Biophysical reprogramming of immunosuppressive neutrophils by cold atmospheric plasma reinvigorates antitumor immunity

Ganyu Wang; Zhiwei Cui; Xin Jiang; Ying Sun; Yifan Liu; Xinyuan Li; Wushan Li; Ruirui Sun; Xuemei Liu; Nengwang Yu; Xiaolong Wang; Benkang Shi; Weiqiang Jing

doi:10.1126/sciadv.aee2483

. 2026 Aug 5;12(32):eaee2483. doi: 10.1126/sciadv.aee2483

Biophysical reprogramming of immunosuppressive neutrophils by cold atmospheric plasma reinvigorates antitumor immunity

Ganyu Wang ^1,^†, Zhiwei Cui ^1,^†, Xin Jiang ^1,^†, Ying Sun ², Yifan Liu ², Xinyuan Li ¹, Wushan Li ³, Ruirui Sun ², Xuemei Liu ¹, Nengwang Yu ^1,^*, Xiaolong Wang ^2,^*, Benkang Shi ^1,^*, Weiqiang Jing ^1,^*

^¹Department of Urology, Qilu Hospital of Shandong University, Cheeloo College of Medicine, Shandong University, 107 Cultural West Road, Jinan, Shandong Province, 250012, China.

^²School of Electrical Engineering, Shandong University, Jinan, Shandong Province, 250061, China.

^³Department of Obstetrics, Jinan Maternity and Child Care Hospital Affiliated to Shandong First Medical University, Jinan, Shandong Province, 250000, China.

Corresponding author. Email: yunengwangsdu@163.com (N.Y.);

Corresponding author. Email: wangxiaolong@sdu.edu.cn (X.W.);

Corresponding author. Email: shibenkangsdu@163.com (B.S.);

Corresponding author. Email: wjing1@sdu.edu.cn (W.J.)

^†

These authors contributed equally to this work.

Roles

Ganyu Wang: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing - original draft, Writing - review & editing

Zhiwei Cui: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing - original draft, Writing - review & editing

Xin Jiang: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing - original draft, Writing - review & editing

Ying Sun: Methodology, Software, Validation

Yifan Liu: Methodology, Validation

Xinyuan Li: Methodology, Software

Wushan Li: Methodology, Software, Validation

Ruirui Sun: Methodology, Validation

Xuemei Liu: Methodology, Software, Validation

Nengwang Yu: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing

Xiaolong Wang: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing

Benkang Shi: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing - original draft, Writing - review & editing

Weiqiang Jing: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing

PMCID: PMC13440387 PMID: 42555723

Abstract

Neutrophils are increasingly recognized as key orchestrators of the immunosuppressive tumor microenvironment, yet their intrinsic plasticity, short lifespan and resistance to genetic manipulation have impeded therapeutic targeting. Here, we report a non-pharmacological, biophysical immunomodulatory strategy based on cold atmospheric plasma (CAP) to reprogram tumor-associated neutrophils and restore antitumor immunity. We show that CAP simultaneously delivers reactive oxygen species and redox cues that inhibit mitophagy, thereby restoring mitochondrial membrane potential and oxidative metabolism in neutrophils. This metabolic reinstatement drives a shift from immunosuppressive to immunostimulatory phenotypes. In both syngeneic and humanized bladder cancer models, intravesical CAP reshapes the myeloid landscape, enhances T cell infiltration, suppresses tumor progression, and sensitizes tumors to PD-1 blockade. These findings establish CAP as a locoregional, drug-free biophysical modality capable of overcoming neutrophil-mediated immune suppression and provide a materials-based framework for modulating innate immunity in solid tumors.

Cold atmospheric plasma reprograms neutrophils to restore antitumor immunity and enhance immunotherapy.

INTRODUCTION

Immunotherapy has revolutionized the treatment landscape of cancer, offering durable clinical responses in a subset of patients. However, its efficacy remains limited in many solid tumors due to the presence of an immunosuppressive tumor microenvironment (TME), which blunts antitumor T cell activity and fosters immune evasion (1, 2). While much attention has been given to regulatory T cells, tumor-associated macrophages, and myeloid-derived suppressor cells, growing evidence highlights tumor-associated neutrophils (TANs) as critical yet underappreciated drivers of immune suppression and disease progression (3, 4).

Neutrophils are the most abundant leukocytes in circulation and are among the earliest immune cells recruited to tumor sites (5, 6). Within the TME, TANs undergo phenotypic and functional reprogramming, adopting immunosuppressive features that inhibit cytotoxic T cell responses, support angiogenesis, and promote metastasis (7–9). Despite their profound impact on cancer immunity, neutrophils have remained largely untargeted in clinical oncology. Their short lifespan, phenotypic plasticity, and resistance to genetic manipulation pose significant barriers to therapeutic intervention, and no approved therapies to date directly reprogram or eliminate immunosuppressive TANs (10–12).

Overcoming neutrophil-mediated immune suppression requires innovative strategies that can modulate neutrophil function in a temporally precise and spatially confined manner, without relying on conventional pharmacological or genetic interventions. Cold atmospheric plasma (CAP), a partially ionized gas containing reactive oxygen and nitrogen species (ROS/RNS), photons, and charged particles, has emerged as a versatile biophysical modality with selective antitumor activity (13, 14). Unlike small-molecule drugs or immune checkpoint inhibitors, CAP exerts its effects through multimodal mechanisms, including the induction of oxidative stress, disruption of redox homeostasis, and modulation of cell fate pathways (15, 16). While its direct cytotoxic effects on microbe and cancer cell are increasingly well-characterized, the capacity of CAP to reprogram immune cells within the tumor microenvironment remains largely unexplored (17–20). Given the central role of redox signaling in neutrophil metabolism and plasticity, we hypothesized that CAP could serve as a controllable platform to modulate neutrophil immunophenotypes and enhance the efficacy of existing immunotherapies (21–23).

In this study, we demonstrate that intravesical CAP treatment induces a phenotypic switch in TANs toward a cytotoxic state by inhibiting mitophagy and restoring mitochondrial homeostasis. This reprogramming enhances T cell infiltration, reduces tumor burden, and sensitizes tumors to immune checkpoint blockade in both syngeneic and humanized models. These findings establish CAP as a drug-free, locoregional strategy to overcome innate immune resistance and provide a framework for targeting neutrophil-driven immune suppression in cancer.

RESULTS

Tumor-associated neutrophils exhibit an immunosuppressive phenotype correlated with tumor invasiveness in bladder cancer

Increasing evidence indicates that neutrophils play a pivotal yet context-dependent role in shaping the tumor immune microenvironment and modulating response to immunotherapy (24–26). In particular, TANs can undergo profound reprogramming within the tumor milieu, transitioning from their homeostatic, cytotoxic state in peripheral blood to an immunosuppressive, pro-tumor phenotype (27, 28). Understanding the phenotypic and functional diversity of neutrophils across tissue compartments and tumor types is critical for advancing neutrophil-targeted therapies (29–31).

To systematically characterize tumor-driven neutrophil reprogramming, we leveraged public single-cell RNA sequencing datasets (GSE267718 and GSE201425) from human bladder cancer and cholangiocarcinoma (fig. S1, A to E) (32, 33). Comparative analysis of canonical neutrophil markers revealed that, relative to peripheral blood neutrophils (PBNs), TANs exhibit downregulation of activation-related genes (S100A9, TLR4, NCF1, NCF2) and altered chemokine receptor expression (CX3CR1, CCR2), but upregulate the inhibitory molecule HAVCR2 (Tim-3), which impedes T cell responses and prolongs neutrophil survival (Fig. 1A and fig. S1F). These findings suggest a tumor-driven functional polarization of neutrophils toward an immunoregulatory phenotype.

Fig. 1. — (A) Violin plots showing the relative expression levels of indicated genes in matched PBNs and TANs from patients. Gene expression was derived from scRNA-seq and normalized using SCTransform. (B) Schematic overview of the experimental workflow for human bladder tumor tissue processing, including tissue preparation, probe hybridization, and Xenium in situ spatial transcriptomic analysis. Created in BioRender. A, H. (2026) https://BioRender.com/dcf1ofc. (C) Markers for cell type recognition. List of differentially expressed genes used for cell type identification across all immune and non-immune populations. (D) Representative images of bladder tumor sections from NMIBC (#1 and #2) and MIBC (#3 and #4) bladder tumor samples. t-SNE analysis displaying cell population clusters of neutrophils, B cell, T cell, macrophage, endothelial cell, fibroblast, tumor cell populations across all samples. Unprocessed images with corresponding processed images with lineage assignment are shown. (E) Balloon plot representing the difference in average signal intensity of selected marker within each indicated cell populations between MIBC and NMIBC samples. Circle size represents relative expression level; color intensity reflects the magnitude of difference. Avg. exp., average expression. (F) Violin plots displaying immune activation scores of neutrophils from the Xenium in situ data. (G) Representative flow cytometry plot and quantification of human CD45⁺CD66b⁺ neutrophils in adjacent normal margins and matched bladder tumor tissue from patients (n = 6 donors). (H) Representative confocal microscopy images of human neutrophils isolated from bladder tumor tissue, stained with CD66b and DAPI. Scale bar, 5 μm. (I) Quantification of intracellular ROS in isolated PBNs and TANs from human samples (n = 8 donors) and MB49-bladder tumor bearing mice (n = 8 mice). Data are presented as means ± SD. **P < 0.01, ***P < 0.001, ****P < 0.0001. Statistical values were calculated using Wilcoxon Rank Sum Test (A), Student’s t-test [(G) and (I)].

Given our previous findings that bladder cancer is particularly enriched for TANs, we selected this tumor type as a model to dissect the link between neutrophil state and disease invasiveness (34). Using high-throughput spatial transcriptomics (Xenium 5 K, 10× Genomics), we performed single-cell gene expression profiling on tissue samples from non-muscle-invasive bladder cancer (NMIBC, #1 and #2) and muscle-invasive bladder cancer (MIBC, #3 and #4) patients, annotating key immune and stromal cell types using established markers (Fig. 1, B and C). We present representative cystoscopic, H&E, and IF images, as well as cell-type annotation maps and tSNE projections, for both NMIBC and MIBC cases (Fig. 1D and fig. S1, G and H).

Transcriptomic comparison of neutrophil subsets revealed that TANs from NMIBC exhibited relatively higher expression of immune-activation and oxidative stress–associated genes compared to those from MIBC (Fig. 1E). In contrast, TANs in MIBC did not display a marked increase in classical immunosuppressive gene expression, but rather showed comparatively reduced activation signatures and a transcriptional profile consistent with a less metabolically active state (Fig. 1F). This relative attenuation of immune-activation programs in MIBC-associated TANs may contribute to impaired antitumor immune responses in advanced disease. To further validate these findings, we assessed neutrophil abundance by flow cytometry in clinical specimens. Tumor tissues showed a 6.45-fold increase in neutrophil infiltration compared with adjacent normal margins (Fig. 1, G and H), while murine bladder tumors recapitulated this pattern with a 10.15-fold higher neutrophil fraction than normal bladder (fig. S1, I and J). Crucially, functional assessment revealed that both human and mouse TANs generated less ROS than their peripheral counterparts (Fig. 1I), consistent with their reduced metabolic activity and immunosuppressive potential.

Taken together, these data indicate that the bladder tumor microenvironment induces a profound reprogramming of neutrophils toward a long-lived, metabolically quiescent, immunosuppressive phenotype. This effect is especially pronounced in advanced, muscle-invasive disease, highlighting a critical barrier to effective antitumor immunity. Targeting ROS-related metabolic pathways in TANs thus represents a promising avenue for therapeutic intervention aimed at reversing neutrophil-mediated immune suppression in cancer.

Engineering and validation of a localized CAP delivery system for intravesical immunotherapy

Building on our finding that TANs in bladder cancer adopt a metabolically quiescent, immunosuppressive phenotype characterized by diminished ROS production, we sought to develop a strategy to restore neutrophil cytotoxicity through targeted metabolic reprogramming within the tumor microenvironment. CAP offers a unique biophysical modality for this purpose. As a source of diverse reactive oxygen and nitrogen species, CAP can non-pharmacologically modulate cellular redox states with exceptional spatial and temporal precision. Unlike systemic ROS donors or conventional drugs, CAP enables the local generation of ROS/RNS directly at the tumor site, allowing for tunable and safe manipulation of the immune microenvironment. This rationale prompted us to engineer a CAP delivery platform optimized for intravesical (bladder-localized) application, with the goal of achieving in situ restoration of neutrophil cytotoxic capacity.

The schematic configuration of the CAP device system was illustrated in Fig. 2 (A and B). A typical discharge image of the CAP was shown in Fig. 2C, as it demonstrates a form of focused plasma jet, a dense collection of ionized and neutral particles mixture with reactive species inside, including ·OH, Ar*, N₂*. Figure 2D shows the characteristic voltage and current signals of the CAP device and they were observed as impulsive time-dependent wave-forms. The rising edge of the voltage wave-form exhibits a steeper slope than its falling edge, which is favorable to initiate the discharge process. The sharp current impulse corresponded to the electric ionization and plasma development process in the working gas. Also, optical emission spectroscopy was employed to identify excited species present in the plasma jet in the wavelength range from 200 to 900 nm. It can be clearly seen that the spectra are dominated by N₂ lines from 315 to 405 nm and Ar lines from 696 to 842 nm, with ·OH line also observed at 308 nm (Fig. 2E). The temperature of the CAP probe was kept below 30°C during a 15-min irradiation period (Fig. 2, F and G). Component analysis indicated that CAP was used as an external source to generate ROS. The concentration levels of reactive oxygen and nitrogen species increased in a time-dependent manner over 5 mins of CAP irradiation. The appearance of the diagnostic reagents changed from colorless to polychromatic due to the reaction, and the absorbance was increased upon CAP irradiation. The fluorescence intensity of HTA, used as indicative of ·OH radical generation, also increased upon exposure to CAP irradiation (Fig. 2H).

Intravesical instillation is an efficient drug delivery route for the local treatment of various urological conditions. Here, we propose an intravesical device equipped with inhalation anesthetic vaporizer, which is suitable for short-term local CAP treatment without a retrieval procedure (Fig. 2I). For in vivo experiments, ultra-fine polytetrafluoroethylene (PTFE) catheters were employed for transurethral delivery of externally-generated reactive species from plasma to the mouse bladder, relying on its excellent chemical resistance, high thermal stability, electrical insulation and biocompatibility. In brief, the short catheter was inserted into the urethra of the mouse to provide structural support. Subsequently, a longer and narrower catheter was advanced through the short catheter into the bladder for CAP delivery (Fig. 2J). The mice were placed in a supine position with the head tilted downward to ensure the intravesical catheter remained submerged below the liquid surface. T2-weighted magnetic resonance imaging (T2W-MRI) for mouse bladder was performed and the plasma generator was turned on after insertion. Thereby, observable plasma current flowing through the inner tube into the mouse bladder, and the annular gap between the dual tubes permitted venting of excess gas, ensuring bladder pressure homeostasis (Fig. 2K and movie S1).

We further developed a disk-shaped plasma array system to optimize experimental throughput in order to enhance the efficiency of animal experiments. The configuration features four independently controlled plasma generation units arranged in parallel, allowing synchronous processing of four mice with standardized treatment parameters (Fig. 2L). The CAP device also exhibited similar electrical outputs with what have been seen in Fig. 2D, for pulse waveform patterns both voltage and current measurements (Fig. 2M). Results of the real-time thermal monitoring probe showed that the temperature was maintained at physiologically safe levels (<30°C) (Fig. 2, N and O), throughout a 15-min treatment protocol.

Collectively, these engineering advances establish a robust and precisely controlled intravesical CAP delivery platform, enabling safe, reproducible, and localized generation of ROS/RNS within the bladder tumor microenvironment. This system provides a technological foundation for subsequent mechanistic investigations and therapeutic studies focused on restoration of neutrophil cytotoxic capacity and immune modulation in situ.

CAP induces oxidative activation and functional reprogramming of neutrophils toward an antitumor phenotype

Having established a robust intravesical CAP delivery system, we next sought to determine whether CAP could functionally reprogram TANs within the bladder tumor microenvironment (Fig. 3A). To directly assess the impact of CAP on neutrophil biology, we first analyzed ROS generation in both mouse and human neutrophils following CAP exposure. Flow cytometry revealed a marked increase in intracellular ROS levels in human neutrophils treated with CAP-activated media, compared to untreated controls (Fig. 3B). Similar results were observed in mouse neutrophils isolated from bladder cancer tissue, where CAP treatment significantly elevated ROS production (Fig. 3C). This effect was both dose- and time-dependent, as revealed by a progressive increase in ROS with extended CAP exposure (Fig. 3D). Confocal imaging using DCFH-DA staining further confirmed enhanced intracellular ROS, an effect that was abrogated by glutathione (GSH) supplementation, implicating redox-dependent mechanisms in neutrophil activation (Fig. 3E).

To comprehensively characterize the functional state of CAP-activated neutrophils, we performed a series of complementary assays. First, to rule out the possibility that CAP-induced activation was merely a consequence of cellular stress or impending apoptosis, we assessed neutrophil viability. Flow cytometric analysis of annexin V/PI staining confirmed that CAP treatment did not compromise neutrophil viability compared to untreated controls (fig. S2A). Next, we evaluated two key effector functions of activated neutrophils including degranulation and NETosis. Flow cytometry analysis of the degranulation marker CD63 revealed that CAP treatment significantly increased surface CD63 expression on neutrophils, indicative of enhanced degranulation potential (fig. S2B). As expected, the potent activator PMA induced a more pronounced degranulation response. Furthermore, immunofluorescence staining for neutrophil extracellular traps (NETs) demonstrated that CAP exposure promoted NETs formation (fig. S2C). While PMA served as a strong positive control, CAP-induced NETosis was evident, suggesting its role in modulating this specialized antimicrobial and potentially cytotoxic pathway. To further assess whether CAP also modulates immunosuppressive features of TANs, we examined the expression of representative immune suppressive markers after CAP treatment. Flow cytometric analysis showed that CAP significantly reduced PD-L1 expression on TANs, whereas TIM-3 expression was not significantly altered (fig. S2, D and E). In addition, when co-cultured with CD8⁺ T cells, CAP-treated TANs significantly increased the proportion of IFN-γ-positive CD8⁺ T cells compared with untreated TANs (fig. S2F). These results suggest that, in addition to enhancing cytotoxic effector functions, CAP treatment can partially alleviate the immune suppressive features of TANs and promote CD8⁺ T cell effector activity.

Having established CAP’s ability to induce global ROS (Fig. 3, B to E), we sought to define the specific subcellular sources of this oxidative burst. Using MitoSOX Green, a mitochondrial superoxide-specific probe, we found that CAP treatment specifically and significantly elevated mitochondrial ROS (mitoROS) in neutrophils (fig. S2G). Concurrently, measurement of extracellular H₂O₂ release using Amplex Red assay confirmed that CAP-activated neutrophils also produced substantially higher levels of extracellular ROS (fig. S2H). These data collectively demonstrate that CAP triggers a comprehensive oxidative response encompassing both intracellular (mitochondrial) and extracellular compartments.

To contextualize the potency of CAP as a neutrophil activator, we compared its ROS-inducing capacity against a panel of classic pharmacological and physiological stimuli. Flow cytometric analysis of intracellular ROS levels showed that CAP-induced ROS production exceeded that observed with H₂O₂, CXCL1, and IL-1β, and was comparable to that induced by LPS and PMA (fig. S2I). Finally, to define the kinetics of this activation, we performed a time-course analysis. CAP-induced ROS generation increased in a time-dependent manner, plateauing after approximately 10–15 minutes of exposure (fig. S2J), identifying the optimal window for neutrophil oxidative reprogramming.

To investigate the functional consequences of this oxidative activation, we evaluated the secretory and cytotoxic profiles of CAP-treated neutrophils. Cytokine array profiling demonstrated robust upregulation of pro-inflammatory and effector molecules, including CXCL2, IL1RA, LCN2, MPO, and MMP9, consistent with a transition toward an immunostimulatory phenotype (Fig. 3, F and G, and fig. S3A). We next asked whether CAP-induced neutrophil activation could translate into antitumor effects. In 3D tumor spheroid co-culture assays, CAP treatment reversed the tumor-promoting effect of TANs, as CAP-exposed neutrophils no longer promoted the growth of bladder cancer spheroids to the same extent as untreated TANs, an effect that was reversible with GSH supplementation (Fig. 3H and fig. S3B). These data indicate that CAP not only enhances ROS production but also restores the cytotoxic effector function of neutrophils against tumor cells.

Prompted by these findings, we performed transcriptomic profiling of tumor-derived TANs following CAP exposure. RNA-seq analysis revealed that CAP-treated TANs (TANs-CAP) displayed a distinct molecular signature relative to untreated TANs, as demonstrated by unsupervised hierarchical clustering and PCA (Fig. 3, I and J). TANs-CAP showed upregulation of oxidative stress response genes (Ncf1, Ncf2, Ncf4, Nos2, Acp5) and persistent downregulation of antioxidant genes (Prdx4, Prdx2, Sod1, Sod3, Gpx8, Glrx5), indicative of a shift toward an oxidative-activated state (Fig. 3K and table S2). GO and gene set enrichment analysis (GSEA) further highlighted significant enrichment of pathways related to reactive oxygen species metabolism and oxidative damage in the CAP-treated group (Fig. 3, L and M, and table S3).

To determine whether these antitumor effects could be attributed solely to CAP-induced reprogramming of neutrophils, rather than direct cytotoxicity against tumor or normal epithelial cells, we exposed bladder cancer cell lines (MB49, T24) and normal bladder epithelial cells (SV-HUC-1) to CAP-activated media in the absence of neutrophils. We observed no significant changes in cell viability, apoptosis, or proliferation between CAP-treated and control groups (fig. S3, C to H), indicating that CAP’s antitumor efficacy in this context is mediated primarily through functional reprogramming of neutrophils rather than direct tumor cell killing.

Building upon this finding, we next sought to delineate the specific functional consequences of CAP-mediated restoration of neutrophil cytotoxic capacity on tumor cells. Using a direct coculture assay, we assessed the ability of CAP-activated neutrophils to inhibit the growth of MB49 tumor cells. Crystal violet staining revealed that coculture with neutrophils significantly promoted MB49 cell growth compared to tumor cells alone (fig. S3, I and J). In contrast, CAP-reprogrammed neutrophils potently reversed this pro-tumoral effect and instead exerted a robust growth-inhibitory activity. This antitumor effect was comparable in magnitude to that induced by neutrophils maximally activated with PMA. To investigate the contribution of neutrophil NETs to this cytotoxicity, we performed coculture experiments in the presence of DNase I (to degrade NETs) and PAD4 inhibitor (to prevent NETs formation). Both interventions partially but significantly rescued tumor cell growth suppressed by CAP-activated neutrophils (fig. S3, I and J), indicating that NETosis constitutes an important, though not exclusive, effector mechanism in CAP-induced neutrophil cytotoxicity.

Furthermore, to define the kinetics of this functional reprogramming, we performed a time-course experiment in which neutrophils were pretreated with CAP for varying durations prior to coculture. The tumor-suppressive capacity of CAP-activated neutrophils increased in a time-dependent manner, with significant inhibition detectable after just 5 minutes of CAP exposure and reaching a plateau at 10–15 minutes (fig. S3, K and L). This kinetic profile closely mirrored the time-dependent induction of intracellular ROS, reinforcing the link between rapid oxidative activation and the acquisition of antitumor function. Collectively, these data demonstrate that CAP rapidly and efficiently reprograms neutrophils from a pro-tumoral to a potent antitumor state, employing NETosis as a key mechanism to directly mediate tumor cell cytotoxicity.

In aggregate, these findings demonstrate that CAP robustly induces an oxidative, immunostimulatory phenotype in neutrophils, restoring their tumoricidal potential and providing a drug-free strategy for reversing neutrophil-mediated immune suppression in bladder cancer.

CAP-mediated inhibition of neutrophil mitophagy restores mitochondrial function and drives immune activation

To elucidate the mechanistic basis of CAP-induced restoration of neutrophil cytotoxic capacity, we next investigated the neutrophil changing in response to CAP treatment. Given that mitochondrial metabolism is central to both ROS generation and immune cell activation, we hypothesized that CAP might modulate neutrophil function by targeting mitochondrial homeostasis. Functional metabolic profiling revealed that CAP-treated neutrophils exhibited a significant increase in both oxygen consumption rate (OCR) and ATP production compared to untreated TANs (Fig. 4A and fig. S4A), suggesting restoration of oxidative phosphorylation and a shift toward a more metabolically active state. Assessment of mitochondrial membrane potential (MMP, Δψm) showed that CAP exposure robustly preserved mitochondrial polarization, as demonstrated by increased red/green fluorescence ratios even in the presence of the depolarizing agent CCCP (Carbonyl cyanide 3-chlorophenylhydrazone) (Fig. 4B and fig. S4B).

Fig. 4. — (A) OCR of mouse TANs following indicated treatments. Cells were treated with control medium, CAP-activated medium, CCCP, or CCCP+CAP (n = 3). (B) Representative confocal images and statistical analysis of mitochondrial transmembrane potential (ΔΨm) which were labeled by JC-1 monomer (green fluorescent probe) and JC-1 aggregate (red fluorescent probe) in human TANs under different treatments (n = 3). Scale bar, 5 μm. (C) Sankey dot plot showing enriched autophagy-related pathways from RNA-seq data of control versus CAP-treated TANs. (D) Heatmap depicting expression changes of mitophagy-related DEGs in CAP-treated TANs compared to control TANs. Data are derived from bulk RNA-seq (n = 3). (E) Representative TEM images of mitochondria in mouse TANs treated with control medium or CAP-activated medium. Red dashed square frame indicate autophagic vacuoles containing mitochondrial remnants (mitophagosomes) or indicate intact mitochondria. Scale bar, 500 nm. (F and G) Evaluation of mitophagy in human TANs by confocal microscopy using the Mitophagy Detection Kit. Mitophagy is indicated by red fluorescence; lysosomes are labeled with green fluorescence. Scale bar, 2 μm. (F) Neutrophils were treated as indicated: control, CAP, CCCP, or CCCP+CAP. (G) Neutrophils were treated as indicated: control, cocultured with bladder cancer cells, CAP, or coculture+CAP. (H) Fluorescence semi-quantification of the mitophagy levels under different treatments (n = 3). (I) Western blot analysis of mitophagy-related proteins in mouse TANs under indicated conditions. Actin served as a loading control. (J) Schematic model summarizing the proposed mechanism of CAP-mediated neutrophil reprogramming. CAP exposure inhibits PINK1/Parkin-dependent mitophagy, leading to mitochondrial accumulation and enhanced ROS production. Created in BioRender. A, H. (2026) https://BioRender.com/dcf1ofc. Data are presented as means ± SD. **P < 0.01, ****P < 0.0001. Statistical values were calculated using one-way ANOVA [(B) and (H)] and two-way ANOVA (A). n indicates biological replicates.

To further interrogate the regulatory pathways involved, transcriptomic analysis revealed that autophagy and mitophagy related signaling pathways were significantly enriched in untreated TANs, but markedly suppressed following CAP exposure (Fig. 4C). Key mitophagy-associated genes, including Fundc1, Phb2, Atg101, Pink1, and Bnip3, were downregulated in CAP-treated neutrophils (Fig. 4D), implicating inhibition of mitophagic activity as a central mechanism. Corroborating these results, transmission electron microscopy (TEM) demonstrated that CAP-treated neutrophils harbored intact, well-preserved mitochondria, whereas untreated TANs contained abundant autophagic vesicles (Fig. 4E).

Direct visualization of mitochondrial quality control further supported this finding. Under basal conditions, mitophagy was rare in neutrophils, but strongly induced by CCCP. Notably, CAP treatment significantly suppressed CCCP-induced colocalization of mitochondrial autophagosomes and lysosomes (Fig. 4F and fig. S4C). Similarly, neutrophils co-cultured with tumor cells showed elevated mitophagy, which was effectively reversed by CAP intervention (Fig. 4, G and H, and fig. S4D). Western blotting and immunofluorescence confirmed that mitophagy related protein levels (LC3I/II, SQSTM1/p62, Parkin, PINK, ATG5) were increased upon tumor co-culture and downregulated following CAP exposure (Fig. 4I).

To further substantiate the specificity of CAP’s ability to inhibit mitophagy, we performed a series of comparative and mechanistic studies. First, we tested whether other potent neutrophil activators could mimic this effect. Flow cytometric analysis revealed that while CCCP robustly induced mitophagy, neither the inflammatory stimulus LPS nor the strong PKC activator PMA was able to suppress CCCP-induced mitophagy when administered in combination (fig. S4E). This result underscores the unique property of CAP to interfere with the mitochondrial quality control pathway, distinguishing it from classical inflammatory or activating signals. Next, we employed pharmacological inhibitors to dissect the upstream signaling pathways linking CAP to mitophagy inhibition. When neutrophils were cotreated with CCCP and CAP, the characteristic suppression of mitophagy was observed. Strikingly, this suppression was profoundly reversed by co-treatment with a HIF-1α inhibitor, but not by inhibitors targeting NF-κB or Nrf2. This finding identifies HIF-1α activation as a critical upstream molecular event in CAP-mediated mitophagy blockade (fig. S4F). Together, these data delineate a specific pathway wherein CAP, through a mechanism involving HIF-1α but not classical inflammatory transcription factors, disrupts mitochondrial turnover to restoration of neutrophil cytotoxic capacity.

Collectively, these findings demonstrate that CAP reprograms neutrophil metabolism and function through the inhibition of mitophagy, thereby restoring mitochondrial integrity and promoting an oxidative, immunostimulatory state. This mechanistic link between mitophagy suppression and immune activation highlights a previously unrecognized avenue for the functional restoration of TANs and the reversal of tumor-induced immunosuppression (Fig. 4J).

Localized CAP treatment suppresses bladder tumor growth by restoration of neutrophil cytotoxic capacity in vivo

To determine whether the reprogramming of TANs by CAP translates into meaningful antitumor efficacy in vivo, we employed an orthotopic bladder cancer mouse model and administered localized CAP treatment beginning five days post tumor implantation (Fig. 5A). CAP-treated mice exhibited significantly reduced tumor burden and prolonged survival compared to controls, as demonstrated by tumor weight measurements, survival analysis, H&E staining, and ultrasound imaging (Fig. 5, B to E). Immunofluorescence analyses further showed decreased Ki67 staining and increased TUNEL positivity in tumor tissues from CAP-treated animals, indicating effective inhibition of tumor cell proliferation and induction of apoptosis (Fig. 5F).

Fig. 5. — (A) Schematic illustration of the experimental protocol for orthotopic bladder cancer model and intravesical CAP treatment. C57BL/6 mice were inoculated with MB49 bladder cancer cells on day 0. Intravesical CAP treatment (15 min per session) was administered every 5 days starting from day 5 post-tumor inoculation. Created in BioRender. A, H. (2026) https://BioRender.com/dcf1ofc. (B) Representative images of bladder tumors (upper) and tumor weight quantification (bottom) (n = 5 mice for control; n = 6 mice for CAP). Scale bar, 1 cm. (C) Kaplan–Meier survival curves of tumor-bearing mice (n = 10 mice per group). (D) Representative ultrasound images of mouse bladders. Tumor masses are outlined by red dashed lines. Scale bar, 1 mm. (E) Representative H&E staining of bladder tumor sections. Scale bar, 500 μm. (F) Representative immunofluorescence images of bladder tumor sections stained for proliferation marker Ki67 (green) and apoptosis marker TUNEL (red). Scale bar, 20 μm and 70 μm. (G) Immunofluorescence analysis of MPO (green) and Ly6G (magenta) in mouse bladder tumor sections. Scale bar, 20 μm. (H) Quantification of cytokine levels in tumor tissue from tumor-bearing mice following indicated treatments (n = 6 mice per group). (I) Representative mIHC staining of bladder tumor sections. Sections were stained for Ly6G (green), CD8 (red), F4/80 (yellow), iNOS (cyan), and DAPI (blue). Scale bar, 50 μm. (J) Quantification of positively stained cells per high-power field for each marker (n = 4 mice per group). (K) t-SNE projection of scRNA-seq data from CD45⁺ immune cells isolated from MB49 tumors. (L) Quantification of proportional distribution of major immune cell subsets from scRNA-seq data shown in (K), expressed as percentage of total CD45⁺ cells. Data are presented as means ± SD. **P < 0.01, ****P < 0.0001. Statistical values were calculated using Student’s t-test [(B), (H), and (J)] and log-rank (Mantel–Cox) test (C).

To dissect the mechanism underlying CAP’s antitumor effect, we leveraged both scRNA-seq data and functional genetic approaches. Analysis of scRNA-seq data revealed that TANs in MIBC express higher levels of immunosuppressive genes than those in NMIBC, implicating these cells in tumor progression and therapy resistance (Fig. 1). To directly evaluate the requirement for neutrophils in CAP-mediated tumor suppression, we utilized Ly6G^DTR transgenic mice, which enable specific and sustained neutrophil depletion upon diphtheria toxin (DT) administration (fig. S5A). Neutrophil ablation alone significantly suppressed tumor growth, with a reduction rate of 40.13% compared to controls. This finding confirms the protumoral role of TANs in bladder cancer. In contrast, CAP monotherapy yielded an even greater tumor inhibition rate of 67.77%, suggesting that CAP not only neutralizes the pro-tumorigenic effects of TANs but also convert them into antitumor effectors. Strikingly, when CAP was administered to neutrophil-depleted Ly6G^DTR mice, tumor growth inhibition was limited to 35.47%, similar to neutrophil ablation alone, and markedly less than CAP monotherapy (fig. S5, B to E).

Having established the necessity of neutrophils for the full antitumor effect of CAP, we next sought to verify that the key functional phenotypes observed in our in vitro studies were recapitulated in TANs following in vivo CAP treatment. Flow cytometric analysis of TANs isolated directly from bladder tumors confirmed that intravesical CAP treatment significantly enhanced their intracellular ROS production (fig. S5F). This result aligns with our in vitro observations and demonstrates successful functional activation in situ. Consistent with the proposed mechanism, CAP-treated TANs also exhibited a significant reduction in mitophagy levels compared to TANs from control mice, as assessed by flow cytometry (fig. S5G). This suppression of mitochondrial turnover was further corroborated by TEM, which visually demonstrated a decrease in autophagic vacuoles containing mitochondria in CAP-treated TANs (fig. S5H). These data provide direct in vivo evidence that CAP reprograms intratumoral neutrophils by inhibiting mitophagy and elevating oxidative metabolism.

Beyond neutrophils, we investigated the impact of CAP on other key innate immune populations within the TME. Flow cytometric analysis showed that the overall frequency of tumor-infiltrating macrophages (F4/80⁺CD11b⁺) was not significantly altered following CAP treatment (fig. S5I). However, CAP treatment was associated with a shift in macrophage polarization. Specifically, the proportion of CD86⁺ macrophages was significantly increased, whereas the proportion of CD206⁺ macrophages was reduced (fig. S5, J and K), consistent with a transition toward a more pro-inflammatory macrophage phenotype. This repolarization toward a pro-inflammatory, antitumor phenotype indicates that CAP’s immunomodulatory effects extend beyond neutrophils to broadly reshape the innate immune landscape of the TME, fostering a more immunostimulatory milieu.

These findings demonstrate that CAP inhibits tumor growth in a neutrophil-dependent manner, primarily through functional restoration rather than changes in neutrophil abundance or trafficking. Supporting this conclusion, CAP-treated tumors displayed increased expression of MPO, a marker of neutrophil activation, by confocal imaging (Fig. 5G). ELISA analysis of tumor tissue cytokines revealed upregulation of proinflammatory mediators, including IL-12, TNF-α, and IL-6, in the CAP group relative to controls (Fig. 5H), consistent with broad immune activation.

To further delineate the impact of CAP on the tumor immune microenvironment, we performed multiplex immunohistochemistry (mIHC) and scRNA-seq profiling of immune cells in treated tumors. CAP-treated mice exhibited a marked increase in intratumoral CD8⁺ T cells, with marginal effect on other immune cells such as macrophages, neutrophils, and upregulation of iNOS (Fig. 5, I and J). scRNA-seq analysis was performed to delineate the transcriptome atlas of intratumoral immune cells (fig. S6A). Leukocytes (including B cells, NK cells, T cells, neutrophils and macrophages) were identified as the major cellular source of immune cells in the TME post CAP treatment compared to the vehicle group (Fig. 5, K and L, and fig. S6, B to D). Parallel neutrophil subclustering identified six functionally distinct populations defined by distinct molecular signatures: metabolic reprogram subset (Cstb, Esd, Syne1), cytotoxic subset (Pglyrp1, Retnlg, Mmp8), oxidative stress subset (Fos, Txnip), IFN responsive subset (Ifit1/2/3, Isg20, Herc6), immunosuppressive subset (Socs3, Clec7a) and immuno-activated subset (Saa3, Cd38, Icam1) (fig. S6, E and F). And the CAP treatment group exhibited a significant increase in the proportion of oxidative stress subset and immune-activated subset (fig. S6G). KEGG pathway analysis revealed that CAP treatment engaged multiple immune activation pathways within neutrophil populations (fig. S6H).

To establish a definitive causal link between neutrophil restoration and the observed antitumor effects, we performed adoptive transfer experiments. Neutrophils isolated from donor mice were pretreated ex vivo with CAP or control and then transferred into tumor-bearing recipient mice. Strikingly, while the transfer of control neutrophils had no significant effect, the adoptive transfer of CAP-reprogrammed neutrophils was sufficient to significantly inhibit tumor growth, recapitulating a substantial portion of the therapeutic effect seen with direct intravesical CAP treatment (fig. S7A). This experiment provides direct evidence that CAP’s antitumor efficacy can be mediated by systemically altering the functional state of neutrophils. To dissect the relative contribution of the innate and adaptive immune systems, we evaluated CAP’s efficacy in Rag1^−/− mice, which lack mature T cells. In immunocompetent wild-type mice, CAP monotherapy induced robust tumor suppression (fig. S7B). In contrast, the therapeutic effect of CAP was significantly attenuated in Rag1^−/− mice, although a residual, statistically significant inhibition was still observed. This result demonstrates that while the adaptive immune response, particularly T cells, plays a major role in mediating the full therapeutic potential of CAP, a substantial component of its antitumor activity is mediated by the innate immune system, consistent with our findings on restoration of neutrophil cytotoxic capacity.

Clinically, therapeutic interventions are often initiated after tumor establishment. We therefore tested the efficacy of CAP when administered at more advanced disease stages. CAP treatment initiated either at day 5 or day 10 post-tumor inoculation resulted in significant tumor growth inhibition compared to untreated controls (fig. S7C). This confirms that CAP is not merely a prophylactic agent but possesses genuine therapeutic activity against established tumors. Finally, to optimize the treatment protocol, we performed a dose-response study by varying the duration of each intravesical CAP application. Tumor growth inhibition increased with treatment duration from 5 to 15 minutes per session (fig. S7D). Extending the treatment to 25 minutes per session did not yield further significant benefit, defining 15 minutes as the optimal treatment window that balances maximal efficacy with practicality. Collectively, these translational studies solidify the therapeutic relevance of CAP by confirming its activity in established disease, defining its optimal dosing, and mechanistically delineating the contributions of both innate and adaptive immune components.

Together, these results establish that localized CAP therapy elicits robust antitumor effects in vivo by functionally reprogramming TANs, triggering broad activation of both innate and adaptive immunity, and reshaping the immune landscape of the tumor microenvironment. This mechanism highlights CAP as a promising modality for overcoming neutrophil-driven immune resistance in solid tumors.

CAP synergizes with anti-PD-1 therapy by remodeling neutrophil-mediated immunosuppressive niches

Given the central role of neutrophils in orchestrating immunosuppressive niches within the bladder tumor microenvironment, and based on our findings that CAP reprograms TANs toward an antitumor state, we sought to determine whether this strategy could enhance the efficacy of immune checkpoint blockade therapy. Although anti-PD-1 antibodies have transformed the management of bladder cancer, most patients derive limited clinical benefit, which is frequently attributed to insufficient CD8⁺ T cell infiltration and persistent myeloid-mediated immune suppression (35–40). Our earlier data indicated that TANs are major contributors to these obstacles, as they exclude cytotoxic T cells and express high levels of PD-L1, thereby suppressing antitumor immunity (34).

Spatial immunofluorescence and phenotypic analyses demonstrated that PD-L1⁺ TANs are frequently located near PD-1⁺ CD8⁺ T cells within tumor tissues, outnumbering other PD-L1-expressing immune populations and suggesting a direct inhibitory interaction (Fig. 6A). These findings provided a rationale to test whether functional reactivation of neutrophils by CAP could disrupt this local immunosuppressive niche and sensitize tumors to PD-1 blockade.

Fig. 6. — (A) Left: Quantification of the distance from individual CD4⁺T cell, CD8⁺T cell, Tregs to the nearest TAN in human bladder cancer tissue. Right: Representative image showing cell type distribution of TAN (PD-L1⁻/PD-L1⁺) and T cells (PD1⁻/PD1⁺). (B) Schematic illustration of the experimental protocol for combination therapy. Intravesical CAP treatment was administered every 5 days. Anti-PD-1 antibody was administered intraperitoneally every 3 days. Created in BioRender. A, H. (2026) https://BioRender.com/dcf1ofc. (C) Representative images of dissected bladders from mice. Scale bar, 1 cm. (D) Quantification of bladder tumor weights at day 22 post-inoculation (n = 6 mice per group). (E) Representative bioluminescence images of MB49-luci tumor-bearing mice. (F) Quantification of total flux over time (n = 5 mice per group). (G) Kaplan–Meier survival curves of orthotopic MB49 tumor-bearing mice (n = 10 mice per group). (H) Representative ultrasound images (upper) and H&E staining (lower) of bladder tumors from each group. Tumor masses are outlined by red dashed lines. Scale bar for H&E, 500 μm; Scale bar for ultrasound, 1 mm. (I) Representative immunofluorescence images of bladder tumor sections stained for Ki67 (green) and TUNEL (red). Scale bar, 20 μm and 70 μm. (J) Representative mIHC staining and quantification of Ly6G⁺ cells, CD8⁺ T cells, F4/80⁺ cells and iNOS⁺ cells of tumor sections from tumor-bearing mice. Scale bar, 50 μm. (K) Representative plots and histogram showing the proportion of CD3⁺CD8⁺ T cells in mouse bladder tumors (n = 4 mice per group). (L) Representative plots and histogram showing the proportion of granzyme B⁺ cells in CD8⁺ T cells in mouse bladder tumors (n = 4 mice per group). Means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Statistical values were calculated using one-way ANOVA [(D), (K), and (L)], log-rank (Mantel–Cox) test (G), and two-way ANOVA (F).

We next evaluated this therapeutic combination in the orthotopic MB49 bladder cancer model. Mice receiving both CAP and anti-PD-1 antibody treatment exhibited the most significant reduction in tumor growth and the greatest improvement in survival compared to either treatment alone (Fig. 6, B to G). The antitumor effects of the combination were confirmed by histological analysis, ultrasound imaging, decreased Ki67 staining, and increased TUNEL-positive apoptotic cells within tumor tissues (Fig. 6, H to J). Flow cytometry and immunofluorescence further revealed a robust increase in intratumoral effector CD8⁺ T cells, particularly those expressing granzyme B, supporting enhanced cytotoxic T cell responses (Fig. 6, K and L). Beyond CD8⁺ T cells, we assessed the effect of combination therapy on other innate immune populations within the tumor microenvironment. Flow cytometric analysis showed that the overall frequency of tumor-associated neutrophils (Ly6G⁺CD11b⁺) exhibited an increased frequency following CAP treatment, with a more pronounced effect observed in the combination group (fig. S7E). In contrast, the overall frequency of tumor-infiltrating macrophages (F4/80⁺CD11b⁺) was not significantly altered in either the CAP monotherapy or combination therapy groups (fig. S7F). Similarly, the frequency of natural killer cells (NK1.1⁺CD45⁺) did not differ significantly across treatment conditions (fig. S7G). These results indicate that the enhanced antitumor efficacy of CAP combined with anti–PD-1 therapy is accompanied by increased neutrophil presence and augmented CD8⁺ T cell infiltration, while macrophage and NK cell abundance remain largely unchanged.

