Ex vivo activation of the GCN2 pathway metabolically reprograms T cells, leading to enhanced adoptive cell therapy

Abstract

The manipulation of T cell metabolism to enhance anti-tumor activity is an area of active investigation. Here, we report that activating the amino acid starvation response in effector CD8⁺ T cells ex vivo using the general control non-depressible 2 (GCN2) agonist halofuginone (halo) enhances oxidative metabolism and effector function. Mechanistically, we identified autophagy coupled with the CD98-mTOR axis as key downstream mediators of the phenotype induced by halo treatment. The adoptive transfer of halo-treated CD8⁺ T cells into tumor-bearing mice led to robust tumor control and curative responses. Halo-treated T cells synergized in vivo with a 4-1BB agonistic antibody to control tumor growth in a mouse model resistant to immunotherapy. Importantly, treatment of human CD8⁺ T cells with halo resulted in similar metabolic and functional reprogramming. These findings demonstrate that activating the amino acid starvation response with the GCN2 agonist halo can enhance T cell metabolism and anti-tumor activity.

Keywords: 4-1BB; CD8(+) T cell; GCN2; Halofuginone; adoptive Cell therapy; autophagy; immunometabolism; immunotherapy.

Graphical abstract

Figure 1

GCN2 activation enhances T cell effector function and oxidative metabolism (A) Diagram illustrating experimental protocol. P14 CD8⁺ T cells were activated with peptide-pulsed bone marrow dendritic cells in complete media. After 3 days of activation, cells were expanded with IL-2 for another 4 days in arginine-replete or -deficient media. (B) Gene expression of downstream targets of GCN2 as determined by real-time PCR. Values are expressed as fold change relative to T cells expanded in arginine (n = 4 mice). (C) CD98 expression as determined by flow cytometry as well as CD98 mean fluorescence intensity (MFI) from n = 4 mice. (D and E) Cytokine production in CD8⁺ T cells expanded in the presence or absence of arginine (n = 4 mice). (F) Seahorse analysis quantifying the oxygen consumption rate (OCR) of T cells expanded in the presence or absence of arginine (n = 3 mice). (G) Basal OCR, extracellular acidification rate (ECAR), and OCR:ECAR ratio as determined by Seahorse (n = 3 mice). (H) ATP levels in T cells (n = 3 mice). (I) Diagram illustrating experimental protocol for halofuginone (halo) treatment during the last 48 h of T cell activation. (J and K) Cytokine production in halo-treated CD8⁺ T cells (n = 3 mice). (L) Granzyme B expression in halo-treated cells (n = 3 mice). (M) CD98 expression in halo-treated cells (n = 3 mice). (N) Seahorse analysis quantifying the OCR of halo-treated CD8⁺ T cells (n = 5 replicates ± SEM). (O) OCR:ECAR ratio as determined by Seahorse (n = 5 mice). (P) ATP levels in halo-treated T cells (n = 3 mice). Each circle represents a different mouse. Histograms are normalized to mode. Gray histograms represent fluorescent minus one (FMO) controls. Results shown are representative of at least 2–3 independent experiments or pooled from multiple independent experiments. ∗p < 0.05 and ∗∗p < 0.01 as determined by two-tailed t test. n.s., not significant. All error bars are SEM.

