A methylation-specific and SYBR-green-based quantitative polymerase chain reaction technique for O6-methylguanine DNA methyltransferase promoter methylation analysis

Analytical Biochemistry 377 (2008) 62–71

Contents lists available at ScienceDirect

Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio

A methylation-specific and SYBR-green-based quantitative polymerase chain reaction technique for O6-methylguanine DNA methyltransferase promoter methylation analysis Kirsten Hattermann a, H. Maximilian Mehdorn b, Rolf Mentlein a, Susann Schultka b, Janka Held-Feindt b,* a b

Department of Anatomy, University of Kiel, Olshausenstr. 40, 24098 Kiel, Germany Department of Neurosurgery, University Medical Center Schleswig-Holstein, Campus Kiel, Schittenhelmstr. 10, 24105 Kiel, Germany

a r t i c l e

i n f o

Article history: Received 4 December 2007 Available online 14 March 2008 Keywords: Hypermethylation Glioblastoma Real-time PCR MGMT Promoter

a b s t r a c t The O6-methylguanine DNA methyltransferase (MGMT) gene encodes a DNA repair enzyme whose activity is a major mechanism of resistance to alkylating drugs in glioblastoma treatment. Hypermethylation of the MGMT promoter is associated with chemosensitivity because it reduces MGMT activity. Here we present a method combining methylation-specific and SYBR-green-based quantitative PCR (MSQP) for MGMT promoter methylation analysis. This highly specific, sensitive, and reproducible method allows the quantification of fully methylated and fully unmethylated MGMT DNA species in terms of percentage. Values are related to standard curves, corrected for DNA input by an internal standard, and calculated in relation to methylated and unmethylated control DNAs as a percentage share. Finally, values are defined relative to the sum of fully methylated and unmethylated MGMT DNA sample amount to obtain percentage of methylated reference and percentage of unmethylated reference results. We have used this technique to investigate MGMT promoter methylation in relation to MGMT mRNA expression in nine tumor cell lines and 15 primary glioblastoma patients. Presented data confirm that this assay is suitable for detection of low amounts of methylated and unmethylated MGMT promoter DNA. Carefully validated quantitative MSQP assays will be useful in both research and clinical molecular diagnosis. Ó 2008 Elsevier Inc. All rights reserved.

Carcinogenesis processes require sequential molecular events, including mutation of genes and overexpression of growth factors and their receptors [1]. Moreover, aberrant CpG island methylation is an early and frequent marker in carcinogenesis [2], and transcriptional inactivation of CpG-island-containing promoters of tumor suppressor genes by DNA hypermethylation has been well documented in many human cancers [3]. Some of these genes are CDKN2A/p16 [4], MLH1 [5], and MGMT1 [6]. The O6-methylguanine DNA methyltransferase gene encodes a DNA repair enzyme and is located on chromosome 10q26. It removes alkyl groups from the O6 position of guanine in DNA and transfers them to an active cysteine within its own sequence in a reaction that inactivates one MGMT molecule for each lesion repaired [7]. Thus, MGMT activity is a major mechanism of resistance to alkylating drugs [8]. The loss of the MGMT function has been described and is most frequently due to epigenetic changes, specifi-

* Corresponding author. Fax: +49 0 431 597 4918. E-mail address: [email protected] (J. Held-Feindt). 1 Abbreviations used: MGMT, O6-methylguanine DNA methyltransferase; MSP, methylation-specific PCR; MSQP, Methylation-specific quantitative PCR; PMR, percentage of methylated reference; PUR, percentage of unmethylated reference; GBM, glioblastoma. 0003-2697/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2008.03.014

cally promoter-region methylation [8,9]. Recently, this methylation and silencing of the MGMT gene itself has been associated with an increased benefit from temozolomide treatment in glioblastomas (GBM), an aggressive brain tumor with very poor survival rates [10–16]. Methylation-specific PCR (MSP) is the most widely used assay for detection of hypermethylation in CpG islands [6,17,18]. In brief, for a given DNA sample, treatment with a bisulfite reagent leads to the conversion of all unmethylated cytosine to uracil, leaving only methylated cytosine unchanged [19]. Then a selective PCR amplification of sequences corresponding to either fully unmethylated or methylated DNA using primers that anneal specifically with either one of the DNA species is performed. Fundamental investigations of the correlations of MGMT promoter methylation status, chemotherapy, and survival of glioblastoma patients [12,13,15] were done using this MSP technique yielding only qualitative results. In investigations focused on the quantitative analysis of MGMT promoter methylation status only the amount of fully methylated MGMT DNA species were determined [20–22]. Now, we have established a new method combining the original MSP technique with a SYBR-green-based quantitative PCR (methylation-specific quantitative PCR; MSQP). The MSQP technique is

63

Methylation-specific quantitative PCR / K. Hattermann et al. / Anal. Biochem. 377 (2008) 62–71

a highly specific, sensitive, and reproducible method that allows quantitative determination of fully unmethylated and methylated bisulfite-converted MGMT DNA species in parallel and in terms of percentage. In brief, values are calculated in relation to standard curves generated for each primer pair, corrected for input of DNA by an internal standard, and generated in relation to methylated and unmethylated control DNAs as a percentage share (results marked as original values). Finally, to make the results more usable in clinical practice, we defined the portions of fully methylated and unmethylated MGMT DNA to obtain percentage of methylated reference (PMR) and percentage of unmethylated reference (PUR) values (mMGMT + uMGMT = 100%). By achieving fully unmethylated and methylated portions of MGMT promoter in terms of percentage, we enable a more precise and more easily interpretable identification of the hypermethylation status. With this improved MSP technique we offer a molecular tool for performing individualadapted patient-based chemotherapy. Our initial application of this method to tumor cell lines and glioblastoma tumor samples suggests that this format may be suitable for clinical studies. Material and methods Patients and tumor specimens Glioblastoma tissues and clinical information were collected from patients who were operated on between August 1998 and November 2006 at the Department of Neurosurgery, Kiel, Germany (see Table 1). Patients fulfilling the following inclusion criteria were selected for this investigation: (1) histological diagnosis of primary GBM according to World Health Organization criteria performed by a neuropathologist, (2) detailed clinical data at diagnosis and therapy and during follow up, (3) treatment with temozolomide after surgery, and (4) availability of frozen tumor samples. Patients previously treated for low-grade glioma were excluded. The patients included 4 women and 11 men; the mean age at diagnosis was 55.1 ± 15.0 years (women 51.75 ± 15.5 years; men 56.4 ± 15.4 years). Normal brain tissue obtained from the Department of General Pathology, Kiel, Germany served as normal control sample. According to neurosurgical requirements and by using highquality-evidence-based operation procedures, all GBM tissues were extirpated without removal of healthy tissue in the best reproducible and accurate way. Additionally, to demonstrate tumor specimen homogeneity, H&E stainings of all GBM tissue samples were routinely surveyed by a neuropathologist. However, due to tumor inhomogeneity per se, contamination with small amounts of normal cells cannot be absolutely excluded.

All samples were obtained in accordance with approved ethical standards of the responsible committee of the University of Kiel and have therefore been performed in accordance with the ethical standards as formulated in the Helsinki Declaration of 1975 (revised 1983). From all samples we could obtain DNA and RNA except for sample 3, where RNA isolation was impossible due to small sample amount (Table 1). Cells, cell lines, and cell culture Human lymphocytes were isolated from human peripheral blood by Ficoll density gradient centrifugation using lymphocyte separation medium LSM 1077 (PAA Laboratories, Cölbe, Germany). Human glioma cell lines (U118, U343, A172, T98G), human breast cancer cell lines (MDA-MB 231, TD47-D, MCF-7, BT-549), and a human colon adenocarcinoma cell line (SW480) were obtained from the European Collection of Cell Cultures ECACC, Salisbury, UK. All cell lines were cultured in glutamine-supplemented Dulbecco’s modified Eagle’s medium (Gibco/Invitrogen, Karlsruhe, Germany) plus 10% fetal calf serum (Gibco/Invitrogen) as described previously [23]. Purity of cell cultures was demonstrated routinely by immunostaining for the cell-type-specific markers glial fibrillary acidic protein (antibody from Boehringer, Mannheim, Germany) and cytokeratin (antibody from DAKO, Glostrup, Denmark) [23,24]. Contamination by Mycoplasma was checked by Mycoplasma-specific PCR using Mycoplasma Detection Kit (VenorGeM; Minerva Biolabs GmbH, Berlin, Germany). Methylation-specific quantitative PCR DNA was isolated from cell lines and from glioblastoma tissue samples using the Qiagen DNA Mini Kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. Bisulfite conversion was performed with the EpiTect Bisulfite Kit (Qiagen) as described by the manufacturer. During this bisulfite conversion all unmethylated cytosine is converted to uracil while methylated cytosine within CpG sites remains unchanged. For each conversion reaction 1 lg DNA was employed, and after conversion and purification DNA was eluted in 20 ll of the provided elution buffer. For positive controls (100% values) and standard curves CG Genome Universal Unmethylated DNA (uDNA) (S7822; Vial A; Millipore, Schwalbach, Germany) and CG Genome Universal Methylated DNA (mDNA) (S7821; Millipore) were used. For methylation-specific quantitative PCR the QuantiTect SYBR green PCR Kit (Qiagen) and primers specific for fully methylated and fully unmethylated MGMT promoter sequences (mMGMT,

Table 1 Patients and tumor specimens Sample number

Gender

Age at first surgery (years)

Tumor localization

Outcome (death) (months)

MSQP

RT-PCR

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Normal brain tissue

Female Male Male Male Female Male Male Male Male Male Male Female Female Male Male

49 23 67 62 65 65 44 46 69 76 68 31 62 46 54

Left parietooccipital Right temporal Right temporal Right temporal Left temporal Left occipital Right temporal Right temporal Left temporal Left parietooccipital Right temporal Right frontal Right temporal Right parietal Left precentral

8.5 5.5 10 12 alive (53) 12 10 10 9.5 alive (14) 10 68 alive (48) 17.5 20

X X X X X X X X X X X X X X X X

X X n/a X X X X X X X X X X X X X

64

Methylation-specific quantitative PCR / K. Hattermann et al. / Anal. Biochem. 377 (2008) 62–71

uMGMT) [6,17] were taken. These primer pairs specifically bind to bisulfite-converted genomic DNA, which means that the uMGMT primers recognize cytosine within CpG sites which were converted to uracil, whereas the mMGMT primers bind to unconverted cytosine. For each 50-ll PCR, 4 ll eluate containing the bisulfite-converted DNA, 25 ll 2x QuantiTect SYBR Green PCR Master Mix, 1.5 ll each forward and reverse primer (concentrations from 10 to 100 pmol/ll resulting in final concentrations of 0.3 to 3 lM), and 18 ll DNase-free water were used. To correct cycles of threshold (CT) values for input of DNA, a primer pair corresponding to a specific b-actin sequence was chosen. Within this ß-actin sequence no CpG sites are present. Thus, the cytosines are always unmethylated, irrespective of methylation signals in other gene regions, e.g., the MGMT promoter, and therefore will always be converted to uracil after bisulfite treatment (ActB) [25]. As a control for efficiency of bisulfite conversion, primers for the same sequence of b-actin, but representing unconverted genomic DNA, were used (ActG). In contrast to ActB primers, the ActG primer pair recognizes the ß-actin sequence from genomic DNA and thus can detect insufficient bisulfite conversion. Primer sequences are as follows: mMGMT Forward 50 -TTTCG ACGTTCGTAGGTTTTCGC-30 , mMGMT Reverse 50 -GCACTCTTCCGA AAACGAAACG-30 , uMGMT Forward 50 -TTTGTGTTTTGATGTTTGT AGGTTTTTGT-30 , uMGMT Reverse 30 -AACTCCACACTCTTCCAAAAA CAAAACA-30 , ActB Forward 50 -TGGTGATGGAGGAGGTTTAGT AAGT-30 , ActB Reverse 50 -AACCAATAAAACCTACTCCTCCCTTAA-30 , ActG Forward 50 -TGGTGATGGAGGAGGCTCAGCAAGT-30 , and ActG Reverse 50 -AGCCAATGGGACCTGCTCCTCCCTTGA-30 . All primers were obtained from MWG-Biotech AG (Ebersberg, Germany). Real-time PCR conditions were 95 °C for 15 min followed by 45 cycles of 94 °C for 15 s, 60 °C for 30 s, 72 °C for 30 s with data acquisition after each cycle. At the end, properties of real-time PCR conditions and amplification products were checked by melting curve analysis: 95 °C for 1 min, 55 °C for 1 min, followed by 80 cycles with increasing incubation temperature for 10 s, starting at 55 °C and ending at 95 °C (increment 0.5 °C) with data acquisition after each cycle. PCRs were done in two replicates of each sample with the MyiQ Single Color Real-time PCR Detection System (BioRad, München, Germany) and the CT values of each sample were determined. For calculation, CT values from serial dilutions (pure up to 1000) of individual pure methylated and unmethylated control DNA samples obtained with mMGMT, uMGMT, and ActB primer pairs were plotted at the negative log of the dilution factor (standard curve). Sample mMGMT, uMGMT, and ActB CT values were calculated using the function of these standard curves, corrected to DNA amounts with ActB values (ratio MGMT/ActB; indicating the input of bisulfite-converted DNA), and calculated in relation to equivalent obtained values of individual pure control DNA as a percentage share. After this step of calculation these now so-called original values represent the fully methylated and unmethylated MGMT promoter portions in individual samples. To make the results more usable in clinical practice another last step of calculation was performed: mMGMT and uMGMT values were defined relative to the sum of MGMT fully methylated and unmethylated DNA sample amounts (mMGMT + uMGMT = 100%). These results are the PMR and PUR values. Correct length and purity of the real-time products of methylated and unmethylated control DNA were verified by polyacrylamide gel electrophoresis run under nondenaturating conditions (6% polyacrylamide). Moreover, sequence analysis of these realtime PCR products was performed by Seqlab Laboratories, Göttingen, Germany. Quantitative RT-PCR RNA was isolated using the TRIzol reagent (Invitrogen) and digested by RNase-free DNase (1 U/ll; Promega, Mannheim, Ger-

many), and cDNA was synthesized with RevertAid H Minus MmuLV Reverse Transcriptase (200 U/ll; Fermentas, Vilnius, Lithuania) [25,26]. Real-time PCR was performed in two replicates of each sample using TaqMan primer probes (assays on demand; Applied Biosystems, Foster City, CA, USA) with a total reaction volume of 20 ll. It contained 10 ll 2x TaqMan Universal PCR Master Mix, 10 or 100 ng cDNA template (diluted in 9 ll DNase-free H2O), and 1 ll 20x Assays-on-Demand Gene Expression Assay Mix. After 2 min at 50 °C and 10 min at 95 °C, 40 cycles of 15 s at 95 °C and 1 min at 60 °C were run and data acquisition followed each cycle. Reactions were performed with the MyiQ Single Color Real-time PCR Detection System (Bio-Rad) and CT values of each sample were determined. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as positive control and normalizer. CT value of each sample was averaged, and DCT values = CTGene of interest  CTGAPDH were calculated. DCT = 3.33 corresponds to one order of magnitude. Low DCT values indicate high expression. TaqMan assays had the following identification numbers and reporter sequences: MGMT, Hs00172470_m1, CCGTTTTCCAGCAAGAGTCG TTCAC and GAPDH, Hs99999905_m1, GGCGCCTGGTCACCAGGG CTGCTTT. Statistical analyses For each severalfold experiment, single results were used to calculate a mean value. Standard deviations of the mean are shown as absolute numbers for CT, DCT, and original mMGMT and uMGMT values. Moreover, bivariate correlation analysis (Pearson correlation coefficient) was determined. Results Validation of methylation-specific quantitative PCR Specificity, cross-reactivity, background analysis, and adjustment of primer concentrations. Initially, specificity and cross-reactivity of fully methylated (mMGMT) and fully unmethylated (uMGMT) primer pairs were demonstrated under real-time PCR conditions. For this purpose a MSQP assay with methylated and unmethylated control DNA using mMGMT and uMGMT primer pairs, respectively, was performed (primer concentrations 3 lM). As demonstrated by melting curve analysis, the methylated and unmethylated control DNA yielded pure products only with mMGMT and uMGMT primers, respectively (melting temperatures: 81 °C for mMGMT, 77 °C for uMGMT). No extra peaks indicative of unspecific products or primer dimerization were seen. No CT values or those above 37 were regarded as background signals. Moreover, correct fragment length and purity of real-time PCR products were analyzed by polyacrylamide gel electrophoresis and subsequent DNA staining with ethidium bromide. Amplificates of methylated control DNA with mMGMT primers showed one specific band at 81 bp, and unmethylated control DNA amplificates had a length of 93 bp (not shown). These results are in accordance with investigations of Esteller et al. [6]. Finally, the fragment sequences were determined by cycle sequencing (performed by Seqlab Laboratories) and they were similar to the predicted sequence. To exclude cross-reactivity with contaminating mRNA, MSQP assays with RNA instead of DNA were performed. No specific amplificates could be measured. Finally, to determine background of unconverted DNA, MSQP assays with ActG primers (primer concentrations 3 lM) and methylated/unmethylated bisulfite-converted control DNA were carried out. Measured CT results did not exceed values of 37 which have to be regarded as background. In contrast, when using non-bisulfite-treated glioma DNA under real-time PCR conditions, specific amplification products with ActG primers could be determined (melting temperature 85 °C, product length

Methylation-specific quantitative PCR / K. Hattermann et al. / Anal. Biochem. 377 (2008) 62–71

133 bp). Finally, to check specificity of ActB primers, MSQP with methylated and unmethylated bisulfite-converted control DNA and ActB primers (3 lM) was performed. Specific amplification products without extra peaks indicative of unspecific products for methylated and unmethylated control DNA could be received (melting temperatures 76 °C; product length 133 bp), verifying that ActB primers could be used as internal controls to normalize for input of DNA, for both methylated and unmethlylated bisulfite-converted DNA. Analysis of all following experiments included close inspection of melting curve analysis and ActG background signals. Subsequently, we ascertained the optimal primer concentrations by performance of a methylation-specific quantitative PCR on methylated and unmethylated control DNA with different primer concentrations (0.3, 0.6, 1.2, 2.4, and 3 lM) of ActB, mMGMT and uMGMT pairs, and accordingly on non-bisulfite-converted glioma DNA with ActG primers (0.3, 1.2, and 2.4 lM). Optimal primer concentrations should achieve the maximal possible product amount necessary for evaluation at minimal possible primer input to avoid unspecific bindings and primer dimerization. Optimal primer concentrations were determined at 1.2 lM for ActB, uMGMT, and ActG primer pairs and 0.6 lM for mMGMT (Fig. 1). Further increase of primer concentrations did not increase the amount of PCR products nor the generation of unspecific products. These optimal primer concentrations were used for all subsequent experiments. Summarized, these results show that the MSQP assay can clearly discriminate between the methylated and the unmethylated status of the MGMT promoter. In addition, the MSQP reactions are specific to bisulfite-converted DNA, which precludes the generation of false positive results caused by incomplete bisulfite conversion. Sensitivity and repeatability of standard curves. To test the sensitivity of MSQP, we serially diluted methylated with unmethylated control DNA up to 1:10,000 (and vice versa dilutions of unmethylated DNA with methylated DNA) and performed methylation-specific quantitative PCR on these dilutions. Both mMGMT and uMGMT primers could easily detect even dilutions up to

65

1:10,000 within the detection range of 45 PCR cycles. In both cases ActB values remained constant, indicating equal DNA amounts in all dilutions. MGMT and ActB CT values were plotted against the dilution factors (Fig. 2). Correlation coefficients were r = 0.998 for mMGMT and r = 0.991 for uMGMT. Thus, methylated and unmethylated DNA portions of as low as 0.01% of whole DNA amounts can be detected and good linearity of both primer pairs is achieved. Because standard curves should be measured in every experiment of MSQP to relate CT values to DNA concentrations, we also evaluated reproducibility of standard curves, which affects runto-run precision. Standard curves were generated by preparing dilutions of methylated and unmethylated DNA in H2O (up to 1:1000). We analyzed two different conversion reactions in six real-time PCR runs in duplicates of methylated and unmethylated control DNA. Means and standard deviations of the mean of every dilution are shown in Table 2. In general, standard deviation did not exceed 2.3% (uMGMT 1:1000), so that these data indicate good reproducibility in standard curves. Quantitative accuracy tested with mixtures of known DNA portions. To assess quantitative accuracy of MSQP, we prepared mixtures with defined portions of bisulfite-treated methylated and unmethylated control DNA (10:90, 30:70, 50:50, 70:30, 90:10) and performed methylation-specific quantitative PCR. Analysis was performed in duplicates. In Fig. 3 measured DNA portions are plotted against real portions, where bottom and left scales indicate methylated DNA and DNA portion detected with methylationspecific primers (mMGMT) and upper and right scales indicate unmethylated DNA and primers specific for unmethylated regions (uMGMT). The detected portions of methylated or unmethylated DNA showed only slight variances from the real mixtures, indicating good quantitative accuracy of the MSQP technique. Correlation coefficients were r = 0.992 (mMGMT) and r = 0.996 (uMGMT). These results demonstrate that this new methylation-specific quantitative PCR technology is suited to amplify and detect methylated and unmethylated sequences of the MGMT promoter and display the correct proportions of methylated and unmethylated DNA amounts. Application of methylation-specific quantitative PCR

Fig. 1. Determination of optimal primer concentrations. Optimal primer concentrations were determined using bisulfite-converted methylated and unmethylated control DNA (ActB, mMGMT, uMGMT) and non-bisulfite-converted genomic DNA isolated from a glioblastoma cell line (ActG). Final concentrations of 0.3, 0.6, 1.2, 2.4, and 3 lM of primer pairs ActB, mMGMT, and uMGMT in quantitative PCRs were tested; ActG was tested with 0.3, 1.2, and 2.4 lM. Optimal primer concentrations were achieved when further increase of primer concentration did not lead to lower CT values (indicating larger quantity of detected fragment), and 1.2 lM for ActB, uMGMT, and ActG primer pairs and 0.6 lM for mMGMT were thus chosen. Additionally, we made sure that there were no unspecific products with all tested primer concentrations by melting curve analysis. mMGMT, primers specific for fully methylated MGMT promoter region; uMGMT, primers specific for fully unmethylated MGMT promoter region; ActB, primers specific for unmethylated b-actin gene region; ActG, primers specific for unconverted genomic b-actin gene region.

Analysis of different cancer cell lines. Nine different cancer cell lines of three different tissue origins were subjected to methylation-specific PCR and quantitative RT-PCR based on TaqMan technology to investigate MGMT promoter methylation status in relation to MGMT mRNA expression. We analyzed four glioma cell lines (U118, U343, A172, and T98G), four breast cancer cell lines (MDA-MB 231, TD47-D, BT-549, and MCF-7), and one colon carcinoma cell line (SW480). Two different methylation-specific quantitative PCR runs each in duplicates and two independent quantitative RT-PCRs (in duplicates) were performed. For analysis of CT values obtained in methylation-specific quantitative PCR, we measured standard curves of methylated and unmethylated bisulfite-converted control DNA (serial dilutions up to 1:1000). CT values of standard curve reactions were plotted against the dilution factors and regression functions served for transformation of sample CT values to DNA amounts. MGMT CT values (either mMGMT or uMGMT) were related to DNA input detected by ActB primers (ratio MGMT/ActB) and corrected to equivalent obtained values of individual pure methylated or unmethylated control DNA as a percentage share. After this step of calculation these now so-called original values represent the fully methylated and unmethylated MGMT promoter portions in individual samples. Finally, we defined uMGMT/mMGMT relative to the sum of fully methylated and unmethylated sample MGMT DNA amount (mMGMT + uMGMT = 100%) to obtain PMR and PUR values. Quantitative RT-PCR was performed with GAPDH as

66

Methylation-specific quantitative PCR / K. Hattermann et al. / Anal. Biochem. 377 (2008) 62–71

Fig. 2. Sensitivity of methylation-specific quantitative PCR. Methylated and unmethylated control DNA were subjected to bisulfite treatment and converted methylated DNA solution was diluted in converted unmethylated DNA solution (1:10, 1:100, 1:1,000, 1:10,000) and vice versa. Dilutions were analyzed by methylation-specific quantitative PCR on fully methylated and unmethylated MGMT promoter regions. CT values of fully methylated MGMT promoter (mMGMT; A), unmethylated MGMT promoter (uMGMT; B), and ActB (b-actin, as control for DNA input) are plotted against the dilution (logarithmic scale). Both primer pairs for fully methylated and unmethylated MGMT promoter detected dilutions up to 1:10,000. Thus, methylation-specific quantitative PCR is sufficiently sensitive to detect 0.01% of fully methylated MGMT promoter DNA in a DNA sample. Correlation coefficients were r = 0.998 for mMGMT and r = 0.991 for uMGMT. mMGMT, primers specific for fully methylated MGMT promoter region; uMGMT, primers specific for fully unmethylated MGMT promoter region; ActB, primers specific for unmethylated b-actin gene region.

Table 2 Reproducibility of standard curves mDNA

pure

1:10

1:100

1:1000

ActB mean (CT) standard deviation mMGMT mean (CT) standard deviation

23.7 0.36 23.5 0.41

26.3 0.21 26.7 0.54

29.7 0.23 29.8 0.46

33.1 0.45 33.4 0.51

uDNA

pure

1:10

1:100

1:1000

ActB mean (CT) standard deviation uMGMT mean (CT) standard deviation

23.7 0.42 23.2 0.46

26.7 0.33 26.1 0.41

30.1 0.47 29.6 0.55

33.5 0.73 33.1 0.75

normalizer for total DNA amount using the DCT method described under Material and methods. Increasing DCT values indicate decreasing mRNA expression. Mean original MGMT (either mMGMT or uMGMT), PMR/PUR, and DCT values and respective standard deviations of different samples are shown in Table 3. Additionally, in Fig. 4 PUR values obtained from methylation-specific quantitative PCR (left scale, gray dots) and DCT values received from quantitative RT-PCR (right scale, white rhombs; inversely plotted) are demonstrated. When observing PMR and PUR values of investigated glioma cell lines with MSQP in U343 and A172 cell lines, no or only a small portion of fully unmethylated MGMT promoter DNA (PUR 0.1 and 3.3%; PMR 99.9 and 96.7%, respectively) was measurable, and—in accordance with this result—no MGMT mRNA expression was detectable. The other two glioma cell lines U118 and T98G contained considerable amounts of fully unmethylated MGMT promoter DNA, namely 62.2 and 80.6%. For both cell lines clearly detectable MGMT mRNA expression as displayed by DCT values of 5.88 and 8.46 was obtained, even though, in relation to U118, T98G was characterized by a higher PUR value but a 10-fold lower MGMT mRNA expression (indicated by a higher DCT value). When analyzing the original mMGMT and uMGMT values it becomes clear that the sum of these values did not reach 100%. This indicates that, among all glioma cell lines investigated, considerable amounts of partly methylated MGMT promoter portions were present; these values represent the rest up to 100% of the original mMGMT and uMGMT values.

Fig. 3. Defined mixtures of methylated and unmethylated control DNAs. Methylated and unmethylated control DNAs were subjected to bisulfite treatment and converted DNAs were mixed in defined proportions (10:90, 30:70, 50:50, 70:30, 90:10). DNA mixtures were analyzed by methylation-specific quantitative PCR on fully methylated and unmethylated MGMT promoter regions. CT values were evaluated by standard curves of pure methylated or unmethylated control DNA and normalized to ActB (b-actin) CT values. Both methylated and unmethylated DNA portions could reasonably well be reproduced. Correlation coefficients were r = 0.992 (mMGMT) and r = 0.996 (uMGMT). mMGMT, primers specific for methylated MGMT promoter region; uMGMT, primers specific for unmethylated MGMT promoter region.

Among the four analyzed breast cancer cell lines, MDA-MB 231 did not show any detectable MGMT mRNA expression nor could a fully unmethylated MGMT promoter fragment be amplified (PUR 0.2%). The other breast cancer cell lines TD47-D, MCF-7, and BT549 had completely fully unmethylated MGMT promoter portions (PUR 99.97, 99.9, and 100%, respectively). Accordingly, they were characterized by a high MGMT mRNA expression, lying in the same range as those values obtained for human lymphocytes, which

67

Methylation-specific quantitative PCR / K. Hattermann et al. / Anal. Biochem. 377 (2008) 62–71 Table 3 Mean original, PMR/PUR, and DCT values of different cancer cell lines MGMT methylation original mMGMT/uMGMT values (%)

U118 U343 A172 T98G MDA-MB TD47-D BT-549 MCF-7 SW480 Lymphocytes

MGMT methylation PMR/PUR values (%)

MGMT mRNA expression

Mean mMGMT

Standard deviation

Mean uMGMT

Standard deviation

PMR

PUR

Mean DCT

Standard deviation

17.2 22.7 8.8 6.5 41.5 0.01 0 0.001 18.5 0.004

±1.6 ±4.7 ±2.3 ±0.7 ±4.9 ±0.01 n/a ±0.001 ±0.7 ±0.004

28.3 0.03 0.3 27 0.07 33 28.8 65.2 25 99.9

±2.7 ±0.03 ±0.2 ±1.4 ±0.05 ±4.2 ±6.1 ±5.9 ±2.8 ±15.1

37.8 99.9 96.7 19.4 99.8 0.03 0 0.001 42.5 0.004

62.2 0.1 3.3 80.6 0.2 99.97 100 99.9 57.5 99.9

5.88 undet. undet. 8.46 undet. 1.72 5.89 4.82 8.7 3.82

±0.09 n/a n/a ±0.08 n/a ±2.9 ±0.53 ±0.94 ±1 ±0.23

Fig. 4. Portion of fully unmethylated MGMT promoter DNA and MGMT gene expression in human cancer cell lines. Portion of fully unmethylated MGMT promoter DNA was determined by methylation-specific quantitative PCR and MGMT mRNA expression was analyzed by quantitative RT-PCR. For methylation-specific quantitative PCR, DNA was isolated from tumor cell lines of different origin (glioblastoma: U118, U343, A172, T98G; breast cancer: MDA-MB 231, TD47-D, BT-549, MCF-7; colon carcinoma: SW480) and subjected to bisulfite treatment. Afterward, methylation-specific quantitative PCR was performed and CT values were evaluated using standard curves of pure methylated and unmethylated control DNA. mMGMT and uMGMT values were corrected to ActB (b-actin, as control to DNA input), related to pure equivalent obtained values of methylated or unmethylated control DNA as a percentage share (original values), and used to obtain PMR and PUR values relative to the sum of fully methylated and unmethylated MGMT sample DNA (mMGMT + uMGMT = 100%). PUR (percentage of fully unmethylated DNA portion) values are plotted at the left-hand scale. Detection limits of methylationspecific quantitative PCR means no detectable signal after 45 PCR cycles. For quantitative RT-PCR, RNA was isolated and reversely transcribed and MGMT mRNA expression analyzed by specific TaqMan primers and probes. DCT values relative to glyceraldehyde-3-phosphate dehydrogenase are plotted at the right hand-scale. DCT = 3.33 corresponds to one order of magnitude. Low DCT values indicate high expression. Below detection limit means that no signal was seen from 10 ng RNA after 40 amplification cycles. In all cell lines investigated there was a general tendency that PUR values correlate with MGMT mRNA expression.

served as a control sample. Nevertheless, as indicated by the original mMGMT and uMGMT values, breast cancer cell lines were also characterized by considerable amounts of partly methylated MGMT DNA portions. SW480, a colon carcinoma cell line, had a PUR of 57.5%, meaning that, in relation to summarized total amounts of fully unmethylated and methylated values, 57.5% of the detected MGMT promoter DNA portions were fully unmethylated, whereas 42.5% were fully methylated. Because a difference of DCT = 3.33 corresponds to one order of magnitude and low DCT values indicate high mRNA expression, in SW480 MGMT mRNA expression was up to 100 fold lower than that in cell lines with completely unmethylated MGMT promoter, e.g., TD47-D. For all investigated samples, statistical analyzes revealed low standard deviations for both methylation-specific quantitative PCR and quantitative RT-PCR results. Summarized, MSQP is suitable for detection of the MGMT promoter methylation status in tumor cell lines of different origin. In all investigated cell lines there was a general tendency that unmethylated MGMT promoter status was detected along with high mRNA levels and methylated MGMT promoter status along with low mRNA levels. Analysis of tumor specimens. Primary glioblastoma samples of 15 patients and a normal brain tissue sample were subjected to

DNA and RNA isolation and analyzed by MSQP and quantitative RT-PCR analysis for MGMT mRNA expression. Calculation of original mMGMT/uMGMT, PMR/PUR, and DCT values were performed as described above. Mean original MGMT, PMR/PUR, and DCT values and respective standard deviations of different samples are shown in Table 4. Additionally, PURs and DCT values are plotted against sample numbers (Fig. 5), with gray dots indicating PURs and referring to the left scale and white rhombs (inversely plotted) indicating MGMT mRNA expression and referring to the right scale. Among all 15 tissue samples there were 4 samples with PUR values below 99% (Table 4). Three of these samples (5, 6, and 10) also show comparatively low gene expression (indicated by high DCT values) which are reduced up to 1000-fold compared to the normal brain tissue sample. By contrast, for 1 of these samples (12) a distinct portion of fully methylated MGMT promoter was detectable (PMR 9.7%; PUR 90.3%), but this sample was characterized by a relatively high MGMT mRNA expression. Also, sample 13 was fully unmethylated but only a low MGMT mRNA expression could be determined (DCT 8.45) with regard to the original mMGMT and uMGMT values, it becomes clear that, with the exception of samples 3 and 4, all glioblastoma samples contained individually variable amounts of partly methylated MGMT DNA promoter portions.

68

Methylation-specific quantitative PCR / K. Hattermann et al. / Anal. Biochem. 377 (2008) 62–71

Table 4 Mean original, PMR/PUR, and DCT values of different tumor samples Sample number

MGMT methylation original mMGMT/uMGMT values (%) Mean mMGMT

Standard deviation

Survival time after first surgery <12 month 1 0.0006 ±0.0009 2 0.2 ±0.04 3 undet. n/a 7 0.04 ±0.06 8 0.009 ±0.008 9 0.003 ±0.001 11 0.005 ±0.001 Survival time after first surgery P12 month 4 0.05 ±0.03 5 6.2 ±0.8 6 0.9 ±0.04 10 1.2 ±0.3 12 2.5 ±0.7 13 0.01 ±0.01 14 0.002 ±0.003 15 0.003 ±0.005 Normal brain tissue NBT 0.011 ±001

MGMT methylation PMR/PUR values (%)

MGMT mRNA expression

PMR

PUR

Mean DCT

0.001 0.5 0 0.05 0.01 0.003 0.006

99.99 99.5 100 99.95 99.99 99.99 99.99

6.24 ±0.24 6.36 ±0.11 not determined 7.06 ±0.17 5.64 ±0.31 5.36 ±0.002 6.60 ±0.09

Mean uMGMT

Standard deviation

60.6 44 100.5 87.7 76.8 88.2 75.5

±7 ±0.8 ±27.6 ±3.3 ±15.9 ±18.7 ±4.3

99.5 7.1 36.5 36.5 23.3 4.3 51.5 25.3

±5.7 ±1.6 ±4.8 ±1.9 ±3 ±0.6 ±3.2 ±1.8

0.05 46.6 2.4 3.2 9.7 0.2 0.004 0.01

99.95 53.4 97.6 96.8 90.3 99.8 99.99 99.98

5.56 10.29 9.40 7.81 4.67 8.45 3.95 3.82

±0.32 ±0.18 ±0.39 ±0.24 ±0.50 ±0.59 ±0.99 ±0.74

100.7

±9.8

0.01

99.99

3.71

±0.33

Standard deviation

Fig. 5. Portion of fully unmethylated MGMT promoter DNA and MGMT gene expression in human glioblastoma and normal brain tissue samples. Portion of fully unmethylated MGMT promoter DNA was determined by methylation-specific quantitative PCR and MGMT mRNA expression was analyzed by quantitative RT-PCR. For methylationspecific quantitative PCR, DNA was isolated from 15 glioblastoma specimens (samples 1–15) and a sample of normal brain tissue (NBT) and subjected to bisulfite treatment. Methylation-specific quantitative PCR and evaluation was performed and CT values were evaluated using standard curves of pure methylated and unmethylated control DNA. mMGMT and uMGMT values were corrected to ActB (b-actin, as control to DNA input), related to equivalent obtained values of pure methylated or unmethylated control DNA as a percentage share (original values), and used to obtain PMR and PUR values relative to the sum of fully methylated and unmethylated MGMT sample DNA (mMGMT + uMGMT = 100%). PUR (percentage of fully unmethylated DNA portion) values are plotted at the left-hand scale. Quantitative RT-PCR was performed and evaluated using the DCT method. DCT values relative to glyceraldehyde-3-phosphate dehydrogenase are plotted at the right-hand scale. A clear correlation between PUR values and MGMT mRNA expression could not be observed. Although there were some tissue samples with low PUR values and low mRNA expression (indicated by high CT values, e.g., samples 5 and 6) other samples with low PUR value showed high mRNA levels (e.g., sample 12).

Considering the outcome data in this small study there is a hint that patients with a PUR reduction to 99% or less showed a prolonged survival time after surgery: in a group of eight long-time survivors (P12 months) four had tissue samples with PURs <99% whereas none of seven tested samples from patients with survival times 612 months had a PUR smaller than 99%. Nevertheless, it has to be clearly taken into account that this study was designed to test the applicability of MSQP to MGMT promoter methylation analysis in glioblastoma tissue samples and its small scope does not allow statistical analysis on methylation status correlated to outcome data. Summarized, MSQP allows a highly sensitive detection of the methylation level of the MGMT promoter in solid human glioblastoma samples by combining methylation-specific and quantitative PCR techniques. Interestingly, compared to cultured cell lines in solid tumor specimens, a broader spectrum of the MGMT promoter methylation status in relation to the MGMT mRNA expression was obtained. Moreover, using MSQP also small portions of fully methylated MGMT promoter were detectable, and it seems that patients with these epigenetic characteristics had a better outcome.

Nevertheless, it has to be taken into account also that some patients without any detectable full methylation status of the MGMT promoter showed prolonged survival times. Discussion O6-methylguanine DNA methyltransferase promoter methylation may be a useful marker for predicting prognosis and monitoring efficacy of adjuvant therapy in human glioblastoma patients [10– 16]. In accordance with this, significantly more frequent MGMT hypermethylation was found in a group of glioblastoma long-term survivors [27]. A variety of assays to measure DNA methylation have been developed, and many of these methods rely on sodium bisulfite treatment of genomic DNA from tumor tissue [25,28–43]. In this paper we present an advanced method, MSQP, combining methylation-specific and SYBR-green-based quantitative PCR for MGMT promoter methylation analysis. This MSQP technique is a highly specific, sensitive, and reproducible method that allows the quantitative determination of not only fully methylated but also fully unmethylated bisulfite-converted MGMT DNA species

Methylation-specific quantitative PCR / K. Hattermann et al. / Anal. Biochem. 377 (2008) 62–71

in terms of percentage. This precise quantitative identification is extremely relevant for the determination of the MGMT promoter methylation status within the scope of clinical diagnosis. The completely unmethylated status of the MGMT promoter is responsible for MGMT protein synthesis and by this for the relevance of this enzyme in the progression of chemotherapy resistance during tumor therapy. On the other hand, amounts of fully methylated MGMT promoter portions have been associated with an increased benefit from chemotherapy [10–16]. With the MSQP technique, fully unmethylated and methylated portions of MGMT promoter can now be achieved in terms of percentage, allowing a more precise and easily interpretable identification of the hypermethylation status. MSQP values are obtained using the original methylation-specific PCR primer pairs [6,17,18] in a real-time setting, calculated in relation to standard curves generated for each primer pair, respectively, corrected for input of DNA by an internal standard, and calculated in relation to equivalent obtained values of methylated or unmethylated control DNA as a percentage share. To make the results more usable in clinical practice, we defined values relative to the sum of fully methylated and unmethylated MGMT DNA sample amount to obtain percentage of methylated reference and percentage of unmethylated reference. Because the methylated and unmethylated control DNA yielded pure products only with mMGMT and uMGMT primers, we could show that the MSQP assay can clearly discriminate between the methylated and the unmethylated status of the MGMT promoter. However, it has to be mentioned that in all PCR-related techniques possible ‘‘wobble” activities of sequence-specific primers may occur and thus recognition of partly methylated DNA portions cannot absolutely be excluded. Next, we were unable to amplify nonbisulfite-converted DNA with the methylated (mMGMT), the unmethylated (uMGMT), or the control (ActB) oligonucleotides, demonstrating that MSQP reactions are specific to bisulfite-converted DNA. Considering that bisulfite-treated DNA samples gave no clearly detectable signal with control oligonucleotides designed to recognize a non-bisulfite-converted sequence (ActG), we could confirm that the bisulfite conversion process had been successful in every experiment. In contrast, control oligonucleotides designed to recognize a bisulfite-converted sequence of b-actin (ActB) yielded identical reproducible signals for methylated and unmethylated bisulfite-converted DNA, respectively, verifying that ActB primers could be used as internal controls to correct for input of DNA. Additionally, we could show that variations of standard curves used for calculation of MSQP values in each run were small and acceptable, indicating a good reproducibility. Moreover, we performed optimization of the primer concentrations to achieve the maximal possible product amount necessary for evaluation at minimal possible primer input to avoid unspecific bindings and primer dimerization. Combining these highly accurate settings, the MSQP technique was particularly suitable for detecting fully methylated and unmethylated DNA portions as low as 0.01% of whole DNA amounts and was able to display the correct proportions of methylated and unmethylated DNA amounts with excellent precision. Like MethyLight, MSQP is a quantitative and relatively simple SYBR-green-based PCR technique which allows in contrast to bisulfite-pyrosequencing [28,40,41] visualization of the PCR procedure over a logarithmic scale and not only with end-point analysis. Moreover, with MSQP several tumor samples can be investigated in one real-time run in parallel, which is time saving and therefore clearly less expensive. In accordance with this, the ability of SYBRgreen-based PCR technology for detection of hypermethylated gene expression has been determined as a valuable tool for diagnosis of early nonaggressive carcinogenesis in benign prostate hyperplasia [44]. Nevertheless, in this study only methylated portions of

69

DNA were analyzed. In comparison to MSP [6,17,18] and MethyLight [25] methods, MSQP offers the ability to investigate fully unmethylated and methylated bisulfite-converted MGMT DNA species in parallel in a quantitative manner. Employing original MSP primers, these results can be easily compared to previous pioneering studies focused on the correlation of MGMT promoter methylation status, chemotherapy, and survival of glioblastoma patients [12,13,15]. Moreover, by defining a relation between fully unmethylated and methylated portions of the MGMT promoter in terms of percentage, MSQP provides more detailed information about the overall methylation status of the samples, and by this it offers the possibility to specify the correlation of MGMT promoter hypermethylation and for example survival of glioblastoma patients in a more exact manner. Finally, by using different gene-specific primer sets, MSQP could easily be adapted for analysis of methylation portions of any other gene promoters, making this novel technique multifaceted. To assess the clinical applicability of MSQP, we used methylation-specific PCR and quantitative RT-PCR based on TaqMan technology to investigate MGMT gene promoter methylation status in relation to MGMT mRNA expression in nine different tumor cell lines and 15 different primary glioblastoma patients. As unmethylated control samples human peripheral lymphocytes and normal brain tissue were used. Two different methylation-specific quantitative PCR runs each in duplicates and two independent quantitative RT-PCRs (in duplicates) were performed. For all investigated samples obtained original mMGMT and uMGMT values and PUR and PMR values were sufficient to definitively differentiate methylation-positive samples from methylation-negative samples. For tumor cell lines, there was a general tendency that unmethylated MGMT promoter status was detected along with high mRNA levels, demonstrating that MSQP was suitable for detection of real portions of MGMT promoter methylation status. For glioblastoma samples four patients were found with PUR lower than 99%, which means that portions of fully methylated MGMT promoter DNA were also existing. Compared to cultured cell lines in solid tumor specimens, a broader spectrum of the MGMT methylation status in relation to the MGMT mRNA expression was obtained. These results are in accordance with previous results indicating an association of MGMT promoter methylation and loss of protein expression [21,45], but cases with loss of protein expression without promoter methylation and cases with intact protein expression and a clearly detectable methylation status also exist [21,46]. However, because it cannot absolutely be excluded that the tumor specimens contain some amount of normal cells (due to tumor inhomogeneity per se), the broader variety of methylation status compared to MGMT mRNA expression may also be contributed by low contaminations with healthy cells. Without any statistical comparison, in our small study set, we found a trend toward longer survival combined with the occurrence of MGMT promoter hypermethylation. Thus, our data obtained with MSQP are in line with recent studies reporting that MGMT promoter hypermethylation is associated with prolonged overall survival in patients treated with alkylating agents [10,12,27]. On the other hand, it is important to note that the absence of MGMT promoter methylation is still compatible with long-term survival in individual patients. Moreover, it should be kept in mind that, next to MGMT, other factors might be involved. For example, high activity of poly-ADP ribose polymerase and base excision repair may be cytoprotective for cells [47,48]. Summarized, we have developed a new method for the quantitative, sensitive and specific analysis of the MGMT promoter methylation status. Presented data confirm that this assay is useful for detection of low amounts of both fully methylated and unmethylated MGMT promoter DNA within glioblastoma samples. Carefully

70

Methylation-specific quantitative PCR / K. Hattermann et al. / Anal. Biochem. 377 (2008) 62–71

validated quantitative MSQP assays will be useful in both research and clinical molecular diagnosis.

[21]

Acknowledgments We thank Mrs. B. Rehmke, Mrs. U. Malkus-Coskun, and Mr. J. Krause for expert technical assistance and Mr. C. Franke for drawing figures. This work was supported by the University of Kiel and the Family Mehdorn Foundation.

[22]

References

[24]

[1] M. Cross, T.M. Dexter, Growth factors in development, transformation, and tumourigenesis, Cell 64 (1991) 271–280. [2] J.F. Costello, M.C. Fruhwald, D.J. Smiraglia, L.J. Rush, G.P. Robertson, X. Gao, F.A. Wright, J.D. Feramisco, P. Peltomaki, J.C. Lang, D.E. Schuller, L. Yu, C.D. Bloomfield, M.A. Caligiuri, A. Yates, R. Nishikawa, H. Su Huang, N.J. Petrelli, X. Zhang, M.S. O´Dorisio, W.A. Held, W.K. Cavenee, C. Plass, Aberrant CpG-island methylation has non-random and tumour-type specific patterns, Nat. Genet. 24 (2000) 132–138. [3] P.A. Jones, P.W. Laird, Cancer epigenetics comes of age, Nat. Genet. 21 (1999) 163–167. [4] S.K. Myohanen, S.B. Baylin, J.G. Herman, Hypermethylation can selectively silence individual p16ink4A alleles in neoplasia, Cancer Res. 58 (1998) 591– 593. [5] J.G. Herman, A. Umar, K. Polyak, J.R. Graff, N. Ahuja, J.P. Issa, S. Markowitz, J.K. Willson, S.R. Hamilton, K.W. Kinzler, M.F. Kane, R.D. Kolodner, B. Vogelstein, T.A. Kunkel, S.B. Baylin, Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma, Proc. Natl. Acad. Sci. USA 95 (1998) 6870–6875. [6] M. Esteller, S.R. Hamilton, P.C. Burger, S.B. Baylin, J.G. Herman, Inactivation of the DNA repair gene O6-methylguanne-DNA methyltransferase by promoter hypermethylation is a common event in primary human neoplasia, Cancer Res. 59 (1999) 793–797. [7] A.E. Pegg, Mammalian O6-alkylguanine-DNA alkyltransferase regulation and importance in response to alkylating carcinogenic and therapeutic agents, Cancer Res. 50 (1990) 6119–6129. [8] M. Esteller, J.G. Herman, Generating mutations by providing chemosensitivity: the role of O6-methylguanine DNA methyltransferase in human cancer, Oncogene 23 (2004) 1–8. [9] M.R. Middleton, G.P. Margison, Improvement of chemotherapy efficacy by inactivation of a DNA-repair pathway, Lancet Oncol. 4 (2003) 37–44. [10] M. Esteller, J. Garcia-Foncillas, E. Andiou, S.N. Goodman, O.F. Hidalgo, V. Vanaclocha, S.B. Baylin, J.G. Herman, Inactivation of the DNA-repair gene MGMT and the clinical response of glioma alkylating agents, N. Engl. J. Med. 343 (2000) 1350–1354. Erratum N. Engl. J. Med. 343 (2000) 1740. [11] J.P. Fruehauf, H. Brem, S. Brem, A. Sloan, G. Barger, W. Huang, R. Parker, In vitro drug response and molecular markers associated with drug resistance in malignant gliomas, Clin. Cancer Res. 12 (2006) 4523–4532. [12] M.E. Hegi, A.-C. Diserens, T. Gorlia, M.-F. Hamou, N. de Tribolet, M. Weller, J.M. Kros, J.A. Hainfellner, W. Mason, L. Mariani, J.E.C. Bromberg, P. Hau, R.O. Mirimanoff, J.G. Cairncross, R.C. Janzer, R. Stupp, MGMT gene silencing and benefit from temozolomide in glioblastoma, N. Engl. J. Med. 352 (2005) 997– 1003. [13] M.E. Hegi, A.-C. Diserens, S. Godard, P.-Y. Dietrich, L. Regli, S. Ostermann, P. Otten, G. Van Melle, N. de Tribolet, R. Stupp, Clinical trial substantiates the predictive value of O-6-methylguanine DNA methyltransferase promoter treated with temozolamide, Clin. Cancer Res. 10 (2004) 1871–1874. [14] W.P. Mason, J.G. Cairncross, Drug insight: temozolomide as a treatment for malignant glioma– impact of a recent trial, Nat. Clin. Pract. Neurol. 1 (2005) 88–95. [15] M.F. Paz, R. Yaya-Tur, I. Rojas-Marcor, G. Reynes, M. Pollan, L. Aguirre-Cruz, J.L. Garcia-Lopez, J. Piqer, M.-J. Safont, C. Balana, M. Sanchez-Cespedes, M. GarciaVillanueva, L. Arribas, M. Esteller, CpG island hypermethylation of the DNA repair enzyme methyltransferase predicts response to temozolomide in primary gliomas, Clin. Cancer Res. 10 (2004) 4933–4938. [16] R. Stupp, P.Y. Dietrich, S. Ostermann Kraljevic, A. Pica, I. Maillard, P. Maeder, R. Meuli, R. Janzer, G. Pizzolato, R. Miralbell, F. Porchet, L. Regli, N. de Tribolet, R.O. Mirimanoff, S. Leyvraz, Promising survival for patients with newly diagnosed glioblastoma multiforme treated with concomitant radiation plus temozolomide followed by adjuvant temozolomide, J. Clin. Oncol. 20 (2002) 1375–1382. [17] J.G. Herman, J.R. Graff, S. Myöhänen, B.D. Nelkin, S.B. Baylin, Methylationspecific PCR: a novel PCR assay for methylation status of CpG islands, Proc. Natl. Acad. Sci. USA 93 (1996) 9821–9826. [18] W.A. Palmisano, K.K. Divine, G. Saccomanno, F.D. Gilliland, S.B. Baylin, J.G. Herman, S.A. Belinsky, Predicting lung cancer by detecting aberrant promoter methylation in sputum, Cancer Res. 60 (2000) 5954–5958. [19] A. Olek, J. Oswald, J. Walter, A modified and improved method for bisulfite based cytosine methylation analysis, Nucleic Acids Res. 24 (1996) 5064–5066. [20] M.O. Hoque, S. Begum, O. Topaloglu, A. Chatterjee, E. Rosenbaum, W.V. Criekinge, W.H. Westra, M. Schoenberg, M. Zahurak, S.N. Goodman, D.

[23]

[25]

[26] [27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41] [42]

[43]

[44]

[45]

Sidransky, Quantification of promoter methylation of multiple genes in urine DNA and bladder cancer detection, J. Natl. Cancer Inst. 98 (2006) 996–1004. S. Ogino, T. Kawasaki, M. Brahmandam, M. Cantor, G.J. Kirkner, D. Spiegelman, G.M. Makrigiorgos, D.J. Weisenberger, P.W. Laird, M. Loda, C.S. Fuchs, Precision and performance characteristics of bisulfite conversion and real-time PCR (MethyLight) for quantitative DNA methylation analysis, J. Mol. Diagn. 8 (2006) 209–217. M. Widschwendter, K.D. Siegmund, H.M. Müller, H. Fiegl, C. Marth, E. MüllerHolzner, P.A. Jones, P.W. Laird, Association of breast cancer DNA methylation profiles with hormone receptor status and response to tamoxifen, Cancer Res. 64 (2004) 3807–3813. J. Feindt, I. Becker, U. Blömer, H.-H. Hugo, H.M. Mehdorn, B. Krisch, R. Mentlein, Expression of somatostatin receptor subtypes in cultured astrocytes and gliomas, J. Neurochem. 65 (1995) 1997–2005. A.M. Stark, B. Anuszkiewicz, R. Mentlein, T. Yoneda, H.M. Mehdorn, J. HeldFeindt, Differential expression of matrix metalloproteinases in brain- and bone-seeking clones of metastatic MDA-MB-231 breast cancer cells, J. NeuroOncol. 81 (2007) 39–48. C.A. Eads, K.D. Danenberg, K. Kawakami, L.B. Saltz, C. Blake, D. Shibata, P.V. Danenberg, P.W. Laird, MethyLight: a high-troughput assay to measure DNA methylation, Nucleic Acids Res. 28 (2000) e32. 1-8. J. Held-Feindt, L. Dörner, G. Sahan, H.M. Mehdorn, R. Mentlein, Cannabinoid receptors in human astroglial tumours, J. Neurochem. 98 (2006) 886–893. D. Krex, B. Klink, C. Hartman, A. von Deimling, T. Pietsch, M. Simon, M. Sabel, J.P. Steinbach, O. Heese, G. Reifenberger, M. Weller, G. Schackert for the German Glioma Network, Long term survival with glioblastoma multiforme, Brain (2007) DOI 10.1093/brain/awm204. S. Colella, L. Shen, K.A. Baggerly, J.P. Issna, R. Krahe, Sensitive and quantitative universal pyrosequencing methylation analysis of CpG sites, Biotechniques 35 (2003) 146–150. S.E. Cottrell, J. Distler, N.S. Goodman, S.H. Mooney, A. Kluth, A. Olek, I. Schwope, R. Tetzner, H. Ziebarth, K. Berlin, A real-time assay for DNAmethylation using methylation-specific blockers, Nucleic Acids Res. 32 (2004) e10. 1-8. D. Deng, G. Deng, M.F. Smith, J. Zhou, H. Xin, S.M. Powell, Y. Lu, Simultaneous detection of CpG methylation and single nucleotide polymorphism by denaturing high performance liquid chromatography, Nucleic Acids Res. 30 (2002) e13. 1-6. C.A. Eads, K.D. Danenberg, K. Kawakami, L.B. Saltz, P.V. Danenberg, P.W. Laird, CpG island hypermethylation in human colorectal tumours is not associated with DNA methyltransferase overexpression, Cancer Res. 59 (1999) 2302– 2306. C.A. Eads, R.V. Lord, S.K. Kurumboor, K. Wickramasinghe, M.L. Skinner, T.I. Long, J.H. Peters, T.R. DeMeester, K.D. Danenberg, P.V. Danenberg, P.W. Laird, K.A. Skinner, Fields of aberrant CpG island hypermethylation in Barrett´s esophagus and associated adenocarcinoma, Cancer Res. 60 (2000) 5021–5026. M. Gonzalgo, P.A. Jones, Quantitative methylation analysis using methylationsensitive single-nucleotide primer extension (MsSNuPE), Methods 27 (2002) 128–133. J.W.M. Jeuken, S.J.B. Cornelissen, M. Vriezen, M.M.G. Dekkers, A. Errami, A. Sijben, S.H.E. Boots-Sprenger, P. Wesseling, MS-MLPA: an attractive alternative laboratory assay for robust, reliable, and semiquantitative detection of MGMT promoter hypermethylation in gliomas, Lab. Invest. 87 (2007) 1055–1065. M.M. Matin, A. Baumer, D.P. Hornby, An analytical method for the detection of methylation differences at specific chromosomal loci using primer extension and ion pair reverse phase HPLC, Human Mutat. 20 (2002) 305–311. K. Rand, W. Qu, T. Ho, S.J. Clak, P. Molloy, Conversion-specific detection of DNA methylation using real-time polymerase chain reaction (ConLight-MSP) to avoid false positives, Methods 27 (2002) 114–120. P. Schatz, D. Dietrich, M. Schuster, Rapid analysis of CpG methylation patterns using RNase T1 cleavage and MALDI-TOF, Nucleic Acids Res. 32 (2004) e167. 17. S.D. Steigerwald, G.P. Pfeifer, A.D. Riggs, Ligation-mediated PCR improves the sensitivity of methylation by restriction enzymes and detection of specific DNA strand breaks, Nucleic Acids Res. 18 (1990) 1435–1439. H. Thomassin, C. Kress, T. Grange, MethylQuant: a sensitive method for quantifiying methylation of specific cytosines within the genome, Nucleic Acids Res. 32 (2004) e168. 1-9. J. Tost, J. Dunker, I.G. Gut, Analysis and quantification of multiple methylation variable positions in CpG islands by pyrosequencing, Biotechniques 35 (2003) 152–156. Z. Xiong, P.W. Laird, COBRA: a sensitive and quantitative DNA methylation assay, Nucleic Acids Res. 25 (1997) 2532–2534. M. Zeschnigk, S. Böhringer, E.A. Price, Z. Onadim, L. Maßhöfer, D.R. Lohmann, A novel real-time PCR assay for quantitative analysis of methylated alleles (QAMA): analysis of the retinoblastoma locus, Nucleic Acids Res. 32 (2004) e125. 1-5. K. Uhlmann, A. Brinckmann, M.R. Toliat, H. Ritter, P. Nurnberg, Evaluation of potential epigenetic biomarker by quantitative methyl-single nucleotide polymorphism analysis, Electrophoresis 23 (2002) 4072–4079. P.J. Bastian, J. Ellinger, N. Wernert, L.C. Heukamp, S.C. Müller, A. von Ruecker, Diagnostic and prognostic information in prostate cancer with the help of a small set of hypermethylated gene loci, Clin. Cancer Res. 11 (2005) 4097–4106. O.L. Chinot, M. Barrie, S. Fuentes, N. Eudes, S. Lancelot, P. Metellus, X. Muracciole, D. Braguer, L. Ouafik, P.M. Martin, H. Dufour, D. Figarella-Branger,

Methylation-specific quantitative PCR / K. Hattermann et al. / Anal. Biochem. 377 (2008) 62–71 Correlation between O6-methylguanine-DNA methyltransferase and survival in inoperable newly diagnosed glioblastoma patients treated with neoadjuvant temozolamide, J. Clin. Oncol. 25 (2007) 1470–1475. [46] E.M. Grasbon-Frodl, F.W. Kreth, M. Ruiter, M. Schnell, K. Bise, J. Felsburg, G. Reifenberger, J.-C. Tonn, H.A. Kretschmar, Intratumoral homogeneity of MGMT promoter hypermethylation as demonstrated in serial stereotactic specimens from anaplastic astrocytomas and glioblastomas, Int. J. Cancer 121 (2007) 2458–2464.

71

[47] N.J. Curtin, L.Z. Wang, A. Yiakouvaki, S. Kyle, C.A. Arris, S. Canan-Koch, S.E. Webber, B.W. Durkacz, H.A. Calvert, Z. Hostomsky, D.R. Newell, Novel poly(ADP-ribose) polymerase-1 inhibitor, AG14361, restores sensitivity to temozolamide in mismatch repair-deficient cells, Clin. Cancer Res. 10 (2004) 881–889. [48] R.N. Trivedi, K.H. Almeida, J.L. Fornsaglio, S. Schamus, R.W. Sobol, The role of base excision repair in the sensitivity and resistance to temozolamidemediated cell death, Cancer Res. 65 (2005) 6394–6400.