To validate these findings in a clinically relevant setting, we established a humanized patient-derived xenograft (PDX) model of bladder cancer in NOG-EXL mice reconstituted with human hematopoietic stem cells and myeloid-supporting cytokines (Fig. 7A). CAP treatment alone was sufficient to induce substantial tumor growth inhibition in this model, demonstrating its capacity to modulate human immune cell-tumor interactions (Fig. 7, B and C). Importantly, combined CAP and anti-PD-1 therapy was well tolerated, as indicated by stable body weight, preserved organ histology, and normal hematologic parameters (fig. S8, A to D).

Fig. 7. — (A) Schematic illustration of the experimental workflow for establishing humanized hu-HSC-NOG mice and evaluating CAP therapy. NOG-EXL mice were transplanted with human hematopoietic stem cells (hu-HSCs). After 7 weeks, successful human immune cell reconstitution was confirmed by flow cytometric detection of human CD45⁺ cells in peripheral blood. Humanized mice were then orthotopically inoculated with patient-derived tumor. Starting from day 7 post-tumor inoculation, mice received intravesical CAP treatment (15 min per session, every 5 days). Mice were euthanized on day 22 for tumor analysis. (B) Representative ultrasound images of mouse bladders from each treatment group at day 3, 7, 14, 21. Tumor masses are outlined by red dashed lines. Scale bar, 1 mm. (C) Antitumor efficacy of CAP therapy in humanized hu-HSC-NOG mice. Left: Representative images of dissected bladders from mice treated with control or CAP. Scale bar, 1 cm. Right: Quantification of bladder tumor weights at day 22 post-inoculation (n = 4 mice for control group; n = 5 mice for CAP). (D) Schematic illustration summarizing the proposed mechanism of CAP-mediated neutrophil reprogramming and its synergy with anti-PD-1 checkpoint blockade. Local intravesical CAP delivery induces oxidative stress in TANs, leading to inhibition of PINK1/Parkin-mediated mitophagy. This results in mitochondrial accumulation, sustained oxidative metabolism, enhanced ROS production, thereby reprogramming TANs from an immunosuppressive to an immunostimulatory phenotype. Reprogrammed neutrophils promote CD8⁺ T cell recruitment and activation, remodel the tumor immune microenvironment, and sensitize tumors to anti-PD-1 therapy. Created with BioRender.com. Means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Statistical values were calculated using one-way ANOVA (C). Schematic illustration in (A) and (D) were created in BioRender. A, H. (2026) https://BioRender.com/dcf1ofc.

Taken together, these results demonstrate that CAP can enhance the antitumor efficacy of immune checkpoint blockade by remodeling the neutrophil-mediated immunosuppressive niche and promoting effective CD8⁺ T cell-mediated tumor clearance. This work provides a compelling rationale for the clinical evaluation of CAP-based combination immunotherapy strategies in patients with bladder cancer.

DISCUSSION

The tumor immune microenvironment remains a formidable barrier to effective immunotherapy, particularly in solid tumors characterized by dense myeloid infiltration (3, 4). Among these, neutrophils have emerged as highly plastic yet underappreciated regulators of tumor progression (5, 6). Our study provides evidence that CAP, a non-pharmacological and locoregionally deliverable modality, can functionally reprogram TANs and thereby reinvigorate antitumor immunity. By integrating in vitro and in vivo therapeutic studies across both syngeneic and humanized tumor models, we demonstrate that CAP acts through a mitochondria-centered mechanism to shift neutrophils from an immunosuppressive state toward cytotoxic and immunostimulatory phenotype.

The heterogeneity of neutrophils in cancer has posed a significant challenge in understanding their dualistic roles in tumor immunity (3, 4, 6). Our spatial transcriptomic and single-cell analyses revealed distinct phenotypic profiles between neutrophils in NMIBC versus MIBC. Neutrophils in MIBC exhibited a transcriptional signature consistent with immunosuppression and metabolic quiescence. These findings underscore the functional diversity of TANs and reveal their potential contribution to immunotherapy resistance, particularly in tumors marked by T cell exclusion and high neutrophil burden.

Previous efforts to modulate neutrophil function in cancer have largely focused on systemic depletion using CXCR2 antagonists or anti-Ly6G antibodies, or on blockade of neutrophil recruitment through chemokine targeting (41, 42). Notably, in contrast to depletion-based or recruitment-blocking strategies that aim to eliminate neutrophils systemically, CAP strategy emphasizes functional re-education of tumor-associated neutrophils in situ, preserving their host-protective capacity while redirecting their activity toward tumor control. More recent strategies have attempted to polarize neutrophils toward antitumor states using pharmacologic agents such as IFN-β, TGF-β inhibitors, or nanocarrier-based cytokine delivery (10–12). However, such approaches are limited by transient effects, systemic toxicity, and poor spatiotemporal control (43). In contrast, our study establishes a drug-free, locoregionally applicable strategy that leverages physical modulation of immune cell metabolism to reprogram tumor-associated neutrophils in situ. By exploiting the unique redox sensitivity of neutrophils, CAP circumvents the need for systemic manipulation and offers a tunable, clinically compatible platform for overcoming myeloid-driven immunosuppression. Several lines of evidence indicate that CAP differs from classical pharmacological neutrophil stimuli in both mode and duration of activation, supporting its potential as a biophysical approach to modulate neutrophil function within the tumor microenvironment.

In vivo, localized CAP treatment markedly reduced bladder tumor burden and prolonged survival in orthotopic models, with minimal systemic toxicity. The antitumor effect was dependent on the presence of neutrophils, as genetic ablation of these cells abrogated CAP efficacy. Interestingly, while CAP did not significantly alter neutrophil abundance in the tumor, it profoundly reshaped their activation state, consistent with a functional reprogramming rather than trafficking-based mechanism. This distinction is crucial, as most previous attempts to target TANs have relied on systemic depletion or trafficking inhibition, approaches that lack precision and carry risk of immune compromise.

The translational potential of this approach is supported by our studies in a humanized patient-derived xenograft (PDX) model, which demonstrated robust tumor inhibition and enhanced CD8⁺ T cell function without notable systemic toxicity. The intravesical delivery method employed here mimics clinical administration routes in bladder cancer, providing a feasible path toward first-in-human studies. Given the broad presence of neutrophil-rich, immune-cold phenotypes in many solid tumors, including lung, pancreatic, and colorectal cancers, CAP may offer a generalized strategy for myeloid reprogramming across multiple malignancies.

Nevertheless, this study has several limitations. While we demonstrate clear evidence of mitophagy inhibition and oxidative activation in neutrophils following CAP exposure, the upstream molecular sensors and signaling pathways mediating this response warrant further investigation. Moreover, although our murine and humanized models recapitulate key aspects of the human immune-tumor interface, clinical validation in patients is required to fully assess safety, durability, and efficacy. Future studies should also explore the kinetics of CAP-induced reprogramming, potential effects on other stromal populations, and combinatorial regimens beyond PD-1 blockade. Although our data support CAP-mediated restoration of TANs cytotoxic activity and partial alleviation of T cell-suppressive features, the current study cannot yet fully determine the extent to which these effects reflect direct immunosuppression reprogramming of TANs in vivo. Further studies will be required to better define the underlying cellular and molecular mechanisms.

In conclusion, our findings define a drug-free strategy to reprogram tumor-associated neutrophils through physical modulation of mitochondrial dynamics. By shifting neutrophils toward a cytotoxic, immunostimulatory state, CAP restores both innate and adaptive immune responses and enhances the efficacy of immune checkpoint therapy. This work provides a mechanistic framework and preclinical rationale for harnessing biophysical interventions to overcome myeloid-driven immune resistance in cancer (Fig. 7D).

MATERIALS AND METHODS

Mice

Female C57BL/6 mice (6–8 weeks old) were purchased from Charles River Laboratories (RRID: MGI:2159769). Ly6G-DTR transgenic mice (C57BL/6 background) and Rag1^−/− mice (C57BL/6 background) were obtained from Cyagen Biosciences. Humanized hu-HSC-NOG-EXL mice (10 weeks after hematopoietic stem cell engraftment) were purchased from Charles River Laboratories. All animals were housed under specific pathogen-free conditions with controlled temperature and humidity on a 12-hour light/dark cycle. All animal procedures were conducted in accordance with the Guidelines for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of Qilu Hospital of Shandong University (Approval No. KYLL-202402-016-1).

Orthotopic bladder tumors were established in 8–10-week-old female immunocompetent mice by intravesical instillation of 5 × 10⁵ bladder cancer cells under isoflurane anesthesia using aseptic technique. Neutrophil-depleted conditions were generated in parallel experiments using Ly6G-DTR mice treated with diphtheria toxin (DT; Calbiochem, #322326) at 20 μg/kg via intraperitoneal injection every 72 hours to maintain sustained neutrophil ablation. Tumor growth was monitored longitudinally by bioluminescence imaging following intraperitoneal administration of D-luciferin (150 mg/kg) under isoflurane anesthesia using an IVIS Lumina system (PerkinElmer). Tumor burden was further assessed by high-resolution transabdominal ultrasonography (Vevo LAZR, VisualSonics) following bladder distension with PBS via catheterization.

For therapeutic studies, mice were randomly assigned to treatment groups using a random number generator. Transvesical CAP treatment and/or intraperitoneal anti–PD-1 antibody (10 mg/kg) were administered on alternating three-day schedules as indicated. Body weight was recorded throughout the study, and bladder tumors were harvested at experimental endpoints for histological and molecular analyses. Sample sizes (5–10 mice per group) were determined based on prior experience with orthotopic bladder tumor models and published literature; no formal power calculation was performed. No predefined exclusion criteria were applied. No animals were excluded from analysis unless technical failure occurred. Investigators were not blinded during treatment administration; however, tumor measurements and histological analyses were performed in a blinded manner. Mice were monitored daily, and humane endpoints were applied if tumor burden exceeded ethical limits or animals exhibited signs of distress.

Patients

Bladder cancer patients undergoing radical cystectomy or transurethral resection at Qilu Hospital of Shandong University were prospectively enrolled in this study. Whole blood samples and tumor tissues were collected for neutrophil isolation. The study cohort included adult patients (age range 57–86 years; 12 males and 3 females). Written informed consent was obtained from all participants prior to sample collection. The study was approved by the Institutional Ethics Committee of Qilu Hospital of Shandong University (Approval Nos. KYLL-202409-052-1 and KYLL-202402-016-1) and conducted in accordance with the Declaration of Helsinki. Information on ethnicity was not systematically collected.

Cell lines

The mouse bladder carcinoma cell line MB49 (C57BL/6 origin; RRID: CVCL_7076) was obtained from Millipore. The human bladder carcinoma cell line T24 (ATCC Cat# HTB-4; RRID: CVCL_0554) and the human uroepithelial cell line SV-HUC-1 (ATCC Cat# CRL-9520; RRID: CVCL_3798) were purchased from ATCC. MB49 cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and 1% (v/v) penicillin–streptomycin. T24 and SV-HUC-1 cells were cultured according to ATCC recommendations in DMEM supplemented with 10% FBS and 1% penicillin–streptomycin. All cell lines were maintained at 37°C in a humidified incubator with 5% CO₂. All cell lines were regularly tested for mycoplasma contamination and were confirmed to be negative. Cells were used at low passage numbers after thawing and were not cultured beyond 20 passages.

Xenium 5 k Experiment and Analyzer

Sample preparation

Tissues were prepared following the Demonstrated Protocols Xenium In Situ for FFPE-Tissue Preparation Guide (CG000578) and Xenium In Situ for FFPE Tissues-Deparaffinization & Decrosslinking (CG000580). Probe hybridization, washing, ligation, amplification, and cell segmentation staining were performed following a development version of the Xenium Prime In Situ Gene Expression with optional Cell Segmentation Staining User Guide (CG000760). Post-instrument processing followed the Demonstrated Protocol Xenium In Situ Gene Expression-Post-Xenium Analyzer H&E Staining (CG000613).

Gene panels

The Xenium Prime 5 K Human Pan Tissue and Pathways Panel was designed to enable comprehensive cell type and cell state identification using publicly available single cell RNA sequencing data. The panel also covers canonical signaling pathways, as well as genes relevant to developmental biology, immuno-oncology, and genes that are well known in biomedical literature.

Xenium Analyzer

The instrument run was performed following the Xenium Analyzer User Guide CG000584. The on-instrument analysis was run with a development version of Xenium Onboard Analysis v3.0.0.

Clustering and analysis

The resulting matrix were further processed by Seurat followed by SCTransform, scaling, feature genes selection, PCA dimension reduction, and clustering.

CAP equipment and treatment

The plasma jet device used in this study consists of two coaxial quartz glass tubes and two metal electrodes, enabling a double-layer dielectric barrier discharge (DBD) configuration. A high-voltage rod-shaped electrode (diameter: 3 mm) is enclosed by the inner quartz tube (outer diameter: 4 mm, inner diameter: 3 mm). The outer quartz tube is fitted with a nozzle at its lower end (outer diameter: 6 mm, inner diameter: 5 mm). A copper coil, wrapped around the exterior of the nozzle, serves as the ground electrode. The distance between the two electrodes is 6 mm, with the ground electrode positioned 6 mm from the nozzle outlet. To prevent direct contact with the working gas and avoid arcing with the high-voltage electrode, the ground electrode is enclosed within a plastic protective cover. Cold atmospheric plasma (CAP) was generated using a high-voltage power supply delivering 15 kV at a frequency of 5 kHz. Argon (99.99%) was used as the working gas at a flow rate of 72 liter hour⁻¹.

For in vitro assay, culture medium was exposed directly to CAP for indicated time, then cells were treated with CAP-activated medium for further analysis. Considering that CAP treatment exhibited the biological effects through reactive species, this study used a constant discharge power and lengthened the treatment time to increase the content of reactive species in different CAP treatment groups. To better describe the dose of CAP, the preset time was used to visually define the actual treatment dose of the CAP treatment (table S1).

