Inhibitors of DNA binding/differentiation (Id1 to Id4) are a family of helix-loop-helix transcription factors, which are highly expressed during embryogenesis and at lower levels in mature tissues. Id4 plays an important role in neuronal stem cell differentiation, and its deregulation has been implicated in glial neoplasia.
The methylation status of Id4 was analyzed by methylation-specific polymerase chain reaction (PCR) in 62 glioblastoma (GBM) cases and in 20 normal brain tissues. Methylation status of Id4 was confirmed by sequencing after subcloning and messenger RNA (mRNA) and protein expression. We also evaluated the mRNA expression of MGP (matrix GLA protein), TGF-β1 (transforming growth factor beta 1), and VEGF (vascular endothelial growth factor) by real-time PCR analysis. Clinical and histological assessment of tumor angiogenesis was performed by evaluating the relative enhancing tumor ratio on magnetic resonance imaging and microvessel density on von Willebrand factor–stained sections, respectively.
The promoter of Id4 was methylated in 23 of 62 (37%) GBMs. In methylated GBMs, Id4 mRNA was significantly reduced, compared with unmethylated GBMs (P = .0002). A significant reduction of protein expression was detected in all hypermethylated cases. GBMs with methylated Id4 showed a significant reduction of MGP, TGF-β1, and VEGF mRNA expression and had significantly lower relative enhancing tumor ratio (P = .0108) and microvessel density (P = .0241) values with respect to unmethylated GBMs. Finally, Id4 methylation was significantly associated with a favorable clinical outcome (P = .0006).
Glioblastoma (GBM), the most frequent brain malignancy, remains an incurable cancer with a prognosis of about 1 year after diagnosis, despite advances in neurosurgery and adjuvant treatments.1,2 One of the most investigated pathways involved in the pathogenesis of GBM concerns angiogenesis, which plays a central role in the aggressiveness of this cancer and which also represents a promising pharmacological target.3-5
In this view, Kuzontkoski et al have recently demonstrated that the inhibitor of DNA binding/differentiation type 4 (Id4) exerts a proangiogenic activity in GBM.6 In fact, tumor-derived cell cultures expressing elevated Id4 levels produced bigger and highly vascularized xenografts in immunodeficient mice, compared with controls. The authors also report that neoangiogenesis was mediated by matrix GLA protein (MGP) which was expressed at higher levels in Id4-transfected cultures and whose inhibition determined smaller and less vascularized tumor xenograft.6 The angiogenic effect of MGP is due to an increasing activity of transforming growth factor beta 1 (TGF-β1) which, in turn, determines an up-regulation of vascular endothelial growth factor (VEGF) in endothelial cells.7
Id4 is a member of a family of helix-loop-helix transcription factors that, having lost the basic DNA-binding domain, acts as dominant-negative regulator by forming inactive heterodimeric complexes with other helix-loop-helix transcription proteins.8-10 To date, 4 structurally similar proteins (Id1 to Id4) have been found that perform different regulatory functions mainly during embryogenesis, when these proteins are highly expressed.10,11 In mature tissues, with the exception of some stem cells, Id proteins are absent; however, they are expressed in several tumors where they seem to play a key role in cellular transformation, immortalization, invasion, and in the metastatic process.10-12 In particular, hyperexpression of Id4 due to functional point mutation or to epigenetic silencing has been extensively reported in colorectal, breast, gastric, and hematologic cancer.4,13-16
Several lines of evidence suggest that Id4 is involved in the neural stem cell, oligodendrocyte, and astrocyte differentiation and that this protein also plays a central role in GBM pathogenesis through an up-regulation of Jagged1 and a subsequent Notch1 activation.4,10,11,17-19 In addition, Id4 seems to regulate the chemoresistance of glioma stem cells by promoting SOX2 (sex-determining region Y box 2)-mediated induction of ATP-binding cassette (ABC) transporters.20
In this report, we analyzed the methylation status and the consequences of promoter hypermethylation on Id4 messenger RNA (mRNA) as well as on protein expression in a series of GBM tumors (GBMs). We also investigated the correlation between Id4 methylation and MGP, TGF-β1, and VEGF expression. Results were correlated with the patients' clinical and biological characteristics.
