Limits...
Reconstructed Metabolic Network Models Predict Flux-Level Metabolic Reprogramming in Glioblastoma.

Özcan E, Çakır T - Front Neurosci (2016)

Bottom Line: Also, our results, in accordance with recent studies, predict a contribution of oxidative phosphorylation to ATP pool via a slightly active TCA cycle in addition to the major contributor aerobic glycolysis.We verified our results by using different computational methods that incorporate transcriptome data with genome-scale models and by using different transcriptome datasets.Correct predictions of flux distributions in glycolysis, glutaminolysis, TCA cycle and lipid precursor metabolism validate the reconstructed models for further use in future to simulate more specific metabolic patterns for GBM.

View Article: PubMed Central - PubMed

Affiliation: Computational Systems Biology Group, Department of Bioengineering, Gebze Technical University Gebze, Turkey.

ABSTRACT
Developments in genome scale metabolic modeling techniques and omics technologies have enabled the reconstruction of context-specific metabolic models. In this study, glioblastoma multiforme (GBM), one of the most common and aggressive malignant brain tumors, is investigated by mapping GBM gene expression data on the growth-implemented brain specific genome-scale metabolic network, and GBM-specific models are generated. The models are used to calculate metabolic flux distributions in the tumor cells. Metabolic phenotypes predicted by the GBM-specific metabolic models reconstructed in this work reflect the general metabolic reprogramming of GBM, reported both in in-vitro and in-vivo experiments. The computed flux profiles quantitatively predict that major sources of the acetyl-CoA and oxaloacetic acid pool used in TCA cycle are pyruvate dehydrogenase from glycolysis and anaplerotic flux from glutaminolysis, respectively. Also, our results, in accordance with recent studies, predict a contribution of oxidative phosphorylation to ATP pool via a slightly active TCA cycle in addition to the major contributor aerobic glycolysis. We verified our results by using different computational methods that incorporate transcriptome data with genome-scale models and by using different transcriptome datasets. Correct predictions of flux distributions in glycolysis, glutaminolysis, TCA cycle and lipid precursor metabolism validate the reconstructed models for further use in future to simulate more specific metabolic patterns for GBM.

No MeSH data available.


Related in: MedlinePlus

Flux values of the in-silico GBM models by GIMME. Values indicate fluxes for “the mean of the three GBM subtypes” (top, based on GSE13041-GPL96), “the metabolic model obtained using different microarray platform but from the same dataset as GBM subtypes” (middle, based on GSE13041-GPL570) and “the metabolic model obtained using the same platform as GBM subtypes but from a different dataset” (down, based on GSE13276-GPL96). Results show that constraining the model with different GBM transcriptome datasets leads to very similar flux profiles. The figure was drawn in PathVisio 3 toolbox (Kutmon et al., 2015).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4834348&req=5

Figure 3: Flux values of the in-silico GBM models by GIMME. Values indicate fluxes for “the mean of the three GBM subtypes” (top, based on GSE13041-GPL96), “the metabolic model obtained using different microarray platform but from the same dataset as GBM subtypes” (middle, based on GSE13041-GPL570) and “the metabolic model obtained using the same platform as GBM subtypes but from a different dataset” (down, based on GSE13276-GPL96). Results show that constraining the model with different GBM transcriptome datasets leads to very similar flux profiles. The figure was drawn in PathVisio 3 toolbox (Kutmon et al., 2015).

Mentions: In order to demonstrate the robustness of the results, transcriptome data from a different microarray platform (GPL570) but from the same dataset (GSE13041) and from the same platform (GPL96) but from a different dataset (GSE13276) were additionally used to derive GBM-specific metabolic models and calculate corresponding flux distributions (see Materials and Methods for details). This is an important issue to be considered to validate our results since platform or laboratory differences may cause serious reproducibility problems in microarray experiments (Draghici et al., 2006). No sub-type differences were accounted in these calculations. The calculated fluxes are depicted in Figure 3.


