCTGTAAGGGTCG-3′ (reverse), yielding PCR products of 122, 232, and 240 bp, respectively. The primers for ARF1, 5′-GACCACGATCCTCTACAAGC-3′ (forward) and 5′-TCCCACACAGTGAAGCTGATG-3′ (reverse), amplify a cDNA fragment of 111 bp. Human brain RNA samples, from corpus callosum and temporal, occipital, frontal, and parietal lobes (BioChain Institute Inc., Hayward, CA), were used as references.

      Statistical Analysis

      All statistical analyses, including Fisher's exact test and the survival analyses, were performed with PRISM software version 4.0c (GraphPad, La Jolla, CA). Clinical and molecular parameters were correlated with survival data available for 50 patients with glioblastoma. Overall survival time was defined as the time between surgery for the primary tumor and death of the patient. Patients alive at their last follow-up were censored. Survival time was estimated by Kaplan-Meier survival curves and compared among patient subsets using log-rank tests.

      Results

      IHC Analysis of 5mC and 5hmC

      The IHC staining for 5mC and 5hmC was performed on 57 glioma samples, including 36 primary glioblastomas, 11 secondary glioblastomas, and 10 anaplastic astrocytomas. Of these 57 investigated tumor samples, 53 (93%) showed even and strong nuclear staining of 5mC [representative staining results of two primary glioblastomas (T2010 and T2486), one secondary glioblastoma (T2727), and one anaplastic astrocytoma (T2526) are shown in Figure 1, A and B]. Only four primary glioblastomas exhibited a significantly reduced number of stained cells, suggesting a global loss of this epigenetic mark in the nuclei of a large fraction of cells. There was no obvious difference in staining intensity between anaplastic astrocytoma, secondary glioblastoma, and primary glioblastoma sections. The IHC analysis of 5hmC showed remarkable differences between the investigated tumor tissues. Of the 57 analyzed tumor samples, 35 (61%) did not show any immunoreactivity for 5hmC, including 4 (40%) of 10 anaplastic astrocytomas, 8 (73%) of 11 secondary glioblastomas, and 23 (64%) of 36 primary glioblastomas. The intensity of the nuclear staining and the percentage of stained nuclei were significantly reduced to 20% to 50% in 12 (21%) of 57 tumor tissues [8 (22%) of 36 primary glioblastomas, 2 (18%) of 11 secondary glioblastomas, and 2 (20%) of 10 anaplastic astrocytomas]. Representative staining results are given in Figure 1, A and B.
      Figure thumbnail gr1
      Figure 1Immunostaining of IDH1, 5mC, and 5hmC in glioma samples. Staining of IDH1, 5mC, and 5hmC in two primary glioblastomas (T2010 and T2486) with wild-type (A) and IDH R132H mutant gliomas (T2727 secondary glioblastoma and T2526 anaplastic astrocytoma) (B). Insets in B: Staining of IDH1 is strongly nuclear.
      Glioblastoma cell lines A172, U178, T98G, LN229, U373MG, and U87MG demonstrated positive immunoreactivity for 5mC but did not contain detectable levels of 5hmC in the nuclei. Cortical and subcortical brain tissue was used as a control. All neurons in the cortex and glial cells in the white matter showed a nuclear positivity with 5mC and 5hmC antibodies. Only rare cells in the white matter appeared not stained with the 5hmC antibody (data not shown).

      Real-Time Analysis of TET1, TET2, and TET3 Transcripts in Gliomas and Glioblastoma Cell Lines

      Real-time RT-PCR analyses were performed to investigate the mRNA expression levels of TET1, TET2, and TET3 in a cohort of 54 glioma samples, of which RNA was available. Compared with five normal white matter tissues, TET1 mRNA expression was increased to at least twofold in 8 (24%) of 33 primary glioblastomas, 5 (45%) of 11 secondary glioblastomas, and 9 (90%) of 10 anaplastic astrocytomas. A total of 16 (48%) of 33 primary glioblastomas, 5 (45%) of 11 secondary glioblastomas, and 9 (90%) of 10 anaplastic astrocytomas showed increased transcription levels of at least twofold for TET2, compared with normal white matter tissues. The highest transcript levels for TET1 and TET2 were observed in anaplastic astrocytomas and secondary glioblastomas, with individual samples showing up to 7- and 12-fold increased transcript levels of TET1 and TET2, respectively (see Supplemental Figure S1, A and B, at http://ajp.amjpathol.org). In 4 (36%) of 11 secondary glioblastomas and 19 (58%) of 33 primary glioblastomas, transcription of TET3 was reduced at least twofold (see Supplemental Figure S1C at http://ajp.amjpathol.org). Transcript levels of TET1, TET2, and TET3 were reduced at least twofold in four of six glioblastoma cell lines (TET1: U178, U87MG, T98G, and A172; TET2: T98G, U178, U373MG, and A172; TET3: U178, U373MG, U87MG, and A172; see Supplemental Figure S1, A–C, at http://ajp.amjpathol.org).

