Network signatures of survival in glioblastoma multiforme

State-transition modeling of blood transcriptome predicts disease evolution and treatment response in chronic myeloid leukemia

Chronic myeloid leukemia (CML) is initiated and maintained by BCR::ABL which is clinically targeted using tyrosine kinase inhibitors (TKIs). TKIs can induce long-term remission but are also not curative. Thus, CML is an ideal system to test our hypothesis that transcriptome-based state-transition models accurately predict cancer evolution and treatment response. We collected time-sequential blood samples from tetracycline-off (Tet-Off) BCR::ABL-inducible transgenic mice and wild-type controls. From the transcriptome, we constructed a CML state-space and a three-well leukemogenic potential landscape. The potential's stable critical points defined observable disease states. Early states were characterized by anti-CML genes opposing leukemia; late states were characterized by pro-CML genes. Genes with expression patterns shaped similarly to the potential landscape were identified as drivers of disease transition. Re-introduction of tetracycline to silence the BCR::ABL gene returned diseased mice transcriptomes to a near healthy state, without reaching it, suggesting parts of the transition are irreversible. TKI only reverted the transcriptome to an intermediate disease state, without approaching a state of health; disease relapse occurred soon after treatment. Using only the earliest time-point as initial conditions, our state-transition models accurately predicted both disease progression and treatment response, supporting this as a potentially valuable approach to time clinical intervention, before phenotypic changes become detectable.

Exploring the Vital Link Between Glioma, Neuron, and Neural Activity in the Context of Invasion

Because of their ability to infiltrate normal brain tissue, gliomas frequently evade microscopic surgical excision. The histologic infiltrative property of human glioma has been previously characterized as Scherer secondary structures, of which the perivascular satellitosis is a prospective target for anti-angiogenic treatment in high-grade gliomas. However, the mechanisms underlying perineuronal satellitosis remain unclear, and therapy remains lacking. Our knowledge of the mechanism underlying Scherer secondary structures has improved over time. New techniques, such as laser capture microdissection and optogenetic stimulation, have advanced our understanding of glioma invasion mechanisms. Although laser capture microdissection is a useful tool for studying gliomas that infiltrate the normal brain microenvironment, optogenetics and mouse xenograft glioma models have been extensively used in studies demonstrating the unique role of synaptogenesis in glioma proliferation and identification of potential therapeutic targets. Moreover, a rare glioma cell line is established that, when transplanted in the mouse brain, can replicate and recapitulate the human diffuse invasion phenotype. This review discusses the primary molecular causes of glioma, its histopathology-based invasive mechanisms, and the importance of neuronal activity and interactions between glioma cells and neurons in the brain microenvironment. It also explores current methods and models of gliomas.

Immune surveillance in glioblastoma: role of the NKG2D system and novel cell-based therapeutic approaches

Glioblastoma, formerly known as Glioblastoma multiforme (GBM) is the most frequent and most aggressive brain tumor in adults. The brain is an immunopriviledged organ and the blood brain barrier shields the brain from immune surveillance. In this review we discuss the composition of the immunosuppressive tumor micromilieu and potential immune escape mechanisms in GBM. In this respect, we focus on the role of the NKG2D receptor/ligand system. NKG2D ligands are frequently expressed on GBM tumor cells and can activate NKG2D-expressing killer cells including NK cells and γδ T cells. Soluble NKG2D ligands, however, contribute to tumor escape from immunological attack. We also discuss the current immunotherapeutic strategies to improve the survival of GBM patients. Such approaches include the modulation of the NKG2D receptor/ligand system, the application of checkpoint inhibitors, the adoptive transfer of ex vivo expanded and/or modified immune cells, or the application of antibodies and antibody constructs to target cytotoxic effector cells in vivo. In view of the multitude of pursued strategies, there is hope for improved overall survival of GBM patients in the future.

An Algorithm to Mine Therapeutic Motifs for Cancer from Networks of Genetic Interactions

Study of pairwise genetic interactions, such as mutually exclusive mutations, has led to understanding of underlying mechanisms in cancer. Investigation of various combinatorial motifs within networks of such interactions can lead to deeper insights into its mutational landscape and inform therapy development. One such motif called the Between-Pathway Model (BPM) represents redundant or compensatory pathways that can be therapeutically exploited. Finding such BPM motifs is challenging since most formulations require solving variants of the NP-complete maximum weight bipartite subgraph problem. In this paper we design an algorithm based on Integer Linear Programming (ILP) to solve this problem. In our experiments, our approach outperforms the best previous method to mine BPM motifs. Further, our ILP-based approach allows us to easily model additional application-specific constraints. We illustrate this advantage through a new application of BPM motifs that can potentially aid in finding combination therapies to combat cancer.

Radiomic Features Associated with Extent of Resection in Glioma Surgery

Radiomics defines a set of techniques for extraction and quantification of digital medical data in an automated and reproducible way. Its goal is to detect features potentially related to a clinical task, like classification, diagnosis, prognosis, and response to treatment, going beyond the intrinsic limits of operator-dependency and qualitative description of conventional radiological evaluation on a mesoscopic scale. In the field of neuro-oncology, researchers have tried to create prognostic models for a better tumor diagnosis, histological and biomolecular classification, prediction of response to treatment, and identification of disease relapse. Concerning glioma surgery, the most significant aid that radiomics can give to surgery is to improve tumor extension detection and identify areas that are more prone to recurrence to increase the extent of tumor resection, thereby ameliorating the patients' prognosis. This chapter aims to review the fundamentals of radiomics models' creation, the latest advance of radiomics in neuro-oncology, and possible radiomic features associated with the extent of resection in the brain gliomas.

Deep Proteomic Analysis on Biobanked Paraffine-Archived Melanoma with Prognostic/Predictive Biomarker Read-Out

The discovery of novel protein biomarkers in melanoma is crucial. Our introduction of formalin-fixed paraffin-embedded (FFPE) tumor protocol provides new opportunities to understand the progression of melanoma and open the possibility to screen thousands of FFPE samples deposited in tumor biobanks and available at hospital pathology departments. In our retrospective biobank pilot study, 90 FFPE samples from 77 patients were processed. Protein quantitation was performed by high-resolution mass spectrometry and validated by histopathologic analysis. The global protein expression formed six sample clusters. Proteins such as TRAF6 and ARMC10 were upregulated in clusters with enrichment for shorter survival, and proteins such as AIFI1 were upregulated in clusters with enrichment for longer survival. The cohort's heterogeneity was addressed by comparing primary and metastasis samples, as well comparing clinical stages. Within immunotherapy and targeted therapy subgroups, the upregulation of the VEGFA-VEGFR2 pathway, RNA splicing, increased activity of immune cells, extracellular matrix, and metabolic pathways were positively associated with patient outcome. To summarize, we were able to (i) link global protein expression profiles to survival, and they proved to be an independent prognostic indicator, as well as (ii) identify proteins that are potential predictors of a patient's response to immunotherapy and targeted therapy, suggesting new opportunities for precision medicine developments.

Regulatory Network of PD1 Signaling Is Associated with Prognosis in Glioblastoma Multiforme

Glioblastoma is an aggressive cancer of the brain and spine. While analysis of glioblastoma 'omics data has somewhat improved our understanding of the disease, it has not led to direct improvement in patient survival. Cancer survival is often characterized by differences in gene expression, but the mechanisms that drive these differences are generally unknown. We therefore set out to model the regulatory mechanisms associated with glioblastoma survival. We inferred individual patient gene regulatory networks using data from two different expression platforms from The Cancer Genome Atlas. We performed comparative network analysis between patients with long- and short-term survival. Seven pathways were identified as associated with survival, all of them involved in immune signaling; differential regulation of PD1 signaling was validated to correspond with outcome in an independent dataset from the German Glioma Network. In this pathway, transcriptional repression of genes for which treatment options are available was lost in short-term survivors; this was independent of mutational burden and only weakly associated with T-cell infiltration. Collectively, these results provide a new way to stratify patients with glioblastoma that uses network features as biomarkers to predict survival. They also identify new potential therapeutic interventions, underscoring the value of analyzing gene regulatory networks in individual patients with cancer. SIGNIFICANCE: Genome-wide network modeling of individual glioblastomas identifies dysregulation of PD1 signaling in patients with poor prognosis, indicating this approach can be used to understand how gene regulation influences cancer progression.

A 3-DNA methylation signature as a novel prognostic biomarker in patients with sarcoma by bioinformatics analysis

BACKGROUND

Tumor-specific DNA methylation can potentially be a useful indicator in cancer diagnostics and monitoring. Sarcomas comprise a heterogeneous group of mesenchymal neoplasms which cause life-threatening tumors occurring throughout the body. Therefore, potential molecular detection and prognostic evaluation is very important for early diagnosis and treatment.

METHODS

We performed a retrospective study analyzing DNA methylation of 261 patients with sarcoma from The Cancer Genome Atlas (TCGA) database. Cox regression analyses were conducted to identify a signature associated with the overall survival (OS) of patients with sarcoma, which was validated in a validation dataset.

RESULTS

Three DNA methylation signatures were identified to be significantly associated with OS. Kaplan-Meier analysis showed that the 3-DNA methylation signature could significantly distinguish the high- and low-risk patients in both training (first two-thirds) and validation datasets (remaining one-third). Receiver operating characteristic (ROC) analysis confirmed that the 3-DNA methylation signature exhibited high sensitivity and specificity in predicting OS of patients. Also, the Kaplan-Meier analysis and the area under curve (AUC) values indicated that the 3-DNA methylation signature was independent of clinical characteristics, including age at diagnosis, sex, anatomic location, tumor residual classification, and histological subtypes.

CONCLUSIONS

The current study showed that the 3-DNA methylation model could efficiently function as a novel and independent prognostic biomarker and therapeutic target for patients with sarcoma.

A novel immune-related prognostic signature in epithelial ovarian carcinoma

The immune response is associated with the progression and prognosis of epithelial ovarian cancer (EOC). However, the roles of infiltrated immune cells and immune-related genes (IRGs) in EOC have not been reported comprehensively. In the current study, the differentially expressed genes (DEGs) were filtered based on the integrated gene expression data acquired from The University of California at Santa Cruz (UCSC) Genome Browser. Then, IRGs and transcriptional factors (TFs) were screened based on the ImmPort database and Cistrome database. A total of 501 differentially expressed IRGs, and 76 TFs were detected. A TF-mediated network was constructed by univariate Cox analysis to reveal the potential regulatory mechanisms of IRGs. Next, a nine immune-based prognostic risk model using nine IRGs (PI3, CXCL10, CXCL11, LCN6, CCL17, CCL25, MIF, CX3CR1, and CSPG5) was established. Based on the risk score worked out from the signature, the EOC patients could be classified into low-risk and high-risk groups. Furthermore, the immune landscapes, elevated by the cell-type identification by estimating relative subsets of RNA transcripts (CIBERSORT) algorithm and the Tumor Immune Estimation Resource (TIMER) database, effectuated different patterns in two groups. Thus, an immune-based prognostic risk model of EOC elucidates the immune status in the tumor microenvironment, and hence, could be used for prognosis.

Design and screening of a novel neuropilin-1 targeted penetrating peptide for anti-angiogenic therapy in glioma

AIMS

This study aimed to design and screen a dual functional fusion peptide that could penetrate the blood-brain barrier and target neuropilin 1 (NRP1) overexpressed in vascular endothelial cells for the anti-angiogenesis of glioma treatment.

MAIN METHODS

At the cellular level, the in vitro anti-angiogenic activity of six NRP1 targeting peptides was screened by testing the ability to inhibit the proliferation and tube formation of HUVECs. Then, the in vitro anti-angiogenic activity of two fusion peptides containing different linkers was screened by testing the ability to inhibit HUVECs proliferation, tube formation and migration. The effect of fusion peptide on VEGFR2 related signal pathway was confirmed by Western-blotting. Surface plasmon resonance technology was used to detect the affinity of the fusion peptide to NRP1. The ability of FITC-labeled peptides to penetrate cells was confirmed by cell uptake assay. By establishing an orthotopic glioma model, we evaluated the ability of FITC-labeled peptides to penetrate the blood-brain barrier and their anti-glioma growth activity in vivo.

KEY FINDINGS

We found that NRP1 targeting peptide RP7 and linker cysteine were the most suitable key components in the fusion peptide. We also found that the fusion peptide Tat-C-RP7 we constructed had the strongest ability to penetrate the blood-brain barrier and anti-angiogenic activity in vitro and in vivo.

SIGNIFICANCE

At present, NRP1 targeting peptide as a drug delivery tool and molecular probe seems to have received more attention. We constructed a fusion peptide Tat-C-RP7 with strong anti-angiogenic activity for the treatment of glioma.

Identification and Validation of Stage-Associated PBMC Biomarkers in Breast Cancer Using MS-Based Proteomics

Background: It is well-described that the transcriptome of peripheral blood mononuclear cells (PBMCs) can be altered in the context of many malignancies to allow them avoid the effective immune response, which leads to cancer invasiveness. Here, we used an MS-based strategy to discover biomarkers in the PBMCs of breast cancer (BC) patients and validated them at different stages of BC. Methods: PBMCs were isolated from the breast cancer patients and were cultured alone or co-cultured with breast cancer cell lines. The role of PBMC in the invasion property of breast cancer cells was explored. NF-kB activity was also measured in the co-cultured breast cancer cells. Identification of protein profiles in the secretome and proteome of the co-cultured PBMCs was performed using SWATH mass spectrometry. Pathway enrichment and gene ontology analyses were carried out to look for the molecular pathways correlated with the protein expression profile of PBMCs in the breast cancer patients. Quantitative real-time polymerase chain reaction (qPCR) was performed to validate the candidate genes in the PBMC fraction of the breast cancer patients at the primary and metastatic stages. In silico survival analysis was performed to assess the potential clinical biomarkers in these PBMC subtypes. Results: PBMCs could significantly increase the invasion property of the BC cells concomitant with a decrease in E-cadherin and an increase in both Vimentin and N-cadherin expression. The NF-kB activity in the BC cells significantly increased following co-culturing implying the role of PBMCs in EMT induction. Enrichment analysis showed that the differentially expressed proteins in PBMCs are mainly associated with IL-17, PI3K-Akt, and HIF-1 signaling pathway, in which a set of seven proteins including TMSB4X, HSPA4, S100A9, SRSF6, THBS1, CUL4A, and CANX were frequently expressed. Finally, in silico analysis confirmed that a gene set consisting of S100A9, SRSF6, THBS1, CUL4A, and CANX were found to provide an insight for the identification of metastasis in breast cancer patients. Conclusion: In conclusion, our study revealed that the protein expression profile in PBMCs is a reflection of the proteins expressed in the BC tissue itself; however, the abundance level is different due to the stage of cancer.

The oncogene BCL6 is up-regulated in glioblastoma in response to DNA damage, and drives survival after therapy

The prognosis for people with the high-grade brain tumor glioblastoma is very poor, due largely to low cell death in response to genotoxic therapy. The transcription factor BCL6, a protein that normally suppresses the DNA damage response during immune cell maturation, and a known driver of B-cell lymphoma, was shown to mediate the survival of glioblastoma cells. Expression was observed in glioblastoma tumor specimens and cell lines. When BCL6 expression or activity was reduced in these lines, increased apoptosis and a profound loss of proliferation was observed, consistent with gene expression signatures suggestive of anti-apoptotic and pro-survival signaling role for BCL6 in glioblastoma. Further, treatment with the standard therapies for glioblastoma—ionizing radiation and temozolomide—both induced BCL6 expression in vitro, and an in vivo orthotopic animal model of glioblastoma. Importantly, inhibition of BCL6 in combination with genotoxic therapies enhanced the therapeutic effect. Together these data demonstrate that BCL6 is an active transcription factor in glioblastoma, that it drives survival of cells, and that it increased with DNA damage, which increased the survival rate of therapy-treated cells. This makes BCL6 an excellent therapeutic target in glioblastoma—by increasing sensitivity to standard DNA damaging therapy, BCL6 inhibitors have real potential to improve the outcome for people with this disease.

Novel Immune-Related Gene Signature for Risk Stratification and Prognosis of Survival in Lower-Grade Glioma

Objective

Despite several clinicopathological factors being integrated as prognostic biomarkers, the individual variants and risk stratification have not been fully elucidated in lower grade glioma (LGG). With the prevalence of gene expression profiling in LGG, and based on the critical role of the immune microenvironment, the aim of our study was to develop an immune-related signature for risk stratification and prognosis prediction in LGG.

Methods

RNA-sequencing data from The Cancer Genome Atlas (TCGA), Genome Tissue Expression (GTEx), and Chinese Glioma Genome Atlas (CGGA) were used. Immune-related genes were obtained from the Immunology Database and Analysis Portal (ImmPort). Univariate, multivariate cox regression, and Lasso regression were employed to identify differentially expressed immune-related genes (DEGs) and establish the signature. A nomogram was constructed, and its performance was evaluated by Harrell’s concordance index (C-index), receiver operating characteristic (ROC), and calibration curves. Relationships between the risk score and tumor-infiltrating immune cell abundances were evaluated using CIBERSORTx and TIMER.

