Editor's evaluation: Metabolic requirement for GOT2 in pancreatic cancer depends on environmental context
Year: 2021
Type: peer-review
Abstract: Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Mitochondrial glutamate-oxaloacetate transaminase 2 (GOT2) is part of the malate-aspartate shuttle, a mechanism by which cells transfer reducing equivalents from the cytosol to the mitochondria. GOT2 is a key component of mutant KRAS (KRAS*)-mediated rewiring of glutamine metabolism in pancreatic ductal adenocarcinoma (PDA). Here, we demonstrate that the loss of GOT2 disturbs redox homeostasis and halts proliferation of PDA cells in vitro. GOT2 knockdown (KD) in PDA cell lines in vitro induced NADH accumulation, decreased Asp and α-ketoglutarate (αKG) production, stalled glycolysis, disrupted the TCA cycle, and impaired proliferation. Oxidizing NADH through chemical or genetic means resolved the redox imbalance induced by GOT2 KD, permitting sustained proliferation. Despite a strong in vitro inhibitory phenotype, loss of GOT2 had no effect on tumor growth in xenograft PDA or autochthonous mouse models. We show that cancer-associated fibroblasts (CAFs), a major component of the pancreatic tumor microenvironment (TME), release the redox active metabolite pyruvate, and culturing GOT2 KD cells in CAF conditioned media (CM) rescued proliferation in vitro. Furthermore, blocking pyruvate import or pyruvate-to-lactate reduction prevented rescue of GOT2 KD in vitro by exogenous pyruvate or CAF CM. However, these interventions failed to sensitize xenografts to GOT2 KD in vivo, demonstrating the remarkable plasticity and differential metabolism deployed by PDA cells in vitro and in vivo. This emphasizes how the environmental context of distinct pre-clinical models impacts both cell-intrinsic metabolic rewiring and metabolic crosstalk with the TME. Editor's evaluation This paper provides evidence that the environmental redox state shapes metabolic liabilities in pancreatic cancer cells. In vitro, pancreatic cancer cells rely on the malate-aspartate shuttle to regenerate oxidized electron acceptors in the cytosol that are required for cell proliferation; stromal cells within tumors secrete metabolites that serve as alternate electron acceptors, thereby reducing cancer cell reliance on the malate-aspartate shuttle. https://doi.org/10.7554/eLife.73245.sa0 Decision letter Reviews on Sciety eLife's review process Introduction Cancer cells depend on deregulated metabolic programs to meet energetic and biosynthetic demands (Pavlova and Thompson, 2016; Vander Heiden and DeBerardinis, 2017; Martínez-Reyes and Chandel, 2021). Metabolic therapies aim to preferentially target these dependencies (Vander Heiden, 2011). This approach has shown promise in preclinical models of pancreatic ductal adenocarcinoma (PDA) – one of the deadliest major cancers, notoriously resistant to anti-cancer therapies (Ryan et al., 2014; Halbrook and Lyssiotis, 2017). Pancreatic tumors are poorly vascularized and nutrient dysregulated (Kamphorst et al., 2015). Therefore, cancer cells commandeer metabolic pathways to scavenge and use nutrients (Halbrook and Lyssiotis, 2017; Encarnación-Rosado and Kimmelman, 2021). A wealth of recent literature has identified that this is mediated predominantly by mutant KRAS (KRAS*), the oncogenic driver in most pancreatic tumors (Commisso et al., 2013; Ying et al., 2012; Son et al., 2013; Viale et al., 2014; Santana-Codina et al., 2018; Humpton et al., 2019). The KRAS* has also been implicated in shaping the pancreatic tumor microenvironment (TME; Tape et al., 2016; Kerk et al., 2021). The PDA tumors exhibit a complex TME (Storz and Crawford, 2020; Zhang et al., 2019) with metabolic interactions between malignant, stromal, and immune cells enabling and facilitating tumor progression (Lyssiotis and Kimmelman, 