      IHC Detection of TET1 in Gliomas and Glioblastoma Cell Lines

      Because TET1 was relevant for the oxidation of 5mC to 5hmC in several cell systems and our real-time PCR expression analysis showed remarkable changes in the abundance of TET1 transcripts, we performed IHC analysis for TET1 in 54 gliomas to assess the protein expression and its subcellular localization. A total of 6 (11%) of 54 investigated tumor tissues [5 (15%) of 34 primary glioblastomas and 1 (10%) of 10 secondary glioblastomas] showed no immunoreactivity for TET1 (eg, primary glioblastoma T2735 illustrated in Figure 2). In 7 (13%) of 54 tumor samples, the protein expression was restricted to the cytoplasm without staining of the nuclei [6 (18%) of 34 primary glioblastomas and 1 (10%) of 10 secondary glioblastomas; representative data are demonstrated for the primary glioblastoma T328 in Figure 2]. In 17 (31%) of 54 cases with predominant cytoplasmic staining, areas with moderate nuclear expression were also detected. These comprise 15 (44%) of 34 primary glioblastomas and 2 (20%) of 10 secondary glioblastomas. A strong nuclear accumulation of the protein was seen in 12 (22%) of 54 investigated tumor specimens, including 7 (70%) of 10 anaplastic astrocytomas, 3 (30%) of 10 secondary glioblastomas, and 2 (6%) of 34 primary glioblastomas. Representative staining of primary glioblastoma T2887 is shown in Figure 2. A total of 12 (22%) of 54 tumor sections demonstrated predominant nuclear accumulation in the presence of regions with moderate cytoplasmic staining [6 (18%) of 34 primary glioblastomas, 3 (30%) of 10 secondary glioblastomas, and 3 (30%) of 10 anaplastic astrocytomas]. Our data showed more frequent cytoplasmatic expression of TET1 in primary glioblastomas and a significant nuclear accumulation of the protein in secondary glioblastomas and anaplastic astrocytomas (Fisher's exact test, P < 0.0001). Interestingly, the loss of 5hmC was significantly more frequent in gliomas showing cytoplasmic or no expression of TET1 (Fisher‘s exact test, P = 0.0042). Among the tumors with TET1 nuclear expression, only a few cases [2 (18%) of 11 cases] had no significant detectable nuclear staining for 5hmC. All six investigated glioblastoma cell lines showed nuclear exclusion of TET1 protein, resembling the situation in tissues of primary glioblastomas (see Supplemental Figure S2, A and B, at http://ajp.amjpathol.org). The TET2 protein that was also involved in the oxidation of 5mC was located in the nucleus in all investigated glioma samples and glioblastoma cell lines (data not shown), arguing for a particular role of TET1 in the determination of the efficiency of 5mC conversion in these tumor specimens. The results of the IHC analyses and clinical information are listed in Supplemental Table S1 (available at http://ajp.amjpathol.org).
      Figure thumbnail gr2
      Figure 2The IHC detection of TET1 in glioma sections. Protein expression of TET1 in three primary glioblastoma samples: T2735, T328, and T2887. In tumor T328, the protein is exclusively expressed in the cytoplasm. Insets: Tumor T2887 shows a prominent nuclear accumulation of TET1, with only weak staining of the cytosol.
      When comparing transcript levels of TET1 with cellular localization, our data showed a significant association between tumors with the highest content of TET1 transcripts (more than threefold) and nuclear or predominantly nuclear expression of the protein (Fisher's exact test, P = 0.0004; Figure 3).
      Figure thumbnail gr3
      Figure 3Real-time expression analysis and subcellular localization of TET1. Relative transcript levels of TET1/ARF1 in the gliomas with regard to the subcellular localization of the protein: C, exclusively cytoplasmic; C > N, predominantly cytoplasmic; N, exclusively nuclear; N > C, predominantly nuclear.

      IHC and Molecular Analysis of IDH1 and IDH2 Mutations

      Because mutant IDH1 was associated with reduced oxidation of 5mC to 5hmC, we determined the mutation status of IDH1 and IDH2 in our patient cohort. For the detection of the most frequent mutant variant of IDH1 (R132H), we used an antibody that specifically binds to the mutant epitope in sections of formalin-fixed, paraffin-embedded tissues. In addition, a pyrosequencing assay was applied, allowing the detection of all possible mutations at codon 132 for IDH1 and codon 172 for IDH2. A total of 18 mutations (30%) were detected in 60 of the investigated tumor samples (39 primary glioblastomas, 11 secondary glioblastomas, and 10 anaplastic astrocytomas). All R132H mutations (n = 16) were detected with the pyrosequencing assay and by IHC. In addition, two rare mutations [CGT → TGT (R132C) and CGT → CTT (R132L)] were detected in an anaplastic astrocytoma (T1329) and a secondary glioblastoma (T3546), respectively. Representative data of the mutation analysis by pyrosequencing are presented in Figure 4. The IHC detection of the R132H mutant IDH1 protein in tumor sections is shown in Figure 1. The pyrosequencing analysis of codon R172 of IDH2 did not show any mutation in our tumor samples.
      Figure thumbnail gr4
      Figure 4Mutation analysis of IDH1 by pyrosequencing. Representative data of the IDH1 mutation analyses by pyrosequencing. A: The peak pattern of the first base of codon 132 for a tumor sample (T625) with a wild-type IDH1 gene. B: A significant increase of the thymine signal with a concomitant reduction of the cytosine peak compatible with a CGT -> TGT (R132C) mutation (sample T1329). C and D: The mutations detected in the second base of codon 132. C: A significant increase of the thymine peak with a concomitant reduction of the guanine signal compatible with a CGT -> CTT (R132L) mutation (sample T3546). D: The most frequent CGT -> CAT (R132H) mutation (sample T1214).
      A total of 7 (39%) of 18 IDH1-mutant glioma samples revealed strong positive nuclear staining for 5hmC, whereas 27 (69%) of the 39 gliomas with wild-type IDH1 did not show nuclear binding of the antibody against 5hmC. The comparison of 5hmC with IDH1 status did not reveal any significant values (Fisher's exact test, P = 0.56). Most interestingly, we found that gliomas with IDH1 mutations showed a strong nuclear accumulation of TET1 (13 of 16), whereas only 11 of 38 gliomas with wild-type IDH1 showed nuclear staining with the antibody against TET1 (Fisher's exact test, P = 0.0007).