For in vivo experiments, a custom-built CAP sprayer was used to treat tumor-bearing mice. This sprayer shares a similar double-layer DBD structure, comprising two coaxial quartz tubes and two metal electrodes. A high-voltage rod electrode (diameter: 0.75 mm) is positioned inside the inner quartz tube (outer diameter: 2 mm, inner diameter: 1 mm), while the outer tube (outer diameter: 4 mm, inner diameter: 3 mm) is equipped with a nozzle (outer diameter: 3 mm, inner diameter: 2 mm) at its bottom end. Plasma is generated through a flexible PTFE catheter (length: 30 mm, inner diameter: 0.6 mm, outer diameter: 0.9 mm). The ground electrode, a copper coil, is wound around the exterior of the nozzle, and the distance between the high-voltage electrode tip and the ground electrode is 6 mm. Operational parameters included an argon gas flow rate of 18 liter hour⁻¹ per tube and a discharge voltage of 15 kV at 1 kHz. Tumor-bearing mice were sequentially treated with the plasma jet, with each session lasting 15 minutes. During the experiment, the plasma sprayer system could be arranged in an array configuration for broader treatment coverage (table S1).

ROS/RNS measurement

The hydrogen peroxide assay kit (Beyotime Biotechnology, S0038) was used to detect H₂O₂ in the solution, according to the manufacturer’s instructions. Briefly, 100 μl of detection reagent was added to 50 μl of samples and standards, followed by a 30 min incubation at room temperature. The absorbance was measured using a microplate reader (Tecan Infinite E plex) with wavelength settings of 560 nm for excitation. Nitrite concentration quantification was performed using the Griess reagent system (Beyotime Biotechnology, S0021S) according to the manufacturer’s protocol, wherein 50 μl aliquots of both Griess Reagent I and II were mixed with 50 μl test samples and incubated for 10 minutes before measuring absorbance at 540 nm using a Tecan Infinite E plex microplate reader. Hydroxyl radical generation following cold atmospheric plasma (CAP) exposure was assessed through the terephthalic acid (TA) fluorescence assay (Sigma), where 2 mM TA solution (pH = 10.16) was irradiated with CAP and the subsequent formation of fluorescent 2-hydroxyterephthalic acid (HTA) was quantified by fluorescence spectroscopy with 310 nm excitation and emission scanning from 350–550 nm.

Cell Isolation and in vitro assay

Neutrophil isolation was performed using magnetic-activated cell sorting (MACS) technology with species-specific markers, where mouse neutrophils were positively selected using Ly6G Microbeads (Miltenyi Biotec, 130–097-658) while human neutrophils were enriched via CD16 Microbeads (Miltenyi Biotec, 130–045-701), both achieving ≥90% purity as confirmed by flow cytometric analysis, with subsequent culture of the isolated cells in complete RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) under standard conditions. Both murine and human neutrophils were resuspended in Hank’s balanced saline solution (HBSS) containing 2% of bovine serum albumin (BSA) and calcium. Cell suspensions contained >95% of viable neutrophils, as established by flow cytometry. For coculture studies, MACS-sorted neutrophils were incubated with bladder cancer cells, with post-treatment fixation in 4% paraformaldehyde (PFA) and crystal violet (0.5%) staining for microscopic imaging analysis, while proliferation assays involved 96-hour incubation with CAP-activated medium followed by sulforhodamine B (SRB) colorimetric assessment performed per standard protocols.

Mouse neutrophils were isolated from female C57BL/6 mice (6–8 weeks old).

Human neutrophils were isolated from tumor specimens or peripheral blood of consented patients.

For pharmacological inhibitor experiments, mouse neutrophils were pre-incubated for 1 hour prior to CAP treatment with the following inhibitors: Nrf2 inhibitor ML385 (5 μM, Selleckchem, S8790), HIF-1α inhibitor KC7F2 (20 μM, Selleckchem, S7946) or NF-κB inhibitor BAY11–7082 (5 μM, Selleckchem, S7352). Following inhibitor pre-incubation, cells were treated with CAP-activated medium. Control groups received an equivalent volume of vehicle (DMSO, final concentration < 0.1%).

Confocal microscopy

For immunofluorescence staining of human TANs, neutrophils were sorted from dissociated human bladder cancer tissues by MACS, incubated onto poly-L-lysine-coated glass slides, and fixed with 4% paraformaldehyde. Cells were permeabilized with 0.1% Triton X-100, blocked with 5% bovine serum albumin, and incubated with anti-human CD66b antibody (Abcam, ab300122) overnight at 4°C, followed by Alexa Fluor 488-conjugated secondary antibody. Nuclei were counterstained with DAPI (Beyotime Biotechnology, C1006). For mouse TANs, Ly6G⁺ cells were sorted from dissociated mouse bladder tumors, processed identically, and stained with anti-mouse Ly6G antibody (Abcam, ab238132) followed by Alexa Fluor 488-conjugated secondary antibody. All slides were mounted and images were acquired using a confocal microscope (Andor Dragonfly 200) and analyzed with ImageJ software.

For mitophagy and mitochondrial function assays, neutrophils or sorted TANs were seeded onto poly-L-lysine-coated coverslips and treated differently. Mitophagy was detected using the Mitophagy Detection Kit (DOJINDO, MD01), and mitochondrial membrane potential was assessed using the JC-1 assay kit (Beyotime Biotechnology, C2003S), following the manufacturers’ protocols. After staining, coverslips were washed, mounted with DAPI-containing mounting medium, and imaged by confocal microscopy. Image analysis was performed using ImageJ.

Histology and immunofluorescence analysis

Standard histological evaluation was performed through hematoxylin and eosin (H&E) staining of 4% paraformaldehyde-fixed, paraffin-embedded tumor sections which were subsequently imaged using a VS120 Olympus microscope equipped with OlyVIA software. While immunofluorescence analysis involved processing of 10-μm cryosections through room temperature dehydration and PBS rehydration followed by 1-hour blocking, overnight incubation with primary antibodies at 4°C, 1-hour room temperature incubation with AlexaFluor-conjugated secondary antibodies, DAPI nuclear counterstaining, and final visualization and quantification using an Andor Dragonfly 200 confocal imaging system.

Western blot

Cells were collected and lysed with RIPA buffer supplemented with PMSF (Beyotime Biotechnology). The protein concentration was determined with BCA Protein Assay Reagent Kit. Total protein of each sample was separated by SDS-PAGE and transferred onto PVDF membranes. Sequentially, the membranes were blocked with 5% (w/v) defatted milk in Tris-buffered saline with 0.05% Tween (TBS-T) for 1 h at room temperature followed by incubation with primary antibodies overnight at 4°C with shaking. Membranes were rinsed in TBS-T and secondary horseradish peroxidase-conjugated antibodies were applied for 1 h at room temperature with shaking. Membranes were rinsed with TBS-T, and bands were visualized using enhanced chemiluminescence reagent (Bio-Rad) and a chemiluminescence apparatus. The following primary antibodies were used for immunoblotting and immunofluorescence analyses: LC3A/B (D3U4C) XP rabbit monoclonal antibody (Cell Signaling Technology, Cat# 12741, RRID: AB_2617131), SQSTM1/p62 (D5E2) rabbit monoclonal antibody (Cell Signaling Technology, Cat# 8025, RRID: AB_10859911), Parkin (Prk8) mouse monoclonal antibody (Cell Signaling Technology, Cat# 4211, RRID: AB_2159920), PINK1 (D8G3) rabbit monoclonal antibody (Cell Signaling Technology, Cat# 6946, RRID: AB_11179069), Atg5 (D5F5U) rabbit monoclonal antibody (Cell Signaling Technology, Cat# 12994, RRID: AB_2630393), β-Actin (13E5) rabbit monoclonal antibody (Cell Signaling Technology, Cat# 4970, RRID: AB_2223172), Myeloperoxidase (MPO; E1E7I) XP rabbit monoclonal antibody (Cell Signaling Technology, Cat# 14569, RRID: AB_2798516).

Intracellular ROS detection

Intracellular ROS were measured with the fluorescent probe DCFH-DA (Beyotime Biotechnology, S0033S). After CAP treatment, neutrophils were incubated with fluorescent probes (10 μM) for 30 min in the cell incubator. Finally, the cells were analyzed with a flow cytometry and the data analysis was performed with Flowjo analysis software.

Mouse XL cytokine array

Neutrophil culture supernatant was obtained as described above, and 111 soluble proteins were quantified using a Mouse XL Cytokine Array Kit (R&D Systems). Protein detection was visualized using a Chemi Reagent Mix (R&D Systems) and a chemiluminescence apparatus. Pixel density was quantified by densiometric analysis using the Image J Software.

Flow cytometry

For the analysis of tumor-infiltrating immune cells, briefly, tumors were dissected from mice, weighed, and digested into single-cell suspension using Type IV Collagenase and DNase. For the neutrophil panel, cells were incubated with CD11b and Ly6G antibodies in the dark for 30 minutes at 4°C and quantified using flow cytometry. For the T cell panel, cells were incubated with CD3 and CD8 antibodies. Only for the T cell function panel, the cells were resuspended in permeabilization solution and stained intracellularly with Granzyme B antibodies in the dark for 1 hour at 4°C. Flow cytometry analysis was performed with a BeckmanCoulter CytoFLEX, and the results were analyzed using CytExpert and FlowJo software. After activating with CAP at 37°C and 5% CO₂, neutrophils were washed in PBS and incubated with Mitophagy Dye (DOJINDO, MD01) or JC-1 dye (Beyotime Biotechnology, C2003S) for 30 min individually. Samples were washed twice in PBS and immediately analyzed by flow cytometry. The fluorescence of samples was measured in a flow cytometer and analyzed using FlowJo software. For flow cytometry, the following fluorochrome-conjugated antibodies were used: APC anti-human CD45 antibody (BioLegend, Cat# 304012, RRID: AB_314400), FITC anti-human CD66b antibody (BioLegend, Cat# 305104, RRID: AB_314496), FITC anti-mouse CD3 antibody (BioLegend, Cat# 100203, RRID: AB_312660), APC anti-mouse CD8a antibody (BioLegend, Cat# 100712, RRID: AB_312751), PerCP/Cy5.5 anti-human/mouse Granzyme B recombinant antibody (BioLegend, Cat# 372212, RRID: AB_2728379), APC anti-mouse/human CD11b antibody (BioLegend, Cat# 101212, RRID: AB_312795), PE anti-mouse Ly-6G antibody (BioLegend, Cat# 127607, RRID: AB_1186104).

Three-dimensional spheroid culture

Three-dimensional spheroid formation was achieved using an optimized hanging drop technique, wherein 15–30 μl aliquots of single-cell suspension containing 0.24% methylcellulose were carefully pipetted onto culture dish lids and inverted over PBS-filled base dishes to maintain humidity, followed by incubation under standard culture conditions (37°C, 5% CO₂) for 5 days with daily monitoring until aggregate formation; mature spheroids were then gently transferred to 96-well plates for confocal microscopic imaging at 10× magnification. While for neutrophil-bladder cancer cell coculture experiments, neutrophils were prelabeled with CFDA-SE [5(6)-carboxyfluorescein diacetate succinimidyl ester] fluorescent dye prior to spheroid formation to enable subsequent tracking and visualization.

ELISA

ELISA kits were used to assay the levels of mouse IL-12 (Cat: M3062; Elabscience), mouse TNF-α (Cat: M3063; Elabscience), and mouse IL-6 (Cat: M0044; Elabscience) in tumor tissue according to the manufacturer’s instructions. Briefly, the standard solution and samples were added to the plate and incubated for 2 hours at room temperature. Then, the wells were aspirated and washed 5 times with wash buffer, and the conjugate was added to each well followed by incubation for 2 hours. After five washes, the substrate solution and stop solution were added successively with a thorough mixture. The optical density of each well was determined using a microplate reader.

Analysis of OCR with the seahorse XF platform

To evaluate the oxygen consumption rate (OCR), neutrophils in Seahorse XF RPMI Medium without glucose (Agilent Technologies) were seeded in a 96-well XF Cell Culture Microplate (Agilent Technologies) pre-coated with poly-L-lysine. Cells were left to adhere for one hour at 37°C without CO₂. Neutrophils were activated with CAP, and OCR were then measured on a Seahorse XF Analyser (Agilent Technologies) in basal conditions, and upon the sequential addition of oligomycin (1.5 μM), Carbonyl cyanide-4-(trifluoromethoxy)-phenylhydrazone (FCCP, 3 μM), and rotenone/antimycin A (1 μM).

Transcriptome sequencing (RNA-seq)

Mouse neutrophils were purified from tumors upon MACS cell sorting, and then neutrophils were treated with CAP and cocultured for 12 hours. Total RNA was extracted using TRIzol reagent (Thermo Fisher, 15596018) following the manufacturer’s procedure. The total RNA quantity and purity were analyzed with a Bioanalyzer 2100 and RNA 6000 Nano LabChip Kit (Agilent, CA, USA, 5067–1511), and high-quality RNA samples with RIN > 7.0 were used to construct a sequencing library. After mRNA purification and cDNA amplification, we performed the 2× 150 bp paired-end sequencing (PE150) on an Illumina Novaseq 6000 (LC-Bio Technology CO., Ltd., Hangzhou, China) following the vendor’s recommended protocol. Gene differential expression analysis was performed by DESeq2 software between two different groups (and by edgeR between two samples). Genes with a false discovery rate (FDR) below 0.05 and absolute fold change ≥2 were considered differentially expressed genes.

Single-cell RNA-seq and data processing

Tumors from MB49-bearing mice with different treatment were digested in tissue dissociation solution according to the manufacturer’s instructions and single-cell suspensions were obtained. Tumor-infiltrating CD45⁺ cells were enriched by FACS. Cell suspensions were counted with Countess 3 automated cell counter to determine cell concentration and viability. Cells were then loaded onto a microfluidic chip and scRNA-seq libraries were constructed according to the manufacturer’s instructions. The scRNA-seq libraries were sequenced on an Illumina HiSeq X10 instrument with 150 bp paired end reads. Raw sequencing reads were processed to generate gene expression profiles using CeleScope v1.5.2 (Singleron Biotechnologies) with default parameters. Adapters and polyA tails were trimmed before aligning reads to the mouse GRCm38. Reads with the same cell barcode, UMI, and gene were grouped together to generate the UMI count matrix.

Single-cell RNA-seq datasets (GSE267718 and GSE201425) were downloaded from the NCBI Gene Expression Omnibus (GEO). Only samples from patients with matched peripheral blood and tumor tissue scRNA-seq profiles were retained for subsequent analyses. Raw UMI count matrices were imported into R(v4.5.0) and analyzed using Seurat (v5.2.1). For each sample, seurat objects were created from raw UMI count matrices. Doublets were identified using the scDblFinder package (v1.24.0) and excluded from downstream analyses. Quality control filtering was applied to remove low-quality cells, retaining cells with nFeature_RNA > 500 and < 6000 and mitochondrial transcript fraction (percent.mt) < 25%. Then, data were normalized using SCTransform with default parameters. Principal component analysis (PCA) was performed, and the top 30 PCs (dims = 1:30) were used for downstream analyses. To mitigate batch effects across samples, batch correction was carried out using the Harmony algorithm on the PCA embeddings.