2017). Recent successes in drug development have provided KRAS* selective inhibitors, and these are in various stages of preclinical and clinical testing. However, consistent with other targeted therapies, resistance inevitably occurs (Janes et al., 2018; Jänne, 2022). Therefore, disrupting downstream metabolic crosstalk mechanisms in PDA is a compelling combinatorial or alternative approach (Cox et al., 2014). In support of this idea, previous work from our lab described that PDA cells are uniquely dependent on KRAS*-mediated rewiring of glutamine metabolism for protection against oxidative stress (Son et al., 2013). Mitochondrial glutamate-oxaloacetate transaminase 2 (GOT2) is implicated in this rewired metabolism in PDA. In normal physiology, GOT2 functions in the malate-aspartate shuttle (MAS), a mechanism by which cells transfer reducing equivalents between the cytosol and mitochondria to balance the two independent NADH pools and maintain redox balance (Figure 1A). The PDA cells driven by KRAS* divert metabolites from the MAS and increase flux through malic enzyme 1 (ME1) to produce NADPH (Son et al., 2013). Since this pathway is critical for PDA, we set out to evaluate GOT2 as a potential therapeutic target. Using metabolomics analyses and manipulation of the redox state in PDA cells, we discovered that loss of GOT2 in vitro induces intracellular NADH accumulation and reductive stress. These metabolic changes impair cellular growth, which can be rescued with chemical or genetic interventions that oxidize NADH. However, loss of GOT2 had no effect on tumor growth or initiation in immunocompromised or immunocompetent mouse models of PDA. Cancer cells use a complex cell-intrinsic rewiring and crosstalk with the TME to maintain redox homeostasis in vivo. These data emphasize an underappreciated role for GOT2 in pancreatic tumor redox homeostasis and illustrate the differential biochemical pathways and metabolic plasticity deployed by cancer cells in vivo. Figure 1 with 2 supplements see all Download asset Open asset Glutamate-oxaloacetate transaminase 2 (GOT2) knockdown (KD) impairs in vitro pancreatic ductal adenocarcinoma (PDA) proliferation. (A) The metabolic roles of mitochondrial GOT2. OAA: oxaloacetate; Glu: glutamate; Mal: malate; Asp: aspartate; αKG: α-ketoglutarate; ME1: malic enzyme 1; TCA: tricarboxylic acid. (B) Immunoblot of GOT2 and VINCULIN loading control in PaTu-8902 cells after 1 µg/mL doxycycline (DOX) induction of two independent GOT2 (sh1, sh2) and non-targeting (NT) shRNAs for 3 days. (C) Representative images from colony formation assays in ishRNA PaTu-8902 cells −Dox (n=3) or +Dox (n=3). (D) Heatmap summarizing the relative colony formation ishRNA PDA cell lines −Dox (n=3) or +Dox (n=3), normalized to −Dox for each indicated shRNA. Representative images from colony formation assays and western blots presented in Figure 1—figure supplement 1. (E) Relative abundances of Asp and αKG in PaTu-8902 (left) and MIAPaCa-2 (right) ishGOT2.1 −Dox (n=3) or +Dox (n=3). (F) Relative proliferation of PaTu-8902 (left) and MIAPaCa-2 (right) ishGOT2.1 −Dox (n=3) or +Dox (n=3) cultured in normal media (Dulbecco's modified Eagle medium, DMEM) or supplemented with 20 mM Asp and 1 mM dimethyl-αKG (dmαKG). For all panels, data represent mean ± SD. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. Figure 1—source data 1 Full western blot images for Figure 1B. https://cdn.elifesciences.org/articles/73245/elife-73245-fig1-data1-v2.zip Download elife-73245-fig1-data1-v2.zip Results Loss of GOT2 impairs PDA cell proliferation in vitro To expand on our previous work studying GOT2 in PDA (Son et al., 2013) and to evaluate GOT2 as a potential therapeutic target, we generated a panel of PDA cell lines with doxycycline-inducible expression of either a control non-targeting shRNA (shNT) or two independent shRNAs (sh1, sh2) targeting the GOT2 transcript. Cells cultured in media containing doxycycline (+Dox) exhibited a marked decrease in GOT2 protein expression compared to cells cultured in media without doxycycline (−Dox) (Figure 1B; Figure 1—figure supplement 1A). This knockdown (KD) was specific for GOT2, relative to the cytosolic aspartate aminotransaminase GOT1 (Figure 1—figure supplement 1B). Having validated GOT2 KD, we tested the importance of GOT2 for cellular proliferation. In general, GOT2 KD in PDA cells impaired colony formation (Figure 1C and D; Figure 1—figure supplement 1C) and proliferation (Figure 1—figure supplement 1D). Consistent with our previous report (Son et al., 2013), GOT2 was not required for the proliferation of non-transformed pancreatic cell types (Figure 1—figure supplement 1E,F). Since GOT2 has several vital metabolic roles in a cell (Figure 1A), the changes caused by decreased GOT2 expression in PDA cells were examined using liquid chromatography-coupled tandem mass spectroscopy (LC-MS/MS). Numerous changes in the intracellular metabolome of GOT2 KD cells were observed (Figure 1—figure supplement 2A,B). Of note, the products of the GOT2-catalyzed reaction, aspartate (Asp) and α-ketoglutarate (αKG), were decreased (Figure 1E), and supplementation of these metabolites rescued growth of GOT2 KD (Figure 1F). While these PDAC cell lines do not express Asp transporters, we confirmed that supraphysiological levels of Asp (20-fold excess) led to an increase in intracellular Asp (Figure 1—figure supplement 2C). In addition to reduced αKG, there was a disruption in tricarboxylic acid (TCA) cycle intermediates, consistent with a role for GOT2 in facilitating glutamine anaplerosis (Figure 1—figure supplement 2D). GOT2 KD perturbs redox homeostasis in PDA cells Aside from the expected decrease in Asp and αKG, and the perturbation of the TCA cycle, closer examination of the GOT2 KD metabolomics dataset revealed an impairment in glycolysis with a node at glyceraldehyde 3-phosphate dehydrogenase (GAPDH), indicative of NADH reductive stress (Figure 2A). Examination of glycolytic rate via Seahorse Flux Analysis confirmed that glycolysis was indeed impaired in GOT2 KD cells (Figure 2—figure supplement 1A). The GAPDH reduces NAD+ to produce NADH, where a buildup of NADH product inhibits GAPDH activity. Indeed, metabolite pools in upstream glycolysis and branch pathways like the pentose phosphate pathway are increased, and those in downstream glycolysis are decreased (Figure 2B). In cultured PDA cells, the MAS transfers glycolytic reducing potential to drive the electron transport chain (ETC) and support the maintenance of cytosolic redox balance (Figure 1A). We thus hypothesized that GOT2 KD interrupted this shuttle, preventing the proper transfer of electron potential in the form of NADH between these two compartments. In support of this, GOT2 KD increased the intracellular ratio of NADH to NAD+ (Figure 2C). Figure 2 with 3 supplements see all Download asset Open asset Glutamate-oxaloacetate transaminase 2 (GOT2) knockdown (KD) induces reductive stress, which can be ameliorated by NADH turnover. (A) Schematic of glycolytic signature induced by GOT2 KD-mediated NADH buildup and reductive stress. G6P: glucose-6-phosphate; F6P: fructose-6-phosphate; FBP: fructose-1,6-bisphosphate; DHAP: dihydroxyacetone phosphate; GA3P: glyceraldehyde-3-phosphate; PEP: phosphoenol pyruvate; oxPPP: oxidative pentose phosphate pathway; non-oxPPP: non-oxidative pentose phosphate pathway; X5P: xylulose-5-phosphate; R5P: ribose-5-phosphate; S7P: seduoheptulose-7-phosphate; HBP: hexosamine biosynthesis pathway; and GlcNAc-1P: N-acetylglucosamine 1-phosphate. (B) Relative fold changes in the indicated metabolites between PaTu-8902 ishGOT2.1 −Dox (n=3) and +Dox (n=3). 