IFN-γ-driven iNOS induction in macrophages mediates CAR T cell resistance in B cell lymphoma

Sae Bom Lee; Allan Pavy; Nolan Beatty; Justin C Boucher; Jay K Mandula; Emiliano Roselli; Darwin C Chang; Rachel V Jimenez; Hiroshi Kotani; Kayla Reid; Julio Vazquez-Martinez; Payal Goala; Max Ilecki; Cooper J Sailer; Shannon L McSain; Showkat Hamid; Muhammad Junaid Tariq; Rosa A Sierra-Mondragon; Min Liu; John M Koomen; Jonathan Nguyen; Mohammad Hussaini; Timothy I Shaw; Xuefeng Wang; Jianmin Wang; Julieta Abraham-Miranda; Rawan Faramand; Michael D Jain; Meredith L Stone; Frederick L Locke; Paulo C Rodriguez; Yun Pyo Kang; Marco L Davila

doi:10.1038/s41467-026-76972-9

. 2026 Aug 31;17:10375. doi: 10.1038/s41467-026-76972-9

IFN-γ-driven iNOS induction in macrophages mediates CAR T cell resistance in B cell lymphoma

Sae Bom Lee ^1,², Allan Pavy ³, Nolan Beatty ^1,², Justin C Boucher ¹, Jay K Mandula ^1,², Emiliano Roselli ⁴, Darwin C Chang ^1,², Rachel V Jimenez ¹, Hiroshi Kotani ⁵, Kayla Reid ⁴, Julio Vazquez-Martinez ¹, Payal Goala ^1,², Max Ilecki ³, Cooper J Sailer ³, Shannon L McSain ³, Showkat Hamid ³, Muhammad Junaid Tariq ³, Rosa A Sierra-Mondragon ¹, Min Liu ⁶, John M Koomen ⁶, Jonathan Nguyen ⁶, Mohammad Hussaini ⁷, Timothy I Shaw ⁸, Xuefeng Wang ⁸, Jianmin Wang ⁹, Julieta Abraham-Miranda ¹⁰, Rawan Faramand ¹⁰, Michael D Jain ¹⁰, Meredith L Stone ³, Frederick L Locke ¹¹, Paulo C Rodriguez ¹, Yun Pyo Kang ¹¹, Marco L Davila ^4,^✉

PMCID: PMC13627683 PMID: 42816487

Abstract

Chimeric antigen receptor (CAR) T cell therapies have revolutionized B cell malignancy treatment, but many patients with large B cell lymphoma (LBCL) experience primary resistance or relapse. To uncover resistance mechanisms, here we examine pre-infusion tumor biopsies and observe that increased immunoregulatory macrophages correlate with poor clinical responses. In murine models, CAR T cell-produced interferon-gamma (IFN-γ) upregulates inducible nitric oxide synthase (iNOS, NOS2) in immunoregulatory macrophages, impairing CAR T cell function. Proteomic profiling reveals that iNOS-expressing macrophages promote apoptosis and cell cycle arrest while downregulating protein synthesis machinery in CAR T cells. Metabolically, CAR T cells exhibit reduced glycolytic intermediates and altered tricarboxylic acid (TCA) cycle activity. Pharmacological inhibition of iNOS enhances CAR T cell treatment efficacy in vivo. Elevated levels of iNOS⁺CD14⁺ monocytes in leukaphereses are associated with non-durable responses to CAR T cells. Targeting iNOS in immunoregulatory macrophages, potentially by modulating CAR T-produced IFN-γ, could improve LBCL outcomes.

Subject terms: Immunotherapy, Haematological cancer, Tumour immunology

Tumor associated macrophages can support cancer progression by suppressing T cell effector functions. Here the authors report that induction of iNOS in tumor-associated macrophages by IFN-gamma secreted from CAR T cells contributes to the development of CAR-T resistance in B cell lymphoma.

Introduction

The success of CD19-targeted CAR T cell therapies has advanced the treatment of B cell malignancies^1–7. However, a substantial proportion of patients with LBCL experience primary resistance or relapse, supporting the need to augment CAR T cell efficacy^8–11. Factors hindering the effectiveness of CAR T cell therapy include a high tumor burden prior to CAR T cell infusion^12,13, loss of or decreased CD19 expression on tumor cells^14,15, tumor genetic alterations^16,17, and the highly differentiated or dysfunctional state of CAR T cells^18,19. Recent studies have also emphasized the importance of the TME in determining clinical outcomes in CAR T therapy patients^20,21.

The TME of B cell lymphoma contains various immune cell types^22,23, including macrophages, myeloid-derived suppressor cells (MDSCs), and regulatory T cells (Tregs), which can impede the recruitment, expansion, and activity of T cells, including endogenous T cells and infused CAR T cells. In LBCL, the pre-infusion TME, characterized by elevated expression of genes associated with immune suppression and diminished T cell-related signatures, is correlated with relapse after CAR T cell therapy²¹. In contrast, higher rates of complete response are associated with a TME that exhibits immune gene signatures linked to cytotoxic T cell activation and is enriched with chemokines and cytokines that potentiate T cell involvement.

Macrophages perform diverse biological functions in response to tissue pathophysiology and environmental cues, acting as central mediators in immune responses²⁴. They are key contributors to immune-mediated toxicities associated with CAR T cell therapy, such as cytokine release syndrome (CRS), which is in part mediated through the release of inflammatory cytokines by myeloid cells^25–27. Activated macrophages secrete proinflammatory cytokines TNF-α, IL-6, and IL-1β and upregulate inducible nitric oxide synthase (iNOS) and nitric oxide (NO), exacerbating CRS through induction of endothelial dysfunction and vascular leakage²⁶. Additionally, macrophages facilitate cancer progression in various cancers by suppressing T cell effector function through multiple processes, including expression of inhibitory checkpoint ligands such as programmed cell death-ligand 1 (PD-L1), secretion of inhibitory cytokines such as TGF-β and IL-10, and depletion of amino acids, including arginine and tryptophan^28–31. However, the crosstalk between CAR T cells and macrophages as a mechanism of therapeutic resistance remains poorly defined.

Here, we report the reciprocal interactions between CAR T cells and macrophages contributing to CAR T cell therapy resistance in B cell lymphoma. Our findings demonstrate that IFN-γ produced by CAR T cells induces phenotypic changes in macrophages, amplifying their immunoregulatory potential. Notably, iNOS induction in macrophages drives CAR T cell dysfunction by impairing their functional and metabolic capacities. This study proposes a critical counter-regulatory mechanism in which IFN-γ-producing CAR T cells restrict therapeutic efficacy. These insights highlight the targetable macrophage-driven resistance mechanisms to improve CAR T cell therapy outcomes in B cell lymphoma.

Results

Macrophages in the pre-CAR T cell treatment TME are linked to therapeutic responses in LBCL patients

We examined the tumor immune infiltrate and its relationship with clinical outcomes in patients with LBCL receiving Axicabtagene ciloleucel (axi-cel). Bulk RNA sequencing (RNA-seq) was performed on patient tumor biopsies taken before lymphodepletion and CAR T cell treatment. Subsequently, CIBERSORTx was used to deconvolute intratumoral immune cell composition (Fig. 1A)^32,33. We found that patients with non-durable responses (NDR) to CAR T cell therapy, characterized by lymphoma relapse or death from disease, exhibited a higher proportion of transcriptionally identified M2-like macrophages compared to patients with durable responses (DR), who remained in remission for at least 6 months following axi-cel infusion (Fig. 1B). Similarly, gene set enrichment analysis (GSEA) revealed the enrichment of M2 macrophage-associated genes in patients with NDR (Fig. 1C). The proportion of nonactivated macrophages (M0) was lower in patients with NDR, while levels of M1-like macrophages were similar between patients with NDR and DR (Fig. 1B). Furthermore, we observed that a higher abundance of M2-like macrophages in patients correlated with worse progression-free survival after axi-cel therapy (Fig. 1D). These findings collectively indicate that the presence of M2-like macrophages within the TME prior to CAR T cell therapy is associated with poor therapeutic responses to axi-cel in patients with LBCL.

Immunoregulatory actions of macrophages on CAR T cells

To explore how macrophages may impact the cellular function of CAR T cells, we employed a syngeneic coculture system (Fig. 2A, Supplementary Fig. 1A). In this model, murine anti-CD19 CAR T cells³⁴, which include CD28 and CD3ζ signaling domains linked to a fluorescent mCherry reporter (1928z) (Supplementary Fig. 1B), were cocultured with a murine malignant B cell line (Eμ-myc cells) in the presence or absence of mouse bone marrow-derived macrophages (BMDMs). The BMDMs were either unpolarized (unMac), polarized with IFN-γ and LPS for “M1-like” macrophages, or polarized with type 2 cytokines IL-4 and IL-10 to exhibit an M2-like phenotype. We have termed these M2-like polarized macrophages as ‘imMac’ to emphasize their distinctive immunoregulatory activity when cocultured with CAR T cells. CAR T cells cocultured with imMac showed increased cell death (Fig. 2B) and reduced proliferation (Fig. 2C) compared to CAR T cells cocultured with unMac, M1-like macrophages, or without macrophages, while diminished CAR T expansion was observed in both imMac and M1-like macrophage cocultures (Fig. 2D and Supplementary Fig. 1C). 48 h was chosen as a time point to measure CAR T cell expansion because nutrient depletion limits CAR T cell survival over a longer timeframe (Supplementary Fig. 1D), while refreshing the media could disrupt CAR T-macrophage interactions. Moreover, CAR T cells cocultured with imMac showed lower total CAR expression (Fig. 2E) as well as reduced surface CAR expression (Fig. 2F) compared to CAR T cells cocultured with unMac, M1-like macrophages, or without macrophages. We next explored the impact of imMac on CAR T cell effector function. To exclude the direct contribution of macrophage effector activities in these functional assays, we first cocultured CAR T cells, Eμ-myc cells, and macrophages for 48 h (Fig. 2G). Next, CAR T cells isolated from the cocultures were evaluated for their effector function against fresh Eμ-myc cells. CAR T cells derived from cocultures with imMac exhibited impaired production of effector cytokines IFN-γ and tumor necrosis factor-alpha (TNF-α) (Fig. 2H, I) and demonstrated decreased ability to lyse target tumor cells (Fig. 2J). In contrast, exposure to unMac did not impact tumor lytic capacity or cytokine secretion, while exposure to M1-like macrophages had a more modest inhibitory effect than imMac (Fig. 2H-J). Collectively, these results show that macrophages polarized towards an M2-like phenotype exert immunoregulatory actions that impair multiple aspects of CAR T cell biology, including survival, expansion, and CAR-dependent effector functions.

CAR T cell-exposed imMac upregulate iNOS

We next interrogated the metabolic crosstalk between CAR T cells and imMac to investigate the potential involvement of immune-metabolic alterations. We conducted a comprehensive analysis of metabolite profiles in the supernatants collected from our coculture model via global metabolomics using liquid chromatography-mass spectrometry (LC-MS) (Fig. 3A). We observed a dramatic increase in citrulline and ornithine and a concomitant reduction in arginine within the supernatants derived from cocultures containing imMac compared to cocultures containing unMac or no macrophages (No Mac) (Fig. 3B–F). Macrophages possess the capacity to metabolize arginine through arginase-1 (ARG-1) or iNOS, producing ornithine and urea or citrulline and NO, respectively³⁵ (Supplementary Fig. 2A). We observed that in the absence of CAR T cells, imMac exhibited high expression levels of ARG-1 but minimal iNOS, whereas unMac displayed minimal expression of both ARG-1 and iNOS (Fig. 3G). When CAR T cells were cocultured with macrophages, there was a significant induction of iNOS in imMac and, to a lesser extent, in unMac. The expression levels of ARG-1 in unMac and imMac remained unchanged, regardless of the presence of CAR T cells in the cocultures. Neither CD3⁺ T cells nor Eμ-myc cells expressed ARG-1 or iNOS, confirming that the expression of these enzymes was limited to macrophages in this model (Supplementary Fig. 2B). Consistent with the enhanced iNOS in imMac, higher levels of NO were produced in cocultures with imMac compared to cocultures with unMac or without macrophages (Fig. 3H). Additionally, coculture with CAR T cells substantially increased the expression of PD-L1 in both unMac and imMac (Supplementary Fig. 2C). Together, these findings demonstrate that exposure of imMac to CAR T cells induces phenotypic changes in imMac, including enhanced arginine metabolism through the upregulation of iNOS.

Fig. 3 — A Schematic for (B–F) where CAR T cells and Eμ-myc cells were cocultured with unMac or imMac or without macrophages (No Mac) for 24 h. Coculture supernatants were analyzed by global metabolomics using LC-MS. B, C Quantification of relative abundance of metabolites in the supernatants derived from cocultures containing imMac versus No Mac (B) or unMac (C) (n = 3 technical replicates from a single experiment). D–F Arginine, ornithine, and citrulline levels in the supernatants of coculture groups (n = 3 technical replicates from a single experiment). The data were normalized to the mean value of the no-macrophage control condition. G Flow cytometry analysis of expression of ARG-1 and iNOS in unMac or imMac cocultured with Eμ-myc cells in the presence or absence of CAR T cells at 24 h. Data are from n = 3 independent experiments. H NO levels in the supernatants derived from cocultures of CAR T cells and Eμ-myc cells with unMac or imMac or without macrophages at 48 h were analyzed by Griess assay. Data are from n = 3 independent experiments. Data in (D–H) are presented as mean ± SD. Statistical significance was determined by unpaired two-sided Student’s t tests (G) or one-way ANOVA with Bonferroni correction for multiple comparisons (H). Exact P values are provided in the Source Data file. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns not significant, FC fold change, LC-MS liquid chromatography-mass spectrometry, iNOS inducible nitric oxide synthase, ARG-1 arginase-1, NO nitric oxide. Source data are provided as a Source Data file.

iNOS upregulation in imMac drives suppression of CAR T cell function

To investigate whether arginine metabolism by imMac contributes to the impairment of CAR T cell function, we examined the effects of ARG-1 and iNOS inhibitors (Supplementary Fig. 3A). Treatment with the ARG-1 inhibitor nor-NOHA³⁶ did not restore CAR T cell expansion in cocultures with imMac (Supplementary Fig. 3B). The supraphysiological arginine concentration (1.15 mM) in culture media could diminish the arginine depletion effect by ARG-1; thus, we repeated the coculture at the physiological arginine concentration (150 μM)³⁷. However, even in the context of physiological arginine levels, treatment with nor-NOHA failed to rescue CAR T cell expansion (Supplementary Fig. 3C), suggesting ARG-1 by itself is not sufficient to inhibit CAR T cells. Next, treatment of cocultures with the iNOS inhibitor L-NIL^25,38 restored CAR T cell expansion, tumor-killing capacity, and production of the effector cytokines IFN-γ and TNF-α in the presence of imMac (Fig. 4A–D). Consistently, imMac developed from BMDMs of iNOS-deficient (iNOS^−/−) mice did not inhibit CAR T cell expansion (Fig. 4E) or impair CAR T cell tumor killing capacity (Fig. 4F). Importantly, neither pharmacologic inhibition nor genetic ablation of iNOS in M1-like macrophages impaired CAR T cell expansion or tumor-killing capacity (Supplementary Fig. 3D). iNOS^−/− imMac expressed similar levels of ARG-1 and PD-L1 as wild-type (WT) imMac (Supplementary Fig. 3E), indicating that these factors were not responsible for the suppression of CAR T cell function by imMac. Inhibition or genetic ablation of iNOS attenuated the production of NO (Supplementary Fig. 3F, G) and citrulline (Supplementary Fig. 3H) in the cocultures with imMac. The levels of arginine and ornithine remained comparable regardless of iNOS ablation or inhibition (Supplementary Fig. 3I, J), ruling out their altered levels as the drivers of CAR T cell impairment. These findings collectively demonstrate that imMac suppress CAR T cell function through the enzymatic activity of iNOS.

Fig. 4 — A CAR T cells and Eμ-myc cells were cocultured with unMac, imMac or no macrophages, +/- L-NIL (50 μM). CAR T cell expansion was measured over 48 h using live cell analysis. Data are from n = 3 independent experiments. Within each independent experiment, technical replicates were averaged, and values were normalized to the corresponding no-macrophage control condition at each time point. B–D After 48 h coculture described in (A), isolated CAR T cells were cocultured with fresh Eμ-myc cells. B Target cell lysis was assessed by luciferase assay at 24 h (n = 4 technical replicates from a single experiment, repeated 2 times with similar results). C, D IFN-γ and TNF-α in coculture supernatants at 36 h were analyzed by ELISA (n = 4 technical replicates from a single experiment). E CAR T cells and Eμ-myc cells were cocultured with WT or iNOS^−/− unMac or imMac or without macrophages. CAR T cell expansion was measured over 48 h by a live cell analysis (n = 4 technical replicates from a single experiment, repeated 2 times with similar results). F After 48 h of coculture described in (E), isolated CAR T cells were cocultured with fresh Eμ-myc cells and target cell lysis was assessed at 36 h (n = 4 technical replicates from a single experiment, repeated 2 times with similar results). G CAR T cells and Eμ-myc cells were cocultured with NCX-4016 or 0.01% DMSO (Veh). CAR T cell expansion was measured over 48 h by a live cell analysis. Data are from n = 3 independent experiments. Within each independent experiment, technical replicates were averaged, and values were normalized to the corresponding +Veh control condition at each time point. H CAR T cells were cocultured with Eμ-myc cells +/- peroxynitrite (PNT). CAR T cell expansion was measured over 48 h by a live cell analysis. Data are from n = 3 independent experiments. Within each independent experiment, technical replicates were averaged, and values were normalized to the corresponding no-treatment control condition at each time point. I After 48 h of initial coculture described in G-H, isolated CAR T cells were cocultured with fresh Eμ-myc cells, and target cell lysis was assessed at 24 h (n = 4 technical replicates from a single experiment). J CAR T cells and Eμ-myc cells were cocultured with or without imMac +/- carboxyl-PTIO (c-PTIO). CAR T cell expansion was measured over 48 h by live cell analysis. Data are from n = 3 independent experiments. Within each independent experiment, technical replicates were averaged, and values were normalized to the corresponding no-macrophage control condition at each time point. All data are presented as mean ± SD. Statistical significance was determined by unpaired two-sided Student’s t tests (A, G, H, J). Exact P values are provided in the Source Data file. *, P < 0.05; **, P < 0.01; ***, P < 0.001; exp exposed CAR T cells isolated from prior coculture. Source data are provided as a Source Data file.

We further investigated whether the iNOS products, citrulline and NO, were responsible for CAR T cell suppression by imMac. Exposure to high levels of citrulline did not impact the expansion of CAR T cells (Supplementary Fig. 3K). NO reacts with superoxide to form peroxynitrite (PNT, ONOO⁻), which leads to protein oxidation, lipid peroxidation, and DNA damage³⁹. Treatment with the NO-donor NCX-4016 or PNT resulted in the inhibition of CAR T cell expansion (Fig. 4G, H) and diminished ability of CAR T cells to kill tumor cells (Fig. 4I) and secrete effector cytokines IFN-γ and TNF-α (Supplementary Fig. 3L). Notably, treatment with NO-scavenger carboxyl-PTIO (c-PTIO) partially rescued the expansion of CAR T cells during cocultures with imMac (Fig. 4J). These data indicate that NO and PNT act as key mediators of iNOS-induced dysfunction of CAR T cells.

CAR T cell-derived IFN-γ induces iNOS in imMac

CAR T cells secrete cytokines, such as IFN-γ and TNF-α, that activate macrophages^40,41. A previous study reported that CAR T cell-derived IFN-γ upregulated iNOS in TAMs and their secretion of chemokines that enabled further recruitment of CAR T cells in a lung adenocarcinoma mouse model⁴². We found that neutralization of IFN-γ with blocking antibodies attenuated CAR T cell-triggered iNOS expression in unMac and imMac (Fig. 5A) and reduced the production of NO (Fig. 5B). Furthermore, treatment with anti-IFN-γ enhanced CAR T cell expansion in cocultures with imMac (Fig. 5C) and preserved the ability of CAR T cells to lyse tumor cells (Fig. 5D) and produce IFN-γ and TNF-α (Fig. 5E, F). Consistently, CAR T cells deficient in IFN-γ (IFN-γ^−/− CAR T cells) neither induced iNOS expression in unMac and imMac nor induced NO production in cocultures, while iNOS expression in previously polarized M1-like macrophages was preserved (Supplementary Fig. 4A–C). Moreover, IFN-γ^−/− CAR T cells exhibited enhanced expansion during cocultures with imMac (Supplementary Fig. 4D). Because IFN-γ signaling can upregulate multiple cytokines that can, in turn, induce iNOS in macrophages²⁵, we next investigated the impact of blocking IFN-γ, IL-1β, or IL-6R on iNOS expression in macrophages cocultured with CAR T cells and Eμ-myc tumor cells. Only treatment with anti-IFN-γ reduced iNOS expression on imMac and unMac (Supplementary Fig. 4E). These findings demonstrate that blocking IFN-γ production by CAR T cells mitigates the counter-regulatory, iNOS-driven inhibitory effects of imMac, while sparing the iNOS expression associated with the anti-tumor potential of M1-like macrophages.

Fig. 5 — A–C CAR T cells and Eμ-myc cells were cocultured with unMac or imMac in the presence of anti-IFN-γ (10 μg/mL) or IsoCon (10 μg/mL). (A) Expression of ARG-1 and iNOS in unMac or imMac was analyzed at 24 h via flow cytometry (n = 3 technical replicates from a single experiment, repeated 2 times with similar results). B NO levels in coculture supernatants at 44 h were analyzed by Griess assay. Data are from n = 3 independent experiments. Within each independent experiment, technical replicates were averaged, and values were normalized to the +imMac+IsoCon control. (C) CAR T cell expansion was measured over 48 h by a live cell analysis system (n = 4 technical replicates from a single experiment, repeated 2 times with similar results). D–F After initial coculture of CAR T cells and Eμ-myc cells with unMac or imMac in the presence of anti-IFN-γ (10 μg/mL) or IsoCon (10 μg/mL) for 48 h, isolated CAR T cells were cocultured with fresh Eμ-myc cells. D Luciferase-based lysis of Eμ-myc cells by CAR T cells was assessed at 24 h (n = 4 technical replicates from a single experiment). E, F Levels of IFN-γ and TNF-α in the coculture supernatants at 24 h were analyzed by ELISA (n = 4 technical replicates from a single experiment). G Schematic for (H) where WT or IFN-γ^−/− CAR T cells were cocultured with Eμ-myc cells in the presence of unMac or imMac. After coculture for 16 h, scRNA-seq was performed on T cells. H Pathway enrichment analysis of differentially expressed genes in CAR T cells from cocultures of WT CAR T cells and imMac, compared to those from WT CAR T cells and unMac or IFN-γ^−/− CAR T cells and imMac cocultures (n = 2 technical replicates from a single experiment). Data in (A–F) are presented as mean ± SD. Statistical significance was determined by one-way ANOVA with Bonferroni correction for multiple comparisons (B). Exact P values are provided in the Source Data file. ***, P < 0.001; ****, P < 0.0001; IsoCon isotype control, NO nitric oxide, exp exposed CAR T cells isolated from prior coculture. Source data are provided as a Source Data file.

To investigate the regulatory genes and pathways underlying the detrimental interactions between CAR T cells and imMac, we performed single-cell RNA sequencing (scRNA-seq) on cocultures of wild-type (WT) or IFN-γ^−/− CAR T cells with unMac or imMac (Fig. 5G). Unsupervised clustering analysis revealed distinct subpopulations of T cells and macrophages (Supplementary Fig. 5A–C). Among the six major macrophage subpopulations identified, cluster 0 was the most predominant subset in imMac cocultured with WT CAR T cells (Supplementary Fig. 5D). This cluster was characterized by elevated expression of Nos2/iNOS and inhibitory checkpoint ligands such as Cd274 (PD-L1) and Pdcd1lg2 (PD-L2) (Supplementary Fig. 5E). Additionally, cluster 0 exhibited increased expression of immunosuppressive genes, including Ptgs2 (COX2), Entpd1 (CD39), Il18bp, Fgl2, and Mertk (Supplementary Fig. 5E). Pathway analysis of genes upregulated in imMac cocultured with WT CAR T cells revealed an enrichment of the IFN-γ and IFN-α response pathways, as well as the hypoxia-inducible factor (HIF)-1α pathway (Supplementary Fig. 5F). Notably, increased Nos2 (iNOS) expression within macrophage subclusters strongly correlated with HIF-1α pathway activation, suggesting HIF-1α signaling mediates Nos2 (iNOS) induction in imMac (Supplementary Fig. 5G, H)^43,44. These findings indicate that CAR T cell-derived IFN-γ reshapes the transcriptomic landscape of imMac, enhancing their immunoregulatory potential. Additionally, pathway analysis revealed enrichment of the HIF-1α and p53 pathways in WT CAR T cells cocultured with imMac, suggesting that imMac induce hypoxia-like conditions and cellular stress responses in CAR T cells, which likely also contributes to their functional impairment (Fig. 5H).

iNOS-expressing imMac induce CAR T cell metabolic dysregulation

To elucidate the mechanisms underlying CAR T cell dysfunction induced by imMac, we performed proteomics analysis on CAR T cells from our culture model (Fig. 6A). Principal component analysis (PCA) revealed distinct protein profiles in CAR T cells cocultured with imMac compared to CAR T cells cocultured with unMac or without macrophages, with these differences reversed by L-NIL treatment (Fig. 6B). Consistent with the scRNA-seq data (Fig. 5H), pathway analysis demonstrated enrichment of the hypoxia and p53 pathways in CAR T cells cocultured with imMac (Fig. 6C). Notably, we observed an increase in p53 target proteins associated with cell cycle arrest (CDKN1A/p21 and GTSE1), apoptosis (BBC3/PUMA, APAF1, FAS, BAX, CASP6, and CASP7), and DNA damage repair (MGMT, ERCC5, POLK, and XPC) (Fig. 6D-G). Phosphoproteomics analysis further revealed increased phosphorylation of CDKN1A/p21 at Serine 78, which can enhance its ability to inhibit cyclin-dependent kinases (CDKs) and halt the cell cycle (Supplementary Fig. 6A–D)⁴⁵. Additionally, pathway analysis identified downregulation of MYC target proteins in CAR T cells cocultured with imMac (Fig. 6C). Specifically, we observed reduced expression of MYC targets involved in ribosome biogenesis (RSL1D1, LSM2, PWP1, NOLC1, RPL18, and RPLP0), protein translation (EIF3J1/EIF3J2, EIF4A1, HNRNPA2B1, EEF1B2, ETF1, ABCE1, RACK1, VBP1, and HSP90AB1), and amino acid uptake (SLC7A5, SLC1A5, SLC3A2) (Fig. 6D–F, H)⁴⁶. Collectively, these findings support that iNOS-expressing imMac induces gene expression programs in CAR T cells that impair protein synthesis, cell cycle progression, and survival, ultimately dysregulating their antitumor efficacy.

Fig. 6 — CAR T cells and Eμ-myc cells were cocultured with unMac, imMac or without macrophages (No Mac), +/- L-NIL (50 μM). After coculture for 48 h, proteomics was performed on T cells. A Experimental design. B Principal component analysis (PCA) plot of proteomics data showing distinct clustering of samples (n = 3 technical replicates from a single experiment). C Pathway enrichment analysis of differentially expressed proteins (DEPs) in CAR T cells cocultured with imMac versus No Mac, unMac, or imMac+L-NIL (50 μM) (n = 3 technical replicates from a single experiment). D–F The p53 and MYC pathway DEPs in CAR T cells cocultured with imMac versus No Mac (D), unMac (E), or imMac+L-NIL treatment (F) (n = 3 technical replicates from a single experiment). G DEPs in the p53 pathway (n = 3 technical replicates from a single experiment). H DEPs in MYC target proteins (n = 3 technical replicates from a single experiment). Exp exposed CAR T cells isolated from prior coculture. Source data are provided as a Source Data file.

Cellular metabolism plays a crucial role in supporting rapid proliferation and effector function of T cells^47,48. To investigate whether metabolic dysregulation of CAR T cells is also induced by imMac, we conducted global metabolomics analysis on CAR T cells from our coculture model. We found profound alterations in glycolytic and TCA cycle intermediates in CAR T cells cocultured with imMac, which were restored with L-NIL treatment (Fig. 7A–D). Specifically, there was a marked depletion of glycolytic intermediates such as fructose 1,6-bisphosphate (F1,6BP), glyceraldehyde 3-phosphate (G3P), and dihydroxyacetone phosphate (DHAP) (Fig. 7E–G). Concurrently, there was an iNOS-dependent accumulation of TCA cycle metabolites citrate, aconitate, and succinate, along with a decrease in malate (Fig. 7H–K). Additionally, we observed a substantial accumulation of itaconate in CAR T cells cocultured with imMac (Fig. 7L). Itaconate is synthesized from aconitate via immune response gene 1 (IRG1), also known as aconitate decarboxylase (ACOD1), in tumor-associated myeloid cells, and uptake of itaconate by CD8⁺ T cells has been shown to suppress their proliferation and cytolytic activity⁴⁹. Exposure of CAR T cells to a cell-permeant form of itaconate, 4-octyl itaconate (4-OI)⁵⁰, impaired their expansion (Fig. 7M). Given the concurrent accumulation of itaconate with citrate and aconitate in CAR T cells cocultured with imMac, we investigated whether CAR T cells can produce itaconate. Through ¹³C₆-glucose tracing on CAR T cells, we identified iNOS-dependent accumulation of ¹³C-labeled citrate, aconitate, and itaconate, as well as a reduction of ¹³C₆-labeled α-ketoglutarate (αKG), fumarate, and malate in CAR T cells cocultured with imMac (Supplementary Fig. 7A–G). Immunoblot analysis confirmed increased expression of IRG1 and decreased expression of isocitrate dehydrogenase 2 (IDH2) in CAR T cells cocultured with imMac (Supplementary Fig. 7H). These results indicate that imMac, via iNOS, deplete glycolytic intermediates and rewire the TCA cycle to divert aconitate towards itaconate production instead of αKG. These metabolic alterations were further corroborated by extracellular flux analysis, which revealed attenuated glycolytic and oxidative metabolic activities in CAR T cells cocultured with imMac, as evidenced by decreased extracellular acidification rate (ECAR) and oxygen consumption rate (OCR), which are proxies for glycolytic rate and mitochondrial oxidative phosphorylation, respectively (Fig. 7N). Importantly, L-NIL treatment preserved glycolytic and oxidative metabolic capacities, mitigating the metabolic disruptions caused by imMac.

