Unveiling the Molecular Mechanisms of Glioblastoma through an Integrated Network-Based Approach

Ali Kaynar; Woonghee Kim; Atakan Burak Ceyhan; Cheng Zhang; Mathias Uhlén; Hasan Turkez; Saeed Shoaie; Adil Mardinoglu

doi:10.3390/biomedicines12102237

Unveiling the Molecular Mechanisms of Glioblastoma through an Integrated Network-Based Approach

Ali Kaynar, Woonghee Kim, Atakan Burak Ceyhan, Cheng Zhang, Mathias Uhlén, Hasan Turkez, Saeed Shoaie, Adil Mardinoglu^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

1 Citation (Scopus)

7 Downloads (Pure)

Abstract

Background/Objectives: Despite current treatments extending the lifespan of Glioblastoma (GBM) patients, the average survival time is around 15–18 months, underscoring the fatality of GBM. This study aims to investigate the impact of sample heterogeneity on gene expression in GBM, identify key metabolic pathways and gene modules, and explore potential therapeutic targets. Methods: In this study, we analysed GBM transcriptome data derived from The Cancer Genome Atlas (TCGA) using genome-scale metabolic models (GEMs) and co-expression networks. We examine transcriptome data incorporating tumour purity scores (TPSs), allowing us to assess the impact of sample heterogeneity on gene expression profiles. We analysed the metabolic profile of GBM by generating condition-specific GEMs based on the TPS group. Results: Our findings revealed that over 90% of genes showing brain and glioma specificity in RNA expression demonstrate a high positive correlation, underscoring their expression is dominated by glioma cells. Conversely, negatively correlated genes are strongly associated with immune responses, indicating a complex interaction between glioma and immune pathways and non-tumorigenic cell dominance on gene expression. TPS-based metabolic profile analysis was supported by reporter metabolite analysis, highlighting several metabolic pathways, including arachidonic acid, kynurenine and NAD pathway. Through co-expression network analysis, we identified modules that significantly overlap with TPS-correlated genes. Notably, SOX11 and GSX1 are upregulated in High TPS, show a high correlation with TPS, and emerged as promising therapeutic targets. Additionally, NCAM1 exhibits a high centrality score within the co-expression module, which shows a positive correlation with TPS. Moreover, LILRB4, an immune-related gene expressed in the brain, showed a negative correlation and upregulated in Low TPS, highlighting the importance of modulating immune responses in the GBM mechanism. Conclusions: Our study uncovers sample heterogeneity’s impact on gene expression and the molecular mechanisms driving GBM, and it identifies potential therapeutic targets for developing effective treatments for GBM patients.

Original language	English
Article number	2237
Journal	Biomedicines
Volume	12
Issue number	10
Early online date	1 Oct 2024
DOIs	https://doi.org/10.3390/biomedicines12102237
Publication status	Published - 1 Oct 2024

Keywords

biomarker
co-expression networks
drug target
GEMs
glioblastoma
immune response
microenvironment

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.3390/biomedicines12102237Licence: CC BY

Unveiling the Molecular Mechanisms_KAYNAR_Publishedonline1October2024_GOLD_VoR (CC BY)Final published version, 5.55 MBLicence: CC BY

Cite this

@article{6ff8882adee64b4e918720cb73443425,

title = "Unveiling the Molecular Mechanisms of Glioblastoma through an Integrated Network-Based Approach",

abstract = "Background/Objectives: Despite current treatments extending the lifespan of Glioblastoma (GBM) patients, the average survival time is around 15–18 months, underscoring the fatality of GBM. This study aims to investigate the impact of sample heterogeneity on gene expression in GBM, identify key metabolic pathways and gene modules, and explore potential therapeutic targets. Methods: In this study, we analysed GBM transcriptome data derived from The Cancer Genome Atlas (TCGA) using genome-scale metabolic models (GEMs) and co-expression networks. We examine transcriptome data incorporating tumour purity scores (TPSs), allowing us to assess the impact of sample heterogeneity on gene expression profiles. We analysed the metabolic profile of GBM by generating condition-specific GEMs based on the TPS group. Results: Our findings revealed that over 90% of genes showing brain and glioma specificity in RNA expression demonstrate a high positive correlation, underscoring their expression is dominated by glioma cells. Conversely, negatively correlated genes are strongly associated with immune responses, indicating a complex interaction between glioma and immune pathways and non-tumorigenic cell dominance on gene expression. TPS-based metabolic profile analysis was supported by reporter metabolite analysis, highlighting several metabolic pathways, including arachidonic acid, kynurenine and NAD pathway. Through co-expression network analysis, we identified modules that significantly overlap with TPS-correlated genes. Notably, SOX11 and GSX1 are upregulated in High TPS, show a high correlation with TPS, and emerged as promising therapeutic targets. Additionally, NCAM1 exhibits a high centrality score within the co-expression module, which shows a positive correlation with TPS. Moreover, LILRB4, an immune-related gene expressed in the brain, showed a negative correlation and upregulated in Low TPS, highlighting the importance of modulating immune responses in the GBM mechanism. Conclusions: Our study uncovers sample heterogeneity{\textquoteright}s impact on gene expression and the molecular mechanisms driving GBM, and it identifies potential therapeutic targets for developing effective treatments for GBM patients.",

keywords = "biomarker, co-expression networks, drug target, GEMs, glioblastoma, immune response, microenvironment",

author = "Ali Kaynar and Woonghee Kim and Ceyhan, {Atakan Burak} and Cheng Zhang and Mathias Uhl{\'e}n and Hasan Turkez and Saeed Shoaie and Adil Mardinoglu",

note = "Publisher Copyright: {\textcopyright} 2024 by the authors.",

year = "2024",

month = oct,

day = "1",

doi = "10.3390/biomedicines12102237",

language = "English",

volume = "12",

journal = "Biomedicines",

issn = "2227-9059",

publisher = "MDPI AG",

number = "10",

}

TY - JOUR

T1 - Unveiling the Molecular Mechanisms of Glioblastoma through an Integrated Network-Based Approach