After Harmony batch correction on the PCA space (dims = 1:30), we constructed the k-nearest neighbor graph using FindNeighbors (dims = 1:30) and performed graph-based clustering using FindClusters (resolution = 1.6 for GSE267718 and 1.0 for GSE201425). UMAP embeddings were generated using RunUMAP (dims = 1:30) for visualization. Differential gene expression analyses were performed using Seurat’s FindMarkers function with the Wilcoxon rank-sum test on the SCTransform-normalized expression values. Multiple testing correction was applied using the Benjamini–Hochberg procedure, and genes were considered differentially expressed if they met all of the following criteria: adjusted P value (padj) < 0.05, log2 fold change (log2FC) ≥ 1, and pct > 0.25. Cell types were annotated based on canonical marker gene expression. Heatmaps were visualized using ComplexHeatmap (v2.26.0).

Statistical analysis

Data analyses in the study were performed using GraphPad Prism. Figure legends specified the means and standard deviations, data presentation and statistical strategies, as well as all experimental repetitions and numbers of specimens. Two-tailed Student’s t-tests were used for two-group comparisons; one-way or two-way ANOVA for multiple groups; log-rank test for survival analyses; Wilcoxon rank-sum test for non-parametric data; RNA-seq analyses were performed using DESeq2 with FDR correction. P < 0.05 was considered as statistically significant (*P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001).

Acknowledgments

We thank the technical support from the Advanced Medical Research Institute/Translational Medicine Core Facility of Shandong University and Research Center for Basic Medical Science of Qilu hospital affiliated to Shandong University for consultation and instrument availability that supported this work. All schematic figures were created with BioRender. A, H. (2026) https://BioRender.com/dcf1ofc.

Funding:

This work was supported by National Natural Science Foundation of China (82303171 and 82473354 to W.J., 52477146 to X.W., 82272834 to N.Y.), Shandong Provincial Natural Science Foundation (ZR2023QH001 and ZR2024YQ053 to W.J., ZR2023LZL003 to N.Y.) and Tai Shan Scholar Foundation (tsqnz20221166 to W.J.).

Author contributions:

Conceptualization: W.J., X.W., B.S., N.Y., G.W., Z.C., X.J., Methodology: W.J., X.W., B.S., N.Y., G.W., Z.C., X.J., X.Y.L., Y.S., W.L., Y.L., R.S., X.M.L., Software: W.J., X.W., N.Y., G.W., Z.C., X.J., X.Y.L., Y.S., W.L., X.M.L, Validation: W.J., X.W., B.S., N.Y., G.W., Z.C., X.J., Y.S., W.L., Y.L., R.S., X.M.L, Formal analysis: W.J., X.W., N.Y., G.W., Z.C., X.J., Investigation: W.J., X.W., N.Y., G.W., Z.C., X.J., Resources: W.J., X.W., B.S., N.Y., Data curation:W.J., X.W., N.Y., G.W., Z.C., X.J., Writing—original draft: W.J., X.W., B.S., N.Y., G.W., Z.C., X.J., Writing—review & editing: W.J., X.W., B.S., N.Y., G.W., Z.C., X.J., Visualization: W.J., X.W., N.Y., G.W., Z.C., X.J., Supervision: W.J., X.W., B.S., N.Y., Project administration: W.J., X.W., B.S., N.Y., Funding acquisition: W.J., X.W., B.S., N.Y.

Competing interests:

All other authors declare they have no competing interests.

Data, code, and materials availability:

RNA-seq data generated in this study have been deposited in the Genome Sequence Archive (GSA) under accession numbers CRA039014 (https://ngdc.cncb.ac.cn/gsa/browse/CRA039014) and CRA039016 (https://ngdc.cncb.ac.cn/gsa/browse/CRA039016). Publicly available single-cell RNA-seq datasets GSE267718 and GSE201425 were obtained from the Gene Expression Omnibus (GEO). All other data and code needed to evaluate and reproduce the conclusions in the paper are present in the paper and/or the Supplementary Materials. This study did not generate new materials.

Supplementary Materials

The PDF file includes:

Figs. S1 to S8

Table S1

Legends for tables S2 and S3

Legend for movie S1

sciadv.aee2483_sm.pdf^{(15.6MB, pdf)}

Other Supplementary Material for this manuscript includes the following:

Tables S2 and S3

sciadv.aee2483_tables_s2_and_s3.zip^{(279KB, zip)}

Movie S1

sciadv.aee2483_movie_s1.zip^{(44MB, zip)}

REFERENCES

1.van Weverwijk A., de Visser K. E., Mechanisms driving the immunoregulatory function of cancer cells. Nat. Rev. Cancer 23, 193–215 (2023). [DOI] [PubMed] [Google Scholar]
2.de Visser K. E., Joyce J. A., The evolving tumor microenvironment: From cancer initiation to metastatic outgrowth. Cancer Cell 41, 374–403 (2023). [DOI] [PubMed] [Google Scholar]
3.Xiong S., Dong L., Cheng L., Neutrophils in cancer carcinogenesis and metastasis. J. Hematol. Oncol. 14, 173 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Hedrick C. C., Malanchi I., Neutrophils in cancer: Heterogeneous and multifaceted. Nat. Rev. Immunol. 22, 173–187 (2022). [DOI] [PubMed] [Google Scholar]
5.Linde I. L., Prestwood T. R., Qiu J., Pilarowski G., Linde M. H., Zhang X., Shen L., Reticker-Flynn N. E., Chiu D. K.-C., Sheu L. Y., Van Deursen S., Tolentino L. L., Song W.-C., Engleman E. G., Neutrophil-activating therapy for the treatment of cancer. Cancer Cell 41, 356–372.e10 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Liu S., Wu W., Du Y., Yin H., Chen Q., Yu W., Wang W., Yu J., Liu L., Lou W., Pu N., The evolution and heterogeneity of neutrophils in cancers: Origins, subsets, functions, orchestrations and clinical applications. Mol. Cancer 22, 148 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Ng M. S. F., Kwok I., Tan L., Shi C., Cerezo-Wallis D., Tan Y., Leong K., Calvo G. F., Yang K., Zhang Y., Jin J., Liong K. H., Wu D., He R., Liu D., Teh Y. C., Bleriot C., Caronni N., Liu Z., Duan K., Narang V., Ballesteros I., Moalli F., Li M., Chen J., Liu Y., Liu L., Qi J., Liu Y., Jiang L., Shen B., Cheng H., Cheng T., Angeli V., Sharma A., Loh Y.-H., Tey H. L., Chong S. Z., Iannacone M., Ostuni R., Hidalgo A., Ginhoux F., Ng L. G., Deterministic reprogramming of neutrophils within tumors. Science 383, eadf6493 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Kalafati L., Kourtzelis I., Schulte-Schrepping J., Li X., Hatzioannou A., Grinenko T., Hagag E., Sinha A., Has C., Dietz S., de Jesus Domingues A. M., Nati M., Sormendi S., Neuwirth A., Chatzigeorgiou A., Ziogas A., Lesche M., Dahl A., Henry I., Subramanian P., Wielockx B., Murray P., Mirtschink P., Chung K.-J., Schultze J. L., Netea M. G., Hajishengallis G., Verginis P., Mitroulis I., Chavakis T., Innate immune training of granulopoiesis promotes anti-tumor activity. Cell 183, 771–785.e12 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Garner H., Martinovic M., Liu N. Q., Bakker N. A. M., Velilla I. Q., Hau C. S., Vrijland K., Kaldenbach D., Kok M., de Wit E., de Visser K. E., Understanding and reversing mammary tumor-driven reprogramming of myelopoiesis to reduce metastatic spread. Cancer Cell 43, 1279–1295.e9 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Xue J., Zhao Z., Zhang L., Xue L., Shen S., Wen Y., Wei Z., Wang L., Kong L., Sun H., Ping Q., Mo R., Zhang C., Neutrophil-mediated anticancer drug delivery for suppression of postoperative malignant glioma recurrence. Nat. Nanotechnol. 12, 692–700 (2017). [DOI] [PubMed] [Google Scholar]
11.Ugolini A., De Leo A., Yu X., Scirocchi F., Liu X., Peixoto B., Scocozza D., Pace A., Perego M., Gardini A., D’Angelo L., Liu J. K. C., Etame A. B., Rughetti A., Nuti M., Santoro A., Vogelbaum M. A., Conejo-Garcia J. R., Rodriguez P. C., Veglia F., Functional reprogramming of neutrophils within the brain tumor microenvironment by hypoxia-driven histone lactylation. Cancer Discov. 15, 1270–1296 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Yu L., Liebenberg K., Shen Y., Liu F., Xu Z., Hao X., Wu L., Zhang W., Chan H. L., Wei B., Lorenzi P. L., Gao Y., Bado I., Becerra-Dominguez L., Rivas C. H., Aguirre S., Pingel B. C., Wu Y.-H., Ding Y., Liu J., Edwards D. G., Eberlin L. S., Zhang X. H., Tumor-derived arachidonic acid reprograms neutrophils to promote immune suppression and therapy resistance in triple-negative breast cancer. Immunity 58, 909–925.e7 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Karthik C., Sarngadharan S. C., Thomas V., Low-temperature plasma techniques in biomedical applications and therapeutics: An overview. Int. J. Mol. Sci. 25, 524 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Chen Z., Chen G., Obenchain R., Zhang R., Bai F., Fang T., Wang H., Lu Y., Wirz R., Gu Z., Cold atmospheric plasma delivery for biomedical applications. Mater. Today 54, 153–188 (2022). [Google Scholar]
15.Peng S., Feng Y., Yu K. N., Wu L., Chen G., Yang M., Zhao L., Cao W., Cui Q., Chen L., Li Q., Huang Y., Cheng C., Zhu F., Han W., Unleashing the power of cold atmospheric plasma: Inducing mitochondria damage-mediated mitotic catastrophe. Adv. Sci. 11, e2401842 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Chen G., Chen Z., Wang Z., Obenchain R., Wen D., Li H., Wirz R. E., Gu Z., Portable air-fed cold atmospheric plasma device for postsurgical cancer treatment. Sci. Adv. 7, eabg5686 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Byun J., Wu Y., Lee J., Kim J. S., Shim G., Oh Y. K., External cold atmospheric plasma-responsive on-site hydrogel for remodeling tumor immune microenvironment. Biomaterials 299, 122162 (2023). [DOI] [PubMed] [Google Scholar]
18.Tavares-da-Silva E., Pereira E., Pires A. S., Neves A. R., Braz-Guilherme C., Marques I. A., Abrantes A. M., Gonçalves A. C., Caramelo F., Silva-Teixeira R., Mendes F., Figueiredo A., Botelho M. F., Cold atmospheric plasma, a novel approach against bladder cancer, with higher sensitivity for the high-grade cell line. Biology 10, 41 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Mahdikia H., Saadati F., Freund E., Gaipl U. S., Majidzadeh-A K., Shokri B., Bekeschus S., Gas plasma irradiation of breast cancers promotes immunogenicity, tumor reduction, and an abscopal effect in vivo. Oncoimmunology 10, 1859731 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Lee C. B., Choi H. G., Gurmessa S. K., Jang I. T., Kumar N., Jiang Z., Kaushik N. K., Kim H. J., Enhancing antitumor immunity in Lewis lung cancer through plasma-treated medium-induced activation of dendritic cells. Cancer Cell Int. 24, 389 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Ueda Y., Kiyonaka S., Selfors L. M., Inoue K., Harada H., Doura T., Onuma K., Uchiyama M., Kurogi R., Yamada Y., Sun J. H., Sakaguchi R., Tado Y., Omatsu H., Suzuki H., Aoun M., Nakayama T., Kajimoto T., Yano T., Holmdahl R., Hamachi I., Inoue M., Mori Y., Takahashi N., Intratumour oxidative hotspots provide a niche for cancer cell dissemination. Nat. Cell Biol. 27, 530–543 (2025). [DOI] [PubMed] [Google Scholar]
22.Crist S. B., Nemkov T., Dumpit R. F., Dai J., Tapscott S. J., True L. D., Swarbrick A., Sullivan L. B., Nelson P. S., Hansen K. C., Ghajar C. M., Unchecked oxidative stress in skeletal muscle prevents outgrowth of disseminated tumour cells. Nat. Cell Biol. 24, 538–553 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.McDowell S. A. C., Luo R. B. E., Arabzadeh A., Doré S., Bennett N. C., Breton V., Karimi E., Rezanejad M., Yang R. R., Lach K. D., Issac M. S. M., Samborska B., Perus L. J. M., Moldoveanu D., Wei Y., Fiset B., Rayes R. F., Watson I. R., Kazak L., Guiot M.-C., Fiset P. O., Spicer J. D., Dannenberg A. J., Walsh L. A., Quail D. F., Neutrophil oxidative stress mediates obesity-associated vascular dysfunction and metastatic transmigration. Nat. Cancer 2, 545–562 (2021). [DOI] [PubMed] [Google Scholar]
24.Camargo S., Moskowitz O., Giladi A., Levinson M., Balaban R., Gola S., Raizman A., Lipczyc K., Richter A., Keren-Khadmy N., Barboy O., Dugach Y., Carmi Y., Sonnenblick A., Cohen M., Neutrophils physically interact with tumor cells to form a signaling niche promoting breast cancer aggressiveness. Nat. Cancer 6, 540–558 (2025). [DOI] [PubMed] [Google Scholar]
25.Benguigui M., Cooper T. J., Kalkar P., Schif-Zuck S., Halaban R., Bacchiocchi A., Kamer I., Deo A., Manobla B., Menachem R., Haj-Shomaly J., Vorontsova A., Raviv Z., Buxbaum C., Christopoulos P., Bar J., Lotem M., Sznol M., Ariel A., Shen-Orr S. S., Shaked Y., Interferon-stimulated neutrophils as a predictor of immunotherapy response. Cancer Cell 42, 253–265.e12 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Wu Y., Ma J., Yang X., Nan F., Zhang T., Ji S., Rao D., Feng H., Gao K., Gu X., Jiang S., Song G., Pan J., Zhang M., Xu Y., Zhang S., Fan Y., Wang X., Zhou J., Yang L., Fan J., Zhang X., Gao Q., Neutrophil profiling illuminates anti-tumor antigen-presenting potency. Cell 187, 1422–1439.e24 (2024). [DOI] [PubMed] [Google Scholar]
27.Xie X., Shi Q., Wu P., Zhang X., Kambara H., Su J., Yu H., Park S. Y., Guo R., Ren Q., Zhang S., Xu Y., Silberstein L. E., Cheng T., Ma F., Li C., Luo H. R., Single-cell transcriptome profiling reveals neutrophil heterogeneity in homeostasis and infection. Nat. Immunol. 21, 1119–1133 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Xue R., Zhang Q., Cao Q., Kong R., Xiang X., Liu H., Feng M., Wang F., Cheng J., Li Z., Zhan Q., Deng M., Zhu J., Zhang Z., Zhang N., Liver tumour immune microenvironment subtypes and neutrophil heterogeneity. Nature 612, 141–147 (2022). [DOI] [PubMed] [Google Scholar]
29.Gong D., Arbesfeld-Qiu J. M., Perrault E., Bae J. W., Hwang W. L., Spatial oncology: Translating contextual biology to the clinic. Cancer Cell 42, 1653–1675 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Maas R. R., Soukup K., Fournier N., Massara M., Galland S., Kornete M., Wischnewski V., Lourenco J., Croci D., Álvarez-Prado Á. F., Marie D. N., Lilja J., Marcone R., Calvo G. F., Mendez R. S., Aubel P., Bejarano L., Wirapati P., Ballesteros I., Hidalgo A., Hottinger A. F., Brouland J.-P., Daniel R. T., Hegi M. E., Joyce J. A., The local microenvironment drives activation of neutrophils in human brain tumors. Cell 186, 4546–4566.e27 (2023). [DOI] [PubMed] [Google Scholar]
31.Galassi C., Chan T. A., Vitale I., Galluzzi L., The hallmarks of cancer immune evasion. Cancer Cell 42, 1825–1863 (2024). [DOI] [PubMed] [Google Scholar]
32.Tran M. A., Youssef D., Shroff S., Chowhan D., Beaumont K. G., Sebra R., Mehrazin R., Wiklund P., Lin J. J., Horowitz A., Farkas A. M., Galsky M. D., Sfakianos J. P., Bhardwaj N., Urine scRNAseq reveals new insights into the bladder tumor immune microenvironment. J. Exp. Med. 221, e20240045 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Shi X., Li Z., Yao R., Cheng Q., Li W., Wu R., Xie Z., Zhu Y., Qiu X., Yang S., Zhou T., Hu J., Zhang Y., Wu T., Zhao Y., Zhang Y., Wu J., Wang H., Jiang X., Chen L., Single-cell atlas of diverse immune populations in the advanced biliary tract cancer microenvironment. NPJ Precis. Oncol. 6, 58 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Jing W., Wang G., Cui Z., Li X., Zeng S., Jiang X., Li W., Han B., Xing N., Zhao Y., Chen S., Shi B., Tumor-neutrophil cross talk orchestrates the tumor microenvironment to determine the bladder cancer progression. Proc. Natl. Acad. Sci. U.S.A. 121, e2312855121 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Chronopoulos J., Crespo M., Chavakis T., Central trained immunity in the context of bladder cancer immunotherapy. Cancer Cell 43, 1394–1396 (2025). [DOI] [PubMed] [Google Scholar]
36.Nadal R., Valderrama B. P., Bellmunt J., Progress in systemic therapy for advanced-stage urothelial carcinoma. Nat. Rev. Clin. Oncol. 21, 8–27 (2024). [DOI] [PubMed] [Google Scholar]
37.Meeks J. J., Black P. C., Galsky M., Grivas P., Hahn N. M., Hussain S. A., Milowsky M. I., Steinberg G. D., Svatek R. S., Rosenberg J. E., Checkpoint inhibitors in urothelial carcinoma-future directions and biomarker selection. Eur. Urol. 84, 473–483 (2023). [DOI] [PubMed] [Google Scholar]
38.Debatin N. F., Bady E., Mandelkow T., Huang Z., Lurati M. C. J., Raedler J. B., Müller J. H., Vettorazzi E., Plage H., Samtleben H., Klatte T., Hofbauer S., Elezkurtaj S., Furlano K., Weinberger S., Bruch P. G., Horst D., Roßner F., Schallenberg S., Marx A. H., Fisch M., Rink M., Slojewski M., Kaczmarek K., Ecke T. H., Hallmann S., Koch S., Adamini N., Lennartz M., Minner S., Simon R., Sauter G., Zecha H., Schlomm T., Blessin N. C., Prognostic impact and spatial interplay of immune cells in urothelial cancer. Eur. Urol. 86, 42–51 (2024). [DOI] [PubMed] [Google Scholar]
39.Goubet A. G., Lordello L., Alves Costa Silva C., Peguillet I., Gazzano M., Mbogning-Fonkou M. D., Thelemaque C., Lebacle C., Thibault C., Audenet F., Pignot G., Gravis G., Helissey C., Campedel L., Roupret M., Xylinas E., Ouzaid I., Dubuisson A., Mazzenga M., Flament C., Ly P., Marty V., Signolle N., Sauvat A., Sbarrato T., Filahi M., Davin C., Haddad G., Khalil J. B., Bleriot C., Danlos F.-X., Dunsmore G., Mulder K., Silvin A., Raoult T., Archambaud B., Belhechmi S., Boneca I. G., Cayet N., Moya-Nilges M., Mallet A., Daillere R., Rouleau E., Radulescu C., Allory Y., Fieschi J., Rouanne M., Ginhoux F., Teuff G. L., Derosa L., Marabelle A., Van Dorp J., Van Dijk N., Van Der Heijden M. S., Besse B., Andre F., Merad M., Kroemer G., Scoazec J.-Y., Zitvogel L., Loriot Y., Escherichia coli-specific CXCL13-producing TFH are associated with clinical efficacy of neoadjuvant PD-1 blockade against muscle-invasive bladder cancer. Cancer Discov. 12, 2280–2307 (2022). [DOI] [PubMed] [Google Scholar]
40.Wei W., Li H., Tian S., Zhang C., Liu J., Tao W., Cai T., Dong Y., Wang C., Lu D., Ai Y., Zhang W., Wang H., Liu K., Fan Y., Gao Y., Huang Q., Ma X., Wang B., Zhang X., Huang Y., Asparagine drives immune evasion in bladder cancer via RIG-I stability and type I IFN signaling. J. Clin. Invest. 135, e186648 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Yang C., Geng H., Yang X., Ji S., Liu Z., Feng H., Li Q., Zhang T., Zhang S., Ma X., Zhu C., Xu N., Xia Y., Li Y., Wang H., Yu C., Du S., Miao B., Xu L., Wang H., Cao Y., Li B., Zhu L., Tang X., Zhang H., Zhu C., Huang Z., Leng C., Hu H., Chen X., Yuan S., Jin G., Bernards R., Sun C., Zheng Q., Qin W., Gao Q., Wang C., Targeting the immune privilege of tumor-initiating cells to enhance cancer immunotherapy. Cancer Cell 42, 2064–2081.e19 (2024). [DOI] [PubMed] [Google Scholar]
42.Guo C., Sharp A., Gurel B., Crespo M., Figueiredo I., Jain S., Vogl U., Rekowski J., Rouhifard M., Gallagher L., Yuan W., Carreira S., Chandran K., Paschalis A., Colombo I., Stathis A., Bertan C., Seed G., Goodall J., Raynaud F., Ruddle R., Swales K. E., Malia J., Bogdan D., Tiu C., Caldwell R., Aversa C., Ferreira A., Neeb A., Tunariu N., Westaby D., Carmichael J., de la Maza M. D. F., Yap C., Matthews R., Badham H., Prout T., Turner A., Parmar M., Tovey H., Riisnaes R., Flohr P., Gil J., Waugh D., Decordova S., Schlag A., Calì B., Alimonti A., de Bono J. S., Targeting myeloid chemotaxis to reverse prostate cancer therapy resistance. Nature 623, 1053–1061 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Kwak J. W., Houghton A. M., Targeting neutrophils for cancer therapy. Nat. Rev. Drug Discov. 24, 666–684 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

