Targeting Cysteinyl Leukotriene Receptor 1 reprograms tumor-promoting myelopoiesis and overcomes immune checkpoint therapy resistance

Abstract

The “emergency” myelopoiesis usually occurs to facilitate malignancy progression. Immunosuppressive neutrophils, also deemed as PMN-MDSCs (polymorphonuclear myeloid-derived suppressor cells), are potent mediators of immunosuppression and tumor promotion. Here, we demonstrate that the STAT3-dependent induction of CysLTR1 (Cysteinyl Leukotriene Receptor 1), a key player for asthma, and other inflammatory conditions, sustains tumor-promoting emergency myelopoiesis. Genetic ablation and pharmacological inhibition of CysLTR1 diminishes tumor growth with enhanced antitumor immunity. Antitumor effect of CysLTR1 inhibition is linked with transcriptomic rewiring of granulopoiesis and neutrophil reprogramming toward an antitumor immune phenotype; this process requires the distinct transcriptional factors MXD1 and NFE2 to specifically dictate myeloid progenitor commitment and differentiation in association with controlled de novo synthesis of granule cargoes. Targeting CysLTR1 with clinically available antagonists overcomes resistance to anti-PD-1 in multiple mouse tumor models. These findings thus open the opportunity for rapid translation of CysLTR1 inhibitors targeting emergency granulopoiesis and unveil a new programmed “disarming” neutrophil strategy to improve cancer immunotherapy.

INTRODUCTION

Inflammation critically shapes tumor development by integrating tumor-intrinsic alterations with systemic immune responses¹. A major consequence of chronic tumor-associated inflammation is the induction of emergency myelopoiesis, a process characterized by expansion and skewing of bone marrow progenitors toward myeloid lineages. This shift promotes accumulation of myeloid-derived suppressor cells (MDSCs), which suppress adaptive immunity and facilitate tumor progression^{2, 3}.

MDSCs comprise heterogeneous immature myeloid populations. In mice, they are broadly defined as CD11b⁺Gr-1⁺ cells and include Ly6G⁺ polymorphonuclear cells (PMN-MDSCs) and Ly6C⁺ monocytic cells (M-MDSCs)⁴. In cancer patients, analogous populations share suppressive properties but lack clear phenotypic uniformity⁵. Beyond T-cell inhibition, MDSCs promote angiogenesis, tumor invasion, metastasis, and resistance to therapy^{5, 6}.

Although expansion of suppressive myeloid populations is well documented, the molecular regulators that drive lineage commitment, functional polarization, and granule formation during tumor-associated granulopoiesis remain incompletely defined. In particular, transcriptional programs linking inflammatory cues to neutrophil differentiation and effector molecule production require clarification^7–9.

Here we identify Cysteinyl Leukotriene Receptor 1 (CysLTR1), a leukotriene receptor implicated in inflammatory diseases, as a central regulator of tumor-driven granulopoiesis. We demonstrate that CysLTR1 governs neutrophil differentiation, suppressive programming, and granule synthesis through transcriptional control involving MXD1 and NFE2. Genetic deletion or pharmacologic inhibition of CysLTR1 reprograms neutrophils toward an immunostimulatory phenotype and enhances responsiveness to PD-1 blockade across multiple tumor models.

RESULTS

CysLTR1 ablation mitigates emergency myelopoiesis and inhibits tumor progression

To compare tumor-associated CD11b⁺ myeloid cells to splenic/circulating myeloid cells of tumor-free mice, we analyzed microarray datasets GSE21927¹⁰ and GSE39807¹¹, and NanoString dataset GSE223028¹² (Fig. 1a, b). Using FC ≥ 0.5, 145, 338, and 32 DEGs were identified, respectively. Cysltr1 was the only gene shared across all three datasets (Fig. 1a, b) and was selected for further investigation.

Figure 1. Upregulated expression of CysLTR1 primarily in tumor-associated neutrophils.

(a) Venn diagram depicting differentially expressed genes (DEGs; Log2FC≥ 0.5 or ≤ −0.5, −Log10(p value) > 4) identified from the deposited datasets of GSE21927 (n=145), GSE39807 (n=338) and GSE223028 (n=32). FC, Fold Change.

(b) Volcano plots illustrating the upregulated DEGs (Log2FC≥ 0.5 or ≤ −0.5, −Log10(p value) in GSE21927 (left, n=145), GSE39807 (middle, n=338) and GSE223028 (right, n=32), with Cysltr1 highlighted in red.

(c) qPCR analysis of relative expression of Cysltr1 in sorted CD11b⁺Gr1⁺ cells (GC), macrophages (Mφ, CD11b⁺Gr1⁻F4/80⁺), dendritic cells (DC, CD11b⁺Gr1⁻CD11c⁺), B cells (B, CD3⁻B220⁺), NK cells (NK, CD3⁻NK1.1⁺), and T cells (T, CD3⁺B220⁻) from tumor-free and 4T1-bearing BALB/c mice (n=3).

(d) qPCR analysis of Cysltr1 expression in CD11b⁺Gr1⁺ cells from various tumor-bearing mice, and tumor-free mice (n=3).

(e) Relative mean fluorescence intensity (MFI) of CysLTR1 on tumor-infiltrating CD11b⁺Ly6G⁺ neutrophils and CD11b⁺Ly6C⁺ monocytic cells from E0771 tumor-bearing C57BL/6 mice, determined by flow cytometry (n=5).

(f) Frequency of CysLTR1⁺ neutrophils in peripheral blood (PB) (Days 0, 7, 14, and 21) and tumors (Day 21) of E0771-bearing mice (n = 5).

(g) Hierarchical differentiation tree illustrating the expression pattern of Cysltr1 in mouse progenitor and differentiated cells from the deposited datasets of GSE14833 (n=47) and GSE6506 (n=20).

(h) qPCR analysis of Cysltr1 expression in each sorted population during granulopoiesis from tumor-free C57BL/6 mice (n=4).

(i) ELISA analysis of GM-CSF, G-CSF, and IL-6 levels in tumor lysates (TL) and plasma (PL) from E0771-bearing C57BL/6 mice.

(j) ELISA analysis of cysteinyl leukotrienes (CysLTs) in C57BL/6 splenocyte supernatants or tumor explant supernatants (TES) from E0771, MC38, LLC1, and BPS1 tumors over 24h (n=3).

(k) ELISA analysis of CysLTs in TES from C57BL/6 mice bearing 1-week-old, 2-week-old or 3-week-old E0771 tumors (n=3).

(l) ELISA analysis of CysLTs in TES of single cell tumor digests, sorted intratumoral CD11b⁺Gr1⁺ cells (GC) and GC-depleted fractions from E0771-bearing C57BL/6 mice (n=3).

(m) qPCR analysis of Cysltr1 expression in sorted BM LSK, CMP, GMP, and splenic CD11b⁺Gr1⁺ GC from tumor-free and EMT6 tumor-bearing BALB/c mice (n=4).

(n) qPCR analysis of Cysltr1 expression in BM-derived cells cultured with GM-CSF + IL-6. C57BL/6 BM cells cultured in medium alone served as control (n=4).

Results are representative of two (c, d, f, j, m), or three (e, h, i, k, l, n) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (e, i, n), or one-way ANOVA (c, d, f, j-m).

RT-qPCR confirmed that Cysltr1 expression was restricted to the myeloid lineage, with highest levels in tumor-infiltrating Gr1⁺CD11b⁺ cells (Fig. 1c). Infiltrating Gr1⁺ cells from multiple tumor models (4T1, E0771, MC38, LLC1, BPS1) showed elevated Cysltr1 compared with naïve splenic counterparts (Fig. 1d). Ly6G⁺ neutrophils expressed higher CysLTR1 than Ly6C⁺ monocytic cells (Fig. 1e). A progressive increase in circulating CysLTR1⁺ neutrophils was observed during tumor growth, with further enrichment in tumors by day 21 (Fig. 1f), indicating dynamic upregulation during tumor progression.

Analysis of hematopoietic datasets (GSE14833, GSE6506) revealed highest Cysltr1 expression in granulocyte-macrophage progenitors (GMPs) (Fig. 1g). RT-qPCR confirmed increasing Cysltr1 expression during early neutrophil maturation (Fig. 1h), supporting a role for CysLTR1 in steady-state granulopoiesis.

In tumor-free Cysltr1^−/− mice, total bone marrow (BM) cellularity was reduced (Extended Data Fig. 1a), with decreased LSK cells but unchanged HPCs (Extended Data Fig. 1b). GMP frequencies increased, whereas CMP and MEP populations decreased (Extended Data Fig. 1b), findings consistent across BALB/c and C57BL/6 backgrounds (Extended Data Fig. 1c).

Tumors contained elevated GM-CSF, G-CSF, and IL-6^{2, 3, 10, 13} (Fig. 1i), along with increased cysteinyl leukotrienes (CysLTs), including LTD4 (Fig. 1j), which rose during tumor growth (Fig. 1k). Gr1⁺ cells and tumor cells were major sources of CysLTs (Fig. 1l). Cysltr1 expression was higher in Gr1⁺ cells and neutrophil progenitors from tumor-bearing mice (Fig. 1m) and was induced by GM-CSF + IL-6 during BM differentiation (Fig. 1n).

Primary cultures of Cysltr1^−/− BM progenitors yielded fewer total cells accompanying a decrease in CFU-G but an increase in CFU-M counts (Extended Data Fig. 1d–f). Additionally, CysLTR1 deficiency enhanced early apoptosis (Extended Data Fig. 1g) and altered cell cycle (Extended Data Fig. 1h).

Lin⁻ BM cells stimulated with GM-CSF, GM-CSF + IL-6, or GM-CSF + EMT6-conditioned medium generated fewer Ly6G⁺ neutrophils and a reduced Ly6G⁺/Ly6C⁺ ratio from Cysltr1^−/− mice, with increased macrophage and DC differentiation (Extended Data Fig. 1i–k). Under GM-CSF ± IL-6, Cysltr1^−/− cultures accumulated early-stage neutrophil precursors (MB, PM, MC, MM)^{14, 15} but showed reduced mature band and segmented neutrophils (Extended Data Fig. 1l, m), indicating impaired differentiation upon cancer-associated proinflammatory cytokines.

Sorted Cysltr1^−/− GMPs similarly generated fewer Ly6G⁺ cells and fewer mature neutrophils (Extended Data Fig. 1n, o). Secondary stimulation of GM-CSF + IL-6-primed Gr1⁻ cells confirmed reduced CD11b⁺Gr1⁺ and Ly6G⁺ differentiation from Cysltr1^−/− cells (Extended Data Fig. 1p, q). Across G-CSF, G-CSF + IL-6, M-CSF, and M-CSF + IL-6 conditions, Cysltr1^−/− cells consistently produced fewer Ly6G⁺ neutrophils with reduced Ly6G⁺/Ly6C⁺ ratios (Extended Data Fig. 1r–u).

Accumulating evidence reveals a crucial role of Neutrophil Extracellular Traps (NETs) in promoting tumor cell progression^{16, 17} and immune evasion^{18, 19}.Cysltr1^−/− neutrophils exhibited impaired NET formation, reduced MPO and CitH3 staining (Extended Data Fig. 1v), and decreased MPO-DNA complexes and CitH3 release after PMA or tumor-conditioned medium stimulation (Extended Data Fig. 1w, x).

In EMT6-bearing mice, Cysltr1 deficiency reduced LSK and HPC frequencies (Fig. 2a, b), with increased GMPs but decreased CMPs and MEPs (Fig. 2c). Similar findings were observed in E0771 tumors and in montelukast-treated mice (Extended Data Fig. 2a, b). Neutrophil maturation in vivo was similarly impaired, with accumulation of MC, MM, and BC but fewer segmented neutrophils (Fig. 2d).

Figure 2. CysLTR1 deficiency attenuates tumor-induced myelopoiesis and inhibits tumor progression.

(a) Frequency of Lin⁻Sca1⁺c-Kit⁺ (LSK) cells in lineage-negative cells and total LSK cell count in EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(b) Frequency of Lin⁻Sca1⁻c-Kit⁺ hematopoietic progenitor cells (HPC) in lineage-negative cells and total HPC count in EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(c) Frequency of CD34⁺CD16/32⁻ common myeloid progenitors (CMP), CD34⁺CD16/32⁺ granulocytic myeloid progenitors (GMP), and CD34⁻CD16/32⁻ megakaryocyte erythrocyte progenitors (MEP) within HPC (left) and total CMP, GMP, and MEP cell counts (right) in BM of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(d) Frequency of myeloblasts (MB), promyelocytes (PM), myelocytes (MC), metamyelocytes (MM), band cells (BC), and segmented neutrophils (SN) in BM neutrophils of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(e, f) Frequencies (left), cell counts (middle) of Ly6G⁺ neutrophils and Ly6C⁺ monocytic cells, and Ly6G⁺/Ly6C⁺ cell ratio (right) in CD45⁺ cells from bone marrow (e) and spleen (f) from EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(g, h) Representative flow dot plots showing Ly6G⁺ neutrophil and Ly6C⁺ monocytic cell subsets within tumor-infiltrating CD11b⁺ cells (g), and statistical analysis of frequencies (left), cell counts (middle), and the Ly6G⁺/Ly6C⁺ cell ratio (right) in CD45⁺ tumor infiltrates (h) of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(i) Frequencies (left) and cell counts (right) of macrophage (Mφ) and dendritic cell (DC) in CD45⁺ tumor infiltrates of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(j) Tumor growth curve and survival of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=8).

(k) Tumor growth curve and survival of E0771 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ mice (n=8).

(l) Tumor growth curve and survival of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ bone marrow (BM)-transplanted mice (n=8).

(m) Tumor growth curve and survival of E0771 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ BM-transplanted mice (n=8).

(n) Tumor growth curve and survival of E0771 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice (n=5).

(o) Tumor growth curve and survival of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice with Ly6G⁺ neutrophil depletion using anti (α)-Ly6G neutralizing antibodies (n=5).

Results are representative of three (a-i), or two (j-o) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (a, b, e (right), f (right), h (right)), log-rank test (j-o (right)), one-way ANOVA (c, d, e (left, middle), f (left, middle), h (left, middle), i) or two-way ANOVA (j-o (left)).

Flow cytometry revealed marked reductions in Ly6G⁺ neutrophils and Ly6G⁺/Ly6C⁺ ratios in BM, spleen, and tumors of Cysltr1^−/− and montelukast-treated mice (Fig. 2e–i; Extended Data Fig. 2c–e). Macrophage and DC frequencies were modestly increased (Fig. 2i; Extended Data Fig. 2c).

In LPS-induced emergency myelopoiesis, Cysltr1^−/− mice exhibited reduced LSK, HPC, CMP, and GMP populations (Extended Data Fig. 2f–h), impaired neutrophil maturation (Extended Data Fig. 2i), and reduced Ly6G⁺ accumulation (Extended Data Fig. 2j), indicating a broader requirement for CysLTR1 in inflammatory myelopoiesis.

Cysltr1^−/− mice bearing EMT6, E0771, CT26, and YUMMER1.7 tumors exhibited delayed tumor growth and improved survival (Fig. 2j, k; Extended Data Fig. 2k–p; Supplementary Fig. 2a–d). Montelukast similarly suppressed spontaneous mammary tumors in FVB-MMTV-PyMT mice (Extended Data Fig. 2q).