Fig. 7 — A–L CAR T cells and Eμ-myc cells were cocultured with unMac or imMac or without macrophages (No Mac), +/- L-NIL (50 μM). After coculture for 48 h, global metabolomics was performed on T cells. Quantification of relative abundance of metabolites in CAR T cells derived from cocultures with imMac versus No Mac (A), unMac (B), or imMac+L-NIL treatment (C) (n = 4 technical replicates from a single experiment). D Schematic depicting altered metabolites associated with glycolysis pathway and TCA cycle. (E–L) Levels of F1,6BP, G3P, DHAP, citrate, aconitate, succinate, malate, and itaconate (n = 4 technical replicates from a single experiment). The data were normalized to the mean value of CAR T cells derived from cocultures without macrophages and L-NIL treatment. M CAR T cells were cocultured with Eμ-myc cells in the presence or absence of 4-OI. CAR T cell expansion was measured over 48 h by a live cell analysis system (1928z CAR T ( + Veh), n = 4; +4-OI (500 µM), n = 3; +4-OI (1 mM), n = 4; +4-OI (2 mM), n = 4; technical replicates from a single experiment, repeated 2 times with similar results). N CAR T cells and Eμ-myc cells were cocultured with imMac or without macrophages (No Mac), +/- L-NIL (50 μM). After coculture for 48 h, seahorse assay was performed on CAR T cells (n = 6 technical replicates from a single experiment). ECAR was measured in response to glucose (Glc), ATP synthase inhibitor (Oligo), or hexokinase II inhibitor (2-DG). OCR was measured in response to Oligo, mitochondrial oxidative phosphorylation uncoupler (FCCP), or electron transport chain complex I/III inhibitor (Rot/AA). Data in (E–N) are presented as mean ± SD. F1,6BP fructose 1,6-bisphosphate, G3P glyceraldehyde 3-phosphate, DHAP dihydroxyacetone phosphate, 4-OI 4-octyl itaconate, ECAR extracellular acidification rate, OCR oxygen consumption rate, Oligo oligomycin A, 2-DG 2-deoxy-D-glucose, FCCP carbonyl cyanide p-trifluoromethoxyphenylhydrazone, Rot/AA rotenone/antimycin A, exp exposed CAR T cells isolated from prior coculture. Source data are provided as a Source Data file.

iNOS inhibition improves CAR T cell therapeutic efficacy

We proceeded to investigate whether iNOS can limit the effectiveness of CAR T cell therapy in vivo. To eliminate the effects of tumor antigen spreading on bystander T cells and secondary sources of IFN-γ production, we utilized Rag1^−/− mice, which lack endogenous T and B cells. Eμ-myc B cell tumors were established in the peritoneal cavity (Fig. 8A). CAR T cells carrying a truncated CD3ζ signaling domain (19dz) were used as non-functional controls (Supplementary Fig. 1A). Intraperitoneal transfer of WT 1928z, IFN-γ^−/− 1928z, or WT 19dz CAR T cells was performed to facilitate direct interaction with macrophages at the tumor site. We found that the frequency of iNOS⁺ macrophages was significantly elevated in WT 1928z CAR T cell-treated mice compared to mice treated with IFN-γ^−/− 1928z or WT 19dz CAR T cells (Fig. 8B). The frequencies of ARG1⁺ macrophages and total F4/80⁺ CD11b⁺ macrophages were similar across all groups (Fig. 8C, D). Furthermore, CD11b⁺ myeloid cells from the peritoneal cavity of WT 1928z CAR T cell-treated tumor-bearing mice suppressed the expansion of fresh antigen-naïve CAR T cells ex vivo in an iNOS-dependent manner (Fig. 8E). To investigate the in vivo impact of therapeutic IFN-γ neutralization on CAR T cell efficacy and iNOS expression, we evaluated both IFN-γ deletion in CAR T cells or systemic IFN-γ blockade in immunocompetent C57BL/6 mice bearing Eμ-myc tumors. As expected, 1928z CAR T cells significantly inhibited tumor growth and extended survival compared to 19dz CAR T cells across all settings (Supplementary Fig. 8A–F). Notably, neither genetic ablation of IFN-γ in CAR T cells nor systemic anti-IFNγ treatment significantly compromised the overall anti-tumor efficacy of CAR T cell therapy. Mechanistically, IFN-γ deletion in CAR T cells but not anti-IFN-γ effectively reduced iNOS expression in macrophages (Supplementary Fig. 8G). The failure of the blocking antibody to reduce iNOS likely reflects the pharmacokinetic challenges of achieving sufficient neutralizing concentrations within the local TME. Consistent with the lack of local effect, systemic anti-IFN-γ treatment did not alter CAR T cell frequencies in the peripheral blood or spleen, nor did it impact systemic TNF-α levels (Supplementary Fig. 8H, I). We next assessed whether direct inhibition of iNOS could improve the therapeutic efficacy of CAR T cells (Fig. 8F). Rag1^−/− mice treated with a combination of 1928z CAR T cells and L-NIL exhibited significantly improved survival compared to mice treated with 1928z CAR T cells alone (Fig. 8G). Similarly, in C57BL/6 immune-competent mice bearing Eμ-myc B cell tumors, combinatorial treatment with 1928z CAR T cells and L-NIL resulted in superior tumor control (Supplementary Fig. 8J, K). These results demonstrate that IFN-γ-producing CAR T cells stimulate iNOS in macrophages at the tumor site, and inhibition of iNOS enhances the therapeutic efficacy of CAR T cells.

Fig. 8 — A Experimental settings for (B–E) where Eμ-myc cells were intraperitoneally injected into *Rag1*^−/− mice. Seven days later, WT 1928z, IFN-γ^−/− 1928z, or WT 19dz CAR T cells were transferred into mice. Peritoneal lavage cells were obtained 24–40 h after CAR T cell transfer. B–D Flow cytometry analysis of iNOS⁺ cells among CD11b⁺F4/80⁺ macrophages (B), ARG-1⁺ cells among CD11b⁺F4/80⁺ macrophages (C), and CD11b⁺F4/80⁺ macrophages among CD45⁺ cells (D) at 24 h after CAR T cell transfer (n = 6 mice/group). E CD11b⁺ peritoneal myeloid cells were obtained 40 h following WT 1928z CAR T cell transfer. CD11b⁺ cells were then ex vivo cocultured with fresh antigen-naïve CAR T cells and Eμ-myc cells in the presence or absence of L-NIL (50 μM) (coculture ratio, 1:1:1 = CD11b⁺ cell:CAR T:Eμ-myc cell). Expansion of CAR T cells was measured over 48 h by a live cell analysis system (technical replicates from a single experiment, +Veh, n = 4; +L-NIL, n = 4; +CD11b⁺ cells+Veh, n = 5; + CD11b⁺ cells+L-NIL, n = 5). (F) Experimental settings for (G). G Percentage of survival of tumor-bearing mice treated with WT 19dz or WT 1928z CAR T cells receiving L-NIL or PBS (vehicle). Results are from two pooled independent experiments (PBS, n = 12 mice; L-NIL, n = 12 mice; 19dz CAR T + PBS, n = 11 mice; 19dz CAR T + L-NIL, n = 12 mice; 1928z+PBS, n = 28 mice; 1928z+L-NIL, n = 34 mice). H Percentage of iNOS⁺ cells among CD45⁺CD3^-CD20^-CD56^-CD11b⁺CD14⁺ monocytes in pre-lymphodepletion leukapheresis (DR, 19 patients; NDR, 32 patients). I Human CAR T cells were cocultured with Raji cells in the presence of human monocyte-derived unpolarized macrophages (unMac; n = 4 technical replicates from 1 human donor), IL-4, IL-10, IL-13-activated macrophages (imMac; n = 4 technical replicates from 1 human donor) or macrophages activated with serum from patients in complete remission (CR) at 3 months post-treatment or from non-responders (NR) (CR, 3 patients; NR, 7 patients). CAR T cell expansion was measured over 48 h by a live cell analysis system. Area under the curve (AUC) values were compared. Data in (B–D) are presented as mean ± SEM and (E, I) are presented as mean ± SD. For the box plot in (H), the center line represents the median, the bounds of the box indicate the 25th and 75th percentiles, and the whiskers extend to the minimum and maximum values, with all individual data points shown. Statistical significance was determined by one-way ANOVA with Bonferroni correction for multiple comparisons (B–D), unpaired two-sided Student’s t tests (I), two-sided log-rank Mantel-Cox test between the 1928z CAR T + PBS and 1928z CAR T + L-NIL groups (G), or two-sided Mann-Whitney U test (H). Exact P values are provided in the Source Data file. *, P < 0.05. Source data are provided as a Source Data file.

A recent clinical study reported that iNOS in pretreatment LBCL tumors is negatively associated with clinical outcomes following axi-cel treatment⁵¹. These intratumoral iNOS+ cells may originate from the peripheral circulation; consistent with this, we observed that NDR patients have increased iNOS⁺ circulating CD11b⁺CD14⁺ monocytes within pre-lymphodepletion leukaphereses (Fig. 8H). Upon trafficking to the tumor, these monocytes can differentiate into TAMs within the TME. The immune-suppressive landscape of the TME is likely shaped by systemic host factors, including baseline inflammation and lymphoma biology. Consistently, we found that pre-lymphodepletion serum from DLBCL patients directly modulated the immunoregulatory activity of macrophages. Specifically, conditioning macrophages with serum from non-responsive (NR) patients enhanced their suppressive capacity, dampening human CD19 CAR T cell expansion to levels comparable to our imMac model. In contrast, macrophages conditioned with serum from complete response (CR) patients resulted in CAR T cell expansion at levels similar to unMac (Fig. 8I).

Discussion

Our findings demonstrate that iNOS upregulation in imMac, provoked by IFN-γ secreted from CAR T cells, impairs various aspects of CAR T cell biology, including expansion, effector function, and metabolism, all of which can reduce the therapeutic efficacy of CAR T cells. Classically, iNOS is considered an antitumor macrophage-associated marker because of its direct tumoricidal effects. Furthermore, its expression is typically co-upregulated alongside key molecules crucial for antigen presentation and costimulation, a process driven by pro-inflammatory cytokines such as IFN-γ, TNF-α, and IL-1β⁵². Our study highlights IFN-γ induction of iNOS in imMac as a critical factor impairing CAR T cell function. This finding aligns with clinical observations across various cancers where elevated iNOS expression in tumors correlated with unfavorable prognoses, highlighting the tumor-promoting potential of iNOS^53–56. Notably, elevated expression of iNOS in pre-CAR T cell treatment TME has been linked to unfavorable outcomes in LBCL patients treated with axi-cel⁵¹. Our data show that unMac also express iNOS but they do not suppress CAR T cell activity to the same extent as imMac. This disparity might be attributed to the lower extent of iNOS induction and subsequent NO production in unMac compared to imMac after exposure to CAR T cells. Moreover, the co-expression of additional immunosuppressive markers, including ARG1, in imMac likely amplifies their inhibitory effects on CAR T cells. Thus, excessive NO production, acting in concert with co-expressed molecules, drives a profoundly immunosuppressive microenvironment that facilitates CAR T therapy failure and tumor progression. By closely mimicking this tumor niche, our experimental design provides critical insights into how CAR T cells inadvertently stimulate imMac toward a suppressive phenotype in the presence of tumor-derived signals.

Mechanistically, the IFN-γ-driven induction of iNOS within imMac upregulates the p53 pathway while downregulating MYC targets in CAR T cells. The resulting accumulation of NO exposes CAR T cells to oxidative and nitrosative stress, triggering damage to DNA and other cellular components. This damage activates the p53 signaling cascade, which promotes cell cycle arrest, apoptosis, and DNA damage repair mechanisms, ultimately resulting in proliferative arrest⁵⁷. Simultaneously, the downregulation of MYC target proteins, which are critical for ribosome biogenesis, amino acid uptake, and protein translation, hinders the protein synthesis required for robust immune activation⁵⁸. Furthermore, this impaired protein synthesis machinery may contribute to the observed reduction in CAR expression on CAR T cells cocultured with imMac.

Our study further reveals that IFN-γ-dependent, iNOS-mediated CAR T cell dysfunction involves profound repression of their glycolytic and oxidative metabolic capacity. The metabolic profiles of CAR T cells are crucial for their antitumor activity, persistence, and differentiation into memory T cells^59,60. Notably, iNOS-expressing imMac triggered rewiring of the TCA cycle, leading to itaconate accumulation in CAR T cells. While IRG1 expression and itaconate production have been previously identified in macrophages and MDSCs as immunosuppressive mediators, their stabilization and functional roles in T cells remain unexplored^49,61. This metabolic rewiring likely arises from the NO-mediated disruption of iron-sulfur (Fe-S) clusters within key TCA cycle enzymes, such as aconitase and succinate dehydrogenase (complex II). Inhibition of these enzymes leads to the accumulation of citrate and succinate, both of which stabilize HIF1-α and trigger a hypoxia-like response^62,63. Furthermore, NO-mediated inhibition of complexes I and IV in the mitochondrial electron transport chain (ETC) impairs oxidative phosphorylation, further exacerbating CAR T cell metabolic dysfunction and compromising their antitumor efficacy.

Our work highlights IFN-γ as a key initiator of the iNOS-dependent inhibitory circuit between CAR T cells and imMac. The role of IFN-γ in the TME and its impact on CAR T cells is pleiotropic. While it can promote tumor cell apoptosis and activate cellular immunity, it also upregulates inhibitory molecules, such as PD-L1, PD-L2, indoleamine 2,3-dioxygenase 1 (IDO), FAS, and FAS ligand (FASL)⁶⁴. During CAR T cell therapy, IFN-γ enhances host antitumor immunity and potentiates CAR T cell-mediated tumor control^40,42,65. The efficacy of IFN-γ-driven antitumor responses depends on the intrinsic sensitivity of cancer cells to IFN-γ-induced cell death⁶⁶ and the capacity of IFN-γ receptor signaling to stabilize the immunologic synapse between CAR T cells and their targets⁶⁷. Despite the anti-tumor effects of IFN-γ, our data and work from others demonstrate that IFN-γ can be inhibited with preservation of CAR T function^41,68. Beyond its role in efficacy, IFN-γ induces iNOS in macrophages critical for driving or maintaining CRS toxicity^25,41. We previously determined that IFN-γ also induces cytopenias and its blockade can rescue mice from CRS and cytopenias⁶⁸, confirming pilot studies of IFN-γ blockade in patients with severe CRS⁶⁹. Notably, blocking intermediary signaling through IL-6R fails to prevent toxicity in mice. Likewise, we now report that blocking IFN-γ, but not IL-6R or IL-1β, mitigates the suppressive induction of iNOS in imMac. In our model (Supplementary Fig. 9), we posit that an initial burst of IFN-γ secretion from CAR T cells drives iNOS expression in imMac, which then dampens further CAR T cell effector functions and promotes proteomic and metabolomic dysfunction. Therapeutic intervention targeting IFN-γ interrupts this feedback circuit by preventing iNOS induction in imMac during the initial CAR T cell activation phase while preserving iNOS expression in pre-existing M1-like macrophages, thereby protecting the therapeutic efficacy of CAR T cell therapy. Deletion of IFN-γ from CAR T cells, but not systemic IFN-γ blockade, inhibited iNOS induction in macrophages in vivo, suggesting that IFN-γ^−/− CAR T cells may represent the most viable clinical strategy. However, optimization of CAR T therapy for B cell malignancies will likely require combination strategies to overcome additional immunosuppressive mechanisms present in the TME that may still limit efficacy. The clinical impact of IFN-γ-mitigating strategies will likely depend on host-specific factors such as inflammatory status and TME composition, as well as variables related to CAR T product and IFN-γ blockade dosing and timing. Our work supports the evaluation of targeting IFN-γ to improve clinical responses to CAR T therapy, but strategic patient selection should be guided by information derived from transcriptomic, genomic, or inflammatory status, as we have reported^16,20,70,71. Translational efforts targeting IFN-γ to improve clinical outcomes in patients treated with CAR T cells are under clinical investigation (NCT06550141) and hold promise.

Methods

Patient samples

Our research complies with all relevant ethical regulations and was conducted in accordance with IRB-approved protocols at H. Lee Moffitt Cancer Center and Research Institute (protocol number 00000971) or the Roswell Park Comprehensive Cancer Center (protocol number I-57217). All samples were prospectively obtained from patients with relapsed or refractory LBCL who underwent axi-cel treatment at the H. Lee Moffitt Cancer Center or the Roswell Park Comprehensive Cancer Center and provided informed consent to participate in an IRB-approved protocol. Patients were not compensated. Pre-treatment tumor biopsies were obtained within 1 month prior to axi-cel infusion and before lymphodepletion. Serum samples were also collected at indicated timepoints. Patients who achieved sustained remission for at least 3 months or 6 months following axi-cel infusion were classified as complete responders (CR) or durable responders (DR), respectively. Patients who either experienced lymphoma relapse or passed away due to recurrent disease within 6 months of treatment were classified as non-durable responders (NDR) or non-responders (NR). Patient characteristics data corresponding to Fig. 1 have been described⁷⁰. Patient characteristics data for Fig. 8H, I have been provided in the Source Data file. Any remaining patient samples have been stored at Roswell Park Comprehensive Cancer Center or the H. Lee Moffitt Cancer Center and are available to be shared upon request of the corresponding author.

Mice

Animal studies performed at Roswell Park Comprehensive Cancer Center were conducted in accordance with Institutional Animal Care and Use Committee (IACUC)-approved protocol 1482 M. Animal studies performed at the University of South Florida and Moffitt Cancer Center were conducted in accordance with IACUC-approved protocol IS00009823. C57BL/6 J mice, iNOS^−/− (B6.129P2-Nos2^tm1Lau/J; strain # 002609) mice, IFN-γ^−/− (B6.129S7-Ifng^tm1Ts/J; strain # 002287) mice, and Rag1^−/− mice (B6.129S7-Rag1^tm1Mom/J; Strain #002216) were purchased from Jackson Laboratories. Knockout strains are all on the C57BL/6J background. Rag1^−/− and IFN-γ^−/− mice were then bred in-house for experiments. Mice were housed in a specific pathogen-free environment within a barrier facility maintained on a 12-h light/dark cycle at ~70 °F and 39% humidity.

Cell lines

Eμ-myc cells were derived from the axillary lymph node of tumor-bearing Eμ-myc transgenic mice, a spontaneous lymphoma model^72–74. For some experiments, Eμ-myc cells that were retrovirally transduced to express GFP-firefly luciferase (Eμ-myc-GFP-FFL) were used. Eμ-myc cells were maintained on irradiated (30 Gy) NIH-3T3 feeder cells in RPMI-1640/IMDM (1/1, v/v) supplemented with 10% heat-inactivated fetal bovine serum (HI-FBS), 2 mM L-glutamine, 100 U/ml Penicillin/Streptomycin, and 22.5 μM β-mercaptoethanol. Prior to use as feeder cells, NIH/3T3 cells were maintained in DMEM supplemented with 10% HI-FBS, 2 mM L-glutamine, and 100 U/ml Penicillin/Streptomycin. Raji cells were maintained in RPMI1640 supplemented with 10% FBS, 2 mM L-glutamine, and 100 U/ml Penicillin/Streptomycin. Cell lines were routinely tested for the absence of mycoplasma contamination using the Universal Mycoplasma Detection kit (ATCC) or MycoAlert PLUS mycoplasma detection kit (Lonza).

Patient tumor bulk RNA-sequencing

RNA-sequencing was performed as described^20,70. Formalin-fixed paraffin-embedded (FFPE) or snap-frozen samples were obtained and examined by a hematologist for tumor content. RNA was extracted and RNA-sequencing libraries were prepared using NuGen RNA-Seq Multiplex System (Tecan US) according to the manufacturer’s protocols. The libraries were then sequenced on the Illumina NextSeq 500 system with a 75-base paired-end run at 80 to 100 million read pairs per sample. Raw count matrices and transcripts per million (TPM) matrices were generated from the sequencing data. Batch correction was not applied. To determine the immune cell composition in bulk RNA-seq profiles of tumor biopsies, TPM values were used as input for CIBERSORTx v.1.0.41 (https://cibersortx.stanford.edu) with the LM22 signature matrix. For gene set enrichment analysis (GSEA), raw counts were normalized using DESeq2, and enrichment of M1- and M2-associated gene signatures⁷⁵ was performed using the Broad GSEA software (v4.1.0) with default settings and phenotype labels as the permutation type. The Molecular Signatures Database (MSigDB) version 6.2 gene set collections were evaluated in the analysis. Statistical significance was determined by a standard empirical permutation test implemented in the GSEA software.

Patient leukapheresis sample flow cytometry analysis

Cryopreserved cells from apheresis were removed from liquid nitrogen and rapidly thawed in a 37 °C water bath prior to being transferred to 10 ml of pre-warmed complete media to remove excess DMSO. Cells were centrifugated 5 min at 300 × g, the cell pellet was washed twice with PBS and resuspended in 100 µl of a solution containing 1X Live/Dead Fixable NIR stain (Invitrogen, ThermoFisher Scientific) and human Fc-receptor block (BD Biosciences, 564219, 1/100). Cells were incubated for 30 min at room temperature. Surface staining was performed for 30 min at 4 °C with antibody mix in MACS buffer with 0.5% BSA (Miltenyi Biotec). Cells were then fixed using IC Fixation Buffer (eBioscience) for 30 min at RT, washed with 1X Permeabilization Buffer (eBioscience), and intracellular staining was performed for 30 min at 4 °C with antibody mix in 1X Permeabilization Buffer (eBioscience). Samples were analyzed with a Symphony flow cytometer (BD Biosciences) and data analyzed using FlowJo software. The following monoclonal antibodies were obtained from BD Biosciences: anti-CD45 (HI30; 564106, 1/200) and anti-CD11b (ICRF44; 562632, 1/200). From Biolegend: anti-CD14 (HCD14; 325617, 1/200), anti-iNOS (W16030C; 696805, 1:100), anti-CD20 (2H7; 302316, 1/200), and anti-CD56 (MEM-188; 304611, 1:200). From R&D Systems: Anti-CD3e (UCHT1; FAB100G, 1/100). Anti-CD3e, anti-CD20 and anti-CD56 were used to dump non-myeloid cells.

Generation of retroviral constructs

Plasmids encoding 19dz and 1928z CAR constructs in SFG γ-retroviral vectors have been described³⁴. The murine 1928z CAR construct includes anti-murine CD19 scFv (1D3), murine CD8α transmembrane and hinge domains, murine CD28 intracellular domain, and murine CD3ζ intracellular domain, followed by the mCherry reporter via glycine-serine linker. The murine 19dz CAR construct includes the same sequence as 1928z construct except for absence of CD28 intracellular domain and having a truncated CD3ζ intracellular domain. Human 1928z CAR construct includes FMC63 scFv with CD8α transmembrane and hinge domain, followed by human CD28 and CD3ζ intracellular domains, and a GFP reporter linked via glycine-serine linker. For retrovirus production, plasmids were transfected to H29 cell lines using a calcium phosphate transfection kit (Invitrogen) to produce vesicular stomatitis virus G-glycoprotein-pseudotyped retroviral supernatants. These retroviral supernatants were subsequently used to transduce Phoenix-ECO or RD114 cell lines, which stably produce retroviral particles pseudotyped with Moloney murine leukemia virus or feline endogenous virus, respectively.

Mouse T cell isolation and CAR T cell generation

Mouse spleens from C57BL/6 or IFN-γ^−/− mice were excised, mechanically disrupted, and filtered through a 40 μm cell strainer. CD3⁺ T cells were enriched via negative selection using EasySep Mouse T Cell Isolation Kit (STEMCELL Technologies). T cells were activated and expanded with anti-CD3/28 Dynabeads (Gibco) at a bead-to-cell ratio of 0.8:1. T cells were spinoculated (2000 × g, 1 h, room temperature) twice, 24 h and 48 h after initial T cell activation, with viral supernatants collected from Phoenix-ECO cells on retronectin (Takara) coated plates. Following the second spinoculation, T cells were maintained for one day. On day 5, anti-CD3/28 Dynabeads were removed, and CAR T cells were used for in vitro or in vivo experiments. CAR transduction efficiency was determined by flow cytometry as a percentage of mCherry⁺ cells in live cells. Mouse T cells were cultured in RPMI-1640 supplemented with 10% HI-FBS, 2 mM L-glutamine, 100 U/ml Penicillin/Streptomycin, 1× nonessential amino acids, 1 mM sodium pyruvate, 10 mM HEPES, 55 μM 2-mercaptoethanol, and 100 IU/ml recombinant human IL−2.

Human T cell isolation and CAR T cell generation

Human peripheral blood mononuclear cells (PBMCs) were obtained from STEMCELL Technologies (catalog # 70500.1). T cells were enriched via negative selection using the EasySep Human T cell Isolation Kit (STEMCELL Technologies). T cells were activated and expanded with anti-CD3/28 Dynabeads (Gibco) at a bead-to-cell ratio of 0.8:1. Spinoculation was performed twice at 24 h and 48 h post activation (2000 × g, 1 h, room temperature) using viral supernatants collected from RD114 cells on retronectin (Takara) coated plates. Following the second spinoculation, T cells were maintained for one day. On day 5, anti-CD3/28 Dynabeads were removed, and CAR T cells were maintained for two additional days before being used for in vitro experiments on day 8. CAR transduction efficiency was determined by flow cytometry, measuring the percentage of GFP⁺ cells in live cells. Human T cell complete medium consists of RPMI1640, 10% FBS, 2 mM L-glutamine, 100 U/ml Penicillin/Streptomycin, and 100 IU/ml recombinant human IL−2.

Animal experiments

Six- to 10-week-old Rag1^−/− mice or C57BL/6 of both sexes were intraperitoneally (i.p.) injected with 3 × 10⁶ Eμ-myc-GFP-FFL cells to generate tumors localized in the peritoneal cavity. Tumor engraftment was verified by bioluminescence imaging one day before CAR T cell transfer. Mice were randomized to different treatment groups without differences in pre-treatment tumor load. C57BL/6 mice received 150 or 300 mg/kg cyclophosphamide (CTX) intraperitoneally one day before CAR T cell transfer. Seven days after tumor cell inoculation, mice were injected i.p. with 5 × 10⁶ CAR T cells (derived from C57BL/6 or IFN-γ^−/− mice) in 300 μl PBS. For survival experiments, L-NIL or PBS was administered i.p. once per day at 20 mg/kg body weight starting on the same day of tumor cell injection. For experiments evaluating IFN-γ blockade, mice were administered either anti–IFN-γ antibody (clone XMG1.2, BioXCell) or isotype IgG control (clone HRPN, BioXCell) at 250 μg/mouse, i.p., three times per week beginning on the same day as CAR T cell transfer. Experimental endpoints were achieved when mice demonstrated signs of morbidity or hind-limb paralysis, or when solid tumor masses reached the maximal tumor size permitted by the Institutional Animal Care and Use Committee (2000 mm³ for some mice that developed palpable masses). The maximal tumor size was not exceeded. Bioluminescence imaging was performed by IVIS Lumina III In Vivo Imaging System (PerkinElmer) with Living Image software (PerkinElmer). Sex of the mice is reported in the Source Data File.

Mouse peritoneal cell collection

Peritoneal cells were obtained by peritoneal lavage as described²⁵. After euthanizing mice, 5–8 ml ice-cold PBS/2 mM EDTA was i.p. injected. Abdomens were massaged for one minute and subsequently incised to drain the lavage fluid in a collection tube, or fluid was retrieved through the same syringe. Cells were filtered through a 40 μm cell strainer. Following red blood cell lysis with ACK lysing buffer, peritoneal cells were used for analyses. For ex vivo coculture experiments with CAR T cells, EasySep Mouse CD11b Positive Selection Kit II (STEMCELL Technologies) was used to isolate CD11b⁺ myeloid cells.

Murine macrophage development and polarization

BMDMs were generated from bone marrow cells harvested from femurs and tibias of WT or iNOS^−/− mice. Following red blood cell lysis by ACK (Ammonium-Chloride-Potassium) lysis buffer, 1 × 10⁷ bone marrow cells were cultured in 10-cm tissue culture dish in 10 ml of RPMI-1640 supplemented with 10% HI-FBS, 2 mM L-glutamine, 100 U/ml Penicillin/Streptomycin, and 20 ng/ml M-CSF (R&D systems) for 7 days. On day 3, 10 ml of fresh medium with 20 ng/ml M-CSF was added. On day 5, the culture medium was entirely discarded and replaced by 15 ml of fresh medium with 20 ng/ml M-CSF. On day 6, BMDMs were activated for 24 h with 20 ng/ml of IL-4 and IL-10 (Peprotech)^76,77 to develop imMac, with 100 ng/mL LPS (invivoGen) and 20 ng/mL IFNγ (Peprotech) to develop M1-like macrophages, or cultured in media only to use as unMac. M-CSF (20 ng/ml) was added during activation with cytokines. On day 7, adherent cells were harvested by gentle scraping and used for in vitro experiments.

Human macrophage development and polarization

Monocytes were enriched from PBMCs using the EasySep Human Monocyte Enrichment Kit (STEMCELL Technologies) and were seeded in a 96-well plate at the concentration of 2.5 × 10⁵ cells/ml in 200 μl of RPMI-1640 supplemented with 10% HI-FBS, 2 mM L-glutamine, 100 U/ml Penicillin/Streptomycin, and 25 ng/ml M-CSF (R&D systems). On day 3, 100 μl of the culture medium was removed and replaced with 100 μl of fresh medium containing 25 ng/ml M-CSF. On day 5, the culture medium was entirely discarded and replaced with 150 μl of fresh medium with 25 ng/ml M-CSF. On day 7, monocyte-derived macrophages (MDMs) were treated for 24 h with 20 ng/ml of IL-4, IL-10, and IL-13 (Peprotech) to develop imMac, or were cultured in media only as unMac, or in patient serum. M-CSF (25 ng/ml) was added during activation with cytokines or patient serum. On day 8, all the supernatants were removed, and MDMs were left on original plates and briefly washed with PBS before coculturing with CAR T cells and Raji cells.

Expansion assay

The expansion of mCherry⁺ CAR T cells was determined by an IncuCyte S3 live cell analysis system (Essen Bioscience). 2 × 10⁴ Eμ-myc cells and 2 × 10⁴ CAR T cells were cocultured in the absence or presence of 0.5 × 10⁴ macrophages (CAR T:Eμ-myc cell:Macrophage=1:1:0.25, unless otherwise indicated in the figure legends) in a 96-well black-walled clear bottom plate in 120 μl of media. Cell images were captured at 4X magnification. The expansion index was calculated by dividing the total integrated red intensity (RCU × μm²/mm²) at each time point by the first time point. The expansion of cells was monitored for 48 h after starting the coculture assays, unless otherwise noted.

The expansion of human GFP⁺ CAR T cells was determined by an IncuCyte S3 live cell analysis system (Essen Bioscience). 2 × 10⁴ Raji cells and 2 × 10⁴ CAR T cells were cocultured in the presence of macrophages from the same donor in a 96-well black-walled clear bottom plate in 120 μl of media. Cell images were captured at 4X magnification. The expansion index was calculated by dividing the total integrated green intensity (GCU × μm²/mm²) at each time point by the first time point.

Griess assay

2 × 10⁴ Eμ-myc cells and 2 × 10⁴ CAR T cells were cocultured in the presence or absence of 0.5 × 10⁴ macrophages in a 96-well plate in 120 μl of media. Coculture supernatants were harvested, and nitric oxide levels were measured using Griess reagent system (Promega) according to manufacturer’s instructions. Absorbance was read at 560 nm using a microplate reader (GloMax, Promega), and NO₂^– concentrations were determined by standard curve. Standard curve was prepared by diluting 0.1 M sodium nitrite standard (provided in the kit) with the culture media used for experiments.

BrdU incorporation assay

2 × 10⁵ Eμ-myc cells and 2 × 10⁵ CAR T cells were cocultured in the absence or presence of 0.5 × 10⁵ macrophages in a 24-well plate in 1200 μl of media. At 24 h of coculture, BrdU was added to each well at 10 μM. After an additional incubation for 18 h, cells were harvested. BrdU staining was performed according to APC BrdU flow kit (BD Pharmingen) and BrdU incorporation was analyzed by flow cytometry.

Flow cytometry

The following fluorophore-conjugated anti-mouse antibodies were used. From BD Horizon: anti-CD45 (30-F11; 749889, 1/400 or 1/200), anti-CD19 (1D3; 612781, 1/400), anti-CD11b (M1/70; 563553, 1/400 or 1/200) and anti-CD3ε (145−2C11; 564378, 1/100). From BioLegend: anti-CD3 (17A2; 100276, 1/200), anti-CD3ε (145−2C11; 100351, 1/50 or 1/100, and 100357, 1/100), anti-CD45 (17A2; 100217, 1/100), anti-CD115 (AFS98; 135510, 1/50), anti-NOS2 (intracellular, W16030C, 696806, 1:36), anti-PD-L1 (10 F.9G2; 124319 or 124331, 1/200), Ly6C (HK1.4; 128041, 1/100), Ly6G (1A8; 127615, 1/100), and anti-F4/80 (BM8; 123115 and 123137, 1/200 and 123131, 1/40 and 123147, 1/200). From eBioscience: anti-mouse ARG-1 (A1exF5; 48-3697-82, 1/50 or 1/100 and 17-3697-82, 1/100) and anti-NOS2 (CXNFT; 25-5920-82, 1/400). From Waters Biosciences, anti-CD11b (M1/70; 612800, 1/50), anti-CD206 (intracellular, Y17-505, 568818, 1/36). Fc receptors were blocked using FcR Blocking Reagent (anti-mouse CD16/CD32 antibody, eBioscience, 14-0161-82, 1/100). DAPI (BD Pharmingen) and Zombie NIR Fixable Viability Kit (BioLegend, 423105, 1/500 or 1/800 or 1/1000) were used as viability dyes. For intracellular staining, surface-labeled cells were fixed and permeabilized with Cytofix/Cytoperm kit (BD Biosciences) according to the manufacturer’s instructions and then stained with intracellular antibodies. For cell surface CAR staining, protein L-biotin conjugate followed by Alexa488-conjugated streptavidin (Invitrogen, S32354, 1/100) was used. Flow cytometry was performed on a LSR II, FACSymphony instrument (BD Biosciences), or Cytek Aurora (Cytek Biosciences). Data were analyzed with the FlowJo software (FlowJo LLC). Gating strategies can be found in Supplementary Figs. 10–12.

Coculture assays and CAR T cell isolation for subsequent downstream assays

CAR T cells (5 × 10⁵ cells/ml) and Eμ-myc-GFP-FFL cells (5 × 10⁵ cells/ml) were cocultured with or without macrophages (1.25 × 10⁵ cells/ml) for 48 h. For these coculture experiments, T cells were isolated from multiple mice, pooled, transduced in separate wells, and then pooled after transduction to generate CAR T cells. Bone marrow cells were isolated from multiple mice, pooled, plated in separate dishes, and differentiated with cytokines to generate macrophage populations. After differentiation, each macrophage population was pooled separately and used for coculture with CAR T cells and tumor cells. Coculture replicates refer to independent coculture wells or dishes established from these pooled CAR T cell and macrophage preparations. Cocultures were prepared in either a 6-well plate with 2 ml of media per cell type (CAR T cells, Eμ-myc cells, and macrophages), totaling 6 ml, or in a 10-cm dish with 12 ml of media per cell type, totaling 36 ml. In certain experiments, neutralizing antibodies were used at a final concentration of 1ug/mL (anti-IL-6R, anti-IFN-γ, and anti-IL-1β; Thermofisher). For use in subsequent downstream assays, after initial coculture, cells were harvested, and T cells were isolated using Mouse T Cell Isolation Kit (STEMCELL Technologies). T cell purity was 100% as tested by flow cytometry. Percentage of CAR-expressing T cells was determined with flow cytometry.

Luciferase-based killing assay

2 × 10⁴ Eμ-myc-GFP-FFL cells were cocultured with CAR T cells at different effector-to-target ratios in a 96-well white-walled plate in 100 μl of media. Following incubation, 100 μl luciferase substrate reagent (ONE-Glo Luciferase assay system, Promega) was added to each well. Target cells alone were plated at the same cell density to determine maximum luciferase signals. Emitted luminescence was detected in the microplate reader (GloMax, Promega). Percent lysis was determined as (1- sample signal/maximum signal)×100.

Cytokine secretion assay

2 × 10⁴ Eμ-myc-GFP-FFL cells were cocultured with 2 × 10⁴ CAR T cells in a 96-well plate in a total volume of 100 μl of media. Supernatants were collected and analyzed for IFN-γ and TNF-α secretion using Ella automated immunoassay system (Proteinsimple Bio-techne) according to manufacturer’s instructions.

Immunoblotting

T cells were lysed in ELB lysis buffer (50 mM HEPES, pH 7.5, 250 mM NaCl, 5 mM EDTA, 0.5 mM DTT, 0.1% NP-40 alternative, 10 µg/ml aprotinin, 10 µg/ml leupeptin, and 100 µg/ml trypsin/chymotrypsin inhibitor). Following protein quantification with the Pierce BCA protein assay (ThermoFisher), the samples were mixed with a loading buffer containing 2-mercaptoethanol. The proteins were electrophoresed in 4–20% Tris-Glycine gels (Novex-Invitrogen) and transferred to PVDF membrane with a Bio-Rad Trans-Blot SD Semi-Dry Transfer Cell. The membrane was blocked in 5% bovine serum albumin (BSA) in TBST and subsequently blotted with primary and secondary antibodies in 5% BSA in TBST. The following antibodies were used: IDH1 (clone D2H1; Cell signaling, #8137S; 1/1000), IDH2 (clone D2E3B; Cell signaling, #56439S; 1/1000), IRG1 (clone E5B2G; Cell signaling, #19857S; 1/1000), β-actin (clone AC-74, Sigma-Aldrich, # A2228; 1/100000), and horseradish peroxidase-conjugated secondary antibodies (Donkey-anti-Rabbit, Cytiva, #NA934-1ML; 1/5000); Sheep-anti-Mouse, Cytiva, #NA931-1ML; 1/5000). Membranes were imaged with a ChemiDoc Imaging System (BioRad, #17001401) and exported through ImageLab (Bio-Rad #12012931).

Metabolomics and ¹³C₆-labeled glucose tracing analyses

For global metabolomics analysis of cell-cultured medium, the cell-free medium was obtained by performing rapid centrifugation (17,000 × g, 10 sec, room temperature) to collect the supernatant. The metabolites present in 20 μl of the cell-cultured medium were then extracted using 80 μl of ice-cold MeOH. Following a 30 min incubation on ice and subsequent centrifugation (17,000 × g, 20 min, 4 °C), the supernatant was subjected to LC-HRMS analysis.

For global metabolomic profiling and ¹³C₆-labeled glucose tracing of CAR T cells, 1 × 10⁶T cells were resuspended in either RPMI-1640 medium (RPMI + 10% heat-inactivated dialyzed FBS) or ¹³C₆-glucose substituted RPMI-1640 medium (glucose-free RPMI + 10% heat-inactivated dialyzed FBS + 11.1 mM ¹³C₆-glucose). After 4 h incubation, cells were collected, rapidly centrifuged (17,000 × g, 10 sec, room temperature), and medium was removed. T cells were washed with 1 ml of ice-cold PBS, and metabolites were extracted with 300 μl of 80% methanol via incubation at −80 °C for 15 min. Samples were centrifuged (17,000 × g, 20 min, 4 °C), and supernatants were transferred to an Eppendorf tube and dried in a vacuum evaporator overnight. The dried extracts were resuspended in 20 μl of aqueous 50% methanol, clarified by centrifugation (17,000 × g, 20 min, room temperature), and analyzed by LC-HRMS.

LC-HRMS analysis was performed on a Vanquish UPLC coupled with a Q-Exactive HF mass spectrometer, employing the same conditions as the previously established methods⁷⁸. A ZIC-pHILIC LC column (4.6 mm inner diameter × 150 mm length, 5 μm particle size, MilliporeSigma, Burlington, MA) with a ZIC-pHILIC guard column (4.6 mm inner diameter × 20 mm length, MilliporeSigma, Burlington, MA) was used for chromatographic separation at a column temperature of 30 °C. The mobile phases consisted of 10 mM (NH₄)₂CO₃ and 0.05% NH₄OH in H₂O for mobile phase A, and 100% acetonitrile (ACN) for mobile phase B. The LC gradient conditions were as follows: 0 to 13 min: a decreasing of 80% to 20% of mobile phase B, 13 to 15 min: 20% of mobile phase B. The ionization was set to negative mode, with the MS scan range set to 60 to 1000 m/z. The mass resolution was 70,000, and the AGC target was 1 × 10⁶. The sample loading volume was 5 μl. The unlabeled or ¹³C-labeled metabolite peaks were extracted using EL-Maven with a metabolite standard-based in-house library. For global metabolomic profiling, peak areas of metabolites were normalized by the median value of the total for identified metabolite peak areas in each sample. For the ¹³C-labeled metabolite peaks, the natural isotope peak area was corrected using IsoCor (Version 2.2)⁷⁹.

Proteomics and phospho-peptide analyses

Proteomics and phospho-proteomics were performed using protocols adapted from previously described methods^80,81, with study-specific sample preparation, TMT channel allocation, fractionation, phospho-peptide enrichment, LC–MS/MS acquisition, and data processing parameters described below.

For each sample, 1.5 × 10⁷T cells were lysed under denaturing conditions in a buffer composed of 8 M urea, 20 mM HEPES (pH 8), 1 mM sodium orthovanadate, 2.5 mM sodium pyrophosphate, and 1 mM β-glycerophosphate. Protein concentrations were measured using a Bradford assay, and 200 μg of total protein was processed per sample. Proteins were reduced with 4.5 mM DTT at 60 °C for 30 min, followed by alkylation of cysteine residues with 10 mM iodoacetamide. Trypsin digestion was carried out overnight at room temperature with an enzyme-to-substrate ratio of 1:20. The resulting tryptic peptides were acidified with 1% trifluoroacetic acid (TFA) and desalted using C18 Sep-Pak cartridges according to the manufacturer’s instruction. Peptides from individual samples were labeled with TMTPro18plex reagent. Labeling efficiency was evaluated by LC-MS/MS and spectral counting, and all channels achieved at least 95% label incorporation. The 18 samples were combined and dried by lyophilization. Legacy labels M0 and M2 in the original proteomics acquisition/sample annotation files correspond to unMac and imMac, respectively, as defined in the manuscript. The TMT channel layouts were as follows: No Macrophages (126 C/127 C/128 C), No Macrophages+L-NIL (127 N/128 N/129 N), Post coculture with unMac (129 C/130 C/131 C), Post coculture with unMac+L-NIL (130 N/131 N/132 N), Post coculture with imMac (132 C/133 C/134 C), and Post coculture with imMac+L-NIL (133 N/134 N/135 N).

After lyophilization, the peptides were re-dissolved in 400 μl of 20 mM ammonium formate (pH 10.0). High-pH reversed-phase separation was performed on a XBridge column, 4.6 mm × 100 mm, packed with BEH C18 resin, 3.5 µm particle size and 130 Å pore size (Waters). Peptides were eluted as follows: 5% solvent B, consisting of 5 mM ammonium formate and 90% acetonitrile (pH 10.0) for 10 min, 5–15% solvent B over 5 min, 15–40% solvent B over 47 min, 40–100% solvent B over 5 min, 100% solvent B for 10 min, and re-equilibration at 1% solvent B. The flow rate was 0.6 ml/min. Fractions were collected, concatenated into 24 fractions for global proteomics and 12 fractions for phospho-proteomics, and dried by vacuum centrifugation.

Following lyophilization, the peptides were re-dissolved in IMAC loading buffer containing 0.1% TFA and 85% acetonitrile. Phospho-peptides in each fraction were enriched using IMAC resin (Cell Signaling Technology, #20432) on a KingFisher robot (Thermo Fisher Scientific). The IMAC resin was first equilibrated with loading buffer. Peptides were incubated with the IMAC resin for 30 min at room temperature with gentle agitation, using 10 μl IMAC resin per sample. After incubation, the IMAC resin was washed twice with loading buffer and once with wash buffer containing 80% acetonitrile and 0.1% TFA. Bound phospho-peptides were eluted with elution buffer containing 50% acetonitrile and 2.5% ammonia, and the eluate volume was reduced to 20 µl by vacuum centrifugation.

LC-MS/MS analyses were performed using a nanoflow ultra-high performance liquid chromatograph (RSLC, Dionex, Sunnyvale, CA) coupled to an electrospray benchtop Orbitrap mass spectrometer (Orbitrap Exploris 480 with FAIMS, Thermo Fisher Scientific, San Jose, CA). Peptide samples were first loaded onto a pre-column, 2 cm × 100 µm inner diameter, packed with C18 reversed-phase resin, 5 µm particle size and 100 Å pore size, and washed for 8 min with aqueous 2% acetonitrile and 0.04% TFA. The trapped peptides were then separated on an analytical column, 25 cm × 75 µm inner diameter, packed with C18 resin, 2 µm particle size and 100 Å pore size (Dionex, Sunnyvale, CA). The 120-min gradient was programmed as follows: 95% solvent A, consisting of 2% acetonitrile and 0.1% formic acid, for 8 min; solvent B, consisting of 90% acetonitrile and 0.1% formic acid, from 5% to 38.5% over 90 min; solvent B from 50% to 90% B over 7 min; 90% solvent B for 5 min; solvent B from 90% to 5% over 1 min; and re-equilibrate for 10 min. The flow rate on analytical column was 300 nl/min. Data-dependent acquisition was performed using two FAIMS compensation voltage values, −45 and −65, with 1.5 s cycle time for each compensation voltage. The spray voltage was 2100 V, and the capillary temperature was 300 °C. MS and MS/MS scans were acquired at resolutions of 120,000 and 45,000, respectively. Dynamic exclusion was 15 s for previously sampled peptide peaks.

Raw files were analyzed with MaxQuant⁸² (version 1.6.14.0) for peptide identification and TMT reporter ion intensity quantification. Searches were performed against the mouse UniProt protein database downloaded in March 2023. Mass tolerances were set to 20 ppm for the first search and 4.5 ppm for the main search. The reporter ion mass tolerance was 0.003 Da, and the minimum precursor intensity fraction was 0.75. A maximum of two missed trypsin cleavages was allowed. Carbamidomethylation of cysteine was set as a fixed modification, whereas methionine oxidation and phosphorylation of serine, threonine, and tyrosine were specified as variable modifications. Peptide spectral match (PSM) and protein false discovery rate (FDR) were set to 0.05. The “Match between Runs” feature was enabled. MaxQuant-normalized and log-transformed data were used for downstream bioinformatics analyses. Principal component analysis (PCA) was performed using PCAtools⁸³ (v2.14.0) to evaluate the overall data quality and sample variability in the proteomics and phosphopeptide datasets.

Differentially expressed proteins were identified using the limma R package (v3.58.1)⁸⁴. The limma empirical Bayes framework was used to stabilize individual protein variance estimates. Proteins were considered significantly altered if they showed a log2 fold change greater than 1 and a Benjamini-Hochberg-adjusted p-value (FDR) < 0.05. Differential phospho-peptide analysis was performed using a linear mixed-effects model implemented with the lme4 (v1.1-35.2)⁸⁵ and lmerTest (v3.1-3)⁸⁶ R packages. In this model, processing run was modeled as a random effect and experimental group was modeled as a fixed effect. Pathway enrichment analysis was performed using the preranked Gene Set Enrichment Analysis (GSEA) algorithm implemented in the clusterProfiler R package (v4.10.1)⁸⁷. Genes were ranked using the moderated t-statistic from the limma differential expression analysis. MSigDB gene sets, including H, C2CP and C5, were obtained using the msigdbr R package (v7.5.1). Pathways with an FDR < 0.05 were considered significantly enriched. All bioinformatics analyses were performed in R (v4.3.3).

Seahorse assay

ECAR and OCR were measured using a Seahorse Extracellular Flux Analyzer (Agilent Technologies). XF96 microplates were coated with CellTak a day before analyses. To assay glycolytic function, T cells were resuspended in glucose-free XF medium supplemented with 2 mM L-glutamine and 1 mM sodium pyruvate and seeded at 2 × 10⁵ cells in 180 μl per well. Following incubation in a CO₂-free incubator for 60 min at 37 °C for pH stabilization, ECAR was measured in response to 10 mM glucose, 1 μM oligomycin, and 50 mM 2-deoxyglucose. To assay mitochondrial function, T cells were resuspended in XF medium supplemented with 2 mM L-glutamine,1 mM sodium pyruvate, and 10 mM glucose and seeded at 2 × 10⁵ cells in 180 μl per well. Following incubation in a CO₂-free incubator for 60 min at 37 °C for pH stabilization, OCR was measured in response to 1 μM oligomycin, 1 μM FCCP, and 0.5 μM rotenone and antimycin A.

Single cell RNA-sequencing analysis

T cells were isolated using Mouse T Cell Isolation Kit (STEMCELL Technologies) and macrophages were isolated using Mouse F4/80 positive selection kit (STEMCELL).

Single-cell gene expression libraries were prepared using the 10X Genomics platform with Chromium Next GEM Single Cell 3’ Kit v3.1. Before loading, cell suspensions were evaluated using ViaStain AOPI and a Cellometer K2 automated cell counter (Nexcelom) to measure cell concentration and viability and to confirm that samples were free of clumps and debris that could compromise single-cell capture. Cells were then loaded onto the Chromium X Controller (10X Genomics) for partitioning into nanoliter-scale Gel Beads-in-emulsion, enabling cell-specific barcoding. After reverse transcription, cDNA was amplified and used to construct gene expression libraries through enzymatic fragmentation, end repair, A-tailing, adapter ligation, and PCR amplification to introduce Illumina-compatible sequencing adapters. Library size distribution and quality were assessed using D1000 ScreenTape on a TapeStation 4200 (Agilent Technologies). Libraries were quantified using the Kapa qPCR quantitation kit for Illumina platforms (Roche), pooled, denatured, and diluted to 300 pM with 1% PhiX control library added. The final library pool was loaded into the appropriate NovaSeq Reagent cartridge and sequenced on a NovaSeq 6000 according to the manufacturer’s instruction (Illumina Inc.). A minimum of 20,000 reads per cell were generated for downstream bioinformatics analysis, which was performed using Cell Ranger (v7.0.0).

Count matrices were generated using Cell Ranger⁸⁸ and analyzed with the Seurat R package⁸⁹. To remove low-quality cells, potential doublets, and dying cells, cells with fewer than 1000 or more than 7500 detected features, or with >15% mitochondrial reads, were excluded. UMI counts were normalized and scaled using the SCTransform method, with the percentage of mitochondrial reads regressed out during normalization. Dimensionality reduction, including PCA, UMAP, and t-SNE, was performed using highly variable genes. Shared nearest neighbor (SNN)-based clustering was performed using the first 30 principal components. Immune cell types were annotated using the SingleR⁹⁰ package with the ImmGen reference dataset. Pathway scores were calculated using the AddModuleScore function in Seurat.

Differentially expressed genes (DEGs) were identified using the Wilcoxon rank-sum test implemented in the FindMarkers function in Seurat, and P values were adjusted by Bonferroni correction for multiple testing. GSEA was performed using the fgsea R package to calculate normalized enrichment scores (NES). For selected gene signatures, pathway scores were computed using the AddModuleScore function in the Seurat package. Pearson correlation analysis was used to assess the correlation between selected gene expression levels and pathway scores.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Supplementary Information^{(2.8MB, pdf)}

Reporting Summary^{(119.4KB, pdf)}

Transparent Peer Review file^{(1MB, pdf)}

Source data

Source Data^{(6.4MB, xlsx)}

Acknowledgements

This work was in part supported by the following–FACCA Grant (M.L.D.), Moffitt Clinical Science Award (M.D.J.), NIH/NCI Grant R01CA244328 (F.L.L.), NIH Grant R01HL167232 (M.L.D.), The Rustum Family Endowed Chair in Translational Research (M.L.D.), Leukemia and Lymphoma Society Clinical Scholar Award (F.L.L.), Hawkins Family Endowed Chair (F.L.L.), generous donations from the Hyer Family Foundation and the Thiel Family. This work has also been supported in part by Total Cancer Care, Tissue Core, Molecular Genomics Core, Biostatistics and Bioinformatics Core, Flow Cytometry Core, Analytic Microscopy Core, Proteomics and Metabolomics Core, Small Animal Imaging Laboratory Core, and Advanced Analytical and Digital Pathology Core at the H. Lee Moffitt Cancer Center & Research Institute, an NCI designated Comprehensive Cancer Center (P30-CA076292).

Author contributions

Conceptualization, S.B.L., Y.P.K., P.C.R., M.L.S. and M.L.D.; methodology-investigation, S.B.L., Y.P.K., J.K.M., J.C.B., E.R., H.K., D.C.C., R.V.J., P.G., A.P., N.B., M.I., J.V.M., C.J.S., S.L.M., R.A.S., M.L., J.M.K, J.N., M.H., T.I.S, X.W., J.W., J.A. and K.R.; patient samples and data, S.H., M.J.T., M.D.J., R.F., F.L.L. and M.L.D.; funding acquisition, M.L.D., P.C.R, and Y.P.K; supervision, M.L.D. and P.C.R.; and all authors contributed to the writing of the manuscript.

Peer review

Peer review information

Nature Communications thanks Saar Gill and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

This work has been supported by funds from Roswell Park Comprehensive Cancer Center and National Cancer Institute (NCI) grant, P30CA016056, the Roswell Park Alliance Foundation, Moffitt Cancer Center, and Seoul National University as well as the following grants/contracts: National Institutes of Health grants R01HL167232-01 to M.L.D; R01-CA184185, R01-CA233512, R01-CA262121, P01-CA250984 Project no. 4, and P30-CA076292 to P.C.R; Florida Department of Health grant no. 20B04 to P.C.R; Kite Pharma/Gilead to R.F., M.D.J., F.L.L; Creative-Pioneering Researchers Program through Seoul National University (370-2025011) and National Research Foundation of Korea (RS-2022-NR071704) to Y.P.K.

Data availability

The RNA-seq data generated in this study have been deposited in the Gene Expression Omnibus database under accession code GSE338087. The scRNA-seq data generated in this study have been deposited in the Gene Expression Omnibus database under accession code GSE293878. The mass spectrometry proteomics expression data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD079073. The phospho-proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD079082. The conditioned media metabolomics data in this study have been deposited in the MetaboLights database under the accession code MTBLS14634. The cellular metabolites metabolomics data in this study have been deposited in the MetaboLights database under the accession code MTBLS14640. The glucose isotope tracer data in this study have been deposited in the MetaboLights database under the accession code MTBLS14641. The remaining data are available within the Article, Supplementary Information or Source Data file. Source data are provided with this paper.

Competing interests

M.L.D. has received research funding from Kite/Gilead. M.L.D. has also received research funding from Novartis, CRISPR, and Bristol Myers Squibb (BMS) that are unrelated to this work. M.L.D. has received consulting fees or honoraria from Adicet, Kite, BMS, A2 Biotherapeutics, Legend, Janssen (J&J), Synthekine, Bellicum, Capstan, and CARGO. M.L.D. holds stock or stock options in Adaptive Biotechnologies, A2 Biotherapeutics, and Adicet. M.L.D. has intellectual property relevant to this work and has licensed CAR technology to CRISPR and Atara. M.H. has received consultancy fees, speaker’s bureau honoraria, and/or other compensation from Adaptive Biotechnologies, Amgen, Aptitude Health, Blueprint Oncology, Celgene, Decibio, Diaceutics, Guidepoint, Seattle Genetics, Stemline, Tegus, Janssen, and BMS. J.M.K. has been funded by a sponsored research agreement with Bristol Myers Squibb unrelated to this project. M.D.J. has served in consultancy or advisory roles for Kite/Gilead, Myeloid Therapeutics, and Allogene. He has received research funding from Kite/Gilead and the Florida Academic Cancer Center Alliance (FACCA). M.D.J. has also received research funding from Loxo@Lilly, Incyte, the Mark Foundation, and the Bankhead-Coley Cancer Research Program unrelated to this project. F.L.L. has research funding from Kite Pharma. He also has financial and professional relationships with following organizations that are unrelated to this work: consulting or scientific advisory roles with A2, Allogene, Amgen, Bluebird Bio, BMS/Celgene, Calibr, Caribou, Cellular Biomedicine Group, Cowen, Daiichi Sankyo, EcoR1, Emerging Therapy Solutions, GammaDelta Therapeutics, Gerson Lehrman Group (GLG), Iovance, Kite Pharma, Janssen, Legend Biotech, Novartis, Sana, Takeda, Wugen, Umoja, and Pfizer; research contracts or grants to his institution from Kite Pharma, Allogene, CERo Therapeutics, Novartis, Bluebird Bio, 2SeventyBio, BMS, the National Cancer Institute (PI), and the Leukemia and Lymphoma Society (PI); patents and other intellectual property (unlicensed) in the field of cellular immunotherapy; and educational or editorial activities with Aptitude Health, ASH, BioPharma Communications CARE Education, Clinical Care Options Oncology, Imedex, and the Society for Immunotherapy of Cancer. R.F. has received research funding from Kite/Gilead. The remaining authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41467-026-76972-9.

References

1.Maude, S. L. et al. Chimeric antigen receptor T cells for sustained remissions in leukemia. N. Engl. J. Med. 371, 1507–1517 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Neelapu, S. S. et al. Axicabtagene ciloleucel CAR T-cell therapy in refractory large B-cell lymphoma. N. Engl. J. Med. 377, 2531–2544 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Schuster, S. J. et al. Tisagenlecleucel in adult relapsed or refractory diffuse large B-cell lymphoma. N. Engl. J. Med. 380, 45–56 (2019). [DOI] [PubMed] [Google Scholar]
4.Park, J. H. et al. Long-term follow-up of CD19 CAR therapy in acute lymphoblastic leukemia. N. Engl. J. Med. 378, 449–459 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Maude, S. L. et al. Tisagenlecleucel in children and young adults with B-cell lymphoblastic leukemia. N. Engl. J. Med. 378, 439–448 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Locke, F. L. et al. Phase 1 results of ZUMA-1: a multicenter study of KTE-C19 anti-CD19 CAR T cell therapy in refractory aggressive lymphoma. Mol. Ther.25, 285–295 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Locke, F. L. et al. Axicabtagene ciloleucel as second-line therapy for large B-cell lymphoma. N. Engl. J. Med. 386, 640–654 (2022). [DOI] [PubMed] [Google Scholar]
8.Majzner, R. G. & Mackall, C. L. Clinical lessons learned from the first leg of the CAR T cell journey. Nat. Med. 25, 1341–1355 (2019). [DOI] [PubMed] [Google Scholar]
9.Locke, F. L. et al. Long-term safety and activity of axicabtagene ciloleucel in refractory large B-cell lymphoma (ZUMA-1): a single-arm, multicentre, phase 1-2 trial. Lancet Oncol.20, 31–42 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Cappell, K. M. & Kochenderfer, J. N. Long-term outcomes following CAR T cell therapy: what we know so far. Nat Rev Clin Oncol, 10.1038/s41571-023-00754-1 (2023). [DOI] [PMC free article] [PubMed]
11.Roselli, E. Faramand, R. & Davila, M. L. Insight into next-generation CAR therapeutics: designing CAR T cells to improve clinical outcomes. J.Clin.Invest. 10.1172/JCI142030 (2021). [DOI] [PMC free article] [PubMed]
12.Locke, F. L. et al. Tumor burden, inflammation, and product attributes determine outcomes of axicabtagene ciloleucel in large B-cell lymphoma. Blood Adv.4, 4898–4911 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Dean, E. A. et al. High metabolic tumor volume is associated with decreased efficacy of axicabtagene ciloleucel in large B-cell lymphoma. Blood Adv.4, 3268–3276 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Majzner, R. G. et al. Tuning the antigen density requirement for CAR T-cell activity. Cancer Discov.10, 702–723 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Plaks, V. et al. CD19 target evasion as a mechanism of relapse in large B-cell lymphoma treated with axicabtagene ciloleucel. Blood138, 1081–1085 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Jain, M. D. et al. Whole-genome sequencing reveals complex genomic features underlying anti-CD19 CAR T-cell treatment failures in lymphoma. Blood140, 491–503 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Shouval, R. et al. Impact of TP53 genomic alterations in large B-cell lymphoma treated with CD19-chimeric antigen receptor T-cell therapy. J. Clin. Oncol.40, 369–381 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Deng, Q. et al. Characteristics of anti-CD19 CAR T cell infusion products associated with efficacy and toxicity in patients with large B cell lymphomas. Nat. Med. 26, 1878–1887 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Jackson, Z. et al. Sequential single-cell transcriptional and protein marker profiling reveals TIGIT as a marker of CD19 CAR-T cell dysfunction in patients with non-Hodgkin lymphoma. Cancer Discov.12, 1886–1903 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Jain, M. D. et al. Tumor interferon signaling and suppressive myeloid cells are associated with CAR T-cell failure in large B-cell lymphoma. Blood137, 2621–2633 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Scholler, N. et al. Tumor immune contexture is a determinant of anti-CD19 CAR T cell efficacy in large B cell lymphoma. Nat. Med28, 1872–1882 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Scott, D. W. & Gascoyne, R. D. The tumour microenvironment in B cell lymphomas. Nat. Rev. Cancer14, 517–534 (2014). [DOI] [PubMed] [Google Scholar]
23.Steen, C. B. et al. The landscape of tumor cell states and ecosystems in diffuse large B cell lymphoma. Cancer Cell39, 1422–1437 e1410 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Mantovani, A., Allavena, P., Marchesi, F. & Garlanda, C. Macrophages as tools and targets in cancer therapy. Nat. Rev. Drug Discov.21, 799–820 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Giavridis, T. et al. CAR T cell-induced cytokine release syndrome is mediated by macrophages and abated by IL-1 blockade. Nat. Med24, 731–738 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Morris, E. C., Neelapu, S. S., Giavridis, T. & Sadelain, M. Cytokine release syndrome and associated neurotoxicity in cancer immunotherapy. Nat. Rev. Immunol.22, 85–96 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Norelli, M. et al. Monocyte-derived IL-1 and IL-6 are differentially required for cytokine-release syndrome and neurotoxicity due to CAR T cells. Nat. Med. 24, 739–748 (2018). [DOI] [PubMed] [Google Scholar]
28.Kloosterman, D. J. & Akkari, L. Macrophages at the interface of the co-evolving cancer ecosystem. Cell186, 1627–1651 (2023). [DOI] [PubMed] [Google Scholar]
29.Christofides, A. et al. The complex role of tumor-infiltrating macrophages. Nat. Immunol.23, 1148–1156 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Munn, D. H. et al. Inhibition of T cell proliferation by macrophage tryptophan catabolism. J. Exp. Med. 189, 1363–1372 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Rodriguez, P. C. et al. Arginase I production in the tumor microenvironment by mature myeloid cells inhibits T-cell receptor expression and antigen-specific T-cell responses. Cancer Res. 64, 5839–5849 (2004). [DOI] [PubMed] [Google Scholar]
32.Newman, A. M. et al. Robust enumeration of cell subsets from tissue expression profiles. Nat. Methods12, 453–457 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Newman, A. M. et al. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat. Biotechnol.37, 773–782 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Boucher, J. C. et al. CD28 costimulatory domain-targeted mutations enhance chimeric antigen receptor T-cell function. Cancer Immunol. Res. 9, 62–74 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Bronte, V. & Zanovello, P. Regulation of immune responses by L-arginine metabolism. Nat. Rev. Immunol.5, 641–654 (2005). [DOI] [PubMed] [Google Scholar]
36.Tenu, J. P. et al. Effects of the new arginase inhibitor N(omega)-hydroxy-nor-L-arginine on NO synthase activity in murine macrophages. Nitric Oxide3, 427–438 (1999). [DOI] [PubMed] [Google Scholar]
37.Rodriguez, P. C. & Ochoa, A. C. Arginine regulation by myeloid derived suppressor cells and tolerance in cancer: mechanisms and therapeutic perspectives. Immunol. Rev.222, 180–191 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Moore, W. M. et al. L-N6-(1-iminoethyl)lysine: a selective inhibitor of inducible nitric oxide synthase. J. Med Chem.37, 3886–3888 (1994). [DOI] [PubMed] [Google Scholar]
39.Harris, I. S. & DeNicola, G. M. The complex interplay between antioxidants and ROS in cancer. Trends Cell Biol.30, 440–451 (2020). [DOI] [PubMed] [Google Scholar]
40.Alizadeh, D. et al. IFNgamma is critical for CAR T cell-mediated myeloid activation and induction of endogenous immunity. Cancer Discov.11, 2248–2265 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Bailey, S. R. et al. Blockade or deletion of IFNgamma reduces macrophage activation without compromising CAR T-cell function in hematologic malignancies. Blood Cancer Discov.3, 136–153 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Srivastava, S. et al. Immunogenic chemotherapy enhances recruitment of CAR-T cells to lung tumors and improves antitumor efficacy when combined with checkpoint blockade. Cancer Cell39, 193–208 e110 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Takeda, N. et al. Differential activation and antagonistic function of HIF-alpha isoforms in macrophages are essential for NO homeostasis. Genes Dev.24, 491–501 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Keith, B., Johnson, R. S. & Simon, M. C. HIF1alpha and HIF2alpha: sibling rivalry in hypoxic tumour growth and progression. Nat. Rev. Cancer12, 9–22 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Hodeify, R., Tarcsafalvi, A., Megyesi, J., Safirstein, R. L. & Price, P. M. Cdk2-dependent phosphorylation of p21 regulates the role of Cdk2 in cisplatin cytotoxicity. Am. J. Physiol. Ren. Physiol.300, F1171–F1179 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Marchingo, J. M. & Cantrell, D. A. Protein synthesis, degradation, and energy metabolism in T cell immunity. Cell Mol. Immunol.19, 303–315 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Bantug, G. R., Galluzzi, L., Kroemer, G. & Hess, C. The spectrum of T cell metabolism in health and disease. Nat. Rev. Immunol.18, 19–34 (2018). [DOI] [PubMed] [Google Scholar]
48.Reina-Campos, M., Scharping, N. E. & Goldrath, A. W. CD8(+) T cell metabolism in infection and cancer. Nat. Rev. Immunol.21, 718–738 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Zhao, H. et al. Myeloid-derived itaconate suppresses cytotoxic CD8(+) T cells and promotes tumour growth. Nat. Metab.4, 1660–1673 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Mills, E. L. et al. Itaconate is an anti-inflammatory metabolite that activates Nrf2 via alkylation of KEAP1. Nature556, 113–117 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Locke, F. L. et al. Impact of tumor microenvironment on efficacy of anti-CD19 CAR T cell therapy or chemotherapy and transplant in large B cell lymphoma. Nat. Med. 30, 507–518 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Chen, S. et al. Macrophages in immunoregulation and therapeutics. Signal Transduct. Target Ther.8, 207 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Chen, C. N., Hsieh, F. J., Cheng, Y. M., Chang, K. J. & Lee, P. H. Expression of inducible nitric oxide synthase and cyclooxygenase-2 in angiogenesis and clinical outcome of human gastric cancer. J. Surg. Oncol.94, 226–233 (2006). [DOI] [PubMed] [Google Scholar]
54.Eyler, C. E. et al. Glioma stem cell proliferation and tumor growth are promoted by nitric oxide synthase-2. Cell146, 53–66 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Glynn, S. A. et al. Increased NOS2 predicts poor survival in estrogen receptor-negative breast cancer patients. J. Clin. Invest120, 3843–3854 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Granados-Principal, S. et al. Inhibition of iNOS as a novel effective targeted therapy against triple-negative breast cancer. Breast Cancer Res. 17, 25 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Watanabe, M., Moon, K. D., Vacchio, M. S., Hathcock, K. S. & Hodes, R. J. Downmodulation of tumor suppressor p53 by T cell receptor signaling is critical for antigen-specific CD4(+) T cell responses. Immunity40, 681–691 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Wang, R. et al. The transcription factor Myc controls metabolic reprogramming upon T lymphocyte activation. Immunity35, 871–882 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Fraietta, J. A. et al. Determinants of response and resistance to CD19 chimeric antigen receptor (CAR) T cell therapy of chronic lymphocytic leukemia. Nat. Med. 24, 563–571 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Kawalekar, O. U. et al. Distinct signaling of coreceptors regulates specific metabolism pathways and impacts memory development in CAR T cells. Immunity44, 380–390 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Chen, Y. J. et al. Targeting IRG1 reverses the immunosuppressive function of tumor-associated macrophages and enhances cancer immunotherapy. Sci. Adv.9, eadg0654 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Read, A. D., Bentley, R. E., Archer, S. L. & Dunham-Snary, K. J. Mitochondrial iron-sulfur clusters: Structure, function, and an emerging role in vascular biology. Redox Biol.47, 102164 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Tannahill, G. M. et al. Succinate is an inflammatory signal that induces IL-1beta through HIF-1alpha. Nature496, 238–242 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Gocher, A. M., Workman, C. J. & Vignali, D. A. A. Interferon-gamma: teammate or opponent in the tumour microenvironment?. Nat. Rev. Immunol.22, 158–172 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Boulch, M. et al. A cross-talk between CAR T cell subsets and the tumor microenvironment is essential for sustained cytotoxic activity. Sci Immunol 10.1126/sciimmunol.abd4344 (2021). [DOI] [PubMed]
66.Boulch, M. et al. Tumor-intrinsic sensitivity to the pro-apoptotic effects of IFN-gamma is a major determinant of CD4(+) CAR T-cell antitumor activity. Nat Cancer. 10.1038/s43018-023-00570-7 (2023). [DOI] [PMC free article] [PubMed]
67.Larson, R. C. et al. CAR T cell killing requires the IFNgammaR pathway in solid but not liquid tumours. Nature604, 563–570 (2022). [DOI] [PubMed] [Google Scholar]
68.Goala, P. et al. IFN-gamma-driven skewing towards Th1 over Th17 differentiation underlies CRS and neutropenia in CAR-T therapy. J. Clin. Invest. 10.1172/JCI194631 (2026). [DOI] [PMC free article] [PubMed]
69.Schuelke, M. R. et al. Emapalumab for the treatment of refractory cytokine release syndrome in pediatric patients. Blood Adv.7, 5603–5607 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
70.Faramand, R. et al. Tumor microenvironment composition and severe cytokine release syndrome (CRS) influence toxicity in patients with large B-cell lymphoma treated with axicabtagene ciloleucel. Clin. Cancer Res. 26, 4823–4831 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
71.Faramand, R. G. et al. Baseline serum inflammatory proteins predict poor CAR T outcomes in diffuse large B-cell lymphoma. Blood Cancer Discov.5, 106–113 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Davila, M. L., Kloss, C. C., Gunset, G. & Sadelain, M. CD19 CAR-targeted T cells induce long-term remission and B Cell Aplasia in an immunocompetent mouse model of B cell acute lymphoblastic leukemia. PLoS One8, e61338 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
73.Adams, J. M. et al. The c-myc oncogene driven by immunoglobulin enhancers induces lymphoid malignancy in transgenic mice. Nature318, 533–538 (1985). [DOI] [PubMed] [Google Scholar]
74.Harris, A. W. et al. The E mu-myc transgenic mouse. A model for high-incidence spontaneous lymphoma and leukemia of early B cells. J. Exp. Med. 167, 353–371 (1988). [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Martinez, F. O., Gordon, S., Locati, M. & Mantovani, A. Transcriptional profiling of the human monocyte-to-macrophage differentiation and polarization: new molecules and patterns of gene expression. J. Immunol.177, 7303–7311 (2006). [DOI] [PubMed] [Google Scholar]
76.Cassetta, L. & Pollard, J. W. A timeline of tumour-associated macrophage biology. Nat. Rev. Cancer23, 238–257 (2023). [DOI] [PubMed] [Google Scholar]
77.Singhal, S. et al. Human tumor-associated monocytes/macrophages and their regulation of T cell responses in early-stage lung cancer. Sci. Transl. Med. 10.1126/scitranslmed.aat1500 (2019). [DOI] [PMC free article] [PubMed]
78.Kang, Y. P. et al. Non-canonical glutamate-cysteine ligase activity protects against ferroptosis. Cell Metab.33, 174–189.e177 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
79.Millard, P. et al. IsoCor: isotope correction for high-resolution MS labeling experiments. Bioinformatics35, 4484–4487 (2019). [DOI] [PubMed] [Google Scholar]
80.Chen, L. et al. PROTAC-mediated degradation of TAF1 induces apoptosis in AML cells and inhibits tumor growth in vivo. Mol. Cancer Ther.24, 1712–1721 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
81.Son, J. et al. Dual targeting of RET and SRC synergizes in RET fusion-positive cancer cells. Mol. Oncol.20, 1041–1060 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]
82.Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol.26, 1367–1372 (2008). [DOI] [PubMed] [Google Scholar]
83.Blighe, K., Lewis, M., Lun, A. & Blighe, M. K. Package ‘PCAtools.’ REAGENT or (2019).
84.Smyth, G. K. in Bioinformatics and cOmputational Biology Solutions Using R and Bioconductor 397–420 (Springer, 2005).
85.Kuznetsova, A., Brockhoff, P. B. & Christensen, R. H. B. lmerTest package: tests in linear mixed effects models. J. Stat. Softw.82, 1–26 (2017).
86.Bates, D. et al. Package ‘lme4’. convergence12, 2 (2015). [Google Scholar]
87.Yu, G., Wang, L.-G., Han, Y. & He, Q.-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics: a J. Integr. Biol.16, 284–287 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
88.Zheng, G. X. et al. Massively parallel digital transcriptional profiling of single cells. Nat. Commun.8, 14049 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
89.Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell184, 3573–3587 e3529 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
90.Aran, D. et al. Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage. Nat. Immunol.20, 163–172 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Information^{(2.8MB, pdf)}

Reporting Summary^{(119.4KB, pdf)}

Transparent Peer Review file^{(1MB, pdf)}

Source Data^{(6.4MB, xlsx)}

Data Availability Statement

[CR1] 1.Maude, S. L. et al. Chimeric antigen receptor T cells for sustained remissions in leukemia. N. Engl. J. Med. 371, 1507–1517 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR2] 2.Neelapu, S. S. et al. Axicabtagene ciloleucel CAR T-cell therapy in refractory large B-cell lymphoma. N. Engl. J. Med. 377, 2531–2544 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.Schuster, S. J. et al. Tisagenlecleucel in adult relapsed or refractory diffuse large B-cell lymphoma. N. Engl. J. Med. 380, 45–56 (2019). [DOI] [PubMed] [Google Scholar]

[CR4] 4.Park, J. H. et al. Long-term follow-up of CD19 CAR therapy in acute lymphoblastic leukemia. N. Engl. J. Med. 378, 449–459 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Maude, S. L. et al. Tisagenlecleucel in children and young adults with B-cell lymphoblastic leukemia. N. Engl. J. Med. 378, 439–448 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR6] 6.Locke, F. L. et al. Phase 1 results of ZUMA-1: a multicenter study of KTE-C19 anti-CD19 CAR T cell therapy in refractory aggressive lymphoma. Mol. Ther.25, 285–295 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR7] 7.Locke, F. L. et al. Axicabtagene ciloleucel as second-line therapy for large B-cell lymphoma. N. Engl. J. Med. 386, 640–654 (2022). [DOI] [PubMed] [Google Scholar]

[CR8] 8.Majzner, R. G. & Mackall, C. L. Clinical lessons learned from the first leg of the CAR T cell journey. Nat. Med. 25, 1341–1355 (2019). [DOI] [PubMed] [Google Scholar]

[CR9] 9.Locke, F. L. et al. Long-term safety and activity of axicabtagene ciloleucel in refractory large B-cell lymphoma (ZUMA-1): a single-arm, multicentre, phase 1-2 trial. Lancet Oncol.20, 31–42 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Cappell, K. M. & Kochenderfer, J. N. Long-term outcomes following CAR T cell therapy: what we know so far. Nat Rev Clin Oncol, 10.1038/s41571-023-00754-1 (2023). [DOI] [PMC free article] [PubMed]

[CR11] 11.Roselli, E. Faramand, R. & Davila, M. L. Insight into next-generation CAR therapeutics: designing CAR T cells to improve clinical outcomes. J.Clin.Invest. 10.1172/JCI142030 (2021). [DOI] [PMC free article] [PubMed]

[CR12] 12.Locke, F. L. et al. Tumor burden, inflammation, and product attributes determine outcomes of axicabtagene ciloleucel in large B-cell lymphoma. Blood Adv.4, 4898–4911 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Dean, E. A. et al. High metabolic tumor volume is associated with decreased efficacy of axicabtagene ciloleucel in large B-cell lymphoma. Blood Adv.4, 3268–3276 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Majzner, R. G. et al. Tuning the antigen density requirement for CAR T-cell activity. Cancer Discov.10, 702–723 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.Plaks, V. et al. CD19 target evasion as a mechanism of relapse in large B-cell lymphoma treated with axicabtagene ciloleucel. Blood138, 1081–1085 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR16] 16.Jain, M. D. et al. Whole-genome sequencing reveals complex genomic features underlying anti-CD19 CAR T-cell treatment failures in lymphoma. Blood140, 491–503 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Shouval, R. et al. Impact of TP53 genomic alterations in large B-cell lymphoma treated with CD19-chimeric antigen receptor T-cell therapy. J. Clin. Oncol.40, 369–381 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Deng, Q. et al. Characteristics of anti-CD19 CAR T cell infusion products associated with efficacy and toxicity in patients with large B cell lymphomas. Nat. Med. 26, 1878–1887 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Jackson, Z. et al. Sequential single-cell transcriptional and protein marker profiling reveals TIGIT as a marker of CD19 CAR-T cell dysfunction in patients with non-Hodgkin lymphoma. Cancer Discov.12, 1886–1903 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Jain, M. D. et al. Tumor interferon signaling and suppressive myeloid cells are associated with CAR T-cell failure in large B-cell lymphoma. Blood137, 2621–2633 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Scholler, N. et al. Tumor immune contexture is a determinant of anti-CD19 CAR T cell efficacy in large B cell lymphoma. Nat. Med28, 1872–1882 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Scott, D. W. & Gascoyne, R. D. The tumour microenvironment in B cell lymphomas. Nat. Rev. Cancer14, 517–534 (2014). [DOI] [PubMed] [Google Scholar]

