Acetyl-CoA carboxylase obstructs CD8+ T cell lipid utilization in the tumor microenvironment

Abstract

The solid tumor microenvironment (TME) imprints a compromised metabolic state in tumor-infiltrating T cells (TILs), hallmarked by the inability to maintain effective energy synthesis for antitumor function and survival. T cells in the TME must catabolize lipids via mitochondrial fatty acid oxidation (FAO) to supply energy in nutrient stress, and it is established that T cells enriched in FAO are adept at cancer control. However, endogenous TILs and unmodified cellular therapy products fail to sustain bioenergetics in tumors. We reveal that the solid TME imposes perpetual acetyl-coenzyme A (CoA) carboxylase (ACC) activity, invoking lipid biogenesis and storage in TILs that opposes FAO. Using metabolic, lipidomic, and confocal imaging strategies, we find that restricting ACC rewires T cell metabolism, enabling energy maintenance in TME stress. Limiting ACC activity potentiates a gene and phenotypic program indicative of T cell longevity, engendering T cells with increased survival and polyfunctionality, which sustains cancer control.

Keywords: T cell; endoplasmic reticulum; immunotherapy; lipid; metabolism; mitochondria; tumor microenvironment.

Figure 1.. CD8⁺ TILs exhibit steatosis

(A–C) GSEA of (A) RNA sequencing (n = 3/3 pooled samples) and (B and C) UPLC-MS lipidomics of CD8⁺ T cells isolated from spleens or tumors of MCA-205 fibrosarcoma-bearing mice (n = 5/5 pooled samples). (D–I) LDs assessed by confocal imaging (D, n = 70/49/70; F, 56/55/50; H, n = 50/50/50) and (E, G, and I) neutral lipid content assessed by spectral flow cytometry in CD8⁺ T cells from spleens, tumor-draining lymph nodes, and tumors of mice bearing MCA-205 fibrosarcoma (n = 7), MC-38 colon adenocarcinoma (n = 6), and B16 melanomas (n = 10). Scale bars, 5μm. Data are represented as mean ± SEM. Statistical analysis was performed using a Benjamini-Hochberg procedure (A), unpaired two-sided Student’s t test (B and C), or ordinary one-way ANOVA (D–I). **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2.. The dysfunctional CD8⁺ TIL pool exhibits elevated lipid storage

(A and B) (A) FACS quantification of neutral lipid expression in endogenous and transferred tumor-antigen-specific CD8⁺ TILs (n = 3/3), and (B) LDs assessed by confocal imaging in transferred tumor-antigen-specific CD8⁺ T cells isolated from spleens, TDLNs, or tumors of B16-OVA tumor-bearing hosts (n = 50/50/50). Scale bars, 5 μm. (C and D) (C) Dimension reduction and UMAP projection and (D) quantification of neutral lipid content assessed by high-dimensional spectral flow cytometry 1 and 2 weeks post MCA-205 fibrosarcoma implantation (n = 5/5). (E) Clustering analysis from high-dimensional spectral flow cytometry of phenotypic markers from MCA-205 sarcoma 2 weeks post-implantation (n = 8). (F–H) Neutral lipid accumulation by PD-1/Tim-3 expression (n = 5/6/6). Data are represented as mean ± SEM. Statistical analysis was performed using a paired two-sided Student’s t test (A), ordinary one-way ANOVA (B), unpaired two-sided Student’s t test (D), or RM one-way ANOVA (F–H). *p < 0.05, **p < 0.01, ***p < 0.001. See also Figure S1.

Figure 3.. ACC dictates lipid composition of T cells

(A and B) (A) Differential gene expression (n = 3/3) and (B) label-free quantitative LC-MS proteomics (n = 4/4) of CD8⁺ T cells isolated from spleens or tumors of MCA-205 fibrosarcoma-bearing mice. (C) Gene expression from n = 4 high-grade deep pleiomorphic undifferentiated sarcoma patient samples. (D–F) UPLC-MS lipidomics of OT-1 T cells cultured ± ACCi (n = 5/5). (G and H) LDs assessed by confocal imaging in vehicle versus ACCi (n = 50/50) or sgNT control versus ACC1^−/− OT-1 T cells (n = 95/93). Scale bars indicate 5 μm in high-magnification and 20 μm in low-magnification images. Data are represented as mean ± SEM. Statistical analysis was performed using unpaired two-sided Student’s t test. **p < 0.01, ***p < 0.001, ****p < 0.0001. See also Figure S2.

Figure 4.. ACC inhibition rewires CD8⁺ T cell metabolism

(A–E) High-performance liquid chromatography (HPLC)-MS/MS metabolomics (n = 5/5). (F–H) Energetic profiling, OCR trace with SRC quantifications of OT-1 T cells cultured ± ACCi (n = 12/12). (I and J) Control, ACCi, or ACC1^−/− OT-1 T cells pulsed with BODIPY-C16 then visualized for co-localization with mitochondria (MitoTracker) by confocal imaging and quantified using relative fluorescent intensity (RFI). See Figures S3C and S3D for quantification of multiple cells. Scale bars, 2 μm. Data are represented as mean ± SEM. Statistical analysis was performed using principal component analysis (A), hypergeometric test (C), unpaired two-sided Student’s t test (B, D–F, and H), or one-way ANOVA (G). **p < 0.01, ***p < 0.001, ****p < 0.0001. See also Figure S3.

Figure 5.. ACC limits energy synthesis in T cells in nutrient stress

(A) Western blotting, (B) Seahorse real-time ATP rate assay, (C) LDs assessed by confocal imaging (n = 70/49), and (D–I) energetic profiling of ATP rates in OT-1 T cells co-cultured ± B16 melanoma cells for 36 h ± ACCi. Scale bars, 5 μm. Data are represented as mean ± SEM. Statistical analysis was performed using unpaired two-sided Student’s t test (B and D–I). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. See also Figure S4.

Figure 6.. ACCi promotes T cell immunity against tumors

(A–D) (A) GSVA of RNA sequencing (n = 4/4), (B) spectral flow cytometry (n = 4/4), and (C and D) representative images of TEM with indicated quantifications of OT-1 T cells cultured ± ACCi (n = 80/80, 20/20, and 126/126). Scale bars indicate 1 μm and 2 μm in vehicle and ACCi low-magnification images, respectively, and 0.5 μm in high-magnification images. (E–H) (E) Tumor growth (n = 4/8/8), (F) engraftment (n = 10/9), (G) polyfunctionality (n = 5/5), and (H) TCF-1 expression (9/9) of OT-1 T cells cultured ± ACCi, then infused into B16-OVA melanoma-bearing mice. (I–K) (I) p-ACC expression in indicated CD8⁺ T cell subsets from n = 7 normal donors, and (J) CD8⁺ and (K) CD4⁺ T cell subsets in response to vehicle or ACCi in n = 7 and n = 5 normal donors, respectively. Data are represented as mean ± SEM. Statistical analysis was performed using unpaired two-sided Student’s t test (D and F–H), log-rank (Mantel-Cox) test (E), and paired two-sided Student’s t test (I–K). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. See also Figure S5.