MATERIALS AND METHODS
This study includes 62 consecutive adult patients who underwent craniotomy for resection of histologically confirmed GBM (World Health Organization grade IV)21 in the supratentorial compartment, and who were treated postoperatively with adjuvant radiotherapy and temozolomide (TMZ) at the Università Cattolica del Sacro Cuore (UCSC), Rome, Italy. All patients provided written informed consent according to the research proposals approved by the Ethical Committee of the UCSC. Patients of pediatric age and patients with secondary GBM, ie, patients with a previous diagnosis of lower grade glioma, were not included in this study.
The patients were aged 20 to 80 years at the time of primary surgery (median age, 60 years; mean age, 59.9 ± 9.3 years); 40 were men and 22 were women. The extent of tumor resection was evaluated on Gd-enhanced axial T1-weighted magnetic resonance imaging obtained 1 month after surgery. All patients received radiotherapy to limited fields (2 Gy per fraction, once a day, 5 days a week, 60 Gy total dose) and adjuvant TMZ after surgery.22
In order to minimize contamination by normal cells, the tumor areas selected for DNA and mRNA extraction contained at least 80% disease-specific cells. Overall survival (OS) was calculated from the date of surgery to death or end of follow-up. Immunohistochemical patterns were established as reported.23 The Ki67 labeling index was defined as the percentage of positive nuclei of a total of 2000 tumor cells counted using an eyepiece grid. The positive nuclei were counted by 2 pathologists (L.M.L. and M.M.) without prior knowledge of the patient prognosis-related information.
Methylation-Specific Polymerase Chain Reaction for ID4
DNA was extracted from three 10 μm-slides from paraffin-embedded tissues using QIAamp DNA mini kit (Qiagen, Milan, Italy), following the manufacturer's protocol. After DNA modification with EpiTect Bisulfite Kits (Qiagen), methylation-specific polymerase chain reaction (MS-PCR) for Id4 was conducted with specific primers for methylated or unmethylated template.13MGMT (O-6-methylguanine-DNA methyltransferase) methylation patterns were determined as described.23 Briefly, bisulfite-modified DNA (100-200 ng) was amplified in a mixture containing 1 × PCR buffer (20 mM Tris [pH 8.3], 50 mM KCl, 1.5 mM MgCl2), deoxynucleotide triphosphates (0.2 mM each), primers (20 pM each), and 0.75 U GoTaq Hot Start polymerase (Promega, Madison, Wis) in a final volume of 25 μL. PCR conditions were: an initial denaturation of 95°C for 8 minutes, followed by 35 cycles of 95°C for 60 seconds, 60°C for 60 seconds, and 72°C for 60 seconds. PCR products were electrophoresed in a 2.5% agarose gel, stained with ethidium bromide, and visualized under ultraviolet illumination. MS-PCR analysis was performed in duplicate for all samples. Normal lymphocyte DNA supermethylated with SssI methyltransferase (New England Biolabs, Beverly, Mass) and treated with bisulfite was used as the unmethylated and methylated control, water as a negative control, and untreated DNA as internal PCR control. We also carried out MS-PCR on granulocyte DNA obtained from 10 healthy individuals, as the control group.
Bisulfite Sequencing for Id4
In order exclude false-positive results due to overestimating and misinterpreting background signals or to an incomplete modification from bisulfite treatment, all Id4 methylated and 15 Id4 unmethylated samples were also examined by promoter sequencing directly or after subcloning. Briefly, after bisulfite modification described above, DNA was amplified with specific primers designed by Chan et al.15 PCR products, encompassing the Id4 promoter region from nucleotides −189 to +128 (the start codon ATG is defined as +1), were purified by agarose gel with MinElute Gel Extraction kit (Qiagen) and thus cloned into a pGEM-T easy vector system (Promega) using JM109 high-efficiency competent cells. Five randomly picked clones were sequenced on an ABI-Prism 3100 Genetic Analyzer (Applied Biosystems, Foster City, Calif). The methylation status of 37 cytosine-guanine dinucleotides (CpG) of the Id4 promoter region was evaluated by sequencing analysis. Normal lymphocyte DNA from 5 healthy donors, treated with bisulfite, was used as unmethylated control.