Bone marrow transplantation experiments (Supplementary Fig. 1a) demonstrated that Cysltr1^−/− > WT chimeras showed reduced tumor growth and prolonged survival (Fig. 2l, m; Extended Data Fig. 2r, s; Supplementary Fig. 2e, f), accompanied by decreased intratumoral Ly6G⁺ accumulation and lower Ly6G⁺/Ly6C⁺ ratios (Extended Data Fig. 2t).

In competitive repopulation assays (Supplementary Fig. 2b), Cysltr1^−/− donor cells progressively declined in tumor-infiltrating CD45⁺ and CD11b⁺ compartments, significantly with competitive disadvantage in Ly6G⁺ neutrophils (Extended Data Fig. 2u–y).

Myeloid-specific deletion using Cysltr1^f/f Lyz2-cre mice (Supplementary Fig. 3a, b) recapitulated tumor suppression (Fig. 2n; Extended Data Fig. 3a–c; Supplementary Fig. 2g, h), reduced BM cellularity and neutrophil maturation (Extended Data Fig. 3d–f), decreased intratumoral Ly6G⁺ accumulation and increased macrophages and DCs (Extended Data Fig. 3g). Anti-Ly6G depletion suppressed tumor growth in WT but not Cysltr1^−/− mice (Fig. 2o; Supplementary Fig. 2i), indicating that neutrophils are likely the key mediators of CysLTR1-dependent tumor progression.

In line with the established role of the spleen in extramedullary myelopoiesis, we also observed impaired splenic myelopoiesis in LPS-stimulated Cysltr1^−/− emergency models (Extended Data Fig. 3h) and MTK-treated tumor-bearing mice (Extended Data Fig. 3i). Similar to the effects of Ly6G⁺ neutrophil depletion (Figure 2o) or CysLTR1 ablation, splenectomy (Supplementary Fig. 1c) reduced tumor growth and circulating neutrophils in WT mice (Extended Data Fig. 3j, k, Supplementary Fig. j) and further enhanced tumor control in Cysltr1^−/− mice, suggesting both CysLTR1-dependent and independent splenic contributions, In line with the established role of the spleen in extramedullary myelopoiesis^20–22.

Tumor-associated Ly6G⁺ neutrophils expressed higher CD14²³, CD84²⁴, CD300ld²⁵, and DcTRAIL-R1²⁶ than naïve neutrophils (Extended Data Fig. 3l). CysLTR1 expression was elevated in tumor-associated neutrophils (Extended Data Fig. 3m) and correlated with CD84 and CD300ld at mRNA and protein levels (Extended Data Fig. 3n–p). These markers were reduced in Cysltr1^−/−, Cysltr1^f/f Lyz2-cre, and Montelukast (MTK, a selective CysLTR1 antagonist)-treated models (Extended Data Fig. 3q–u), supporting CysLTR1 as a functional marker of tumor-promoting neutrophils.

Collectively, the data confirmed the tumor-promoting role of CysLTR1 in driving emergency myelopoiesis, highlighting a tighter dependence of neutrophil development on increased CysLTR1 expression during tumor progression.

CysLTR1 ablation impairs PMN-MDSC activity and improves antitumor T cell immunity rendering tumors sensitive to anti-PD-1 therapy

In size-matched EMT6 (Fig. 3a–e) and E0771 tumors (Extended Data Fig. 4a–c), CysLTR1 ablation significantly increased CD3⁺ T cell infiltration (Fig. 3a; Extended Data Fig. 4a), including CD4⁺ (Fig. 3b; Extended Data Fig. 4b) and CD8⁺ subsets (Fig. 3c; Extended Data Fig. 4b), while reducing Ly6G⁺ neutrophil accumulation (Fig. 2h). Regulatory T cells were also decreased (Fig. 3d; Extended Data Fig. 4c). Similar enhancement of T cell infiltration was observed in Cysltr1^f/f Lyz2-cre mice (Extended Data Fig. 4d–f) and MTK-treated mice (Extended Data Fig. 4g–i).

Figure 3. CysLTR1 ablation mitigates PMN-MDSC/neutrophil activity and improves antitumor T cell immunity rendering tumors sensitive to anti-PD-1 therapy.

(a-c) Frequencies and cell counts of CD3⁺ (a), CD4 (b), CD8 (c) T cells in CD45⁺ tumor infiltrates of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(d) Frequency of CD25⁺ Treg within infiltrating CD4 T cells of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(e) Frequency of PD1⁺Tim3⁺ exhausted-like cells within infiltrating CD4 and CD8 T cells of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(f) Suppression of CD8 T cells (T) proliferation by Cysltr1^−/− and Cysltr1^+/+ infiltrating CD11b⁺Ly6G⁺ neutrophils (E) isolated from EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(g) Suppression of CD8 T cells (T) proliferation by Cysltr1^−/− and Cysltr1^+/+ CD11b⁺Ly6G⁺ neutrophils (E) differentiated in vitro from BALB/c Cysltr1^−/− and Cysltr1^+/+ BM cells with GM-CSF + IL-6 (n=4).

(h) Frequencies of CD40⁺ and CD86⁺ neutrophils (left) and relative MFI of H2Kb-OVAI within tumor-infiltrating neutrophils from LLC1-OVA tumor bearing C57BL/6 mice (n=4).

(i) Frequency of proliferated OTI CD8 T cells after co-culture with tumor-infiltrating neutrophils from LLC1-OVA tumor bearing C57BL/6 mice (left) and frequencies of CD69⁺, IFN-γ⁺ and Granzyme B⁺ cells after co-culture (right) (n=3).

(j) Relative MFI of CD163 (left) and PD-L1 (right) on tumor-infiltrating neutrophils of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(k) Frequencies of CD44⁺ (left), IFN-γ⁺ (middle), and TNF-α⁺ (right) cells in EMT6-infiltrating CD8⁺ T cells co-cultured with splenic CD11b⁺ cells from tumor-free BALB/c WT mice with antigenic peptides as indicated (n=4).

(l) ELISPOT analysis of IFN-γ-producing CD8⁺ T cells isolated from EMT6 tumors co-cultured with spleen CD11b⁺ cells from tumor-free BALB/c mice and EMT6 lysate (EMT6) or peptides (n=4).

(m) ELISA analysis of MPO-DNA complex (top) and CitH3 (bottom) in tumor explant supernatants (TES) from EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice over 24h (n=4).

(n) Tumor growth (top) and survival (bottom) of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ with or without CD8 depletion (n=8).

(o) Tumor growth (top) and survival (bottom) of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ with or without anti-PD-1 therapy (n=8).

(p) Tumor growth (top) and survival (bottom) of EMT6 tumor-bearing BALB/c mice with or without montelukast (MTK) and anti-PD-1 therapy (n=8).

Results are representative of three (a-e, k, n-p), or two (f-i, m) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (a-d, f, g, h (right), i (left), j, m), log-rank test (n-p (bottom)), one-way ANOVA (e, h (left), i (right), k, l) or two-way ANOVA (n-p (top)).

Tumor-infiltrating CD4⁺ and CD8⁺ T cells from Cysltr1^−/− mice showed reduced exhaustion, with lower PD-1 and TIM3 expression (Fig. 3e). Functionally, Ly6G⁺ neutrophils from EMT6- (Fig. 3f) and E0771-bearing Cysltr1^−/− mice (Extended Data Fig. 4j) exhibited diminished suppression of T cell proliferation compared to wild-type controls. A decreased suppressive activity was also detected in GM-CSF + IL-6-induced Cysltr1^−/− BM-derived cells (Fig. 3g; Extended Data Fig. 4k). In co-culture assays, CysLTR1^hi Ly6G⁺ neutrophils were more suppressive than CysLTR1^low counterparts (Extended Data Fig. 4l). Furthermore, inhibition of cysteinyl leukotriene production with Zileuton (ZIL) or blockade of CysLTR1 with MTK abolished neutrophil suppressive activity (Extended Data Fig. 4m), confirming reliance on the CysLT-CysLTR1 axis.

We next evaluated antigen-presenting features of neutrophils. In LLC1-OVA tumors, Ly6G⁺ neutrophils from Cysltr1^−/− mice expressed higher levels of CD40 and CD86 and displayed increased H2Kb-OVA-I complexes (Fig. 3h). Sorted Cysltr1^−/− Ly6G⁺ neutrophils more effectively induced OTI CD8⁺ T cell proliferation and effector responses, including IFN-γ and Granzyme B production (Fig. 3i). Thus, CysLTR1 ablation not only reduces suppressive activity but also promotes a shift toward an antigen-presenting phenotype.

Consistently, intratumoral CD11b⁺ cells from Cysltr1^−/− mice showed reduced M2-like polarization, with lower CD163 and PD-L1 expression (Fig. 3j). Tumor-reactive CD8⁺ T cells displayed greater activation (CD44 expression) and higher IFN-γ and TNF-α production in response to AH1 peptide compared to OVA-I control (Fig. 3k). ELISPOT assays confirmed increased IFN-γ-producing CD8⁺ T cells upon stimulation with EMT6 lysate or AH1 peptide (Fig. 3l; Extended Data Fig. 4n). Similar enhancement of antigen-specific CD8⁺ responses was observed in E0771-SIY tumors (Extended Data Fig. 4o–q) and in Cysltr1^f/f Lyz2-cre or MTK-treated mice (Extended Data Fig. 4r, s), supporting a broad improvement in tumor-specific T cell immunity.

ELISA analysis showed reduced MPO-DNA complexes and citrullinated histone H3 (CitH3) in tumor-infiltrating Cysltr1^−/− neutrophils (Fig. 3m), confirming a role for CysLTR1 in promoting tumor NET formation.

To test whether CD8⁺ T cells mediate tumor control in Cysltr1^−/− mice, we depleted CD8⁺ T cells in EMT6-bearing hosts. CD8⁺ depletion eliminated the survival advantage and accelerated tumor growth in Cysltr1^−/− mice (Fig. 3n; Supplementary Fig. 2k), confirming CD8⁺ T-cell dependency.

Because MDSCs contribute to resistance to anti-PD-1 therapy²⁷ and CysLTR1 supports PMN-MDSC suppression (Fig. 3f, g; Extended Data Fig. 4j, k), we evaluated combination strategies. Anti-PD-1 monotherapy had minimal effect in EMT6-bearing wild-type mice but significantly reduced tumor growth and prolonged survival in Cysltr1^−/− mice (Fig. 3o; Supplementary Fig. 2l). MTK similarly sensitized tumors to anti-PD-1 (Fig. 3p; Supplementary Fig. 2l), while showing no additional effect in Cysltr1^−/− mice (Extended Data Fig. 4t; Supplementary Fig. 2m), indicating on-target activity. Enhanced efficacy of MTK plus anti-PD-1 was also observed in Myc-Cap prostate tumors, BPS1 melanoma, and ID8-agg ovarian cancer models (Extended Data Fig. 4u–w), indicating that CysLTR1 blockade can mitigate MDSC-driven resistance across multiple tumor models, with potential involvement of other cell types.

Transcriptional regulation of de novo synthesis of neutrophil granules by CysLTR1 in emergency granulopoiesis

To define mechanisms underlying CysLTR1-driven emergency myelopoiesis, we performed bulk RNA-seq on BM-derived Gr1⁺CD11b⁺ cells induced with GM-CSF, GM-CSF + IL-6, or GM-CSF + EMT6 conditioned medium (Supplementary Fig. 1e). Principal component analysis revealed clear separation between Cysltr1^−/− and Cysltr1^+/+ cells under all conditions (Fig. 4a), underscoring CysLTR1’s key role in shaping the myeloid transcriptome. Sixty-two DEGs were shared across the three conditions (Fig. 4b).

Figure 4. Transcriptional regulation of de novo synthesis of neutrophil granules by CysLTR1 during emergency granulopoiesis.

(a-c) Principal component analysis (PCA) (a), Venn plot of DEGs (b), and Gene Ontology (GO) analysis (c) of the bulk RNA-sequencing dataset of BALB/c BM-derived CD11b⁺ Gr1⁺ cells induced in vitro with GM-CSF, GM-CSF + IL-6, or GM-CSF + EMT6-conditioned medium (n=3).

(d) Heatmap of representative DEGs from GO term: Neutrophil Degranulation (n=3).

(e) qPCR analysis of the relative mRNA levels of Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 in BALB/c BM-derived CD11b⁺Ly6G⁺ neutrophils induced in vitro with GM-CSF, GM-CSF + IL-6 or GM-CSF + EMT6-conditioned medium (n=4).

(f) Relative MFI of MPO and S100A8/9 in neutrophils from BM (top), spleen (middle) and tumor (bottom) of E0771 tumor-bearing Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(g) Relative MFI of MPO and S100A8/9 of neutrophils from spleen (top) and tumor (bottom) of E0771 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice (n=4).

(h) Relative MFI of MPO and S100A8/9 of neutrophils from spleen (top) and tumor (bottom) of FVB-MMTV-PyMT mice with or without MTK treatment (n=4).

(i) Bulk ATAC sequencing analysis of BALB/c BM-derived CD11b⁺Ly6G⁺ neutrophils induced in vitro with GM-CSF, GM-CSF + IL-6 or GM-CSF + EMT6-conditioned medium (n=3), depicting the specific peaks of Elane, Mpo, and S100a8.

(j) Normalized counts of Mxd1 and Nfe2 of bulk RNA sequencing dataset of in vitro induced BM-derived CD11b⁺ Gr1⁺ cells (n=3).

(k) qPCR analysis of the relative mRNA levels of Mxd1 and Nfe2 in in vitro induced BALB/c BM-derived neutrophils (n=3).

(l) Western blot analysis of the protein levels of MXD1 and NFE2 in BALB/c BM-derived neutrophils induced in vitro with GM-CSF + IL-6.

(m) ChIP-qPCR analysis confirming the binding capacity of MXD1 and NFE2 on Elane, Mpo, S100a8, and S100a9 in in vitro GM-CSF + IL-6-induced BALB/c BM-derived neutrophils (n=4).

Results are representative of two (e, k-m) or three (g-h) independent experiments. Data are mean ± s.d. Data were analyzed one-way ANOVA (e-h, j, k, m).

Gene Ontology analysis of DEGs highlighted pathways related to innate immunity, cell cycle, and neutrophil degranulation (Fig. 4c; Extended Data Fig. 5a, b). Notably, “Neutrophil Degranulation” was consistently enriched in Cysltr1^+/+ but not Cysltr1^−/− cells, particularly under tumor-associated inflammatory conditions (GM-CSF + IL-6; GM-CSF + EMT6 CM). Hallmark genes linked to neutrophil degranulation were downregulated in Cysltr1^−/− cells across all conditions (Fig. 4d). In parallel, genes associated with “Negative Regulation of Proliferation” and an M2-like signature were reduced, whereas M1-associated genes were increased in Cysltr1^−/− cells (Extended Data Fig. 5c–e), consistent with a shift toward a less suppressive phenotype.

Consistent with prior flow cytometry findings (Extended Data Fig. 3q–s), expression of Cd14, Cd84, Cd300ld, and Tnfrsf23 (DcTRAIL-R1) was decreased in Cysltr1^−/− cells (Extended Data Fig. 5f, g), supporting a role for CysLTR1 in promoting suppressive PMN-MDSC differentiation.

