Gene regulatory network topology governs resistance and treatment escape in glioma stem-like cells.

Poor prognosis and drug resistance in glioblastoma (GBM) can result from cellular heterogeneity and treatment-induced shifts in phenotypic states of tumor cells, including dedifferentiation into glioma stem-like cells (GSCs). This rare tumorigenic cell subpopulation resists temozolomide, undergoes proneural-to-mesenchymal transition (PMT) to evade therapy, and drives recurrence. Through inference of transcriptional regulatory networks (TRNs) of patient-derived GSCs (PD-GSCs) at single-cell resolution, we demonstrate how the topology of transcription factor interaction networks drives distinct trajectories of cell-state transitions in PD-GSCs resistant or susceptible to cytotoxic drug treatment. By experimentally testing predictions based on TRN simulations, we show that drug treatment drives surviving PD-GSCs along a trajectory of intermediate states, exposing vulnerability to potentiated killing by siRNA or a second drug targeting treatment-induced transcriptional programs governing nongenetic cell plasticity. Our findings demonstrate an approach to uncover TRN topology and use it to rationally predict combinatorial treatments that disrupt acquired resistance in GBM.

An atlas of causal and mechanistic drivers of interpatient heterogeneity in glioma.

Grade IV glioma, formerly known as glioblastoma multiforme (GBM) is the most aggressive and lethal type of brain tumor, and its treatment remains challenging in part due to extensive interpatient heterogeneity in disease driving mechanisms and lack of prognostic and predictive biomarkers. Using mechanistic inference of node-edge relationship (MINER), we have analyzed multiomics profiles from 516 patients and constructed an atlas of causal and mechanistic drivers of interpatient heterogeneity in GBM (gbmMINER). The atlas has delineated how 30 driver mutations act in a combinatorial scheme to causally influence a network of regulators (306 transcription factors and 73 miRNAs) of 179 transcriptional "programs", influencing disease progression in patients across 23 disease states. Through extensive testing on independent patient cohorts, we share evidence that a machine learning model trained on activity profiles of programs within gbmMINER significantly augments risk stratification, identifying patients who are super-responders to standard of care and those that would benefit from 2 nd line treatments. In addition to providing mechanistic hypotheses regarding disease prognosis, the activity of programs containing targets of 2 nd line treatments accurately predicted efficacy of 28 drugs in killing glioma stem-like cells from 43 patients. Our findings demonstrate that interpatient heterogeneity manifests from differential activities of transcriptional programs, providing actionable strategies for mechanistically characterizing GBM from a systems perspective and developing better prognostic and predictive biomarkers for personalized medicine.

Gene regulatory network topology governs resistance and treatment escape in glioma stem-like cells.

Poor prognosis and drug resistance in glioblastoma (GBM) can result from cellular heterogeneity and treatment-induced shifts in phenotypic states of tumor cells, including dedifferentiation into glioma stem-like cells (GSCs). This rare tumorigenic cell subpopulation resists temozolomide, undergoes proneural-to-mesenchymal transition (PMT) to evade therapy, and drives recurrence. Through inference of transcriptional regulatory networks (TRNs) of patient-derived GSCs (PD-GSCs) at single-cell resolution, we demonstrate how the topology of transcription factor interaction networks drives distinct trajectories of cell state transitions in PD-GSCs resistant or susceptible to cytotoxic drug treatment. By experimentally testing predictions based on TRN simulations, we show that drug treatment drives surviving PD-GSCs along a trajectory of intermediate states, exposing vulnerability to potentiated killing by siRNA or a second drug targeting treatment-induced transcriptional programs governing non-genetic cell plasticity. Our findings demonstrate an approach to uncover TRN topology and use it to rationally predict combinatorial treatments that disrupts acquired resistance in GBM.

The potential role of human endogenous retrovirus K in glioblastoma.

The most active human endogenous retrovirus K (HERV-K) subtype, HML-2, has been implicated as a driver of oncogenesis in several cancers. However, the presence and function of HML-2 in malignant gliomas has remained unclear. In this issue of the JCI, Shah and colleagues demonstrate HML-2 overexpression in glioblastoma (GBM) and its role in maintaining the cancer stem cell phenotype. Given that stem-like cells are considered responsible for GBM heterogeneity and treatment resistance, targeting the stem cell niche may reduce tumor recurrence and improve clinical outcomes. The findings provide a foundation for future studies to determine whether antiretroviral and/or immunotherapy approaches targeting HML-2 could be used as therapeutics for GBM.