      Methylation Analysis of MGMT and DUSP4/MKP2 by Pyrosequencing

      To investigate the relationship between TET1 localization and DNA methylation, we selected a CpG island 5′ of the DUSP4/MKP2 gene, the methylation of which we identified earlier in gliomas containing mutant IDH1 alleles. In addition, the methylation status of MGMT was determined, which is a well-established epigenetic marker for the estimation of response to chemotherapy with the alkylating drug, temozolomide (see Supplemental Figure S3, A and B, at http://ajp.amjpathol.org). The methylation of DUSP4/MKP2 was detected in 23 (38%) of the 60 investigated tumor specimens [10 (100%) of 10 anaplastic astrocytomas, 8 (73%) of 11 secondary glioblastomas, and 5 (13%) of 39 primary glioblastomas] and correlated significantly with the mutation status of IDH1 (Fisher's exact test, P < 0.0001). In addition, methylation of this gene was significantly more abundant in gliomas showing nuclear accumulation of TET1 (Fisher's exact test, P = 0.00030). The correlation of DUSP4/MKP2 methylation with the presence or absence of 5hmC did not show any significant correlation. The methylation of MGMT was observed in 34 (57%) of the 60 investigated glioma samples [8 (80%) of 10 anaplastic astrocytomas, 8 (73%) of 11 secondary glioblastomas, and 18 (46%) of 39 primary glioblastomas]. The correlation of MGMT methylation with IDH1 mutation status was highly significant (Fisher's exact test, P = 0.0014). MGMT methylation was significantly more frequent in gliomas showing nuclear TET1 expression (Fisher's exact test, P = 0.0275). The presence or absence of 5hmC from tumor nuclei did not correlate with MGMT methylation status.

      Relationship Between Molecular Findings and Clinical Data

      Comparison of the 5hmC content with survival did not show a statistically significant correlation (P = 0.0613); however, a trend toward shorter survival for patients with 5hmC-positive glioblastomas was noted. The average survival was 240 days for patients with 5hmC-positive tumors and 480 days for those with 5hmC-negative tumors. A possible explanation for not reaching statistical significance might be the few 5hmC-positive glioblastomas.

