Blocking P2X7-Mediated Macrophage Polarization Overcomes Treatment Resistance in Lung Cancer

Lung cancers are the most frequently diagnosed cancer and a leading cause of cancer-related deaths (1, 2). Despite the availability of molecular targeted therapies and immunotherapy for patients with lung cancer, recurrence and disease progression due to drug resistance are common (3). Immune cells in the tumor microenvironment (TME) can support tumor growth, metastasis, and resistance to cancer therapy, highlighting the clinical potential of TME-targeted therapies (4, 5). Tumor-associated macrophages (TAM) with an M2-like phenotype play an important role in facilitating tumor progression and metastasis (6). More subtypes of TAMs have been identified as M2a, M2b, and M2c based on the cytokines and immune responses that were produced (7). However, the drivers of TAM polarization and the modulation of the immune responses in lung cancer have not been fully elucidated.

The concentration of ATP in the extracellular milieu of solid tumors is in the range of 100 to 500 mmol/L, much higher than the interstitial fluid concentration of healthy tissues, suggesting a role of ATP in tumor proliferation and communication with immune cells (8–10). ATP can be degraded into adenosine by ectonucleotidase CD39 and CD73 in TME. Adenosine ligates A2A and A2B receptors to inhibit antigen presentation by dendritic cells (DC) and drive M2 macrophage differentiation, which results in immunosuppression (8). Genetic deletion or pharmacologic blockade of CD39 and CD73 increases antitumor immunity (11). The effects of extracellular ATP are mainly mediated by the P2X7 receptor, which is widely expressed on most immune cells, as well as tumor cells. P2X7 positively regulates NLRP3 inflammasome activation, cytokine and chemokine release, T lymphocyte survival and differentiation, transcription factor activation, and cell death (12). Host P2X7 expression supports antitumor immune response via NLRP3/IL1β signaling in DCs in melanoma B16 and colon cancer CT26 mouse tumor models (13). P2X7 blockade increases colitis-associated cancer (in a mouse model) by enhancing proliferation of intestinal epithelial cells and altering immune cell infiltration (14). A variety of cancer cells have a high expression of P2X7, which promotes tumor growth, invasiveness, and metastasis (15, 16). Consequently, P2X7 blockade has been considered for clinical trials (8). P2X7 is also highly expressed in immunosuppressive cells such as myeloid-derived suppressor cells (17) and TAMs, all which can facilitate tumor progression and metastasis. P2X7 is involved in both tumor-promoting and tumor-eradicating inflammatory responses in different types of cancers (18). However, the function and potential applications of P2X7 and TAMs in regulating immunosuppressive responses in the lung TME are still unclear.

In this study, we found that the expression of P2X7 mainly colocalized with TAMs in lung cancer, suggesting a potential role of P2X7 in TAMs. P2X7 facilitated M2 polarization of TAMs by activating the STAT6/IRF4 axis to promote tumor cell proliferation, angiogenesis, and T-cell immunosuppression in lung tumors. Inhibition or deletion of P2X7 impaired M2 polarization of TAMs and significantly restricted lung cancer growth in vivo. P2X7 inhibition overcame the resistance of lung cancers to both immunotherapy and chemotherapy, indicating a therapeutic approach for recurrent and refractory lung cancer.

Antibodies used for Western blotting, immunofluorescence, and IHC are as follows: rabbit anti-STAT6 (ab44718), rabbit anti-p-STAT6 (ab28829), rat anti-F4/80 (ab6640), and mouse anti-CD68 (ab955) were from Abcam. Rabbit anti-CD206 (18704–1-AP), rabbit anti-Arg1 (16001–1-AP), and rabbit anti-GAPDH (10494–1-AP) were from Proteintech. CREB (#4820), p-CREB (#9198), PPARγ (#2443), PI3K (#4257), p-PI3K (#4228), IRF4 (#15106), and AlexaFluor 488/555 mouse, rat, or rabbit secondary antibodies were purchased from Cell Signaling Technology (CST). Rabbit anti-P2X7 (APR-008) was from Alomone labs (Israel), and goat anti-rabbit horseradish peroxidase–conjugated IgG was from Jackson ImmunoResearch. Live/dead dye and antibodies used for flow cytometry were purchased from Biolegend unless indicated otherwise: Fixable Viability Dye (Zombie NIR Fixable Viability Kit, catalog# 423105), allophycocyanin (APC)-conjugated Rat anti-mouse F4/80 (clone BM8), phycoerythin (PE)-conjugated Rat anti-mouse CD206 (clone C068C2), FITC-conjugated Armenian Hamster anti-mouse CD3 (clone 145–2C11), PE-conjugated Rat anti-mouse P2X7 (clone 1F11), FITC-conjugated Rat anti-mouse CD11b (clone M1/70), APC-conjugated Rat anti-mouse Gr1 (clone RB6–8C5), FITC-conjugated Armenian Hamster anti-mouse CD11c (clone N418), PE-conjugated Rat anti-mouse MHCII (I-A/I-E; clone M5/114.15.2), APC-conjugated Rat anti-mouse CD4 (clone GK1.5), APC-conjugated Rat anti-mouse CD8 (clone 53–6.7), PE-conjugated mouse anti-mouse T-bet (clone 4B10), PE-conjugated Rat anti-mouse IFNγ (clone XMG1.2), PE-conjugated Rat anti-mouse TNFα (clone MP6-XT22), FITC-conjugated Rat anti-mouse CD45 (clone 30-F11), BV421-conjugated Rat anti-mouse CD3 (clone 17A2), PE-Cy7–conjugated Rat anti-mouse MHCII (clone M5/114.152), APC-conjugated mouse anti-mouse NK1.1 (Invitrogen, clone PK136), PE-eFluor610–conjugated Rat anti-mouse F4/80 (Invitrogen, clone BM8), PE-Cy5–conjugated Rat anti-mouse CD11b (Invitrogen, clone M1/70), PE-Cy5.5–conjugated Rat anti-mouse CD19 (Invitrogen, clone eBio1D3), Super Bright 780–conjugated Armenian Hamster anti-mouse CD11c (Invitrogen, clone N418), AF700-conjugated Rat anti-mouse Gr-1 (BD, clone RB6–8c5). Lipopolysaccharide (LPS), urethane, O-ATP, and A-740003 were purchased from Sigma-Aldrich. Cisplatin, A-438079 hydrochloride, and AS1517499 were purchased from MedChemExpress.

All cells were maintained in DMEM or RPMI1640 medium (Gibco) containing 10% FBS, 100 U/mL penicillin, and 100 mg/mL streptomycin, at 5% CO₂ and 37°C. Lewis lung carcinoma (LLC) cells, mouse ovarian cancer cell line ID8 cells, and mouse alveolar macrophage (AM) cell line MH-S cells were obtained from the American Type Culture Collection (LLC and ID8 obtained in 2017 and MH-S in 2016). All cell lines were routinely tested to confirm the absence of Mycoplasma contamination using the MycAwayTM-Color One-Step Mycoplasma Detection Kit (Yeasen Bio-technol), and the most recent test date for all cells was March 28, 2020. All cell lines were used within ten generations after thawing in all experiments. LLC cells stably expressing firefly luciferase were generated by our lab as described (19) and cultured in complete DMEM medium with 200 ng/mL G418 (Gibco).

Primary bone marrow–derived macrophages (BMDM) and primary AMs from mice were prepared as previously described (20, 21). To polarize macrophages into the M1 subset, BMDMs, MH-S cells, or AMs were incubated with 100 ng/mL LPS and 20 ng/mL IFNγ (Peprotech). To polarize macrophages into the M2 subset, BMDMs, MH-S cells, or AMs were incubated with 50 ng/mL recombinant mouse IL4 (R&D Systems) or LLC conditioned medium (LLC-CM; 50% of the final culture medium, prepared as previously described; ref. 22).

For P2X7 inhibitors assay, all cells were pretreated with 100 μmol/L O-ATP, 20 μmol/L A-438079 hydrochloride, and 20 μmol/L A-740003 for 1 hour respectively and then stimulated with IL4 or LLC-CM. Then cells were harvested and assessed by quantitative real-time PCR (Q-PCR), immunoblotting, or flow cytometry.

Naïve CD4⁺ or CD8⁺ T cells were isolated from spleens of wild-type (WT) mice by MojoSort mouse CD4 (catalog# 480006) or CD8 (catalog# 480008) T-cell isolation kit (Biolegend) according to the manufacturer's instructions and expanded or activated by Dynabeads mouse T-activator CD3/CD28 (Gibco, catalog# 11452D) according to the manufacturer's instructions and 30 U/mL recombinant mouse IL2 (R&D Systems). F4/80⁺ cells were isolated from single-cell suspensions of LLC tumors from P2X7^+/+ or P2X7^−/− donor mice by MagniSort mouse F4/80-positive magnetic bead isolation (Invitrogen, Thermo Fisher Scientific, catalog# 8802–6863) according to the manufacturer's instructions. For T-cell activity assay, CD4⁺ or CD8⁺ T cells were mixed 1:1 with or without P2X7^+/+ or P2X7^−/− tumor-derived macrophages and cocultured for 3 days. Then, cells were harvested and stimulated with Leukocyte activation cocktail (with BD Golgiplug; BD, catalog# 550583) at 37°C for 4 to 6 hours according to the manufacturer's instructions and analyzed by flow cytometry.

For cell viability studies, LLC cells were seeded into 96-well plates (Corning) at 5 × 10⁴ cells/well in a volume of 100 μL complete DMEM containing 10% FBS, 100 U/mL penicillin, and 100 mg/mL streptomycin overnight. Then diluent O-ATP or cisplatin was added into each well at the indicated concentration and incubated for 24 hours. The cell viability of cultured LLC cells was quantified using cell counting kit-8 assay according to the manufacturer's instructions (Dojindo Molecular Technologies, catalog# CK04). The percentage growth inhibition of O-ATP- or cisplatin-treated cells relative to vehicle-treated cells was calculated. For cell proliferation studies, LLC cells were prestained with carboxyfluorescein diacetate succinimidyl ester (CFSE) according to the manufacturer's CFSE Cell Labeling Protocol (eBioscience, catalog# 65–0850). Then, cells were stimulated with WT or P2X7^−/− TAM conditioned medium or cocultured with M1, M2 BMDMs, and TAMs from tumors of WT or P2X7^−/− mice.

P2X7 knockout mice were generated by CRISPR/Cas9 systems as described previously (23), and corresponding C57BL/6 WT mice were purchased from the Shanghai Laboratory Animal Company. For urethane induction of lung tumors, 6-week-old P2X7^+/+ and P2X7^−/− mice were i.p. injected with 200 μL urethane (1 g/kg, ethyl carbamate) dissolved in PBS once per week for 10 weeks. All mice were euthanized for lung tumor examination at 6 months after injection of carcinogen. For tumor challenge and tumor therapy experiments, subcutaneous tumor transplantation and orthotopic lung transplantation mouse models of lung cancer were established as described previously (24). Briefly, 6- to 8-week-old mice were grouped and anesthetized by i.p. injection with pentobarbital sodium dissolved in PBS (50 mg/kg). Then, 1 × 10⁶ LLC cells in 200 μL PBS were subcutaneously inoculated into the back or 1 × 10⁶ LLC cells in 100 μL PBS were orthotopically injected into the lungs of 6- to 8-week-old C57BL/6 mice as described previously (24).

In tumor therapeutic studies, each group of mice bearing tumors was treated by i.p. injection with 7 mg/kg of P2X7 inhibitor O-ATP, 2 mg/kg of A-438079 hydrochloride, or 0.2 mg/kg of A-740003 every other day, 1 mg/kg of anti–PD-1 antibody (BioXcell) or 3 mg/kg of cisplatin every 3 days or equivalent volumes of PBS as control for 22 or 30 days, respectively. In combination groups, mice bearing tumors were treated with 7 mg/kg of O-ATP plus 1 mg/kg of anti–PD-1 or 3 mg/kg of cisplatin, 2 mg/kg of A-438079 hydrochloride plus 3 mg/kg of cisplatin, or 0.2 mg/kg of A-740003 plus 3 mg/kg of cisplatin respectively for 22 or 30 days.

BMDMs from P2X7^+/+ mice (treated with or without P2X7 receptor inhibitor O-ATP) or BMDMs from P2X7^−/− mice were mixed 1:4 with LLC cells in 200 μL PBS, and 1 × 10⁶ total cells were injected subcutaneously into WT C57BL/6 mice. Tumor dimensions were measured with a caliper (length × width²/2) once every 5 days. For in vivo imaging, mice were anesthetized with isoflurane and injected with 150 mg/kg substrate (D-luciferin) as described previously (19). Luminescence was detected and analyzed using Living Image software (PerkinElmer). In another macrophage adoptive transfer study, F4/80⁺ cells were isolated from single-cell suspensions of LLC tumors from donor mice by MagniSort mouse F4/80-positive magnetic bead isolation (Invitrogen, Thermo Fisher Scientific, catalog# 8802–6863). Then LLC cells were mixed 1:1 with or without P2X7^+/+ or P2X7^−/− F4/80⁺ cells in 200 μL PBS, and 1 × 10⁶ total cells were injected subcutaneously into P2X7^+/+ or P2X7^−/− C57BL/6 mice, respectively.

Single-cell suspensions were prepared for flow cytometry as described previously (19), with slight changes. Briefly, cultured BMDMs, MH-S cells, or LLC cells were digested with trypsin and filtered with a 70 μm mesh to obtain single-cell suspensions. For tumor-infiltrating leukocytes, mice were sacrificed at specific time points, and excised tumors from the host mice were minced and digested in RPMI1640 with 10 U/mL collagenase I, 400 U/mL collagenase IV, and 30 U/mL DNase I at 37°C for 30 minutes. Then, tissue suspensions were passed through a 70 μm strainer and washed with PBS containing 2% FBS. All samples were labeled with indicated fluorescence-conjugated antibodies after blocking FcγII/III with anti-CD16/32 (Biolegend, catalog# 101302) for 40 minutes at 4°C. Intracellular staining of IFNγ and TNFα in T cells was determined as follows: cells were stained with surface marker then fixed and permeated with Fixation and Permeabilization Solution (BD Bioscience, catalog# 555028) overnight, washed 3 times, and stained with TNFα and IFNγ for 40 minutes at 4°C. Data were acquired on LSRFortessa or Caliber and analyzed using FlowJo software (Tree Star).

Cells were treated with indicated substances and then total RNA was isolated by applying RNAiso plus reagent (TAKARA, catalog# 9109), and RNA content was measured by NanoDrop2000 (Thermo Fisher Scientific). cDNA was synthesized with 500 ng RNA using the Prime Script RT Master Mix Perfect Real Time kit (TAKARA, catalog# RR037A), according to the manufacturer's protocol. The mRNA expression of each gene was detected by Q-PCR using SYBR Green premix (TAKARA, catalog# RR420A), and the data were analyzed by the QuantStudio 3 Real-Time PCR System (Applied Biosystems). The expression of each gene (P2X7, CD206, Arg1, Ym1, Fizzl, Mrc2, IL10, iNOS, TNFα, and Cxcl10) was normalized to the expression level of GAPDH and reported as relative mRNA expression (2^−ΔΔCt) or fold change. The sequence-specific primers are shown in Supplementary Table S1.

To assay secreted VEGFA, WT or P2X7^−/− BMDMs and LLC cells were seeded into 12-well plates at 5 × 10⁵ cells/well respectively, and LLC cells were also cocultured with WT or P2X7^−/− BMDMs in 12-well plates. All cells were cultured in DMEM containing 10% FBS for 24 hours, and the supernatants were collected. The concentration of VEGFA in the supernatant was measured by an ELISA kit (Dakewe Biotech; catalog# CL89125K-96) according to the manufacturer's instructions.

For protein extraction, BMDMs stimulated with LPS, IFNγ, IL4, or LLC-CM were harvested and lysed with RIPA buffer (CoWin Biosciences; catalog# CW2333) supplemented with complete Mini Protease and Phosphatase inhibitor Cocktail (Roche, catalog# 4693159001 and 4906837001). Cell lysates were quantified by BCA protein assay (Thermo Fisher, catalog# 23227) and separated by standard SDS-PAGE. Then proteins were electrotransferred to nitrocellulose membranes (Millipore, catalog# N8645) and blocked with 5% BSA (Sigma-Aldrich, catalog# A4161) for 1 hour. Membranes were incubated with primary antibodies, and protein bands were developed by incubating with appropriate fluorophore-conjugated secondary antibodies, followed by imaging using the Odyssey laser digital imaging system (Gene Company). Data were quantified by Image J, and the ratio of each protein expression was normalized to the expression level of GAPDH.

Organs and tumors were harvested, fixed in 4% paraformaldehyde overnight, and embedded into paraffin. Tissue blocks were sectioned at 6-μm-thick, and slides were deparaffinized and rehydrated, then stained with hematoxylin and eosin (H&E) solution. For immunofluorescence assay, human lung adenocarcinoma tissue microarray (Hlug-Ade060PG-01, Outdo Biotech Company) and mouse lung tumor sections were subjected to antigen retrieval, permeated with 0.1% Triton X-100, blocked with goat serum, and stained with the primary antibodies of F4/80, CD68, P2X7, CD206, CD31, p-STAT6, and IRF4 overnight at 4°C. Then, the corresponding secondary antibodies were incubated for 1 hour at room temperature, and DAPI was used to stain nuclei. All stained sections were examined and visualized by microscopy (Leica Microsystems).

Statistical analyses were analyzed by GraphPad Prism 8.0.2 (GraphPad Software). Statistical differences between groups were analyzed using unpaired two-tailed Student t test, or one-way or two ANOVA analysis. Survival analysis was performed using the Kaplan–Meier method and the log-rank test. All data were shown as mean ± SEM. P value ≤ 0.05 was considered to be statistically significant (ns, no significant difference; *, P < 0.05; **, P < 0.01; and ***, P < 0.001).

All animal experiments conformed to the regulations drafted by the Association for the Assessment and Accreditation of Laboratory Animal Care in Shanghai and were performed in direct accordance with the animal care guidelines of the Ministry of Science and Technology of the People's Republic of China. The protocol was approved by the East China Normal University Center for Animal Research (AR2013/08002). All surgeries were performed under anesthesia, and all efforts were made to minimize suffering.

To investigate the potential role of P2X7 in lung cancer formation, we examined the expression of P2X7 in various tumor-infiltrating leukocytes in mouse LLC-implanted tumor tissues by flow cytometry. Approximately 80% of macrophages were P2X7 positive, with little expression in other cell types (Fig. 1A; Supplementary Fig. S1A). We performed immunostaining of mouse lung tumor tissues and a human lung adenocarcinoma tissue microarray to detect the expression of P2X7 in lung cancer tissues. P2X7 (red) was predominantly expressed by F4/80⁺ (green) or CD68⁺ (green) macrophages in lung tumors (Fig. 1B; Supplementary Fig. S1B). As M2-like TAMs play a vital role in promoting tumorigenesis (25), we checked the expression of P2X7 in M1/M2 polarized macrophages and in primary BMDMs, the AM cell line MH-S cells, and primary AMs. P2X7 was significantly upregulated in IL4-induced M2 macrophages and consistent with the expression of CD206 and Arg1, both the surface markers for M2 macrophages (Fig. 1C–F). In addition, compared with LLC tumor cells, P2X7 had higher expression in TAMs from mouse lung tumors (Fig. 1G). P2X7 was predominantly expressed in TAMs and DCs rather than LLC tumor cells, neutrophils, and inflammatory monocytes (Neu/mono; Fig. 1H). In particular, the expression of P2X7 was increased nearly 4-fold in TAMs within the LLC TME as compared with their counterparts in the spleen (Fig. 1I). These results indicated that P2X7 expression was upregulated in TAMs, which might play fundamental roles in the lung cancer immune microenvironment.

To investigate the fundamental roles of P2X7 in lung tumorigenesis, we established three different tumor models including a urethane-induced lung tumor model, the LLC subcutaneous transplantation tumor model, and the LLC lung orthotopic implant tumor model by using WT and P2X7-deficient (P2X7^−/−) mice. The number of urethane-induced tumor nodes in P2X7 knockout mice was significantly reduced as compared with WT mice (Fig. 2A and B). H&E staining also showed that there were numerous urethane-induced tumor lesions in the lungs of WT mice, whereas lungs from P2X7 knockout mice showed little or no tumor lesions (Fig. 2C). P2X7 deletion obviously regressed tumor growth and tumor angiogenesis, and extended survival rate of tumor-bearing mice in the LLC subcutaneous transplantation model (Fig. 2D–G) and in the LLC lung orthotopic implant model (Supplementary Fig. S2A–S2C).

As described above, P2X7 was highly expressed particularly in TAMs. To determine if blockade of P2X7 in macrophages disrupted tumor formation, we subcutaneously coinjected WT mice with both luciferase-expressing LLC cells and either WT BMDMs, P2X7^−/− BMDMs, or WT BMDMs treated with the P2X7 inhibitor O-ATP. P2X7 inhibition or loss in macrophages restrained tumor progression. After injection for 10 days, tumor growth rate of LLC cells coinjected with WT macrophages was 2-fold higher than that of LLC cells with P2X7 inhibited or deleted macrophages (Fig. 2H and I). P2X7-inhibited or -deleted macrophages prolonged the survival of tumor mice compared with WT macrophages (Fig. 2J). We also performed adoptive transfer tests by subcutaneously inoculating LLC cells with WT or P2X7^−/− BMDMs in WT or P2X7 knockout host mice. Tumor growth was significantly suppressed when P2X7-deficient macrophages were inoculated in either WT or P2X7 knockout host mice, whereas WT macrophages promoted tumor growth in both WT and P2X7 knockout host mice (Fig. 2K). Collectively, P2X7 in macrophages was essential for lung cancer formation with P2X7 deficiency potentially impairing the protumoral function of M2-like TAMs.

To further examine the role of P2X7 in macrophage polarization, we isolated murine BMDMs and AMs from WT and P2X7^−/− mice. We treated BMDMs or AMs with LPS plus IFNγ to induce M1 polarization or with IL4 or LLC-CM to induce M2 polarization. Although P2X7 deficiency had little effect on the expression of M1 markers (iNOS, TNFα, and CXCL10) in BMDMs (Supplementary Fig. S3A–S3C), it resulted in a remarkable decrease of M2 marker expression, including Arg1, Ym1, Mrc2, CD206, and Fizzl in IL4 or LLC-CM–stimulated BMDMs or AMs (Fig. 3A–D and F–H). Flow cytometry demonstrated that P2X7 deficiency impaired polarization of BMDMs to F4/80⁺CD206⁺ M2 cells (Fig. 3E) but had little influence on F4/80⁺CD16/32⁺ M1 cells (Supplementary Fig. S3D). We treated the M2-like BMDMs or MH-S cells induced by IL4 or LLC-CM with the P2X7 inhibitor O-ATP, A-438079 hydrochloride, and A-740003. P2X7 inhibition significantly reduced the expression of classical M2 markers Arg1, CD206, and Fizzl both in BMDMs (Supplementary Fig. S4A–S4I) and in MH-S cells (Supplementary Fig. S4J–S4M).

We characterized M2 TAMs in vivo by sorting the TAMs (CD45⁺CD11c⁻CD11b⁺Gr1⁻F4/80⁺) in subcutaneously implanted LLC tumors from WT or P2X7 knockout mice and examining the polarized phenotypes of such cells by Q-PCR. The expressions of CD206, Arg1, and IL10 were all significantly reduced in P2X7-deficient TAMs, which was consistent with the in vitro data of macrophages (Fig. 3I–K). Next, we performed the immunofluorescence staining and observed a decrease of F4/80⁺CD206⁺ TAMs in urethane-induced lung tumors (Fig. 3L) and LLC orthotopic implanted lung tumors (Supplementary Fig. S2D) of P2X7 knockout mice. In order to verify the key role of P2X7 in TAMs, we also established a murine ID8 epithelial ovarian cancer model which typically has abundant infiltrating macrophages and is a good model to test the innate immune response of myeloid cells (26). As expected, the average volume of ascites was significantly reduced in P2X7 knockout mice (3.24 mL) when compared with WT mice (8.32 mL; Supplementary Fig. S2E and S2F). Notably, flow cytometry analysis showed that the proportion of F4/80⁺CD206⁺ TAMs approximately dropped from 16.2% (WT mice) to 5.7% (P2X7^−/− mice) in ascites of i.p. implanted ID8 tumor model (Supplementary Fig. S2G). Collectively, these results indicated that P2X7 contributes to the M2 polarization of TAMs both in vitro and in vivo, which facilitated the progression of lung cancer.

To understand the mechanism of P2X7 in regulating TAM polarization, we first stimulated WT and P2X7^−/− BMDMs with IL4 and showed that both the phosphorylation of PI3K or CREB and the expression of PPAR-γ were barely changed in the context of P2X7 deficiency (Fig. 4A and B). However, P2X7-deficient macrophages undergoing IL4 stimulation showed a dramatic decrease in phosphorylation of STAT6 at tyrosine 641 site and decreased expression of IRF4 (Fig. 4A and B). To mimic the polarization process of macrophages in the TME, we treated BMDMs with LLC-CM, and both p-STAT6 and IRF4 were markedly reduced in P2X7-deficient macrophages, which was consistent with the results from IL4-induced M2 cells (Fig. 4C). As the IL4-STAT6 pathway induces IRF4 activation in M2-polarized macrophage to facilitate glycolysis (27), we treated IL4-induced BMDMs with or without the STAT6 inhibitor AS1517499 to test the expression of IRF4. As shown in Fig. 4D, the activity of Arg1 and IRF4 activated by IL4 at 60 minutes was abrogated by STAT6 inhibition, which indicated that STAT6 is required for IRF4 expression to increase M2 polarization of macrophages. To confirm the role of P2X7-involved STAT6/IRF4 signaling in macrophage polarization in lung tumors, we examined STAT6 phosphorylation and IRF4 expression by immunofluorescence staining. We observed a dramatic reduction in numbers of F4/80⁺p-STAT6⁺ and F4/80⁺IRF4⁺ TAMs in lung tumor tissues of urethane-induced or orthotopic LLC-implanted P2X7 knockout mice (Fig. 4E and F), which indicated that the P2X7-STAT6/IRF4 signaling pathway was crucial for TAM polarization and lung tumor progression.

TAMs play crucial roles in multiple aspects of tumor development, such as angiogenesis, tumor cell proliferation, cancer stem-cell niche formation, and recruitment or immunosuppression of T cells (28). Some studies have reported that disruption of the CSF1–CSF-1R axis, the CCL2-CCR2 axis, PI3Kγ, or the BTK signaling pathway by antibodies or specific inhibitors can deplete or reprogram TAMs and further increase T-cell infiltration and activation (29–31). To investigate the effect of P2X7-deficient TAMs on educating T cells in the TME, we examined the proportion of CD4⁺ and CD8⁺ T cells by flow cytometry in tumors of mice coinjected with macrophages and LLC cells. Compared with the WT macrophage and LLC coinjection group, P2X7-deficient TAMs promoted CD8⁺ T-cell mobilization and the ratio of CD8⁺/CD4⁺ T cells, but not CD4⁺ T-cell influx (Fig. 5A–D).

To further assess whether P2X7 deletion in macrophages could alleviate the immunosuppressive phenotype of lung tumor–infiltrating T cells, we treated naïve CD4⁺ and CD8⁺ T cells with CD3/CD28 coligation activated beads and tested the activity of inducible T cells. Both IFNγ and TNFα expressions were upregulated in activated CD4⁺ and CD8⁺ T cells (Fig. 5E; Supplementary Fig. S5A). However, when we cocultured CD4⁺ or CD8⁺ T cells with F4/80⁺ macrophages harvested from LLC-implanted tumors, we found that WT macrophages significantly inhibited the expression of IFNγ and TNFα in CD8⁺ T cells, but had little effect on CD4⁺ T cells (Supplementary Fig. S5A and S5B). Conversely, P2X7-deficient macrophages diminished their capacity to suppress CD4⁺ or CD8⁺ T-cell activation as compared with WT macrophages (Fig. 5E; Supplementary Fig. S5B). Accordingly, P2X7 deletion reprogrammed M2-like TAMs and relieved their suppression of T-cell antitumor immune responses.

TAMs promote tumor cell proliferation and angiogenesis (25). To explore the functional role of P2X7-deficient TAMs in the lung TME, we cocultured CFSE-labeled LLC cells with different subtypes of BMDMs (M1 macrophages stimulated with LPS and IFNγ, M2 macrophages stimulated with IL4) and found that M2 macrophages were more effective than M1 macrophages at promoting LLC cell proliferation (Fig. 6A). Then, we cultured the CFSE-labeled LLC cells with TAMs which had been sorted from LLC-induced tumors or with TAM conditioned medium (TAM-CM). Compared with WT TAMs and their conditioned medium, both P2X7-deficient TAMs and their conditioned medium lost the abilities to promote LLC cell proliferation (Fig. 6B and C). H&E staining and CD31 immunofluorescence showed that the number of vessels in tumors coinjected with WT macrophages and LLCs was higher than when P2X7^−/− macrophages were used (Fig. 6D and F). VEGFA secretion was also increased in the cell supernatant of LLCs cocultured with WT macrophages but not with P2X7^−/− macrophages (Fig. 6E). Collectively, P2X7 deficiency impaired the polarization of macrophages to the M2 phenotype and reduced tumor cell proliferation and angiogenesis.

As macrophage P2X7 promoted lung tumor progression and inhibition of P2X7 impaired macrophage M2 polarization, we speculated that blockade of P2X7 could influence the outcome of lung cancer therapy. To evaluate this, we established the subcutaneous implanted LLC lung tumor model and treated mice bearing lung tumors with the P2X7 inhibitor O-ATP, PBS as a control, anti–PD-1 therapy, cisplatin, or the combination of O-ATP with anti–PD-1 or cisplatin. O-ATP, anti–PD-1, or cisplatin substantially restrained lung tumor growth to approximately half that of PBS-treated mice and correspondingly extended mouse survival. Surprisingly, although O-ATP, anti–PD-1, and cisplatin similarly inhibited tumor progression respectively, the combination therapy (O-ATP plus anti–PD-1 or O-ATP plus cisplatin) dramatically suppressed tumor growth on both size and weight (Fig. 7A–F). The combination of O-ATP and cisplatin significantly increased the survival of mice bearing tumors to 70%, whereas the survival of other treatment groups was less than 40% (Fig. 7G). Because tumor growth was inhibited in O-ATP–treated mice, we examined whether O-ATP kills tumor cells directly. Cisplatin significantly reduced LLC survival (Fig. 7H), whereas O-ATP had little effect on the cell viability of LLC cells (Fig. 7I). M2-like F4/80⁺CD206⁺ TAMs were significantly reduced in O-ATP–treated mice (Fig. 7J), similar to our in vitro results showing that O-ATP impaired M2 polarization of macrophages. To further confirm the effects of P2X7 blockade, we also treated mice bearing lung tumors with two another P2X7 inhibitors A-438079 hydrochloride and A-740003. As expected, both A-438079 hydrochloride and A-740003 suppressed tumor growth and synergistically improve the effect of chemotherapy in lung tumor (Supplementary Fig. S6A–S6C). Collectively, P2X7 blockade impaired M2 macrophage polarization and overcame the resistance of anti–PD-1 immunotherapy and chemotherapy in treating lung cancer.

Tumor heterogeneity is a key characteristic of lung cancer, in which carcinoma cells are closely associated with endothelial cells, fibroblasts, infiltrating immune cells, and the extracellular matrix. Among these, myeloid cells including TAMs and tumor-associated neutrophils are implicated in regulating the tumor immune microenvironment (32, 33). In particular, TAMs frequently constitute the largest population of leukocytes in most tumors and can produce multiple cytokines and chemokines to facilitate tumor progression by promoting tumor cell proliferation, matrix degradation, angiogenesis, and immunosuppression (28, 29). Switching macrophage polarization could restore efficacy of antitumor checkpoint immunotherapy (34, 35). Thus, re-education of TAMs toward an antineoplastic phenotype has become a promising approach to boost the effectiveness of checkpoint inhibitors, chemotherapy, and other treatments. In this study, we identified P2X7 as a key regulator that promoted M2-like macrophage polarization by STAT6/IRF4 signaling to accelerate tumor cell proliferation, angiogenesis, and immunosuppression in three lung cancer mouse models. Our data suggested that blockade of P2X7 by pharmacologic inhibitors or therapeutic antibodies may yield a promising therapy for patients with lung cancer.

Unlike in M1 macrophages, P2X7 activation is uncoupled to NLRP3-inflammasome activation in M2 phenotype macrophages and mediates chronic rejection of heart allografts (36, 37), suggesting that the function of P2X7 in macrophage-mediated inflammation varies during the dynamics of macrophage polarization. Here, we demonstrated that the expression of P2X7 was specifically increased in TAMs and in M2-polarized macrophages as compared with M1-like or resting macrophages, which reinforced the M2-like polarization of TAMs both in vitro and in vivo. Deletion or blocking P2X7 in macrophages significantly reduced the expression of M2-related biomarkers, such as CD206, Arg1, IL10, and Ym1. The number of M2-phenotype TAMs was reduced in tumors with P2X7 deletion. All together, these results demonstrated that P2X7 would accelerate lung cancer formation via controlling the polarization of TAMs. Most strikingly, pharmacologic inhibition of P2X7 by O-ATP, A-438079 hydrochloride, and A-740003 markedly suppressed tumor growth, and synergized with anti–PD-1 immunotherapy and chemotherapy. Thus depletion of P2X7 was sufficient to increase CD8⁺ T-cell infiltration in tumors and that P2X7 inhibition in mice induced a remarkable response to anti–PD-1 antibody treatment accompanied by impaired M2 macrophage activation. In addition, low concentrations of ATP also stimulate P2X7 in tumor cells to promote proliferation and neoplastic transformation (38), implying that inhibition of P2X7 not only reversed the immunosuppressive TME but also restricted the proliferation of tumor cells. Thus, these data indicated the potential of P2X7 inhibition as a method to overcome therapy resistance in lung cancer.

Elevated STAT6 signaling was observed in recurrent CSF-1R inhibitor–treated patients with glioblastoma multiforme, indicating the necessity of combination therapy to expose STAT6 signaling dependency in recurrent disease (39). STAT6 signaling in TAMs promotes tumor growth and metastatic niche formation in breast cancer (40), and nonhematopoietic epithelial STAT6 activation accelerates intestinal inflammation and colorectal cancer (41). Here, P2X7 was required for M2 TAM polarization which was induced by IL4 or lung tumor cell supernatant through the activation of STAT6/IRF4 signaling. Thus, it was possible that P2X7 inhibition could help overcoming intrinsic resistance to CSF-1R inhibition, and consequently, those patients may benefit from combinatorial inhibition of these pathways from the outset. IL4- and IL13-induced STAT6 signaling is also responsible for the hearing impairment caused by cisplatin (42). This implies that blockade of P2X7 not only facilitates lung cancer therapy through reversing M2 macrophage polarization but also may reduce cisplatin-induced hearing impairment.

P2X7 played a vital role in lung tumor progression by enhancing the M2-like polarization of TAMs through activation of the STAT6/IRF4 axis. P2X7 blockade reprogramed protumoral TAMs to an antitumor phenotype which helped overcomes resistance to anti–PD-1 immunotherapy and chemotherapy. Therefore, blockade of P2X7 could be new immunotherapy for lung cancer that warrants clinical investigation.

J. Qin reports grants from National Natural Science Foundation of China, Shanghai Super Postdoctoral Incentive Program, China Postdoctoral Science Foundation, and Shanghai Sailing Program during the conduct of the study. M. Qian reports grants from National Natural Science Foundation of China during the conduct of the study. B. Du reports grants from National Key R&D Program of China and National Natural Science Foundation of China during the conduct of the study. No potential conflicts of interest were disclosed by the other authors.

J. Qin: Data curation, investigation, writing–original draft. X. Zhang: Investigation. B. Tan: Investigation. S. Zhang: Investigation. C. Yin: Investigation. Q. Xue: Investigation. Z. Zhang: Investigation. H. Ren: Formal analysis. J. Chen: Supervision. M. Liu: Supervision, funding acquisition. M. Qian: Supervision. B. Du: Conceptualization, supervision, funding acquisition, writing–original draft.

This work was supported in part by grants from the National Key R&D Program of China (2018YFA0507001); National Natural Science Foundation of China (81672811, 31770969, 81871250, and 81902892); Shanghai Super Postdoctoral Incentive Program; China Postdoctoral Science Foundation (2018M640364); Shanghai Sailing Program (19YF1414400); and Innovation Program of Shanghai Municipal Education Commission (2017-01-07-00-05-E00011).

The authors thank Dr. Stefan Siwko for his kind suggestions and for proofreading this article.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.