Susie Ahn

Epigenetic-Mediated Dysfunction of the Bone Morphogenetic Protein Pathway Inhibits Differentiation of Glioblastoma-Initiating Cells

Despite similarities between tumor-initiating cells with stem-like properties (TICs) and normal neural stem cells, we hypothesized that there may be differences in their differentiation potentials. We now demonstrate that both bone morphogenetic protein (BMP)-mediated and ciliary neurotrophic factor (CNTF)-mediated Jak/STAT-dependent astroglial differentiation is impaired due to EZH2-dependent epigenetic silencing of BMP receptor 1B (BMPR1B) in a subset of glioblastoma TICs. Forced expression of BMPR1B either by transgene expression or demethylation of the promoter restores their differentiation capabilities and induces loss of their tumorigenicity. We propose that deregulation of the BMP developmental pathway in a subset of glioblastoma TICs contributes to their tumorigenicity both by desensitizing TICs to normal differentiation cues and by converting otherwise cytostatic signals to proproliferative signals.

Identification of competence pheromone responsive genes in Streptococcus pneumoniae by use of DNA microarrays

Natural genetic transformation in Streptococcus pneumoniae is controlled in part by a quorum‐sensing system mediated by a peptide pheromone called competence‐stimulating peptide (CSP), which acts to coordinate transient activation of genes required for competence. To characterize the transcriptional response and regulatory events occurring when cells are exposed to competence pheromone, we constructed DNA microarrays and analysed the temporal expression profiles of 1817 among the 2129 unique predicted open reading frames present in the S. pneumoniae TIGR4 genome (84%). After CSP stimulation, responsive genes exhibited four temporally distinct expression profiles: early, late and delayed gene induction, and gene repression. At least eight early genes participate in competence regulation including comX, which encodes an alternative sigma factor. Late genes were dependent on ComX for CSP‐induced expression, many playing important roles in transformation. Genes in the delayed class (third temporal wave) appear to be stress related. Genes repressed during the CSP response include ribosomal protein loci and other genes involved in protein synthesis. This study increased the number of identified CSP‐responsive genes from approximately 40 to 188. Given the relatively large number of induced genes (6% of the genome), it was of interest to determine which genes provide functions essential to transformation. Many of the induced loci were subjected to gene disruption mutagenesis, allowing us to establish that among 124 CSP‐inducible genes, 67 were individually dispensable for transformation, whereas 23 were required for transformation.

Unsupervised Analysis of Transcriptomic Profiles Reveals Six Glioma Subtypes

Gliomas are the most common type of primary brain tumors in adults and a significant cause of cancer-related mortality. Defining glioma subtypes based on objective genetic and molecular signatures may allow for a more rational, patient-specific approach to therapy in the future. Classifications based on gene expression data have been attempted in the past with varying success and with only some concordance between studies, possibly due to inherent bias that can be introduced through the use of analytic methodologies that make a priori selection of genes before classification. To overcome this potential source of bias, we have applied two unsupervised machine learning methods to genome-wide gene expression profiles of 159 gliomas, thereby establishing a robust glioma classification model relying only on the molecular data. The model predicts for two major groups of gliomas (oligodendroglioma-rich and glioblastoma-rich groups) separable into six hierarchically nested subtypes. We then identified six sets of classifiers that can be used to assign any given glioma to the corresponding subtype and validated these classifiers using both internal (189 additional independent samples) and two external data sets (341 patients). Application of the classification system to the external glioma data sets allowed us to identify previously unrecognized prognostic groups within previously published data and within The Cancer Genome Atlas glioblastoma samples and the different biological pathways associated with the different glioma subtypes offering a potential clue to the pathogenesis and possibly therapeutic targets for tumors within each subtype.

Prediction of Associations between microRNAs and Gene Expression in Glioma Biology

Despite progress in the determination of miR interactions, their regulatory role in cancer is only beginning to be unraveled. Utilizing gene expression data from 27 glioblastoma samples we found that the mere knowledge of physical interactions between specific mRNAs and miRs can be used to determine associated regulatory interactions, allowing us to identify 626 associated interactions, involving 128 miRs that putatively modulate the expression of 246 mRNAs. Experimentally determining the expression of miRs, we found an over-representation of over(under)-expressed miRs with various predicted mRNA target sequences. Such significantly associated miRs that putatively bind over-expressed genes strongly tend to have binding sites nearby the 3′UTR of the corresponding mRNAs, suggesting that the presence of the miRs near the translation stop site may be a factor in their regulatory ability. Our analysis predicted a significant association between miR-128 and the protein kinase WEE1, which we subsequently validated experimentally by showing that the over-expression of the naturally under-expressed miR-128 in glioma cells resulted in the inhibition of WEE1 in glioblastoma cells.

Predicting in vitro drug sensitivity using Random Forests

MOTIVATION Panels of cell lines such as the NCI-60 have long been used to test drug candidates for their ability to inhibit proliferation. Predictive models of in vitro drug sensitivity have previously been constructed using gene expression signatures generated from gene expression microarrays. These statistical models allow the prediction of drug response for cell lines not in the original NCI-60. We improve on existing techniques by developing a novel multistep algorithm that builds regression models of drug response using Random Forest, an ensemble approach based on classification and regression trees (CART). RESULTS This method proved successful in predicting drug response for both a panel of 19 Breast Cancer and 7 Glioma cell lines, outperformed other methods based on differential gene expression, and has general utility for any application that seeks to relate gene expression data to a continuous output variable. IMPLEMENTATION Software was written in the R language and will be available together with associated gene expression and drug response data as the package ivDrug at http://r-forge.r-project.org.

Correlation Analysis between Single-Nucleotide Polymorphism and Expression Arrays in Gliomas Identifies Potentially Relevant Target Genes

Primary brain tumors are a major cause of cancer mortality in the United States. Therapy for gliomas, the most common type of primary brain tumors, remains suboptimal. The development of improved therapeutics will require greater knowledge of the biology of gliomas at both the genomic and transcriptional levels. We have previously reported whole genome profiling of chromosome copy number alterations (CNA) in gliomas, and now present our findings on how those changes may affect transcription of genes that may be involved in tumor induction and progression. By calculating correlation values of mRNA expression versus DNA copy number average in a moving window around a given RNA probe set, biologically relevant information can be gained that is obscured by the analysis of a single data type. Correlation coefficients ranged from -0.6 to 0.7, highly significant when compared with previous studies. Most correlated genes are located on chromosomes 1, 7, 9, 10, 13, 14, 19, 20, and 22, chromosomes known to have genomic alterations in gliomas. Additionally, we were able to identify CNAs whose gene expression correlation suggests possible epigenetic regulation. This analysis revealed a number of interesting candidates such as CXCL12, PTER, and LRRN6C, among others. The results have been verified using real-time PCR and methylation sequencing assays. These data will further help differentiate genes involved in the induction and/or maintenance of the tumorigenic process from those that are mere passenger mutations, thereby enriching for a population of potentially new therapeutic molecular targets.

G-CIMP Status Prediction of Glioblastoma Samples Using mRNA Expression Data

Glioblastoma Multiforme (GBM) is a tumor with high mortality and no known cure. The dramatic molecular and clinical heterogeneity seen in this tumor has led to attempts to define genetically similar subgroups of GBM with the hope of developing tumor specific therapies targeted to the unique biology within each of these subgroups. Recently, a subset of relatively favorable prognosis GBMs has been identified. These glioma CpG island methylator phenotype, or G-CIMP tumors, have distinct genomic copy number aberrations, DNA methylation patterns, and (mRNA) expression profiles compared to other GBMs. While the standard method for identifying G-CIMP tumors is based on genome-wide DNA methylation data, such data is often not available compared to the more widely available gene expression data. In this study, we have developed and evaluated a method to predict the G-CIMP status of GBM samples based solely on gene expression data.

Identification of molecular pathways facilitating glioma cell invasion in situ.

Gliomas are mostly incurable secondary to their diffuse infiltrative nature. Thus, specific therapeutic targeting of invasive glioma cells is an attractive concept. As cells exit the tumor mass and infiltrate brain parenchyma, they closely interact with a changing micro-environmental landscape that sustains tumor cell invasion. In this study, we used a unique microarray profiling approach on a human glioma stem cell (GSC) xenograft model to explore gene expression changes in situ in Invading Glioma Cells (IGCs) compared to tumor core, as well as changes in host cells residing within the infiltrated microenvironment relative to the unaffected cortex. IGCs were found to have reduced expression of genes within the extracellular matrix compartment, and genes involved in cell adhesion, cell polarity and epithelial to mesenchymal transition (EMT) processes. The infiltrated microenvironment showed activation of wound repair and tissue remodeling networks. We confirmed by protein analysis the downregulation of EMT and polarity related genes such as CD44 and PARD3 in IGCs, and EFNB3, a tissue-remodeling agent enriched at the infiltrated microenvironment. OLIG2, a proliferation regulator and glioma progenitor cell marker upregulated in IGCs was found to function in enhancing migration and stemness of GSCs. Overall, our results unveiled a more comprehensive picture of the complex and dynamic cell autonomous and tumor-host interactive pathways of glioma invasion than has been previously demonstrated. This suggests targeting of multiple pathways at the junction of invading tumor and microenvironment as a viable option for glioma therapy.

Micro-Environment Causes Reversible Changes in DNA Methylation and mRNA Expression Profiles in Patient-Derived Glioma Stem Cells

In vitro and in vivo models are widely used in cancer research. Characterizing the similarities and differences between a patients tumor and corresponding in vitro and in vivo models is important for understanding the potential clinical relevance of experimental data generated with these models. Towards this aim, we analyzed the genomic aberrations, DNA methylation and transcriptome profiles of five parental tumors and their matched in vitro isolated glioma stem cell (GSC) lines and xenografts generated from these same GSCs using high-resolution platforms. We observed that the methylation and transcriptome profiles of in vitro GSCs were significantly different from their corresponding xenografts, which were actually more similar to their original parental tumors. This points to the potentially critical role of the brain microenvironment in influencing methylation and transcriptional patterns of GSCs. Consistent with this possibility, ex vivo cultured GSCs isolated from xenografts showed a tendency to return to their initial in vitro states even after a short time in culture, supporting a rapid dynamic adaptation to the in vitro microenvironment. These results show that methylation and transcriptome profiles are highly dependent on the microenvironment and growth in orthotopic sites partially reverse the changes caused by in vitro culturing.

Inositol Polyphosphate-5-Phosphatase F (INPP5F) inhibits STAT3 activity and suppresses gliomas tumorigenicity

Glioblastoma (GBM), the most common type of primary malignant brain tumors harboring a subpopulation of stem-like cells (GSCs), is a fast-growing and often fatal tumor. Signal Transducer and Activator of Transcription 3 (STAT3) is one of the major signaling pathways in GSCs maintenance but the molecular mechanisms underlying STAT3 deregulation in GSCs are poorly defined. Here, we demonstrate that Inositol Polyphosphate-5-Phosphatase F (INPP5F), one of the polyphosphoinositide phosphatases, is differentially expressed in GSCs from glioma patients, and is identified as an inhibitor of STAT3 signaling via interaction with STAT3 and inhibition of its phosphorylation. Constitutively expressed INPP5F showed to suppress self-renewal and proliferation potentials of glioblastoma cells and reduced tumorigenicity of glioblastoma. In addition, loss of INPP5F gene in gliomas is significantly correlated with lower overall patient survivals. These findings suggest that INPP5F is a potential tumor suppressor in gliomas via inhibition of STAT3 pathway, and that deregulation of INPP5F may lead to contribution to gliomagenesis.