      Discussion

      The recent discovery of mutations in IDH1 and IDH2 genes in glioblastomas added an entirely new direction to the field of glioma research.
      • Parsons D.W.
      • Jones S.
      • Zhang X.
      • Lin J.C.-H.
      • Leary R.J.
      • Angenendt P.
      • Mankoo P.
      • Carter H.
      • Siu I.-M.
      • Gallia G.L.
      • Olivi A.
      • McLendon R.
      • Rasheed B.A.
      • Keir S.
      • Nikolskaya T.
      • Nikolsky Y.
      • Busam D.A.
      • Tekleab H.
      • Diaz L.A.
      • Hartigan J.
      • Smith D.R.
      • Strausberg R.L.
      • Marie S.K.N.
      • Shinjo S.M.O.
      • Yan H.
      • Riggins G.J.
      • Bigner D.D.
      • Karchin R.
      • Papadopoulos N.
      • Parmigiani G.
      • Vogelstein B.
      • Velculescu V.E.
      • Kinzler K.W.
      An integrated genomic analysis of human glioblastoma multiforme.
      • Yan H.
      • Parsons D.W.
      • Jin G.
      • McLendon R.
      • Rasheed B.A.
      • Yuan W.
      • Kos I.
      • Batinic-Haberle I.
      • Jones S.
      • Riggins G.J.
      • Friedman H.
      • Friedman A.
      • Reardon D.
      • Herndon J.
      • Kinzler K.W.
      • Velculescu V.E.
      • Vogelstein B.
      • Bigner D.D.
      IDH1 and IDH2 mutations in gliomas.
      A comprehensive study performed by the human Cancer Genome Atlas project indicates that IDH1 mutation status is associated with increased DNA methylation at various positions.
      • Noushmehr H.
      • Weisenberger D.J.
      • Diefes K.
      • Phillips H.S.
      • Pujara K.
      • Berman B.P.
      • Pan F.
      • Pelloski C.E.
      • Sulman E.P.
      • Bhat K.P.
      • Verhaak R.G.W.
      • Hoadley K.A.
      • Hayes D.N.
      • Perou C.M.
      • Schmidt H.K.
      • Ding L.
      • Wilson R.K.
      • Van Den Berg D.
      • Shen H.
      • Bengtsson H.
      • Neuvial P.
      • Cope L.M.
      • Buckley J.
      • Herman J.G.
      • Baylin S.B.
      • Laird P.W.
      • Aldape K.
      Cancer Genome Atlas Research Network
      Identification of a CpG island methylator phenotype that defines a distinct subgroup of glioma.
      The metabolite, 2-HG, that is produced by the mutant IDH1 protein has inhibited α-KG–dependent enzymes relevant in cancer.
      • Xu W.
      • Yang H.
      • Liu Y.
      • Yang Y.
      • Wang P.
      • Kim S.-H.
      • Ito S.
      • Yang C.
      • Wang P.
      • Xiao M.-T.
      • Liu L.-X.
      • Jiang W.-Q.
      • Liu J.
      • Zhang J.-Y.
      • Wang B.
      • Frye S.
      • Zhang Y.
      • Xu Y.-H.
      • Lei Q.-Y.
      • Guan K.-L.
      • Zhao S.-M.
      • Xiong Y.
      Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of α-ketoglutarate-dependent dioxygenases.
      In particular, dioxygenases that act on histones and on methylated DNA have gained interest because their inhibition would have a substantial effect on gene expression by altering the epigenetic shape of the tumor cells. 2-HG inhibits TET1, TET2, and TET3 hydroxymethylases that oxidize 5mC to 5hmC, leading to a reduction of 5hmC. The lack of conversion of 5mC into 5hmC might lead to an accumulation of methylated CpG islands. Recently, 2-HG–producing IDH1 or IDH2 alleles are sufficient to induce global DNA hypermethylation and reduced 5hmC in 293T cells.
      • Figueroa M.E.
      • Abdel-Wahab O.
      • Lu C.
      • Ward P.S.
      • Patel J.
      • Shih A.
      • Li Y.
      • Bhagwal N.
      • Vasanthakumar A.
      • Fernandez H.F.
      • Tallman M.S.
      • Sun Z.
      • Wolniak K.
      • Peeters J.K.
      • Liu W.
      • Choe S.E.
      • Fantin V.R.
      • Paietta E.
      • Löwenberg B.
      • Licht J.D.
      • Godley L.A.
      • Delwel R.
      • Valk P.J.
      • Thompson C.B.
      • Levine R.L.
      • Melnick A.
      Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and impair hematopoietic differentiation.
      Taken together, the data available argue for a scenario in which those tumors showing IDH1 mutations may inactivate TET enzymes, rendering them inactive for the conversion of 5mC to 5hmC.
      Herein, we localized these epigenetic marks in the nuclei in glioma tissues and determined the relation to IDH1 status and the expression of TET enzymes. The IHC analysis of 5hmC in our glioma cohort showed that a loss of this epigenetic mark is a frequent molecular event and is not restricted to gliomas with a mutant IDH1 gene. Within the group of IDH1-mutant gliomas, 41% of the samples revealed strong positive nuclear staining for 5hmC, indicating that the molecular mechanisms involved in the oxidation of 5mC to 5hmC are still functional. On the other hand, 68% of gliomas with wild-type IDH1 do not contain detectable levels of 5hmC, suggesting that there must be additional mechanisms distinct from IDH1 mutations that hamper the conversion of 5mC to 5hmC. Staining of 5mC was significantly reduced in four primary glioblastomas that are wild type for IDH1, indicating the occurrence of hypomethylation in these nuclei. Because 5mC is the substrate for the production of 5hmC by TET enzymes, its global loss from glioma cells might be an independent cause for the lack of 5hmC in respective nuclei. The methylation analysis of MGMT and DUSP4/MKP2 revealed a significant correlation between hypermethylation and mutant IDH1 status and hypermethylation and nuclear localization of TET1. For DUSP4/MKP2 hypermethylation was restricted to gliomas with mutant IDH1 and does not occur in wild-type tumors. The investigation of 5hmC in nuclei of established glioblastoma cell lines did not show any reactivity for this epigenetic mark in the tumor cells. This indicates alternative inhibitory mechanisms for the 5mC oxidation to be present in gliomas because all glioma cell lines do not show mutant IDH1 alleles. Also, hypermethylation of DUSP4/MKP2 was identified in all established glioblastoma cell lines that do not contain mutant IDH1 alleles, confirming our previous observation
      • Waha A.
      • Felsberg J.
      • Hartmann W.
      • von dem Knesebeck A.
      • Mikeska T.
      • Joos S.
      • Wolter M.
      • Koch A.
      • Yan P.S.
      • Endl E.
      • Wiestler O.D.
      • Reifenberger G.
      • Pietsch T.
      • Waha A.
      Epigenetic downregulation of mitogen-activated protein kinase phosphatase MKP-2 relieves its growth suppressive activity in glioma cells.
      and suggesting that target genes whose methylation is caused by mechanisms associated with IDH mutations in primary gliomas are also prone for hypermethylation in cell culture, independent of IDH1 mutations. The IHC data of TET1 expression show a remarkable pattern of subcellular localization. Nuclear accumulation of TET1 was exclusively found in gliomas with an IDH1 mutation (P = 0.0007) and more frequently in secondary glioblastomas and anaplastic astrocytomas, compared with primary glioblastomas (P < 0.0001). It is tempting to hypothesize that, in primary glioma tissues, the oncometabolite, 2-HG, directly or indirectly influences the subcellular localization, because it is specifically abundant in IDH1-mutant cells and is bound by TET enzymes.
      Strikingly, most gliomas with wild-type IDH1 status show nuclear exclusion of the protein in most of the tumor cells. The nuclear exclusion of TET1 is significantly associated with the loss of 5hmC from respective tumor cells (P = 0.0122). If the oxidation of 5mC to 5hmC is considered a nuclear event, this suggests nuclear exclusion of TET1 to lead to a reduction of 5hmC in the tumor cells. Consistent with this hypothesis, only one case lacking TET1 expression shows a high content of 5hmC in the tumor cells.
      A hypermethylation phenotype was described in gliomas with mutant IDH1 alleles, and these tumors correspond to the proneural subgroup of glioblastomas with a high prevalence among lower-grade gliomas (low-grade diffuse astrocytomas and anaplastic astrocytomas) and secondary glioblastomas.
      • Noushmehr H.
      • Weisenberger D.J.
      • Diefes K.
      • Phillips H.S.
      • Pujara K.
      • Berman B.P.
      • Pan F.
      • Pelloski C.E.
      • Sulman E.P.
      • Bhat K.P.
      • Verhaak R.G.W.
      • Hoadley K.A.
      • Hayes D.N.
      • Perou C.M.
      • Schmidt H.K.
      • Ding L.
      • Wilson R.K.
      • Van Den Berg D.
      • Shen H.
      • Bengtsson H.
      • Neuvial P.
      • Cope L.M.
      • Buckley J.
      • Herman J.G.
      • Baylin S.B.
      • Laird P.W.
      • Aldape K.
      Cancer Genome Atlas Research Network
      Identification of a CpG island methylator phenotype that defines a distinct subgroup of glioma.
      Xu et al
      • Xu W.
      • Yang H.
      • Liu Y.
      • Yang Y.
      • Wang P.
      • Kim S.-H.
      • Ito S.
      • Yang C.
      • Wang P.
      • Xiao M.-T.
      • Liu L.-X.
      • Jiang W.-Q.
      • Liu J.
      • Zhang J.-Y.
      • Wang B.
      • Frye S.
      • Zhang Y.
      • Xu Y.-H.
      • Lei Q.-Y.
      • Guan K.-L.
      • Zhao S.-M.
      • Xiong Y.
      Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of α-ketoglutarate-dependent dioxygenases.
      reported that 2-HG is a competitive inhibitor of α-KG–dependent dioxygenases, including histone demethylases and the TET family of 5mC hydroxylases. They demonstrate that ectopic expression of wild-type IDH1 and TET1 and TET2 in glioblastoma cells leads to an increase, whereas cotransfection of R132H-mutant IDH1 and TET1 and TET2 results in a dramatic decrease, of 5hmC. These elegant studies argue for a direct association between the loss of 5hmC and the presence of IDH1 mutation in gliomas. Our IHC data show exclusive cytoplasmic expression of TET1 in all glioma cell lines. The authors reported that all 20 glioma samples that they investigated showed 5hmC, with gliomas harboring IDH1 mutations containing significantly lower 5hmC levels. In our cohort of gliomas, 61% of the cases did not exhibit 5hmC and 33% of IDH1-mutant gliomas showed strong nuclear staining of 5hmC. Recent data support our findings that there is no significant correlation between IDH1 status and 5hmC content.
      • Jin S.-G.
      • Jiang Y.
      • Qiu R.
      • Rauch T.A.
      • Wang Y.
      • Schackert G.
      • Krex D.
      • Lu Q.
      • Pfeifer G.P.
      5-Hydroxymethylcytosine is strongly depleted in human cancers but its levels do not correlate with IDH1 mutations.
      Promoter methylation of TET2 was recently found in low-grade diffuse astrocytomas, suggesting that this might be an alternative pathomechanism in low-grade diffuse astrocytomas with wild-type IDH1/2.
      • Kim Y.-H.
      • Pierscianek D.
      • Mittelbronn M.
      • Vital A.
      • Mariani L.
      • Hasselblatt M.
      • Ohgaki H.
      TET2 promoter methylation in low-grade diffuse gliomas lacking IDH1/2 mutations.
      Interestingly, five primary glioblastomas and one secondary glioblastoma investigated in our study did not show expression of TET1, indicating that inactivation of gene expression might also be involved in the regulation of TET1 activity in gliomas.
      The high incidence of loss of 5hmC from nuclei of gliomas independent from their IDH1 status argues for the presence of alternative mechanisms that regulate the conversion of 5mC into 5hmC. Furthermore, it argues for a general contribution of this modification to the molecular pathological characteristics of this tumor entity. Thus far, there is little information on the biological consequences of 5hmC loss from the genome of cancer cells. Proteins binding 5mC do not bind stably to 5hmC and dissociate from DNA, presumably altering the epigenetic state of respective chromatin domains.
      • Jin S.G.
      • Kadam S.
      • Pfeifer G.P.
      Examination of the specificity of DNA methylation profiling techniques towards 5-methylcytosine and 5-hydroxymethylcytosine.
      In addition, several lines of evidence suggest that the oxidation of 5mC to 5hmC leads to demethylation by passive mechanisms in proliferating cells (eg, because it is not recognized by DNMT1 after DNA replication
      • Valinluck V.
      • Sowers L.C.
      Endogenous cytosine damage products alter the site selectivity of human DNA maintenance methyltransferase DNMT1.
      or by active mechanisms in nondividing cells involving DNA repair mechanisms).
      • Guo J.U.
      • Su Y.
      • Zhong C.
      • Ming G.-L.
      • Song H.
      Emerging roles of TET proteins and 5-hydroxymethylcytosines in active DNA demethylation and beyond.
      These demethylation processes may remove false methylation events and serve to reinforce accurate DNA methylation patterns.
      • Williams K.
      • Christensen J.
      • Pedersen M.T.
      • Johansen J.V.
      • Cloos P.A.
      • Rappsilber J.
      • Helin K.
      TET1 and hydroxymethylcytosine in transcription and DNA methylation fidelity.
      The inactivation of this mechanism might lead to hypermethylation events in cancer cells that accumulate in tumors if they are associated with a growth advantage. The observations that 5hmC-negative tumor cells show enhanced staining of Ki-67 argue for an implication in processes regulating cellular proliferation.
      • Jin S.-G.
      • Jiang Y.
      • Qiu R.
      • Rauch T.A.
      • Wang Y.
      • Schackert G.
      • Krex D.
      • Lu Q.
      • Pfeifer G.P.
      5-Hydroxymethylcytosine is strongly depleted in human cancers but its levels do not correlate with IDH1 mutations.
      The results of our IHC study of TET1 expression in anaplastic astrocytomas and glioblastomas showing a high incidence of nuclear exclusion of TET1 may provide an alternative route for the decrease in 5hmC from the tumor cells that is independent from IDH1 status (see Supplemental Figure S4, A–C, at http://ajp.amjpathol.org). The correlation between cytoplasmic localization and no expression of TET1 with absent or reduced 5hmC staining proves to be significant (P = 0.0122). Thus far, there is little information about post-translational modifications of TET enzymes that may be involved in the determination of their subcellular localization. TET1 has three nuclear localization signals, suggesting a mainly nuclear localization of the protein. The investigation of regulatory pathways influencing the activity or localization of TET enzymes is, therefore, an important task for future studies. In addition to the inability to oxidize 5mC and other nuclear targets, a prominent expression of TET1 in the cytoplasm might also lead to accelerated activities of the enzyme on proteins in the cytosol. Also, TET1 has repressed gene transcription, independent of its catalytic activity, by recruiting the Sin3A complex
      • Williams K.
      • Christensen J.
      • Pedersen M.T.
      • Johansen J.V.
      • Cloos P.A.
      • Rappsilber J.
      • Helin K.
      TET1 and hydroxymethylcytosine in transcription and DNA methylation fidelity.
      ; in addition, it has promoted binding of the polycomb-repressive complex 2,
      • Wu H.
      • Zhang Y.
      Tet1 and 5-hydroxymethylation: a genome-wide view in mouse embryonic stem cells.
      arguing for a complex role of TET1 protein activity in the transcriptional activation via demethylation of gene promoters and transcriptional repression.