2PG: 2-phosphoglycerate; Pyr: pyruvate; and Lac: lactate. (C) Relative NADH/NAD+ ratio in PaTu-8902 ishGOT2.1 −Dox (n=3) and +Dox (n=3). (D) Relative colony formation of MIAPaCa-2 ishRNA −Dox (n=3) or +Dox (n=3) cultured in normal media (Dulbecco's modified Eagle medium, DMEM) or DMEM with 1 mM pyruvate, normalized to −Dox for each condition. (E) Relative colony formation of MIAPaCa-2 ishGOT2.1 −Dox (n=3) or +Dox (n=3) cultured in normal media (DMEM) or DMEM with the indicated concentrations of pyruvate (mM), normalized to −Dox for each condition. (F) Relative proliferation of PaTu-8902 (left) and MIAPaCa-2 (right) ishGOT2.1 −Dox (n=3) or +Dox (n=3) cultured in normal media (DMEM) or DMEM with 1 mM Pyr or α-ketobutyrate (αKB), normalized to Day 0 for each condition. (G) Relative proliferation of PaTu-8902 (left) and MIAPaCa-2 (right) ishGOT2.1 −Dox (n=3) or +Dox (n=3) expressing doxycycline-inducible empty vector (EV), cytosolic Lactobacillus NADH oxidase (LbNOX), or mitochondrial LbNOX (mLbNOX), normalized to Day 0 for each condition. (H) Relative NADH/NAD+ ratio of PaTu-8902 (left) and MIAPaCa-2 (right) ishGOT2.1 −Dox (n=3) or +Dox (n=3) expressing EV, LbNOX, or mLbNOX. (I) Schematic of 13C3-pyruvate into relevant metabolic pathways. 13C-carbon labels in blue, non-labeled carbon in gray. LDH: lactate dehydrogenase; GPT: glutamate-pyruvate transaminase; MPC: mitochondrial pyruvate carrier; and PC: pyruvate carboxylase. (J–K) Fractional labeling of intracellular pyruvate (J) or lactate (K) in PaTu-8902 and MIAPaCa-2 ishGOT2.1 −Dox (n=3) or +Dox (n=3) cultured in 1 mM 13C3-pyruvate and treated with DMSO vehicle control or 5 µM UK5099 (MPC inhibitor). Unlabeled controls presented at right. For all panels, data represent mean ± SD. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. NADH accumulation leads to reductive stress, which can be relieved if the cell has access to electron acceptors (Gui et al., 2016). Pyruvate is a notable metabolite in this regard, as it can accept electrons from NADH, producing lactate and regenerating NAD+, in a reaction catalyzed by lactate dehydrogenase (LDH). Therefore, we hypothesized that pyruvate could rescue the defect in cellular proliferation mediated by GOT2 KD. Indeed, culturing GOT2 KD cells in pyruvate rescued proliferation in a dose-dependent manner (Figure 2D and E; Figure 2—figure supplement 1B). Additionally, cells expressing a genetically encoded, fluorescent ATP sensor indicated that ATP levels dropped with GOT2 KD and were restored with pyruvate supplementation (Figure 2—figure supplement 1C), reflecting the link between TCA cycle activity, respiration, and oxidative phosphorylation. Furthermore, having identified a metabolite that permits in vitro proliferation of PDA cells without GOT2, we engineered CRISPR-Cas9 GOT2 knockout (KO) cells for further investigation (Figure 2—figure supplement 1D). In support of the data generated using the doxycycline-inducible shRNA, GOT2 KO impaired colony formation of PDA cells, which was similarly restored through extracellular pyruvate supplementation (Figure 2—figure supplement 1E,F). The α-ketobutyrate (αKB) is another electron acceptor that turns over NADH in a mechanism analogous to pyruvate but without entering downstream metabolism in the same fashion as pyruvate (Sullivan et al., 2015). The αKB also rescued proliferation after GOT2 KD (Figure 2F; Figure 2—figure supplement 1F). This mechanism is dependent on NADH turnover and not NAD+ synthesis, as the NAD+ precursor nicotinamide mononucleotide (NMN) failed to rescue GOT2 KD (Figure 2—figure supplement 1G). To test this further, GOT2 KD cells were engineered to express either doxycycline-inducible cytosolic or mitochondrial Lactobacillus brevis NADH oxidase (LbNOX), which uses molecular oxygen to oxidize NADH and produce water and NAD+ (Figure 2—figure supplement 2A; Goodman