Table 3.
Application of GEMs in metabolic cancer research.
| Reference | Category | Concept | Tools Used | Databases | Type of Validation |
|---|---|---|---|---|---|
| [170] | CSGEMs to generate metabolic signatures for drug repositioning | PC GEM to explore PC metabolism and repurpose new drugs. Reconstruction performed combining personalized GEMs from individual patient’s transcriptome and PC-specific proteomics data from the HCA. | RAVEN, FastGeneSL [171], DIRAC, TCGAbiolinks, DESeq, gcrma | HPA, HMA, ConnectivityMap2 | In silico cell viability assay and in vitro cell assay. |
| [172] | CSGEMs to predict biomarkers and drug targets | GEM of transcriptional regulator-metabolite associations with mixed computational and wet lab experiments integrating intracellular metabolic profiles of NCI-60 (4) 54 CCLs with transcriptomic and proteomic data. Perform metabolic profiling of CCLs and resolve signaling across multiple regulatory layers. | fitlm (Matlab) sparseNCA | Gene Expression Omnibus NCI-60, HMD, TRRUST, KEGG | In vivo metabolite fold-changes between normal and cancer tissues. |
| [83] | CSGEMs to predict biomarkers and drug targets | CSGEM to study the role of metabolic alterations for novel therapy targets. Predicts 52 cytostatic drug targets (40% by known drugs). Analyze synthetic lethal drug targets to identify drug synergies. | NCI-60 | shRNA data, cytostatic scores for single and double drug target predictions, synergistic drug targets via yeast orthologs. | |
| [173] | GEMs to identify antimetabolites for drug design | Assess anticancer effects of drugs structurally similar to DrugBank [174]. Uses Tanimoto scores from OpenBabel [175] to assess structural similarity between DrugBank drugs and metabolites CSGEM predicted to be essential for maximal growth rate. Developed pyTARG to constrain the HMR, using 34 CCLs and 26 healthy tissue RNA-Seqs. Implemented FBA within PyTARG to quantify the original drug affecting reactions rates decrease. Model the impact of a relative inhibition on global cell metabolism. | OpenBabel [175], pyTARG, COBRApy | DrugBank [174], KEGG, BioProject, HPA, GEO, BioModels | Differential effects of a lipoamide analog on MCF7 and ASM cells. Proof of concept of identification of therapeutic windows. |
| [109] | GEMs to identify antimetabolites for drug design | Proteomics samples from 27 HCC patients and 83 healthy individuals from HPA to reconstruct cancer and healthy GEMs with the tINIT from the HMR 2.0 generic GEM. CSGEMs to identify antimetabolites used as anticancer drugs. Healthy GEMs to explore candidate antimetabolites toxicity on healthy samples. | RAVEN (gap-filling, tINIT, checkTasks) | HPA | Usage of antimetabolites for treatment of HCC demonstrated by the inhibitory effect of the l-carnitine analog, one of the predicted antimetabolites, on the proliferation of the HepG2 CCL. |
| [176] | GEMs to identify metabolic inhibitors to administrate with drug combinations in adaptive therapies | Found that taxane-treated breast cancer cells undergo a metastable transition in which they depend more on oxidative and non-oxidative glucose metabolism conferring them resistance to doxorubicin. Predict that these rewired cancer cells can be effectively targeted when a glucose metabolism inhibitor is co-administered with doxorubicin. | Prism (GraphPad) | In vivo experiments with mouse models, patient explant system. | |
| [177] | GEMs to explore cancer metabolism biology | Central C and N RMGEM to study the interplay between glucose and glutamine for biomass formation in ammonia microenvironment. Perform Warburg effect quantitative. Used the RMGEM to do FBA to study all possible glutamine fates. Found that glutamine can supply C sources for cell energy production and can be used as a C and N source to synthesize essential metabolites. | FAME | ||
| [178] | GEMs to explore cancer metabolism biology | HPaA to explore the prognosis of each protein in 17 major cancers. Uses CSGEMs to identify tumor growth involved genes. Based on transcriptomics of ~ 8000 patients with clinical metadata. Revealed that survival is associated with upregulation mitosis and cell growth genes while downregulated genes are mostly involved in cellular differentiation. | Kaplan–Meier plots, PCA | HPaA [178], BioModels, TCGA, GO, GDC | Immunohistochemistry. |