      Supplementary data

      • Supplemental Figure S1

        Real-time PCR analysis of TET1-3 transcripts in gliomas. This figure shows the relative transcript levels of TET1 (A), TET2 (B), and TET3 (C) in different glioma entities, normal tissues, and glioblastoma cell lines. AAIII, anaplastic astrocytoma, grade 3; NB, normal brain; pGBM, primary glioblastoma multiforme; sGBM, secondary glioblastoma multiforme.

      • Supplemental Figure S2

        The IHC detection of TET1 in glioblastoma cell lines. Localization of the TET1 protein in the glioblastoma cell lines LN229 (A) and T98G (B). Original magnification, ×200.

      • Supplemental Figure S3

        Methylation analyses of DUSP4/MKP2 and MGMT. Representative pyrograms are shown for the detection of hypermethylation of DUSP4/MKP2 (A) and MGMT (B). Top panels: Pyrograms showing unmethylated tumor samples (T172 and T327). Bottom panels: Pyrograms showing hypermethylated tumor samples (T1214 and T2481).

      • Supplemental Figure S4

        Influence of IDH1 status on TET activity and 5mC/5hmC levels. This illustration summarizes the role of IDH1 in the regulation of TET activity. Wild-type (wt) IDH1 produces aKG, which is required for the activity of TET enzymes, and then oxidizes 5mC into 5hmC (A). Mutant IDH1 produces 2HG, which inactivates TET function (B). Activity on 5mC in the absence of IDH1 mutations (C).

