BCAT1 affects mitochondrial metabolism independently of leucine transamination in activated human macrophages

Abstract

In response to environmental stimuli, macrophages change their nutrient consumption and undergo an early metabolic adaptation that progressively shapes their polarization state. During the transient, early phase of pro-inflammatory macrophage activation, an increase in tricarboxylic acid (TCA) cycle activity has been reported, but the relative contribution of branched-chain amino acid (BCAA) leucine remains to be determined. Here, we show that glucose but not glutamine is a major contributor of the increase in TCA cycle metabolites during early macrophage activation in humans. We then show that, although uptake of BCAAs is not altered, their transamination by BCAT1 is increased following 8 h lipopolysaccharide (LPS) stimulation. Of note, leucine is not metabolized to integrate into the TCA cycle in basal or stimulated human macrophages. Surprisingly, the pharmacological inhibition of BCAT1 reduced glucose-derived itaconate, α-ketoglutarate and 2-hydroxyglutarate levels without affecting succinate and citrate levels, indicating a partial inhibition of the TCA cycle. This indirect effect is associated with NRF2 (also known as NFE2L2) activation and anti-oxidant responses. These results suggest a moonlighting role of BCAT1 through redox-mediated control of mitochondrial function during early macrophage activation.

Keywords: BCAT1; Immunometabolism; Macrophages; Mitochondria; Redox biology; TCA cycle.

Fig. 1.

Glucose but not glutamine is a major carbon source for TCA cycle metabolites during early human macrophage activation. (A) Diagram of uniformly labelled [U-¹³C]-glucose catabolism, highlighting the TCA cycle metabolites and their expected glucose-derived ¹³C atoms (filled circles). (B,C) Glucose-derived and M+0 metabolites measured by LC-MS in control and LPS (8 h; 100 ng/ml)-stimulated hMDMs; n=6 donors. See Fig. S1 for individual data points. Glu, glutamate; Succ, succinate. (D) Diagram of uniformly labelled [U-¹³C]-glutamine catabolism, highlighting the TCA cycle metabolites and their expected glutamine-derived ¹³C atoms (filled circles). (E,F) Glutamine-derived and M+0 metabolites measured by LC-MS in control and LPS (8 h; 100 ng/ml)-stimulated hMDMs; n=6 donors. See Fig. S1 for individual data points.

Fig. 2.

Short-term LPS exposure increases leucine transamination without inducing BCAA uptake and catabolism. (A) Schematic of glucose-derived pyruvate and lactate adducts in macrophages treated with uniformly labelled [U-¹³C]-glucose. Filled circles indicate glucose-derived ¹³C atoms. (B) Uptake and secretion profiles by LC-MS shown for M+6 glucose, M+3 pyruvate and M+3 lactate in basal (control) and LPS (8 h; 100 ng/ml)-stimulated hMDMs. (C) Tryptophan, arginine and BCAA uptake in basal (control) and LPS (8 h; 100 ng/ml)-stimulated hMDMs. (D) Leucine transamination by BCAT1 (left) and [¹⁵N]-glutamate amount measured by LC-MS in hMDMs incubated with [¹⁵N]-leucine in control and LPS (8 h; 100 ng/ml)-stimulated hMDMs. Mean±s.e.m and individual data points are shown for at least n=5 donors. *P<0.05; ***P<0.001; ns, not significant (paired t-test).

Fig. 3.

Leucine is not metabolized to integrate into the TCA cycle in basal or stimulated human macrophages. (A) LC-MS for unlabelled and [U-13C]-leucine-derived (Leu) metabolites in control (Ctrl) and LPS (100 ng/ml; 8 h)-treated hMDMs. (B) Diagram of uniformly labelled [U-¹³C]-leucine catabolism, highlighting the possibly entry to the TCA cycle through acetyl-coA (AcoA). The TCA cycle is shown in grey to indicate there were no leucine-derived ¹³C atoms (filled circles) found in any of the metabolites. Mean±s.e.m and individual data points are shown for at least n=5 donors. **P<0.01; ***P<0.001; ****P<0.0001; ns, not significant (one-way ANOVA followed by Tukey's test).

Fig. 4.

BCAT1 inhibition results in partial inhibition of the TCA cycle between citrate and succinate. (A) Glucose-derived isotopologues (citrate M+2, itaconate M+1, αKG M+2, glutamate M+2, 2-HG M+2, succinate M+2, malate M+2) levels were measured in control, ERG240-treated, LPS-stimulated (8 h; 100 ng/ml) and LPS-stimulated and ERG240-treated (8 h; LPS+ERG240) hMDMs. (B) Diagram of the TCA cycle metabolites and their derivative metabolites, highlighting the reactions affected by BCAT1 inhibition (grey box). For each metabolite, the statistical significance of the effect of BCAT1 inhibition is shown. Mean±s.e.m and individual data points are shown for n=5 donors. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001; ns, not significant (one sample t-test).

Fig. 5.

BCAT1 inhibition activates NRF2 and is anti-oxidant in human macrophages. (A) Volcano plot for RNA-seq analysis of hMDMs treated with LPS (8 h; 100 ng/ml) or LPS and ERG240 (8 h; LPS+ERG240); n=3 donors. The dotted lines show the thresholds used to define differentially expressed genes (fold change<1.5, fold change>1.5; P_adj<0.01). (B) Gene Set Enrichment Analysis (GSEA) for RNA-seq analysis (left panel; LPS versus LPS+ERG240) and NRF2 western blotting (right panel) in hMDMs. Actin is shown as a loading control. (C) Schematic representation of the pathways affected in the RNA-seq analysis. The genes in red are upregulated in LPS+ERG240 when compared to LPS in hMDMs. (D) LC-MS quantification of serine, glycine, methionine and GSH/glutathione disulfide (GSSG) in basal (control), LPS-treated (8 h; 100 ng/ml) and LPS+ERG240-treated hMDMs. Mean±s.e.m and individual data points are shown for n=12 donors. (E) Flow cytometry analysis of live hMDMs (n=2 donors). Mean fluorescence intensity (MFI) was quantified as a measure of cellular ROS production (Cell ROX) in control, PMA only (PMA) and ERG240 pre-treated (ERG240+PMA) cells. ERG240 pre-treatment was for 16 h; PMA stimulation was for 30 min. (F) Ferritin heavy chain (Ferritin H) western blotting in control (Ctrl), ERG240, LPS (16 h; 100 ng/ml) and LPS+ERG240 hMDMs. Actin is shown as a loading control. *P<0.05; **P<0.01; ***P<0.001; ns, not significant (one sample t-test).