Macrophage-intrinsic and IL-9–dependent arginine metabolism promotes lung tumor growth

Abstract

Tumor-associated macrophages are an abundant, tumor-infiltrating cell population that supports the evasion of tumor cells from antitumoral immune cell detection by generating an immunosuppressive tumor-immune microenvironment (TIME). The immunosuppressive function of macrophages is dictated by the cytokine environment. IL-9 is a pleiotropic cytokine that can be a positive or negative regulator of tumor growth. Our lab previously identified a protumoral role of IL-9 by expanding lung interstitial macrophage (IM) populations and inducing the expression of arginase 1 (ARG1) to enhance tumor growth. However, the underlying mechanism by which IL-9 receptor/ARG1⁺ IMs promote tumor progression remains incomplete. Here, we demonstrate that macrophage-targeting nanoparticles containing Arg1 siRNA can therapeutically reduce tumor burden and reduce protumor arginine-derived metabolite production. Furthermore, using bulk RNA sequencing of lung macrophages isolated from Il9r^−/−:wild-type mixed-bone marrow chimeric mice, we demonstrate that IL-9 intrinsically alters the transcriptomic landscape of lung IMs. Mechanistically, IL-9 promotes intrinsic Arg1 expression through an IRF4-dependent regulatory pathway and modulates arginine and polyamine concentration within IMs and lung tissue, resulting in increased lung tumor growth and altered macrophage phenotypes. Thus, our work defines a protumor function of IL-9–responsive macrophages mediated by altered intrinsic arginine metabolism in lung IMs that enhances lung tumor growth.

Introduction

Lung cancer and metastasis remain the leading causes of cancer-related mortality worldwide, with survival rates remaining low despite advancements in early detection and treatment strategies. A hurdle to efficacious treatment of lung tumors is overcoming the immunosuppressive tumor-immune microenvironment (TIME) generated by the complex interactions of cytokines and the surrounding stromal and immune cells. Cytokines play an instructive role, either promoting or inhibiting tumor growth and immune evasion by influencing communication between cancer cells and immune cells, altering immune responses, and promoting inflammation that supports tumor progression.

Interleukin (IL-9) has potent antitumor or protumor activity depending on the cancer type and the responsive cells found in the TIME.^1–4 In antitumor roles, IL-9 directly inhibits tumor growth, enhances immune surveillance, and promotes anticancer immune responses. The antitumor effects of IL-9 in colorectal, breast, and gastric cancers^1,5–9 are mediated by stimulating the innate and adaptive immune responses. Leveraging multiple mechanisms of action, adoptive cell therapy using IL-9–producing T cells and chimeric antigen receptor (CAR) Th9 cells have potent antitumor activity across various murine tumor models, including melanoma, glioma, lymphoma, liver, and ovarian cancers.^1,10–15 The cell type responding to IL-9 in each scenario is likely key to the differing effects.

In contrast, IL-9 facilitates tumor progression by enhancing immunosuppression, angiogenesis, cancer cell proliferation, and metastasis.^1,3 In hematological malignancies, IL-9 directly promotes the proliferation of IL-9R⁺ lymphoma cells with concomitant decreases in apoptosis in vitro and in vivo.^16–19 Furthermore, Kumar et al demonstrated an IL-9–dependent shift toward aerobic glycolysis in a model of cutaneous T-cell lymphoma.²⁰ In solid tumors, IL-9 has both direct and indirect effects. Tian et al demonstrated that IL-9 directly stimulates malignant colonic epithelial cells enhancing proliferation by upregulating c-MYC and CyclinD1 in RKO and Caco-2 colon carcinoma cell lines.²¹ Other groups have demonstrated indirect IL-9–dependent increases in metastasis,^22,23 cellular proliferation,^24–26 immunosuppression,^22,23,27 angiogenesis,²⁸ and epithelial-mesenchymal transition²⁹ in models of hepatocellular carcinoma, colorectal cancer, bladder cancer, breast cancer, and pancreatic cancer.

With IL-9 demonstrating diverse roles in tumor progression across various cancer types, its specific involvement in lung cancer pathogenesis has remained a subject of considerable debate and investigation. In patient-derived data, elevated IL-9 or IL-9R expression is associated with decreased overall survival.^28,30 In murine models, IL-9 contributes to direct and indirect antitumor immunity in the lung by promoting tumor cell–derived MHC class I presentation and enhancing cytotoxic T-cell immunity,⁶ while other studies have demonstrated increased STAT3-dependent angiogenesis, cell proliferation, and immunosuppression.^28,31,32 Moreover, in murine models of non-small cell lung cancer (NSCLC), elevated levels of IL-9 resulted in increased regulatory T cell–mediated immunosuppression.²⁷ Additionally, our recently published work demonstrated that IL-9 promotes lung tumor growth by increasing populations of ARG1⁺ interstitial macrophages (IMs)³⁰; however, the underlying mechanism of tumor control remained unclear.

In this study, we define the mechanism by which IL-9–responsive macrophages use to enhance lung tumor growth. Our findings reveal that IL-9 intrinsically regulates Arg1-dependent metabolism thereby enhancing the tumor-supporting capabilities of pulmonary macrophages. These findings offer mechanistic insights supporting the use of IL-9–responsive macrophages and arginine metabolism as potential therapeutic targets and biomarkers in lung cancer.

Materials and methods

Patient samples

Human lung cancer tissue from the Indiana University School of Medicine Biospecimen Collection and Banking Core was obtained for analysis of arginine metabolites with the written consent of the human participants. Tumor samples were collected from patients with stage II or III adenocarcinoma. Control tissues were patient-matched and isolated from adjacent, tumor-free tissue. Information regarding patient samples is listed in Table S1. The use of all patient samples conforms to the principles outlined in the Declaration of Helsinki and was approved by the Indiana University Institutional Review Board, Tissue Procurement Core, the Komen Tissue Bank, and the Oncology Research Information Exchange Network.

Mice

All mice utilized were derived from the C57BL/6 background. Wild-type (WT) mice (C57BL/6, strain #000664) were purchased from The Jackson Laboratory. Il9r^−/− mice (C57BL/6) were a gift from Dr Jean-Christophe Renauld.³³  Irf4 fl/fl Lyz2-Cre, Odc1 fl/fl Lyz2-Cre, IL9r fl/fl Lyz2-Cre, and Arg1 fl/fl Lyz2-Cre mice were generated by crossing floxed mice to B6.129P2-Lyz2^tm1^(cre)Ifo/J (C57BL/6, strain #004781). Irf4 fl/fl mice were purchased from The Jackson Laboratory. Odc1 fl/fl mice were provided by Dr Teresa Mastracci,³⁴ and Il9r fl/fl mice were generated by Dr Daniella Schwartz. Female mice between the ages of 8 and 16 weeks were used.³⁵ Previous studies have demonstrated no difference between males and females.^30,36 All mice were maintained in specific pathogen–free animal facilities (ambient room temperature 21 °F to −22 °C, humidity 50%, 12-hour light/dark cycle). Early euthanasia of mice occurred if body weight loss exceeded 20% or if mice displayed signs of physical pain or distress. All experiments were approved by Indiana University’s Laboratory Animal Resource Center and University Institutional Animal Care and Use Committee.

Cell lines

B16-F10 melanoma (CRL-6475) and Lewis lung carcinoma (LLC) (LL/2, CRL-1642) cell lines were purchased from American Type Culture Collection (ATCC). Both cell lines were maintained according to ATCC’s recommended protocols. B16-F10 and LLC cells were cultured in DMEM containing 10% fetal bovine serum (FBS), 1% (5000 μg/mL) penicillin-streptomycin, and 1 mM sodium pyruvate.

Mixed bone marrow chimera generation

F1 B6xBoyJ mice (CD45.1/CD45.2) mice were irradiated twice at 550 cGy. Following irradiation, 8 million Il9r^−/− and WT bone marrow cells were injected intravenously (i.v.) into the recipient mice at a 1:1 ratio. Mice were left to rest for 12 weeks before further use.

Tumor model and tissue processing

In lung models using B16-F10 or LLC cells, mice were i.v. injected with 0.15 or 0.3 million cells in 200 µL. For both models, mice were euthanized 21 days postinjection unless mice lost 20% of their body weight or were visibly distressed. Euthanasia was performed by CO₂ asphyxiation followed by secondary cervical dislocation. Mice were dissected and lung, bronchoalveolar lavage fluid (BALF), and blood were taken for further processing. Lungs were perfused and washed with 4 °C PBS, and lungs were weighed and tumor burden was quantified. Lungs were then diced and digested in DMEM containing 2% to 5% BSA, 1 mg/mL collagenase D, and 100 μg/mL DNaseI. Single-cell suspensions were made following passage through 40-µM filter strainers. Cells were pelleted and red blood cells lysed. Cells were further filtered using a 40%/80% Percoll density centrifugation step. Cells at the interface were collected and washed with flourescence-activated cell sorting (FACS) buffer. BALF cells were collected in 4 °C PBS and centrifuged at 500 × g. Cells were separated from the supernatant, and supernatant and serum were kept for additional tests.

Mouse bone marrow–derived macrophage culture and polarization

Bone marrow–derived macrophages (BMDMs) were generated by isolating bone marrow from donor mice. Red blood cells were lysed, and cells were plated based on well size. Cells were cultured in DMEM containing 10% FBS, 1% penicillin-streptomycin, 1% glutamax, and 50 ng/mL M-CSF for 7 days. Macrophage medium was replaced on day 3. On day 7, macrophages were polarized using IL-4 (20 ng/mL) or IFN-γ (50 ng/mL).

In vivo cell stimulation for cytokine staining

Following single cell isolation, cells were washed and filtered. Cells were then counted and 5 × 10⁵ cells were plated in 96-well, flat-bottom plates in RPMI 1640 containing 10 ng/mL of LPS for 24 hours and incubated for 24 hours at 37 °C.

In vitro and in vivo polyamine stimulation

For in vivo experiments involving polyamine supplementation, mice were treated intranasally every 72 hours with 15 µL of a PBS solution containing spermine (25 mM), spermidine (30 mM), and putrescine (3 mM) starting 4 days postinjection. The dose was determined by physiological concentration defined previously.^37,38 During intranasal administration, mice were anesthetized with isoflurane and the metabolite solution was applied to the external portion of the nares. For polyamine stimulation in vitro, BMDMs were generated, polarized, and stimulated with putrescine (3 μM), spermidine (10 μM), and spermine (5 μM)³⁹ for 24 hours.

Bulk RNA sequencing

Lung IMs from tumor-bearing, mixed bone marrow chimeric mice were isolated using FACS (CD45.1 or CD45.2, CD64⁺, MERTK⁺, SIGLECF⁻, CD11c⁻/⁺). Sorted cells from 3 mice were pooled together prior to further processing. RNA extraction was performed using the RNeasy micro kit according to the manufacturer’s instructions (Qiagen). Purified RNA quality and quantity was determined using Agilent Bioanalyzer 2100. Each sample required a minimum RNA integrity number of 7.5. Bulk RNA sequencing (RNA-seq) was performed by Indiana University’s Medical Genomics Core. In brief, dual-indexed cDNA library preparation was performed following the Clontech SMART-seq v4 + Nextera XT protocol. Paired-end sequencing was performed using 1 ng of RNA on Illumina’s NovaSeq 6000 platform resulting in approximately 40 to 70 M reads per library.

Bulk RNA-seq analysis

Sequenced reads were first quality controlled using FastQC (v.0.11.5, Babraham Bioinformatics, Cambridge, UK) and mapped to the reference genome (UCSCmm10) using Spliced Transcripts Alignment to a Reference (STAR 2.7.10a).⁴⁰ Evaluation of sequencing data was performed by determining the number of reads that fall into different annotated regions of the reference genome using bamUtils (v.0.4.17).⁴¹ Uniquely mapped reads were assigned to mm10 reGene genes using featureCounts (subread v.2.0.3).⁴² Low-quality mapped reads were excluded. Differential gene expression analysis was performed using negative bionomical generalized linear modeling with likelihood ratio tests in EdgeR (v.3.42.4).⁴³ False discovery rate (FDR) was computed using the Benjamini–Hochberg procedure. Visualizations of gene expression were generated using z-score normalized read counts in R software version 4.3.2.

Histology

Tissues were fixed with 10% formalin for 12 to 24 hours at room temperature. Following fixation, tissues were embedded in paraffin, sectioned, and stained with hematoxylin-eosin by Indiana University’s Histology core.

Flow cytometry

Following tissue processing, single cell suspensions were stained with a fixable viability dye, surface markers, and intracellular markers. For surface staining, cells were stained in FACS buffer (containing antibodies) for 30 to 60 minutes at 4 °C. Following surface staining, cells were fixed using an IC fixation buffer (Invitrogen) for 20 minutes at 4 °C in the dark. For intracellular staining, cells were washed and stained in permeabilization buffer for 5 minutes and 60 minutes, respectively. Stained cells were washed 3 times with FACS buffer and analyzed using an Attune flow cytometer and FlowJo software. Negative median flourescence intensity (MFI) values were treated as zero during analyses. Antibodies used for staining are listed in Table S2.

Cell sorting

Murine total lung macrophages were sorted with FACSAria or SORPAria following staining with CD45.1 (BoyJ) or CD45.2 (C57BL/6), CD64, and MERTK. Interstitial macrophages were identified using SIGLECF and CD11c. All cells were stained with a viability dye. Sorted cells were collected in 10% FBS and used for further experiments.

In vivo siRNA knockdown

Mice were i.v. injected with B16-F10 tumor cells. Four days after tumor cell injection, mice were treated i.v. once or twice a week with nanoparticles containing 50 μg of target siRNA or control siRNA. Nanoparticles were synthesized using the extrusion method as previously described.^44–46 In brief, all lipid nanoparticle components were individually prepared, purified, and then combined at the desired stoichiometric ratios in chloroform. Using a siRNA-containing buffer, lipid films were hydrated and extruded to form liposomes. For targeting, an SIRPα-targeting peptide was used to preferentially target macrophages as previously reported.⁴⁷

Total polyamine assay

Polyamine quantification was performed using Abcam’s Total Polyamine Assay Kit following the manufacturer’s instructions. In brief, a proprietary enzyme acts on polyamines (putrescine, spermine, and spermidine) to generate hydrogen peroxide that is then reacted with a fluorometric probe (Excitation/Emission = 535/587). Fluorescence was measured using Molecular Devices’ Filter Max F5 multi-Mode Microplate Reader with FilterMax F3 or F5 lens filters yielding a signal that is proportional to the amount of polyamines present.

Total l-arginine and ornithine assays

The l-arginine and ornithine levels were measured using fluorometric assays from Abcam. Assays were performed following the manufacturer’s instruction. In brief, samples were prepared using the sample cleanup mix provided. Samples were then enzymatically processed into a series of intermediates and further reacted with a fluorescent probe (Ex/Em = 535/587). Fluorescence was measured using Molecular Devices’ Filter Max F5 Multi-Mode Microplate Reader with FilterMax F3 or F5 lens filters.

In vivo anti-IL-9 neutralizing antibody treatment

Mice were injected intraperitoneally with anti-IL-9 (100 μg/mouse, BioXcell, catalog #9C1) or isotype control (100 μg/mouse, BioXcell, catalog #BP0088).

Statistical analysis

Statistical analyses were performed using GraphPad Prism software. Analyses are presented as means ± SEM unless otherwise noted. Samples were excluded if identified as outliers by the Grubbs test and observed deviation was attributed to technical, rather than biological, causes. All data presented are representative of at least 2 independent experiments unless otherwise noted in the figure legends. Unpaired or paired Student t-tests, one-way ANOVA, or 2-way ANOVA were used for data analysis unless otherwise noted in the figure legends. P values <0.05 were considered statistically significant. Data normality was assessed using the Shapiro–Wilk Test.

Results

Targeted suppression of Arg1-expressing macrophages reduces tumor burden

Previous work demonstrated that elevated IL9 or IL9R expression correlates with decreased survival in lung cancer patients.^28,30 In mouse models, the adoptive transfer of WT ARG1-expressing pulmonary macrophages from tumor-bearing mice into Il9r^−/− mice previously injected with B16-F10 cells restored tumor growth in the lung.³⁰ Moreover, deletion of Arg1 in myeloid cells resulted in attenuated lung tumor growth.³⁰ Based on these findings, we targeted Arg1-expressing lung macrophages to determine their utility as a possible therapeutic target using nanoparticles conjugated with a peptide that recognizes SIRPα on myeloid cells.^30,45,46 Mice were injected with B16 cells i.v. and treated with nanoparticles loaded with siRNA targeting Arg1 or control siRNA every 72 hours or once a week for 3 weeks (Fig. 1A). Consistent with previous studies,³⁰ flow cytometric analysis of AF647-conjugated siRNA delivered by SIRPα-targeting nanoparticles revealed that approximately 50% of lung macrophages received siRNA (Fig. S1A). In both treatment groups, mice treated with Arg1 siRNA containing nanoparticles (siArg1) had significantly reduced lung tumor burden and lung weight compared to their respective scramble siRNA (siScr) controls (Fig. 1B, C). Notably, weekly treatment of siArg1 did not significantly differ from mice treated every 72 hours with siArg1. We next explored the impacts of siArg1 treatment on lung macrophage populations using flow cytometric analysis. Our previous research identified alveolar macrophages (AMs) (CD64⁺, MerTK⁺, SiglecF⁺, CD11c⁺), CD11c⁻ IMs (CD64⁺, MerTK⁺, SiglecF⁻, CD11c⁻), and CD11c⁺ IMs (CD64⁺, MerTK⁺, SiglecF⁻, CD11c⁺) as IL-9–responsive lung macrophage populations involved in lung tumor progression³⁰ (Fig. S1B). Interestingly, similar to naïve mice, AMs in siArg1-treated mice remained the predominant macrophage population (Fig. 1D). Concomitantly, CD11c⁻ and CD11c⁺ IM expansion was significantly reduced in mice that received siArg1when compared to siScr-treated mice (Fig. 1E). Furthermore, intracellular staining of ARG1 revealed that the proportion of ARG1-expressing macrophages was significantly reduced in siArg1-treated mice compared to controls (Fig. 1F) but was more pronounced in the CD11c⁻ IM population compared to the CD11c⁺ or AM populations (Fig. S1C).

Figure 1.

Targeted suppression of Arg1-expressing macrophages reduces tumor burden. (A) Workflow of siRNA-containing nanoparticle treatment. Mice received 3 treatments (days 7, 14, 20) or 6 treatments (days 4, 7, 10, 14, 17, 20) intravenously. (B, C) Tumor burden was assessed on day 21. (D–F) Flow cytometric analysis of lung macrophages from the 6-treatment group. Data for the 3-treatment group are not shown. The naïve group did not receive tumor cells or nanoparticle treatment (n = 4–6). (G) Quantification of BALF polyamine concentration from siArg1- and siScr-treated mice (n = 6). Data represent the mean ± SEM. Statistical analyses were performed using one-way ANOVA with Tukey multiple comparison test for figures with comparisons of 3 or more groups. Two-tailed unpaired Student-t test was used for comparison of 2 groups. Data are representative of 2 independent experiments that used 4 to 6 mice per group per experiment.

Arginase 1 is a critical enzyme involved in the catabolism of arginine to polyamines.⁴⁸ Furthermore, elevated polyamine content is associated with advanced tumor progression and reduced survival in many cancers.^49,50 Based on our initial findings, we hypothesized that treatment with siArg1-containing nanoparticles would decrease polyamine content. Fluorometric analysis of the total polyamine content of BALF from siScr-treated mice revealed that they had greater levels of polyamines compared to siArg1-treated mice, which had undetectable polyamine concentrations in the BALF (Fig. 1G). No significant changes were observed in the sera of siArg1-treated mice compared to siScr-treated mice (Fig. S1D). The absence of significant changes in serum polyamine concentration siArg1-treated mice relative to siScr controls suggests that the therapeutic effects are predominantly restricted to the tumor microenvironment (TME) rather than systemically. Together, these findings underscore the critical role of ARG1-expressing IMs in promoting lung tumor growth and suggest a role for macrophage-intrinsic arginine metabolism in influencing lung tumor growth.

IL-9–dependent regulation of ARG1 dysregulates lung polyamine concentration in tumor-burdened lungs

In the context of cancer, metabolic reprogramming of cancer cells and tumor-supporting cells, such as macrophages, creates a metabolically favorable environment for tumor growth. Deprivation of arginine and inversely, the production of arginine-derived metabolites in the TME, is correlated to decreased immune surveillance and enhanced tumor growth, respectively.^51–53 Based on our previous findings showing decreased tumor growth in Il9r^−/− mice (Fig. S2A–C), and the data from Fig. 1 showing a dramatic loss of polyamines with Arg1 targeting, we hypothesized that an IL-9/ARG1/polyamine axis may promote lung tumor growth. To define the connection of IL-9 and polyamines in IMs to lung tumor progression, we utilized Il9r fl/fl Lyz2-Cre mice in the B16-F10 tumor model. Quantification of lung tumor foci indicated that IL-9 promoted lung tumor growth (Fig. 2A; Fig. S2D) with parallel expansion of ARG1-expressing lung IMs (Fig. 2B–D). To confirm this was not a cell line–dependent effect, we i.v. injected 1.5 × 10⁵ LLC cells into Il9r fl/fl Lyz2-Cre mice as previously described.³⁰ Compared to Il9r fl/fl Lyz2-Cre⁻ controls, Il9r fl/fl Lyz2-Cre⁺ mice had reduced tumor number (Fig. 2E), tumor size, and lung weight (Fig. S2E, F). Furthermore, using flow cytometric analysis, Il9r fl/fl Lyz2-Cre⁺ mice had reduced expansion of lung IMs (Fig. 2F, G) and reduced ARG1-expressing IMs compared to littermate Il9rfl/fl Lyz2-Cre⁻ controls (Fig 2H; Fig. S1G). Thus, using conditional mutant mice, these findings demonstrate that IL-9–dependent tumor growth in the lung is an intrinsic effect of IL-9 signaling in myeloid cells including macrophages.

Figure 2.

IL-9–dependent regulation of ARG1 dysregulates lung polyamine concentration in tumor-burdened lungs. (A–D) B16 melanoma cells were intravenously injected into Il9r fl/fl Lyz2-Cre and control mice (n = 3 to 5). (A) Analysis of tumor burden 21 days after injection. (B–D) Identification and analysis of lung macrophages by flow cytometry. (E–H) Il9r fl/fl Lyz2-Cre and control mice were intravenously injected with LLC cells. (E) Analysis of tumor burden 21 days after injection (n = 3 to 5). Identification and analysis of lung macrophage populations (n = 3 to 5). (F–H) Flow cytometric analysis of lung macrophages. (I, J) Quantification of arginine-derived metabolites from human lung and tumor tissue and serum (n = 5 to 6). (K–N) Arginine metabolite quantification from tumor-bearing (i.v. B16 or LLC) WT, Il9r^−/−, and Il9r fl/fl Lyz2-Cre mice. (K, L) Polyamine quantification of lung lysate (n = 4) and BALF (n = 4) from tumor-bearing WT and Il9r^−/− mice. (M, N) Polyamine quantification of lung lysate (n = 3 to 4) and BALF (n = 4) from Il9r fl/fl Lyz2-Cre mice from 2 independent experiments. Data represent the mean ± SEM. Statistical analyses were performed using 2-tailed Mann–Whitney test (A–H), paired Student t-tests (I), and Welch t-test (J–N). Data are representative of 2 independent experiments that used 3 to 5 mice per group per experiment.

We next sought to determine if deletion of IL-9 signaling impacted downstream arginine and arginine-derived metabolite concentration. We hypothesized that IL-9–dependent upregulation of ARG1 would decrease environmental and intracellular arginine concentrations. To test this, we performed the i.v. B16 tumor model and assayed for environmental arginine concentration, the substrate of ARG1, in the lung lysate and BALF of tumor-bearing WT and Il9r^−/− mice. WT mice had significantly lower levels of arginine in lung lysate and BALF compared with Il9r^−/− mice (Fig. S2H, I) suggesting that loss of IL-9 signaling decreases arginine consumption by ARG1. Supporting these data, the treatment of WT tumor-bearing mice with IL-9 neutralizing antibodies significantly increased arginine concentration in BALF (Fig. S2J). We next sought to investigate if IL-9 affected intracellular macrophage arginine metabolism. To test this, we isolated total lung macrophages (CD64⁺MerTK⁺) using FACS and assayed for the intracellular arginine content. Parallel to our findings in the lung lysate and BALF, the intracellular level of arginine in WT macrophages was significantly lower when compared to Il9r^−/− macrophages (Fig. S2K). To determine if IL-9 signaling in myeloid cells contributes to environmental arginine concentration modulation, quantification of arginine in the lung lysate of Il9r fl/fl Lyz2-Cre⁺ mice injected with LLC cells had increased arginine concentration when compared to Il9r fl/fl Lyz2-Cre⁻ control mice (Fig. S2L, M).

Metabolism of arginine by ARG1 results in the production of ornithine and subsequently polyamines. In cancer, polyamine levels are elevated when compared to normal or adjacent, tumor-free tissue. To demonstrate this is also true for the lung, we measured the polyamine content of paired tumor-free and tumor-bearing tissue and serum from human patients with stage II or III lung adenocarcinoma. Polyamine levels in the tumor-bearing tissue were significantly higher compared to the paired, tumor-adjacent controls (Fig. 2I; Table S1). Similar to our findings in Fig. 1, there was no observed difference in polyamine content in the serum, suggesting that polyamine content fluctuation is confined to the local environment (Fig. 2J).

Since our findings show that IL-9 promotes ARG1 expression and decreases environmental and intracellular arginine concentration, we hypothesized that downstream metabolite concentrations, particularly polyamines, would be increased. To address this, we assayed total polyamine levels in the lung lysate and BALF collected from tumor-bearing WT and Il9r^−/− mice. As expected, the total polyamine levels were significantly higher in the lung lysate (Fig. 2K) and BALF (Fig. 2L) of WT mice i.v. injected with B16 cells when compared to Il9r^−/− mice. To verify that these results are not tumor model dependent, we utilized the LLC tumor model to measure polyamine concentrations in lung lysates and BALF from Il9r fl/fl Lyz2-Cre⁺ and Il9r fl/fl Lyz2-Cre⁻ tumor-bearing mice. Il9r fl/fl Lyz2-Cre⁺ mice had significantly reduced levels of total polyamines compared with Cre-negative control mice (Fig. 2M, N). Collectively, these data suggest that IL-9–dependent ARG1 expression in IMs alters lung arginine metabolism and increases the concentration of polyamines in the lung TME.

IL-9 intrinsically alters arginine metabolism in lung IMs to drive lung tumor growth

Previous studies using Il9r^−/− and WT mice i.v. injected with B16 cells demonstrated that Arg1 expression is greater in WT IMs when compared to Il9r^−/− IMs.³⁰ Importantly, in these studies, significant loss of tumor burden is associated with the loss of IL-9 signaling in Il9r^−/− mice, making it challenging to discriminate transcriptomic effects due to the loss of IL-9 signaling and those arising as a consequence of diminished tumor burden in the lung. To investigate the intrinsic effects of IL-9 and ARG1 on lung IMs, we generated Il9r^−/− (CD45.2):WT (CD45.1) and Arg1fl/fl Lyz2-Cre (CD45.2):WT (CD45.1) mixed bone marrow chimeras and performed bulk RNA-seq of sorted WT, Il9r^−/−, and Arg1 fl/fl Lyz2-Cre⁺ IMs isolated from tumor-bearing lungs (Fig. 3A). In this setting, using FACS, WT and Il9r^−/− or Arg1 fl/fl Lyz2-Cre⁺ IMs can be isolated from an environment where tumor burden, and the associated tumor-derived factors, are controlled for. Interestingly, flow cytometric analysis of mixed bone marrow chimeras revealed a higher proportion of CD45.2-derived bone marrow engraftment from Il9r^−/− mice compared to CD45.1-derived bone marrow from BoyJ mice following immune system reconstitution (Fig. S3A).

Figure 3.

IL-9 intrinsically alters arginine metabolism in lung IMs to drive lung tumor growth. (A–D) Bulk RNA-seq workflow of IMs isolated from tumor-bearing Arg1 fl/fl Lyz2-Cre:WT chimeras and Il9r^−/−:WT chimeras. Three mice were pooled per biological replicate before RNA extraction. (B) Heatmap of log₂FC values for each gene based on genotype and macrophage subpopulation. (C) Venn diagram of Il9r^−/−, Arg1 fl/fl Lyz2-Cre, and WT differentially expressed genes. (D) Heat map of enzymes involved in arginine metabolism generated from bulk RNA-seq analysis on lung macrophages isolated from mixed bone marrow chimeras. Data represent an average of triplicates for each group. Three mice were pooled per biological replicate before RNA extraction. (E–H) Paired comparison of flow cytometry data of macrophages from tumor-bearing WT:Il9r^−/− mixed bone marrow chimeras (n = 22). (I) Paired comparison of intracellular polyamine content of IMs sorted from tumor-bearing mixed bone marrow chimeras (n = 6, 3 mice pooled per biological replicate before sorting). Statistical analyses were performed using 2-tailed paired Student t-tests. Data are representative of 2 independent experiments, with the exception of RNA-seq, which utilized pooled samples from 10 mice per group.

We next performed differential gene expression analysis of CD11c⁻ and CD11c⁺ IMs from each set of chimeras to determine the transcriptomic effects unique to IL-9 and ARG1. Comparison of the transcriptomes based on genotype or IM subset revealed several distinct differential expression patterns dependent on IL-9 and ARG1 in lung IMs (Fig. 3B). Comparison of the differentially expressed genes (DEGs) (FDR <0.05) between CD11c⁻ and CD11c⁺ IM subsets from Arg1 fl/fl Lyz2-Cre⁺:WT and Il9r^−/−:WT chimeras indicated a higher number of DEGs in the Il9r^−/−:WT chimeras (CD11c⁻: 109; CD11c⁺: 30) when compared to the Arg1 fl/fl Lyz2-Cre⁺:WT chimeras (CD11c⁻: 16; CD11c⁺: 9) (Fig. 3C; Fig. S3B). While only a few DEGs were commonly dysregulated in both chimeric conditions (Fig. 3C; Tables S3 and S4), these included inflammatory pathway genes. These results suggest that the impact of Arg1 deficiency is at least partially due to changes in the microenvironment that are obscured in the mixed bone marrow chimera model. Differential gene expression analysis revealed that IMs derived from Il9r^−/− bone marrow had reduced Arg1 transcript expression in both CD11c⁻ and CD11c⁺ populations when compared to WT-derived IMs obtained from the same tumor-bearing chimeric mouse (Fig. 3D; Fig. S3B). Furthermore, Arg1 expression was most significantly impacted in CD11c⁺ IMs, which have been previously demonstrated to express higher IL-9R than CD11c⁻ IMs.^30,36 Importantly, there were no significant changes in inducible nitric oxide synthase 2 (Nos2), which has been previously described to be impacted by IL-9 in tumor-associated macrophages (TAMs) in a subcutaneous model of B16,⁵⁴ suggesting that different tissue environments may impact macrophage responses to IL-9 or that these effects are not dependent on IL-9 (Fig. 3D; Fig. S3D).

To further investigate the intrinsic impacts of IL-9 signaling on lung IM arginine metabolism, we performed flow cytometric analysis of immune cells isolated from the lungs of Il9r^−/−:WT mixed bone marrow chimeric mice as previously described in Fig. 3A. Flow cytometric analysis of cells isolated from the lungs of chimeric mice confirmed previous observations that, upon loss of IL-9 signaling, IM expansion is attenuated and AMs are proportionally dominant (Fig. 3E, F). Furthermore, flow analysis of IMs confirmed that the proportion of IMs that expressed ARG1 and the overall staining intensity of ARG1 was significantly decreased in the Il9r^−/− IMs (Fig. 3G, H). Together, these results demonstrate that IL-9 intrinsically upregulates ARG1 expression at both the transcript and protein levels in lung IMs.

Based on our results demonstrating the IL-9–dependent regulation of ARG1, we next investigated whether the effects of arginine metabolism were dependent on IL-9 or an effect of the factors in the tumor environment. To confirm the intrinsic impact of IL-9 on macrophage arginine metabolism, IMs were sorted based on CD45.1 (WT) or CD45.2 (Il9r^−/−) expression from Il9r^−/−:WT mixed bone marrow chimeric mice previously i.v. injected with B16 cells. Supporting previous findings, IMs derived from Il9r^−/− bone marrow had reduced intracellular polyamine content compared to WT IMs derived from the same mouse (Fig. 3I). Interestingly, these effects are mirrored in BMDMs polarized toward an M2 phenotype using IL-4, but not unpolarized BMDMs or BMDMs polarized toward an M1 phenotype using IFN-γ (Fig. S3C). Together, these data suggest that IL-9 intrinsically upregulates ARG1 expression and consequently leads to altered arginine metabolism in lung IMs that potentially enhances tumor growth in the lung.

Intrinsic regulation of ARG1 by IL-9 is mediated by IRF4

IRF4 regulates type 2 immune responses across various cell types, including dendritic cells, T cells, monocytes, and macrophages, following induction by STAT6 and STAT5 in macrophages and T cells, respectively.^55–57 Furthermore, both IRF4 and STAT6 promote a protumoral macrophage phenotype and induce Arg1 expression in macrophages.^56,58–60 However, the connection from IL-9, which activates STAT1, STAT3, and STAT5 in macrophages,^54,61 to ARG1 remains unelucidated. Based on this, we hypothesized that an IL-9/IRF4 regulatory pathway drives IL-9–dependent Arg1 expression.

To test this, we generated Irf4 fl/fl Lyz2-Cre mice and performed the i.v. B16 tumor model. In Irf4 fl/fl Lyz2-Cre⁺ mice, tumor burden was significantly decreased compared to Irf4 fl/fl Lyz2-Cre⁻ controls (Fig. 4A, B). Flow cytometric analysis revealed that, proportionally, CD45⁺ cells were enriched in the Irf4 fl/fl Lyz2-Cre⁺ group, but the proportion and total number of macrophages were unchanged (Fig. S4A–J). Furthermore, the proportion of IMs and AMs was not significantly impacted between Irf4 fl/fl Lyz2-Cre⁻ or Irf4 fl/fl Lyz2-Cre⁺ mice (Fig. 4C–E), suggesting that IRF4 is not required for expansion of IMs in the lung and that there are IL-9–dependent effects on macrophages that are IRF4 independent. Importantly, in flow cytometric analysis, ARG1 expression in IMs from Irf4 fl/fl Lyz2-Cre⁺ mice was significantly reduced compared to controls. Visualization of CD11c⁺ and CD11c⁻ IMs using t-distributed stochastic neighbor embedding (tSNE) highlighted a greater impact of IRF4 deficiency on ARG1 in CD11c⁺ IMs (Fig. 4F–H). Thus, loss of IRF4 expression in macrophages uncouples the effect of IL-9–dependent ARG1 expression from IL-9–induced expansion of IMs.

Figure 4.

Intrinsic regulation of ARG1 by IL-9 is mediated by IRF4. (A–H) B16-F10 cells were intravenously injected into Irf4 fl/fl Lyz2-Cre mice (n = 4). Lungs were harvested 21 days postinjection. (A, B) Tumor burden of Irf4 fl/fl Lyz2-Cre mice. (D–G) Flow cytometric analysis of macrophages isolated from tumor-bearing Irf4 fl/fl Lyz2-Cre mice. (H) tSNE plots represent data derived from multiparameter flow cytometry, capturing expression patterns of ARG1 (color-coded by expression level: low, blue; high, yellow/red). Statistical analyses were performed using a 2-tailed Mann–Whitney test. Data represent the mean ± SEM. Data are representative of 2 independent experiments that used 4 mice per group per experiment.

Polyamines alter lung macrophage IL-6 expression to enhance tumor growth

Polyamines are known to increase tumor cell proliferation and progression.^49,62 However, much less is known about the impact of polyamines on tumor-infiltrating immune cells. To further determine the direct impact of polyamines on macrophages, we generated BMDMs and polarized them toward an immunostimulatory (M1-like) or immunosuppressive (M2-like) phenotype using IFN-γ or IL-4, respectively, and then stimulated them with polyamines (putrescine, spermine, and spermidine). Using flow cytometry, we quantified the expression of surface and intracellular proteins associated with antitumoral and protumoral phenotypes. We hypothesized that polyamine supplementation would reprogram immunostimulatory macrophages toward an immunosuppressive phenotype. Interestingly, we found that macrophages, regardless of prior polarization, had enhanced expression of both protumor (Fig. 5A–H) and antitumor markers (Fig. 5I–P). Most notably, we found significant polyamine-induced IL-6, CD206, and PDL1 expression in unpolarized (M0), M1-like, and M2-like macrophages (Fig. 5A–H).

Figure 5.

Polyamines (PAs) enhance protumor and antitumor macrophage gene expression. (A–P) Flow cytometric analysis of PA (putrescine [3 μM], spermidine [10 μM], and spermine [5 μM])–stimulated BMDMs after polarization with IFN-γ or IL-4 (n = 5). Data represent the mean ± SEM. Statistical analyses were performed using a one-way ANOVA with mixed-effects analysis or Friedman test. Data are representative of 2 independent experiments. Each experiment used bone marrow from 5 individual C57BL/6 mice paired across different treatment groups per experiment.

An important enzyme controlling polyamine biosynthesis is the rate-limiting enzyme ODC1, which catalyzes the product of ARG1, ornithine, into putrescine. Since IL-9 leads to increased levels of polyamines in tumor-burdened lungs and IMs, we next sought to determine the impact of polyamines on pulmonary macrophage populations in the context of IL-9. To test this, we compared Odc1 fl/fl Lyz2-Cre⁺, Il9r^−/−, Il9r^−/− supplemented with polyamines, and WT mice in the i.v. B16 tumor model. In polyamine-supplemented groups, mice were intranasally treated every 72 hours with a PBS solution containing putrescine, spermidine, and spermine at a concentration based on intratumor polyamine levels previously described^37,38 for 3 weeks starting 4 days post–i.v. injection with B16 cells (Fig. S5A). To confirm that Odc1 deletion impacts lung polyamine content, we assayed polyamine content of lungs from tumor-bearing Odc1 fl/fl Lyz2-Cre⁺ and Odc1 fl/fl Lyz2-Cre⁻ control mice. The loss of Odc1 expression resulted in decreased polyamine content in the lung lysate of Odc1 fl/fl Lyz2-Cre⁺ tumor-bearing mice but not control mice (Fig. S5B). Compared to PBS-treated controls, Il9r^−/− mice that received treatments with the polyamine-supplemented solution had significantly higher lung tumor burden (Fig. 6A). Furthermore, we found that compared to untreated WT mice, all groups had significant reductions in tumor burden, with the most notable decreases found in the Odc1 fl/fl Lyz2-Cre⁺ and Il9r^−/− untreated control. Notably, Odc1 fl/fl Lyz2-Cre⁺ mouse tumor burden was decreased when compared to WT (Fig. 6A), highlighting the impact of myeloid-derived polyamines on lung tumor growth.

Figure 6.

Polyamines (PAs) enhance lung macrophage IL-6 to promote tumor development. B16-F10 cells were intravenously injected into the indicated mice. (A) Lung tumor burden (n = 3 to 5). (B–H) Flow cytometric analysis of lung cells isolated from tumor-bearing WT, Il9r^−/−, Il9r^−/− mice supplemented with PA (spermine [25 mM], spermidine [30 mM], putrescine [3 mM]), and Odc1 fl/fl Lyz2-Cre mice. (I) Flow plots of WT (n = 5), Il9r^−/− (n = 4), Il9r^−/− plus polyamine supplementation (n = 4), and Odc1 fl/fl Lyz2-Cre (n = 3) interstitial macrophages. Data represent the mean ± SEM. Statistical analyses were performed using a one-way ANOVA with Tukey or Dunnett multiple comparison test. Data are representative of 2 independent experiments that used 3 to 5 mice per group per experiment.

We next used flow cytometry to examine changes in the macrophage populations. We found that the loss of Odc1 expression in macrophages did not completely restore AM or IM proportions, similar to that of Il9r^−/− control mice, further suggesting that IL-9 may influence IM expansion independent of ARG1 (Fig. 6B–D). Interestingly, the loss of Odc1 in macrophages impacted CD11c⁻ IM proportions to a greater extent than AMs or CD11c⁺ IMs (Fig. 6C, D), suggesting a more apparent role of polyamines in the expansion of CD11c⁻ IMs. This observation is supported by the reciprocal addition of polyamines in the Il9r^−/− polyamine-treated group where there is an expansion of CD11c⁻ IMs but not CD11c⁺ IMs (Fig. 6C, D). We observed minimal effects on T-cell populations isolated from tumor-bearing Il9r^−/− polyamine-treated mice when compared to PBS control Il9r^−/− mice (Fig. S5C–N). Consistent with previous reports on the impact of polyamine supplementation on T-cell populations, CD8⁺ T-cell IFN-γ expression was reduced with polyamine supplementation (Fig. S5O, P)⁶³.

Consistent with previous studies examining IM IL-6 production in the context of IL-9,³⁰ flow cytometric analysis of IMs from WT and Il9r^−/− mice confirmed a loss of IL-6 production primarily in CD11c⁻ macrophages (Fig. 6E–I). The loss of IL-6 was recovered in Il9r^−/− supplemented with polyamines (Fig. 6E–I). Similar to loss of IL-9R signaling, loss of ODC1 in the Odc1 fl/fl Lyz2-Cre mice exhibited decreased IL-6 production (Fig. 6E–I). This finding is consistent with our in vitro analyses investigating the role of polyamines on BMDM polarization (Fig. 5) and suggests a regulatory role of polyamines on IL-6 expression in lung macrophages. Furthermore, supplementation of polyamines restored CD11c⁻ IM, but not CD11c⁺ IM, proportions and IL-6 production in Il9r^−/− mice (Fig. 6E–I). These findings demonstrate that IL-9–dependent arginine-derived metabolites are important effector molecules for IL-6 production in lung CD11c⁻ IMs.

Discussion

In the context of tumor immunity, the mechanisms of IL-9–dependent antitumor and protumor functions are poorly defined. Accumulating evidence demonstrates that IL-9 can function as a positive regulator of tumor growth, specifically in the lung,^1,25,30 which is in contrast to many studies highlighting the antitumor functions of IL-9 associated with CAR Th9 cells and adoptive cell therapies.^1,64 Multiple factors are likely to influence tumor outcomes in the context of IL-9, including the cell type–specific IL-9 responses within the TIME, the tumor type, and potentially the concentration of IL-9 within the TME. Additionally, previous studies have demonstrated that IL-9 signaling is linked to the expansion of ARG1-expressing IMs and the loss of tumor burden.³⁰ In this study, we define the intrinsic effects of IL-9 on lung IM arginine metabolism that drives lung tumor progression. Our findings reveal that IL-9 intrinsically regulates ARG1 expression at both the transcript and protein levels, resulting in altered arginine and arginine metabolite concentrations in macrophages that foster lung tumor growth.

In this study, we utilized mixed bone marrow Il9r^−/−:WT chimeras to demonstrate that ARG1 is intrinsically regulated by IL-9 in IMs. Our findings suggest that IL-9 regulates Arg1 through an IRF4-dependent regulatory pathway that operates independently of IL-9–mediated expansion of IMs. This is evidenced by the observation that IRF4 expression uncouples IL-9–driven IM expansion from ARG1 upregulation, positioning IRF4 downstream of IL-9 signaling. Mechanistically, IL-9 binds to the IL-9R, activating STAT1, STAT3, and STAT5 in lymphocytes.^65,66 Previous studies have established that STAT5 induces IRF4 expression in T cells and plasmacytoid dendritic cells to drive Th2 responses.^57,67 Expanding on this, studies using human blood monocyte-derived macrophages reported that IL-9 also induced phosphorylation of STAT1, STAT3, and STAT5,⁶¹ raising the possibility that IL-9–dependent Arg1 expression may be mediated through STAT5/IRF4 axis. Notably, chromatin immunoprecipitation–qPCR analyses of BMDMs indicate that IRF4 does not directly interact with the Arg1 promoter, suggesting that IRF4-dependent regulation of Arg1 is indirect. Additional studies are required to clarify the role of IRF4 in modulating IM transcriptional profiles.

Cancer cells require arginine for cell proliferation and protein generation. Similarly, arginine is required to mount an effective antitumor immune response and is particularly important for the adaptive immune response.^51,68,69 A recent study noted that in solid tumors, cancer cells catabolize arginine, generating an arginine-depleted TIME that suppressed CAR T-cell antitumor functions, and that the addition of arginine restored their antitumor efficacy.⁷⁰ One component of arginine metabolism in solid tumors is tumor-associated myeloid cells. Studies have demonstrated that myeloid-derived ARG1 expression is associated with myeloid cell populations such as TAMs and myeloid derived suppressor cells (MDSCs) that act to deplete arginine in the TIME and prevent effective antitumor immunity through suppression of T-cell activation, proliferation, and differentiation.^69,71 In macrophages, arginine metabolism is influenced by the type 2 cytokines IL-4 and IL-13 via JAK1/JAK3/STAT6-dependent gene expression.^72,73 Accumulating evidence suggests that IL-9 potentially reinforces Th2-associated metabolic programming to further modulate immunometabolism and macrophage function in the cancer TIME. A recent study by LaMarche and colleagues highlighted that loss of IL-4 signaling in early myeloid progenitor cells in the bone marrow—but not in tumor-infiltrating mature myeloid cells—resulted in NSCLC tumor reduction.⁷⁴ While IL-4 and IL-13 are well-established drivers of alternative macrophage activation characterized by arginine metabolism and tissue remodeling changes, IL-9 occupies a nuanced position within the type 2 cytokine milieu. In the context of lung cancer, IL-9 may act as a secondary modulator within the type II cytokine network. Specifically, IL-4–primed myeloid precursor cells could acquire increased responsiveness to IL-9 when in the lung TME, potentially amplifying IL-9–mediated changes in macrophage metabolism and function.

In our lung tumor growth models, IL-9–driven ARG1 expression in lung macrophages reduced both environmental and intracellular arginine levels while significantly increasing polyamine concentrations across extracellular and intracellular compartments. In the context of cancer, these findings underscore the importance of polyamines as downstream metabolites of ARG1 that contribute to IL-9–dependent tumor growth, extending our understanding beyond the well-characterized effects of arginine. Supporting our previous findings,³⁰ IL-9–dependent changes in environmental arginine had minimal impact on T-cell phenotypes in our studies. It is possible that the beneficial effects of increased arginine in the lung could be mitigated by a robust immunosuppressive TIME, warranting further investigations into combination therapies targeting arginine metabolism and checkpoint inhibitors.

Polyamines are protumor factors that are found at elevated concentrations in the tumor. Under normal physiological conditions, polyamines are tightly regulated and found in low concentrations. Functionally, polyamines are critical for normal cellular processes including modulation of chromatin structure, regulation of ion channels, membrane stability, immune modulation, scavenging free radicals, and translational integrity, but in lung, breast, colon, prostate, and skin cancers^48,75 intracellular and environmental polyamine levels are dysregulated, elevated in concentration,⁷⁶ and correlate with high levels of cell proliferation, altered gene expression and epigenetic profiles, and tumorigenesis.^48,77

In macrophages, the individual contributions of the polyamines putrescine, spermidine, and spermine have been extensively studied; however, the exact effects of each remain challenging to distinguish due to the myriad cellular processes influenced by polyamines and by metabolic conversion among the polyamines. Studies using myeloid-specific Odc1 knockout mice infected with Helicobacter pylori revealed that the absence of ODC1 decreased putrescine levels and led to epigenetic modifications, including increased H3K9 acetylation and decreased H3K9 methylation, which enhanced the expression of proinflammatory genes including Il1b, Il6, Il12a, Il12b, Tnfa, and Nos2.⁷⁸ These findings are inconsistent with our in vivo and in vitro observations, where polyamine supplementation was associated with increased IL-6 production in lung CD11c⁻ IMs and BMDMs, respectively. These inconsistencies are likely due to the different environmental contexts or differences in experimental models and support the notion that the effects of polyamines are influenced by additional factors found in the environment. Furthermore, spermidine has been demonstrated to promote the expression of NOS2, an arginine-metabolizing enzyme associated with antitumor responses.⁷⁹ Conversely, other studies have demonstrated that spermidine induces M2 polarization by activating AMPK and upregulating HIF-1α.^39,80,81 Building on these findings, IL-9 has been previously associated with enhanced angiogenesis,^28,30 suggesting that IL-9–dependent modulation of polyamine metabolism may promote angiogenic-supporting responses in macrophages. Consistent with our findings, a recent study determined that elevated spermine in the TIME is linked to enhanced immunosuppressive macrophage polarization in hepatocellular carcinoma and is characterized by increases in TGF-β and IL-10.⁸² Together, these results suggest that polyamines are important effector molecules that enhance protumor gene expression of IMs in the lung TME. Importantly, the implications of these results are likely nuanced, given that the interactions of other factors in the TME with polyamines that contribute to different tumor immunity outcomes are likely dependent on cancer type and tissue localization. Further studies are necessary to investigate the kinetics of polyamines and their mechanisms of action in different diseases and to determine whether their effects are specific to macrophages or extend to other cell populations in the context of IL-9.

The conclusions from this study are restricted to the i.v. models of lung tumor growth using B16-F10 melanoma and LLC cell lines. In previous studies,³⁶ we demonstrated that loss of IL-9 signaling in subcutaneous models of B16-F10 does not impair tumor growth. Our current and prior studies³⁶ indicate that macrophage IL-9R expression is elevated in the lung compared to TAMs isolated from subcutaneous B16 tumors, suggesting that factors within the lung microenvironment are critical for IL-9–dependent responses. This key distinction points to a potentially greater role for IL-9–responsive macrophages in the context of lung metastasis. Our studies do not address the role of IL-9 in spontaneous, primary lung tumor growth, which warrants further investigation. Models of spontaneous lung cancer using Kras^+/LSL-K12D × p53^fl/ mice could help address the central question of whether IL-9–responsive macrophages influence primary tumor growth in a manner similar to their roles in metastatic models. Notably, the observed reduction in lung tumor growth following loss of IL-9 signaling in macrophages suggests that IL-9 may be important for tumor seeding or tumor progression, but a role in tumor development in a spontaneous lung cancer model is still unclear.

Therapeutic targeting of arginine metabolism, particularly through the modulation of polyamine synthesis, has emerged as a promising avenue in cancer immunotherapy. One such approach involves the use of difluoromethylornithine (DFMO), a potent inhibitor of ODC1, which is currently undergoing evaluation in a phase 2 clinical trial for the treatment of neuroblastoma. Beyond neuroblastoma, recent studies underscore the potential of DFMO in combination therapies. Notably, polyamine blockade using DFMO alongside a trimeric polyamine transport inhibitor significantly enhanced antitumor immune responses mediated by macrophages in a subcutaneous B16 melanoma model.⁵³ Further supporting the therapeutic relevance of targeting the IL-9/ARG1/polyamine axis, targeted delivery of nanoparticles containing Il9r-siRNA reduced tumor burden in an i.v. B16 tumor model.³⁰ Importantly, systemic targeting of IL-9 pathways may present significant challenges due to the potential of hampering antitumor effects mediated by IL-9 and other IL-9–responsive cell types within different TMEs. Instead, our findings suggest targeting specific protumor, IL-9–responsive cell populations or downstream targets such as ARG1 may offer a more viable and strategic approach that mitigates unintended, immune-related adverse effects. Specifically, we demonstrate that macrophage-targeting Arg1-siRNA–containing nanoparticles have considerable efficacy in reducing tumor burden, particularly in lung metastasis models. Overall, this study identifies lung IMs as critical effector cells driving tumor growth by altering the metabolic balance of arginine and its derivatives in the TIME. Nonetheless, the accumulating evidence of IL-9 as a tumor-supporting factor highlights the need for further research to establish its context-dependent effects across various cell and cancer types and to optimize its therapeutic potential.

Author contributions

Conceptualization and methodology: A.C. and M.H.K. Reagents and mouse strains: S.K., E.B., G.H., B.B., D.M.S., T.L.M., and M.H.K. Investigation: A.C., J.Z., C.C.L.C., M.L.N., A.M.A., A.P., and M.K. Visualization: A.C. Supervision: M.H.K. and J.R. Writing—original draft: A.C. and M.H.K. Writing—review & editing: A.C., M.H.K., and A.P. All authors have read and approved of submission of this manuscript.

Anthony Cannon (Conceptualization [Lead], Data curation [Lead], Formal analysis [Lead], Funding acquisition [Lead], Investigation [Lead], Methodology [Lead], Visualization [Lead], Writing—original draft [Lead], Writing—review & editing [Lead]), Jilu Zhang (Formal analysis [Supporting], Investigation [Supporting], Methodology [Supporting], Writing—review & editing [Supporting]), James Ropa (Formal analysis [Supporting], Writing—review & editing [Supporting]), Abigail Pajulas (Formal analysis [Supporting], Writing—review & editing [Supporting]), Cherry C.L. Cheung (Formal analysis [Equal]), Michelle Liu Niese (Formal analysis [Supporting]), Ahmed M. Abdelaal (Formal analysis [Supporting]), Sabrina Khan (Methodology [Supporting], Resources [Supporting]), Emily Bromley (Methodology [Supporting], Resources [Supporting]), Gyoyeon Hwang (Methodology [Supporting], Resources [Supporting]), Maya Krishnan (Formal analysis [Supporting]), Basar Bilgicer (Methodology [Supporting], Resources [Supporting], Supervision [Supporting]), Teresa L. Mastracci (Resources [Supporting]), Daniella Muallem Schwartz (Resources [Supporting]), and Mark H. Kaplan (Conceptualization [Lead], Funding acquisition [Lead], Project administration [Lead], Resources [Lead], Supervision [Lead], Writing—original draft [Lead], Writing—review & editing [Lead])

Supplementary material

Supplementary material is available at The Journal of Immunology online.

Funding

The work was supported by grants from the National Institutes of Health (T32 DK007910 to A.C., R00 HL166790 to J.R., T32 AI060519 to A.P., T32 HL091816 to M.L.N., F30AI174762 to M.L.N.) and from the Indiana University Simon Comprehensive Cancer Center and the Brown Center for Immunotherapy (P30 CA082709 and U54 DK106846 to A.C. and M.H.K.).

Conflicts of interest

None declared.

Data availability

All sequencing data are available on the Gene Expression Omnibus (GEO accession number GSE29222). All other data underlying the article are presented in the article and online material.

Ethical statement

Indiana University’s Biospecimen Collection and Banking Core approved the collection of patient samples. All research was performed in compliance with Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines.