Fig 1.
Synthetic Lethality in 2 different environmental contexts.
(a) Example metabolic network under Culture Medium 1 (CM1). We have 5 reactions (rE1, r1, r2, r3, r4), 4 metabolites (M1, A, B, C) and 3 genes (g1, g2, g3). rE1 is an input exchange reaction and represents the availability of M1; (b) g1 and g2 are synthetic lethals under CM1. (c) If g2 is not active (red dotted line), the inhibition of g1 is essential under CM1. (d) Example metabolic network under Culture Medium 2 (CM2). With respect to the metabolic network in (a), we have 2 additional reactions (rE2, r5) and 1 additional metabolite (M2). rE2 is an input exchange reaction and represents the availability of M2; (e) g1 and g2 are not synthetic lethal under CM2 due to the alternative pathway via M2 degradation (in green). (f) The inhibition of g1 and g2 and deprivation of M2 in the metabolic network from (d) renders cellular proliferation impossible. (g) gMCSs and ngMCSs for CM1 and CM2. Gray dotted lines stand for not active/not present reactions/metabolites.
Fig 2.
Predicted synthetic lethals involving DHFR with thymidine and hypoxanthine in the growth medium.
(a) 4 genetic Minimal Cut Sets (gMCSs) and 2 nutrient-genetic Minimal Cut Sets (ngMCSs) involving DHFR. They were derived from Recon3D under the RPMI1640 growth medium plus thymidine and hypoxanthine; (b) Simplified network of metabolites and enzymes implied in the synthesis of purines and pyrimidines, emphasizing the role of DHFR, Thymidine and Hypoxanthine. Abbreviations: DHF: dihydrofolate; THF: tetrahydrofolate; IMP: inosinic acid; dTMP: 5-Thymidylic acid; DHFR: dihydrofolate reductase; PNP: purine nucleoside phosphorylase; HPRT1: hypoxanthine phosphoribosyltransferase 1; XDH: xanthine dehydrogenase; GMPS: Guanine Monophosphate Synthase; TK1: thymidine kinase 1; TK2: thymidine kinase 2; SLC29A2: solute carrier family 29 member 2.
Fig 3.
In-vitro experimental validation of DHFR inhibition with hypoxanthine and thymidine in the growth medium.
a) GI50 values of MTX for PF-382, JVM-2 and HT-29 cell lines. b) Proliferation of PF-382, JVM-2 and HT-29 cell lines treated with different doses of MTX for 96h in presence or absence of Hypoxanthine-Thymidine. Data represent mean ± standard deviation of four experiments. c) Proliferation of PF-382, JVM-2 and HT-29 cell lines nucleofected with siRNAs targeted to DHFR gene in presence or absence of Hypoxanthine-Thymidine was studied by MTS at day 6 after nucleofection. The proliferation percentage refers to cells nucleofected with a negative control siRNA. Data represent mean ± standard deviation of at least three experiments. d) mRNA expression of DHFR gene 48 h after nucleofection with the specific siRNAs. Data are referred to GUS gene and an experimental group nucleofected with negative control siRNA. Data represent mean ± standard deviation of four experiments. All p-values (p) were determined by one-tailed Wilcoxon tests. Abbreviations: MTX: methotrexate; HT: Hypoxanthine-Thymidine.
Fig 4.
Extracellular nutrient dependences of in vitro cancer cell lines.
a) Relevant ngMCSs involving L-Asparagine, L-Glutamine, L-Arginine, Myo-Inositol and Cholesterol. b) Expression levels of the most limiting gene in the ngMCSs associated with Cholesterol in 18 auxotrophic and 18 prototrophic cell lines. Auxotrophic cell lines were defined as those where the most limiting gene has an expression level below 1 TPM. For comparison, we took the same number of prototrophic cell lines; c) CERES scores for LDLR in the cell lines shown in b). The more negative CERES scores are, the more cellular proliferation will be decreased; d) Expression levels of the most limiting gene in the ngMCSs associated with Myo-Inositol in 45 auxotrophic and 45 prototrophic cell lines. e) CERES scores for SLC5A3 in the cell lines shown in d). Abbreviations: ‘ALL’ and ‘AML’ refer to Acute Lymphoblastic Leukemia and Acute Myeloid Leukemia, respectively. In addition, ‘Aux.’ refers to auxotrophic cell lines, e.g. Aux. ALL refers to auxotrophic ALL cell lines. Note here that in panels (b)-(d) gene expression data was obtained from CCLE [9] and in panels (c)-(e) CERES scores were obtained from the DepMap platform [8]. The significance of the difference in LDLR and SLC5A3 CERES scores in the auxotrophic groups against the corresponding prototrophic group was assessed via a one-tailed Wilcoxon test.