Cell-free plasma microRNAs that identify patients with glioblastoma.

Glioblastoma (GBM) is still one of the most commonly diagnosed advanced stage primary brain tumors. Current treatments for patients with primary GBM (pGBM) are often not effective and a significant proportion of the patients with pGBM recur. The effective treatment options for recurrent GBM (rGBM) are limited and survival outcomes are poor. This retrospective multicenter pilot study aims to determine potential cell-free microRNAs (cfmiRs) that identify patients with pGBM and rGBM tumors. 2,083 miRs were assessed using the HTG miRNA whole transcriptome assay (WTA). CfmiRs detection was compared in pre-operative plasma samples from patients with pGBM (n = 32) and rGBM (n = 13) to control plasma samples from normal healthy donors (n = 73). 265 cfmiRs were found differentially expressed in plasma samples from pGBM patients compared to normal healthy donors (FDR < 0.05). Of those 193 miRs were also detected in pGBM tumor tissues (n = 15). Additionally, we found 179 cfmiRs differentially expressed in rGBM, of which 68 cfmiRs were commonly differentially expressed in pGBM. Using Random Forest algorithm, specific cfmiR classifiers were found in the plasma of pGBM, rGBM, and both pGBM and rGBM combined. Two common cfmiR classifiers, miR-3180-3p and miR-5739, were found in all the comparisons. In receiving operating characteristic (ROC) curves analysis for rGBM miR-3180-3p showed a specificity of 87.7% and a sensitivity of 100% (AUC = 98.5%); while miR-5739 had a specificity of 79.5% and sensitivity of 92.3% (AUC = 90.2%). This study demonstrated that plasma samples from pGBM and rGBM patients have specific miR signatures. CfmiR-3180-3p and cfmiR-5739 have potential utility in diagnosing patients with pGBM and rGBM tumors using a minimally invasive blood assay.

A Systems Approach to Brain Tumor Treatment.

Brain tumors are among the most lethal tumors. Glioblastoma, the most frequent primary brain tumor in adults, has a median survival time of approximately 15 months after diagnosis or a five-year survival rate of 10%; the recurrence rate is nearly 90%. Unfortunately, this prognosis has not improved for several decades. The lack of progress in the treatment of brain tumors has been attributed to their high rate of primary therapy resistance. Challenges such as pronounced inter-patient variability, intratumoral heterogeneity, and drug delivery across the blood-brain barrier hinder progress. A comprehensive, multiscale understanding of the disease, from the molecular to the whole tumor level, is needed to address the intratumor heterogeneity resulting from the coexistence of a diversity of neoplastic and non-neoplastic cell types in the tumor tissue. By contrast, inter-patient variability must be addressed by subtyping brain tumors to stratify patients and identify the best-matched drug(s) and therapies for a particular patient or cohort of patients. Accomplishing these diverse tasks will require a new framework, one involving a systems perspective in assessing the immense complexity of brain tumors. This would in turn entail a shift in how clinical medicine interfaces with the rapidly advancing high-throughput (HTP) technologies that have enabled the omics-scale profiling of molecular features of brain tumors from the single-cell to the tissue level. However, several gaps must be closed before such a framework can fulfill the promise of precision and personalized medicine for brain tumors. Ultimately, the goal is to integrate seamlessly multiscale systems analyses of patient tumors and clinical medicine. Accomplishing this goal would facilitate the rational design of therapeutic strategies matched to the characteristics of patients and their tumors. Here, we discuss some of the technologies, methodologies, and computational tools that will facilitate the realization of this vision to practice.

Brain metastasis DNA methylomes, a novel resource for the identification of biological and clinical features.