et al., 2020; Titov et al., 2016). Cytosolic LbNOX, but not mitochondrial LbNOX (mLbNOX), rescued proliferation of GOT2 KD cells (Figure 2G; Figure 2—figure supplement 2B). We confirmed the mLbNOX construct encoded a functional enzyme as defects in proliferation incurred by treatment with complex I inhibitor piericidin could be rescued with mLbNOX as reported previously (Figure 2—figure supplement 2C; Titov et al., 2016). LbNOX reversed the increased NADH/NAD+ ratio induced by GOT2 KD via an overall decrease in NADH levels (Figure 2H; Figure 2—figure supplement 2D). Furthermore, the secreted pyruvate/lactate ratios in the media of LbNOX-expressing cells dramatically increased, indicating a resolution of the cytosolic NADH stress induced by GOT2 KD (Figure 2—figure supplement 2E,F). Moreover, the metabolic defects observed following GOT2 KD were ameliorated by cytosolic LbNOX activity, including an increase in Asp and αKG, normalization of TCA cycle metabolites, and the release of the glycolytic block at GAPDH (Figure 2—figure supplement 2G). The spatial control of the LbNOX system indicated that KD of mitochondrial GOT2 could be rescued by balancing the cytosolic NADH/NAD+ pool. Next, we traced the metabolic fate of U13C-pyruvate in cells with GOT2 KD. Importantly, we also assessed the impact and metabolism of pyruvate in this system following inhibition of the mitochondrial pyruvate carrier (MPC) inhibitor UK5099, which blocks entry of pyruvate into the mitochondria. (Figure 2I). Media containing equimolar unlabeled glucose and U13C-pyruvate was used to prevent dilution of the pyruvate label by unlabeled pyruvate generated via glycolysis from the high glucose concentration used in normal media. This media formulation had no impact on the pyruvate GOT2 KD rescue phenotype (Figure 2—figure supplement 3A). In this experiment, roughly 50% of the intracellular pyruvate was labeled (Figure 2J), in line with the other half of the unlabeled pyruvate coming from glucose, and most of the labeled pyruvate was converted to lactate (Figure 2K). We observed modest labeling of citrate, Asp, and alanine from pyruvate (Figure 2—figure supplement 3B-D). While the labeled TCA cycle and branching pathway intermediates decreased dramatically with UK5099, this MPC inhibition had no effect on the pyruvate GOT2 KD rescue phenotype (Figure 2—figure supplement 3E). Both lactate and alanine can fuel oxidation via the TCA cycle (Hui et al., 2017; Sousa et al., 2016), and Asp is critical for nucleotide production. Pyruvate can be converted into all three of these metabolites, yet supplementation with exogenous lactate, alanine, or nucleoside bases failed to rescue GOT2 KD (Figure 2—figure supplement 3F-H). These data collectively suggest a redox role for pyruvate-mediated rescue of GOT2 KD, as opposed to entering the mitochondria for ATP production or biosynthesis. These findings provide clear evidence using several orthogonal strategies that GOT2 KD results in accumulation of NADH pools and reductive stress in PDA cells. GOT2 is not required for PDA tumor growth in vivo To test the effect of GOT2 KD on in vivo tumor growth, PDA cell lines were injected subcutaneously into the flanks of immunocompromised (non-obese diabetic (NOD) scid gamma; NSG) mice. We allowed the tumors to establish for 7 days, after which the mice were fed normal chow or doxycycline chow (Dox chow) ad libitum. Surprisingly, despite the inhibitory in vitro phenotype, and robust suppression of GOT2 expression in vivo, PDA tumors from five different cell lines grew unimpeded with GOT2 KD (Figure 3A and B; Figure 3—figure supplement 1A,B). Nuclear Ki67 staining confirmed that tumors lacking GOT2 were proliferative and actually displayed a modest, but significant, increase in Ki67-positive nuclei (Figure 3C and D). To further examine the role of GOT2 