Molecular Subtypes Of Glioblastoma Research Articles

Migrating glioma cells express stem cell markers and give rise to new tumors upon xenografting.

Glioblastoma (GBM) is the most frequent and malignant brain tumor with an overall survival of only 14.6 months. Although these tumors are treated with surgery, radiation and chemotherapy, recurrence is inevitable. A critical population of tumor cells in terms of therapy, the so-called cancer stem cells (CSCs), has been identified in gliomas and many other cancers. These tumor cells have a stem cell-like phenotype and are suggested to be responsible for tumor growth, chemo- and radio-resistance as well as recurrence. However, functional evidence for migrating glioma cells having a stem cell-like phenotype is currently lacking. In the present study, the aim was to characterize the phenotype of migrating tumor cells using a novel migration assay based on serum-free stem cell medium and patient-derived spheroid cultures. The results showed pronounced migration of five different GBM spheroid cultures, but not of the commercial cell line U87MG. An in vitro limiting dilution assay showed preserved but reduced spheroid formation capacity of migrating cells. Orthotopic xenografting in mice showed preserved but reduced tumorigenic capacity. Profiling of mRNAs revealed no significant deregulation of 16 predefined CSC-related genes and the HOX-gene list in migrating cells compared to spheroids. Determination of GBM molecular subtypes revealed that subtypes of spheroids and migrating cells were identical. In conclusion, migrating tumor cells preserve expression of stem cell markers and functional CSC characteristics. Since CSCs are reported to be highly resistant to therapy, these results emphasize that the CSC phenotype should be taken into consideration in future treatment of GBMs.

Extracellular Vesicles in Brain Tumor Progression.

Brain tumors can be viewed as multicellular 'ecosystems' with increasingly recognized cellular complexity and systemic impact. While the emerging diversity of malignant disease entities affecting brain tissues is often described in reference to their signature alterations within the cellular genome and epigenome, arguably these cell-intrinsic changes can be regarded as hardwired adaptations to a variety of cell-extrinsic microenvironmental circumstances. Conversely, oncogenic events influence the microenvironment through their impact on the cellular secretome, including emission of membranous structures known as extracellular vesicles (EVs). EVs serve as unique carriers of bioactive lipids, secretable and non-secretable proteins, mRNA, non-coding RNA, and DNA and constitute pathway(s) of extracellular exit of molecules into the intercellular space, biofluids, and blood. EVs are also highly heterogeneous as reflected in their nomenclature (exosomes, microvesicles, microparticles) attempting to capture their diverse origin, as well as structural, molecular, and functional properties. While EVs may act as a mechanism of molecular expulsion, their non-random uptake by heterologous cellular recipients defines their unique roles in the intercellular communication, horizontal molecular transfer, and biological activity. In the central nervous system, EVs have been implicated as mediators of homeostasis and repair, while in cancer they may act as regulators of cell growth, clonogenicity, angiogenesis, thrombosis, and reciprocal tumor-stromal interactions. EVs produced by specific brain tumor cell types may contain the corresponding oncogenic drivers, such as epidermal growth factor receptor variant III (EGFRvIII) in glioblastoma (and hence are often referred to as 'oncosomes'). Through this mechanism, mutant oncoproteins and nucleic acids may be transferred horizontally between cellular populations altering their individual and collective phenotypes. Oncogenic pathways also impact the emission rates, types, cargo, and biogenesis of EVs, as reflected by preliminary analyses pointing to differences in profiles of EV-regulating genes (vesiculome) between molecular subtypes of glioblastoma, and in other brain tumors. Molecular regulators of vesiculation can also act as oncogenes. These intimate connections suggest the context-specific roles of different EV subsets in the progression of specific brain tumors. Advanced efforts are underway to capture these events through the use of EVs circulating in biofluids as biomarker reservoirs and to guide diagnostic and therapeutic decisions.