Brain metastases (BM) are one the most lethal and poorly managed clinical complications in cancer patients. These secondary tumors represent the most common intracranial neoplasm in adults, most frequently originating from lung cancer, breast cancer, and cutaneous melanoma. In primary brain tumors, such as gliomas, recent advances in DNA methylation profiling have allowed for a comprehensive molecular classification. Such data provide prognostic information, in addition to helping predict patient response to specific systemic therapies. However, epigenetic alterations of metastatic brain tumors with specific biological and translational relevance still require much further exploration. Using the widely employed Illumina Infinium HumanMethylation 450K platform, we have generated a cohort of genome-wide DNA methylomes from ninety-six needle-dissected BM specimens from patients with lung cancer, breast cancer, and cutaneous melanoma with clinical, pathological, and demographic annotations. This resource offers an unprecedented and unique opportunity to identify novel DNA methylation features influencing the behavior of brain metastasis, and thus accelerate the discovery of BM-specific theranostic epigenetic alterations.

Epigenetic profiling for the molecular classification of metastatic brain tumors.

Optimal treatment of brain metastases is often hindered by limitations in diagnostic capabilities. To meet this challenge, here we profile DNA methylomes of the three most frequent types of brain metastases: melanoma, breast, and lung cancers (n = 96). Using supervised machine learning and integration of DNA methylomes from normal, primary, and metastatic tumor specimens (n = 1860), we unravel epigenetic signatures specific to each type of metastatic brain tumor and constructed a three-step DNA methylation-based classifier (BrainMETH) that categorizes brain metastases according to the tissue of origin and therapeutically relevant subtypes. BrainMETH predictions are supported by routine histopathologic evaluation. We further characterize and validate the most predictive genomic regions in a large cohort of brain tumors (n = 165) using quantitative-methylation-specific PCR. Our study highlights the importance of brain tumor-defining epigenetic alterations, which can be utilized to further develop DNA methylation profiling as a critical tool in the histomolecular stratification of patients with brain metastases.

Modified carbazoles destabilize microtubules and kill glioblastoma multiform cells.

Small molecules that target microtubules (MTs) represent promising therapeutics to treat certain types of cancer, including glioblastoma multiform (GBM). We synthesized modified carbazoles and evaluated their antitumor activity in GBM cells in culture. Modified carbazoles with an ethyl moiety linked to the nitrogen of the carbazole and a carbonyl moiety linked to distinct biaromatic rings exhibited remarkably different killing activities in human GBM cell lines and patient-derived GBM cells, with IC50 values from 67 to >10,000 nM. Measures of the activity of modified carbazoles with tubulin and microtubules coupled to molecular docking studies show that these compounds bind to the colchicine site of tubulin in a unique low interaction space that inhibits tubulin assembly. The modified carbazoles reported here represent novel chemical tools to better understand how small molecules disrupt MT functions and kill devastating cancers such as GBM.

An anatomic transcriptional atlas of human glioblastoma.

Glioblastoma is an aggressive brain tumor that carries a poor prognosis. The tumor's molecular and cellular landscapes are complex, and their relationships to histologic features routinely used for diagnosis are unclear. We present the Ivy Glioblastoma Atlas, an anatomically based transcriptional atlas of human glioblastoma that aligns individual histologic features with genomic alterations and gene expression patterns, thus assigning molecular information to the most important morphologic hallmarks of the tumor. The atlas and its clinical and genomic database are freely accessible online data resources that will serve as a valuable platform for future investigations of glioblastoma pathogenesis, diagnosis, and treatment.

Precision knockdown of EGFR gene expression using radio frequency electromagnetic energy.

Electromagnetic fields (EMF) in the radio frequency energy (RFE) range can affect cells at the molecular level. Here we report a technology that can record the specific RFE signal of a given molecule, in this case the siRNA of epidermal growth factor receptor (EGFR). We demonstrate that cells exposed to this EGFR siRNA RFE signal have a 30-70% reduction of EGFR mRNA expression and ~60% reduction in EGFR protein expression vs. control treated cells. Specificity for EGFR siRNA effect was confirmed via RNA microarray and antibody dot blot array. The EGFR siRNA RFE decreased cell viability, as measured by Calcein-AM measures, LDH release and Caspase 3 cleavage, and increased orthotopic xenograft survival. The outcomes of this study demonstrate that an RFE signal can induce a specific siRNA-like effect on cells. This technology opens vast possibilities of targeting a broader range of molecules with applications in medicine, agriculture and other areas.

TGFß-Responsive HMOX1 Expression Is Associated with Stemness and Invasion in Glioblastoma Multiforme.

Glioblastoma multiforme (GBM) is the most common and lethal adult brain tumor. Resistance to standard radiation and chemotherapy is thought to involve survival of GBM cancer stem cells (CSCs). To date, no single marker for identifying GBM CSCs has been able to capture the diversity of CSC populations, justifying the needs for additional CSC markers for better characterization. Employing targeted mass spectrometry, here we present five cell-surface markers HMOX1, SLC16A1, CADM1, SCAMP3, and CLCC1 which were found to be elevated in CSCs relative to healthy neural stem cells (NSCs). Transcriptomic analyses of REMBRANDT and TCGA compendiums also indicated elevated expression of these markers in GBM relative to controls and non-GBM diseases. Two markers SLC16A1 and HMOX1 were found to be expressed among pseudopalisading cells that reside in the hypoxic region of GBM, substantiating the histopathological hallmarks of GBM. In a prospective study (N = 8) we confirmed the surface expression of HMOX1 on freshly isolated primary GBM cells (P0). Employing functional assays that are known to evaluate stemness, we demonstrate that elevated HMOX1 expression is associated with stemness in GBM and can be modulated through TGFß. siRNA-mediated silencing of HMOX1 impaired GBM invasion-a phenomenon related to poor prognosis. In addition, surgical resection of GBM tumors caused declines (18% ± 5.1SEM) in the level of plasma HMOX1 as measured by ELISA, in 8/10 GBM patients. These findings indicate that HMOX1 is a robust predictor of GBM CSC stemness and pathogenesis. Further understanding of the role of HMOX1 in GBM may uncover novel therapeutic approaches. Stem Cells 2016;34:2276-2289.

Tamoxifen improves cytopathic effect of oncolytic adenovirus in primary glioblastoma cells mediated through autophagy.

Oncolytic gene therapy using viral vectors may provide an attractive therapeutic option for malignant gliomas. These viral vectors are designed in a way to selectively target tumor cells and spare healthy cells. To determine the translational impact, it is imperative to assess the factors that interfere with the anti-glioma effects of the oncolytic adenoviral vectors. In the current study, we evaluated the efficacy of survivin-driven oncolytic adenoviruses pseudotyping with adenoviral fiber knob belonging to the adenoviral serotype 3, 11 and 35 in their ability to kill glioblastoma (GBM) cells selectively without affecting normal cells. Our results indicate that all recombinant vectors used in the study can effectively target GBM in vitro with high specificity, especially the 3 knob-modified vector. Using intracranial U87 and U251 GBM xenograft models we have also demonstrated that treatment with Conditionally Replicative Adenovirus (CRAd-S-5/3) vectors can effectively regress tumor. However, in several patient-derived GBM cell lines, cells exhibited resistance to the CRAd infection as evident from the diminishing effects of autophagy. To improve therapeutic response, tumor cells were pretreated with tamoxifen. Our preliminary data suggest that tamoxifen sensitizes glioblastoma cells towards oncolytic treatment with CRAd-S-5/3, which may prove useful for GBM in future experimental therapy.

Global analysis of H3K4me3 and H3K27me3 profiles in glioblastoma stem cells and identification of SLC17A7 as a bivalent tumor suppressor gene.

Epigenetic changes, including H3K4me3 and H3K27me3 histone modification, play an important role in carcinogenesis. However, no genome-wide histone modification map has been generated for gliomas. Here, we report a genome-wide map of H3K4me3 and H3K27me3 histone modifications for 8 glioma stem cell (GSC) lines, together with the associated gene activation or repression patterns. In addition, we compared the genome-wide histone modification maps of GSC lines to those of astrocytes to identify unique gene activation or repression profiles in GSCs and astrocytes. We also identified a set of bivalent genes, which are genes that are associated with both H3K4me3 and H3K27me3 marks and are poised for action in embryonic stem cells. These bivalent genes are potential targets for inducing differentiation in glioblastoma (GBM) as a therapeutic approach. Finally, we identified SLC17A7 as a bivalent tumor suppressor gene in GBM, as it is down-regulated at both the protein and RNA levels in GBM tissues compared with normal brain tissues, and it inhibits GBM cell proliferation, migration and invasion.

VEGFR inhibitors upregulate CXCR4 in VEGF receptor-expressing glioblastoma in a TGFßR signaling-dependent manner.

The failure of standard treatment for patients diagnosed with glioblastoma (GBM) coupled with the highly vascularized nature of this solid tumor has led to the consideration of agents targeting VEGF or VEGFRs, as alternative therapeutic strategies for this disease. Despite modest achievements in survival obtained with such treatments, failure to maintain an enduring survival benefit and more invasive relapsing tumors are evident. Our study suggests a potential mechanism by which anti-VEGF/VEGFR therapies regulate the enhanced invasive phenotype through a pathway that involves TGFßR and CXCR4. VEGFR signaling inhibitors (Cediranib and Vandetanib) elevated the expression of CXCR4 in VEGFR-expressing GBM cell lines and tumors, and enhanced the in vitro migration of these lines toward CXCL12. The combination of VEGFR inhibitor and CXCR4 antagonist provided a greater survival benefit to tumor-bearing animals. The upregulation of CXCR4 by VEGFR inhibitors was dependent on TGFß/TGFßR, but not HGF/MET, signaling activity, suggesting a mechanism of crosstalk among VEGF/VEGFR, TGFß/TGFßR, and CXCL12/CXCR4 pathways in the malignant phenotype of recurrent tumors after anti-VEGF/VEGFR therapies. Thus, the combination of VEGFR, CXCR4, and TGFßR inhibitors could provide an alternative strategy to halt GBM progression.

Expression and functional heterogeneity of chemokine receptors CXCR4 and CXCR7 in primary patient-derived glioblastoma cells.

Glioblastoma (GBM) is the most common primary brain tumor in adults. The poor prognosis and minimally successful treatments of these tumors indicates a need to identify new therapeutic targets. Therapy resistance of GBMs is attributed to heterogeneity of the glioblastoma due to genetic alterations and functional subpopulations. Chemokine receptors CXCR4 and CXCR7 play important roles in progression of various cancers although the specific functions of the CXCL12-CXCR4-CXCR7 axis in GBM are less characterized. In this study we examined the expression and function of CXCR4 and CXCR7 in four primary patient-derived GBM cell lines of the proliferative subclass, investigating their roles in in vitro growth, migration, sphere and tube formation. CXCR4 and CXCR7 cell surface expression was heterogeneous both between and within each cell line examined, which was not reflected by RT-PCR analysis. Variable percentages of CXCR4+CXCR7- (CXCR4 single positive), CXCR4-CXCR7+ (CXCR7 single positive), CXCR4+CXCR7+ (double positive), and CXCR4-CXCR7- (double negative) subpopulations were evident across the lines examined. A subpopulation of slow cell cycling cells was enriched in CXCR4 and CXCR7. CXCR4+, CXCR7+, and CXCR4+/CXCR7+ subpopulations were able to initiate intracranial tumors in vivo. CXCL12 stimulated in vitro cell growth, migration, sphere formation and tube formation in some lines and, depending on the response, the effects were mediated by either CXCR4 or CXCR7. Collectively, our results indicate a high level of heterogeneity in both the surface expression and functions of CXCR4 and CXCR7 in primary human GBM cells of the proliferative subclass. Should targeting of CXCR4 and CXCR7 provide clinical benefits to GBM patients, a personalized treatment approach should be considered given the differential expression and functions of these receptors in GBM.

High-throughput chemical screens identify disulfiram as an inhibitor of human glioblastoma stem cells.

Glioblastoma Multiforme (GBM) continues to have a poor patient prognosis despite optimal standard of care. Glioma stem cells (GSCs) have been implicated as the presumed cause of tumor recurrence and resistance to therapy. With this in mind, we screened a diverse chemical library of 2,000 compounds to identify therapeutic agents that inhibit GSC proliferation and therefore have the potential to extend patient survival. High-throughput screens (HTS) identified 78 compounds that repeatedly inhibited cellular proliferation, of which 47 are clinically approved for other indications and 31 are experimental drugs. Several compounds (such as digitoxin, deguelin, patulin and phenethyl caffeate) exhibited high cytotoxicity, with half maximal inhibitory concentrations (IC50) in the low nanomolar range. In particular, the FDA approved drug for the treatment of alcoholism, disulfiram (DSF), was significantly potent across multiple patient samples (IC50 of 31.1 nM). The activity of DSF was potentiated by copper (Cu), which markedly increased GSC death. DSF-Cu inhibited the chymotrypsin-like proteasomal activity in cultured GSCs, consistent with inactivation of the ubiquitin-proteasome pathway and the subsequent induction of tumor cell death. Given that DSF is a relatively non-toxic drug that can penetrate the blood-brain barrier, we suggest that DSF should be tested (as either a monotherapy or as an adjuvant) in pre-clinical models of human GBM. Data also support targeting of the ubiquitin-proteasome pathway as a therapeutic approach in the treatment of GBM.

Pressure effects on enzyme-catalyzed quantum tunneling events arise from protein-specific structural and dynamic changes.

The rate and kinetic isotope effect (KIE) on proton transfer during the aromatic amine dehydrogenase-catalyzed reaction with phenylethylamine shows complex pressure and temperature dependences. We are able to rationalize these effects within an environmentally coupled tunneling model based on constant pressure molecular dynamics (MD) simulations. As pressure appears to act anisotropically on the enzyme, perturbation of the reaction coordinate (donor-acceptor compression) is, in this case, marginal. Therefore, while we have previously demonstrated that pressure and temperature dependences can be used to infer H-tunneling and the involvement of promoting vibrations, these effects should not be used in the absence of atomistic insight, as they can vary greatly for different enzymes. We show that a pressure-dependent KIE is not a definitive hallmark of quantum mechanical H-tunneling during an enzyme-catalyzed reaction and that pressure-independent KIEs cannot be used to exclude tunneling contributions or a role for promoting vibrations in the enzyme-catalyzed reaction. We conclude that coupling of MD calculations with experimental rate and KIE studies is required to provide atomistic understanding of pressure effects in enzyme-catalyzed reactions.

Driving force analysis of proton tunnelling across a reactivity series for an enzyme-substrate complex.

Quantitative structure-activity relationships are widely used to probe C-H bond breakage by quinoprotein enzymes. However, we showed recently that p-substituted benzylamines are poor reactivity probes for the quinoprotein aromatic amine dehydrogenase (AADH) because of a requirement for structural change in the enzyme-substrate complex prior to C-H bond breakage. This rearrangement is partially rate limiting, which leads to deflated kinetic isotope effects for p-substituted benzylamines. Here we report reactivity (driving force) studies of AADH with p-substituted phenylethylamines for which the kinetic isotope effect (approximately 16) accompanying C-H/C-(2)H bond breakage is elevated above the semi-classical limit. We show bond breakage occurs by quantum tunnelling and that within the context of the environmentally coupled framework for H-tunnelling the presence of the p-substituent places greater demand on the apparent need for fast promoting motions. The crystal structure of AADH soaked with phenylethylamine or methoxyphenylethylamine indicates that the structural change identified with p-substituted benzylamines should not limit the reaction with p-substituted phenylethylamines. This is consistent with the elevated kinetic isotope effects measured with p-substituted phenylethylamines. We find a good correlation in the rate constant for proton transfer with bond dissociation energy for the reactive C-H bond, consistent with a rate that is limited by a Marcus-like tunnelling mechanism. As the driving force becomes larger, the rate of proton transfer increases while the Marcus activation energy becomes smaller. This is the first experimental report of the driving force perturbation of H-tunnelling in enzymes using a series of related substrates. Our study provides further support for proton tunnelling in AADH.

Correction of pre-steady-state KIEs for isotopic impurities and the consequences of kinetic isotope fractionation.

We show, both experimentally and by kinetic modeling, that enzymatic single-turnover (pre-steady-state) H-transfer reactions can be significantly complicated by kinetic isotope fractionation. This fractionation results in the formation of more protiated than deuterated product and is a unique problem for pre-steady-state reactions. When observed rate constants are measured using rapid-mixing (e.g., stopped flow) methodologies, kinetic isotope fractionation can lead to a large underestimation of both the magnitude and temperature dependence of kinetic isotope effects (KIEs). This fractionation is related to the isotopic purity of the substrates used and highlights a major problem with experimental studies which measure KIEs with substrates that are not isotopically pure. As it is not always possible to prepare isotopically pure substrates, we describe two general methods for the correction, for known isotope impurities, of KIEs calculated from pre-steady-state measurements.