in the proper tissue context, PDA cells were injected orthotopically into the pancreas of NSG mice, and tumors were allowed to establish for 7 days before feeding the mice regular or Dox chow. Similar to the flank model, GOT2 KD had no effect on the growth of orthotopic tumors (Figure 3—figure supplement 1C). Figure 3 with 1 supplement see all Download asset Open asset Glutamate-oxaloacetate transaminase 2 (GOT2) is not required for in vivo growth of pancreatic ductal adenocarcinoma (PDA) xenografts. (A) Tumor volumes of PaTu-8902 ishRNA flank xenografts in NOD scid gamma (NSG) mice fed normal chow (−Dox, n=6) or doxycycline chow (+Dox, n=6). Arrows indicate administration of Dox chow 1 week after PDA cell injection. (B) Tumor volumes of four additional PDA cell line ishGOT2.1 flank xenografts in NSG mice fed normal chow (−Dox, n=6) or doxycycline chow (+Dox, n=6). Arrows indicate administration of Dox chow 1 week after PDA cell injection. (C) Immunohistochemistry for Ki67 in flank xenograft tissue from PaTu-8902 ishRNA −Dox (n=6) or +Dox (n=6). Scale bar is 100 µm. (D) Quantification of Ki67+ cells in tissue depicted in (C). (E) Relative abundances of Asp and αKG in PaTu-8902 ishGOT2.1 −Dox (n=6) or +Dox (n=6) flank xenografts. (F) Relative fold changes in the indicated metabolites between PaTu-8902 ishGOT2.1 −Dox (n=6) and +Dox (n=6) flank xenografts. G6P: glucose-6-phosphate; F6P: fructose-6-phosphate; FBP: fructose-1,6-bisphosphate; DHAP: dihydroxyacetone phosphate; GA3P: glyceraldehyde-3-phosphate; 2PG: 2-phosphoglycerate; PEP: phosphoenol pyruvate; Pyr: pyruvate; Lac: Lactate; X5P: xylulose-5-phosphate; R5P: ribose-5-phosphate; S7P: seduoheptulose-7-phosphate; and P: pentose phosphate pathway. For all panels, data represent mean ± SD. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. Having observed a discrepancy between in vitro and in vivo dependence on GOT2 for proliferation, the relative abundances of intracellular metabolites from flank tumors were analyzed via LC-MS/MS to compare the metabolic changes between cell lines and tumors following loss of GOT2. While GOT2 KD induced some changes in tumor metabolite levels, the affected metabolic pathways were distinct from those observed in vitro, bearing in mind we were comparing homogenous cell lines with heterocellular xenografts (Figure 3E and F; Figure 3—figure supplement 1D-F). Asp abundance was significantly decreased, yet αKG levels remained constant (Figure 3E), and TCA cycle intermediates were unaffected (Figure 3—figure supplement 1F). This led us to initially hypothesize that PDA cells rewire their metabolism in vivo to maintain αKG levels when GOT2 is knocked down. However, upon examination of the expression of other αKG-producing enzymes in GOT2 KD tumors, we did not a increase in expression (Figure 3—figure supplement 1G). expression not metabolic but these data led us to an to the different in vitro and in vivo GOT2 KD the glycolytic signature indicative of NADH stress was not observed in the metabolomics from flank GOT2 KD tumors, further the differential dependence on GOT2 in PDA in vitro and in vivo (Figure To evaluate the role of GOT2 in an immunocompetent model, we the mouse with mice to an mouse (Figure Figure supplement 1A). Loss of had no effect on the of the non-transformed pancreas (Figure supplement This is in support of our data demonstrating loss of GOT2 was not in pancreatic cell types (Figure 1—figure supplement 1F). Figure with 1 supplement see all Download asset Open asset Glutamate-oxaloacetate transaminase 2 (GOT2) not impact on pancreatic ductal adenocarcinoma (PDA) in an autochthonous (A) with expression of mutant driven by on an immunocompetent (B) Representative for in from or Scale bar is (C) Representative and staining of from or (n=6) mice. Scale bar is (D) of or (n=6) mice. (E) of staining from (C) of tissue with cells, or pancreatic (F) Representative staining of from or mice. Scale bar is (G) of or mice. (H) of staining from (F) of tissue with cells, or (I) Representative staining of from (n=3) or mice. Scale bar is 100 (J) of or tissue that had to at 1 For all panels, data represent mean ± SD. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. and controls were to and at which were We confirmed loss of GOT2 in the via Figure were observed in the of between and mice (Figure the and from these by a revealed that mice had a significantly of cells compared to controls (Figure and However, no were observed in the of or pancreatic between and mice (Figure Additionally, we mice to and compared the to a decrease in for pancreatic (Figure and A by a did not a in or of (Figure In both and mice had to after for 1 (Figure and This that loss of not the progression of PDA following by oncogenic conditioned media colony formation in GOT2 KD cells in vitro PDA tumors a complex microenvironment of a immune a robust response of stromal cell and a extracellular Zhang et al., 2019). While the flank tumor in immunocompromised mice is complex that of a PDA staining revealed that mouse fibroblasts a of the microenvironment in tumors of GOT2 (Figure supplement 1A). Additionally, we and have previously reported mechanisms by which cancer-associated fibroblasts in the in metabolic crosstalk with pancreatic cancer cells et al., 2016; et al., 2016; et al., 2019). we hypothesized that PDA metabolism following GOT2 KD. To potential metabolic crosstalk in a PDA cells were cultured in vitro with from In support of our colony formation in PDA cells with GOT2 KD in a dose-dependent manner (Figure and B; Figure supplement 1B). Furthermore, displayed a colony formation rescue phenotype compared to from or from PDA cells (Figure supplement 1C). Figure 5 with 2 supplements see all Download asset Open asset Pancreatic cancer-associated fibroblasts release pyruvate and for loss of glutamate-oxaloacetate transaminase 2 (GOT2) in vitro. (A) Relative colony formation of MIAPaCa-2 ishGOT2.1 −Dox (n=3) or +Dox (n=3) cultured in normal media (Dulbecco's modified Eagle medium, DMEM) or CAF conditioned media (CM) generated after of normalized to −Dox for each condition. (B) Relative colony formation of MIAPaCa-2 ishGOT2.1 −Dox (n=3) or +Dox (n=3) cultured in DMEM or indicated of normalized to −Dox for each condition. (C) Relative colony formation of MIAPaCa-2 ishGOT2.1 −Dox (n=3) or +Dox (n=3) cultured in 3 or that to Data normalized to −Dox for each condition. (D) fold changes of metabolite abundances in compared to normal (E) of pyruvate concentrations in three generated of or mouse CAF including a DMEM (F) Relative colony formation of MIAPaCa-2 ishGOT2.1 −Dox (n=3) or +Dox (n=3) cultured in normal media (DMEM) or normalized to −Dox for each condition. (G) abundance of extracellular or intracellular pyruvate (left) or lactate (right) from cultured with mM (H) Fractional labeling of intracellular glycolysis tricarboxylic acid (TCA) cycle and pentose phosphate pathway metabolites in cultured with mM F6P: fructose-6-phosphate; FBP: fructose-1,6-bisphosphate; DHAP: dihydroxyacetone phosphate; glyceraldehyde-3-phosphate; 2PG: 2-phosphoglycerate; PEP: phosphoenol pyruvate; αKG: α-ketoglutarate; Mal: malate; Asp: aspartate; S7P: seduoheptulose-7-phosphate; and R5P: For all panels, data represent mean ± SD. *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. To to the in for this was through a 3 or to of and In each of these colony formation in GOT2 KD cells, the relevant was a metabolite
Source: eLife
Author: Lydia WS Finley
Institution: Memorial Sloan Kettering Cancer Center
Language: English
Subfield: Cancer Research
Domain: Life Sciences
Open Access status: gold
Estimated APC paid: $2,500
OpenAlex ID: W4288073100