ATPS-16IDENTIFICATION OF CLINICALLY RELEVANT DRUGS FOR GLIOBLASTOMA WITH SPECIFIC MOLECULAR SIGNATURE USING CHEMICAL BIOLOGY FINGERPRINTING

Translating molecular subtyping of glioblastoma (GBM) into therapeutically actionable guidance remains an unfulfilled opportunity. A central challenge is to discover therapeutic targets matched to agents (drugs or tool compounds) that are predictably present in subclasses of GBM. Transcriptional patterns in GBM cases in TCGA can identify 12 distinct molecular contexts (mC), where, across a subset of genes and tumors, conditional dependencies of gene expression (signal & echo relationships) consistently emerge. We deployed a panel of 64 patient-derived xenografts from GBM patients whose gene expression profiles allow mapping against the same 12 molecular contexts identified from TCGA cases. These clinically-relevant, preclinical models afford bioinformatics mining for actionable targets as well as chemical library screening for activity against short-term cultures derived from the PDX models. Utilizing a technique we term Chemical Biology Fingerprinting (CBF), we interrogated a series of GBM PDX models using a small chemical library (650 compounds) of clinically relevant anti-cancer agents to uncover context-specific chemovulnerabilities. Preliminary data demonstrated that, mC14 GBM (GBM59, SF7300, GBM116), characterized by mt-P53 and transcriptional similarity to GBM proneural subtype, show distinct vulnerability to Arsenic Trioxide (ATO) as compared to mC4 GBM (GBM91, GBM102), characterized by CHEK2 and NF1 mutations and transcriptional patterns similar to GBM mesenchymal subtype. To clinically-validate an ATO vulnerability signature, we acquired 22 treatment naive archival patient samples, who were part of Phase I/II clinical trial to study efficacy of ATO and Temozolomide (TMZ) in combination with radiation in treatment of high grade gliomas (NCT00275067) and which exhibited varied survival with ATO treatment (91 days to >1000 days), and determined their molecular context classification using whole transcriptome sequencing (RNAseq). In summary, we demonstrate a subclassification of GBM into novel contexts and we also show that these contexts are differentially sensitive to clinically relevant drugs. Supported by NIH U01CA168397.

Imaging patterns predict patient survival and molecular subtype in glioblastoma via machine learning techniques.

MRI characteristics of brain gliomas have been used to predict clinical outcome and molecular tumor characteristics. However, previously reported imaging biomarkers have not been sufficiently accurate or reproducible to enter routine clinical practice and often rely on relatively simple MRI measures. The current study leverages advanced image analysis and machine learning algorithms to identify complex and reproducible imaging patterns predictive of overall survival and molecular subtype in glioblastoma (GB). One hundred five patients with GB were first used to extract approximately 60 diverse features from preoperative multiparametric MRIs. These imaging features were used by a machine learning algorithm to derive imaging predictors of patient survival and molecular subtype. Cross-validation ensured generalizability of these predictors to new patients. Subsequently, the predictors were evaluated in a prospective cohort of 29 new patients. Survival curves yielded a hazard ratio of 10.64 for predicted long versus short survivors. The overall, 3-way (long/medium/short survival) accuracy in the prospective cohort approached 80%. Classification of patients into the 4 molecular subtypes of GB achieved 76% accuracy. By employing machine learning techniques, we were able to demonstrate that imaging patterns are highly predictive of patient survival. Additionally, we found that GB subtypes have distinctive imaging phenotypes. These results reveal that when imaging markers related to infiltration, cell density, microvascularity, and blood-brain barrier compromise are integrated via advanced pattern analysis methods, they form very accurate predictive biomarkers. These predictive markers used solely preoperative images, hence they can significantly augment diagnosis and treatment of GB patients.

Modulation of miR-21 signaling by MPS1 in human glioblastoma.

Monopolar spindle 1 (MPS1) is an essential spindle assembly checkpoint (SAC) kinase involved in determining spindle integrity. Beyond its mitotic functions, it has been implicated in several other signaling pathways. Our earlier studies have elaborated on role of MPS1 in glioblastoma (GBM) radiosensitization. In this study using reverse phase protein arrays (RPPAs), we assessed MPS1 mediated cell signaling pathways and demonstrated that inhibiting MPS1 could upregulate the expression of the tumor suppressor PDCD4 and MSH2 genes, by down regulating micro RNA-21 (miR-21). In GBMs miR-21 expression is significantly elevated and is associated with chemo and radioresistance. Both MPS1 and miR-21 depletion suppressed GBM cell proliferation, whereas, ectopic expression of miR-21 rescued GBM cell growth from MPS1 inhibition. Further, we demonstrate that MPS1 mediates phosphorylation of SMAD3 but not SMAD2 in GBM cells; A possible mechanism behind miR-21 modulation by MPS1. Collectively, our results shed light onto an important role of MPS1 in TGF-β/SMAD signaling via miR-21 regulation. We also, show the prognostic effect of miR-21, PDCD4 and MSH2 levels to patient survival across different GBM molecular subtypes. This scenario in which miR-21 is modulated by MPS1 inhibition may be exploited as a potential target for effective GBM therapy.

SC-22 * GLIOMA STEM CELLS INTERNALIZE BEVACIZUMAB FOUND IN THE PERIVASCULAR NICHE AND TARGET IT FOR RECYCLING OR DEGRADATION

Antiangiogenic therapy shows great promise for treatment of cancer. Bevacizumab is a humanized mAb that blocks VEGF-A, thereby inhibiting angiogenesis. It has received FDA approval for patients with recurrent glioblastoma (GBM); however, approximately 30% of patients are non-responsive and the underlying mechanism for the lack of response is not known. It has been assumed that Bevacizumab solely targets circulating VEGF-A; however, VEGF-A found in the perivascular space of GBM tumors could also be an important target. We hypothesized that Bevacizumab is transcytosed across brain endothelial cells (ECs), gains access to the perivascular niche containing glioma stem cells (GSCs), and that differences in GSC endocytosis and targeting of Bevacizumab for recycling or degradation determines a patients' response. We found that Bevacizumab is internalized by normal and tumor-isolated brain ECs in a time-dependent manner that is partially inhibited by amiloride and stimulated by EGF, suggesting Bevacizumab enters the cells by macropinocytosis. ECs transcytose 62% of Bevacizumab over two-hours. Importantly, we found that 95% of GSCs internalize Bevacizumab in 30 minutes. Internalized Bevacizumab in GSCs can be recycled to the extracellular space or degraded; at 5 minutes we found that Bevacizumab was localized to the Rab4+ fast recycling compartment (67%); whereas at 30 minutes only 7% localized to the Rab4+ compartment and 16% localized to the LAMP1+ compartment (late endosome/lysosome). Furthermore, in an orthotopic xenograft mouse model of GBM administered Bevacizumab, we found a gradient of Bevacizumab extending from the vessel into the tumor and it was readily detected within perivascular tumor cells. Our data indicate that the trafficking of Bevacizumab by perivascular GSCs in GBM may influence Bevacizumab effectiveness in patients. Understanding the mechanism of internalization and trafficking of Bevacizumab in GSCs from different molecular subtypes of GBM may help us understand which patients will be responsive to Bevacizumab therapy.

AI-07 * GLIOBLASTOMA SUBTYPES DO NOT EXPLOIT DIFFERENTIAL ANGIOGENIC SIGNALING PATHWAYS NOR ARE THE VESSELS FUNCTIONALLY DIFFERENT

Glioblastoma is the most aggressive type of primary brain tumor in adults with dismal survival and for which novel therapeutic regimens are warranted. Over the last decade molecular subtypes of glioblastoma related to clinical outcome and response to therapy have been described. Of these subtypes, especially the mesenchymal type has been ascribed high angiogenic activity. The feasibility and efficacy of therapy targeted at the vascular compartment is now studied widely, but in clinical trials evaluating anti-angiogenic therapy molecular subtype analyses were not included. In this study we analyzed the acitivation of angiogenic signaling pathways and the vascular status of molecular subtypes of glioblastoma. Tissue samples of primary glioblastomas of proneural, classical and mesenchymal subtypes, 10 samples per subclasses, were utilized for all experiments. The activation of angiogenic signaling pathways, including the VEGF pathway, the Angiopoietin/TIE-2 pathway, and DLL4-Notch signaling, was determined through quantitative PCR analyses by assessing targets from these pathways. These analyses revealed that no specific pathway activation confined to one of the subtypes could be observed and none of the addressed targets was expressed differentially between the subtypes. Morphometric analyses were performed to determine the vascular status of the molecular subtypes. These analyses were confined to the vital tissue, which was itself subdivided into hypoxic and normoxic regions. Mesenchymal tumors, in comparison to proneural tumors, had a relatively larger hypoxic area and the vessels were larger as was determined by volumetric measurements and Chalkley grid overlays. The number and maturation of vessels, assessed by manual counting, α-smooth muscle actin and Collagen IV staining, was found not to vary between the subtypes. This study provides evidence that mesenchymal tumors have larger vessels, but targets suitable for subtype-specific anti-angiogenic treatment were not discovered.