Given the importance of granule cargo proteins in tumor-associated granulopoiesis (Supplementary Fig. 1f), we next examined granule-related gene expression. Cysltr1^−/− Ly6G⁺ neutrophils generated under all induction conditions exhibited reduced expression of key granule-associated genes, including Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9²⁸ (Fig. 4e; Extended Data Fig. 5h). Pharmacologic inhibition with MTK reduced expression of these genes in a dose-dependent manner (Extended Data Fig. 5i).

Flow cytometric analysis confirmed decreased MPO and S100A8/9 protein levels in Ly6G⁺ neutrophils from BM, spleen, and tumors of Cysltr1^−/− mice in both EMT6 and E0771 models (Extended Data Fig. 6a; Fig. 4f). Similar reductions were observed in Cysltr1^f/f Lyz2-cre mice (Fig. 4g) and MTK-treated tumor-bearing mice (Fig. 4h). Decreased MPO and S100A8/9 were also evident in induced BM-derived neutrophils (Extended Data Fig. 6b), and S100A8/9 release was reduced from Cysltr1^−/− cells (Extended Data Fig. 6c). In BM chimeras, tumor-infiltrating neutrophils from Cysltr1^−/− donors expressed lower MPO and S100A8/9 than controls (Extended Data Fig. 6d), and competitive transplantation assays confirmed an intrinsic disadvantage of Cysltr1^−/− cells in generating MPO⁺S100A8/9⁺ neutrophils (Extended Data Fig. 6e). Together, these data indicate that CysLTR1 intrinsically regulates neutrophil granule synthesis and mobilization.

Because inflammatory granule components (e.g., MMP8, MMP9, MPO) and S100A8/9 promote tumor invasion^{29, 30}, we tested whether CysLTR1-dependent neutrophil secretions affect tumor cell migration. Conditioned medium from Cysltr1^+/+ neutrophils enhanced tumor cell invasion, whereas medium from Cysltr1^−/− neutrophils significantly reduced invasion (Extended Data Fig. 6f, g). The MMP inhibitor doxycycline abrogated invasion, confirming a functional role for neutrophil-derived MMPs.

To evaluate tumor-promoting effects independent of adaptive immunity, Cysltr1^+/+ or Cysltr1^−/− neutrophils were adoptively transferred into SCID and NSG mice bearing EMT6 tumors (Supplementary Fig. 1d). Only Cysltr1^+/+ neutrophils enhanced tumor growth and reduced survival (Extended Data Fig. 6h, i; Supplementary Fig. 2n, o), indicating that CysLTR1-dependent neutrophil programs promote tumor progression through both immunosuppressive and paracrine mechanisms.

ATAC-seq analysis revealed reduced chromatin accessibility near granule-associated genes (Elane, Mpo, S100a8) in Cysltr1^−/− Ly6G⁺ neutrophils (Fig. 4i), suggesting transcriptional regulation downstream of CysLTR1. CHIP-X enrichment analysis identified several candidate transcription factors (Extended Data Fig. 6j–l), including SPI1 (PU.1), CEBPB, CEBPE, MXD1, and NFE2. While SPI1 and C/EBP family members regulate emergency granulopoiesis¹⁴, the roles of MXD1 and NFE2 in tumor-associated neutrophils are less defined.

Mxd1 and Nfe2 expressions increased during early neutrophil maturation (Extended Data Fig. 6m) and were reduced in Cysltr1^−/− GMPs (Extended Data Fig. 6n). Both were induced by GM-CSF + IL-6 (Extended Data Fig. 6o). RNA-seq and RT-qPCR confirmed decreased Mxd1 and Nfe2 transcripts in Cysltr1^−/− cells (Fig. 4j, k), with corresponding reductions at the protein level (Fig. 4l; Supplementary Fig. 3c, d).

Knockdown of either Mxd1 or Nfe2 reduced expression of S100a8, S100a9, and granule genes (Extended Data Fig. 6p, q). ChIP-qPCR demonstrated direct binding of MXD1 and, more prominently, NFE2 to Elane, Mpo, S100a8, and S100a9 loci (Fig. 4m), establishing a transcriptional mechanism for granule gene regulation.

Furthermore, Nfe2 knockdown (KD) in Cysltr1^+/+ BM cells reduced granule gene expression, whereas Nfe2 overexpression (OE) restored expression in Cysltr1^−/− neutrophils (Extended Data Fig. 7a). In BM chimeras, Nfe2 KD (Fig. 5a) phenocopied CysLTR1 deficiency: reduced BM cellularity, altered progenitor composition (Extended Data Fig. 7b, c), impaired neutrophil maturation (Extended Data Fig. 7d), decreased Ly6G⁺ differentiation (Extended Data Fig. 7e, f), and lower MPO/S100A8/9 expression (Extended Data Fig. 7g, h). Consequently, Nfe2 KD reduced tumor-infiltrating Ly6G⁺ neutrophils (Fig. 5b), decreased granule protein levels (Fig. 5c), inhibited tumor growth, and improved survival (Fig. 5d, e; Supplementary Fig. 2p). Nfe2 KD in Cysltr1^−/− cells had no additional effect (Extended Data Fig. 7i, j; Supplementary Fig 2q), confirming NFE2 acts downstream of CysLTR1.

Figure 5. NFE2-dependent neutrophil reprogramming regulates tumor progression and CD8⁺ T cell immunity.

(a) qPCR analysis of relative mRNA levels of Nfe2 in peripheral blood of reconstituted mice following transplantation of Cysltr1^−/− or Cysltr1^+/+ BM with or without Nfe2 knockdown (KD) or overexpression (OE) as indicated (n=4).

(b) Frequencies of Ly6G⁺ neutrophils, Ly6C⁺ monocytic cells (left) and Ly6G⁺/Ly6C⁺ cell ratio (right) in tumor-infiltrating CD45⁺ cells of EMT6-bearing reconstituted mice following transplantation of Cysltr1^−/− and Cysltr1^+/+ BM with or without Nfe2 knockdown (KD) or overexpression (OE) as indicated (n=4).

(c) Relative MFI of MPO and S100A8/9 in tumor-infiltrating neutrophils of EMT6 tumor-bearing transplantation mice as b (n=4).

(d) Tumor weights of EMT6 tumor-bearing reconstituted mice as b (n=4).

(e) Tumor growth (left) and survival (right) of EMT6 tumor-bearing reconstituted mice as b (n=4).

(f) Frequencies of CD86⁺ and CD206⁺ myeloid cells (left), and the CD86⁺/CD206⁺ cell ratio (right) in tumor-infiltrating neutrophils of EMT6 tumor-bearing reconstituted mice (n=4).

(g) Frequency of CD8 T cells in tumor CD45⁺ tumor infiltrates of EMT6-bearing reconstituted mice as b (n=4).

(h, i) MFI of CD44, CD69 (p), and frequencies of IFN-γ⁺ cells and TNF-α⁺ cells (q) in tumor-infiltrating CD8⁺ cells of EMT6 tumor-bearing BALB/c reconstituted mice as b.

(j) Suppression of CD8 T cell (T) proliferation by EMT6-infiltrating neutrophils (E) from reconstituted mice at E:T ratio 1:1 (n=4).

(k) Suppression of CD8 T cell (T) proliferation by EMT6-infiltrating neutrophils (E) from reconstituted mice at different E:T ratio as indicated (n=4).

Results are representative of two (a-k) independent experiments. Data are mean ± s.d. Data were analyzed by log-rank test (e (right)), one-way ANOVA (a-d, f-k) or two-way ANOVA (e (left)).

Nfe2 KD also promoted a phenotypic shift from N2 (CD206⁺) to N1 (CD86⁺) neutrophils (Fig. 5f), increased CD8⁺ T cell infiltration (Fig. 5g), enhanced activation (CD44 and CD69) (Fig. 5h), and boosted IFN-γ and TNF-α production (Fig. 5i). Suppressive activity of Ly6G⁺ neutrophils was reduced in a dose-dependent manner (Fig. 5j, k). These changes closely mirrored CysLTR1 deficiency and were reversed by Nfe2 OE (Fig. 5; Extended Data Fig. 7).

Collectively, these findings demonstrate that CysLTR1 orchestrates emergency granulopoiesis through transcriptional regulation of granule biogenesis and mobilization. The CysLTR1-MXD1/NFE2 axis represents a key intrinsic pathway linking inflammatory cues to neutrophil functional programming within the tumor microenvironment.

Single-cell RNA sequence reveals distinct transcription modules of CysLTR1-mediated granulocytic differentiation and granule synthesis

To gain an in-depth understanding of the role of CysLTR1 in granulopoiesis, we performed single-cell RNA sequencing of CD45⁺ cells isolated from BM, spleen (SP), and tumors of EMT6-bearing Cysltr1^+/+ and Cysltr1^−/− mice (Supplementary Fig. 1g). The uniform manifold approximation and projection (UMAP) analysis revealed distinct myeloid clustering patterns across tissues (Fig. 6a), while overall architecture remained comparable between genotypes (Extended Data Fig. 8a).

Figure 6. Single-cell RNA-seq analysis of tumor-associated neutrophils.

(a) Uniform Manifold Approximation and Projection (UMAP) displaying neutrophils from bone marrow (BM, n=3), spleen (SP, n=3), and tumor (Tu, n=3) of EMT6 tumor-bearing mice, colored by tissue origin (left top), cysltr1 status (left bottom), and inferred cluster identity (right), respectively.

(b) Bubble dot plot analysis of the scaled expression of top five signature genes for each cluster. Color intensity indicates the average expression of each gene scaled across all clusters. Dot size represents the percentage of cells in each cluster with more than one read of the corresponding gene.

(c) Proportions of different clusters among BM, SP, and Tu, respectively.

(d) Proportions of Cysltr1^−/− and Cysltr1^+/+ cells in total clusters and each cluster.

(e) Proportions of clusters across BM, SP, and Tu, respectively.

(f) Monocle trajectory analysis of myeloid cells, colored by tissue origin (top left), pseudotime (top right), and cluster identity (bottom). The pseudotime sequence here represented the differentiated state of cells, and the cells on the right side of the tree (dark blue) are less differentiated than those on two branches of the left side (light blue).

(g) Comparison of functional scores (neutrophil maturation and antigen presentation) between Cysltr1^−/− and Cysltr1^+/+ cells within each cluster. Boxes represent the interquartile range (25^th-75th percentiles), the central line indicates the median, whiskers extend to 1.5x the interquartile range, and points beyond the whiskers represent individual outliers.

(h) Signature score of different granules/vesicles (azurophil, specific, gelatinase, and secretory) in each cluster, determined by Z-score.

(i) Heatmap revealing the row-scaled expression of top related regulons across clusters with the indicated DEG counts for each regulon.

(j) Correlation of Mxd1 or Nfe2 expression with the early granulopoiesis signature (left) and late granulopoiesis signature (right).

Data were analyzed by two-sided Wilcoxon rank-sum test (g) or Pearson correlation (j).

Focusing on neutrophils, we identified seven clusters (C0-C6) with distinct transcriptional signatures (Fig. 6b). Two major differentiation branches emerged, originating from BM and SP, respectively (Fig. 6a), suggesting parallel granulopoietic trajectories. Terminal clusters displayed divergent identities: C2 showed a mature granulocyte signature, whereas C6 was enriched for antigen-presentation genes (Fig. 6b; Supplementary Fig. 4a).

PMN-MDSC-associated markers including Cd14, Cd84, Cd300ld, Tnfrsf23 (DcTRAIL-R1) were highly enriched in more differentiated clusters such as splenic C2, and tumoral C4 (Supplementary Fig. 4b), with reduced expression in Cysltr1^−/− samples (Extended Data Fig. 8b). Cysltr1 expression positively correlated with these suppressive markers, particularly Cd84, Cd300ld, and Tnfrsf23 (Extended Data Fig. 8c). Consistent with previous reports linking activated PMN signatures to tumor neutrophil persistence and suppressive function²³, the activated PMN signature was enriched in clusters C2 and C4 (Extended Data Fig. 8d). Cysltr1 expression aligned with activated PMN programs, especially within C2 and C4 (Extended Data Fig. 8d). DcTRAIL-R1⁺ PMN-MDSC signatures were primarily enriched in C4 cells (Extended Data Fig. 8e; Supplementary Fig. 4c).

Functional enrichment analysis revealed “Neutrophil degranulation” in C2 and “Lymphocyte activation” in C6 (Supplementary Fig. 4d). Tissue origin analysis showed C3, C1, and C0 derived mainly from BM; C2 and C6 from spleen; and C4 and C5 enriched in tumors (Fig. 6c; Extended Data Fig. 8f). Importantly, Cysltr1^−/− samples were enriched in C0 and C6, whereas Cysltr1^+/+ samples preferentially populated C1 and C2 (Fig. 6d, e) with sharper distinctions across tissues.

Trajectory analysis using Monocle mapped a putative granulopoietic path beginning with proliferative C3 BM cells (Mki67⁺, Pcna⁺, Top2a⁺) (Fig. 6f; Supplementary Fig. 4e), progressing through C1 and C0, then maturing into splenic C2 and C6 cells. These splenic populations appeared to seed tumor clusters C4 and C5 (Fig. 6f), consistent with splenic contribution to tumor-associated granulopoiesis²⁴.

Neutrophil maturation scoring showed higher scores in C2 and C6 (Fig. 6g). Cysltr1^−/− cells exhibited reduced maturation scores in C2, which was enriched for N2-like signatures (Extended Data Fig. 8g). Conversely, Cysltr1^−/− cells displayed increased cytotoxicity and antigen-presentation scores in C4 and C6, reflecting an N2-to-N1 shift³¹ and enhanced antigen-presenting potential^{32, 33} (Fig. 6g; Extended Data Fig. 8h). In contrast, Cysltr1^+/+ neutrophils exhibited higher recruitment, ECM remodeling, and angiogenesis scores (Extended Data Fig. 8h), supporting a direct contribution of CysLTR1 by neutrophils to tumor progression.

Granule gene programs mapped clearly along differentiation trajectories. Early clusters (C1, C0) expressed azurophil (primary) granule genes, whereas mature C2 cells expressed secondary, tertiary, and secretory granule subsets (Fig. 6h; Extended Data Fig. 8i), consistent with staged granule formation during granulopoiesis^{29, 30}. Expression of S100a8, S100a9, and other granule cargo genes followed similar patterns (Supplementary Fig. 4f). Notably, granule signatures were significantly reduced in Cysltr1^−/− neutrophils across C1, C2, C4, and C6 (Extended Data Fig. 8i), confirming impaired granule synthesis.

Regulon analysis identified degranulation-associated transcriptional networks (Fig. 6i), including established regulators of myelopoiesis such as Cebpb, Spi1, and Cebpe, as well as Mxd1 and Nfe2. Mxd1 was enriched mainly in C2, consistent with a late differentiation role, whereas Nfe2 was broadly expressed across C3-C2 clusters, indicating their nvolvement throughout maturation (Fig. 6i).

