An Essential Role of the Mitochondrial Electron Transport Chain in Cell Proliferation Is to Enable Aspartate Synthesis

Abstract

The mitochondrial electron transport chain (ETC) enables many metabolic processes, but why its inhibition suppresses cell proliferation is unclear. It is also not well understood why pyruvate supplementation allows cells lacking ETC function to proliferate. We used a CRISPR-based genetic screen to identify genes whose loss sensitizes human cells to phenformin, a complex I inhibitor. The screen yielded GOT1, the cytosolic aspartate aminotransferase, loss of which kills cells upon ETC inhibition. GOT1 normally consumes aspartate to transfer electrons into mitochondria, but, upon ETC inhibition, it reverses to generate aspartate in the cytosol, which partially compensates for the loss of mitochondrial aspartate synthesis. Pyruvate stimulates aspartate synthesis in a GOT1-dependent fashion, which is required for pyruvate to rescue proliferation of cells with ETC dysfunction. Aspartate supplementation or overexpression of an aspartate transporter allows cells without ETC activity to proliferate. Thus, enabling aspartate synthesis is an essential role of the ETC in cell proliferation.

Figure 1. A CRISPR-based genetic screen identifies metabolic genes whose loss sensitizes human cells to phenformin

(A) Dose-dependent effects of phenformin on Jurkat cell proliferation (mean ± SD, n=3). (B) Schematic depicting the pooled CRISPR-based screen. (C) Gene scores in untreated versus phenformin-treated (0.5 uM) Jurkat cells. The gene score is the median log2 fold change in the abundance of all sgRNAs targeting that gene during the culture period. Most genes, as well as non-targeting control sgRNAs, have similar scores in the presence or absence of phenformin. (D) Top 25 genes scoring as differentially required upon phenformin treatment (top). Genes linked to the GOT1-catalyzed transamination reaction are indicated in red, the ETC in blue, and to nucleotide biosynthesis in green. The top-scoring gene, GOT1, catalyzes the transamination of aspartate to alpha-ketoglutarate yielding L-glutamate and oxaloacetate (OAA) and requires PLP as a cofactor (bottom). (E) Changes in abundance in the primary screen of the individual GOT1 sgRNAs in the presence (gray) or absence (black) of phenformin. (F) GOT1-null cells die upon phenformin treatment. Immunoblot analysis for indicated proteins of wild type and GOT1-null Jurkat cells (top). Akt was used as a loading control. Fold change in cell number (log2) of wild type (black) and GOT1-null (blue) Jurkat cells after treatment with indicated phenformin concentrations for 5 days (mean ± SD, n=3) (bottom). Representative bright-field micrographs of indicated cells after a 5-day phenformin treatment (right). (G) Expression of an sgRNA-resistant GOT1 cDNA rescues phenformin sensitivity of the GOT1-null Jurkat cells. Immunoblot analysis of wild type, GOT1-null, and rescued null cells (top). Raptor was used as a loading control. Fold change in cell number (log2) of wild type (black), GOT1-null (blue), and rescued GOT1-null (gray) cells after a 5-day treatment with indicated phenformin concentrations (mean ± SD, for n=3) (bottom).

Figure 2. ETC inhibition kills cells lacking GOT1

(A) GOT1 loss sensitizes various human cell types to phenformin treatment. Immunoblot analysis of wild type and GOT1-null Raji and KMS-26 cells (left). Raptor was used as a loading control. Fold change in cell number (log2) of wild type (black) and GOT1-null (blue) KMS-26 and Raji cells after a 5-day treatment with indicated phenformin concentrations (mean ± SD, n=3) (right). (B) GOT1-null cells die upon ETC dysfunction induced with various ETC inhibitors. Graphical scheme depicting effects of phenformin (complex I inhibitor), piericidin (complex I inhibitor), and antimycin (complex III inhibitor) (left). Fold change in cell number (log2) of wild type (black), GOT1-null (blue), and rescued GOT1-null (gray) Jurkat cells after a 5-day treatment with indicated piericidin and antimycin concentrations (mean ± SD, n=3) (right).