BI-11 * PROGNOSTIC microRNAS IN MALIGNANT GLIOMA

MicroRNAs are short non-coding RNAs that act as micro-managers of cellular function. Their dysregulation is well documented in many cancers. TCGA microRNA expression data is available to identify the prognostic potential of microRNAs in malignant glioma. Using the LASSO statistical approach for 475 glioblastomas in the TCGA dataset we identified a nine-microRNA signature that predicts overall (p = 2.26e-09) and progression-free survival (p = 9.91e-08) with superior power than conventional molecular markers, including MGMT methylation status. This signature has significance in all molecular subtypes of glioblastoma with the exception of the non-G-CIMP proneural group and validated in a separate validation cohort (p = 4.50E-02). The signature also predicts survival in grade II and III disease (p = 5.20e-03). The great majority of microRNAs in the signature have previously been shown to play a role in glioma biology. In addition, when we compared patient groups at the extremes of survival in grade III astrocytoma (poor prognosis n = 10, survival 48 months) and glioblastoma (poor prognosis n = 14, survival 48 months), we found that microRNA profiles aligned with specific neuro-developmental pathways were associated with outcomes. Specifically, poorer prognosis disease was associated with profiles aligned towards glial-restricted and oligodendrocyte precursor (OP) cells. Furthermore, correlation of microRNA expression patterns of OPs with expression of 39 grade III astrocytomas and 558 glioblastomas predicts survival with high significance (p= 5.50e-06), indicating that the more tightly a grade III or IV tumor correlates with an OP profile, the poorer the outcome. These observations indicate that microRNAs have predictive power in malignant glioma. MicroRNA expression patterns may also reveal important aspects of underlying tumor biology. This data, coupled with the superior stability of microRNAs over mRNAs in clinical specimens, suggests they may represent attractive candidates as novel biomarkers for the disease.

TGF-β is an inducer of ZEB1-dependent mesenchymal transdifferentiation in glioblastoma that is associated with tumor invasion.

Different molecular subtypes of glioblastoma (GBM) have been recently identified, of which the mesenchymal subtype is associated with worst prognoses. Here, we report that transforming growth factor-β (TGF-β) is able to induce a mesenchymal phenotype in GBM that involves activation of SMAD2 and ZEB1, a known transcriptional inducer of mesenchymal transition in epithelial cancers. TGF-β exposure of established and newly generated GBM cell lines was associated with morphological changes, enhanced mesenchymal marker expression, migration and invasion in vitro and in an orthotopic mouse model. TGF-β-induced mesenchymal differentiation and invasive behavior was prevented by chemical inhibition of TGF-β signaling as well as small interfering RNA (siRNA)-dependent silencing of ZEB1. Furthermore, TGF-β-responding and -nonresponding GBM neurospheres were identified in vitro. Interestingly, nonresponding cells displayed already high levels of pSMAD2 and ZEB1 that could not be suppressed by inhibition of TGF-β signaling, suggesting the involvement of yet unknown mechanisms. These different GBM neurospheres formed invasive tumors in mice as well as revealed mesenchymal marker expression in immunohistochemical analyses. Moreover, we also detected distinct zones with overlapping pSMAD2, elevated ZEB1 and mesenchymal marker expression in GBM patient material, suggestive of the induction of local, microenvironment-dependent mesenchymal differentiation. Overall, our findings indicate that GBM cells can acquire mesenchymal features associated with enhanced invasive potential following stimulation by secretory cytokines, such as TGF-β. This property of GBM contributes to heterogeneity in this tumor type and may blur the boundaries between the proposed transcriptional subtypes. Targeting TGF-β or downstream targets like ZEB1 might be of potential benefit in reducing the invasive phenotype of GBM in a subpopulation of patients.

CELL INTRINSIC SEXUAL DIMORPHISM IN THE RB AND P21 PATHWAYS UNDERLIES MALE PREDOMINANCE IN GBM

BACKGROUND: Male prevalence across histological subtypes of brain tumors is well known but unexplained. The effects of sex on human disease frequently result from acute actions of circulating sex hormones on cellular physiology and immune function. However, differences in brain tumor rates between males and females are comparable in prepubertal children, and pre and post-menopausal adults, suggesting that alternate mechanisms may underlie this disparity. Here we show that sex-specific differences exist in thresholds for malignant transformation within the astrocyte lineage and that this can contribute to male predominance in the most common brain tumor, glioblastoma (GBM). METHODS: Analysis of three publically accessible databases was utilized to determine whether sex differences are equally present in each of the four molecular subtypes of GBM. Stepwise transformation was modeled for Mesenchymal GBM in separate male and female murine astrocyte cultures by sequential inactivation of neurofibromin (NF1) and p53. Thresholds for transformation were measured in soft agar assays and with in vivo tumorigenesis assays. The molecular basis for differences in thresholds for transformation between male and female astrocytes was investigated in evaluations of stem cell induction, Rb inactivation and p21 activation. RESULTS: Significant male predominance was consistently observed in Proneural and Mesenchymal subtypes of GBM. We modeled Mesenchymal subtype GBM with combined loss of Nf1 and p53 function through cre-mediated knock-out of Nf1 (Nf1-/-) and expression of a dominant negative p53 construct (DNp53). These changes induced a subpopulation of cells with clonogenic, stem cell function in male but not female astrocytes. Male Nf1-/-,DNp53 astrocytes exhibited significantly greater growth compared to female Nf1-/-,DNp53 astrocytes. In addition, male but not female Nf1-/-,DNp53 astrocytes were transformed with either EGF treatment or in vivo implantation. Male Nf1-/-,DNp53 astrocytes exhibited greater in vivo tumorigencitiy regardless of the sex of the recipient mouse. Induced intracranial tumors were diagnosed as grade 4 astrocytomas based on standard histological features. Enhanced stem cell function and tumorigenicity in male Nf1-/-,DNp53 astrocytes was associated with greater inactivation of Rb and less p21 activation compared to female Nf1-/-,DNp53 astrocytes. CONCLUSIONS: Sex differences in cancer, including GBM can involve cell intrinsic sexual dimorphism in tumor suppressor pathways resulting in different thresholds for transformation in males and females. These results suggest that tumor suppressor and oncogene mechanisms should be evaluated in a sex-specific fashion and that screening and treatment of GBM and other cancers may need to be sex-specific. SECONDARY CATEGORY: n/a.

Molecular Subtypes of Glioblastoma Are Relevant to Lower Grade Glioma

BackgroundGliomas are the most common primary malignant brain tumors in adults with great heterogeneity in histopathology and clinical course. The intent was to evaluate the relevance of known glioblastoma (GBM) expression and methylation based subtypes to grade II and III gliomas (ie. lower grade gliomas).MethodsGene expression array, single nucleotide polymorphism (SNP) array and clinical data were obtained for 228 GBMs and 176 grade II/II gliomas (GII/III) from the publically available Rembrandt dataset. Two additional datasets with IDH1 mutation status were utilized as validation datasets (one publicly available dataset and one newly generated dataset from MD Anderson). Unsupervised clustering was performed and compared to gene expression subtypes assigned using the Verhaak et al 840-gene classifier. The glioma-CpG Island Methylator Phenotype (G-CIMP) was assigned using prediction models by Fine et al.ResultsUnsupervised clustering by gene expression aligned with the Verhaak 840-gene subtype group assignments. GII/IIIs were preferentially assigned to the proneural subtype with IDH1 mutation and G-CIMP. GBMs were evenly distributed among the four subtypes. Proneural, IDH1 mutant, G-CIMP GII/III s had significantly better survival than other molecular subtypes. Only 6% of GBMs were proneural and had either IDH1 mutation or G-CIMP but these tumors had significantly better survival than other GBMs. Copy number changes in chromosomes 1p and 19q were associated with GII/IIIs, while these changes in CDKN2A, PTEN and EGFR were more commonly associated with GBMs.ConclusionsGBM gene-expression and methylation based subtypes are relevant for GII/III s and associate with overall survival differences. A better understanding of the association between these subtypes and GII/IIIs could further knowledge regarding prognosis and mechanisms of glioma progression.

Survival benefit from bevacizumab in newly diagnosed glioblastoma (GBM) according to transcriptional subclasses.

2057 Background: Genome-wide transcriptional studies (TCGA and others) have identified distinct GBM molecular subtypes, but to date this has not translated into prognostic or therapeutic implications. Bevacizumab has emerged as a new treatment option for GBMs, although a survival benefit has yet to be demonstrated in unselected patients (pts). We analyzed outcomes from a prospective phase II trial in newly diagnosed GBM treated with hypofractionated stereotactic radiotherapy (HFSRT) combined with temozolomide and bevacizumab, and correlated with GBM transcriptional subclasses. Methods: Pts with newly diagnosed GBM with tumor volume < 60cc were eligible. Treatment consisted of HFSRT (6x6 Gy to contrast-enhancing tumor and 6x4 Gy to FLAIR hypersignal with dose painting), concomitant with bevacizumab (10 mg/kg Q2 weeks) and temozolomide (75mg/m2 daily), followed by standard adjuvant bevacizumab/ temozolomide. Primary endpoint was 1-y overall survival (OS). To establish TCGA transcriptional subclasses, mRNA from formalin-fixed paraffin-embedded tissue blocks was analyzed with a validated, 151-probe Nanostring gene expression assay. Results: A total of 40 evaluable pts were accrued, achieving a 1-y OS of 90% (95% CI 76-96), 2-y OS of 32% (95%CI 19-47) and median OS of 17.4m. The molecular subclass could be defined in 31 pts, as follows: Mesenchymal: 14 (45%) pts, pro-neural: 8 (26%), classical: 7 (23%), neural: 2 (6%). The pro-neural phenotype was associated with reduced overall survival: median OS of 13.5m vs 21.2m for non pro-neural tumors (univariate: p = 0.015; multivariate: p = 0.003). MGMT promoter methylation did not predict survival (p = 0.13). Conclusions: We provide proof-of-principle that GBM transcriptional classification is biologically and therapeutically relevant, identifying non pro-neural GBMs as the best candidates for bevacizumab treatment. Our findings imply that angiogenesis and tumor invasion mechanisms in proneural tumors may be distinct from other subtypes, and we suggest such pts should not be offered bevacizumab treatment upfront. Future randomized trials focusing on non-proneural tumors may finally demonstrate a survival advantage for bevacizumab in GBM. Clinical trial information: NCT01392209.

Murine cell line model of proneural glioma for evaluation of anti-tumor therapies

Molecular subtypes of glioblastoma (GBM) with distinct alterations have been identified. There is need for reproducible, versatile preclinical models that resemble specific GBM phenotypes to facilitate preclinical testing of novel therapies. We present a cell line-based murine proneural GBM model and characterize its response to radiation therapy. Proneural gliomas were generated by injecting PDGF-IRES-Cre retrovirus into the subcortical white matter of adult mice that harbor floxed tumor suppressors (Pten and p53) and stop-floxed reporters. Primary cell cultures were generated from the retrovirus induced tumors and maintained in vitro for multiple passages. RNA sequencing-based expression profiling of the resulting cell lines was performed. The tumorigenic potential of the cells was assessed by intracranial injection into adult naïve mice from different strains. Tumor growth was assessed by bioluminescence imaging (BLI). BLI for tumor cells and brain slices were obtained and compared to in vivo BLI. Response to whole-brain radiation was assessed in glioma-bearing animals. Intracranial injection of Pdgf(+)Pten(-/-)p53(-/-)luciferase(+) glioma cells led to formation of GBM-like tumors with 100 % efficiency (n = 48) and tumorigenesis was retained for more than 3 generations. The cell lines specifically resembled proneural GBM based on expression profiling by RNA-Seq. Pdgf(+)Pten(-/-)p53(-/-)luciferase(+) cell number correlated with BLI signal. Serial BLI measured tumor growth and correlated with size and location by ex vivo imaging. Moreover, BLI predicted tumor-related mortality with a 93 % risk of death within 5 days following a BLI signal between 1 × 10(8) and 5 × 10(8) photons/s cm(2). BLI signal had transient but significant response following radiotherapy, which corresponded to a modest survival benefit for radiated mice (p < 0.05). Intracranial injection of Pdgf(+)Pten(-/-)p53(-/-)luciferase(+) cells constitutes a novel and highly reproducible model, recapitulating key features of human proneural GBM, and can be used to evaluate tumor-growth and response to therapy.

Unique genome-wide map of TCF4 and STAT3 targets using ChIP-seq reveals their association with new molecular subtypes of glioblastoma

Aberrant activation of beta-catenin/TCF4 and STAT3 signaling in glioblastoma multiforme (GBM) has been reported. However, the molecular mechanisms related to this process are still poorly understood. Genome-wide screening of the binding characteristics of the transcription factors TCF4 and STAT3 in GBM cells was performed by chromatin immunoprecipitation sequencing (ChIP-seq) assay. Hierarchical clustering was used to analyze the association of TCF4 and STAT3 coregulated genes with The Cancer Genome Atlas (TCGA) GBM subtypes (classical, mesenchymal, neural, and proneural). New molecular classification of GBM was proposed and validated in Western and Asian populations. We identified 1250 overlapping putative target genes that were coregulated by TCF4 and STAT3. Further, the coregulated genes had the potential to guide TCGA GBM subtypes. Finally, we proposed a new molecular classification of GBM into 2 subtypes (proneural-like and mesenchymal-like) and showed that the new classification could be applied to both Western and Asian populations. In addition, the GBM response to temozolomide therapy differed depending on its subtype; mesenchymal-like GBM benefited, while there was no benefit for proneural-like GBM. This is the first comprehensive study to combine a ChIP-seq assay of TCF4 and STAT3 and data mining of patient cohorts to derive molecular subtypes of GBM.

Future Directions in Glioblastoma Therapy

The standard of care for both newly diagnosed and recurrent glioblastoma (GBM) patients has changed significantly in the past 10 years. Surgery followed by radiation and concurrent and adjuvant temozolomide is now the well-established standard treatment for newly diagnosed GBM. More recently, bevacizumab has become a mainstay of treatment for recurrent GBM. However, despite these advances and significant improvements in patient outcomes, the management and treatment of GBM patients remains a challenging and frustrating endeavor. Difficulties in interpretation of imaging changes after initial treatment, as well as the effects of antiangiogenic agents like bevacizumab on MRI characteristics, can make even the determination of disease progression complicated in multiple situations. Although a high percentage of patients benefit from antiangiogenic therapy in terms of radiographic response and progression-free survival, the effects of bevacizumab on prolonging overall survival remain controversial. Furthermore, tumor progression after treatment with antiangiogenic agents carries a particularly poor prognosis and there is a general lack of effective therapies for this group of patients. These limitations in terms of standard treatments contrast with a relative wealth of new information regarding the molecular underpinnings of GBM. Data from several large-scale efforts to molecularly profile GBM tumors including The Cancer Genome Atlas (TCGA) project have helped define specific molecular subtypes of GBM with distinct biology and clinical outcomes. These findings are helping to refine our understanding of the molecular heterogeneity and pathogenesis of these tumors and provide a basis for the future development of rational and targeted therapies for specific tumor subtypes.

Abstract 2990: Molecular subtyping of glioblastoma based on isoform level network modeling

Abstract Glioblastoma multiforme (GBM) is the most common primary central nervous system tumor, accounting for nearly 80% of the 22,000 malignant brain tumors diagnosed annually in the United States. Patients who receive a diagnosis of glioblastoma have a dismal prognosis and median survival of 12-14 months following the standard therapy. Therefore, there is a need for improved molecular classification of GBM to identify highly sensitive and specific biomarkers and therapeutic targets. Recent genome wide transcriptome studies, by us and others, have discovered that one gene gives rise to multiple protein isoforms, which could be involved in multiple functional pathways. Based on this model, we hypothesize that the isoform-level expression profiling will lead to significantly improved classification of molecular sub-types and identification of more specific isoform based regulatory modules that are disrupted in each GBM molecular sub-type. Using the GBM data sets from TCGA data portal, we identified 2903 genes, which differ significantly in expression between the normal brain and GBM. In contrast, similar analysis using transcript-level expression showed that the expression of 7639 transcripts (correspond to 4401 genes) was significantly altered in GBMs, suggesting that measuring isoform/mRNA variant expression is more sensitive to identify molecular changes in GBM. Furthermore, we identified 51 genes (e.g. RTN3, DCLK2, PKM2, ACTN1), mostly associated with cellular assembly and organization, whose transcript variants show opposite expression pattern, an information lost in a gene centric approach. Although GBMs have been grouped into four subtypes (Proneural, Neural, Classical, Mesenchymal) based on gene expression and mutation profiling, these sub-types do not show any survival differences. Since we observed higher expression differences at isoform-level, we investigated whether isoform-level expression estimates can improve the grouping in terms of disease prognosis. We observed that isoform level clustering always generated a better survival grouping and effectively grouped the 173 core samples into 4 sub-groups with significant (p=0.024) survival differences between the groups. Interestingly, we identified a subgroup, consisting of selected proneural samples, which exhibits higher survival and is characterized by isoform specific overexpression of oligodendroglial markers including Olig1, Olig2, Sox 4, Sox6, Sox8, and Sox10. Gene level expression fails to identify Olig1, Sox6 and Sox10 as significantly misregulated in this subgroup. Similarly, we identify an isoform specific network in another subgroup that could confer resistance to EGFR tyrosine kinase inhibitor, erlotinib. Our analyses suggests that the quest for genomics based personalized medicine requires molecular classification and systematic understanding of the gene regulatory networks at the isoform level. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 2990. doi:1538-7445.AM2012-2990

Magnetic Resonance Imaging Characteristics Predict Epidermal Growth Factor Receptor Amplification Status in Glioblastoma

Two clinical-molecular glioblastoma subtypes have been described: "primary" glioblastomas arise de novo in older patients and often overexpress epidermal growth factor receptor (EGFR); "secondary" glioblastomas progress from lower-grade tumors in younger patients and commonly have TP53 mutations. EGFR overexpression correlates in experimental gliomas with increased angiogenesis, edema, and invasion. No radiographic predictors of molecular glioblastoma subtype are known. We retrospectively reviewed 75 glioblastomas, classified as TP53-mutated (n=11), EGFR-amplified (n=31), or neither (non-TP53/non-EGFR; n=33). Four variables were derived from preoperative magnetic resonance imaging: (a) T2/T1, the ratio of T2-bright volume to enclosed T1-enhancing volume; (b) percentage of tumor volume that was necrosis; and (c and d) T1 and T2 border sharpness coefficients (BSC), the rates of change in grayscale intensity of adjacent 0.02-cm2 voxels traversing the anterior, posterior, and lateral borders on T1-enhanced and T2 images. Mean T2/T1 was 4.7 for EGFR-amplified glioblastomas, greater than that of TP53-mutated glioblastomas (2.3) or non-TP53/non-EGFR glioblastomas (2.6; P<0.00005). All four tumors with T2/T1>7.2 were EGFR-amplified; 0 of 15 with T2/T1<4.7 underwent gross total resection. The mean T2 BSC of EGFR-amplified glioblastomas was 33.7, less sharp (P<0.0000005) than TP53-mutated (72.2) and non-TP53/non-EGFR glioblastomas (81.2). All 15 glioblastomas with T2 BSC<30.8 were EGFR-amplified. Percentage necrosis and T1 BSC did not differ between glioblastoma subtypes. The increased T2/T1 ratio and decreased T2 BSC in EGFR-overexpressing tumors are the first radiographic distinctions described between glioblastoma molecular subtypes. These findings may reflect increased angiogenesis, edema, and/or invasion in EGFR-overexpressing tumors.

Commentary

Background: Gliomas represent the most common primary brain tumor. Despite recent advances in diagnostic imaging, neurosurgical technique, radiation therapy, and chemotherapy, significant advances in accurate prognosis and improved survival have not been achieved. Nevertheless, new developments in molecular biology could have potential impact on the clinical management of patients with these brain tumors. This review will describe the technological advances being used to enrich the classification of gliomas, present specific studies that have successfully used the new technologies to identify molecular subtypes of glioblastoma, and discuss the implications of such enhanced classification and molecular characterizations for the prediction of therapeutic response and the design of future brain tumor therapies. Results: Innovative techniques using complementary DNA and oligonucleotide microarrays (gene chips), tissue microarrays (tissue chips), and differential immunoabsorption have provided high throughput and potentially comprehensive approaches for the molecular characterization of human gliomas. Alterations of several tumor suppressor genes and oncogenes have already been identified as being critical to glioma transformation and progression. These approaches have led to the subclassification of glioblastoma multiforme into distinct subtypes based on the molecular signatures of the tumors. Conclusions: Classifications of gliomas can now be enhanced with new techniques for comprehensive'molecular characterization. Improved and efficient molecular profiling of brain tumors is advancing diagnosis/prognosis and identifying targets for novel and rational therapeutic approaches. Neurosurgeons and neuro-oncologists should be aware of these new developments so they can better advise and treat their patients.