Comparison with established early and late granulopoiesis gene sets^34–36 showed early genes enriched in C1 and late genes in C2 and C4 (Supplementary Fig. 4g). Mxd1 correlated strongly with late granulopoiesis signatures, whereas Nfe2 correlated with both early and late programs, albeit more moderately with late signatures (Fig. 6j). Several additional transcription factors (Cebpa, Irf8, Runx1, Junb, Klf2) were modestly reduced in Cysltr1^−/− samples (Extended Data Fig. 8j). Expression patterns suggested partial overlap between Mxd1/Junb and Nfe2/Runx1/Irf5 modules (Supplementary Fig. 4g). More specifically, it has been reported that RUNX1 and KLF6 regulate neutrophil maturation, whereas RELB, IRF5, and JUNB drive neutrophil effector responses, and RFX2 and RELB promote survival³⁷. We analyzed the relationship of these genes with Mxd1 and Nfe2 in our neutrophils (Supplementary Fig. 4h). Both Mxd1 and Nfe2 correlated with maturation (Runx1, Klf6) and effector-response signatures (Relb, Irf5, Junb), but not survival programs (Rfx2, Relb)³⁷ (Extended Data Fig. 8k, l). Consistently, Mxd1 exhibited a similar expression pattern to Junb, while the expression pattern of Nfe2 more closely resembled that of Runx1 and Irf5 (Supplementary Fig. 4h, i).

Transition-state regulon analysis defined five transcriptional modules: early granulocytic, late granulocytic, monocytic-like, antigen-presentation, and tumor granulocytic (Supplementary Fig. 4j). Mxd1 was central to late granulopoiesis, whereas Nfe2 appeared particularly important for tumor-associated granulocytic differentiation. Late granulocytic scores were higher in Cysltr1^+/+ C2 cells, while Cysltr1^−/− cells showed increased monocytic-like and antigen-presentation module scores (Extended Data Fig. 8m), reflecting functional reprogramming of neutrophils upon Cysltr1 loss.

Extension of single-cell analysis to tumor-infiltrating T cells revealed enrichment of exhausted (C2) and pro-apoptotic (C5) CD8⁺ clusters in Cysltr1^+/+ tumors, whereas cytotoxic and IFN-responsive clusters (C3, C6) predominated in Cysltr1^−/− tumors (Extended Data Fig. 8n–p). Exhaustion markers were consistently higher in Cysltr1^+/+ T cells (Extended Data Fig. 8q).

These findings delineates distinct transcriptional modules through which CysLTR1 promotes granulocytic maturation, granule gene programs, and tumor-associated suppressive phenotypes. Together, these data link CysLTR1-dependent regulation of MXD1/NFE2-centered networks to enhanced neutrophil granule synthesis, tumor-promoting differentiation states, and impaired antitumor T cell immunity in tumor-bearing hosts.

STAT3-mediated upregulation of CysLTR1 facilitates granulocytic differentiation and maturation via activation of AKT and ERK pathways

CysLTR1 signaling through leukotriene D4 (LTD4) activates AKT and MAPK pathways in inflammatory settings³⁸. We therefore examined whether these pathways mediate CysLTR1-dependent transcriptional programming in neutrophils. Phosphorylation of AKT and ERK1/2 was reduced in induced Cysltr1^−/− myeloid cells compared with Cysltr1^+/+ cells, even after LTD4 stimulation (Fig. 7a, Extended Data Fig. 9a).

Figure 7. STAT3-mediated upregulation of CysLTR1 promotes neutrophil granule/vesicle gene synthesis via activation of AKT and ERK pathways.

(a) Frequency of p-AKT⁺ (left) and p-ERK1/2⁺ (right) cells in GM-CSF+IL-6 induced BM-derived Cysltr1^−/− and Cysltr1^+/+ neutrophils upon LTD4 stimulation over time (n=4).

(b) Western blot analysis of the protein levels of MXD1, NFE2, and MPO in GM-CSF+IL-6 induced BM-derived Cysltr1^−/− and Cysltr1^+/+ neutrophils with or without AKT inhibitor (AKTi) or ERK inhibitor (ERKi).

(c) qPCR analysis of relative mRNA levels of Mxd1 (left) and Nfe2 (right) in GM-CSF+IL-6 induced BM-derived Cysltr1^−/− and Cysltr1^+/+ neutrophils with or without AKTi or ERKi (n=4).

(d) qPCR analysis of relative mRNA levels of Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 in GM-CSF+IL-6 induced BM-derived Cysltr1^−/− and Cysltr1^+/+ neutrophils with or without AKTi or ERKi (n=4).

(e) qPCR analysis of relative mRNA levels of Cysltr1 in cytokine-induced BM-derived Cysltr1^−/− and Cysltr1^+/+ neutrophils as indicated (n=4).

(f) qPCR analysis of relative mRNA levels of Cysltr1 in GM-CSF-treated WT BM-derived neutrophils with or without anti-GM-CSF (left) or anti-CSF2Rα (right) antibodies (n=4).

(g) qPCR analysis of relative mRNA levels of Cysltr1 in neutrophils from BM (left) or tumor (right) of EMT6 tumor-bearing mice with or without anti-GM-CSF and anti-IL-6 neutralizing antibodies (n=6).

(h) Flow cytometry analysis of expression levels of CysLTR1 and pSTAT3 in 3 subsets of in vitro GM-CSF + IL-6-induced BM-derived neutrophils based on differential expression of CSF2Rα (n=4).

(i) qPCR analysis of relative mRNA levels of Cysltr1 in in vitro GM-CSF + IL-6-induced BM-derived neutrophils treated with or without WP1066 (a STAT3 inhibitor) (n=4).

(j) ChIP-qPCR analysis illustrating the binding capacity of STAT3 on the promoter region of Cysltr1 (n=4).

(k) qPCR analysis of relative mRNA levels of Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 in in vitro GM-CSF + IL-6-induced BM-derived neutrophils treated with or without WP1066 (n=4).

Results are representative of two (a-e) or three (f-k) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (g), or one-way ANOVA (a, c-f, i-k).

Pharmacologic inhibition of AKT or ERK significantly decreased MXD1 and NFE2 expression at both mRNA and protein levels (Fig. 7b, c, Supplementary Fig. 3e, f), along with reduced MPO expression (Fig. 7b, Supplementary Fig. 3f). In parallel, expressions of S100a8, S100a9, and downstream granule cargo genes were markedly suppressed upon AKT or ERK inhibition (Fig. 7d), recapitulating the transcriptional effects observed with CysLTR1 ablation. These results suggest that CysLTR1 activation mediates the transcriptional regulation of neutrophil granule synthesis and mobilization via an AKT/ERK pathway.

Although GM-CSF + IL-6 robustly induced Cysltr1 expression, GM-CSF alone also increased Cysltr1 mRNA through its receptor CSF2RA (Fig. 7e, f), albeit to a lesser extent, suggesting cooperative cytokine signaling. In vivo neutralization of GM-CSF and IL-6 in tumor-bearing mice reduced Cysltr1 expression in BM and tumor neutrophils (Fig. 7g) and slowed tumor growth while prolonging survival (Extended Data Fig. 9b; Supplementary Fig. 2r).

Given the established role of STAT3 downstream of IL-6, we examined the relationship among CysLTR1, CSF2RA, and STAT3. Expression of CysLTR1 correlated positively with phosphorylated STAT3 (pSTAT3) and CSF2RA in GM-CSF + IL-6-induced cells (Fig. 7h). CSF2RA^hi neutrophils exhibited higher levels of pSTAT3 and CysLTR1 protein and transcripts (Extended Data Fig. 9c, d), together with increased Mxd1, Nfe2, S100a8, S100a9, and granule cargo gene expression compared with CSF2RA^low cells in both BALB/c (Extended Data Fig. 9e) and C57BL/6 backgrounds (Extended Data Fig. 9f–h). These findings suggest that STAT3 contributes to Cysltr1 transcription during cytokine-driven granulocytic differentiation.

Consistently, treatment with the STAT3 inhibitor WP1066 markedly reduced Cysltr1 expression (Fig. 7i). ChIP-qPCR demonstrated STAT3 binding to the Cysltr1 promoter in GM-CSF + IL-6-induced myeloid cells (Fig. 7j), which was abrogated by WP1066 (Fig. 7j). WP1066 also dose-dependently decreased S100a8, S100a9, and granule cargo gene expression (Fig. 7k).

Overall, STAT3-driven upregulation of Cysltr1 under inflammatory conditions fuels emergency granulopoiesis and granule-gene regulation via AKT and ERK activation.

CysLTR1 expression and neutrophil granule formation correlate with patient outcomes in human cancers

Similar to the mouse cell system, we differentiated myeloid cells from human PBMCs using GM-CSF and IL-6 induction³⁹ with or without CYSLTR1 knockdown (Fig. 8a). CYSLTR1 knockdown reduced CD15⁺CD11b⁺ neutrophil differentiation and lowered the CD15⁺/CD14⁺ ratio (Fig. 8b), accompanied by decreased S100A8/9 and MPO expression (Fig. 8c). Similar effects were observed in cells from patient donors (Fig. 8d) and upon MTK treatment (Extended Data Fig. 10a, b). CYSLTR1 knockdown also impaired the T cell-suppressive capacity of induced CD15⁺CD11b⁺ neutrophils (Fig. 8e, Extended Data Fig. 10c). MTK treatment promoted a phenotypic shift from N2 (CD206) toward N1 (CD86) (Extended Data Fig. 10d) and enriched early differentiation stages (PM, MC) while reducing later stages (MM, BC, SN) (Extended Data Fig. 10e).

Figure 8. CysLTR1 expression and neutrophil granule/vesicle synthesis correlate with patient outcomes in human cancers.

(a) qPCR analysis of relative mRNA levels of CYSLTR1 of human PBMC with or without CYSLTR1 knockdown (KD) (n=4).

(b) Frequencies of GM-CSF + IL-6 induced CD15⁺ neutrophils, CD14⁺ monocytic cells (left) and CD15⁺/CD14⁺ cell ratio (right) in CD45⁺ cells from human PBMC with or without CYSLTR1 knockdown (KD) (n=4).

(c) Relative MFI of MPO and S100A8/9 in GM-CSF + IL-6 induced neutrophils from human PBMC with or without CYSLTR1 knockdown (KD) (n=4).

(d) CD15⁺/CD14⁺ cell ratio (left) and relative MFI of MPO and S100A8/9 (right) in CD45⁺ cells GM-CSF + IL-6 induced from human PBMC of ovarian cancer patients with or without CYSLTR1 knockdown (KD) (n=3).

(e) Suppression of CD8 T cells (T) proliferation by GM-CSF + IL-6 induced neutrophils (E) from human PBMC of ovarian cancer patients with or without CYSLTR1 knockdown (KD) at an E:T ratio of 1:1 (n=3).

(f) Flow cytometry analysis of expression levels of CysLTR1 and pSTAT3 in GM-CSF + IL-6-induced neutrophils from human PBMC with differential expression of CSF2Rα (n=4).

(g) Frequency of GM-CSF + IL-6-induced CD15⁺ neutrophils, CD14⁺ monocytic cells (left) and CD15⁺/CD14⁺ cell ratio (right) in CD45⁺ cells from human PBMC with or without MXD1 or NFE2 knockdown (n=4).

(h) Frequency of CD86⁺, or CD206⁺ cells (left) and CD86⁺/CD206⁺ cell ratio (right) in GM-CSF + IL6-induced neutrophils from CD45⁺ human PBMC with or without MXD1 or NFE2 knockdown (n=4).

(i) Relative MFI of MPO and S100A8/9 in GM-CSF + IL-6-induced neutrophils from human PBMC with or without MXD1 or NFE2 knockdown (n=4).

(j, k) Overall survival curve of patients with high (left) or low (right) CYSLTR1 level related to CTL abundance from METABRIC breast cancer dataset (j), or TCGA melanoma dataset (k).

(l) Correlation of CYSLTR1 expression and granule synthesis and mobilization (GSM) score in BRCA and LUSC patients from TCGA database.

(m) Correlation between CYSLTR1 expression and GSM score across a variety of human cancers.

(n) Survival analysis of BRCA and LUSC patients from TCGA database with comparison between high and low expression of GSM score.

(o) Comparison of CYSLTR1 expression between responding (CR/PR) patients and non-responding (SD/PD) to immune checkpoint therapy from 2 published datasets (GSE78220, n=28 and PHS000452, n=122). The central dotted line represents the median. The two additional dotted lines denote the first and third quartiles (25th and 75th percentiles).

(p) Survival analysis of bladder cancer patients treated with anti-PD-L1 therapy from Mariathasan2018 dataset⁴² with comparison between high and low expression of GSM score.

(q) Overall survival (OS) curves of patients with high (left) or low (right) CYSLTR1 level related to CTL abundance from VanAllen 2015 melanoma⁴³ treated with anti-CTLA4 dataset.

(r) Progression-free survival (PFS) curves of patients with high (left) or low (right) CYSLTR1 level related to CTL abundance from Liu 2019 melanoma⁴¹ treated with anti-PD1 dataset.

Results are representative of two (a-i) independent experiments. Data are mean ± s.d. Analyses (j, k, q, r) were done with tumor immune dysfunction and exclusion (TIDE) tools (http://tide.dfci.harvard.edu). Data were analyzed by two-tailed unpaired t-test (a, b (right), d (left), e), log-rank test (j, k, n, p-r), two-sided Mann-Whitney test (o), Pearson correlation (l, m), or one-way ANOVA (b (left), c, d (right), g-i).

Consistent with mouse data, expression of CysLTR1 correlated positively with pSTAT3 and CSF2RA in induced human CD15⁺ neutrophils (Fig. 8f). Knockdown of MXD1 or NFE2 impaired CD15⁺ neutrophil differentiation (Fig. 8g), promoted an N1-like phenotype, reduced N2 markers (Fig. 8h), and decreased MPO and S100A8/9 production (Fig. 8i), supporting a conserved CysLTR1-MXD1/NFE2 axis.

Multiplex immunohistochemistry in triple-negative breast cancer (TNBC) tissues (Extended Data Fig. 10f) identified CD15⁺CD11b⁺ neutrophils with differential CysLTR1 and S100A9 expression. S100A9 intensity was higher in CD15⁺CD11b⁺CysLTR1⁺ cells than in CysLTR1⁻ counterparts (Extended Data Fig. 10g). CD8⁺ T cells were more abundant around CysLTR1⁺ neutrophils (Extended Data Fig. 10g). Similar correlations were observed in publicly available datasets across multiple cancers (Extended Data Fig. 10h, i).

Analysis of TCGA BRCA and LUSC datasets showed that higher CysLTR1 expression correlated with worse overall survival (Extended Data Fig. 10j), consistent across additional tumor types (Extended Data Fig. 10j). In breast and melanoma cohorts, low CysLTR1 expression was associated with improved survival and increased CTL abundance (Fig. 8j, k). CYSLTR1 expression correlated positively with myeloid lineage markers (ITGAM, CD14, FUT4 and FCGR3A) (Extended Data Fig. 10k, l), myelopoietic transcription factors (Extended Data Fig. 10m), and neutrophil activation genes (Extended Data Fig. 10n).

We established a putative granule synthesis and mobilization (GSM) score (Supplementary Table 2), which correlated with CYSLTR1 expression in BRCA, LUSC (Fig. 8l), and other tumor types (Fig. 8m). Lower GSM scores were associated with improved survival in BRCA and LUSC (Fig. 8n). In immunotherapy datasets (GSE78220⁴⁰ and PHS000452⁴¹), responders with complete response (CR) or partial response (PR) exhibited lower CYSLTR1 expression than non-responders with stable disease (SD) or progressive disease (PD) (Fig. 8o). In bladder cancer patients treated with anti-PD-L1 therapy⁴², lower GSM scores were associated with improved survival (Fig. 8p). Across melanoma cohorts^{41, 43}, lower CysLTR1 expression correlated with better survival following immune checkpoint blockade and increased CTL abundance (Fig. 8q, r).

Together, these findings reveal a conserved CysLTR1-mediated immune-evasion pathway and support its potential as an immunotherapy target.

DISCUSSION

CysLTs, potent inflammatory lipid mediators^{44, 45}, are produced predominantly by innate immune cells and drive a range of proinflammatory responses, functioning as key mediators in asthma, allergic rhinitis, and other inflammatory conditions primarily through CysLTR1 activation^{46, 47}. Our findings identify CysLTR1 as a central node linking leukotriene signaling to granulocytic lineage progression, granule synthesis, and immune suppression in cancer.

Leukotriene engagement of CysLTR1 subsequently activates AKT and ERK pathways, amplifying transcriptional programs required for late-stage neutrophil maturation. Consistent with prior studies highlighting AKT and ERK as critical regulators of granulopoiesis^48–50, our data further suggest that CysLTR1 may influence additional myeloid behaviors, including chemotaxis of CysLTR1 expressing neutrophils and monocytes noted in inflammatory diseases^{51, 52}—processes likely relevant to tumor progression.

Functionally, loss of CysLTR1 limits neutrophil accumulation and suppressive activity while restoring intratumoral CD8⁺ T cell infiltration and effector function. Across global and myeloid specific knockouts and BM chimera models, Cysltr1 deficiency exerts its strongest effects on neutrophils. Neutrophil depletion phenocopied these effects, and adoptive transfer further confirmed the neutrophil centric mechanism.

Mechanistically, CysLTR1 drives a transcriptional program for neutrophil granule synthesis governed by MXD1 and, most prominently, NFE2. These transcription factors directly regulate hallmark granule cargo genes (Elane, Mpo, Mmp8/9, S100a8/a9). NFE2 knockdown impaired granulopoiesis and granule production, whereas NFE2 overexpression rescued defects in Cysltr1-deficient chimera models. ATAC-seq data support CysLTR1-dependent chromatin accessibility at granule loci. Limiting granule formation likely reduces the direct tumor promoting impact of neutrophil secretions and reshapes myeloid-T cell interactions. Consistent with this, CysLTR1 ablation inhibited tumor growth in both immunodeficient and immunocompetent settings and enhanced antigen presentation features and an immunostimulatory myeloid phenotype.

Upstream, tumor-derived cytokines including GM-CSF and IL-6 activate STAT3, which binds the Cysltr1 promoter and upregulates its transcription, linking inflammatory cues to emergency granulopoiesis via AKT/ERK and MXD1/NFE2 modules. Single cell transcriptomic analysis revealed neutrophil diversity and differentiation trajectories, with distinct regulons implicating NFE2 as a key driver of de novo granule synthesis during tumor progression. The phenotype continuum observed aligns with recently described neutrophil subsets in mice and humans^{15, 53, 54}.

Clinically, higher CYSLTR1 expression and a neutrophil granule synthesis/mobilization signature correlate with poorer survival and reduced response to immune checkpoint blockade. These data support combining CysLTR1 antagonists with anti-PD (L)1 to “disarm” neutrophils by attenuating granule biogenesis and protumoral functions during differentiation (see the model in Extended Data Fig. 10o). Given the pleiotropy of S100A8/9 and granule cargoes, pathway level blockade of CysLTR1 may be more effective than targeting individual mediators. A neutrophil disarming strategy merits clinical evaluation as an adjunct to cancer immunotherapy.

METHODS

Ethics statement

All animal experiments were conducted in accordance with relevant institutional and national guidelines and were approved by the Institutional Animal Care and Use Committee (IACUC) of Northwestern University (Chicago, IL) (approval no. IS00013856). The use of archival human specimens was approved by the Institutional Review Board Office of Northwestern University (Chicago, IL) (approval no. STU00211852).

Reagents

Complete reagent names, suppliers and catalog numbers are detailed in Supplementary Table 1. The cell culture media, cytokines, small molecules, plasmids, and sequencing library preparation kits were used according to the manufacturers’ instructions unless otherwise specified.

Cell lines

YUMMER1.7 cell line was provided by Dr. Marcus Bosenberg (Yale University); Myc-CaP cell by Dr. Sarki A Abdulkadir (Northwestern University). BPS1⁵⁵ and ID8agg⁵⁶ cells were described previously. All other tumor cell lines were obtained from the ATCC. All lines were mycoplasma-free (Biotool) and used within 10 passages after thawing. Cells were cultured in complete medium consisting of RPMI1640 supplemented with 5% FBS, 1% penicillin/streptomycin (Thermo Fisher Scientific) and 1% MEM non-essential amino acid solution (Gibco) at 37°C and 5% CO₂. For conditional medium, tumor cells were replated, and their supernatants were collected after 24 hours. These supernatants were then mixed with fresh media at a 1:1 ratio to generate the conditional medium.

Experimental animals

C57BL/6J (000664), BALB/cJ (000651), B6 CD45.1 (B6.SJL-Ptprca Pepcb/BoyJ) (002014), NOD scid (NOD.Cg-Prkdcscid/J) (001303), NSG (NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ) (005557), FVB/N-TgN (MMTV-PyMT)^634Mul-transgenic strain (002374), Lyz2-cre (B6.129P2-Lyz2tm1(cre)Ifo/J) (004781) and Cysltr1^−/− strains (C57BL/6N-Cysltr1tm1Ykn/J, and C.B6-Cysltr1tm1Ykn/J) (030814, and 030960), were purchased from Jackson Laboratories. Cysltr1-flox mice were developed in the Transgenic & Targeted Mutagenesis Laboratory and crossed with Lyz2-cre mice in barrier facilities at Northwestern University. Cysltr1-flox mice generated in this study can be shared from the lead contact upon request. B6 CD45.1⁺CD45.2⁺ mice were bred by crossing C57BL/6J and B6 CD45.1 mice. Mice were housed under specific pathogen-free conditions with a 12-hour light/12-hour dark cycle, controlled ambient temperature (22±2 °C), and relative humidity of 40–60%, with ad libitum access to food and water. Phenotypic analysis and tumor challenges were conducted on mice aged between 8 to 12 weeks. No significant differences were observed between male and female mice in terms of tumor growth or immune responses. Thus, results were pooled from both sexes without further stratification by gender.

Tumor challenge and treatments

Tumor cells (10⁶) were injected subcutaneously unless otherwise indicated. ID8-agg cells (5×10⁶) were intraperitoneally injected. MycCaP cells (10⁶) were mixed with Matrigel before implantation. Depleting antibodies (anti-CD8a or anti-Ly6G) were administered intraperitoneally (5 mg/kg) starting 1 day before tumor inoculation and continued twice weekly. For checkpoint treatment, 7 days after tumor inoculation, mice were randomly assigned to control or treatment groups and received intraperitoneal injection of isotypes (150 μg per mouse), anti-PD-1 (150 μg per mouse), Montelukast (MTK, 10 mg/kg, daily), a CysLTR1 inhibitor, or both of anti-PD-1 and MTK. Antibody treatments were conducted every 3 days. For the MMTV-PyMT model, mice with palpable tumors at 7–8 weeks old received treatment with control vehicles or MTK (10 mg/kg, daily) until reaching the experimental endpoint. Tumor volumes were measured every 2–3 days. Measurements were taken along two perpendicular axes (a and b), and tumor volumes were calculated using the formula: a×b²/2. Tumor volumes were maintained within the limits (≤ 2.5 cm³) approved by the IACUC. In some cases, this limit has exceeded at the last day of measurement, and the mice were immediately euthanized.

LPS-induced emergency myelopoiesis

Mice were intraperitoneally injected with 35 μg ultrapure LPS (E. coli, InvivoGen) in sterile PBS to induce emergency myelopoiesis. Control mice received PBS following the same schedule. A second injection was administered 48 h after the first, and mice were euthanized 24 h later for tissue collection. Emergency myelopoiesis was assessed by flow cytometry.

BM reconstitution

BM chimeric mice were generated as described previously⁵⁷. Briefly, Cysltr1^−/− mice and Cysltr1^+/+ mice were exposed to 10 Gy irradiation followed by 1×10⁷ donor BM cells intravenously. Reconstitution was verified by flow cytometry. Mice were housed for 8 weeks after BM transplantation before experimentation. For BM competitive reconstitution, irradiated recipient CD45.1⁺CD45.2⁺ mice received BM cells from CD45.2⁺ Cysltr1^−/− and CD45.1⁺ Cysltr1^+/+ mice with a 1:1 mixture. Mice were closely monitored, and peripheral blood was collected weekly to measure BM chimerism by flow cytometry.

In vitro induction of neutrophils by proinflammatory cytokines

Murine BM-derived myeloid cells were induced in vitro as described previously⁵⁸. In brief, BM cells from both Cysltr1^−/− mice and Cysltr1^+/+ mice were induced with GM-CSF (40 ng/ml) alone, GM-CSF+IL-6 (40 ng/ml), or GM-CSF+EMT6 or E0771 tumor cell conditional medium in the presence or absence of MTK at 20 μM for 4 to 5 days. BM-derived myeloid cells were enriched using CD11b selection kit or Gr1 MicroBeads, while neutrophils were enriched using Ly6G MicroBeads. The cytokine-induced myeloid cells including neutrophils (CD11b⁺CD15⁺CD14⁻CD33⁺) and monocytic cells (CD11b⁺CD14⁺HLADR^−/low) subsets from human PBMCs of healthy donors and ovarian cancer patients were generated as described previously³⁹.

Flow Cytometry and cell sorting

The flow cytometry analysis was conducted as previously described⁵⁷. In brief, tumor tissues were enzymatically digested (Collagenase IV (1 mg/mL)/DNase I (20 μg/mL)), filtered, and RBC-lysed. Cells were blocked with Fc receptors and stained with fluorochrome-conjugated antibodies (see Supplementary File for antibody list). For intracellular cytokine staining, tumor samples were restimulated with a combination of PMA (10 ng/mL), ionomycin (100 ng/mL), and Golgi Stop. After surface staining, cells were fixed and permeabilized, followed by intracellular staining. Data acquisition was carried out using a BD LSR II analyzer. Alternatively, cells were sorted using a BD FACS Aria II sorter. CD11b⁺Gr1⁺ cells were isolated by a FACS sorter from dissociated tumor tissues. Granulocytes-depleted samples were generated by removing the entire CD11b⁺Gr1⁺ population from tumor single-cell suspensions via FACS prior to CysLT quantification. Dead cells were excluded using Live/Dead Fixable UV/Yellow/Red as per the manufacturer’s instructions. Data analysis was performed using FlowJo software. The gating strategy for all tested subpopulations and activation markers was provided in Supplementary Fig. 5.

Colony forming assay

BM cells were resuspended in Methylcellulose-based Media at a concentration of 1–2×10⁴ cells/mL, plated and cultured at 37°C with 5% CO₂. Colony types and total cell numbers were assessed on day 7. Cells were then resuspended and replated for second and third rounds on days 7 and 14, respectively, with colony counts and total cells measured each time.

T cell-suppression assay & priming assay

T cell-suppression assay was performed as described previously⁵⁸. In brief, purified splenic CD8⁺ T cells were initially labeled with the proliferation dye eFluor450, and subsequently stimulated in vitro with anti-CD3/CD28 in the absence or presence of in vitro-induced myeloid cells or cells isolated from tumor tissues at various indicated ratios for 3 days. For T cell priming assay, proliferation dye labeled splenic purified CD8⁺ T cells from OTI mice were co-cultured with Ly6G⁺CD11b⁺ neutrophils sorted from LLC1-OVA tumor tissues for 3 days. T cell proliferation and activation was measured by flow cytometric eFluor450 dye dilution and activation markers.

Gene knockdown and overexpression

BM cells were harvested and transfected with the specific siRNA of either Mxd1 or Nfe2 by electroporation using the Neon transfection system (Thermo Fisher Scientific) following the manufacturer’s instructions. The parameters employed were a pulse voltage of 1400 V, pulse width of 10 ms, and three pulses. For long-term knockdown and overexpression of Nfe2, the corresponding GFP-tagged Piggybac vector was combined with the PBase vector for transfection according to the manufacturer’s guidelines. Electroporation was conducted using the aforementioned parameters. Transfection efficiency was assessed by flow cytometry based on GFP (6–10%), followed by GFP sorting to enrich Nfe2 knockdown (KD) or overexpression (OE) cells. Similar gene knockdown was performed in freshly isolated PBMCs from healthy donors or ovarian cancer patients (n>3) with siRNA targeting CYSLTR1, MXD1, or NFE2. The electroporation parameters were 2150 V, 20 ms, 1 pulse.

NETosis measurement

BM-induced or tumor-infiltrating neutrophils were harvested and seeded onto poly-L-lysine-coated coverslips in 24-well plates, in the presence or absence of 100 nM PMA. Cells were subsequently fixed with 4% paraformaldehyde (PFA) and permeabilized with 0.1% Triton X-100. NETs were stained using anti-MPO and anti-citH3 monoclonal antibodies, while DNA was counterstained with DAPI. Immunofluorescence images were acquired using fluorescence microscopy and ELISA was performed to quantify MPO-DNA complexes. Briefly, 96-well plates were coated overnight at 4 °C with anti-MPO antibody (2 μg/mL; 0400–0002, Bio-Rad) and blocked with 2.5% BSA in PBS for 2 h at room temperature. Plates were washed and incubated for 90 min with pre-treated mouse BM cells or tumor infiltrated neutrophils diluted in blocking buffer. Plates were then incubated for 90 min with anti-DNA antibody (1:10; Cell Death Detection ELISA, 11544675001, Sigma-Aldrich). Plates were then developed with ABTS substrate (10102946001, Sigma-Aldrich). CitH3 levels were measured using a commercial ELISA kit (Cayman Chemical, Ann Arbor, MI, USA).

Splenectomy

Splenectomy was performed on tumor-bearing mice on day 24 post-tumor implantation, when tumors reached ~1000 mm³. Mice were anesthetized and maintained under sterile conditions. A 2-cm left subcostal lateral laparotomy was made, and the spleen exteriorized. The splenic vessels were ligated with 6–0 coated vicryl sutures (Ethicon), and the spleen was removed. The laparotomy was closed in two layers, and mice were administered subcutaneous Hartmann’s solution and tramadol for hydration and analgesia. Sham surgeries were performed similarly, except the spleen was returned to the abdominal cavity without removal.

In vivo GM-CSF and IL-6 neutralization

EMT6-bearing mice were treated with 150 μg of anti-GM-CSF and anti-IL-6 antibodies or isotype controls via intraperitoneal injection every 7 days, starting when tumors became palpable (~day 7). Tumors were harvested on day 14, and single-cell suspensions were prepared for downstream analyses. CD11b⁺Ly6G⁺ neutrophils were sorted for Cysltr1 mRNA quantification by qPCR.

ELISA assay

Supernatants collected from neutrophils were utilized to measure S100a8/9 release levels via ELISA kits, following the manufacturer’s guidelines (eBioscience, San Diego, CA, USA). The levels of CysLTs were measured via ELISA kits (Cayman Chemical, Ann Arbor, MI, USA). The levels of G-CSF, GM-CSF and IL-6 were measured via ELISA kits (Biolegend, San Diego, CA, USA).

ELISPOT assay

Tumor infiltrating CD45⁺ cells or splenocytes obtained from EMT6- or E0771-bearing mice were collected between day 14 and day 21 post-tumor cell injection. These cells were then stimulated ex vivo with/without anti-CD3 or specified peptides at a concentration of 2×10⁵ cells per well. After a 24-hour incubation period, an ELISPOT assay was conducted according to the manufacturer’s instructions (eBioscience), as described previously⁵⁶.

Transwell assay

Tumor cell suspensions were added to the top chamber of each Matrigel-coated transwell insert in the presence or absence of conditional medium of GM-CSF+IL-6-induced BM-derived Cysltr1^−/− or Cysltr1^+/+ neutrophils with/without Doxycycline (0.1 μg/mL or 1 μg/mL), while 500 μL of complete medium was added to the bottom chamber of each well for 24 hours culture. Following incubation, non-invaded cells atop the transwell insert were removed using a cotton swab. The invaded cells on the bottom of the transwell insert were fixed with 4% paraformaldehyde for 15 minutes, then stained with crystal violet per the manufacturer’s protocol. The invaded cells were observed and imaged using a microscope.

RNA extraction and real-time qPCR

Total RNA extraction was conducted following the manufacturer’s protocol using Trizol reagent (Invitrogen). Subsequently, cDNA synthesis was carried out utilizing SuperScript One-Step RT-PCR (Invitrogen). Quantitative real-time PCR was employed to measure gene expression levels using SYBR Green (Bio-Rad

Bulk RNA seq analysis

Total RNA was extracted using the Direct-zol RNA mini prep kit (Zymo Research), following the manufacturer’s protocol. The extracted RNA served as the basis for constructing sequencing libraries. The NEBNext Poly(A) mRNA Magnetic Isolation Module, Ultra II RNA Library Prep Kit, and Multiplex Oligos (NEB) were utilized according to the manufacturer’s instructions. Sequencing was performed on an Illumina NovaSeq 6000, generating 50 bp single-end reads, with an average depth of 30.0 million reads per sample. The obtained clean reads were aligned to the mouse genome (mm10) using HISAT2 version 2.1.0. Quantification of gene expression was carried out employing HTSeq-count version 0.13.5, and normalization of gene expression levels was executed using DESeq2 version 1.34.0. A comprehensive set of 47,727 genes was detected, with extensive transcript coverage and even distribution of reads across the transcripts.

Quantification of gene expression was conducted using the htseq-count package. Differential gene expression analysis was performed utilizing DESeq2. Significantly differentially expressed genes (DEGs) between groups were identified based on an FC>2, FDR-adjusted P value < 0.05. The up- or down-regulated DEGs were subjected to Gene Ontology (GO) enrichment analysis and KEGG pathway analysis using the Metascape platform (metascape.org). Enrichment of Biological Process GO terms and KEGG pathways was determined based on an FDR-adjusted P value < 0.05. DEGs associated with neutrophil degranulation underwent further analysis to elucidate upstream regulator activation using ChEA3 (maayanlab.cloud/chea3/). The top significantly activated transcription factors and upstream regulators are presented. Principal Component Analysis (PCA) plots were generated via a custom R script. Volcano plots, bar plots, and heat maps were created using GraphPad Prism.

Validation datasets used in Fig. 7 include data from TIDE, accessible at http://tide.dfci.harvard.edu/. Further details and analytical methodologies related to TIDE are described in “Tumor Immune Dysfunction and Exclusion”⁵⁹.

Single-cell RNA library preparation

Single cell suspensions were prepared from bone marrow, spleen, and tumor tissues of EMT6-bearing Cysltr1^+/+ or Cysltr1^−/− mice (n=3). Cells were then pooled prior to downstream analysis. Hematopoietic progenitor cells were enriched from bone marrow using the Hematopoietic Progenitor Cell Isolation Kit (Stemcell), while CD45-positive cells were enriched from spleen and tumor samples using the CD45 Positive Selection Kit (Stemcell). Tumor-infiltrating leukocytes were further purified by FACS. Approximately 10,000 cells from each sample were loaded for single-cell RNA-seq library preparation.

Single-cell RNA-seq libraries were generated using the 10x Genomics Chromium Single Cell 3’ v3 platform according to the manufacturer’s instructions. Briefly, single-cell suspensions were loaded onto the Chromium Controller, where individual cells were encapsulated into Gel Bead-in-Emulsions (GEMs) containing barcoded oligonucleotides. Within each GEM, cells were lysed, and mRNA was reverse transcribed to generate barcoded cDNA. Following amplification, sequencing libraries were prepared utilizing the Chromium Single Cell 3′ Library and Gel Bead Kit v3 and the Chromium i7 Multiplex Kit. Single-cell sequencing was conducted using an Illumina NovaSeq 6000 platform. Raw sequencing data was demultiplexed and processed using Cell Ranger version 6.0.0 software. The original data generated in this study are publicly accessible in the GEO database.

Data processing and analysis of scRNA Seq

The quality assessment of sequencing reads was conducted using FastQC and MultiQC. Raw FASTQ files underwent processing by Cell Ranger 7.2.0 with the mm10 reference genome. This generated a gene expression matrix for each sample, capturing the number of UMIs associated with each gene for every cell barcode. Downstream analyses were performed using R v4.3.2 and Seurat v5.0.0, unless stated otherwise.

Mouse neutrophils typically exhibit a median of ~1,455 UMIs and ~515 detected genes per cell³⁴. Based on these benchmarks, we applied commonly used droplet-based quality control (QC) criteria—excluding cells with >5% mitochondrial transcripts or fewer than 500 total transcripts—to remove empty droplets and damaged cells. Cells showing transcript numbers above the 99th percentile, potentially indicative of doublets, were also excluded. After QC, 47,430 high-quality single cells—including intratumoral CD45⁺ immune cells—were retained, among which 10,045 cells were identified as neutrophils.

Normalized counts per ten thousand (CP10K) were employed, unless specified otherwise. Gene set scores were calculated by determining the average expression of genes within each gene set of interest minus the average expression of a control gene set. Z-scores resulting from this calculation were used as a measure of relative gene expression for each gene. Refer to Supplementary Table 2 for the listed gene sets in this study.

Simple Western Blotting assay

Protein extraction from in vitro cytokine-induced BM-derived neutrophils was conducted, with 1 μg of total protein loaded into each well of the Simple Western blot plate. Experiments were performed using the Bio-Techne Simple Western system following the manufacturer’s protocol. Exposure time was adjusted individually for each target protein to ensure clear and visible imaging. Chemiluminescent signals were detected using the imaging system provided with the Simple Western system.

Multiplexed Immunohistochemistry (mIHC)

The TNBC tissue array (TMA) (US Biomax Inc) was utilized for mIHC staining of CD11b, CD8, S100A9, CysLTR1 and CD15 by using the Opal multiple color IHC kit (AKOYA Biosciences) as described previously⁶⁰. The de-identified tissue arrays were IRB-approved and fully consented. The current study was approved by the Institutional Review Board Office of Northwestern University (the IRB exemption approval letter will be provided upon request). to the provider, all samples were collected under Institutional Review Board (IRB)-approved protocols with informed consent from donors. As the samples were fully de-identified, no additional IRB approval was required for this study. Briefly, tissue sections were deparaffinized, rehydrated, and subjected to heat-mediated antigen retrieval in AR9 buffer. For each marker, slides were incubated with primary antibodies followed by HRP-conjugated secondary antibodies and visualization using the appropriate Opal fluorophore. After each staining round, antibody complexes were removed by heat treatment in AR6 buffer while preserving the deposited fluorophore signal. This process was repeated sequentially for all targets. After completion of staining cycles, nuclei were counterstained with DAPI. Single-stained control slides for each fluorophore were used to generate spectral reference libraries for downstream unmixing. Antibody details and fluorophore assignments are provided in the Supplementary Information.

Multispectral images acquisition and analysis

Fluorescent signals from Opal-stained tissue microarrays were imaged using the Vectra 3 automated quantitative pathology imaging system (PerkinElmer) at 200x magnification. Multispectral images were processed using InForm software (Akoya Biosciences) to separate overlapping fluorescence signals based on reference emission spectra generated from single-stained controls. Following spectral separation, nuclei were identified using DAPI staining, and cell segmentation was performed accordingly. Cell phenotypes were assigned using a supervised machine-learning algorithm trained on marker expression patterns within InForm. Quantitative data, including cell counts, marker intensity, and spatial metrics, were exported and analyzed using the R packages phenoptrReports and phenoptr (Akoya Biosciences).

Correlation analysis and survival assessment in publicly available datasets

Expression profiles for CYSLTR1, myelopoiesis-related genes, neutrophil degranulation genes, and patient survival data were sourced from the TCGA database, underwent Z-score normalization. A putative signature score for neutrophil granule synthesis and mobilization (GSM) was computed by subtracting the mean gene abundance from the raw expression for a list of specific genes related to neutrophil GSM (see Supplementary Table 2) and then divided by the standard deviation of all measured counts across the samples. Pearson’s correlation coefficient was calculated to analyze the relationship between CYSLTR1, other genes of interest, and the GSM signature score. ANOVA was utilized to evaluate the significance of these correlations. For survival analysis, Kaplan-Meier curves were generated to depict survival trends associated with varying levels of CYSLTR1 expression and the GSM signature score across multiple cancer types. Additionally, Cox proportional hazards regression analysis was employed to explore the relationship between patient survival and CYSLTR1 expression or the GSM signature score.

QUANTIFICATION AND STATISTICAL ANALYSIS

The data were analyzed using GraphPad Prism Software (version 9.5.1). Mean/median values were compared using a two-tailed Student’s t-test, either unpaired or paired, or Mann-Whitney U-test as appropriate. When more than two groups were assessed, data were determined by two-way analysis of variance (ANOVA) followed by Tukey’s or Sidak’s multiple comparison tests. The statistical differences between the survival of groups were calculated using the log-rank test. No animals or data points were excluded from the analyses. P values greater than 0.05 were considered non-significant. Additional statistical methods are described where appropriate. No statistical methods were used to pre-define sample sizes, sample sizes were selected based on prior experience and are consistent with those reported in similar studies. Data distribution was assumed to be normal, but this was not formally tested. Mice were randomly assigned to treatment groups after tumor inoculation when tumors became palpable. Allocation was performed to ensure comparable tumor sizes across groups at the start of treatment. Data collection and analysis were not performed blind to the conditions of the experiments. Further information on research design is available in the Nature Research Reporting Summary linked to this article.

RESOURCE AVAILABILITY

Data availability

Bulk RNA-seq data generated in this study have been deposited in the Gene Expression Omnibus (GEO) under accession code GSE223123 [https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE223123]. Single-cell RNA-seq data have been deposited in GEO under accession code GSE223315 [https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE223315].

The previously published datasets re-analysed in this study are publicly available in the Gene Expression Omnibus (GEO) database under accession codes: GSE21927 [https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE21927], GSE39807 [https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE39807], GSE223028 [https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE223028], GSE14833 [https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE14833], GSE6506 [https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE6506], GSE78220 [https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE78220] and in the Database of Genotypes and Phenotypes (dbGaP) under accession code phs000452 [https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000452].

TCGA datasets were obtained from the TCGA Research Network (https://www.cancer.gov/tcga). TIDE data were obtained from http://tide.dfci.harvard.edu/.

All other data supporting the findings of this study are available from the corresponding author upon reasonable request. Source data underlying Figs. 1–8 and Extended Data Figs. 1–10 are provided with this paper.

Extended Data

Extended Data Fig. 1. CysLTR1 regulates BM myeloid commitment, neutrophil differentiation, and NET formation in vitro.

(a) Total BM cell counts (from 2 femurs and 2 tibia) of BALB/c Cysltr1^−/− mice and Cysltr1^+/+ wild-type (WT) mice (n=10).

(b) Total cell numbers of LSK (Lin⁻Sca1⁺c-Kit⁺) (left), HPC (Lin⁻Sca1⁻c-Kit⁺) (middle), CMP (Lin⁻Sca1⁻c-Kit⁺CD34⁺CD16/32⁻), GMP (Lin⁻Sca1⁻c-Kit⁺CD34⁺CD16/32⁺) and MEP (Lin⁻Sca1⁻c-Kit⁺CD34⁻CD16/32⁻) (right) per femur in BALB/c Cysltr1^−/− mice and Cysltr1^+/+ mice (n=4).

(c) Total cell numbers of LSK (left), HPC (middle), CMP, GMP and MEP (right) per femur in C57BL/6 Cysltr1^−/− mice and Cysltr1^+/+ mice (n=4).

(d-f) Colony-forming unit assays. Representative images of BM colonies (d), total colony counts of CFU-G (granulocyte), CFU-M (macrophage), CFU-E (erythroid), CFU-GM (granulocyte/macrophage), and CFU-GEMM (granulocyte/erythroid/macrophage/megakaryocyte) colonies from 1^st round culture of BALB/c Cysltr1^−/− mice and Cysltr1^+/+ mice (n=4) (e). Total colony counts (left) and total cell numbers (right) from the 1^st (day 7), 2^nd (day 14), and 3^rd (day 21) round of cultures (n=4) (f).

(g) Apoptosis analysis of HPCs from BALB/c Cysltr1^−/− mice and Cysltr1^+/+ mice (n=4). Early apoptosis, Annexin V⁺7AAD⁻; and late apoptosis, Annexin V⁺7AAD⁺.

(h) Cell cycle analysis of HPCs from BALB/c Cysltr1^−/− mice and Cysltr1^+/+ mice (n=4). G0, Ki67⁻7AAD⁻; G1, Ki67⁺7AAD⁻; and S/G2/M, Ki67⁺7AAD⁺.

(i-k) Frequencies of Ly6G⁺ neutrophil and Ly6C⁺ monocytic cell (left), Ly6G⁺/Ly6C⁺ cell ratio (middle), and frequency of macrophage (Mφ) and dendritic cell (DC) (right) in CD45⁺ cells differentiated from BALB/c Cysltr1^−/− and Cysltr1^+/+ BM cells stimulated with GM-CSF (i), GM-CSF+IL-6 (j), or GM-CSF+EMT6-conditionedmedium (k) in vitro (n=4).

(l, m) Frequencies of myeloblast (MB), promyelocyte (PM), myelocytes (MC), metamyelocytes (MM), blast cell (BC), and segmented neutrophil (SN) subsets in neutrophils differentiated from BALB/c Cysltr1^−/− and Cysltr1^+/+ BM cells stimulated with GM-CSF (l), or GM-CSF+IL-6 (m) in vitro (n=4).

(n) Frequencies of Ly6G⁺ neutrophil and Ly6C⁺ monocytic cell (left), Ly6G⁺/Ly6C⁺ cell ratio (middle), and frequency of macrophage and DC (right) differentiated from sorted BALB/c Cysltr1^−/− and Cysltr1^+/+ GMP cells with GM-CSF+IL-6 in vitro (n=4).

(o) Frequencies of MB, PM, MC, MM, BC, and SN subsets in neutrophils differentiated from sorted BALB/c Cysltr1^−/− and Cysltr1^+/+ GMP cells with GM-CSF+IL-6 in vitro (n=4).

(p) Representative flow dot plots of CD11b⁺Gr1⁺ cells, Ly6G⁺ and Ly6C⁺ cells differentiated from BALB/c Cysltr1^−/− and Cysltr1^+/+ Gr1⁻ BM cells stimulated with GM-CSF + IL-6 in vitro (n=4).

(q) Frequencies of Ly6G⁺ neutrophils and Ly6C⁺ monocytic cells (left), Ly6G⁺/Ly6C⁺ cell ratio (middle), and frequencies of macrophage and DC (right) in CD45⁺ cells differentiated from BALB/c Cysltr1^−/− and Cysltr1^+/+ Gr1⁻ BM cells stimulated with GM-CSF + IL-6 in vitro (n=4).

(r-u) Frequencies of Ly6G⁺ neutrophil and Ly6C⁺ monocytic cell (left), Ly6G⁺/Ly6C⁺ cell ratio (middle), and frequency of macrophage and DC (right) in CD45⁺ cells differentiated from BALB/c Cysltr1^−/− and Cysltr1^+/+ BM cells stimulated with G-CSF (r), G-CSF+IL-6 (s), M-CSF (t), or M-CSF+IL-6 (u) in vitro (n=4).

(v) Immunofluorescence staining of DNA, MPO and CitH3 (left), and quantification (right) of NET-forming Ly6G⁺neutrophils (triple-colocalizing DNA-MPO-citH3 events) from Cysltr1^−/− and Cysltr1^+/+ BALB/c BM cells stimulated with PMA (n=4). Blue: DAPI; Purple: MPO; Red: CitH3.

(w) Quantification of MPO–DNA complexes (n=4) and CitH3 (n=3) by ELISA in Ly6G⁺neutrophils from BALB/c Cysltr1^−/− and Cysltr1^+/+ BM stimulated with PMA.

(x) Quantification of MPO–DNA complexes (n=3) and CitH3 (n=4) by ELISA in Ly6G⁺neutrophils from BALB/c Cysltr1^−/− and Cysltr1^+/+ BM cultured in EMT6 tumor-conditioned medium (TCM) compared to regular medium (Med).

Results are representative of three (a-c, i-m) or two (d-h, n-x) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (a, b (left, middle), c (left, middle), i-k (middle), n (left, middle), q-u (middle), v (right)) or one-way ANOVA (b (right), c (right), e-h, i-k (left, right), l, m, n (left, right), o, q-u (left, right), w, x).

Extended Data Fig. 2. CysLTR1 promotes emergency granulopoiesis, intrinsic myeloid output, and tumor progression across mouse models.

(a) Frequencies of LSK cells (left) and HPC (middle) in lineage-negative BM cells and frequencies of CMP, GMP, and MEP within HPC (right) from E0771 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(b) Frequencies of LSK cells (left) and HPC (middle) in lineage-negative BM cells and frequencies of CMP, GMP, MEP within HPC (right) from MTK-treated EMT6 tumor bearing BALB/c mice (n=4).

(c) Frequencies of Ly6G⁺ neutrophil and Ly6C⁺ monocytic cell (left), Ly6G⁺/Ly6C⁺ cell ratio (middle) and frequencies of macrophage and DC (right) in BM CD45⁺ cells from E0771 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(d) Ly6G⁺/Ly6C⁺ cell ratio in CD45⁺ cells from BM, SP and tumor of MTK-treated EMT6 tumor-bearing mice (n=4).

(e) Frequencies of Ly6G⁺ neutrophils in peripheral blood of EMT6-bearing mice at day 7, 14, and 21 (n=4).

(f-h) Frequencies and cell counts of LSK cells (f) and HPC (g) in lineage-negative BM cells and CMP, GMP, MEP within HPC from LPS-treated BALB/c mice (n=4).

(i) Frequencies of MB, PM, MC, MM, BC, and SN subsets in BM neutrophils from LPS-treated BALB/c mice (n=4).

(j) Frequencies (left), cell counts (middle) of Ly6G⁺ neutrophils and Ly6C⁺ monocytic cells, and Ly6G⁺/Ly6C⁺ cell ratio (right) from BM of LPS-treated BALB/c mice (n=4).

(k) Tumor weights of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=8). The top and bottom solid lines indicate the maximum and minimum values, respectively. The central dotted line represents the median. The two additional dotted lines denote the first and third quartiles (25th and 75th percentiles).

(l) Individual tumor growth curves of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=8).

(m) Tumor weights of CT26 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=8).

(n) Tumor growth curve and survival of CT26 tumor-bearing Cysltr1^−/− and Cysltr1^+/+ mice (n=8).

(o) Tumor weight of YUMMER1.7 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ mice (n=8).

(p) Tumor growth curve and survival of YUMMER1.7 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ mice (n=8).

(q) Tumor weights of FVB-MMTV-PyMT mice with or without montelukast (MTK) treatment (n=4).

(r) Tumor weights of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ BM-transplanted mice (n=8).

(s) Tumor weights of E0771 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ BM-transplanted mice (n=8).

(t) Frequencies of Ly6G⁺ neutrophils, Ly6C⁺ monocytic cells (left), Ly6G⁺/Ly6C⁺ cell ratio (middle) and frequencies of macrophage and DC in tumor infiltrating CD45⁺ cells of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ BM-transplanted mice (n=4).

(u-w) Representative flow dot plots and quantification of Cysltr1^−/− donor (CD45.2⁺) and Cysltr1^+/+ donor (CD45.1⁺) cells in peripheral blood on day 7 (u), month 1 (v) after BM transplantation, and in CD45⁺ tumor infiltrates cells at ~ month 3 (w) (1 month post E0771 tumor challenge) (n=4).

(x) Frequencies of Cysltr1^−/− donor (CD45.2⁺) and Cysltr1^+/+ donor (CD45.1⁺) cells in total CD45⁺ cells of peripheral blood (PB), bone marrow (BM), spleen (SP) and tumor infiltrates (Tu) over time (n=4).

(y) Frequencies of Cysltr1^−/− donor (CD45.2⁺) and Cysltr1^+/+ donor (CD45.1⁺) cells in E0771-infiltrating CD11b⁺ cells (left), and Ly6G⁺ or Ly6C⁺ cells (right) (n=4).

Results are representative of three (a-d, k, l) or two (e-j, m-y) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (a (left, middle), b (left, middle), c (middle), d, f, g, k-m, o, q-s, t (middle), y (left)), two-tailed paired t-test (u-w (right)), one-way ANOVA (a (right), b (right), c (left, right), e, h, I, j (left, middle), t (left, right), x, y (right)), two-way ANOVA (n (left), p (left)) or long-rank test (n (right), p (right)).

Extended Data Fig. 3. CysLTR1 promotes splenic emergency myelopoiesis, neutrophil commitment, and tumor progression.

(a) Tumor weights of E0771 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice (n=5).

(b) Individual tumor growth curves of MC38 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice (n=4).

(c) Tumor growth curve and survival of MC38 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice (n=4).

(d) Total bone marrow cell counts (from 2 femurs and 2 tibia) of C57BL/6 Cysltr1^f/f Lyz2-cre mice and Cysltr1^f/f mice (n=3).

(e) Total cell numbers of neutrophils per femur in C57BL/6 Cysltr1^f/f Lyz2-cre mice and Cysltr1^f/f mice (n=3).

(f) Frequencies of MB, PM, MC, MM, BC, and SN subsets in BM neutrophils from in C57BL/6 Cysltr1^f/f Lyz2-cre mice and Cysltr1^f/f mice (n=3).

(g) Frequency of Ly6G⁺ neutrophils, Ly6C⁺ monocytic cells (left), Ly6G⁺/Ly6C⁺ cell ratio (middle) and frequencies of macrophage and DC (right) in tumor-infiltrating CD45⁺ cells of E0771 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice (n=3).

(h) Frequencies of LSK cell (left) HPC (middle) in lineage negative cells and CMP, GMP, MEP in HPC (right) from spleens of LPS treated BALB/c Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(i) Frequencies of LSK cell (left) HPC (middle) in lineage negative cells and CMP, GMP, MEP within HPC (right) from spleens of MTK-treated EMT6 tumor-bearing BALB/c mice (n=4).

(j) Tumor growth curve and survival of EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice with splenectomy or sham surgery (n=5).

(k) Frequencies of Ly6G⁺ neutrophils (left) and Ly6G⁺/Ly6C⁺ cell ratio (right) in PB from EMT6 tumor-bearing BALB/c Cysltr1^−/− and Cysltr1^+/+ mice with splenectomy or sham surgery (n=4).

(l) Scaled MFI of PMN-MDSC-associated markers including CD14, CD84, CD300ld and DcTRAIL-R1 in splenic neutrophils of E0771 tumor-bearing C57BL/6 WT mice (n=3).

(m) Flow cytometry analysis of CysLTR1 expression levels in splenic neutrophils from E0771 tumor-bearing C57BL/6 WT mice compared to tumor-free controls (n=4).

(n) qPCR analysis of relative expression of Cysltr1 in sorted CD84 hi compared to low/negative (left), and CD300ld hi compared to low/negative (right) populations within tumor-infiltrating neutrophils of E0771 tumor-bearing C57BL/6 mice (n=4).

(o, p) Flow cytometry analysis of expression levels of CysLTR1 in CD84 hi, middle or low/negative (o) or CD300ld hi, middle or low/negative (p) populations within tumor-infiltrating neutrophils of E0771 tumor-bearing C57BL/6 Cysltr1^+/+ mice (n=4).

(q) Scaled MFI of CD14, CD84, CD300ld and DcTRAIL-R1 in tumor-infiltrating neutrophils of E0771 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ mice (n=3).

(r) Scaled MFI of CD14, CD84, CD300ld and DcTRAIL-R1 in tumor-infiltrating neutrophils of E0771 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice(n=3).

(s) Scaled MFI of CD14, CD84, CD300ld and DcTRAIL-R1 in tumor-infiltrating neutrophils of tumor-bearing FVB-MMTV-PyMT mice with or without MTK treatment (n=3).

(t, u) Flow cytometry analysis of expression levels of CD84 (t) and CD300ld (u) in CysLTR1 hi, middle or low/negative populations within tumor-infiltrating neutrophils of E0771 tumor-bearing C57BL/6 Cysltr1^+/+ mice (n=4).

Results are representative of three (a, d-i, l, m, q-s) or two (b, c, j, k, n-p, t, u) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (a, b, d, e, g (middle), h (left, middle), i (left, middle), l, m (right), n, q-s), one-way ANOVA (f, g (left, right), h (right), i (right), k, o (right), p (right), t (right), u (right)), two-way ANOVA (c (left), j (left)) or long-rank test (c (right), j (right)).

Extended Data Fig. 4. Genetic ablation or pharmacologic inhibition of CysLTR1 enhances antitumor CD8⁺ T cell responses and sensitizes tumors to anti–PD-1 therapy.

(a, b) Frequencies and cell counts of CD3⁺ (a), CD4 and CD8 (b) T cells in CD45⁺ tumor infiltrates of E0771 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(c) Frequency of FoxP3⁺ Treg cells in infiltrating CD4 T cells of E0771 tumor-bearing C57BL/6 Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(d, e) Frequencies and cell counts of CD3⁺ (d), CD4 and CD8 (e) T cells in CD45⁺ tumor infiltrates of E0771 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice (n=4).

(f) Frequency of FoxP3⁺ Treg cells in infiltrating CD4 T cells of E0771 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice (n=4).

(g, h) Frequencies and cell counts of CD3⁺ (g), CD4 and CD8 (h) T cells in CD45⁺ tumor infiltrates of FVB-MMTV-PyMT mice treated with or without MTK (n=4).

(i) Frequency of FoxP3⁺ Treg cells in infiltrating CD4 T cells of FVB-MMTV-PyMT mice treated with or without MTK (n=4).

(j) Suppression of CD8 T cell (T) proliferation by Cysltr1^−/− and Cysltr1^+/+ tumor-infiltrating neutrophils (E) from E0771 tumor-bearing mice (n=4).

(k) Suppression of CD8 T cell (T) proliferation by Cysltr1^−/− and Cysltr1^+/+ neutrophils (E) differentiated in vitro from C57BL/6 Cysltr1^−/− and Cysltr1^+/+ BM cells with GM-CSF + IL-6 (n=4).

(l) Suppression of CD8 T cell (T) proliferation by CysLTR1 hi or low tumor-infiltrating neutrophils (E) from EMT6 tumor-bearing mice (n=4).

(m) Suppression of CD8 T cell proliferation by neutrophils differentiated in vitro from C57BL/6 WT BM cells with GM-CSF + IL-6 in the presence of montelukast (MTK) or zileuton (ZIL) (n=3).

(n) Representative ELISPOT images of IFN-γ-producing CD8⁺ T cells isolated from EMT6 tumors.

(o) Frequencies of CD44⁺ (top), IFN-γ⁺ (middle), and TNF-α⁺ (bottom) cells in E0771-infiltrating CD8⁺ T cells cocultured with splenic CD11b⁺ cells from tumor-free C57BL/6 WT mice with antigenic peptides as indicated (n=4).

(p, q) Representative ELISPOT images (p) and quantification (q) of IFN-γ-producing CD8⁺ T cells isolated from E0771 tumors co-cultured with spleen CD11b⁺ cells from tumor-free C57BL/6 WT mice and antigenic peptides as indicated (n=4).

(r) Frequencies of CD44⁺, IFN-γ⁺, TNF-α⁺ and GzmB⁺ cells in tumor-infiltrating CD8⁺ T cells from E0771 tumor-bearing C57BL/6 Cysltr1^f/f Lyz2-cre and Cysltr1^f/f mice (n=4).

(s) Frequencies of CD44⁺, IFN-γ⁺, TNF-α⁺ and GzmB⁺ cells in tumor-infiltrating CD8⁺ T cells from FVB-MMTV-PyMT mice treated with or without MTK (n=4).

(t) Tumor growth curve and survival of EMT6 tumor-bearing C57BL/6 Cysltr1^−/− mice with or without MTK treatment (n=8).

(u) Survival of Myc-Cap tumor-bearing C57BL/6 mice with or without MTK and anti-PD1 therapy (n=8).

(v) Tumor growth of BPS1 tumor-bearing C57BL/6 mice with or without MTK and anti-PD1 therapy (n=5).

(w) Survival of ID8-agg tumor-bearing C57BL/6 mice with or without MTK and anti-PD1 therapy (n=10).

Results are representative of three (a-s) or two (t-w) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (a, c, d, f, g, i-l, r, s), one-way ANOVA (b, e, h, m, o, q), two-way ANOVA (t (left), v) or log-rank test (t (right), u, w).

Extended Data Fig. 5. CysLTR1 regulates transcriptional and phenotypic programs of cytokine-induced CD11b⁺Gr1⁺ cells and neutrophil granule synthesis in vitro.

(a, b) Gene Ontology (GO) analysis of the bulk RNA-sequencing datasets of BALB/c BM-derived CD11b⁺Gr1⁺ cells induced in vitro with GM-CSF (a) or GM-CSF + EMT6-conditioned medium (b) (n=3).

(c, d) Heatmap of selected DEGs from GO term: Negative Regulation of Proliferation (c) and selected DEGs from a M1/M2 Signature (d).

(e) Volcano plot of DEGs of the bulk RNA sequencing dataset of BALB/c BM-derived CD11b⁺Gr1⁺ cells induced in vitro with GM-CSF (top), GM-CSF + IL-6 (middle) or GM-CSF + EMT6-conditioned medium (bottom).

(f) Heatmap showing the relative expression of Cd14, Cd84, Cd300ld and Tnfrsf23 (dcTRAIL-R1) of bulk RNA sequencing dataset of BALB/c BM-derived CD11b⁺Gr1⁺ cells induced in vitro with GM-CSF (top), GM-CSF + IL-6 (middle), or GM-CSF + EMT6-conditioned medium (bottom).

(g) Normalized counts of cd14, cd84, cd300ld and tnfrsf23(dcTRAIL-R1) of bulk RNA sequencing dataset of BALB/c BM-derived CD11b⁺Gr1⁺ cells induced in vitro with GM-CSF, GM-CSF + IL-6, or GM-CSF + EMT6-conditioned medium (n=3)

(h) Normalized counts of Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 of bulk RNA sequencing datasets of BALB/c BM-derived neutrophils induced in vitro with GM-CSF, GM-CSF + IL-6 or GM-CSF + EMT6-conditioned medium (n=3).

(i) qPCR analysis of relative mRNA level of Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 in GM-CSF +IL-6 induced neutrophils from BALB/c BM with or without MTK at different concentrations (n=4).

Results are representative of two (i) independent experiments. Data are mean ± s.d. Data were analyzed by one-way ANOVA (g-i).

Extended Data Fig. 6. CysLTR1 is required for transcriptional control of neutrophil granule synthesis and mobilization during tumor progression.

(a) Relative MFI of MPO and S100A8/9 in neutrophils from BM, spleen and tumor of EMT6 tumor-bearing Cysltr1^−/− and Cysltr1^+/+ mice (n=4).

(b) Relative MFI of MPO and S100A8/9 in cytokine-induced BM-derived neutrophils (n=4).

(c) ELISA analysis of S100A8/9 release by GM-CSF + IL-6-induced BM-derived neutrophils on both BALB/c (Bc) and C57BL6 (B6) backgrounds (n=4).

(d) Relative MFI of MPO and S100A8/9 in neutrophils of EMT6 tumor-bearing reconstituted mice with transplantation of Cysltr1^−/− or Cysltr1^+/+ BM cells (n=4).

(e) Frequency of Cysltr1^−/− donor (CD45.2⁺) and Cysltr1^+/+ donor (CD45.1⁺) cells in tumor-infiltrating MPO⁺S100A8/9⁺ cells of EMT6-bearing reconstituted mice (n=4).

(f, g) Representative images (f) and quantification (g) of transwell invasion assays of EMT cells in the presence or absence of conditioned medium of GM-CSF + IL-6-induced BM-derived Cysltr1^−/− or Cysltr1^+/+ neutrophils with doxycycline at different concentrations as indicated.

(h) Tumor growth curve and survival of EMT6 tumor-bearing NOD-SCID mice with or without co-transfer of BM-derived Cysltr1^−/− or Cysltr1^+/+ neutrophils (n=8).

(i) Tumor growth curve and survival of EMT6 tumor-bearing NSG mice with or without co-transfer of BM-derived Cysltr1^−/− or Cysltr1^+/+ neutrophils (n=8).

(j) CHIP-X enrichment analysis showing the potential upstream transcription factors of S100a8/9 and other selected granule genes. No sample size (n) is applicable as this is a computational enrichment analysis.

(k) Interaction networks functional enrichment analysis of the top 10 potential transcription factors (TF) for neutrophil granule genes.

(l) Cumulative weighted mean TF ranks via different analytic resource libraries.

(m) qPCR analysis of Mxd1 (left) and Nfe2 (right) expression during granulopoiesis (n=4).

(n) qPCR analysis of Mxd1 (left) and Nfe2 (right) expression in GMPs from BALB/c Cysltr1^+/+ and Cysltr1^−/− BM cells (n=4).

(o) qPCR analysis of Mxd1 (left) and Nfe2 (right) expression in GM-CSF + IL-6 induced GMP-derived neutrophils (n=4).

(g) qPCR analysis of Mxd1, Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 in GM-CSF + IL-6-induced BM-derived neutrophils with or without Mxd1 knockdown (KD) (n=4).

(h) qPCR analysis of Nfe2, Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 in GM-CSF+IL-6-induced BM-derived neutrophils with or without Nfe2 knockdown (KD) (n=4)

Results are representative of three (a, b, f, g) or two (c-e, h-i, m-q) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (c, e, n-q), one-way ANOVA (a, b, d, g), two-way ANOVA (h (left), i (left)) or log-rank test (h (right), i (right)).

Extended Data Fig. 7. Modulation of NFE2 expression regulates CysLTR1-mediated emergency granulopoiesis in BM-reconstituted tumor-bearing mice.

(a) qPCR analysis of relative mRNA levels of Nfe2, Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 in GM-CSF+IL-6-induced BM-derived Cysltr1^−/− or Cysltr1^+/+ neutrophils with or without Nfe2 knock down (KD) or overexpression (OE) (n=4).

(b) Total BM cell counts of EMT6 tumor-bearing reconstituted mice as a (n=4).

(c) Frequencies of CMP, GMP, and MEP in HPC of EMT6 tumor-bearing reconstituted mice as a (n=4).

(d) Frequencies of MB, PM, MC, MM, BC, and SN in BM neutrophils of EMT6 tumor-bearing reconstituted mice as a (n=4).

(e, f) Frequency of Ly6G⁺ neutrophils, Ly6C⁺ monocytic cells (left) and the Ly6G⁺/Ly6C⁺ cell ratio (right) in CD45⁺ cells of BM (e) and spleen (f) from EMT6 tumor-bearing reconstituted mice as a (n=4).

(g, h) Relative MFI of MPO and S100A8/9 in neutrophils of BM (g) and spleen (h) from EMT6 tumor-bearing reconstituted mice as a (n=4).

(i) qPCR analysis of relative mRNA levels of Nfe2 in BALB/c Cysltr1^−/− mice with or without Nfe2 knockdown (n=3).

(j) Tumor growth curve and survival of EMT6 tumor-bearing BALB/c Cysltr1^−/− mice with or without Nfe2 knockdown (n=5).

Results are representative of two (a-j) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (i), one-way ANOVA (a-h), two-way ANOVA (j (left)) or log-rank test (j (right)).

Extended Data Fig. 8. Single-cell transcriptomic analysis reveals neutrophil differentiation programs, transcriptional regulation, and T cell states in tumor-bearing mice.

(a) Uniform Manifold Approximation and Projection (UMAP) of neutrophils from BM, SP, and Tu, colored by tissue origin (n=3).

(b) Comparisons of the expression of Cd14, Cd84, Cd300ld and Tnfrsf23 between Cysltr1^−/− and Cysltr1^+/+ cells within each cluster.

(c) Correlation between expressions of individual neutrophil-associated marker Cysltr1, Cd14, Cd84, Cd300ld or Tnfrsf23.

(d) Correlation of Cysltr1 expression with the PMN-MDSC signature (left) and activated PMN-MDSC signature (right).

(e) Correlation of Cysltr1 expression with the DcTRAIL-R1 signature score.

(f) Proportions of Cysltr1^−/− and Cysltr1^+/+ neutrophils in total clusters and each cluster across BM, SP, and Tu.

(g) Comparison of functional scores (N1 and N2 signatures) between Cysltr1^−/− and Cysltr1^+/+ cells within each cluster.

(h) Comparisons of myeloid functional scores (neutrophil cytotoxicity, myeloid cell recruitment, EC remodeling, and angiogenesis) between Cysltr1^−/− and Cysltr1^+/+ cells within each cluster. Boxes represent the interquartile range (25th–75th percentiles), the central line indicates the median, whiskers extend to 1.5 × the interquartile range, and points beyond the whiskers represent individual outliers.

(i) Comparison of the overall granule/vesicle signature score between Cysltr1^−/− and Cysltr1^+/+ cells within each cluster.

(j) Violin plots illustrating the expression levels of selected transcription factors known to be involved in granulopoiesis between Cysltr1^−/− and Cysltr1^+/+ neutrophils.

(k, l) Correlation of Mxd1 (k) or Nfe2 (l) expression with the maturation signature (top), effector response signature (middle), and survival signature (bottom).

(m) Comparisons of functional transcription module scores (early granulocytic, late granulocytic, monocytic-like, antigen presentation, and tumor granulocytic) between Cysltr1^−/− and Cysltr1^+/+ cells within each cluster.

(n) UMAP displaying all CD45⁺ cells within tumor of EMT6 tumor-bearing mice, delineated by inferred cluster identity. T cells were identified according to the lineage specific DEGs (top) and sub-clustered UMAP displaying both CD4 and CD8 T cells (bottom).

(o) Heatmap revealing distinct signatures pertaining to T cell activities such as apoptosis, differentiation, function, and metabolism within each cluster.

(p) Distribution of different clusters across Cysltr1^−/− and Cysltr1^+/+ T cells.

(q) Bubble dot plots showing the scaled expression of representative exhaustion genes for Cysltr1^−/− and Cysltr1^+/+ T cells. Color intensity indicates the average expression of each gene scaled across all clusters. Dot size represents the percentage of cells in each cluster with more than one read of the corresponding gene.

Data were analyzed by Pearson correlation (c-e, k, l) or two-tailed unpaired t-test (j).

Extended Data Fig. 9. STAT3-mediated induction of Cysltr1 transcription links cytokine signaling to neutrophil differentiation and granule/vesicle synthesis.

(a) Mean fluorescence intensity (MFI) of p-AKT and p-ERK1/2 in GM-CSF + IL-6-induced BM-derived Cysltr1^−/− and Cysltr1^+/+ neutrophils upon LTD4 stimulation over time (n=4).

(b) Tumor growth curve and survival of EMT6 tumor-bearing BALB/c WT mice with or without anti-GM-CSF and anti-IL-6 antibodies (n=5).

(c) Flow cytometry analysis of expression levels of CysLTR1 and pSTAT3 in 3 subsets of in-vitro GM-CSF + IL-6-induced BALB/c BM-derived neutrophils based on differential expression of CSF2Rα (n=4).

(d) qPCR analysis of relative mRNA levels of Cysltr1 and Stat3 in CSF2Rα^hi and CSF2Rα^low subsets of in-vitro GM-CSF + IL-6-induced BALB/c BM-derived neutrophils (n=4).

(e) qPCR analysis of relative mRNA levels of Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 in CSF2Rα^hi and CSF2Rα^low subsets of in-vitro GM-CSF + IL-6-induced BALB/c BM-derived neutrophils (n=4).

(f) Flow cytometry analysis of the expression of CysLTR1 and pSTAT3 in 3 subsets of in-vitro GM-CSF + IL-6-induced C57BL/6 BM-derived neutrophils based on differential expression of CSF2Rα (n=4).

(g) qPCR analysis of relative mRNA levels of Cysltr1 and Stat3 in CSF2Rα^hi and CSF2Rα^low subsets of in-vitro GM-CSF+IL-6-induced C57BL/6 BM-derived neutrophils (n=4).

(h) qPCR analysis of relative mRNA levels of Mxd1, Nfe2, Elane, Mmp8, Mmp9, Mpo, S100a8, and S100a9 in CSF2Rα^hi and CSF2Rα^low subsets of in-vitro GM-CSF + IL-6-induced C57BL/6 BM-derived neutrophils (n=4).

Results are representative of two (a-h) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (d, e, g, h), two-way ANOVA (b (left)) or log-rank test (b (right)).

Extended Data Fig. 10. CysLTR1 expression in neutrophils correlates with CTL infiltration and MDSC-associated gene signatures in human cancers.

(a) Frequencies of GM-CSF + IL-6 induced CD15⁺ neutrophils, CD14⁺ monocytic cells (left) and CD15⁺/CD14⁺ cell ratio (right) in CD45⁺ cells from human PBMC of healthy donors with or without MTK treatment (n=4).

(b) Relative MFI of MPO and S100A8/9 in GM-CSF + IL-6 induced neutrophils from human PBMC of healthy donors with or without MTK treatment (n=4).

(c) Representative flow cytometry plots showing suppression of CD8 T cell (T) proliferation by GM-CSF + IL-6–induced neutrophils (E) derived from human PBMCs with or without CYSLTR1 knockdown (KD) at an E:T ratio of 1:1.

(d) Frequencies of GM-CSF + IL-6 induced CD86⁺, and CD206⁺ cells (left) and CD86⁺/CD206⁺ cell ratio (right) in CD45⁺ cells from human PBMC with or without MTK treatment (n=4).

(e) Neutrophil differentiation stages of GM-CSF + IL-6-induced cells from human PBMC of healthy donors with or without MTK.

(f, g) Representative composite image of a breast cancer specimen stained by multiplex immunohistochemistry (mIHC) comprising CD11b (green), CD15 (white), CysLTR1 (purple), S100A9 (yellow), CD8α (red), and DAPI (blue). Arrows indicated representative CD11b⁺CD15⁺CysLTR1⁺S100A9⁺ cells (f). Scale bar; 100 μM. Comparison of S100A9 intensity between CD15⁺CysLTR1⁻ cells and CD15⁺CysLTR1⁺ cells (g, top) (n=31, 6). Comparison of CD8 density between regions with CD15⁺CysLTR1⁻ cells and CD15⁺CysLTR1⁺ cells (g, bottom) (n=32, 6).

(h, i) Correlation of CYSLTR1 expression with cytotoxic T cells (CTL) distribution from the METABRIC breast cancer dataset (n=175), GSE16446 breast cancer dataset (n=120) (h), TCGA melanoma dataset (n=470), or VanAllen2015 melanoma dataset (n=58) (i).

(j) Survival analysis of different human cancers from TCGA database with comparison between high and low expression of CYSLTR1. BRCA: Breast invasive carcinoma; LUSC: Lung squamous cell carcinoma; STAD: stomach adenocarcinoma; UVM: uveal melanoma.

(k) Correlation between CYSLTR1 and ITGAM (CD11b) expression in BRCA (n=1090) and LUSC (n=452) patients from TCGA database.

(l) Correlation between CYSLTR1 with CD14 (left), FUT4 (CD15) (middle) and FCGR3A (CD16) (right) expression in BRCA and LUSC patients from TCGA database.

(m) Correlation between CYSLTR1 and GFI1, IRF8, or RORC in BRCA and LUSC patients from the TCGA database.

(n) Correlation between CYSLTR1 and MYO1F, PTAFR, STX11 in BRCA and LUSC patients from the TCGA database.

(o) Working model of CysLTR1-driven neutrophil reprogramming that promotes tumor progression and limits antitumor CD8⁺ T cell immunity. Created with BioRender.com.

Results are representative of two (a-g) independent experiments. Data are mean ± s.d. Data were analyzed by two-tailed unpaired t-test (a (right), d (right), g), one-way ANOVA (a (left), b, d (left)), log-rank test (j) or Pearson correlation (h, i, k-n).

Supplementary Material

SUPPLEMENTAL INFORMATION

Extended Data Fig. 1–10, Supplementary Fig. 1–5 and Supplementary Table 1–2

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