Figure 3. Upon ETC inhibition, GOT1 reverses and generates aspartate, which is limiting for cell proliferation

(A) Schematic depicting the malate-aspartate shuttle. Normally, the malateaspartate shuttle runs in the forward direction to transfer reducing equivalents across the mitochondrial membrane. GOT1 is part of the malate-aspartate shuttle and consumes aspartate to generate oxaloacetate (OAA). Aspartate produced by mitochondria is a precursor for protein and nucleotide biosynthesis. (B) Upon ETC inhibition, GOT1 reverses and consumes aspartate. Relative abundance of indicated amino acids in wild type and GOT1-null Jurkat cells after a 24-hour treatment with (gray) or without (black) phenformin (mean ± SD, n=3, **p<0.05). All measurements are relative to untreated wild type Jurkat cells. (C) Tfam knock-out hearts have a lower ratio of aspartate to leucine than wild type hearts. Immunoblot analyses of TFAM and COXI in wild type and TFAM-null murine hearts (left). S6K1 was used as a loading control. Relative ratio of aspartate to leucine in wild type and TFAM-null mouse hearts (mean ± SD, n=7 (wild type) and n=6 (TFAM null), **p<0.05). (D) Aspartate supplementation rescues death of GOT1-null cells upon ETC inhibition. Fold change in cell number (log2) of wild type (black), GOT1-null (blue) and rescued GOT1-null (gray) Jurkat cells in the absence and presence (10 mM) of aspartate after treatment with the indicated phenformin concentrations for 5 days (mean ± SD, n=3, **p<0.05) (top). Representative bright-field micrographs of indicated cells after a 5-day phenformin treatment in the absence or presence of aspartate (bottom). (E) Expression of a glutamate-aspartate transporter (SLC1A3) rescues the phenformin-induced death of GOT1-null cells cultured in standard RPMI media, which contains only 150 uM aspartate. Fold change in cell number (log2) of GOT1-null (blue) and SLC1A3-overexpressing GOT1-null (gray) Jurkat cells in RPMI (150 uM aspartate) after a 5-day treatment with 10 uM phenformin (mean ± SD, n=3, **p<0.05).

Figure 4. Metabolic routes of aspartate synthesis in cells with ETC inhibition

(A) Schematic depicting oxidative and reductive glutamine metabolism pathways (top). Green and blue arrows indicate oxidative and reductive arms of the TCA cycle, respectively. Filled circles represent ¹³C atoms derived from [U-¹³C]-L-glutamine. (B) Upon ETC inhibition, aspartate is mainly synthesized by reductive metabolism of glutamine in a GOT1-dependent manner. Mass isotopomer analysis of aspartate in wild type and GOT1-null Jurkat cells cultured for 7 hours with [U-¹³C]-L-glutamine in the presence or absence of phenformin (10 uM). Aspartate pool sizes (middle) and fraction of labeled aspartate derived from labeled glutamine (bottom) for each sample are indicated in separate graphs (mean ± SD, for n=3, **p<0.05). OAA, oxaloacetate.

Figure 5. In cells with ETC inhibition pyruvate stimulates aspartate synthesis in a GOT1-dependent fashion

(A) Pyruvate cannot rescue death of GOT1-null cells induced by ETC inhibitors. Fold change in cell number (log2) of wild type (black) and GOT1-null (blue) Jurkat cells in the presence or absence of pyruvate (1 mM) after treatment with phenformin (10 uM), antimycin (1 uM) and piericidin (1 uM) for 5 days (mean ± SD, n=3, **p<0.05). (B) In cells with ETC inhibition pyruvate addition increases cellular aspartate levels in a GOT1-dependent manner. Relative aspartate levels were measured in wild type (black), GOT1-null (blue), and rescued GOT1-null (gray) Jurkat cells in the presence (1 mM) or absence of pyruvate after a 24-hour phenformin (10 uM) treatment using LC-MS/MS (mean ± SD, for n=3, **p<0.05). All measurements are relative to untreated wild type Jurkat cells. (C) Proposed mechanism of pyruvate-mediated rescue of cell proliferation upon ETC inhibition.