| [179] | CSGEMs to explore cancer metabolism biology | Merged 374 CSGEMs from the HPaA to reconstruct a generic CRC GEM. Identified the mayor differences between tumor and normal samples in terms of highly perturbed metabolites by applying modules reporter metabolite and reporter subnetworks algorithms. Mayor differences were related to the glutathione, arginine and proline metabolic reprogramming. | PIANO (R) (KEGG and GO enrichment analysis), CRC (Bioprofiler analysis), RAVEN, Kaplan-Meier survival analysis, log-rank p-value | HpaA, BioModels, TCGA, GO, GDC | ODC1, SMOX, SRM and SAT validated in vivo and in vitro, using 15 patients and 4 CRC CCLs. |
| [71] | GEMs to explore cancer metabolism biology | HMR 2.0 and proteomics data in HPA to construct consensus hepatocytes GEM (iHepatocytes2322) that improves previous GEMs including an extensive description of lipid metabolism. GEM used to analyze transcriptomics data from non-alcoholic fatty liver disease patients. | INIT, Reporter Subnetwork analysis checkTasks (RAVEN), RAVEN | HPA, Uniprot, GEO | Hepatocytes biological functions of hepatocyte-specific GEM demonstrated by simulating 256 metabolic functions of HepatoNet, using checkTasks/fitTasks (RAVEN). |
| [90] | GEMs to explore cancer metabolism biology | GPR called S-GPR considering transcripts stoichiometry. Investigate PC cells metabolic effects chronic exposure to an endocrine disruptor. | Mann-Whitney test, FASIMU, Gimme, Mat, pFBA MADE | HMD, Lipid Maps, HMA, GEO | Qualitative comparison between predicted metabolic consumption/productions and the metabolomic and lipidomic experimental measurements. |
| [180] | Warburg effect computer metabolic modeling | Study common and robust metabolic pathways supporting cancer cells (glycolysis, TCA cycle, pentose phosphate, glutaminolysis and oxidative phosphorylation). Propose metabolic targets for anticancer treatments by a constraint-based modeling on integrated data. | COBRA toolbox | GEO | Verified that the in silico kinetic growth curve exhibit a comparable behavior with the experimentally obtained from Hela CCLs. |
| [84] | Warburg effect computer metabolic modeling | GEM human metabolic network accounting stoichiometry and enzyme solvent capacity. Demonstrate that Warburg effect happens since the metabolic adaptation of cancer cells to increase biomass production rate. | BRENDA, SABIO-RK | Correlation between enzyme concentration predictions and expression of 1269 metabolic genes from 60 NCI CCLs. Validated against 1000 flux distributions of two models by ACHR sampling. | |
| [181] | Warburg effect computer metabolic modeling | Expand metabolic efficiency notion by ATP production FBM constrained by glucose uptake and solvent capacities in the cell’s cytoplasm. Found that at low glucose uptake rates mitochondrial respiration is the most efficient pathway for ATP generation while when increasing glucose uptake rates a gradual switch to aerobic glycolysis achieves ATP highest rate since it is more efficient for the required solvent capacity. | Agreement between the experimentally determined fluxes and the model predictions. | ||
| [182] | Warburg effect computer metabolic modeling | Constraint-based modeling with E-Flux integrating 13 different cancer cell transcriptomics with Recon1 generic model. Found that metabolic changes distributions are similar in different cancer types, supporting that Warburg effect is a general metabolic adaptation. | E-Flux, GeWorkbench 2.4.0, COBRA | GEO |
Abbreviations: (C) carbon, (CCL) cancer cell line, (CRC) colorectal cancer, (CSGEM) cancer-specific GEM, (DIRAC) DIfferential RAnk Conservation, (GDC) Genomic Data Commons, (GEO) Gene Expression Omnibus, (GO) Gene Ontology, (HCC) Hepatocellular Carcinoma, (HMA) Human Metabolic Atlas, (HMD) Human Metabolome Database, (HPA) Human Protein Atlas, (HpaA) Human Protein Pathology Atlas, (N) nitrogen, (PC) prostate cancer, (RMGEM) reduced metabolic GEM.