Reconstructed Metabolic Network Models Predict Flux-Level Metabolic Reprogramming in Glioblastoma.

Özcan E, Çakır T - Front Neurosci (2016)

Flux values of the in-silico GBM models by GIMME. Values indicate fluxes for “the mean of the three GBM subtypes” (top, based on GSE13041-GPL96), “the metabolic model obtained using different microarray platform but from the same dataset as GBM subtypes” (middle, based on GSE13041-GPL570) and “the metabolic model obtained using the same platform as GBM subtypes but from a different dataset” (down, based on GSE13276-GPL96). Results show that constraining the model with different GBM transcriptome datasets leads to very similar flux profiles. The figure was drawn in PathVisio 3 toolbox (Kutmon et al., 2015).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4834348&req=5

Figure 3: Flux values of the in-silico GBM models by GIMME. Values indicate fluxes for “the mean of the three GBM subtypes” (top, based on GSE13041-GPL96), “the metabolic model obtained using different microarray platform but from the same dataset as GBM subtypes” (middle, based on GSE13041-GPL570) and “the metabolic model obtained using the same platform as GBM subtypes but from a different dataset” (down, based on GSE13276-GPL96). Results show that constraining the model with different GBM transcriptome datasets leads to very similar flux profiles. The figure was drawn in PathVisio 3 toolbox (Kutmon et al., 2015).
Mentions: In order to demonstrate the robustness of the results, transcriptome data from a different microarray platform (GPL570) but from the same dataset (GSE13041) and from the same platform (GPL96) but from a different dataset (GSE13276) were additionally used to derive GBM-specific metabolic models and calculate corresponding flux distributions (see Materials and Methods for details). This is an important issue to be considered to validate our results since platform or laboratory differences may cause serious reproducibility problems in microarray experiments (Draghici et al., 2006). No sub-type differences were accounted in these calculations. The calculated fluxes are depicted in Figure 3.

Bottom Line: Also, our results, in accordance with recent studies, predict a contribution of oxidative phosphorylation to ATP pool via a slightly active TCA cycle in addition to the major contributor aerobic glycolysis.We verified our results by using different computational methods that incorporate transcriptome data with genome-scale models and by using different transcriptome datasets.Correct predictions of flux distributions in glycolysis, glutaminolysis, TCA cycle and lipid precursor metabolism validate the reconstructed models for further use in future to simulate more specific metabolic patterns for GBM.

View Article: PubMed Central - PubMed

Affiliation: Computational Systems Biology Group, Department of Bioengineering, Gebze Technical University Gebze, Turkey.

ABSTRACT
Developments in genome scale metabolic modeling techniques and omics technologies have enabled the reconstruction of context-specific metabolic models. In this study, glioblastoma multiforme (GBM), one of the most common and aggressive malignant brain tumors, is investigated by mapping GBM gene expression data on the growth-implemented brain specific genome-scale metabolic network, and GBM-specific models are generated. The models are used to calculate metabolic flux distributions in the tumor cells. Metabolic phenotypes predicted by the GBM-specific metabolic models reconstructed in this work reflect the general metabolic reprogramming of GBM, reported both in in-vitro and in-vivo experiments. The computed flux profiles quantitatively predict that major sources of the acetyl-CoA and oxaloacetic acid pool used in TCA cycle are pyruvate dehydrogenase from glycolysis and anaplerotic flux from glutaminolysis, respectively. Also, our results, in accordance with recent studies, predict a contribution of oxidative phosphorylation to ATP pool via a slightly active TCA cycle in addition to the major contributor aerobic glycolysis. We verified our results by using different computational methods that incorporate transcriptome data with genome-scale models and by using different transcriptome datasets. Correct predictions of flux distributions in glycolysis, glutaminolysis, TCA cycle and lipid precursor metabolism validate the reconstructed models for further use in future to simulate more specific metabolic patterns for GBM.

No MeSH data available.


Related in: MedlinePlus