Glutamine supports pancreatic cancer growth through a KRAS-regulated metabolic pathway

Abstract

Cancer cells have metabolic dependencies that distinguish them from their normal counterparts. Among these dependencies is an increased use of the amino acid glutamine to fuel anabolic processes. Indeed, the spectrum of glutamine-dependent tumours and the mechanisms whereby glutamine supports cancer metabolism remain areas of active investigation. Here we report the identification of a non-canonical pathway of glutamine use in human pancreatic ductal adenocarcinoma (PDAC) cells that is required for tumour growth. Whereas most cells use glutamate dehydrogenase (GLUD1) to convert glutamine-derived glutamate into α-ketoglutarate in the mitochondria to fuel the tricarboxylic acid cycle, PDAC relies on a distinct pathway in which glutamine-derived aspartate is transported into the cytoplasm where it can be converted into oxaloacetate by aspartate transaminase (GOT1). Subsequently, this oxaloacetate is converted into malate and then pyruvate, ostensibly increasing the NADPH/NADP(+) ratio which can potentially maintain the cellular redox state. Importantly, PDAC cells are strongly dependent on this series of reactions, as glutamine deprivation or genetic inhibition of any enzyme in this pathway leads to an increase in reactive oxygen species and a reduction in reduced glutathione. Moreover, knockdown of any component enzyme in this series of reactions also results in a pronounced suppression of PDAC growth in vitro and in vivo. Furthermore, we establish that the reprogramming of glutamine metabolism is mediated by oncogenic KRAS, the signature genetic alteration in PDAC, through the transcriptional upregulation and repression of key metabolic enzymes in this pathway. The essentiality of this pathway in PDAC and the fact that it is dispensable in normal cells may provide novel therapeutic approaches to treat these refractory tumours.

Figure 1. PDAC utilize a non-canonical glutamine metabolism pathway

a, PDAC proliferation requires both glucose and glutamine. Cells were plated in the complete media (10mM glucose and 2mM Gln) which was replaced the following day with glucose or Gln-free medium supplemented with 10% dialyzed FBS. b, Relative clonogenic growth of 8988T cells expressing a control shRNA (shGFP) or two independent shRNAs to GLS1. c, Schematic overview of GLUD1- or transaminase-mediated Glu metabolism. d, Relative clonogenic growth of 8988T cells. αKG (4mM), NEAA mixture (0.1 mM glycine, alanine, aspartate, asparagine, proline and serine) or the combination was added to media following Gln-withdrawal. αKG, α-ketoglutarate; Glc, glucose; Gln, glutamine; NEAA, non-essential amino acid. Error bars represent s.d. of triplicate wells from a representative experiment. **, p< 0.01.

Figure 2. GOT1 is essential for redox balance and growth in PDAC

a, Relative clonogenic growth of 8988T cells expressing a control shRNA (shGFP) or two independent shRNAs targeting GLUD1, GOT1, GPT2 or PSAT1. Error bars represent s.d. of triplicate wells from a representative experiment. b and h, Relative metabolite abundance in 8988T cells grown in [U-¹³C₅]-Gln upon GOT1 or ME1 knockdown. Data are presented as the total metabolite pool (encompassing both metabolite derived from Gln and that not Gln-derived, left) and the [U-¹³C]-labeled and Gln-derived metabolite pool (right). Error bars represent the s.d. of three independently prepared samples. c, d, e and f, Relative ROS levels in 8988T cells under conditions indicated as determined by DCFDA staining. DCFDA assay was performed 24hr after supplementing Gln-free media with either OAA (4mM) or dimethyl malate (4mM). Each bar represents the mean of three independent experiments with error bars representing the s.d. g, Schematic depiction of the cytoplasmic reactions that convert Asp into pyruvate. i, NADP+/NADPH ratio in 8988T cells expressing a control shRNA (shGFP), or an shRNA to G6PD, IDH1, GOT1 or ME1. Error bars represent s.d. of five replicate wells from a representative experiment. j, Flux of the Gln carbon skeleton into downstream metabolites as a function of time. The reads for uniformly ¹³C-labeled metabolites, presented in ion current, are plotted for cells expressing the shGFP control or shRNA to GOT1. Asp, aspartate; OAA, oxaloacetate; Glu, glutamate; GSH, reduced glutathione; GSSH, oxidized glutathione. *, p< 0.05; **, p< 0.01.

Figure 3. Metabolism of the Gln carbon skeleton through GOT1, MDH1 and ME1 supports PDAC growth by maintaining redox balance

a and b, Relative clonogenic growth of 8988T cells expressing a control shRNA (shGFP) or two independent shRNAs to MDH1 or ME1. c and e, Relative clonogenic growth of 8988T cells under conditions indicated. Cells were plated in complete culture media (10mM glucose and 2mM Gln), which was replaced the following day with Gln-free medium supplemented with OAA (4mM), GSH (4mM) or N-acetylcysteine (NAC) (4mM). d and f, Relative clonogenic growth of 8988T cells expressing a control shRNA (shGFP) or two independent shRNAs to GOT1 with or without OAA (4mM), GSH (4mM) or NAC (4mM). Error bars represent the s.d. of triplicate wells from a representative experiment. g, Xenograft growth of 8988T cells expressing a control shRNA (shGFP), shRNAs to GOT1 (#1 and #2), MDH1 (#1 and #2) or ME1 (#1 and #2) in mice. Error bars represent s.e.m (n=10). *, p< 0.05; **, p< 0.01.

Figure 4. Oncogenic Kras mediates Gln reprogramming in PDAC

a, Expression of GLUD1 and GOT1 was determined by quantitative RT-PCR in 8988T cells expressing a control shRNA (shGFP) or two independent shRNAs targeting Kras. Western blot confirmed knockdown of Kras expression. b, The effect of Kras knockdown on GLUD1 or GOT1 protein levels in Panc1 cells expressing a doxycycline-inducible Kras shRNA. c, Relative clonogenic growth of 8988T cells expressing a control shRNA (shGFP) or shRNA to Kras following treatment with AOA or EGCG. Error bars represent the s.d. of triplicate wells from a representative experiment. d, Relative metabolite abundance in Miapaca2 cells grown in [U-¹³C₅]-Gln upon doxycycline-inducible Kras knockdown. Data are presented as the total metabolite pool (encompassing both metabolite derived from Gln and that not Gln-derived, left) and the [U-¹³C]-labeled and Gln-derived metabolite pool (right). Error bars represent the s.d. of three independently prepared samples. e, Relative cell viability of 8988T cells treated with GLS inhibitors 968 (active) (10µM), 365 (inactive analog) (50µM), or BPTES (100nM) with increasing concentrations of H₂O₂. . Error bars represent the s.d. of triplicate wells from a representative experiment. f, Model depicting the Kras-regulated Gln metabolic reprogramming in PDAC used to maintain redox and support growth. AOA, aminooxyacetate; EGCG, epigallocatechin gallate. *, p< 0.05; **, p< 0.01.