Figure 6. Cells with ETC inhibition require MDH1 for pyruvate to stimulate aspartate synthesis and enable proliferation

(A) Pyruvate does not rescue the proliferation of phenformin-treated MDH1 null cells. Immunoblot analysis of wild type and MDH1-null Jurkat cells along with counterparts expressing an sgRNA-resistant MDH1 cDNA (left). mTOR and Raptor were used as loading controls. Relative fold change in cell number of wild type (black), MDH1-null (light blue), and rescued MDH1-null (gray) Jurkat cells after a 5-day treatment with phenformin (10 uM), piericidin (1 uM), and antimycin (1 uM) in the presence or absence of pyruvate (1 mM) (mean ± SD, n=3) (right). (B) The pyruvate-induced increase in aspartate synthesis depends on MDH1. Relative aspartate levels were determined in wild type (black), GOT1-null (blue), and MDH1-null (light blue) Jurkat cells in the presence or absence of pyruvate (1 mM) after a 24-hour phenformin (10 uM) treatment (mean ± SD, for n=3). All measurements are relative to untreated wild type Jurkat cells. (C) Metabolic routes for pyruvate-induced aspartate synthesis under ETC inhibition. Pyruvate, through the lactate dehydrogenases, regenerates NAD+ in the cytoplasm. This NAD+ can activate the cytoplasmic malate dehydrogenase to provide OAA from malate and drive aspartate synthesis via GOT1. Alternatively, another source for OAA is through ATP-citrate lyase, which catalyzes the conversion of citrate and CoA into acetyl-CoA and OAA in the cytosol. Under ETC inhibition, the latter reaction is likely less dependent on NAD+ and can work even in the absence of pyruvate supplementation. Note that both pathways are dependent on GOT1. (D) Aspartate levels correlate with cellular health upon ETC inhibition. ETC inhibition leads to a decrease in aspartate levels and inhibits cell proliferation in wild type cells. The residual aspartate generated by GOT1 is sufficient to maintain viability, as GOT1 loss results in cell death and corresponds to an almost complete depletion of aspartate. Pyruvate addition rescues aspartate levels and proliferation under ETC inhibition in a GOT1- and MDH1-dependent fashion.

Figure 7. Aspartate supplementation enables the proliferation of patient-derived cybrids with mtDNA mutations and replaces the need for pyruvate

(A) Aspartate can replace pyruvate in enabling the proliferation of patient-derived mtDNA mutant cybrids and 143B ρ⁰ cells devoid of mtDNA. Cell line models of ETC dysfunction were cultured in RPMI (supplemented with 50 ug/ml uridine to bypass the need for complex III in uridine synthesis) with pyruvate (1 mM) or aspartate (10 mM) for 6 days. Relative cell number was determined by normalizing to the pyruvate-supplemented condition. Representative bright-field micrographs of MERRF, CYTB, and 143B ρ⁰ cells after 6 days in indicated conditions (right) (mean ± SD, n=3, **p<0.05). (B) Pyruvate stimulates aspartate biosynthesis in cybrid cells with ETC dysfunction. Mass isotopomer analysis of aspartate in wild type, phenformin-treated wild type, and ETC-defective cybrid cells cultured for 7 hours with [U-¹³C]-L-glutamine in the presence or absence of pyruvate (1 mM). The fraction of labeled aspartate from [U-¹³C]-L-glutamine is indicated (mean ± SD, for n=3, **p<0.05). OAA, oxaloacetate. (C) Pyruvate supplementation enables the proliferation of patient-derived cybrid cells in a GOT1-dependent fashion that can be bypassed by aspartate. Immunoblot analysis of wild type, MERRF, and CYTB cybrid cells expressing sgControl and sgGOT1 (top). Raptor was used as a loading control. Cell line models of ETC dysfunction expressing sgControl or sgGOT1 were cultured in RPMI (supplemented with uridine (50 ug/ml) and pyruvate (1 mM)) with or without aspartate (10 mM) for 6 days (bottom). Relative cell number was determined by normalizing to sgControl expressing cell line (mean ± SD, for n=3, **p<0.05). (D) SLC1A3 overexpression enables CYTB cells to proliferate in standard RPMI media without pyruvate addition. Fold changes in cell number over time of 143B wild type, CYTB cybrid cells, and their SLC1A3-expressing counterparts when cultured in RPMI media lacking aspartate and pyruvate (blue), or supplemented with aspartate (150 uM) (black) or pyruvate (1 mM) (gray) (mean ± SD, n=3, **p<0.05).