Results

Noted, 277 immune-related DEGs were identified. Consecutively, 6 immune genes (CANX, HSPA1B, KLRC2, PSMC6, RFXAP, and TAP1) were identified as risk signature and Kaplan–Meier curve, ROC curve, and risk plot verified its performance in TCGA and CGGA datasets. Univariate and multivariate Cox regression indicated that the risk group was an independent predictor in primary LGG. The prognostic signature showed fair accuracy for 3- and 5-year overall survival in both internal (TCGA) and external (CGGA) validation cohorts. However, predictive performance was poor in the recurrent LGG cohort. The CIBERSORTx algorithm revealed that naïve CD4⁺ T cells were significant higher in low-risk group. Conversely, the infiltration levels of M1-type macrophages, M2-type macrophages, and CD8⁺T cells were significant higher in high-risk group in both TCGA and CGGA cohorts.

Conclusion

The present study constructed a robust six immune-related gene signature and established a prognostic nomogram effective in risk stratification and prediction of overall survival in primary LGG.

Edgetic perturbation signatures represent known and novel cancer biomarkers

Isoform switching is a recently characterized hallmark of cancer, and often translates to the loss or gain of domains mediating protein interactions and thus, the re-wiring of the interactome. Recent computational tools leverage domain-domain interaction data to resolve the condition-specific interaction networks from RNA-Seq data accounting for the domain content of the primary transcripts expressed. Here, we used The Cancer Genome Atlas RNA-Seq datasets to generate 642 patient-specific pairs of interactomes corresponding to both the tumor and the healthy tissues across 13 cancer types. The comparison of these interactomes provided a list of patient-specific edgetic perturbations of the interactomes associated with the cancerous state. We found that among the identified perturbations, select sets are robustly shared between patients at the multi-cancer, cancer-specific and cancer sub-type specific levels. Interestingly, the majority of the alterations do not directly involve significantly mutated genes, nevertheless, they strongly correlate with patient survival. The findings (available at EdgeExplorer: “http://webclu.bio.wzw.tum.de/EdgeExplorer”) are a new source of potential biomarkers for classifying cancer types and the proteins we identified are potential anti-cancer therapy targets.

Proteins inform survival-based differences in patients with glioblastoma

BACKGROUND

Improving the care of patients with glioblastoma (GB) requires accurate and reliable predictors of patient prognosis. Unfortunately, while protein markers are an effective readout of cellular function, proteomics has been underutilized in GB prognostic marker discovery.

METHODS

For this study, GB patients were prospectively recruited and proteomics discovery using liquid chromatography-mass spectrometry analysis (LC-MS/MS) was performed for 27 patients including 13 short-term survivors (STS) (≤10 months) and 14 long-term survivors (LTS) (≥18 months).

RESULTS

Proteomics discovery identified 11 941 peptides in 2495 unique proteins, with 469 proteins exhibiting significant dysregulation when comparing STS to LTS. We verified the differential abundance of 67 out of these 469 proteins in a small previously published independent dataset. Proteins involved in axon guidance were upregulated in STS compared to LTS, while those involved in p53 signaling were upregulated in LTS. We also assessed the correlation between LS MS/MS data with RNAseq data from the same discovery patients and found a low correlation between protein abundance and mRNA expression. Finally, using LC-MS/MS on a set of 18 samples from 6 patients, we quantified the intratumoral heterogeneity of more than 2256 proteins in the multisample dataset.

CONCLUSIONS

These proteomic datasets and noted protein variations present a beneficial resource for better predicting patient outcome and investigating potential therapeutic targets.

The invasive proteome of glioblastoma revealed by laser-capture microdissection

Background

Glioblastomas are heterogeneous tumors composed of a necrotic and tumor core and an invasive periphery.

Methods

Here, we performed a proteomics analysis of laser-capture micro-dissected glioblastoma core and invasive areas of patient-derived xenografts.

Results

Bioinformatics analysis identified enriched proteins in central and invasive tumor areas. Novel markers of invasion were identified, the genes proteolipid protein 1 (PLP1) and Dynamin-1 (DNM1), which were subsequently validated in tumors and by functional assays.

Conclusions

In summary, our results identify new networks and molecules that may play an important role in glioblastoma development and may constitute potential novel therapeutic targets.

NKG2D Controls Natural Reactivity of Vγ9Vδ2 T Lymphocytes against Mesenchymal Glioblastoma Cells

PURPOSE

Cellular immunotherapies are currently being explored to eliminate highly invasive and chemoradioresistant glioblastoma (GBM) cells involved in rapid relapse. We recently showed that concomitant stereotactic injections of nonalloreactive allogeneic Vγ9Vδ2 T lymphocytes eradicate zoledronate-primed human GBM cells. In the present study, we investigated the spontaneous reactivity of allogeneic human Vγ9Vδ2 T lymphocytes toward primary human GBM cells, in vitro and in vivo, in the absence of any prior sensitization.

EXPERIMENTAL DESIGN

Through functional and transcriptomic analyses, we extensively characterized the immunoreactivity of human Vγ9Vδ2 T lymphocytes against various primary GBM cultures directly derived from patient tumors.

RESULTS

We evidenced that GBM cells displaying a mesenchymal signature are spontaneously eliminated by allogeneic human Vγ9Vδ2 T lymphocytes, a reactivity process being mediated by γδ T-cell receptor (TCR) and tightly regulated by cellular stress-associated NKG2D pathway. This led to the identification of highly reactive Vγ9Vδ2 T lymphocyte populations, independently of a specific TCR repertoire signature. Moreover, we finally provide evidence of immunotherapeutic efficacy in vivo, in the absence of any prior tumor cell sensitization.

CONCLUSIONS

By identifying pathways implicated in the selective natural recognition of mesenchymal GBM cell subtypes, accounting for 30% of primary diagnosed and 60% of recurrent GBM, our results pave the way for novel targeted cellular immunotherapies.

A network-based analysis for mining the risk pathways in glioblastoma

The most malignant type of brain tumour is glioblastoma multiforme (GBM). Patients with GBM often have a poor prognosis, as a result of incomplete or inaccurate diagnoses. Regulatory pathways have been demonstrated to serve important roles in complex human diseases. Therefore, deciphering these risk pathways may shed light on the molecular mechanisms underlying GBM progression. In the present study, differentially expressed genes and microRNAs (miRNAs) in a publicly available database were identified between normal and tumour samples. To determine the pathophysiology and molecular mechanisms underlying GBM, integrated network analysis was performed to mine GBM-specific risk pathways. Specifically, a GBM-specific regulatory network was constructed that integrated manually curated GBM-associated transcription and post-transcriptional data resources, including transcription factors and miRNAs. A total of 1,827 differentially expressed genes and 30 miRNAs were identified. The differentially expressed genes were significantly enriched in a number of immune response-associated functions. Based on the GBM-specific regulatory network, 15 risk regulatory pathways containing not only known regulators, but also potential novel targets that might be involved in tumourigenesis were identified. Network analysis provides a strategy for leveraging genomic data to identify potential oncogenic pathways and molecular targets for GBM.

Proteomic Advances in Glial Tumors through Mass Spectrometry Approaches

Being the fourth leading cause of cancer-related death, glial tumors are highly diverse tumor entities characterized by important heterogeneity regarding tumor malignancy and prognosis. However, despite the identification of important alterations in the genome of the glial tumors, there remains a gap in understanding the mechanisms involved in glioma malignancy. Previous research focused on decoding the genomic alterations in these tumors, but due to intricate cellular mechanisms, the genomic findings do not correlate with the functional proteins expressed at the cellular level. The development of mass spectrometry (MS) based proteomics allowed researchers to study proteins expressed at the cellular level or in serum that may provide new insights on the proteins involved in the proliferation, invasiveness, metastasis and resistance to therapy in glial tumors. The integration of data provided by genomic and proteomic approaches into clinical practice could allow for the identification of new predictive, diagnostic and prognostic biomarkers that will improve the clinical management of patients with glial tumors. This paper aims to provide an updated review of the recent proteomic findings, possible clinical applications, and future research perspectives in diffuse astrocytic and oligodendroglial tumors, pilocytic astrocytomas, and ependymomas.

Discovering cooperative biomarkers for heterogeneous complex disease diagnoses

Biomarkers with high reproducibility and accurate prediction performance can contribute to comprehending the underlying pathogenesis of related complex diseases and further facilitate disease diagnosis and therapy. Techniques integrating gene expression profiles and biological networks for the identification of network-based disease biomarkers are receiving increasing interest. The biomarkers for heterogeneous diseases often exhibit strong cooperative effects, which implies that a set of genes may achieve more accurate outcome prediction than any single gene. In this study, we evaluated various biomarker identification methods that consider gene cooperative effects implicitly or explicitly, and proposed the gene cooperation network to explicitly model the cooperative effects of gene combinations. The gene cooperation network-enhanced method, named as MarkRank, achieves superior performance compared with traditional biomarker identification methods in both simulation studies and real data sets. The biomarkers identified by MarkRank not only have a better prediction accuracy but also have stronger topological relationships in the biological network and exhibit high specificity associated with the related diseases. Furthermore, the top genes identified by MarkRank involve crucial biological processes of related diseases and give a good prioritization for known disease genes. In conclusion, MarkRank suggests that explicit modeling of gene cooperative effects can greatly improve biomarker identification for complex diseases, especially for diseases with high heterogeneity.

Exploring Notch Pathway to Elucidate Phenotypic Plasticity and Intra-tumor Heterogeneity in Gliomas

The phenotypic plasticity and self-renewal of adult neural (aNSCs) and glioblastoma stem cells (GSCs) are both known to be governed by active Notch pathway. During development, GSCs can establish differential hierarchy to produce heterogeneous groups of tumor cells belong to different grades, which makes the tumor ecosystem more complex. However, the molecular events regulating these entire processes are unknown hitherto. In this work, based on the mechanistic regulations of Notch pathway activities, a novel computational framework is introduced to inspect the intra-cellular reactions behind the development of normal and tumorigenic cells from aNSCs and GSCs, respectively. The developmental dynamics of aNSCs/GSCs are successfully simulated and molecular activities regulating the phenotypic plasticity and self-renewal processes in normal and tumor cells are identified. A novel scoring parameter “Activity Ratio” score is introduced to find out driver molecules responsible for the phenotypic plasticity and development of different grades of tumor. A new quantitative method is also developed to predict the future risk of Glioblastoma tumor of an individual with appropriate grade by using the transcriptomics profile of that individual as input. Also, a novel technique is introduced to screen and rank the potential drug-targets for suppressing the growth and differentiation of tumor cells.

Identification of key candidate genes involved in melanoma metastasis

Metastasis is the most lethal stage of cancer progression. The present study aimed to investigate the underlying molecular mechanisms of melanoma metastasis using bioinformatics. Using the microarray dataset GSE8401 from the Gene Expression Omnibus database, which included 52 biopsy specimens from patients with melanoma metastasis and 31 biopsy specimens from patients with primary melanoma, differentially expressed genes (DEGs) were identified, subsequent to data preprocessing with the affy package, followed by Gene Ontology and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. A protein-protein interaction (PPI) network was constructed. Mutated genes were analyzed with 80 mutated cases with melanoma from The Cancer Genome Atlas. The overall survival of key candidate DEGs, which were within a filtering of degree >30 criteria in the PPI network and involved three or more KEGG signaling pathways, and genes with a high mutation frequency were delineated. The expression analysis of key candidate DEGs, mutant genes and their associated genes were performed on UALCAN. Of the 1,187 DEGs obtained, 505 were upregulated and 682 were downregulated. ‘Extracellular exosome’ processes, the ‘amoebiasis’ pathway, the ‘ECM-receptor interaction’ pathway and the ‘focal adhesion’ signaling pathway were significantly enriched and identified as important processes or signaling pathways. The overall survival analysis of phosphoinositide-3-kinase regulator subunit 3 (PIK3R3), centromere protein M (CENPM), aurora kinase A (AURKA), laminin subunit α 1 (LAMA1), proliferating cell nuclear antigen (PCNA), adenylate cyclase 1 (ADCY1), BUB1 mitotic checkpoint serine/threonine kinase (BUB1), NDC80 kinetochore complex component (NDC80) and protein kinase C α (PRKCA) in DEGs was statistically significant. Mutation gene analysis identified that BRCA1-associated protein 1 (BAP1) had a higher mutation frequency and survival analysis, and its associated genes in the BAP1-associated PPI network, including ASXL transcriptional regulator 1 (ASXL1), proteasome 26S subunit, non-ATPase 3 (PSMD3), proteasome 26S subunit, non ATPase 11 (PSMD11) and ubiquitin C (UBC), were statistically significantly associated with the overall survival of patients with melanoma. The expression levels of PRKCA, BUB1, BAP1 and ASXL1 were significantly different between primary melanoma and metastatic melanoma. Based on the present study, ‘extracellular exosome’ processes, ‘amoebiasis’ pathways, ‘ECM-receptor interaction’ pathways and ‘focal adhesion’ signaling pathways may be important in the formation of metastases from melanoma. The involved genes, including PIK3R3, CENPM, AURKA, LAMA1, PCNA, ADCY1, BUB1, NDC80 and PRKCA, and mutation associated genes, including BAP1, ASXL1, PSMD3, PSMD11 and UBC, may serve important roles in metastases of melanoma.

NoMAS: A Computational Approach to Find Mutated Subnetworks Associated With Survival in Genome-Wide Cancer Studies

Next-generation sequencing technologies allow to measure somatic mutations in a large number of patients from the same cancer type: one of the main goals in their analysis is the identification of mutations associated with clinical parameters. The identification of such relationships is hindered by extensive genetic heterogeneity in tumors, with different genes mutated in different patients, due, in part, to the fact that genes and mutations act in the context of pathways: it is therefore crucial to study mutations in the context of interactions among genes. In this work we study the problem of identifying subnetworks of a large gene-gene interaction network with mutations associated with survival time. We formally define the associated computational problem by using a score for subnetworks based on the log-rank statistical test to compare the survival of two given populations. We propose a novel approach, based on a new algorithm, called Network of Mutations Associated with Survival (NoMAS) to find subnetworks of a large interaction network whose mutations are associated with survival time. NoMAS is based on the color-coding technique, that has been previously employed in other applications to find the highest scoring subnetwork with high probability when the subnetwork score is additive. In our case the score is not additive, so our algorithm cannot identify the optimal solution with the same guarantees associated to additive scores. Nonetheless, we prove that, under a reasonable model for mutations in cancer, NoMAS identifies the optimal solution with high probability. We also design a holdout approach to identify subnetworks significantly associated with survival time. We test NoMAS on simulated and cancer data, comparing it to approaches based on single gene tests and to various greedy approaches. We show that our method does indeed find the optimal solution and performs better than the other approaches. Moreover, on three cancer datasets our method identifies subnetworks with significant association to survival when none of the genes has significant association with survival when considered in isolation.

EMT network-based feature selection improves prognosis prediction in lung adenocarcinoma

Various feature selection algorithms have been proposed to identify cancer prognostic biomarkers. In recent years, however, their reproducibility is criticized. The performance of feature selection algorithms is shown to be affected by the datasets, underlying networks and evaluation metrics. One of the causes is the curse of dimensionality, which makes it hard to select the features that generalize well on independent data. Even the integration of biological networks does not mitigate this issue because the networks are large and many of their components are not relevant for the phenotype of interest. With the availability of multi-omics data, integrative approaches are being developed to build more robust predictive models. In this scenario, the higher data dimensions create greater challenges. We proposed a phenotype relevant network-based feature selection (PRNFS) framework and demonstrated its advantages in lung cancer prognosis prediction. We constructed cancer prognosis relevant networks based on epithelial mesenchymal transition (EMT) and integrated them with different types of omics data for feature selection. With less than 2.5% of the total dimensionality, we obtained EMT prognostic signatures that achieved remarkable prediction performance (average AUC values >0.8), very significant sample stratifications, and meaningful biological interpretations. In addition to finding EMT signatures from different omics data levels, we combined these single-omics signatures into multi-omics signatures, which improved sample stratifications significantly. Both single- and multi-omics EMT signatures were tested on independent multi-omics lung cancer datasets and significant sample stratifications were obtained.