Real-Time PCR for Id4, MGP, TGF-β1, and VEGF
After being deparaffinized, three 10 μm-slides were digested overnight at 55°C in 200 μL of TENS 1 × (10 mM Tris pH 7.4, 10 mM ethylene diamine tetraacetic acid [EDTA], 100 mM NaCl, 1% SDS) with 100 mg/mL proteinase K, and mRNA was then extracted by RNAsi mini kit (Qiagen), following the manufacturer's protocol. We assessed the quantity and quality of the RNA spectrophometrically (E260, E260/E280 ratio, spectrum 220–320 nm; Biochrom, Cambridge, UK) and by separation on an Agilent 2100 Bioanalyzer (Palo Alto, Calif). RNA was treated with RQ1 RNase-Free deoxyribonuclease (Promega). Levels of Id4, MGP, TGF-β1, and VEGF mRNA were assessed by real-time PCR using SYBR green chemistry in 20 GBMs with methylated Id4 and in 20 GBMs without Id4 methylation. Diluted (1/20) complementary DNA (4 μL) was added to a PCR mix containing 8.4 μL sterile water, 12.5 μL 2 × SYBR mix (Qiagen), and 0.05 μL each of forward and reverse primers (200 mM) to make up a final volume of 25 μL. Cycling conditions were 95°C for 5 minutes, followed by 40 cycles of 95°C for 10 seconds, 60°C for 30 seconds, and 72°C for 30 seconds, and 80 cycles of 55 + 0.5°C per cycle for melting curve analysis in a CFX96 Real-Time PCR Detection System (Bio-Rad, Milan, Italy). Each assay was performed in triplicate, and data were processed by CFX Manager software (Bio-Rad). The average obtained for Id4, MGP, TGF-β1, and VEGF was normalized to the average amount of β-actin for each sample to determine relative changes in mRNA expression. Primers used for Id4, MGP, TGF-β1, VEGF, and β-actin were: Id4 (forward) 5′-CGC TCA CTG CGC TCA ACA C-3′ and (reverse) 5′-CTA ACT TCT GCT CTT CCC CC-3′; MGP (forward) 5′-AGG ACG AAA CCA TGA AGA GC-3′ and (reverse) 5′-CGT TCT CGG ATC CTC TCT TG-3′; TGF-β1 (forward) 5′-AAG TGG ACA TCA ACG GGT TC-3′ and (reverse) 5′-GTC CTT GCG GAA GTC AAT GT-3′; VEGF (forward) 5′-ATG ACG AGG GCC TGG AGT GTG-3′ and (reverse) 5′-CCT ATG TGC TGG CCT TGG TGA-3′; β-actin (forward) 5′-AGC ACT GTG TTG GCG TAC AG-3′ and (reverse) 5′-AGA GCT ACG AGC TGC CTG AC-3′.