[CR23] 23.Steen, C. B. et al. The landscape of tumor cell states and ecosystems in diffuse large B cell lymphoma. Cancer Cell39, 1422–1437 e1410 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Mantovani, A., Allavena, P., Marchesi, F. & Garlanda, C. Macrophages as tools and targets in cancer therapy. Nat. Rev. Drug Discov.21, 799–820 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Giavridis, T. et al. CAR T cell-induced cytokine release syndrome is mediated by macrophages and abated by IL-1 blockade. Nat. Med24, 731–738 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Morris, E. C., Neelapu, S. S., Giavridis, T. & Sadelain, M. Cytokine release syndrome and associated neurotoxicity in cancer immunotherapy. Nat. Rev. Immunol.22, 85–96 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Norelli, M. et al. Monocyte-derived IL-1 and IL-6 are differentially required for cytokine-release syndrome and neurotoxicity due to CAR T cells. Nat. Med. 24, 739–748 (2018). [DOI] [PubMed] [Google Scholar]

[CR28] 28.Kloosterman, D. J. & Akkari, L. Macrophages at the interface of the co-evolving cancer ecosystem. Cell186, 1627–1651 (2023). [DOI] [PubMed] [Google Scholar]

[CR29] 29.Christofides, A. et al. The complex role of tumor-infiltrating macrophages. Nat. Immunol.23, 1148–1156 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Munn, D. H. et al. Inhibition of T cell proliferation by macrophage tryptophan catabolism. J. Exp. Med. 189, 1363–1372 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR31] 31.Rodriguez, P. C. et al. Arginase I production in the tumor microenvironment by mature myeloid cells inhibits T-cell receptor expression and antigen-specific T-cell responses. Cancer Res. 64, 5839–5849 (2004). [DOI] [PubMed] [Google Scholar]

[CR32] 32.Newman, A. M. et al. Robust enumeration of cell subsets from tissue expression profiles. Nat. Methods12, 453–457 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.Newman, A. M. et al. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat. Biotechnol.37, 773–782 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Boucher, J. C. et al. CD28 costimulatory domain-targeted mutations enhance chimeric antigen receptor T-cell function. Cancer Immunol. Res. 9, 62–74 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Bronte, V. & Zanovello, P. Regulation of immune responses by L-arginine metabolism. Nat. Rev. Immunol.5, 641–654 (2005). [DOI] [PubMed] [Google Scholar]

[CR36] 36.Tenu, J. P. et al. Effects of the new arginase inhibitor N(omega)-hydroxy-nor-L-arginine on NO synthase activity in murine macrophages. Nitric Oxide3, 427–438 (1999). [DOI] [PubMed] [Google Scholar]

[CR37] 37.Rodriguez, P. C. & Ochoa, A. C. Arginine regulation by myeloid derived suppressor cells and tolerance in cancer: mechanisms and therapeutic perspectives. Immunol. Rev.222, 180–191 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Moore, W. M. et al. L-N6-(1-iminoethyl)lysine: a selective inhibitor of inducible nitric oxide synthase. J. Med Chem.37, 3886–3888 (1994). [DOI] [PubMed] [Google Scholar]

[CR39] 39.Harris, I. S. & DeNicola, G. M. The complex interplay between antioxidants and ROS in cancer. Trends Cell Biol.30, 440–451 (2020). [DOI] [PubMed] [Google Scholar]

[CR40] 40.Alizadeh, D. et al. IFNgamma is critical for CAR T cell-mediated myeloid activation and induction of endogenous immunity. Cancer Discov.11, 2248–2265 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR41] 41.Bailey, S. R. et al. Blockade or deletion of IFNgamma reduces macrophage activation without compromising CAR T-cell function in hematologic malignancies. Blood Cancer Discov.3, 136–153 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Srivastava, S. et al. Immunogenic chemotherapy enhances recruitment of CAR-T cells to lung tumors and improves antitumor efficacy when combined with checkpoint blockade. Cancer Cell39, 193–208 e110 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Takeda, N. et al. Differential activation and antagonistic function of HIF-alpha isoforms in macrophages are essential for NO homeostasis. Genes Dev.24, 491–501 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Keith, B., Johnson, R. S. & Simon, M. C. HIF1alpha and HIF2alpha: sibling rivalry in hypoxic tumour growth and progression. Nat. Rev. Cancer12, 9–22 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR45] 45.Hodeify, R., Tarcsafalvi, A., Megyesi, J., Safirstein, R. L. & Price, P. M. Cdk2-dependent phosphorylation of p21 regulates the role of Cdk2 in cisplatin cytotoxicity. Am. J. Physiol. Ren. Physiol.300, F1171–F1179 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Marchingo, J. M. & Cantrell, D. A. Protein synthesis, degradation, and energy metabolism in T cell immunity. Cell Mol. Immunol.19, 303–315 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR47] 47.Bantug, G. R., Galluzzi, L., Kroemer, G. & Hess, C. The spectrum of T cell metabolism in health and disease. Nat. Rev. Immunol.18, 19–34 (2018). [DOI] [PubMed] [Google Scholar]

[CR48] 48.Reina-Campos, M., Scharping, N. E. & Goldrath, A. W. CD8(+) T cell metabolism in infection and cancer. Nat. Rev. Immunol.21, 718–738 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Zhao, H. et al. Myeloid-derived itaconate suppresses cytotoxic CD8(+) T cells and promotes tumour growth. Nat. Metab.4, 1660–1673 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR50] 50.Mills, E. L. et al. Itaconate is an anti-inflammatory metabolite that activates Nrf2 via alkylation of KEAP1. Nature556, 113–117 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR51] 51.Locke, F. L. et al. Impact of tumor microenvironment on efficacy of anti-CD19 CAR T cell therapy or chemotherapy and transplant in large B cell lymphoma. Nat. Med. 30, 507–518 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Chen, S. et al. Macrophages in immunoregulation and therapeutics. Signal Transduct. Target Ther.8, 207 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR53] 53.Chen, C. N., Hsieh, F. J., Cheng, Y. M., Chang, K. J. & Lee, P. H. Expression of inducible nitric oxide synthase and cyclooxygenase-2 in angiogenesis and clinical outcome of human gastric cancer. J. Surg. Oncol.94, 226–233 (2006). [DOI] [PubMed] [Google Scholar]

[CR54] 54.Eyler, C. E. et al. Glioma stem cell proliferation and tumor growth are promoted by nitric oxide synthase-2. Cell146, 53–66 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Glynn, S. A. et al. Increased NOS2 predicts poor survival in estrogen receptor-negative breast cancer patients. J. Clin. Invest120, 3843–3854 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Granados-Principal, S. et al. Inhibition of iNOS as a novel effective targeted therapy against triple-negative breast cancer. Breast Cancer Res. 17, 25 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Watanabe, M., Moon, K. D., Vacchio, M. S., Hathcock, K. S. & Hodes, R. J. Downmodulation of tumor suppressor p53 by T cell receptor signaling is critical for antigen-specific CD4(+) T cell responses. Immunity40, 681–691 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Wang, R. et al. The transcription factor Myc controls metabolic reprogramming upon T lymphocyte activation. Immunity35, 871–882 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR59] 59.Fraietta, J. A. et al. Determinants of response and resistance to CD19 chimeric antigen receptor (CAR) T cell therapy of chronic lymphocytic leukemia. Nat. Med. 24, 563–571 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR60] 60.Kawalekar, O. U. et al. Distinct signaling of coreceptors regulates specific metabolism pathways and impacts memory development in CAR T cells. Immunity44, 380–390 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR61] 61.Chen, Y. J. et al. Targeting IRG1 reverses the immunosuppressive function of tumor-associated macrophages and enhances cancer immunotherapy. Sci. Adv.9, eadg0654 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR62] 62.Read, A. D., Bentley, R. E., Archer, S. L. & Dunham-Snary, K. J. Mitochondrial iron-sulfur clusters: Structure, function, and an emerging role in vascular biology. Redox Biol.47, 102164 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR63] 63.Tannahill, G. M. et al. Succinate is an inflammatory signal that induces IL-1beta through HIF-1alpha. Nature496, 238–242 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR64] 64.Gocher, A. M., Workman, C. J. & Vignali, D. A. A. Interferon-gamma: teammate or opponent in the tumour microenvironment?. Nat. Rev. Immunol.22, 158–172 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR65] 65.Boulch, M. et al. A cross-talk between CAR T cell subsets and the tumor microenvironment is essential for sustained cytotoxic activity. Sci Immunol 10.1126/sciimmunol.abd4344 (2021). [DOI] [PubMed]

[CR66] 66.Boulch, M. et al. Tumor-intrinsic sensitivity to the pro-apoptotic effects of IFN-gamma is a major determinant of CD4(+) CAR T-cell antitumor activity. Nat Cancer. 10.1038/s43018-023-00570-7 (2023). [DOI] [PMC free article] [PubMed]

[CR67] 67.Larson, R. C. et al. CAR T cell killing requires the IFNgammaR pathway in solid but not liquid tumours. Nature604, 563–570 (2022). [DOI] [PubMed] [Google Scholar]

[CR68] 68.Goala, P. et al. IFN-gamma-driven skewing towards Th1 over Th17 differentiation underlies CRS and neutropenia in CAR-T therapy. J. Clin. Invest. 10.1172/JCI194631 (2026). [DOI] [PMC free article] [PubMed]

[CR69] 69.Schuelke, M. R. et al. Emapalumab for the treatment of refractory cytokine release syndrome in pediatric patients. Blood Adv.7, 5603–5607 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR70] 70.Faramand, R. et al. Tumor microenvironment composition and severe cytokine release syndrome (CRS) influence toxicity in patients with large B-cell lymphoma treated with axicabtagene ciloleucel. Clin. Cancer Res. 26, 4823–4831 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR71] 71.Faramand, R. G. et al. Baseline serum inflammatory proteins predict poor CAR T outcomes in diffuse large B-cell lymphoma. Blood Cancer Discov.5, 106–113 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR72] 72.Davila, M. L., Kloss, C. C., Gunset, G. & Sadelain, M. CD19 CAR-targeted T cells induce long-term remission and B Cell Aplasia in an immunocompetent mouse model of B cell acute lymphoblastic leukemia. PLoS One8, e61338 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR73] 73.Adams, J. M. et al. The c-myc oncogene driven by immunoglobulin enhancers induces lymphoid malignancy in transgenic mice. Nature318, 533–538 (1985). [DOI] [PubMed] [Google Scholar]

[CR74] 74.Harris, A. W. et al. The E mu-myc transgenic mouse. A model for high-incidence spontaneous lymphoma and leukemia of early B cells. J. Exp. Med. 167, 353–371 (1988). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR75] 75.Martinez, F. O., Gordon, S., Locati, M. & Mantovani, A. Transcriptional profiling of the human monocyte-to-macrophage differentiation and polarization: new molecules and patterns of gene expression. J. Immunol.177, 7303–7311 (2006). [DOI] [PubMed] [Google Scholar]

[CR76] 76.Cassetta, L. & Pollard, J. W. A timeline of tumour-associated macrophage biology. Nat. Rev. Cancer23, 238–257 (2023). [DOI] [PubMed] [Google Scholar]

[CR77] 77.Singhal, S. et al. Human tumor-associated monocytes/macrophages and their regulation of T cell responses in early-stage lung cancer. Sci. Transl. Med. 10.1126/scitranslmed.aat1500 (2019). [DOI] [PMC free article] [PubMed]

[CR78] 78.Kang, Y. P. et al. Non-canonical glutamate-cysteine ligase activity protects against ferroptosis. Cell Metab.33, 174–189.e177 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR79] 79.Millard, P. et al. IsoCor: isotope correction for high-resolution MS labeling experiments. Bioinformatics35, 4484–4487 (2019). [DOI] [PubMed] [Google Scholar]

[CR80] 80.Chen, L. et al. PROTAC-mediated degradation of TAF1 induces apoptosis in AML cells and inhibits tumor growth in vivo. Mol. Cancer Ther.24, 1712–1721 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR81] 81.Son, J. et al. Dual targeting of RET and SRC synergizes in RET fusion-positive cancer cells. Mol. Oncol.20, 1041–1060 (2026). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR82] 82.Cox, J. & Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol.26, 1367–1372 (2008). [DOI] [PubMed] [Google Scholar]

[CR83] 83.Blighe, K., Lewis, M., Lun, A. & Blighe, M. K. Package ‘PCAtools.’ REAGENT or (2019).

[CR84] 84.Smyth, G. K. in Bioinformatics and cOmputational Biology Solutions Using R and Bioconductor 397–420 (Springer, 2005).

[CR85] 85.Kuznetsova, A., Brockhoff, P. B. & Christensen, R. H. B. lmerTest package: tests in linear mixed effects models. J. Stat. Softw.82, 1–26 (2017).

[CR86] 86.Bates, D. et al. Package ‘lme4’. convergence12, 2 (2015). [Google Scholar]

[CR87] 87.Yu, G., Wang, L.-G., Han, Y. & He, Q.-Y. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics: a J. Integr. Biol.16, 284–287 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR88] 88.Zheng, G. X. et al. Massively parallel digital transcriptional profiling of single cells. Nat. Commun.8, 14049 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR89] 89.Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell184, 3573–3587 e3529 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR90] 90.Aran, D. et al. Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage. Nat. Immunol.20, 163–172 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

IFN-γ-driven iNOS induction in macrophages mediates CAR T cell resistance in B cell lymphoma

Sae Bom Lee

Allan Pavy

Nolan Beatty

Justin C Boucher

Jay K Mandula

Emiliano Roselli

Darwin C Chang

Rachel V Jimenez

Hiroshi Kotani

Kayla Reid

Julio Vazquez-Martinez

Payal Goala

Max Ilecki

Cooper J Sailer

Shannon L McSain

Showkat Hamid

Muhammad Junaid Tariq

Rosa A Sierra-Mondragon

Min Liu

John M Koomen

Jonathan Nguyen

Mohammad Hussaini

Timothy I Shaw

Xuefeng Wang

Jianmin Wang

Julieta Abraham-Miranda

Rawan Faramand

Michael D Jain

Meredith L Stone

Frederick L Locke

Paulo C Rodriguez

Yun Pyo Kang

Marco L Davila

Abstract

Introduction

Results

Macrophages in the pre-CAR T cell treatment TME are linked to therapeutic responses in LBCL patients

Fig. 1. Macrophages in the pre-CAR T cell treatment TME are linked to therapeutic responses to CAR T cell therapy in LBCL patients.

Immunoregulatory actions of macrophages on CAR T cells

Fig. 2. Exposure to imMac provokes CAR T cell dysfunction.

CAR T cell-exposed imMac upregulate iNOS

Fig. 3. CAR T cell-exposed imMac upregulate iNOS.

iNOS upregulation in imMac drives suppression of CAR T cell function

Fig. 4. iNOS upregulation in imMac drives suppression of CAR T cell function.

CAR T cell-derived IFN-γ induces iNOS in imMac

Fig. 5. CAR T cell-derived IFN-γ induces iNOS expression in imMac.

iNOS-expressing imMac induce CAR T cell metabolic dysregulation

Fig. 6. iNOS-expressing imMac drive CAR T cell dysregulation.

Fig. 7. iNOS-expressing imMac induce CAR T cell metabolic dysregulation.

iNOS inhibition improves CAR T cell therapeutic efficacy

Fig. 8. iNOS inhibition improves efficacy of CAR T cell therapy.

Discussion

Methods

Patient samples

Mice

Cell lines

Patient tumor bulk RNA-sequencing

Patient leukapheresis sample flow cytometry analysis

Generation of retroviral constructs

Mouse T cell isolation and CAR T cell generation

Human T cell isolation and CAR T cell generation

Animal experiments

Mouse peritoneal cell collection

Murine macrophage development and polarization

Human macrophage development and polarization

Expansion assay

Griess assay

BrdU incorporation assay

Flow cytometry

Coculture assays and CAR T cell isolation for subsequent downstream assays

Luciferase-based killing assay

Cytokine secretion assay

Immunoblotting

Metabolomics and 13C6-labeled glucose tracing analyses

Proteomics and phospho-peptide analyses

Seahorse assay

Single cell RNA-sequencing analysis

Reporting summary

Metabolomics and ¹³C₆-labeled glucose tracing analyses