Figs. S1 to S8

Table S1

Legends for tables S2 and S3

Legend for movie S1

sciadv.aee2483_sm.pdf^{(15.6MB, pdf)}

Tables S2 and S3

sciadv.aee2483_tables_s2_and_s3.zip^{(279KB, zip)}

Movie S1

sciadv.aee2483_movie_s1.zip^{(44MB, zip)}

Data Availability Statement

[R1] 1.van Weverwijk A., de Visser K. E., Mechanisms driving the immunoregulatory function of cancer cells. Nat. Rev. Cancer 23, 193–215 (2023). [DOI] [PubMed] [Google Scholar]

[R2] 2.de Visser K. E., Joyce J. A., The evolving tumor microenvironment: From cancer initiation to metastatic outgrowth. Cancer Cell 41, 374–403 (2023). [DOI] [PubMed] [Google Scholar]

[R3] 3.Xiong S., Dong L., Cheng L., Neutrophils in cancer carcinogenesis and metastasis. J. Hematol. Oncol. 14, 173 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] 4.Hedrick C. C., Malanchi I., Neutrophils in cancer: Heterogeneous and multifaceted. Nat. Rev. Immunol. 22, 173–187 (2022). [DOI] [PubMed] [Google Scholar]

[R5] 5.Linde I. L., Prestwood T. R., Qiu J., Pilarowski G., Linde M. H., Zhang X., Shen L., Reticker-Flynn N. E., Chiu D. K.-C., Sheu L. Y., Van Deursen S., Tolentino L. L., Song W.-C., Engleman E. G., Neutrophil-activating therapy for the treatment of cancer. Cancer Cell 41, 356–372.e10 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.Liu S., Wu W., Du Y., Yin H., Chen Q., Yu W., Wang W., Yu J., Liu L., Lou W., Pu N., The evolution and heterogeneity of neutrophils in cancers: Origins, subsets, functions, orchestrations and clinical applications. Mol. Cancer 22, 148 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R7] 7.Ng M. S. F., Kwok I., Tan L., Shi C., Cerezo-Wallis D., Tan Y., Leong K., Calvo G. F., Yang K., Zhang Y., Jin J., Liong K. H., Wu D., He R., Liu D., Teh Y. C., Bleriot C., Caronni N., Liu Z., Duan K., Narang V., Ballesteros I., Moalli F., Li M., Chen J., Liu Y., Liu L., Qi J., Liu Y., Jiang L., Shen B., Cheng H., Cheng T., Angeli V., Sharma A., Loh Y.-H., Tey H. L., Chong S. Z., Iannacone M., Ostuni R., Hidalgo A., Ginhoux F., Ng L. G., Deterministic reprogramming of neutrophils within tumors. Science 383, eadf6493 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R8] 8.Kalafati L., Kourtzelis I., Schulte-Schrepping J., Li X., Hatzioannou A., Grinenko T., Hagag E., Sinha A., Has C., Dietz S., de Jesus Domingues A. M., Nati M., Sormendi S., Neuwirth A., Chatzigeorgiou A., Ziogas A., Lesche M., Dahl A., Henry I., Subramanian P., Wielockx B., Murray P., Mirtschink P., Chung K.-J., Schultze J. L., Netea M. G., Hajishengallis G., Verginis P., Mitroulis I., Chavakis T., Innate immune training of granulopoiesis promotes anti-tumor activity. Cell 183, 771–785.e12 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R9] 9.Garner H., Martinovic M., Liu N. Q., Bakker N. A. M., Velilla I. Q., Hau C. S., Vrijland K., Kaldenbach D., Kok M., de Wit E., de Visser K. E., Understanding and reversing mammary tumor-driven reprogramming of myelopoiesis to reduce metastatic spread. Cancer Cell 43, 1279–1295.e9 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] 10.Xue J., Zhao Z., Zhang L., Xue L., Shen S., Wen Y., Wei Z., Wang L., Kong L., Sun H., Ping Q., Mo R., Zhang C., Neutrophil-mediated anticancer drug delivery for suppression of postoperative malignant glioma recurrence. Nat. Nanotechnol. 12, 692–700 (2017). [DOI] [PubMed] [Google Scholar]

[R11] 11.Ugolini A., De Leo A., Yu X., Scirocchi F., Liu X., Peixoto B., Scocozza D., Pace A., Perego M., Gardini A., D’Angelo L., Liu J. K. C., Etame A. B., Rughetti A., Nuti M., Santoro A., Vogelbaum M. A., Conejo-Garcia J. R., Rodriguez P. C., Veglia F., Functional reprogramming of neutrophils within the brain tumor microenvironment by hypoxia-driven histone lactylation. Cancer Discov. 15, 1270–1296 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R12] 12.Yu L., Liebenberg K., Shen Y., Liu F., Xu Z., Hao X., Wu L., Zhang W., Chan H. L., Wei B., Lorenzi P. L., Gao Y., Bado I., Becerra-Dominguez L., Rivas C. H., Aguirre S., Pingel B. C., Wu Y.-H., Ding Y., Liu J., Edwards D. G., Eberlin L. S., Zhang X. H., Tumor-derived arachidonic acid reprograms neutrophils to promote immune suppression and therapy resistance in triple-negative breast cancer. Immunity 58, 909–925.e7 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R13] 13.Karthik C., Sarngadharan S. C., Thomas V., Low-temperature plasma techniques in biomedical applications and therapeutics: An overview. Int. J. Mol. Sci. 25, 524 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] 14.Chen Z., Chen G., Obenchain R., Zhang R., Bai F., Fang T., Wang H., Lu Y., Wirz R., Gu Z., Cold atmospheric plasma delivery for biomedical applications. Mater. Today 54, 153–188 (2022). [Google Scholar]

[R15] 15.Peng S., Feng Y., Yu K. N., Wu L., Chen G., Yang M., Zhao L., Cao W., Cui Q., Chen L., Li Q., Huang Y., Cheng C., Zhu F., Han W., Unleashing the power of cold atmospheric plasma: Inducing mitochondria damage-mediated mitotic catastrophe. Adv. Sci. 11, e2401842 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.Chen G., Chen Z., Wang Z., Obenchain R., Wen D., Li H., Wirz R. E., Gu Z., Portable air-fed cold atmospheric plasma device for postsurgical cancer treatment. Sci. Adv. 7, eabg5686 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] 17.Byun J., Wu Y., Lee J., Kim J. S., Shim G., Oh Y. K., External cold atmospheric plasma-responsive on-site hydrogel for remodeling tumor immune microenvironment. Biomaterials 299, 122162 (2023). [DOI] [PubMed] [Google Scholar]

[R18] 18.Tavares-da-Silva E., Pereira E., Pires A. S., Neves A. R., Braz-Guilherme C., Marques I. A., Abrantes A. M., Gonçalves A. C., Caramelo F., Silva-Teixeira R., Mendes F., Figueiredo A., Botelho M. F., Cold atmospheric plasma, a novel approach against bladder cancer, with higher sensitivity for the high-grade cell line. Biology 10, 41 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] 19.Mahdikia H., Saadati F., Freund E., Gaipl U. S., Majidzadeh-A K., Shokri B., Bekeschus S., Gas plasma irradiation of breast cancers promotes immunogenicity, tumor reduction, and an abscopal effect in vivo. Oncoimmunology 10, 1859731 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] 20.Lee C. B., Choi H. G., Gurmessa S. K., Jang I. T., Kumar N., Jiang Z., Kaushik N. K., Kim H. J., Enhancing antitumor immunity in Lewis lung cancer through plasma-treated medium-induced activation of dendritic cells. Cancer Cell Int. 24, 389 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R21] 21.Ueda Y., Kiyonaka S., Selfors L. M., Inoue K., Harada H., Doura T., Onuma K., Uchiyama M., Kurogi R., Yamada Y., Sun J. H., Sakaguchi R., Tado Y., Omatsu H., Suzuki H., Aoun M., Nakayama T., Kajimoto T., Yano T., Holmdahl R., Hamachi I., Inoue M., Mori Y., Takahashi N., Intratumour oxidative hotspots provide a niche for cancer cell dissemination. Nat. Cell Biol. 27, 530–543 (2025). [DOI] [PubMed] [Google Scholar]

[R22] 22.Crist S. B., Nemkov T., Dumpit R. F., Dai J., Tapscott S. J., True L. D., Swarbrick A., Sullivan L. B., Nelson P. S., Hansen K. C., Ghajar C. M., Unchecked oxidative stress in skeletal muscle prevents outgrowth of disseminated tumour cells. Nat. Cell Biol. 24, 538–553 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.McDowell S. A. C., Luo R. B. E., Arabzadeh A., Doré S., Bennett N. C., Breton V., Karimi E., Rezanejad M., Yang R. R., Lach K. D., Issac M. S. M., Samborska B., Perus L. J. M., Moldoveanu D., Wei Y., Fiset B., Rayes R. F., Watson I. R., Kazak L., Guiot M.-C., Fiset P. O., Spicer J. D., Dannenberg A. J., Walsh L. A., Quail D. F., Neutrophil oxidative stress mediates obesity-associated vascular dysfunction and metastatic transmigration. Nat. Cancer 2, 545–562 (2021). [DOI] [PubMed] [Google Scholar]

[R24] 24.Camargo S., Moskowitz O., Giladi A., Levinson M., Balaban R., Gola S., Raizman A., Lipczyc K., Richter A., Keren-Khadmy N., Barboy O., Dugach Y., Carmi Y., Sonnenblick A., Cohen M., Neutrophils physically interact with tumor cells to form a signaling niche promoting breast cancer aggressiveness. Nat. Cancer 6, 540–558 (2025). [DOI] [PubMed] [Google Scholar]

[R25] 25.Benguigui M., Cooper T. J., Kalkar P., Schif-Zuck S., Halaban R., Bacchiocchi A., Kamer I., Deo A., Manobla B., Menachem R., Haj-Shomaly J., Vorontsova A., Raviv Z., Buxbaum C., Christopoulos P., Bar J., Lotem M., Sznol M., Ariel A., Shen-Orr S. S., Shaked Y., Interferon-stimulated neutrophils as a predictor of immunotherapy response. Cancer Cell 42, 253–265.e12 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] 26.Wu Y., Ma J., Yang X., Nan F., Zhang T., Ji S., Rao D., Feng H., Gao K., Gu X., Jiang S., Song G., Pan J., Zhang M., Xu Y., Zhang S., Fan Y., Wang X., Zhou J., Yang L., Fan J., Zhang X., Gao Q., Neutrophil profiling illuminates anti-tumor antigen-presenting potency. Cell 187, 1422–1439.e24 (2024). [DOI] [PubMed] [Google Scholar]

[R27] 27.Xie X., Shi Q., Wu P., Zhang X., Kambara H., Su J., Yu H., Park S. Y., Guo R., Ren Q., Zhang S., Xu Y., Silberstein L. E., Cheng T., Ma F., Li C., Luo H. R., Single-cell transcriptome profiling reveals neutrophil heterogeneity in homeostasis and infection. Nat. Immunol. 21, 1119–1133 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R28] 28.Xue R., Zhang Q., Cao Q., Kong R., Xiang X., Liu H., Feng M., Wang F., Cheng J., Li Z., Zhan Q., Deng M., Zhu J., Zhang Z., Zhang N., Liver tumour immune microenvironment subtypes and neutrophil heterogeneity. Nature 612, 141–147 (2022). [DOI] [PubMed] [Google Scholar]

[R29] 29.Gong D., Arbesfeld-Qiu J. M., Perrault E., Bae J. W., Hwang W. L., Spatial oncology: Translating contextual biology to the clinic. Cancer Cell 42, 1653–1675 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.Maas R. R., Soukup K., Fournier N., Massara M., Galland S., Kornete M., Wischnewski V., Lourenco J., Croci D., Álvarez-Prado Á. F., Marie D. N., Lilja J., Marcone R., Calvo G. F., Mendez R. S., Aubel P., Bejarano L., Wirapati P., Ballesteros I., Hidalgo A., Hottinger A. F., Brouland J.-P., Daniel R. T., Hegi M. E., Joyce J. A., The local microenvironment drives activation of neutrophils in human brain tumors. Cell 186, 4546–4566.e27 (2023). [DOI] [PubMed] [Google Scholar]

[R31] 31.Galassi C., Chan T. A., Vitale I., Galluzzi L., The hallmarks of cancer immune evasion. Cancer Cell 42, 1825–1863 (2024). [DOI] [PubMed] [Google Scholar]

[R32] 32.Tran M. A., Youssef D., Shroff S., Chowhan D., Beaumont K. G., Sebra R., Mehrazin R., Wiklund P., Lin J. J., Horowitz A., Farkas A. M., Galsky M. D., Sfakianos J. P., Bhardwaj N., Urine scRNAseq reveals new insights into the bladder tumor immune microenvironment. J. Exp. Med. 221, e20240045 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.Shi X., Li Z., Yao R., Cheng Q., Li W., Wu R., Xie Z., Zhu Y., Qiu X., Yang S., Zhou T., Hu J., Zhang Y., Wu T., Zhao Y., Zhang Y., Wu J., Wang H., Jiang X., Chen L., Single-cell atlas of diverse immune populations in the advanced biliary tract cancer microenvironment. NPJ Precis. Oncol. 6, 58 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R34] 34.Jing W., Wang G., Cui Z., Li X., Zeng S., Jiang X., Li W., Han B., Xing N., Zhao Y., Chen S., Shi B., Tumor-neutrophil cross talk orchestrates the tumor microenvironment to determine the bladder cancer progression. Proc. Natl. Acad. Sci. U.S.A. 121, e2312855121 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] 35.Chronopoulos J., Crespo M., Chavakis T., Central trained immunity in the context of bladder cancer immunotherapy. Cancer Cell 43, 1394–1396 (2025). [DOI] [PubMed] [Google Scholar]

[R36] 36.Nadal R., Valderrama B. P., Bellmunt J., Progress in systemic therapy for advanced-stage urothelial carcinoma. Nat. Rev. Clin. Oncol. 21, 8–27 (2024). [DOI] [PubMed] [Google Scholar]

[R37] 37.Meeks J. J., Black P. C., Galsky M., Grivas P., Hahn N. M., Hussain S. A., Milowsky M. I., Steinberg G. D., Svatek R. S., Rosenberg J. E., Checkpoint inhibitors in urothelial carcinoma-future directions and biomarker selection. Eur. Urol. 84, 473–483 (2023). [DOI] [PubMed] [Google Scholar]

[R38] 38.Debatin N. F., Bady E., Mandelkow T., Huang Z., Lurati M. C. J., Raedler J. B., Müller J. H., Vettorazzi E., Plage H., Samtleben H., Klatte T., Hofbauer S., Elezkurtaj S., Furlano K., Weinberger S., Bruch P. G., Horst D., Roßner F., Schallenberg S., Marx A. H., Fisch M., Rink M., Slojewski M., Kaczmarek K., Ecke T. H., Hallmann S., Koch S., Adamini N., Lennartz M., Minner S., Simon R., Sauter G., Zecha H., Schlomm T., Blessin N. C., Prognostic impact and spatial interplay of immune cells in urothelial cancer. Eur. Urol. 86, 42–51 (2024). [DOI] [PubMed] [Google Scholar]

[R39] 39.Goubet A. G., Lordello L., Alves Costa Silva C., Peguillet I., Gazzano M., Mbogning-Fonkou M. D., Thelemaque C., Lebacle C., Thibault C., Audenet F., Pignot G., Gravis G., Helissey C., Campedel L., Roupret M., Xylinas E., Ouzaid I., Dubuisson A., Mazzenga M., Flament C., Ly P., Marty V., Signolle N., Sauvat A., Sbarrato T., Filahi M., Davin C., Haddad G., Khalil J. B., Bleriot C., Danlos F.-X., Dunsmore G., Mulder K., Silvin A., Raoult T., Archambaud B., Belhechmi S., Boneca I. G., Cayet N., Moya-Nilges M., Mallet A., Daillere R., Rouleau E., Radulescu C., Allory Y., Fieschi J., Rouanne M., Ginhoux F., Teuff G. L., Derosa L., Marabelle A., Van Dorp J., Van Dijk N., Van Der Heijden M. S., Besse B., Andre F., Merad M., Kroemer G., Scoazec J.-Y., Zitvogel L., Loriot Y., Escherichia coli-specific CXCL13-producing TFH are associated with clinical efficacy of neoadjuvant PD-1 blockade against muscle-invasive bladder cancer. Cancer Discov. 12, 2280–2307 (2022). [DOI] [PubMed] [Google Scholar]

[R40] 40.Wei W., Li H., Tian S., Zhang C., Liu J., Tao W., Cai T., Dong Y., Wang C., Lu D., Ai Y., Zhang W., Wang H., Liu K., Fan Y., Gao Y., Huang Q., Ma X., Wang B., Zhang X., Huang Y., Asparagine drives immune evasion in bladder cancer via RIG-I stability and type I IFN signaling. J. Clin. Invest. 135, e186648 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R41] 41.Yang C., Geng H., Yang X., Ji S., Liu Z., Feng H., Li Q., Zhang T., Zhang S., Ma X., Zhu C., Xu N., Xia Y., Li Y., Wang H., Yu C., Du S., Miao B., Xu L., Wang H., Cao Y., Li B., Zhu L., Tang X., Zhang H., Zhu C., Huang Z., Leng C., Hu H., Chen X., Yuan S., Jin G., Bernards R., Sun C., Zheng Q., Qin W., Gao Q., Wang C., Targeting the immune privilege of tumor-initiating cells to enhance cancer immunotherapy. Cancer Cell 42, 2064–2081.e19 (2024). [DOI] [PubMed] [Google Scholar]

[R42] 42.Guo C., Sharp A., Gurel B., Crespo M., Figueiredo I., Jain S., Vogl U., Rekowski J., Rouhifard M., Gallagher L., Yuan W., Carreira S., Chandran K., Paschalis A., Colombo I., Stathis A., Bertan C., Seed G., Goodall J., Raynaud F., Ruddle R., Swales K. E., Malia J., Bogdan D., Tiu C., Caldwell R., Aversa C., Ferreira A., Neeb A., Tunariu N., Westaby D., Carmichael J., de la Maza M. D. F., Yap C., Matthews R., Badham H., Prout T., Turner A., Parmar M., Tovey H., Riisnaes R., Flohr P., Gil J., Waugh D., Decordova S., Schlag A., Calì B., Alimonti A., de Bono J. S., Targeting myeloid chemotaxis to reverse prostate cancer therapy resistance. Nature 623, 1053–1061 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] 43.Kwak J. W., Houghton A. M., Targeting neutrophils for cancer therapy. Nat. Rev. Drug Discov. 24, 666–684 (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Biophysical reprogramming of immunosuppressive neutrophils by cold atmospheric plasma reinvigorates antitumor immunity

Ganyu Wang

Zhiwei Cui

Xin Jiang

Ying Sun

Yifan Liu

Xinyuan Li

Wushan Li

Ruirui Sun

Xuemei Liu

Nengwang Yu

Xiaolong Wang

Benkang Shi

Weiqiang Jing

Roles

Abstract

INTRODUCTION

RESULTS

Tumor-associated neutrophils exhibit an immunosuppressive phenotype correlated with tumor invasiveness in bladder cancer

Fig. 1. Tumor-associated neutrophils exhibit an immunosuppressive phenotype correlated with tumor invasiveness in bladder cancer.

Engineering and validation of a localized CAP delivery system for intravesical immunotherapy

Fig. 2. Engineering and validation of a localized CAP delivery system for intravesical immunotherapy.

CAP induces oxidative activation and functional reprogramming of neutrophils toward an antitumor phenotype

Fig. 3. CAP induces oxidative activation and functional reprogramming of neutrophils toward an antitumor phenotype.

CAP-mediated inhibition of neutrophil mitophagy restores mitochondrial function and drives immune activation

Fig. 4. CAP-mediated inhibition of neutrophil mitophagy restores mitochondrial function and drives immune activation.

Localized CAP treatment suppresses bladder tumor growth by restoration of neutrophil cytotoxic capacity in vivo

Fig. 5. Localized CAP treatment suppresses bladder tumor growth by reprogramming neutrophils in vivo.

CAP synergizes with anti-PD-1 therapy by remodeling neutrophil-mediated immunosuppressive niches

Fig. 6. CAP synergizes with anti–PD-1 therapy by remodeling neutrophil-mediated immunosuppressive niches.

Fig. 7. CAP therapy displays excellent antitumor activity in hu-HSC-NOG mice.

DISCUSSION

MATERIALS AND METHODS

Mice

Patients

Cell lines

Xenium 5 k Experiment and Analyzer

Sample preparation

Gene panels

Xenium Analyzer

Clustering and analysis

CAP equipment and treatment

ROS/RNS measurement

Cell Isolation and in vitro assay

Confocal microscopy

Histology and immunofluorescence analysis

Western blot

Intracellular ROS detection

Mouse XL cytokine array

Flow cytometry

Three-dimensional spheroid culture

ELISA

Analysis of OCR with the seahorse XF platform

Transcriptome sequencing (RNA-seq)

Single-cell RNA-seq and data processing

Statistical analysis

Acknowledgments

Funding:

Author contributions:

Competing interests:

Data, code, and materials availability:

Supplementary Materials

The PDF file includes:

Other Supplementary Material for this manuscript includes the following:

REFERENCES

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases