Inhibition of BCAT1-mediated cytosolic leucine metabolism regulates Th17 responses via the mTORC1-HIF1α pathway

Abstract

Branched-chain amino acids (BCAAs), particularly leucine, are indispensable AAs for immune regulation through metabolic rewiring. However, the molecular mechanism underlying this phenomenon remains unclear. Our investigation revealed that T-cell receptor (TCR)-activated human CD4⁺ T cells increase the expression of BCAT1, a cytosolic enzyme responsible for BCAA catabolism, and SLC7A5, a major BCAA transporter. This upregulation facilitates increased leucine influx and catabolism, which are particularly crucial for Th17 responses. Activated CD4⁺ T cells induce an alternative pathway of cytosolic leucine catabolism, generating a pivotal metabolite, β-hydroxy β-methylbutyric acid (HMB), by acting on BCAT1 and 4-hydroxyphenylpyruvate dioxygenase (HPD)/HPD-like protein (HPDL). Inhibition of BCAT1-mediated cytosolic leucine metabolism, either with BCAT1 inhibitor 2 (Bi2) or through BCAT1, HPD, or HPDL silencing using shRNA, attenuates IL-17 production, whereas HMB supplementation abrogates this effect. Mechanistically, HMB contributes to the regulation of the mTORC1-HIF1α pathway, a major signaling pathway for IL-17 production, by increasing the mRNA expression of HIF1α. This finding was corroborated by the observation that treatment with L-β-homoleucine (LβhL), a leucine analog and competitive inhibitor of BCAT1, decreased IL-17 production by TCR-activated CD4⁺ T cells. In an in vivo experimental autoimmune encephalomyelitis (EAE) model, blockade of BCAT1-mediated leucine catabolism, either through a BCAT1 inhibitor or LβhL treatment, mitigated EAE severity by decreasing HIF1α expression and IL-17 production in spinal cord mononuclear cells. Our findings elucidate the role of BCAT1-mediated cytoplasmic leucine catabolism in modulating IL-17 production via HMB-mediated regulation of mTORC1-HIF1α, providing insights into its relevance to inflammatory conditions.

Fig. 1. TCR stimulation induces the expression of BCAT1 and SLC7A5 in human CD4⁺ T cells.

a, b The mRNA and protein expression of BCAT1 and BCAT2 in human CD4⁺ T cells from healthy controls (HCs) was analyzed by RT‒qPCR (A; n = 5) and immunoblotting (B; n = 3) at 24 h with or without TCR stimulation. c Public RNA-seq data (GEO No: GSE140244) were analyzed to examine the expression of 42 amino acid transporters at the indicated time points in TCR-stimulated human CD4⁺ memory T cells. Heatmap analysis (left) and fold change in the expression (right) of amino acid transporters at the indicated times after TCR stimulation. d The expression of major BCAA transporters was validated by RT‒qPCR at 24 h poststimulation in human CD4⁺ T cells from HCs (n = 5). e The protein expression of SLC7A5 in human CD4⁺ T cells from HCs was analyzed at 24 h with or without TCR stimulation (n = 3). f Uptake of ³H-leucine by TCR-stimulated human CD4⁺ T cells in the presence of 50 mM BCH and 10 μM JPH203 (n = 5). The graphs show the means ± SEMs. *p < 0.05, **p < 0.01, and ***p < 0.001 according to the Mann‒Whitney U test (a, d, f), two-tailed unpaired t test (b, e), or one-way ANOVA with Tukey’s test (f).

Fig. 2. SLC7A5-mediated leucine influx regulates Th17 responses.

a mRNA expression of SLC7A5, BCAT1, and BCAT2 was analyzed by RT‒qPCR at 24 h poststimulation in human naive and memory CD4⁺ T cells (n = 5). b CFSE-labeled naive and memory CD4⁺ T cells were stimulated with anti-CD3/CD28-coated microbeads for 4 days with or without BCAAs, Bi2 (10 μM), or JPH203 (10 μM) (n = 5). The proportion of proliferating cells was measured by CFSE dilution assay. c, d The amount of cytokines in the culture supernatant of TCR-stimulated CD4⁺ naive T cells (c) and CD4⁺ memory T cells (d) stimulated with anti-CD3/CD28-coated microbeads for 3 days under the indicated conditions (n = 5). e The amount of IL-17A (left) and IFN-γ (right) in culture supernatant from TCR-stimulated CD4⁺ memory T cells supplemented with leucine and Bi2 (n = 7–8). f, g Freshly isolated human CD4⁺ memory T cells were activated and infected with GFP lentivirus containing BCAT1 shRNA for 24 h. shRNA⁺ cells expressing GFP were sorted and cultured for another 3 days with TCR stimulation. BCAT1 expression in sorted shRNA⁺ cells (n = 5) (f). The mRNA (left) and protein (right) levels of IL-17 and IFN-γ in the culture supernatant were analyzed by RT‒qPCR (n = 5) and ELISA (n = 5) (g). The graphs show the means ± SEMs. *p < 0.05 and **p < 0.01 according to the Mann‒Whitney U test.

Fig. 3. HMB, a BCAT1-mediated leucine metabolite, is involved in the regulation of Th17 responses.

a Public RNA-seq data (GSE140244) were analyzed to examine the expression of BCAA catabolic enzymes at the indicated times in TCR-stimulated human CD4⁺ memory T cells. Heatmap analysis illustrating the time course of changes in the expression of major BCAA catabolic enzymes. b The expression of the HPD, HPDL, and BCKDK in human CD4⁺ T cells was validated by RT‒qPCR at 24 h poststimulation (n = 4). c Scheme of cytosolic leucine metabolism. d CD4⁺ memory T cells were pretreated with Bi2 (10 μM), a statin (10 μM), or HMB (0.4 mM) for 1 h and stimulated with anti-CD3/CD28-coated microbeads for 3 days. The accumulation of cholesterol in cell lysates was measured (n = 3). e The amount of IL-17A in the culture supernatant from TCR-stimulated CD4⁺ memory T cells was measured on Day 3 by ELISA (n = 4). f TCR-stimulated CD4⁺ memory T cells were cultured for 3 days with Bi2 and the indicated concentration of acetate. The amount of IL-17A was measured by ELISA (n = 5). g The amount of IL-17A in the culture supernatant of CD4⁺ memory T cells supplemented with HMB was measured by ELISA (n = 5). h, i Freshly isolated human CD4⁺ memory T cells were activated and infected with GFP lentivirus containing HPD or HPDL shRNA for 24 h. shRNA⁺ cells expressing GFP were sorted and cultured for another 3 days with TCR stimulation. Expression of the indicated genes in sorted shRNA⁺ cells (n = 3) (h). The mRNA (left) and protein (right) levels of IL-17 and IFN-γ in the culture supernatant were analyzed by RT‒qPCR (n = 3) and ELISA (n = 3) (i). The graphs show the means ± SEMs. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 according to the Mann‒Whitney U test (b, e), one-way ANOVA with the Kruskal‒Wallis test (d, f, g), or two-tailed unpaired t test (h, i), respectively.

Fig. 4. scRNA-seq analysis reveals unique signaling pathways involved in the BCAT1-mediated regulation of IL-17A production.

Multiplex scRNA-seq analysis of human CD4⁺ memory T cells from three different groups was performed using Seurat in R software (version 4.3.0): no TCR stimulation (TCR-), TCR stimulation for 72 h without Bi2 (TCR+Bi2-), and TCR stimulation for 72 h with Bi2 (10 μM) (TCR+Bi2+), a Individual cells (28,651 cells) were color-coded based on the cluster (n = 31) in a t-distributed stochastic neighbor-joining (t-SNE) plot generated by unsupervised Seurat clustering. b Major CD4⁺ memory T-cell subsets were identified by canonical cell type marker expression. c t-SNE plots segregated on the basis of major CD4⁺ memory T-cell subsets (top, TCR-, n = 10,109 cells; TCR+Bi2-, n = 9506 cells; TCR+Bi2 +, n = 9036 cells). Dotted regions highlight Th17 cluster changes after TCR stimulation with Bi2 treatment. Pie charts showing relative Th subset abundances under different conditions (bottom). Activated T cells were annotated by the average expression of activation markers listed in Fig. S4A. d Pathway enrichment analysis of differentially expressed genes (DEGs) in activated Th17 cells between the TCR+Bi2- and TCR+Bi2+ groups. Representative genes in each pathway are shown. e Volcano plot of DEGs in activated Th17 cells between the TCR⁺Bi2^- and TCR⁺Bi2⁺ groups. Volcano plots were generated using the EnhancedVolcano package (version 1.16.0). f Expression of HIF1A in activated Th17 cells was projected onto the t-SNE plot (top) shown as feature plots (top). Violin plots showing the distribution of HIF1A expression levels, with dots representing individual cells (bottom). g Gene set enrichment analysis (GSEA) revealed 15 pathways enriched in 23,457 DEGs in the activated Th17 cluster with an FDR < 0.25. Red bars represent gene sets with a nominal p value < 0.1. h GSEA was used to examine the significantly enriched pathways. GSEA enrichment plot (left) and heatmap of downregulated DEGs (right) in the PI3K-AKT-mTOR signaling pathway in Bi2-treated cells. All transcripts within annotated genes were uploaded to a locally installed GSEA tool and compared with the hallmark gene sets.

Fig. 5. HMB regulates HIF-1α expression in human CD4⁺ T cells.

a The protein expression of HIF-1α in TCR-stimulated CD4⁺ memory T cells treated with or without Bi2 was analyzed (n = 5). As a control for HIF-1α, PANC-1 cells were incubated for 8 h under 2% or 21% O₂. A representative immunoblot is shown (left). The band intensity in the immunoblots was quantified by densitometry, except for the PANC-1 data. β-Actin was used as a normalization control (right). b Freshly isolated human CD4⁺ memory T cells were activated and infected with GFP lentivirus containing BCAT1, HPD, or HPDL shRNA for 24 h. shRNA⁺ cells expressing GFP were sorted and cultured for another 3 days with TCR stimulation. The mRNA expression of HIF-1α was analyzed by RT‒qPCR (n = 3–5). c HIF-1α expression in human CD4⁺ memory T cells was measured 24 h after TCR stimulation with Bi2 (n = 5). The cells were supplemented with HMB (0.4 mM) for the indicated time before harvest. d Quantitative PCR analysis of HIF-1α target gene expression after 72 h of TCR stimulation in human CD4⁺ memory T cells from HCs (n = 6). HMB (0.4 mM) was used to pretreat the cells for 1 h before TCR stimulation. e IL-17A and IFN-γ production were quantified at 72 h after TCR stimulation with or without Bi2 (10 μM), VH298 (100 nM), or HMB (0.4 mM) (n = 5). f mRNA expression of HIF-1α was measured at 4 h after TCR stimulation. The cells were supplemented with HMB (0.4 mM) for the indicated time before harvest (n = 5). g CD4⁺ memory T cells were stimulated with anti-CD3/CD28-coated microbeads for 1 h in the absence or presence of Bi2. Cell lysates were prepared at the indicated times and immunoblotted with phospho-p70-S6K and total p70-S6K (n = 3 independent experiments). h CD4⁺ memory T cells were stimulated with anti-CD3/CD28 Abs for 1 h. The cells were supplemented with HMB (0.4 mM) for the indicated time before harvest. Cell lysates were prepared at 1 h poststimulation and immunoblotted for phospho-p70-S6K and total p70-S6K (n = 5 independent experiments). The graph shows the band intensity quantified by densitometry. The graphs show the means ± SEMs. *p < 0.05 and **p < 0.01 according to the Mann‒Whitney U test [a, b (left), c–f, h] or two-tailed unpaired t test [b (right)].

Fig. 6. Inhibition of BCAT1 ameliorates EAE severity.

Experimental autoimmune encephalomyelitis (EAE) was induced by MOG_35–55 in a CFA emulsion with PTX. Bi2 (10 mg/kg) was intraperitoneally administered to MOG-immunized mice 4 h before immunization, and the treatment was repeated three times per week for 14 days. a Clinical scores of EAE mice. b Histological analysis of spinal cord tissue stained with Luxol fast blue or Hematoxylin & Eosin on Day 16. The areas of demyelination (left) and inflammatory cell infiltration (right) are marked with dashed black lines. c Quantification of immune cell infiltration in the spinal cord (SCMCs: spinal cord mononuclear cells) among the four different groups. d The proportion of CD3⁺CD4 ⁺ T cells among total CD45⁺ T cells in untreated or Bi2-treated EAE mice (n = 6 per group). e, f Flow cytometric analysis of IL-17A- and IFN-γ-producing CD3⁺CD4⁺ T cells in the SCMCs and inguinal lymph nodes (iLNs) of EAE mice (n = 5–8). g Cell lysates were prepared from SCMCs from DMSO- or Bi2-treated EAE mice and immunoblotted for HIF-α (n = 5 per group). h A representative histogram of intracellular HIF-1α in total CD45⁺ SCMCs (left) and CD3⁺CD4⁺ T cells (right) from EAE mice (n = 5 per group). The graphs show the means ± SEMs. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 by two-way ANOVA (a) or the Mann‒Whitney U test (c–h).

Fig. 7. LβhL, a leucine analog, attenuates IL-17A production.

a Uptake of ³H-leucine by TCR-stimulated CD4⁺ T cells in the presence of L-leucine, D-leucine, or LβhL (400 mg/L for each) (n = 3). b, c CFSE-labeled CD4⁺ naive (b) and memory (c) T cells were stimulated with anti-CD3/CD28-coated microbeads with or without L-leucine, LβhL (400 mg/L), or Bi2 (10 μM) (n = 5). The proportion of proliferating cells was measured by CFSE dilution assay. d The amount of cytokines in culture supernatant from TCR-stimulated CD4⁺ naive T cells (left) and CD4⁺ memory T cells (right) stimulated with anti-CD3/CD28-coated microbeads for 3 days under the indicated conditions (n = 5). Cells were treated with Bi2 and LβhL for 1 h before TCR stimulation. e Clinical scores of EAE mice. Bi2 (10 mg/kg) or LβhL (1 g/kg) was intraperitoneally administered to MOG-immunized mice 4 h before immunization, and the treatment was repeated three times per week for 14 days. f Flow cytometry analysis of IL-17A- and IFN-γ-producing CD3⁺CD4⁺ T cells in the iLN of EAE mice (n = 5 per group). g Quantification of SCMCs from EAE mice in the three groups (n = 6–7 per group). h Flow cytometry analysis of IL-17A- and IFN-γ-producing CD3⁺CD4⁺ T cells in the SCMCs of EAE mice (n = 5 per group). i A representative histogram plot of intracellular HIF-1α in CD3⁺CD4⁺ T cells among CD45⁺ SCMCs from EAE mice (n = 5 per group). The graphs show the means ± SEMs. **p < 0.01, ***p < 0.001, and ****p < 0.0001 by one-way ANOVA with Tukey’s test (a), the Mann‒Whitney U test (b, c–f, g–i), or two-way ANOVA (e).