Elongation factor 1A1 regulates metabolic substrate preference in mammalian cells

Abstract

Eukaryotic elongation factor 1A1 (EEF1A1) is canonically involved in protein synthesis but also has noncanonical functions in diverse cellular processes. Previously, we identified EEF1A1 as a mediator of lipotoxicity and demonstrated that chemical inhibition of EEF1A1 activity reduced mouse liver lipid accumulation. These findings suggested a link between EEF1A1 and metabolism. Therefore, we investigated its role in regulating metabolic substrate preference. EEF1A1-deficient Chinese hamster ovary (2E2) cells displayed reduced media lactate accumulation. These effects were also observed with EEF1A1 knockdown in human hepatocyte-like HepG2 cells and in WT Chinese hamster ovary and HepG2 cells treated with selective EEF1A inhibitors, didemnin B, or plitidepsin. Extracellular flux analyses revealed decreased glycolytic ATP production and increased mitochondrial-to-glycolytic ATP production ratio in 2E2 cells, suggesting a more oxidative metabolic phenotype. Correspondingly, fatty acid oxidation was increased in 2E2 cells. Both 2E2 cells and HepG2 cells treated with didemnin B exhibited increased neutral lipid content, which may be required to support elevated oxidative metabolism. RNA-seq revealed a >90-fold downregulation of a rate-limiting glycolytic enzyme, hexokinase 2, which we confirmed through immunoblotting and enzyme activity assays. Pathway enrichment analysis identified downregulations in TNFA signaling via NFKB and MYC targets. Correspondingly, nuclear abundances of RELB and MYC were reduced in 2E2 cells. Thus, EEF1A1 deficiency may perturb glycolysis by limiting NFKB- and MYC-mediated gene expression, leading to decreased hexokinase expression and activity. This is the first evidence of a role for a translation elongation factor, EEF1A1, in regulating metabolic substrate utilization in mammalian cells.

Keywords: glucose; glycolysis; lipid; lipid metabolism; translation elongation factor.

Figure 1

EEF1A1 deficiency and inhibition reduce media acidification and media lactate.A, light micrographs of WT and EEF1A1-deficient (2E2) CHO-K1 cells at low density. The scale bar represents 50 μm. B, cumulative population doublings for WT and 2E2 cells at the indicated time points, n = 3. C, conditioned media from WT and 2E2 cells over 72 h. D, change in 415/560 (Δ415/560) from 0 h to 72 h for WT and 2E2 cell conditioned media, normalized to total cell number, n = 3. ∗∗∗∗p < 0.0001 for WT 24 h versus WT 0 h; ∗∗∗p < 0.001 for WT 48 h versus WT 24 h; p = 0.06 for WT 72 h versus WT 48 h; ++++ p < 0.0001 for 2E2 versus WT at the corresponding time point. E, lactate concentration in conditioned media from WT and 2E2 cells after 24 h, n = 3. F, Δ415/560 and (G) lactate concentration in conditioned media collected from HepG2 cells expressing scrambled (Scr) and EEF1A1 (EF) shRNA after 48 h, n = 4. H, Δ415/560 and (I) lactate concentration in conditioned media collected from CHO-K1 WT cells treated with DMSO or didemnin B (DB) (20 nM) for 24 h, n = 3. J and L, Δ415/560 and (K and M) lactate concentration in conditioned media collected from HepG2 WT cells treated with DMSO and (J and K) DB (80 nM) or (L and M) plitidepsin (PL) (80 nM) for 24 h, n = 3. Data are means ± SD. ∗p < 0.05, ∗∗p < 0.01. and ∗∗∗p < 0.001. CHO, Chinese hamster ovary; EEF1A, eukaryotic elongation factor 1A.

Figure 2

EEF1A1-deficient CHO-K1 cells exhibit reduced glycolysis and a more oxidative metabolic phenotype.A, proton efflux rate (PER) and (C) oxygen consumption rate (OCR) for WT and EEF1A1-deficient (2E2) CHO-K1 cells under basal conditions and upon treatment with oligomycin (Olig) or rotenone/antimycin A (Rot/AA), n = 3. B, glycolytic, (D) mitochondrial, and (E) total ATP production rates determined from data in A and C. F, total cellular ATP in WT and 2E2 cell lysates, n = 8. G, ATP rate index calculated by dividing the mitochondrial ATP production rate in D by the glycolytic ATP production rate in B. H, fatty acid oxidation in WT and 2E2 cells, n = 4. I, WT and 2E2 cell relative population doublings after treatment with oligomycin (Olig) or 2-deoxyglucose (2-DG) for 16 h, n = 3. J, areas under scratch closure curves from scratch assays performed over 28 h with WT and 2E2 cells treated as in I, n = 3. K, transwell migration of WT and 2E2 cells that were treated as in I for 24 h, n = 3. Data are means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001. CHO, Chinese hamster ovary; EEF1A, eukaryotic elongation factor 1A.

Figure 3

EEF1A1-deficient CHO-K1 cells exhibit increased neutral lipid accumulation.A, triglyceride (TG) and (B) cholesteryl ester (CE) masses measured in WT and EEF1A1-deficient (2E2) CHO-K1 cells after treatment with media containing BSA alone or 0.5 mM palmitate (PA) or 0.5 mM palmitate plus oleate (2:3 ratio, PA/OA) for 6 h, n = 4. C, total neutral lipid mass calculated as the sum of TG and CE, n = 4. D, confocal micrographs of WT and 2E2 cells stained with Oil Red O (neutral lipid, red) and DAPI (nuclei, blue) after 6 h treatment with 0.5 mM PA/OA. The scale bar represents 10 μm. E, lipid droplet (LD) size distributions generated using particle analysis in ImageJ, n = 5. F, images from D colourized based on the LD size bins and corresponding colors from the x-axis in E. The scale bar represents 10 μm. G, median LD size, (H) LD number, and (I) total LD area determined from data in E. Data are means ± SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001.

Figure 4

EEF1A1 inhibition increases neutral lipid accumulation in HepG2 cells.A, triglyceride (TG) and (B) cholesteryl ester (CE) masses measured in HepG2 cells pretreated with DMSO or didemnin B (DB) (80 nM) in basal media for 24 h, followed by treatment with media containing BSA alone or 1 mM palmitate plus oleate (2:3 ratio, PA/OA) with DMSO or DB for 6 h, n = 5. C, total neutral lipid mass, calculated as the sum of TG and CE, n = 5. D, confocal micrographs of HepG2 cells stained with Oil Red O (neutral lipid, red) and DAPI (nuclei, blue) after pretreatment with DMSO or DB (80 nM) in basal media for 24 h, followed by treatment with media containing PA/OA with DMSO or DB for 6 h. The scale bar represents 10 μm. E, lipid droplet (LD) size distributions, generated using particle analysis in ImageJ, n = 4. F, images from D colorized based on the LD size bins and corresponding colors from the x-axis in E. The scale bar represents 10 μm. G, median LD size, (H) LD number, and (I) total LD area determined from data in E. Data are means ± SD. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001. CHO, Chinese hamster ovary; EEF1A, eukaryotic elongation factor 1A.

Figure 5

RNA-seq and gene set enrichment analysis reveal transcriptomic alterations in NFKB and MYC signaling, among others, in EEF1A1-deficient CHO-K1 cells.A, principal component (PC) analysis scores plot of RNA-seq data in WT (blue) and EEF1A1-deficient (2E2) (yellow) CHO-K1 cells, n = 4. B, volcano plot with data points representing gene transcripts altered in 2E2 relative to WT cells. Genes altered at a q value <0.05 and a fold change >1.5 are indicated in blue. Labels correspond to the top five most significantly altered genes and the genes with the top five largest absolute fold changes. C, bubble plot of gene set enrichment analysis representing pathways altered in 2E2 cells relative to WT cells. Normalized enrichment scores (NES) are plotted on the x-axis, and gene sets are plotted on the y-axis. Bubble size corresponds to −log₁₀(q value) (increases as q value decreases). CHO, Chinese hamster ovary; EEF1A, eukaryotic elongation factor 1A; HK2, hexokinase 2.

Figure 6

EEF1A1-deficient CHO-K1 cells exhibit reduced nuclear abundance of RELB and MYC, reduced whole-cell HK2 abundance, and reduced hexokinase activity.A, representative immunoblots of nuclear and cytoplasmic fractions (probed for RELA, RELB, and MYC) prepared from WTand EEF1A1-deficient (2E2) CHO-K1 cells. B, densitometric analysis of RELA, RELB, and MYC in WT and 2E2 nuclear fractions, n = 4. C–E, confocal micrographs of WT and 2E2 cells stained for (C) RELA, (D) RELB, or (E) MYC (red), TUBA (cyan), and DAPI (gray). The scale bar represents 10 μm. F, representative immunoblots of whole-cell lysates probed for HK2 prepared from WT and 2E2 cells. G, densitometric analysis of HK2 in WT and 2E2 whole-cell lysates, n = 4. H, hexokinase activity determined by an enzymatic, colorimetric assay in WT and 2E2 cells, n = 3. Data are means ± SD. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗∗p < 0.0001. Molecular weight markers in kilodaltons indicated on the left side of each membrane image. CHO, Chinese hamster ovary; EEF1A, eukaryotic elongation factor 1A; HK2, hexokinase 2.

Figure 7

Working model for regulation of metabolic substrate preference by EEF1A1.A, EEF1A1 regulates metabolic substrate preference. When EEF1A1 is intact, both glycolysis and oxidative metabolism are used for energy generation. When EEF1A1 is perturbed using genetic or chemical strategies, glycolysis is impaired and oxidative metabolism (OXPHOS) and fatty acid β oxidation (FAO) are preferred for energy generation. B, inset of A (dashed outline) summarizing potential mechanisms through which EEF1A1 regulates metabolic substrate preference. EEF1A1 deficiency impairs glycolysis, reduces hexokinase 2 (HK2) expression and hexokinase activity, and reduces NFKB and MYC signaling. EEF1A1 may promote glycolysis by promoting NFKB and MYC signaling and hexokinase expression and activity. EEF1A1 likely also regulates metabolic substrate preference through other mechanisms. Created with Biorender.com. CHO, Chinese hamster ovary; EEF1A, eukaryotic elongation factor 1A.