      References

        • Kleihues P.
        • Louis D.N.
        • Scheithauer B.W.
        • Rorke L.B.
        • Reifenberger G.
        • Burger P.C.
        • Cavenee W.K.
        The WHO classification of tumors of the nervous system.
        J Neuropathol Exp Neurol. 2002; 61 (discussion 226–229): 215-225
        • Ohgaki H.
        • Kleihues P.
        Genetic pathways to primary and secondary glioblastoma.
        Am J Pathol. 2007; 170: 1445-1453
        • Parsons D.W.
        • Jones S.
        • Zhang X.
        • Lin J.C.-H.
        • Leary R.J.
        • Angenendt P.
        • Mankoo P.
        • Carter H.
        • Siu I.-M.
        • Gallia G.L.
        • Olivi A.
        • McLendon R.
        • Rasheed B.A.
        • Keir S.
        • Nikolskaya T.
        • Nikolsky Y.
        • Busam D.A.
        • Tekleab H.
        • Diaz L.A.
        • Hartigan J.
        • Smith D.R.
        • Strausberg R.L.
        • Marie S.K.N.
        • Shinjo S.M.O.
        • Yan H.
        • Riggins G.J.
        • Bigner D.D.
        • Karchin R.
        • Papadopoulos N.
        • Parmigiani G.
        • Vogelstein B.
        • Velculescu V.E.
        • Kinzler K.W.
        An integrated genomic analysis of human glioblastoma multiforme.
        Science. 2008; 321: 1807-1812
        • Yan H.
        • Parsons D.W.
        • Jin G.
        • McLendon R.
        • Rasheed B.A.
        • Yuan W.
        • Kos I.
        • Batinic-Haberle I.
        • Jones S.
        • Riggins G.J.
        • Friedman H.
        • Friedman A.
        • Reardon D.
        • Herndon J.
        • Kinzler K.W.
        • Velculescu V.E.
        • Vogelstein B.
        • Bigner D.D.
        IDH1 and IDH2 mutations in gliomas.
        N Engl J Med. 2009; 360: 765-773
        • Noushmehr H.
        • Weisenberger D.J.
        • Diefes K.
        • Phillips H.S.
        • Pujara K.
        • Berman B.P.
        • Pan F.
        • Pelloski C.E.
        • Sulman E.P.
        • Bhat K.P.
        • Verhaak R.G.W.
        • Hoadley K.A.
        • Hayes D.N.
        • Perou C.M.
        • Schmidt H.K.
        • Ding L.
        • Wilson R.K.
        • Van Den Berg D.
        • Shen H.
        • Bengtsson H.
        • Neuvial P.
        • Cope L.M.
        • Buckley J.
        • Herman J.G.
        • Baylin S.B.
        • Laird P.W.
        • Aldape K.
        • Cancer Genome Atlas Research Network
        Identification of a CpG island methylator phenotype that defines a distinct subgroup of glioma.
        Cancer Cell. 2010; 17: 510-522
        • Dang L.
        • White D.W.
        • Gross S.
        • Bennett B.D.
        • Bittinger M.A.
        • Driggers E.M.
        • Fantin V.R.
        • Jang H.G.
        • Jin S.
        • Keenan M.C.
        • Marks K.M.
        • Prins R.M.
        • Ward P.S.
        • Yen K.E.
        • Liau L.M.
        • Rabinowitz J.D.
        • Cantley L.C.
        • Thompson C.B.
        • Vander Heiden M.G.
        • Su S.M.
        Cancer-associated IDH1 mutations produce 2-hydroxyglutarate.
        Nature. 2009; 462: 739-744
        • Tahiliani M.
        • Koh K.P.
        • Shen Y.
        • Pastor W.A.
        • Bandukwala H.
        • Brudno Y.
        • Agarwal S.
        • Iyer L.M.
        • Liu D.R.
        • Aravind L.
        • Rao A.
        Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1.
        Science. 2009; 324: 930-935
        • Gu T.-P.
        • Guo F.
        • Yang H.
        • Wu H.-P.
        • Xu G.-F.
        • Liu W.
        • Xie Z.-G.
        • Shi L.
        • He X.
        • Jin S.-G.
        • Iqbal K.
        • Shi Y.G.
        • Deng Z.
        • Szabó P.E.
        • Pfeifer G.P.
        • Li J.
        • Xu G.-L.
        The role of Tet3 DNA dioxygenase in epigenetic reprogramming by oocytes.
        Nature. 2011; 477: 606-610
        • Iqbal K.
        • Jin S.-G.
        • Pfeifer G.P.
        • Szabó P.E.
        Reprogramming of the paternal genome upon fertilization involves genome-wide oxidation of 5-methylcytosine.
        Proc Natl Acad Sci U S A. 2011; 108: 3642-3647
        • Li W.
        • Liu M.
        Distribution of 5-hydroxymethylcytosine in different human tissues.
        J Nucleic Acids. 2011; 2011: 1-5
        • Kriaucionis S.
        • Heintz N.
        The nuclear DNA base 5-hydroxymethylcytosine is present in Purkinje neurons and the brain.
        Science. 2009; 324: 929-930
        • Pastor W.A.
        • Pape U.J.
        • Huang Y.
        • Henderson H.R.
        • Lister R.
        • Ko M.
        • McLoughlin E.M.
        • Brudno Y.
        • Mahapatra S.
        • Kapranov P.
        • Tahiliani M.
        • Daley G.Q.
        • Liu X.S.
        • Ecker J.R.
        • Milos P.M.
        • Agarwal S.
        • Rao A.
        Genome-wide mapping of 5-hydroxymethylcytosine in embryonic stem cells.
        Nature. 2011; 473: 394-397
        • Jin S.G.
        • Wu X.
        • Li A.X.
        • Pfeifer G.P.
        Genomic mapping of 5-hydroxymethylcytosine in the human brain.
        Nucleic Acids Res. 2011; 39: 5015-5024
        • Jones P.A.
        • Baylin S.B.
        The epigenomics of cancer.
        Cell. 2007; 128: 683-692
        • Williams K.
        • Christensen J.
        • Pedersen M.T.
        • Johansen J.V.
        • Cloos P.A.
        • Rappsilber J.
        • Helin K.
        TET1 and hydroxymethylcytosine in transcription and DNA methylation fidelity.
        Nature. 2011; 473: 343-348
        • Jin S.-G.
        • Jiang Y.
        • Qiu R.
        • Rauch T.A.
        • Wang Y.
        • Schackert G.
        • Krex D.
        • Lu Q.
        • Pfeifer G.P.
        5-Hydroxymethylcytosine is strongly depleted in human cancers but its levels do not correlate with IDH1 mutations.
        Cancer Res. 2011; 71: 7360-7365
        • Haffner M.C.
        • Chaux A.
        • Meeker A.K.
        • Esopi D.M.
        • Gerber J.
        • Pellakuru L.G.
        • Toubaji A.
        • Argani P.
        • Iacobuzio-Donahue C.
        • Nelson W.G.
        • Netto G.J.
        • De Marzo A.M.
        • Yegnasubramanian S.
        Global 5-hydroxymethylcytosine content is significantly reduced in tissue stem/progenitor cell compartments and in human cancers.
        Oncotarget. 2011; 2: 627-637
        • Louis D.N.
        • Ohgaki H.
        • Wiestler O.D.
        • Cavenee W.K.
        • Burger P.C.
        • Jouvet A.
        • Scheithauer B.W.
        • Kleihues P.
        The 2007 WHO classification of tumours of the central nervous system.
        Acta Neuropathol. 2007; 114: 97-109
        • Ichimura K.
        • Schmidt E.E.
        • Goike H.M.
        • Collins V.P.
        Human glioblastomas with no alterations of the CDKN2A (p16INK4A, MTS1) and CDK4 genes have frequent mutations of the retinoblastoma gene.
        Oncogene. 1996; 13: 1065-1072
        • Waha A.
        • Felsberg J.
        • Hartmann W.
        • von dem Knesebeck A.
        • Mikeska T.
        • Joos S.
        • Wolter M.
        • Koch A.
        • Yan P.S.
        • Endl E.
        • Wiestler O.D.
        • Reifenberger G.
        • Pietsch T.
        • Waha A.
        Epigenetic downregulation of mitogen-activated protein kinase phosphatase MKP-2 relieves its growth suppressive activity in glioma cells.
        Cancer Res. 2010; 70: 1689-1699
        • Setty P.
        • Hammes J.
        • Rothämel T.
        • Vladimirova V.
        • Kramm C.M.
        • Pietsch T.
        • Waha A.
        A pyrosequencing-based assay for the rapid detection of IDH1 mutations in clinical samples.
        J Mol Diagn. 2010; 12: 750-756
        • Mikeska T.
        • Bock C.
        • El-Maarri O.
        • Hübner A.
        • Ehrentraut D.
        • Schramm J.
        • Felsberg J.
        • Kahl P.
        • Büttner R.
        • Pietsch T.
        • Waha A.
        Optimization of quantitative MGMT promoter methylation analysis using pyrosequencing and combined bisulfite restriction analysis.
        J Mol Diagn. 2007; 9: 368-381
        • Xu W.
        • Yang H.
        • Liu Y.
        • Yang Y.
        • Wang P.
        • Kim S.-H.
        • Ito S.
        • Yang C.
        • Wang P.
        • Xiao M.-T.
        • Liu L.-X.
        • Jiang W.-Q.
        • Liu J.
        • Zhang J.-Y.
        • Wang B.
        • Frye S.
        • Zhang Y.
        • Xu Y.-H.
        • Lei Q.-Y.
        • Guan K.-L.
        • Zhao S.-M.
        • Xiong Y.
        Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of α-ketoglutarate-dependent dioxygenases.
        Cancer Cell. 2011; 19: 17-30
        • Figueroa M.E.
        • Abdel-Wahab O.
        • Lu C.
        • Ward P.S.
        • Patel J.
        • Shih A.
        • Li Y.
        • Bhagwal N.
        • Vasanthakumar A.
        • Fernandez H.F.
        • Tallman M.S.
        • Sun Z.
        • Wolniak K.
        • Peeters J.K.
        • Liu W.
        • Choe S.E.
        • Fantin V.R.
        • Paietta E.
        • Löwenberg B.
        • Licht J.D.
        • Godley L.A.
        • Delwel R.
        • Valk P.J.
        • Thompson C.B.
        • Levine R.L.
        • Melnick A.
        Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and impair hematopoietic differentiation.
        Cancer Cell. 2010; 18: 553-567
        • Kim Y.-H.
        • Pierscianek D.
        • Mittelbronn M.
        • Vital A.
        • Mariani L.
        • Hasselblatt M.
        • Ohgaki H.
        TET2 promoter methylation in low-grade diffuse gliomas lacking IDH1/2 mutations.
        J Clin Pathol. 2011; 64: 850-852
        • Jin S.G.
        • Kadam S.
        • Pfeifer G.P.
        Examination of the specificity of DNA methylation profiling techniques towards 5-methylcytosine and 5-hydroxymethylcytosine.
        Nucleic Acids Res. 2010; 38: e125
        • Valinluck V.
        • Sowers L.C.
        Endogenous cytosine damage products alter the site selectivity of human DNA maintenance methyltransferase DNMT1.
        Cancer Res. 2007; 67: 946-950
        • Guo J.U.
        • Su Y.
        • Zhong C.
        • Ming G.-L.
        • Song H.
        Emerging roles of TET proteins and 5-hydroxymethylcytosines in active DNA demethylation and beyond.
        Cell Cycle. 2011; 10: 2662-2668
        • Wu H.
        • Zhang Y.
        Tet1 and 5-hydroxymethylation: a genome-wide view in mouse embryonic stem cells.
        Cell Cycle. 2011; 10: 2428-2436