Figure 2

Halo promotes the transcriptional regulation of 4-1BB expression and mitochondrial metabolism (A) Flow cytometry of halo- or vehicle-treated cells evaluating the expression of surface markers associated with Tcm lineage. (B–G, I, and J) Total and ribosomal enriched RNA was extracted and sequenced from halo- or vehicle-treated CD8⁺ T cells from 3 mice. (B) Principal-component analysis plot of halo- or vehicle-treated CD8⁺ T cells. (C) Volcano plot of highly up- or down-regulated genes in halo-treated cells as found in total RNA as well as those regulated translationally as determined by translational efficiency (TE). (D and E) Pathways significantly enriched in halo-treated cells in (D) total RNA and (E) TE. (F and G) Log2 fold change levels versus vehicle of total and ribosomal RNA fractions and TE of genes associated with (F) the Trm cell lineage and (G) the T cell co-stimulatory and co-inhibitory molecules. (H) Expression level of 4-1BB in halo- or vehicle-treated CD8⁺ T cells as determined by flow cytometry (n = 3 mice). Histograms are normalized to mode. (I) FPKM levels of Bhlhe40 in halo- and vehicle-treated cells as determined by RNA sequencing (n = 3 mice). (J) Log2 fold change levels versus vehicle of total and ribosomal RNA fractions as well as TE of genes associated with the electron transport chain complexes. ∗∗p < 0.01 as determined by (H) two-tailed t test or (I) ANOVA with Tukey test. n.s., not significant. All bar graphs are mean ± SEM.

Figure 3

Autophagy and the CD98-mTOR axis mediate enhanced OXPHOS and IFN-γ production (A) Amino acid levels as determined by mass spectrometry. Values represent average Z score from n = 4 mice. (B) p-mTOR staining in halo-treated CD8⁺ T cells. Histogram is normalized to mode. (C) p-mTOR MFI after treatment with BCH (n = 3 mice). (D) MFI of 4-1BB, granzyme B, and IFN-γ after treatment with BCH (n = 3 mice). (E) Basal OCR after treatment with BCH (n = 3 mice). (F) Autophagy levels in halo-treated cells (n = 3 mice). (G) Seahorse analysis quantifying OCRs of T cells treated as indicated (n = 3 mice). (H) Basal OCR after treatment with 3MA (n = 3 mice). (I and J) IFN-γ scatterplot (I) and MFI (J) (n = 3 mice) after treatment with 3MA. (K and L) IFN-γ scatterplot (K) and MFI (L) (n = 3 mice) after treatment with oligomycin. (M) Diagram illustrating the proposed mechanism. Results shown are representative of at least 2–3 independent experiments. Bars indicate mean, and each circle represents an individual mouse. ∗p < 0.05 and ∗∗p < 0.01 as determined by ANOVA with Tukey test or (F) two-tailed t test. n.s., not significant. All bar graphs are mean ± SEM.

Figure 4

Halo-treated CD8⁺ T cells demonstrate robust anti-tumor activity (A) Mice bearing day 10 established EG7-OVA tumors were administered 1 × 10⁶ halo- or vehicle-treated OT-1 cells. Each line is a different mouse (n = 6 mice). (B) Survival curve from (A) representing combined survival across all experiments (n = 11–12 mice). (C) Mice bearing day 11 established B16-gp33tumors received an adoptive transfer of 0.5 × 10⁶ halo- or vehicle-treated P14 cells in conjunction with an in vivo administration of 4-1BB agonistic antibody. Each line is a different mouse (n = 5 mice). (D) Day 25 tumor size from (C) pooled from multiple independent experiments (n = 10 mice). (E) OCR Seahorse curve of halo-treated human CD8⁺ T cells (n = 4 replicates). (F–I) Human naive CD8⁺ T cells were activated in the presence of halo and transduced with DMF5 TCR. (F) Fluorescence-activated cell sorting (FACS) plot from a representative donor showing high TCR expression in both treatment groups. (G–I) Representative histogram and pooled MFI of (G) CD98, (H) 4-1BB, and (I) granzyme B gated on TCR⁺ CD8⁺ T cells as shown in (F) (n = 6 donors). Histograms are normalized to mode. Results shown are representative of at least 2 independent experiments (A, C, E, and F) or pooled from multiple independent experiments (B, D, and G–I). ∗p < 0.05 and ∗∗p < 0.01 as determined by (B) log-rank test, (D) ANOVA with Tukey test, or (G–I) Mann-Whitney U test. All bar graphs are mean ± SEM.