Immunohistochemistry for Id4 and von Willebrand Factor
Immunohistochemical analysis was performed on all GBM samples. Briefly, formalin-fixed, paraffin-embedded sections (4 μm thick) were mounted on positive-charged glass slides. For antigen retrieval, deparaffinized and rehydrated sections were treated with citric acid buffer (pH 6.0) for anti-Id4 or with EDTA buffer (pH 8.0) for von Willebrand factor, 2 cycles of 5 minutes each at 750 W, followed by inhibition of endogenous peroxidase with 3% H2O2 for 5 minutes. Then, the sections were incubated overnight at 4°C with rabbit monoclonal antibody anti-Id4 (1:500 dilution; Biocheck, Foster City, Calif) or rabbit polyclonal anti-human von Willebrand factor (1:400 dilution; Dako, Glostrup, Denmark). The primary antibodies were visualized using the avidin–biotin–peroxidase complex method (UltraTek HRP Anti-polyvalent; ScyTek, Logan, Utah) according to the instruction manual. 3,3′-Diaminobenzidine was used as the enzyme substrate to observe the specific antibody localization, and Mayer hematoxylin was used as a nuclear counterstain. In surgical specimens where an en bloc tumor resection was performed, regions of normal brain, which included both the cortex and white matter, were used as internal control. Negative controls were tumor sections stained in the absence of the primary antibody. Positive controls were breast carcinoma samples for Id4 and normal tonsil for von Willebrand factor. Samples were stained more than once, and the results were highly reproducible.
The staining intensity of tissue slides was evaluated independently by 2 observers (L.M.L., M.M.) who were blinded to the patient's characteristics and survival. Cases with disagreement were discussed using a multiheaded microscope until agreement was achieved. To assess differences in staining intensity, an immunoreactivity scoring system was applied. Intensity of staining was classified by both the percentages of the cells stained and the intensity of the staining.24 In this way, the final scores of 0 to 3 were obtained (0, negative; 1, weak; 2, moderate; 3, strong). Only cells with moderate to strong nuclear staining were interpreted as positive. Microvessel density (MVD) of GBMs (n = 55) was measured according to the method described by Leon et al with few modifications.25 Briefly, in areas of most intense neovascularization, individual microvessel counts were made on a ×200 magnification field (equivalent to 0.7386 mm2). Any endothelial cell or endothelial cell cluster was considered as a single countable microvessel. MVD was expressed as the mean number of microvessels per field from 3 highly vascularized areas in each case.
Clinical Assessment of Tumor Angiogenesis
Tumor angiogenesis was assessed preoperatively either on Gd-enhanced axial T1-weighted magnetic resonance images (n = 53) or on contrast-enhanced axial computed tomography scans (n = 9), which were retrieved from or imported into a dedicated software program (Carestream Health Italia, Genova, Italy). In each patient, the relative enhancing tumor ratio (rHTR) was calculated as the ratio between the enhancing area and total tumor area.26
Statistical analysis was performed using GraphPad Prism 5 software (GraphPad Software, La Jolla, Calif) and MedCalc, version 10.2.0.0 (MedCalc Software, Mariakerke, Belgium). Kaplan-Meier survival curves were plotted, and differences in survival between groups of patients were compared using the log-rank test. Statistical comparison of continuous variables was performed by the Mann-Whitney U test, as appropriate. Comparison of categorical variables was performed by chi-square statistic, using the Fisher's exact test. Multivariate analysis was performed using the Cox proportional hazards model, which was adjusted for the major prognostic factors that included Id4 methylation status, age (≤ 60 years versus > 60 years), Karnofsky performance status (KPS; < 70 versus ≥ 70), MGMT methylation status, extent of surgical resection (total versus partial resection), and Ki67 index (≤ 20% vs > 20%) and recursive partitioning analysis (RPA score III and IV versus RPA score V and VI). P values < 0.05 were considered statistically significant.
Id4 Methylation Status and Expression in GBM
Of 62 patients, the Id4 promoter was hypermethylated in 23 cases (37%) (Fig. 1). Id4 hypermethylation was confirmed by sequencing, demonstrating that almost 85% to 95% of the CpG sites analyzed were methylated. On the contrary, almost 80% to 85% of the CpG sites of unmethylated samples and of normal controls were unmethylated, as determined by sequencing (data not shown).
To determine the influence of hypermethylation on transcriptional silencing, we analyzed the Id4 mRNA and protein expression using quantitative real-time PCR and immunohistochemistry, respectively. We found that the level of Id4 mRNA was significantly lowered by about 5 times in the hypermethylated GBMs as compared with those tumors without hypermethylation (P = .0002, Mann-Whitney U test) (Fig. 2A). This reduction was confirmed by immunohistochemistry which showed negative and weak staining in 16 of 23 cases (69.6%) of Id4 hypermethylated GBMs, whereas the unmethylated GBMs showed moderate and intense nuclear staining in 31 of 39 cases (79.5%) (P = .0003, odds ratio = 8.8; 95% confidence interval [CI] = 2.72–28.84; Fisher's exact test) (Fig. 3). Interestingly, brain regions that were not involved by the tumor showed negative Id4 immunostaining, similar to GBMs with hypermethylation of this gene.
Correlation Between Id4 Methylation, MGP, TGF-β1, and VEGF Expression
To better understand the influence of Id4 on GBM neoangiogenesis, we analyzed the expression of MGP, TGF-β1, and VEGF mRNA in 40 GBMs (20 cases with Id4 hypermethylation and 20 without hypermethylation). Expression of MGP and TGF-β1 mRNA were both significantly reduced by about 2 times in GBMs with methylated Id4 in comparison with those tumors that had unmethylated Id4 (P = .0019 for MGP and P = .0317 for TGF-β1; Mann-Whitney U test) (Fig. 2B,C).
Similarly, when we analyzed the expression of VEGF mRNA, we found that GBMs with Id4 hypermethylation had a significantly lower level of this gene (about 2.5-fold) in respect to GBMs with unmethylated Id4 (P = .0353, Mann-Whitney U test) (Fig. 2D).
Correlation Between Id4 Methylation Status and Tumor Angiogenesis
GBMs with Id4 methylation (n = 24) showed a significantly lower rHTR compared with that of Id4 unmethylated tumors (n = 38; 0.3883 ± 0.3018, mean ± standard deviation, versus 0.6 ± 0.3126; P = .0108, Mann-Whitney U test). Likewise, the MVD in GBMs with unmethylated Id4 showed a significantly higher value in respect to GBMs with hypermethylated Id4 (40.55 ± 51.62, mean ± standard deviation, versus 24.38 ± 19.09; P = .0241, Mann-Whitney U test).
Correlation Between Molecular and Clinical Parameters
Follow-up data were available for all patients. The median OS was 8.5 months among patients with unmethylated Id4 as compared with 16 months among those with Id4 hypermethylation (P = .0006; hazard ratio [HR] = 2.96; 95% CI, 95% CI = 1.59–5.52) (Fig. 4). Moreover, GBMs with a Ki67 index ≤ 20% had a significantly longer OS with respect to GBMs with Ki67 index > 20% (P = .0001; HR = 0.22; 95% CI = 0.10–0.47). MGMT promoter methylation conferred significantly longer OS (P = .0474; HR = 1.9; 95% CI = 1.00–3.59) (Fig. 5A,B).
Among the clinical estimates, age ≤ 60 years, preoperative KPS ≥ 70, and RPA score III and IV were significantly correlated with longer OS (P = .0120; HR = 0.42; 95% CI = 0.21–0.82 for age. P = .0001; HR = 3.95; 95% CI = 1.96–7.93 for KPS. P = .0004; HR = 0.29; 95% CI = 0.15–0.58 for RPA; Fig. 5C–E). Patients whose tumors were totally resected trended to survive longer than those who had undergone partial tumor resection, although this difference was not significant (P = .1554). GBMs showing both Id4 and MGMT methylation had significantly longer OS than tumors with methylated Id4 and unmethylated MGMT (P = .0263; HR = 3.75; 95% CI = 1.17–12.04; Fig. 5F).
When we performed a Cox multivariate analysis considering age, KPS, extent of resection, Ki67 index, MGMT methylation status, RPA score, and Id4 methylation status, we found that KPS ≥ 70 (P = .0271), Ki67 index ≤ 20% (P = .0293), and Id4 hypermethylation (P = .0237) emerged as independent prognostic factors for OS (Table 1).
Table 1. Multivariate Analysis
95% CI of Exp(b)
Abbreviations: b, coefficient estimates; 95% CI, 95% confidence interval; Exp(b), hazard ratio value; SE, standard error for coefficient estimates b.
In this study, we demonstrated for the first time that in a subgroup of GBMs, Id4 is silenced by promoter hypermethylation. We also found that the GBM tumors with Id4 promoter hypermethylation have a significant reduction of mRNA and protein expression in comparison with those tumors without methylation and that the epigenetic silencing of this gene is significantly related to a better prognosis.
Although we used a different antibody, our results (moderate to strong immunostaining in approximately 60% of GBMs) are in accordance with Zeng et al, who described a progressive up-regulation of Id4 in the progression of low-grade astrocytoma to high-grade gliomas with a strong nuclear staining for this protein in about 70% of GBMs.24
Recently, it has been reported that Id4 plays an important role in the neoangiogenic process of GBM. In particular, Kutzontkoski et al demonstrated that this proangiogenic function is due to MGP, a member of a vitamin K–dependent family of proteins that is able to stimulate VEGF expression in endothelial cells through a TGF-β1 mediation.6,7,27 High levels of MGP in GBM have been described by van den Boom et al, who performed microarray and real-time analysis to characterize the gene expression profiling associated with glioma progression.28 Similarly, Kjellman et al found an up-regulation of TGF-β1 protein expression correlated with the increasing grade of glioma malignancy.29 In this article, we confirmed that GBMs with unmethylated Id4 have a significantly higher MGP, TGF-β1, and VEGF expression with respect to those GBMs with Id4 hypermethylation, supporting the hypothesis that the epigenetic silencing of Id4 negatively affects tumor angiogenesis in GBM. This hypothesis was reinforced by the clinical and histological assessment of tumor angiogenesis that show significantly lower rHTR and MVD values in GBMs with hypermethylated Id4 compared with unmethylated ones.
Angiogenesis and tumor cell invasion play a central role in the development and growth of malignant gliomas, contributing to tumor progression and resistance to chemoradiotherapy.3,5,30 In light of these findings, we compared the methylation status of Id4 with clinical outcome and found that GBMs with hypermethyled Id4 were significantly associated with a favorable clinical outcome (P = .0006). Interestingly, in our patient cohort that was homogeneously treated with radiotherapy and TMZ after surgery, the association of hypermethylation of both Id4 and MGMT identified a subgroup of patients with better prognosis.
As demonstrated by a number of scientific reports, neoangiogenesis of GBM is an extremely complex process, in which several pathogenetic mechanisms can be identified. Other than recruiting vessels from outside through specific angiogenic factors released by cancer cells (extrinsic pathway), we and others have recently demonstrated that GBM stem-like cells may directly contribute to tumor vascularization via endothelial differentiation (intrinsic pathway).31,32 Most probably, these 2 pathways are interconnected in a network whose complete cross-talking is thus far yet to be understood. Nevertheless, preclinical and clinical investigations have shown that agents targeting angiogenesis may be active against GBMs. Unfortunately, tumor cells rapidly acquire resistance to antiangiogenic agents probably as a consequence of the complexity of the angiogenic pathway.5,33 Thus, more precise knowledge of the mechanisms regulating GBM angiogenesis are warranted in order to develop effective antiangiogenic therapies.
Although our data require confirmatory studies, this report shows that Id4 methylation is a molecular predictor of longer OS in patients with GBM who are treated with surgery followed by adjuvant radiotherapy and TMZ. This effect may be ascribed to the proangiogenic effect of this gene mediated by MGP and VEGF. The inhibition of this pathway may represent a novel therapeutic target.
We thank Dr. Sara Capodimonti for her technical support.
This work was supported by Ministero della Salute (N.ONC_ORD 15/07) to Dr. Pallini, by Fondi d'Ateneo Università Cattolica (Linea D1) to Drs. Pallini and Larocca, and by Banca d'Italia to Dr. D'Alessandris.