Current Challenges and Opportunities in Treating Glioblastoma

Glioblastoma multiforme (GBM), the most common and aggressive primary brain tumor, has a high mortality rate despite extensive efforts to develop new treatments. GBM exhibits both intra- and intertumor heterogeneity, lending to resistance and eventual tumor recurrence. Large-scale genomic and proteomic analysis of GBM tumors has uncovered potential drug targets. Effective and “druggable” targets must be validated to embark on a robust medicinal chemistry campaign culminating in the discovery of clinical candidates. Here, we review recent developments in GBM drug discovery and delivery. To identify GBM drug targets, we performed extensive bioinformatics analysis using data from The Cancer Genome Atlas project. We discovered 20 genes, BOC, CLEC4GP1, ELOVL6, EREG, ESR2, FDCSP, FURIN, FUT8-AS1, GZMB, IRX3, LITAF, NDEL1, NKX3-1, PODNL1, PTPRN, QSOX1, SEMA4F, TH, VEGFC, and C20orf166AS1 that are overexpressed in a subpopulation of GBM patients and correlate with poor survival outcomes. Importantly, nine of these genes exhibit higher expression in GBM versus low-grade glioma and may be involved in disease progression. In this review, we discuss these proteins in the context of GBM disease progression. We also conducted computational multi-parameter optimization to assess the blood-brain barrier (BBB) permeability of small molecules in clinical trials for GBM treatment. Drug delivery in the context of GBM is particularly challenging because the BBB hinders small molecule transport. Therefore, we discuss novel drug delivery methods, including nanoparticles and prodrugs. Given the aggressive nature of GBM and the complexity of targeting the central nervous system, effective treatment options are a major unmet medical need. Identification and validation of biomarkers and drug targets associated with GBM disease progression present an exciting opportunity to improve treatment of this devastating disease.

The modular network structure of the mutational landscape of Acute Myeloid Leukemia

Acute myeloid leukemia (AML) is associated with the sequential accumulation of acquired genetic alterations. Although at diagnosis cytogenetic alterations are frequent in AML, roughly 50% of patients present an apparently normal karyotype (NK), leading to a highly heterogeneous prognosis. Due to this significant heterogeneity, it has been suggested that different molecular mechanisms may trigger the disease with diverse prognostic implications. We performed whole-exome sequencing (WES) of tumor-normal matched samples of de novo AML-NK patients lacking mutations in NPM1, CEBPA or FLT3-ITD to identify new gene mutations with potential prognostic and therapeutic relevance to patients with AML. Novel candidate-genes, together with others previously described, were targeted resequenced in an independent cohort of 100 de novo AML patients classified in the cytogenetic intermediate-risk (IR) category. A mean of 4.89 mutations per sample were detected in 73 genes, 35 of which were mutated in more than one patient. After a network enrichment analysis, we defined a single in silico model and established a set of seed-genes that may trigger leukemogenesis in patients with normal karyotype. The high heterogeneity of gene mutations observed in AML patients suggested that a specific alteration could not be as essential as the interaction of deregulated pathways.

Systems Pharmacology: Defining the Interactions of Drug Combinations

The majority of diseases are associated with alterations in multiple molecular pathways and complex interactions at the cellular and organ levels. Single-target monotherapies therefore have intrinsic limitations with respect to their maximum therapeutic benefits. The potential of combination drug therapies has received interest for the treatment of many diseases and is well established in some areas, such as oncology. Combination drug treatments may allow us to identify synergistic drug effects, reduce adverse drug reactions, and address variability in disease characteristics between patients. Identification of combination therapies remains challenging. We discuss current state-of-the-art systems pharmacology approaches to enable rational identification of combination therapies. These approaches, which include characterization of mechanisms of disease and drug action at a systems level, can enable understanding of drug interactions at the molecular, cellular, physiological, and organismal levels. Such multiscale understanding can enable precision medicine by promoting the rational development of combination therapy at the level of individual patients for many diseases.

Bivalent Chromatin Domains in Glioblastoma Reveal a Subtype-Specific Signature of Glioma Stem Cells

Glioblastoma multiforme (GBM) can be clustered by gene expression into four main subtypes associated with prognosis and survival, but enhancers and other gene-regulatory elements have not yet been identified in primary tumors. Here, we profiled six histone modifications and CTCF binding as well as gene expression in primary gliomas and identified chromatin states that define distinct regulatory elements across the tumor genome. Enhancers in mesenchymal and classical tumor subtypes drove gene expression associated with cell migration and invasion, whereas enhancers in proneural tumors controlled genes associated with a less aggressive phenotype in GBM. We identified bivalent domains marked by activating and repressive chromatin modifications. Interestingly, the gene interaction network from common (subtype-independent) bivalent domains was highly enriched for homeobox genes and transcription factors and dominated by SHH and Wnt signaling pathways. This subtype-independent signature of early neural development may be indicative of poised dedifferentiation capacity in glioblastoma and could provide potential targets for therapy.Significance: Enhancers and bivalent domains in glioblastoma are regulated in a subtype-specific manner that resembles gene regulation in glioma stem cells. Cancer Res; 78(10); 2463-74. ©2018 AACR.

Linearity of network proximity measures: implications for set-based queries and significance testing

Motivation

In recent years, various network proximity measures have been proposed to facilitate the use of biomolecular interaction data in a broad range of applications. These applications include functional annotation, disease gene prioritization, comparative analysis of biological systems and prediction of new interactions. In such applications, a major task is the scoring or ranking of the nodes in the network in terms of their proximity to a given set of 'seed' nodes (e.g. a group of proteins that are identified to be associated with a disease, or are deferentially expressed in a certain condition). Many different network proximity measures are utilized for this purpose, and these measures are quite diverse in terms of the benefits they offer.

Results

We propose a unifying framework for characterizing network proximity measures for set-based queries. We observe that many existing measures are linear, in that the proximity of a node to a set of nodes can be represented as an aggregation of its proximity to the individual nodes in the set. Based on this observation, we propose methods for processing of set-based proximity queries that take advantage of sparse local proximity information. In addition, we provide an analytical framework for characterizing the distribution of proximity scores based on reference models that accurately capture the characteristics of the seed set (e.g. degree distribution and biological function). The resulting framework facilitates computation of exact figures for the statistical significance of network proximity scores, enabling assessment of the accuracy of Monte Carlo simulation based estimation methods.

Availability and Implementation

Implementations of the methods in this paper are available at https://bioengine.case.edu/crosstalker which includes a robust visualization for results viewing.

Contact

stm@case.edu or mxk331@case.edu.

Supplementary information

Supplementary data are available at Bioinformatics online.

Application of pharmacometrics and quantitative systems pharmacology to cancer therapy: The example of luminal a breast cancer. Pharmacol Res. 2017;124:20-33. [PMID: 28735000 DOI: 10.1016/j.phrs.2017.07.015]

Breast cancer (BC) is the most common cancer in women, and the second most frequent cause of cancer-related deaths in women worldwide. It is a heterogeneous disease composed of multiple subtypes with distinct morphologies and clinical implications. Quantitative systems pharmacology (QSP) is an emerging discipline bridging systems biology with pharmacokinetics (PK) and pharmacodynamics (PD) leveraging the systematic understanding of drugs' efficacy and toxicity. Despite numerous challenges in applying computational methodologies for QSP and mechanism-based PK/PD models to biological, physiological, and pharmacological data, bridging these disciplines has the potential to enhance our understanding of complex disease systems such as BC. In QSP/PK/PD models, various sources of data are combined including large, multi-scale experimental data such as -omics (i.e. genomics, transcriptomics, proteomics, and metabolomics), biomarkers (circulating and bound), PK, and PD endpoints. This offers a means for a translational application from pre-clinical mathematical models to patients, bridging the bench to bedside paradigm. Not only can these models be applied to inform and advance BC drug development, but they also could aid in optimizing combination therapies and rational dosing regimens for BC patients. Here, we review the current literature pertaining to the application of QSP and pharmacometrics-based pharmacotherapy in BC including bottom-up and top-down modeling approaches. Bottom-up modeling approaches employ mechanistic signal transduction pathways to predict the behavior of a biological system. The ones that are addressed in this review include signal transduction and homeostatic feedback modeling approaches. Alternatively, top-down modeling techniques are bioinformatics reconstruction techniques that infer static connections between molecules that make up a biological network and include (1) Bayesian networks, (2) co-expression networks, and (3) module-based approaches. This review also addresses novel techniques which utilize the principles of systems biology, synthetic lethality and tumor priming, both of which are discussed in relationship to novel drug targets and existing BC therapies. By utilizing QSP approaches, clinicians may develop a platform for improved dose individualization for subpopulation of BC patients, strengthen rationale in treatment designs, and explore mechanism elucidation for improving future treatments in BC medicine.

Angiogenic Gene Signature Derived from Subtype Specific Cell Models Segregate Proneural and Mesenchymal Glioblastoma

Intertumoral molecular heterogeneity in glioblastoma identifies four major subtypes based on expression of molecular markers. Among them, the two clinically interrelated subtypes, proneural and mesenchymal, are the most aggressive with proneural liable for conversion to mesenchymal upon therapy. Using two patient-derived novel primary cell culture models (MTA10 and KW10), we developed a minimal but unique four-gene signature comprising genes vascular endothelial growth factor A (VEGF-A), vascular endothelial growth factor B (VEGF-B) and angiopoietin 1 (ANG1), angiopoietin 2 (ANG2) that effectively segregated the proneural (MTA10) and mesenchymal (KW10) glioblastoma subtypes. The cell culture preclassified as mesenchymal showed elevated expression of genes VEGF-A, VEGF-B and ANG1, ANG2 as compared to the other cell culture model that mimicked the proneural subtype. The differentially expressed genes in these two cell culture models were confirmed by us using TCGA and Verhaak databases and we refer to it as a minimal multigene signature (MMS). We validated this MMS on human glioblastoma tissue sections with the use of immunohistochemistry on preclassified (YKL-40 high or mesenchymal glioblastoma and OLIG2 high or proneural glioblastoma) tumor samples (n = 30). MMS segregated mesenchymal and proneural subtypes with 83% efficiency using a simple histopathology scoring approach (p = 0.008 for ANG2 and p = 0.01 for ANG1). Furthermore, MMS expression negatively correlated with patient survival. Importantly, MMS staining demonstrated spatiotemporal heterogeneity within each subclass, adding further complexity to subtype identification in glioblastoma. In conclusion, we report a novel and simple sequencing-independent histopathology-based biomarker signature comprising genes VEGF-A, VEGF-B and ANG1, ANG2 for subtyping of proneural and mesenchymal glioblastoma.

RNA processing as an alternative route to attack glioblastoma

Genomic analyses have become an important tool to identify new avenues for therapy. This is especially true for cancer types with extremely poor outcomes, since our lack of effective therapies offers no tangible clinical starting point to build upon. The highly malignant brain tumor glioblastoma (GBM) exemplifies such a refractory cancer, with only 15 month average patient survival. Analyses of several hundred GBM samples compiled by the TCGA (The Cancer Genome Atlas) have produced an extensive transcriptomic map, identified prevalent chromosomal alterations, and defined important driver mutations. Unfortunately, clinical trials based on these results have not yet delivered an improvement on outcome. It is, therefore, necessary to characterize other regulatory routes known for playing a role in tumor relapse and response to treatment. Alternative splicing affects more than 90% of the human coding genes and it is an important source for transcript variation and gene regulation. Mutations and alterations in splicing factors are highly prevalent in multiple cancers, demonstrating the potential for splicing to act as a tumor driver. As a result, numerous genes are expressed as cancer-specific splicing isoforms that are functionally distinct from the canonical isoforms found in normal tissue. These include genes that regulate cancer-critical pathways such as apoptosis, DNA repair, cell proliferation, and migration. Splicing defects can even induce genomic instability, a common characteristic of cancer, and a driver of tumor evolution. Importantly, components of the splicing machinery are targetable; multiple drugs can inhibit splicing factors or promote changes in splicing which could be exploited to begin improving clinical outcomes. Here, we review the current literature and present a case for exploring RNA processing as therapeutic route for the treatment of GBM.

Computational analysis of the mesenchymal signature landscape in gliomas

BACKGROUND

Epithelial to mesenchymal transition, and mimicking processes, contribute to cancer invasion and metastasis, and are known to be responsible for resistance to various therapeutic agents in many cancers. While a number of studies have proposed molecular signatures that characterize the spectrum of such transition, more work is needed to understand how the mesenchymal signature (MS) is regulated in non-epithelial cancers like gliomas, to identify markers with the most prognostic significance, and potential for therapeutic targeting.

RESULTS

Computational analysis of 275 glioma samples from "The Cancer Genome Atlas" was used to identify the regulatory changes between low grade gliomas with little expression of MS, and high grade glioblastomas with high expression of MS. TF (transcription factor)-gene regulatory networks were constructed for each of the cohorts, and 5 major pathways and 118 transcription factors were identified as involved in the differential regulation of the networks. The most significant pathway - Extracellular matrix organization - was further analyzed for prognostic relevance. A 20-gene signature was identified as having prognostic significance (HR (hazard ratio) 3.2, 95% CI (confidence interval) = 1.53-8.33), after controlling for known prognostic factors (age, and glioma grade). The signature's significance was validated in an independent data set. The putative stem cell marker CD44 was biologically validated in glioma cell lines and brain tissue samples.

CONCLUSIONS

Our results suggest that the differences between low grade gliomas and high grade glioblastoma are associated with differential expression of the signature genes, raising the possibility that targeting these genes might prolong survival in glioma patients.

Finding lost genes in GWAS via integrative-omics analysis reveals novel sub-networks associated with preterm birth

Maternal genome influences associate with up to 40% of spontaneous preterm births (PTB). Multiple genome wide association studies (GWAS) have been completed to identify genetic variants associated with PTB. Disappointingly, no highly significant SNPs have replicated in independent cohorts so far. We developed an approach combining protein-protein interaction (PPI) network data with tissue specific gene expression data to "find" SNPs of modest significance to identify candidate genes of functional importance that would otherwise be overlooked. This approach is based on the assumption that "high-ranking" SNPs falling short of genome wide significance may nevertheless indicate genes that have substantial biological value in understanding PTB. We mapped highly-ranked candidate SNPs from a meta-analysis of PTB-GWAS to coding genes and developed a PPI network enriched with PTB-SNP carrying genes. This network was scored with gene expression data from term and preterm myometrium to identify subnetworks of PTB-SNP associated genes coordinately expressed with labour onset in myometrial tissue. Our analysis consistently identified significant sub-networks associated with the interacting transcription factors MEF2C and TWIST1, genes not previously associated with PTB, both of which regulate processes clearly relevant to birth timing. Other genes in the significant sub-networks were also associated with inflammatory pathways, as well as muscle function and ion channels. Gene expression level dysregulation was confirmed for eight of these networks by qRT-PCR in an independent set of term and pre-term subjects. Our method identifies novel genes dysregulated in PTB and provides a generalized framework to identify GWAS SNPs that would otherwise be overlooked.

Evaluation of non-supervised MALDI mass spectrometry imaging combined with microproteomics for glioma grade III classification

An integrated diagnosis using molecular features is recommended in the 2016 World Health Organization (WHO) classification. Our aim was to explore non-targeted molecular classification using MALDI mass spectrometry imaging (MALDI MSI) associated to microproteomics in order to classify anaplastic glioma by integration of clinical data. We used fresh-frozen tissue sections to perform MALDI MSI of proteins based on their digestion peptides after in-situ trypsin digestion of the tissue sections and matrix deposition by micro-spraying. The generated 70μm spatial resolution image datasets were further processed by individual or global segmentation in order to cluster the tissues according to their molecular protein signature. The clustering gives 3 main distinct groups. Within the tissues the ROIs (regions of interest) defined by these groups were used for microproteomics by micro-extraction of the tryptic peptides after on-tissue enzymatic digestion. More than 2500 proteins including 22 alternative proteins (AltProt) are identified by the Shotgun microproteomics. Statistical analysis on the basis of the label free quantification of the proteins shows a similar classification to the MALDI MSI segmentation into 3 groups. Functional analysis performed on each group reveals sub-networks related to neoplasia for group 1, glioma with inflammation for group 2 and neurogenesis for group 3. This demonstrates the interest on these new non-targeted large molecular data combining both MALDI MSI and microproteomics data, for tumor classification. This analysis provides new insights into grade III glioma organization. This specific information could allow a more accurate classification of the biopsies according to the prognosis and the identification of potential new targeted therapeutic options. This article is part of a Special Issue entitled: MALDI Imaging, edited by Dr. Corinna Henkel and Prof. Peter Hoffmann.

A network medicine approach to build a comprehensive atlas for the prognosis of human cancer

The Cancer Genome Atlas project has generated multi-dimensional and highly integrated genomic data from a large number of patient samples with detailed clinical records across many cancer types, but it remains unclear how to best integrate the massive amount of genomic data into clinical practice. We report here our methodology to build a multi-dimensional subnetwork atlas for cancer prognosis to better investigate the potential impact of multiple genetic and epigenetic (gene expression, copy number variation, microRNA expression and DNA methylation) changes on the molecular states of networks that in turn affects complex cancer survivorship. We uncover an average of 38 novel subnetworks in the protein-protein interaction network that correlate with prognosis across four prominent cancer types. The clinical utility of these subnetwork biomarkers was further evaluated by prognostic impact evaluation, functional enrichment analysis, drug target annotation, tumor stratification and independent validation. Some pathways including the dynactin, cohesion and pyruvate dehydrogenase-related subnetworks are identified as promising new targets for therapy in specific cancer types. In conclusion, this integrative analysis of existing protein interactome and cancer genomics data allows us to systematically dissect the molecular mechanisms that underlie unexpected outcomes for cancer, which could be used to better understand and predict clinical outcomes, optimize treatment and to provide new opportunities for developing therapeutics related to the subnetworks identified.

Pooling-Based Genome-Wide Association Study Identifies Risk Loci in the Pathogenesis of Ovarian Endometrioma in Chinese Han Women

Endometriosis, regarded as a complex disease, is influenced by multiple genetic factors. Recent genome-wide association studies (GWASs) in endometriosis have identified several susceptibility loci in Caucasian and Japanese populations. However, the overlapped susceptible loci were few. This case-control study tried to identify risk loci-related genes for ovarian endometrioma in Chinese Han women from central China using DNA pooling-based GWAS. Genome DNA samples were extracted from 3038 participants in central China. Pooling-based genome-wide scan and individual genotyping were performed using Affymetrix Genome-Wide Human SNP Array 6.0 and IPLEX Gold system, which demonstrated 10 ovarian endometrioma-related novel risk loci. There were 3 of them with P value < 5 × 10^-06, separately locating in intron of insulin-like growth factor 1 receptor, chromosome 7 open reading frame 50, and Meis homeobox 1. In conclusion, the pooling-based GWAS for ovarian endometrioma identified some novel single-nucleotide polymorphisms in Chinese Han women of central China. Further assessment in other samples will be crucial to confirm the susceptibility of these results and explore the mechanisms of the related genes in the pathogenesis of ovarian endometrioma.

Molecular Predictors of Long-Term Survival in Glioblastoma Multiforme Patients

Glioblastoma multiforme (GBM) is the most common and aggressive adult primary brain cancer, with <10% of patients surviving for more than 3 years. Demographic and clinical factors (e.g. age) and individual molecular biomarkers have been associated with prolonged survival in GBM patients. However, comprehensive systems-level analyses of molecular profiles associated with long-term survival (LTS) in GBM patients are still lacking. We present an integrative study of molecular data and clinical variables in these long-term survivors (LTSs, patients surviving >3 years) to identify biomarkers associated with prolonged survival, and to assess the possible similarity of molecular characteristics between LGG and LTS GBM. We analyzed the relationship between multivariable molecular data and LTS in GBM patients from the Cancer Genome Atlas (TCGA), including germline and somatic point mutation, gene expression, DNA methylation, copy number variation (CNV) and microRNA (miRNA) expression using logistic regression models. The molecular relationship between GBM LTS and LGG tumors was examined through cluster analysis. We identified 13, 94, 43, 29, and 1 significant predictors of LTS using Lasso logistic regression from the somatic point mutation, gene expression, DNA methylation, CNV, and miRNA expression data sets, respectively. Individually, DNA methylation provided the best prediction performance (AUC = 0.84). Combining multiple classes of molecular data into joint regression models did not improve prediction accuracy, but did identify additional genes that were not significantly predictive in individual models. PCA and clustering analyses showed that GBM LTS typically had gene expression profiles similar to non-LTS GBM. Furthermore, cluster analysis did not identify a close affinity between LTS GBM and LGG, nor did we find a significant association between LTS and secondary GBM. The absence of unique LTS profiles and the lack of similarity between LTS GBM and LGG, indicates that there are multiple genetic and epigenetic pathways to LTS in GBM patients.

Understanding Genotype-Phenotype Effects in Cancer via Network Approaches

Cancer is now increasingly studied from the perspective of dysregulated pathways, rather than as a disease resulting from mutations of individual genes. A pathway-centric view acknowledges the heterogeneity between genomic profiles from different cancer patients while assuming that the mutated genes are likely to belong to the same pathway and cause similar disease phenotypes. Indeed, network-centric approaches have proven to be helpful for finding genotypic causes of diseases, classifying disease subtypes, and identifying drug targets. In this review, we discuss how networks can be used to help understand patient-to-patient variations and how one can leverage this variability to elucidate interactions between cancer drivers.

The Mutational Landscape of Acute Promyelocytic Leukemia Reveals an Interacting Network of Co-Occurrences and Recurrent Mutations

Preliminary Acute Promyelocytic Leukemia (APL) whole exome sequencing (WES) studies have identified a huge number of somatic mutations affecting more than a hundred different genes mainly in a non-recurrent manner, suggesting that APL is a heterogeneous disease with secondary relevant changes not yet defined. To extend our knowledge of subtle genetic alterations involved in APL that might cooperate with PML/RARA in the leukemogenic process, we performed a comprehensive analysis of somatic mutations in APL combining WES with sequencing of a custom panel of targeted genes by next-generation sequencing. To select a reduced subset of high confidence candidate driver genes, further in silico analysis were carried out. After prioritization and network analysis we found recurrent deleterious mutations in 8 individual genes (STAG2, U2AF1, SMC1A, USP9X, IKZF1, LYN, MYCBP2 and PTPN11) with a strong potential of being involved in APL pathogenesis. Our network analysis of multiple mutations provides a reliable approach to prioritize genes for additional analysis, improving our knowledge of the leukemogenesis interactome. Additionally, we have defined a functional module in the interactome of APL. The hypothesis is that the number, or the specific combinations, of mutations harbored in each patient might not be as important as the disturbance caused in biological key functions, triggered by several not necessarily recurrent mutations.

Network-Based Integration of Disparate Omic Data To Identify "Silent Players" in Cancer

Development of high-throughput monitoring technologies enables interrogation of cancer samples at various levels of cellular activity. Capitalizing on these developments, various public efforts such as The Cancer Genome Atlas (TCGA) generate disparate omic data for large patient cohorts. As demonstrated by recent studies, these heterogeneous data sources provide the opportunity to gain insights into the molecular changes that drive cancer pathogenesis and progression. However, these insights are limited by the vast search space and as a result low statistical power to make new discoveries. In this paper, we propose methods for integrating disparate omic data using molecular interaction networks, with a view to gaining mechanistic insights into the relationship between molecular changes at different levels of cellular activity. Namely, we hypothesize that genes that play a role in cancer development and progression may be implicated by neither frequent mutation nor differential expression, and that network-based integration of mutation and differential expression data can reveal these “silent players”. For this purpose, we utilize network-propagation algorithms to simulate the information flow in the cell at a sample-specific resolution. We then use the propagated mutation and expression signals to identify genes that are not necessarily mutated or differentially expressed genes, but have an essential role in tumor development and patient outcome. We test the proposed method on breast cancer and glioblastoma multiforme data obtained from TCGA. Our results show that the proposed method can identify important proteins that are not readily revealed by molecular data, providing insights beyond what can be gleaned by analyzing different types of molecular data in isolation.

Identification of cancer-related genes is an important task, made more difficult by heterogeneity between samples and even within individual patients. Methods for identifying disease-related genes typically focus on individual data sets such as mutational and differential expression data, and therefore are limited to genes that are implicated by each data set in isolation. In this work we propose a method that uses protein interaction network information to integrate mutational and differential expression data on a sample-specific level, and combine this information across samples in ways that respect the commonalities and differences between distinct mutation and differential expression profiles. We use this information to identify genes that are associated with cancer but not readily identifiable by mutations or differential expression alone. Our method highlights the features that significantly predict a gene’s association with cancer, shows improved predictive power in recovering cancer-related genes in known pathways, and identifies genes that are neither frequently mutated nor differentially expressed but show significant association with survival.

Combined Mapping of Multiple clUsteriNg ALgorithms (COMMUNAL): A Robust Method for Selection of Cluster Number, K

In order to discover new subsets (clusters) of a data set, researchers often use algorithms that perform unsupervised clustering, namely, the algorithmic separation of a dataset into some number of distinct clusters. Deciding whether a particular separation (or number of clusters, K) is correct is a sort of ‘dark art’, with multiple techniques available for assessing the validity of unsupervised clustering algorithms. Here, we present a new technique for unsupervised clustering that uses multiple clustering algorithms, multiple validity metrics, and progressively bigger subsets of the data to produce an intuitive 3D map of cluster stability that can help determine the optimal number of clusters in a data set, a technique we call COmbined Mapping of Multiple clUsteriNg ALgorithms (COMMUNAL). COMMUNAL locally optimizes algorithms and validity measures for the data being used. We show its application to simulated data with a known K, and then apply this technique to several well-known cancer gene expression datasets, showing that COMMUNAL provides new insights into clustering behavior and stability in all tested cases. COMMUNAL is shown to be a useful tool for determining K in complex biological datasets, and is freely available as a package for R.

Network-based sub-network signatures unveil the potential for acute myeloid leukemia therapy

Although gene expression profiling studies of acute myeloid leukemia (AML) patients have provided key insights into potential diagnostic and prognostic markers and therapeutic targets, it is not clear that the patterns of molecular heterogeneity affect the tumor biology and respond to the treatment. We hypothesized that network-based gene expression signatures of AML represent the mechanistically important genes and may improve the predicted performance of prognosis and clinical outcome. We provided the random walk with restart (RWR) analysis to discover the sub-network of genomic alterations. The RWR approach integrates the signature genes derived from the random forest (RF) analysis as "seeds" to identify genes critical to the AML recurrence phenotype. To test whether the 81-gene biomarkers could predict AML recurrence, we developed Survival Support Vector Machine (SSVM) models using a gene expression dataset and test on an independent dataset. The random forest classifier was built based on 81-gene biomarkers to separate the AML patients into "recurrence" and "non-recurrence" groups. The 81-gene biomarkers showed significant enrichment related to cancer pathophysiology and provided good coverage of sub-network biomarkers and AML-related signaling pathways. The SSVM-based score was significantly associated with overall survival (hazard ratio [HR], 2.16; 95% confidence interval [CI], 1.18-3.97; p = 0.01). Similar results were obtained with reversed training and testing datasets (hazard ratio [HR], 1.6; 95% confidence interval [CI], 1.08-2.37; p = 0.02). The 81-gene biomarker based RF classifier improved classification performance. Overall, 81-gene biomarkers might be useful prognostic and predictive molecular markers to predict the clinical outcome of AML patients.

Ion channel gene expression predicts survival in glioma patients

Ion channels are important regulators in cell proliferation, migration, and apoptosis. The malfunction and/or aberrant expression of ion channels may disrupt these important biological processes and influence cancer progression. In this study, we investigate the expression pattern of ion channel genes in glioma. We designate 18 ion channel genes that are differentially expressed in high-grade glioma as a prognostic molecular signature. This ion channel gene expression based signature predicts glioma outcome in three independent validation cohorts. Interestingly, 16 of these 18 genes were down-regulated in high-grade glioma. This signature is independent of traditional clinical, molecular, and histological factors. Resampling tests indicate that the prognostic power of the signature outperforms random gene sets selected from human genome in all the validation cohorts. More importantly, this signature performs better than the random gene signatures selected from glioma-associated genes in two out of three validation datasets. This study implicates ion channels in brain cancer, thus expanding on knowledge of their roles in other cancers. Individualized profiling of ion channel gene expression serves as a superior and independent prognostic tool for glioma patients.

Discovering key regulatory mechanisms from single-factor and multi-factor regulations in glioblastoma utilizing multi-dimensional data

Glioblastoma (GBM) is the most common malignant brain cancer in adults. Investigating the regulatory mechanisms underlying GBM is effective for the in-depth study of GBM. The Cancer Genome Atlas (TCGA) project is producing large-scale data and makes the comprehensive study of the diverse regulatory mechanisms underlying GBM possible. Although there have been research studies on GBM with large-scale data, distinguishing different regulatory mechanisms and identifying the key regulation types remain challenging. In this study, we integrated multi-dimensional data of differentially expressed genes in GBM: copy number variation (CNV), gene expression, miRNA expression and methylation, by performing partial correlation analysis with the Lasso technique. Our results showed that there were single-factor and multi-factor regulatory mechanisms in GBM. In further analysis of the regulation subtypes, we discovered that single-factor and multi-factor regulations are potentially distinct in functionality. Moreover, multi-factor regulations especially the key regulation subtypes may be more relevant to GBM and affect many GBM-related genes such as ERBB2 and MAPK1. This study not only verifies the utility of multi-dimensional data integration into GBM research but also distinguishes the key multi-factor regulatory subtypes that may drive pathogenesis of GBM from various regulatory mechanisms.

Survival impact of time to initiation of chemoradiotherapy after resection of newly diagnosed glioblastoma

OBJECT

There are few and conflicting reports on the effects of delayed initiation of chemoradiotherapy on the survival of patients with glioblastoma. The standard of care for newly diagnosed glioblastoma is concurrent radiotherapy and temozolomide chemotherapy after maximal safe resection; however, the optimal timing of such therapy is poorly defined. Given the lack of consensus in the literature, the authors performed a retrospective analysis of The Cancer Genome Atlas (TCGA) database to investigate the effect of time from surgery to initiation of therapy on survival in newly diagnosed glioblastoma.

METHODS

Patients with primary glioblastoma diagnosed since 2005 and treated according to the standard of care were identified from TCGA database. Kaplan-Meier and multivariate Cox regression analyses were used to compare overall survival (OS) and progression-free survival (PFS) between groups stratified by postoperative delay to initiation of radiation treatment.

RESULTS

There were 218 patients with newly diagnosed glioblastoma with known time to initiation of radiotherapy identified in the database. The median duration until therapy was 27 days. Delay to radiotherapy longer than the median was not associated with worse PFS (HR = 0.918, p = 0.680) or OS (HR = 1.135, p = 0.595) in multivariate analysis when controlling for age, sex, KPS score, and adjuvant chemotherapy. Patients in the highest and lowest quartiles for delay to therapy (≤ 20 days vs ≥ 36 days) did not statistically differ in PFS (p = 0.667) or OS (p = 0.124). The small subset of patients with particularly long delays (> 42 days) demonstrated worse OS (HR = 1.835, p = 0.019), but not PFS (p = 0.74).

CONCLUSIONS

Modest delay in initiation of postoperative chemotherapy and radiation does not appear to be associated with worse PFS or OS in patients with newly diagnosed glioblastoma, while significant delay longer than 6 weeks may be associated with worse OS.

Prognostic gene signature identification using causal structure learning: applications in kidney cancer

Identification of molecular-based signatures is one of the critical steps toward finding therapeutic targets in cancer. In this paper, we propose methods to discover prognostic gene signatures under a causal structure learning framework across the whole genome. The causal structures are represented by directed acyclic graphs (DAGs), wherein we construct gene-specific network modules that constitute a gene and its corresponding regulators. The modules are then subsequently used to correlate with survival times, thus, allowing for a network-oriented approach to gene selection to adjust for potential confounders, as opposed to univariate (gene-by-gene) approaches. Our methods are motivated by and applied to a clear cell renal cell carcinoma (ccRCC) study from The Cancer Genome Atlas (TCGA) where we find several prognostic genes associated with cancer progression - some of which are novel while others confirm existing findings.

Predicting cancer prognosis using functional genomics data sets

Elucidating the molecular basis of human cancers is an extremely complex and challenging task. A wide variety of computational tools and experimental techniques have been used to address different aspects of this characterization. One major hurdle faced by both clinicians and researchers has been to pinpoint the mechanistic basis underlying a wide range of prognostic outcomes for the same type of cancer. Here, we provide an overview of various computational methods that have leveraged different functional genomics data sets to identify molecular signatures that can be used to predict prognostic outcome for various human cancers. Furthermore, we outline challenges that remain and future directions that may be explored to address them.

Cluster and Principal Component Analysis of Human Glioblastoma Multiforme (GBM) Tumor Proteome

BACKGROUND

Glioblastoma Multiforme (GBM) or grade IV astrocytoma is the most common and lethal adult malignant brain tumor. Several of the molecular alterations detected in gliomas may have diagnostic and/or prognostic implications. Proteomics has been widely applied in various areas of science, ranging from the deciphering of molecular pathogen nests of discuses.

METHODS

In this study proteins were extracted from the tumor and normal brain tissues and then the protein purity was evaluated by Bradford test and spectrophotometry. In this study, proteins were separated by 2-Dimensional Gel (2DG) electrophoresis method and the spots were then analyzed and compared using statistical data and specific software. Protein clustering analysis was performed on the list of proteins deemed significantly altered in glioblastoma tumors (t-test and one-way ANOVA; P< 0.05).

RESULTS

The 2D gel showed totally 876 spots. We reported, 172 spots were exhibited differently in expression level (fold > 2) for glioblastoma. On each analytical 2D gel, an average of 876 spots was observed. In this study, 188 spots exhibited up regulation of expression level, whereas the remaining 232 spots were decreased in glioblastoma tumor relative to normal tissue. Results demonstrate that functional clustering (up and down regulated) and Principal Component Analysis (PCA) has considerable merits in aiding the interpretation of proteomic data.

CONCLUSION

2D gel electrophoresis is the core of proteomics which permitted the separation of thousands of proteins. High resolution 2DE can resolve up to 5,000 proteins simultaneously. Using cluster analysis, we can also form groups of related variables, similar to what is practiced in factor analysis.