AU - Kaynar, Ali

AU - Kim, Woonghee

AU - Ceyhan, Atakan Burak

AU - Zhang, Cheng

AU - Uhlén, Mathias

AU - Turkez, Hasan

AU - Shoaie, Saeed

AU - Mardinoglu, Adil

PY - 2024/10/1

Y1 - 2024/10/1

N2 - Background/Objectives: Despite current treatments extending the lifespan of Glioblastoma (GBM) patients, the average survival time is around 15–18 months, underscoring the fatality of GBM. This study aims to investigate the impact of sample heterogeneity on gene expression in GBM, identify key metabolic pathways and gene modules, and explore potential therapeutic targets. Methods: In this study, we analysed GBM transcriptome data derived from The Cancer Genome Atlas (TCGA) using genome-scale metabolic models (GEMs) and co-expression networks. We examine transcriptome data incorporating tumour purity scores (TPSs), allowing us to assess the impact of sample heterogeneity on gene expression profiles. We analysed the metabolic profile of GBM by generating condition-specific GEMs based on the TPS group. Results: Our findings revealed that over 90% of genes showing brain and glioma specificity in RNA expression demonstrate a high positive correlation, underscoring their expression is dominated by glioma cells. Conversely, negatively correlated genes are strongly associated with immune responses, indicating a complex interaction between glioma and immune pathways and non-tumorigenic cell dominance on gene expression. TPS-based metabolic profile analysis was supported by reporter metabolite analysis, highlighting several metabolic pathways, including arachidonic acid, kynurenine and NAD pathway. Through co-expression network analysis, we identified modules that significantly overlap with TPS-correlated genes. Notably, SOX11 and GSX1 are upregulated in High TPS, show a high correlation with TPS, and emerged as promising therapeutic targets. Additionally, NCAM1 exhibits a high centrality score within the co-expression module, which shows a positive correlation with TPS. Moreover, LILRB4, an immune-related gene expressed in the brain, showed a negative correlation and upregulated in Low TPS, highlighting the importance of modulating immune responses in the GBM mechanism. Conclusions: Our study uncovers sample heterogeneity’s impact on gene expression and the molecular mechanisms driving GBM, and it identifies potential therapeutic targets for developing effective treatments for GBM patients.

AB - Background/Objectives: Despite current treatments extending the lifespan of Glioblastoma (GBM) patients, the average survival time is around 15–18 months, underscoring the fatality of GBM. This study aims to investigate the impact of sample heterogeneity on gene expression in GBM, identify key metabolic pathways and gene modules, and explore potential therapeutic targets. Methods: In this study, we analysed GBM transcriptome data derived from The Cancer Genome Atlas (TCGA) using genome-scale metabolic models (GEMs) and co-expression networks. We examine transcriptome data incorporating tumour purity scores (TPSs), allowing us to assess the impact of sample heterogeneity on gene expression profiles. We analysed the metabolic profile of GBM by generating condition-specific GEMs based on the TPS group. Results: Our findings revealed that over 90% of genes showing brain and glioma specificity in RNA expression demonstrate a high positive correlation, underscoring their expression is dominated by glioma cells. Conversely, negatively correlated genes are strongly associated with immune responses, indicating a complex interaction between glioma and immune pathways and non-tumorigenic cell dominance on gene expression. TPS-based metabolic profile analysis was supported by reporter metabolite analysis, highlighting several metabolic pathways, including arachidonic acid, kynurenine and NAD pathway. Through co-expression network analysis, we identified modules that significantly overlap with TPS-correlated genes. Notably, SOX11 and GSX1 are upregulated in High TPS, show a high correlation with TPS, and emerged as promising therapeutic targets. Additionally, NCAM1 exhibits a high centrality score within the co-expression module, which shows a positive correlation with TPS. Moreover, LILRB4, an immune-related gene expressed in the brain, showed a negative correlation and upregulated in Low TPS, highlighting the importance of modulating immune responses in the GBM mechanism. Conclusions: Our study uncovers sample heterogeneity’s impact on gene expression and the molecular mechanisms driving GBM, and it identifies potential therapeutic targets for developing effective treatments for GBM patients.

KW - biomarker

KW - co-expression networks

KW - drug target

KW - GEMs

KW - glioblastoma

KW - immune response

KW - microenvironment

UR - http://www.scopus.com/inward/record.url?scp=85207678383&partnerID=8YFLogxK

U2 - 10.3390/biomedicines12102237

DO - 10.3390/biomedicines12102237

M3 - Article

AN - SCOPUS:85207678383

SN - 2227-9059

VL - 12

JO - Biomedicines

JF - Biomedicines

IS - 10

M1 - 2237

ER -

Unveiling the Molecular Mechanisms of Glioblastoma through an Integrated Network-Based Approach

Abstract

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this