Tissue of origin dictates GOT1 dependence and confers synthetic lethality to radiotherapy

Abstract

Background: Metabolic programs in cancer cells are influenced by genotype and the tissue of origin. We have previously shown that central carbon metabolism is rewired in pancreatic ductal adenocarcinoma (PDA) to support proliferation through a glutamate oxaloacetate transaminase 1 (GOT1)-dependent pathway.

Methods: We utilized a doxycycline-inducible shRNA-mediated strategy to knockdown GOT1 in PDA and colorectal cancer (CRC) cell lines and tumor models of similar genotype. These cells were analyzed for the ability to form colonies and tumors to test if tissue type impacted GOT1 dependence. Additionally, the ability of GOT1 to impact the response to chemo- and radiotherapy was assessed. Mechanistically, the associated specimens were examined using a combination of steady-state and stable isotope tracing metabolomics strategies and computational modeling. Statistics were calculated using GraphPad Prism 7. One-way ANOVA was performed for experiments comparing multiple groups with one changing variable. Student's t test (unpaired, two-tailed) was performed when comparing two groups to each other. Metabolomics data comparing three PDA and three CRC cell lines were analyzed by performing Student's t test (unpaired, two-tailed) between all PDA metabolites and CRC metabolites.

Results: While PDA exhibits profound growth inhibition upon GOT1 knockdown, we found CRC to be insensitive. In PDA, but not CRC, GOT1 inhibition disrupted glycolysis, nucleotide metabolism, and redox homeostasis. These insights were leveraged in PDA, where we demonstrate that radiotherapy potently enhanced the effect of GOT1 inhibition on tumor growth.

Conclusions: Taken together, these results illustrate the role of tissue type in dictating metabolic dependencies and provide new insights for targeting metabolism to treat PDA.

Keywords: CRC; Colorectal cancer; Fluxomics; Metabolomics; NADPH; PDA; Pancreatic cancer; Redox; Stable isotope tracing.

Fig. 1

GOT1 dependence exhibits tissue specificity. a Schematic of the GOT1 pathway in PDA. b Colony number after dox treatment in PDA (red) and CRC (blue) cell lines expressing dox-inducible (iDox) shRNAs against GOT1 (two independent hairpins; shGOT1 #1, shGOT1 #3) relative to a non-targeting hairpin (shNT). Error bars represent s.d. from biological replicates (n = 3). Mutations in KRAS, BRAF, and TP53 are presented in the table below the bar graph. WT, wild type; SM, silent mutation. c Western blots (left) and quantification (right) for GOT1 and vinculin (VCL) loading control from iDox-shGOT1 #1 PDA and CRC tumors. d, e Tumor growth curves and f, g final tumor weights from subcutaneous PDA xenografts (n = 8, BxPC-3 +/−dox tumors; n = 6, PA-TU-8902 +/−dox tumors). Error bars represent s.d. h, i Tumor growth curves and j, k final tumor weights from subcutaneous CRC xenografts (n = 5, DLD-1 +/−dox, HCT 116 +dox tumors; n = 4, HCT 116 −dox tumors). Error bars represent s.d. Tumor growth curves for the corresponding iDox-shNT lines are presented in Additional file 1: Figure S2b. l Western blot (left) and quantification (right) for GOT1 pathway components from a in wild-type PDA and CRC cell lines. AcCoA, acetyl-CoA; αKG, alpha-ketoglutarate; Asp, aspartate; Cit, citrate; Fum, fumarate; Glu, glutamate; GOT1, glutamate oxaloacetate transaminase 1; GOT2, glutamate oxaloacetate transaminase 2; Iso, isocitrate; Mal, malate; MDH1, malate dehydrogenase 1; ME1, malic enzyme 1; NADP+, oxidized nicotinamide adenine dinucleotide phosphate; NADPH, reduced nicotinamide adenine dinucleotide phosphate; OAA, oxaloacetate; Pyr, pyruvate; Suc, succinate. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; Student’s t test (unpaired, two-tailed)

Fig. 2

Metabolic profiles of PDA and CRC. a Relative metabolite levels as determined by LC/MS for glycolysis, pentose phosphate pathway (PPP), and serine metabolism in parental PDA (red) and CRC (blue) cell lines. Uniformly labeled (M+3, hashed bars) and unlabeled (M+0, solid bars) metabolite pools derived from U-¹³C-glucose (Glc, left) or U-¹³C-glutamine (Gln, right) for b lactate (Lac) and c alanine (Ala) as determined by LC/MS. d Relative U-¹³C-Glc-labeling of serine (Ser) and e glycine (Gly), as determined by gas chromatography (GC)/MS. f Ion currents for isotopologue distribution of citrate (Cit) derived from U-¹³C-Glc (left) or U-¹³C-Gln (right) in PDA and CRC cell lines. g Schematic summary of metabolic patterns observed in parental PDA and CRC cells. Red represents increased pool sizes in GOT1-sensitive PDA cells and blue represents increased metabolite pools in GOT1-insensitive CRC cells. 3-pSer, 3-phosphoserine; 3PG, 3-phosphogycerate; 6PG, 6-phosphogluconate; Ac-CoA, acetyl-CoA; αKG, alpha-ketoglutarate; B(1,3)PG, 1,3-bisphosphoglycerate; B(2,3)PG, 2,3-bisphosphoglycerate; Cit, citrate; DHAP, dihydroxyacetone phosphate; E4P, erythrose 4-phosphate; F6P, fructose 6-phosphate; FBP, fructose-1,6-bisphosphate; Fum, fumarate; G3P, glycerol-3-phosphate; G6P, glucose 6-phosphate; GA3P, glyceraldehyde 3-phosphate; GdL6P, glucono-delta-lactone 6-phosphate; Iso, isocitrate; Mal, malate; OAA, oxaloacetate; PEP, phosphoenolpyruvate; PHP, phosphohydroxypyruvate; PRPP, phosphoribosyl pyrophosphate; Pyr, pyruvate; R5P, ribose 5-phosphate; S7P, sedoheptulose-7 phosphate; SBP, sedoheptulose-1,7-bisphosphate; Suc, succinate. Error bars represent s.d. from biological replicates (n = 3). Stacked P values are presented for isotopologues in b–e and correspond by color. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; Student t test (unpaired, two-tailed)

Fig. 3

Metabolic profiles of PDA and CRC following GOT1 inhibition. a Relative aspartate (Asp) and malate (Mal) pools as determined by LC/MS in iDox-shGOT1 #1 PDA and CRC, presented as GOT1 knockdown over mock (+dox/−dox). b Relative glycolysis metabolite pools, as presented in a. c Summary of changes to central carbon metabolism upon GOT1 knockdown PDA cells. Red represents increased pool sizes in PDA cells upon GOT1 knockdown, and gray represents decreased metabolite pools. d Basal extracellular acidification rate (ECAR) levels in iDox-shGOT1 #1 and control shNT PDA and CRC cells, as determined by Seahorse Metabolic Flux Analysis. e Ion currents for isotopologue distribution of Asp derived from U-¹³C-Glc (left) or U-¹³C-Gln (right) in iDox-shGOT1 #1 PA-TU-8902 PDA and DLD-1 CRC cell lines. 3PG, 3-phosphogycerate; Ac-CoA, acetyl-CoA; αKG, alpha-ketoglutarate; Ala, alanine; B(1,3)PG, 1,3-bisphosphoglycerate; B(2,3)PG, 2,3-bisphosphoglycerate; Cit, citrate; DHAP, dihydroxyacetone phosphate; F6P, fructose 6-phosphate; FBP, fructose-1,6-bisphosphate; Fum, fumarate; G6P, glucose 6-phosphate; GA3P, glyceraldehyde 3-phosphate; Iso, isocitrate; Lac, lactate; OAA, oxaloacetate; PEP, phosphoenolpyruvate; Pyr, pyruvate; Suc, succinate. Error bars represent s.d. from biological replicates (n = 3, in a, b, e; n = 5 in d). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; Student t test (unpaired, two-tailed)

Fig. 4

GOT1 inhibition disrupts nucleotide metabolism in PDA. a Fold-change versus P value is plotted per metabolite as the average from three iDox-shGOT1 #1 PDA lines (+dox/−dox) over the average from three iDox-shGOT1 #1 CRC lines (+dox/−dox). Metabolites with filled circles were used for the pathway analysis in Additional file 1: Figure S6a. Metabolites identity is indicated for those with P < 0.05 and fold change +/−2. b Relative nucleic acid pools as determined by LC/MS in iDox-shGOT1 #1 PDA and CRC, presented as GOT1 knockdown over mock (+dox/−dox). c Relative IC50 of gemcitabine, 5-fluorouracil (5-FU), and oxaliplatin in iDox-shGOT1 #1 PDA and CRC cells upon GOT1 knockdown. The dose response curves from which the IC50s were derived are presented in Additional file 1: Figure S7e. ADP, adenosine diphosphate; AMP, adenosine monophosphate; ATP, adenosine triphosphate; CDP, cytidine diphosphate; CMP, cytidine monophosphate; CTP, cytidine triphosphate; dAMP, deoxyadenosine monophosphate; dATP, deoxyadenosine triphosphate; dCDP, deoxycytidine diphosphate; dCMP, deoxycytidine monophosphate; dCTP, deoxycytidine triphosphate; dGDP, deoxyguanosine diphosphate; dGMP, deoxyguanosine monophosphate; dGTP, deoxyguanosine triphosphate; dTDP, deoxythymidine diphosphate; dTMP, deoxythymidine monophosphate; dTTP, deoxythymidine triphosphate; dUMP, deoxyuridine monophosphate; GDP, guanosine diphosphate; GMP, guanosine monophosphate; GTP, guanosine triphosphate; IDP, inosine diphosphate; IMP, inosine monophosphate; UDP, uridine diphosphate; UMP, uridine monophosphate; UTP, uridine triphosphate; XMP, xanthosine monophosphate. Error bars in b and c represent s.d. from biological replicates (n = 3). n.s., not significant; *P < 0.05; **P < 0.01; Student t test (unpaired, two-tailed)

Fig. 5

GOT1 inhibition induces redox imbalance and sensitizes PDA to radiation therapy. a Relative reduced glutathione (GSH), oxidized glutathione (GSSG), and the ratio (GSH/GSSG) pools upon GOT1 knockdown in iDox-shGOT1 #1 PDA and CRC cells. Data were obtained by LC/MS and are normalized as GOT1 knockdown over mock (+dox/−dox). b Relative GSH/GSSG upon GOT1 knockdown as determined by enzymatic assay in PDA and CRC and normalized as GOT1 knockdown over mock (+dox/−dox). c Time course of relative GSH/GSSG pools upon GOT1 knockdown in iDox-shGOT1 #1 PA-TU-8902 PDA cells. Data were obtained by LC/MS and are normalized as GOT1 knockdown over mock (+dox/−dox). d GOT1 protein expression during dox-mediated knockdown time course in iDox-shGOT1 #1 PA-TU-8902 cells. GAPDH serves as the protein loading control. e Total aspartate (Asp) pools in PA-TU-8902 cells, as determined by LC/MS and plotted as GOT1 knockdown over mock (+dox/−dox). f Surviving fraction from clonogenic cell survival assays of radiation-treated iDox-shGOT1 #1 PA-TU-8902 and g HCT 116. Gy, Gray. h Enhancement ratio of radiation-treated iDox-shGOT1 #1 PDA and CRC cells. Error bars represent s.d. from biological replicates in a, c–h (n = 3) and in b (n = 4, iDox-shGOT1 #1 PA-TU-8902, PA-TU-8988 T, HCT 116, DLD-1; n = 3, iDox-shGOT1 #1 MIA PaCa-2, LoVo and iDox-shGOT1 #3). i Tumor growth of iDox-shGOT1 #1 PA-TU-8902 or j iDox-shGOT1 #1 HCT116 xenografts treated with dox (solid arrow; maintained for the duration of the experiment) and/or radiation (rad; dashed arrows). n = 12 tumors per arm, except for dox HCT 116 tumors where n = 10. Error bars represent s.d. k Time to tumor tripling of iDox-shGOT1 #1 PA-TU-8902 or l iDox-shGOT1 #1 HCT116 xenografts. n.s., not significant; *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; Student t test (unpaired, two-tailed) (a, h, i, j); one-way ANOVA (b, c, e, k, l)