Abstract

   Black phosphorus (BP) has demonstrated potential as a drug carrier and
   photothermal agent in cancer therapy; however, its intrinsic functions
   in cancer treatment remain underexplored. This study investigates the
   immunomodulatory effects of polyethylene glycol‐functionalized BP
   (BP‐PEG) nanosheets in breast cancer models. Using immunocompetent
   mouse models‐including 4T1 orthotopic BALB/c mice and MMTV‐PyMT
   transgenic mice, it is found that BP‐PEG significantly inhibits tumor
   growth and metastasis without directly inducing cytotoxicity in tumor
   cells. Mass cytometry analysis reveals that BP‐PEG reshapes the tumor
   immune microenvironment by recruiting neutrophils. Neutrophil depletion
   experiments further demonstrate that the antitumor effects of BP‐PEG
   are dependent on neutrophils. Moreover, bulk and single‐cell RNA
   sequencing indicate that BP‐PEG is mainly taken up by macrophages,
   leading to the release of inflammatory factors such as IL1A and CXCL2,
   which enhance neutrophil recruitment and activation, thereby amplifying
   the antitumor immune response. Finally, co‐culture assays confirm that
   BP‐PEG indeed enhances the antitumor activity of neutrophils and
   natural killer (NK) cells. These findings position BP‐PEG as an
   immunomodulatory agent capable of reprogramming the tumor
   microenvironment to promote innate immunity against breast cancer. By
   stimulating neutrophil‐mediated antitumor activity, BP‐PEG offers a
   unique therapeutic approach that can potentially enhance the efficacy
   of existing cancer immunotherapies.

   Keywords: black phosphorus, cancer therapy, immunomodulatory,
   neutrophil
     __________________________________________________________________

   This study explores black phosphorus (BP) nanosheets in breast cancer
   models, showing their ability to inhibit tumor growth. BP‐PEG enhances
   immune responses by promoting neutrophil recruitment and reprogramming
   tumor‐associated neutrophils to an antitumor phenotype, boosting both
   innate and adaptive antitumor immune activities. Combining BP therapy
   with immunotherapies or chemotherapies potentially counteract
   neutrophil‐induced immunosuppression, restoring or enhancing their
   therapeutic efficacy.

   graphic file with name ADVS-12-2414779-g003.jpg

1. Introduction

   Cancer remains a leading cause of mortality worldwide, necessitating
   innovative therapeutic strategies.^[ [44]^1 ^] 2D nanomaterials have
   gained significant attention in oncology due to their unique properties
   and multifunctional applications.^[ [45]^2 , [46]^3 ^] Among them,
   black phosphorus (BP) has emerged as a promising candidate for cancer
   therapy owing to its layered structure, tunable electronic properties,
   and biodegradability.^[ [47]^4 , [48]^5 , [49]^6 ^] BP consists of
   corrugated planes of phosphorus atoms connected by strong intralayer
   P–P bonds and weak interlayer van der Waals forces, allowing
   exfoliation into few‐layer or monolayer nanosheets.^[ [50]^7 ^] This
   exfoliation results in a layer‐dependent bandgap ranging from 0.3 eV
   (bulk BP) to ≈2 eV (monolayer BP), endowing the material with
   remarkable optical and electronic properties suitable for biomedical
   applications. Notably, BP degrades into nontoxic phosphate ions under
   physiological conditions.^[ [51]^8 ^] Its responsiveness to pondus
   hydrogenii（pH） and near‐infrared (NIR) radiation enables controlled
   drug release and reactive oxygen species (ROS) generation, making it
   attractive for photothermal and photodynamic therapies.^[ [52]^9 ,
   [53]^10 ^]

   Despite these attributes, most studies have focused on BP's role as a
   drug carrier or photothermal agent,^[ [54]^9 ^] with limited
   exploration of its direct impact on the tumor microenvironment (TME)
   and the immune system. The TME is a complex milieu of tumor cells,
   stromal cells, immune cells, and extracellular matrix components that
   contribute to tumor progression and therapy resistance.^[ [55]^11 ,
   [56]^12 ^] Immune cells within the TME can suppress or promote tumor
   growth depending on their activation states.^[ [57]^13 , [58]^14 ,
   [59]^15 ^] Advancements in cancer immunotherapy have highlighted the
   potential of modulating the immune system to achieve durable anti‐tumor
   responses.^[ [60]^16 ^] Nanomaterials have been investigated as
   immunomodulatory agents capable of reprogramming the TME to overcome
   immunosuppression.^[ [61]^17 , [62]^18 , [63]^19 ^] Certain
   nanomaterials act as adjuvants, enhancing antigen presentation and
   stimulating immune cell activation.^[ [64]^20 , [65]^21 ^] However,
   BP's immunomodulatory effects, particularly on innate immune cells like
   neutrophils and macrophages, remain underexplored.

   Neutrophils, the most abundant leukocytes in human blood, play critical
   roles in innate immunity and have been implicated in cancer
   progression.^[ [66]^22 ^] While traditionally considered
   tumor‐promoting, emerging evidence suggests that neutrophils can
   exhibit antitumor functions under specific conditions.^[ [67]^23 ^]
   Polarizing neutrophils toward an antitumor phenotype represents a novel
   therapeutic avenue.^[ [68]^24 , [69]^25 ^] Given BP's potential
   interactions with immune cells upon administration,^[ [70]^26 ^]
   understanding its effects on the TME's immune components is crucial.
   BP's degradation products and surface properties may influence immune
   cell behavior, leading to immune stimulation. Functionalization with
   biocompatible polymers like polyethylene glycol (PEG) can enhance BP's
   stability and modulate its biological interactions.^[ [71]^27 ^]

   In this study, we investigated BP's direct immunomodulatory effects in
   a breast cancer model. We synthesized stable BP‐PEG nanosheets and
   evaluated their impact on tumor growth and metastasis in
   immunocompetent mouse models. We focused on how BP‐PEG influences the
   TME, particularly neutrophil recruitment and activation, and whether
   these effects contribute to its antitumor activity. Using in vivo
   experiments, including neutrophil depletion and immunodeficient mouse
   models, as well as advanced analytical techniques like mass cytometry
   and single‐cell RNA sequencing, we provide insights into BP‐PEG's
   immunomodulatory functions. Our findings suggest that BP‐PEG can
   reshape the tumor's immune landscape, promoting innate immune responses
   that inhibit tumor progression.

   This research enhances our understanding of BP's interactions with the
   immune system and highlights BP‐PEG's potential as a novel
   immunotherapeutic agent. By harnessing innate immune responses,
   specifically neutrophil‐mediated actions, BP‐PEG offers a promising
   strategy that complements existing therapies and may overcome
   limitations of conventional chemotherapies, which often promote
   neutrophil infiltration and induce chemoresistance.^[ [72]^28 ^]

2. Results

2.1. BP‐PEG Exhibits Antitumor Activity in Immunocompetent Breast Cancer Mice
Model

   To investigate the antitumor mechanism of black phosphorus (BP), we
   first synthesized BP and modified it with polyethylene glycol (PEG) to
   enhance its stability, given BP's inherent instability. Subsequently,
   we characterized BP‐PEG. Transmission electron microscopy (TEM) images
   revealed typical 2D sheet‐like structures for both BP and BP‐PEG
   (Figure [73]1A). Most BP‐PEG nanosheets were less than 200 nm in size,
   predominantly around 100 nm, which could ensure the stability and
   specific tissue targeting of BP‐PEG (Figure [74]1B). The zeta potential
   decreased from −27.6 to −18.1 mV after PEG modification
   (Figure [75]1C), indicating successful PEG conjugation and enhanced
   stability of BP. Energy‐dispersive X‐ray spectroscopy (EDS) analysis
   confirmed the presence of C, N, O, and P elements in BP‐PEG (Figure
   [76]S1A, Supporting Information). Fourier‐transform infrared (FTIR)
   spectra showed characteristic peaks corresponding to P═O at 1637 cm⁻¹
   and enhanced peaks at 2924 cm⁻¹ due to C─H vibrations from PEG chains
   (Figure [77]S1B, Supporting Information). X‐ray diffraction (XRD)
   patterns and Raman spectra confirmed the successful exfoliation of BP
   into multilayer nanosheets (Figure [78]S1C,D, Supporting Information).
   UV–vis spectroscopy showed a positive correlation between BP
   concentration and absorbance (Figure [79]S1E, Supporting Information).
   Additionally, atomic force microscopy (AFM) images showed that the
   average size and thickness of PEGylated BP nanosheets (BPNSs) slightly
   increased due to the PEG‐NH2 coating on the surface (Figure [80]S1F,
   Supporting Information). Sectional plots revealed that the thickness of
   BPNSs ranged from 5 to 10 nm, while BPNS‐PEG thickness ranged from 10
   to 20 nm (Figure [81]S1F, Supporting Information). Together, these
   findings confirm the successful synthesis of BP‐PEG nanosheets with
   improved physiological stability.

Figure 1.

   Figure 1
   [82]Open in a new tab

   Characterization of BP‐PEG nanosheets and their antitumor activity in
   an immunocompetent breast cancer mouse model. A) Transmission electron
   microscopy (TEM) images of black phosphorus (BP) and polyethylene
   glycol‐modified black phosphorus (BP‐PEG) nanosheets, Scale bar:
   200 nm. B) SEM image and particle size distribution of BP‐PEG, Scale
   bar, 200 µm and 200 nm, respectively. C) Surface charge of BP and
   BP‐PEG. D) Schematic diagram of the experimental design. 4T1 breast
   cancer cells were injected into the mammary fat pad of immunocompetent
   BALB/c mice to establish a syngeneic tumor model. E) BP‐PEG treatment
   significantly inhibits tumor growth in 4T1 tumor‐bearing BALB/c mice.
   Left: Tumor growth curves showing tumor volume over time for
   BP‐PEG‐treated and control groups (n = 6 per group). Right: Final tumor
   weights at the end of the study. F) Schematic diagram of the
   experimental design. BP‐PEG treatment in MMTV‐PyMT transgenic mice, a
   model that closely mimics human breast cancer progression. G) Tumor
   growth and weights of MMTV‐PyMT transgenic mice after BP‐PEG treatment
   (n = 5 per group). Data are presented as mean ± SD; p‐values were
   determined by two‐tailed unpaired t‐test; **p < 0.01, ***p < 0.001.

   We then aimed to explore whether BP‐PEG possesses antitumor activity in
   breast cancer, as several studies have demonstrated the promise of
   BP‐based approaches in clinical breast cancer therapy. To this end, we
   injected 4T1 cells into the fat pad of immunocompetent BALB/C mice to
   establish a syngeneic model, which is widely used in preclinical cancer
   research, particularly for studying triple‐negative breast cancer
   (TNBC), as it replicates many aspects of human TNBC (Figure [83]1D).
   Once the average tumor volume reached ≈100 mm^3, BP‐PEG was
   administered intravenously, as outlined in the experimental design
   (Figure [84]1D). BP‐PEG treatment significantly inhibited the growth of
   4T1 breast tumors compared to the control group (Figure [85]1E), and
   tumor weights were also reduced (Figure [86]1E). Histological analysis
   of major organs showed no obvious toxicity (Figure [87]S1G, Supporting
   Information).

   To determine whether the observed reduction in tumor growth was due to
   BP alone and not PEG, we treated 4T1 tumor‐bearing BALB/c mice with
   saline, 200 µL PEG, 10 mg kg^−1 BP, or 10 mg kg^−1 BP‐PEG. The results
   indicated that PEG administration alone had no significant effect on
   tumor growth when compared to the saline‐treated control group (Figure
   [88]S1H, Supporting Information). In contrast, BP alone was able to
   reduce tumor growth, with an effect comparable to that seen with BP‐PEG
   treatment at the same concentration (Figure [89]S1H, Supporting
   Information). Additionally, we tested whether reducing the BP‐PEG dose
   from 10 to 5 mg kg^−1 would still impact tumor growth. While 5 mg kg^−1
   BP‐PEG also reduced tumor growth, its efficacy was lower than that of
   the higher dose (Figure [90]S1H, Supporting Information). Based on
   these findings, we conclude that 10 mg kg^−1 BP‐PEG is more suitable
   for evaluating antitumor effects in this model. Additionally, BP‐PEG
   treatment markedly suppressed lung metastasis of 4T1 cells (Figure
   [91]S1I, Supporting Information).

   To further confirm the inhibitory effect of BP‐PEG on tumor growth in
   an immunocompetent breast cancer model, we treated MMTV‐PyMT transgenic
   mice, a model that closely mimics human breast cancer progression, with
   BP‐PEG. Similar antitumor effects were observed in MMTV‐PyMT transgenic
   mice (Figure [92]1F,G). Altogether, these results indicate that BP‐PEG
   effectively inhibits breast cancer development and metastasis in models
   that closely mimic human breast cancer.

2.2. BP‐PEG Antitumor Growth Dependent on the Immune Microenvironment

   To investigate the antitumor mechanism of BP‐PEG, we first examined
   whether BP‐PEG exerts its function by directly killing tumor cells. We
   constructed FITC‐labeled BP‐PEG and confirmed that BP‐PEG primarily
   accumulated in 4T1‐derived tumor tissues in vivo using an animal
   imaging system (Figure [93]S2A, Supporting Information). Electron
   microscopy also confirmed that BP‐PEG could be taken up by 4T1 tumor
   cells in vitro (Figure [94]S2B, Supporting Information).

   Next, we used flow cytometry to quantify the amount of BP‐PEG
   internalized by tumor cells in vivo and compared it to the uptake
   observed in vitro (Figure [95]2A). By dissociating tumor tissue to
   obtain single cells and labeling tumor cells with the tumor marker
   EPCAM, we measured the amount of BP‐PEG internalized by tumor cells in
   vivo 24 h after BP‐PEG injection via tail vein (Figure [96]2B). We then
   compared the intensity of FITC internalized by tumor cells in vivo with
   that of 4T1 cells treated with varying concentrations of FITC‐BP‐PEG in
   vitro (Figure [97]2C). The results showed that the majority of tumor
   cells in vivo had similar FITC intensity as 4T1 cells treated with
   5 µg mL^−1 FITC‐BP‐PEG in vitro (Figure [98]2C). These data
   demonstrated that, when treated with 10 mg kg^−1 FITC‐BP‐PEG, tumor
   cells in vivo internalized an amount of BP‐PEG comparable to 4T1 cells
   treated with 5 µg mL^−1 FITC‐BP‐PEG in vitro. We found that, at this
   concentration, BP‐PEG did not significantly affect tumor cell
   proliferation or survival over 72 h (Figure [99]2D), suggesting that
   BP‐PEG does not exert direct cytotoxic effects on tumor cells at
   physiologically relevant concentrations.

Figure 2.

   Figure 2
   [100]Open in a new tab

   BP‐PEG does not directly kill tumor cells but requires an intact immune
   system for antitumor activity. A) Schematic diagram illustrating the
   comparison of BP‐PEG internalization by tumor cells in vivo and in
   vitro using flow cytometry. B) Flow cytometry gating strategy to
   identify EPCAM⁺ tumor cells from dissociated tumor tissues and measure
   FITC‐BP‐PEG internalization 24 h post‐injection. C) Comparison of FITC
   fluorescence intensity in tumor cells treated in vivo with BP‐PEG (10
   mg kg^−1) and 4T1 cells treated in vitro with varying concentrations of
   FITC‐BP‐PEG (0, 2.5, 5, 10 µg mL^−1). The in vivo internalization
   corresponds to ≈5 µg mL^−1 in vitro treatment. D) Cell viability assay
   of 4T1 cells treated with BP‐PEG at concentrations up to 100 µg mL^−1
   for 72 h, showing no significant cytotoxicity. Data are presented as
   mean ± SD, p‐values were determined by two‐tailed Student's t‐test (ns,
   not significant; **p < 0.01). E) Tumor growth in immunodeficient NSG
   mice bearing 4T1 tumors. Left: Tumor growth curves over time (n = 7 per
   group). Right: Final tumor weights. Data are presented as mean ± SD,
   ns, not significant. F) Schematic of the co‐culture assay where human
   MDA‐MB‐231 breast cancer cells were co‐cultured with human PBMCs in the
   presence or absence of BP‐PEG. G) Representative images of tumor cells
   after co‐culture and stained with crystal violet. H) Quantification of
   tumor cell viability after co‐culture. Data are presented as mean ± SD;
   p‐values were determined by two‐tailed unpaired t‐test; ***p < 0.001.

   Since BP‐PEG does not directly kill tumor cells, we propose that it may
   inhibit tumor growth in vivo by modulating the tumor microenvironment,
   particularly the immune environment. To explore whether BP‐PEG exerts
   its function through the immune system, we utilized immunodeficient NOD
   scid gamma (NSG) mice bearing 4T1 tumors to evaluate the effects of
   BP‐PEG (Figure [101]2E). The experimental procedures were consistent
   with those performed in BALB/c mice (Figure [102]1D). We found that
   BP‐PEG treatment failed to inhibit tumor growth and metastasis in
   immunodeficient NSG mice bearing 4T1 tumors (Figures [103]2E and
   [104]S2C, Supporting Information), indicating that an intact immune
   system is required for BP‐PEG's antitumor activity.

   To further demonstrate that BP‐PEG exerts its antitumor effects
   primarily through the immune system in the tumor microenvironment, we
   performed a co‐culture assay. We collected blood from breast cancer
   patients, isolated peripheral blood mononuclear cells (PBMCs), and
   co‐cultured MDA‐MB‐231 cells with PBMCs in the presence of BP‐PEG or
   saline. Flow cytometry was used to analyze the PBMC cell components,
   revealing a composition similar to that of immune cells found within
   the tumor microenvironment (Figure [105]S2D, Supporting Information).
   We then quantified the number of remaining tumor cells after co‐culture
   using microscopy and violet staining (Figure [106]2F). The data showed
   that only minimal tumor cell loss occurred in the BP‐PEG or PBMC
   treatment alone groups, while >90% of tumor cells were eliminated when
   BP‐PEG and PBMCs were combined (Figure [107]2G,H). These findings
   further suggest that BP‐PEG modulates the immune microenvironment to
   exert its antitumor effects.

2.3. BP‐PEG Specifically Activates Tumor Immune Microenvironments Reshaping
Pathway in Immunocompetent Mice

   Then, we wanted to understand how BP‐PEG exerts antitumor effects in
   the presence of an intact immune system. To identify the pathways or
   genes specifically regulated under BP‐PEG treatment, we collected tumor
   tissues from both BALB/c and NSG mice and performed bulk tissue RNA
   sequencing for differentially expressed genes (DEGs) analysis (Figure
   [108]3A). Our results demonstrated that BP‐PEG altered the gene
   expression profiles of tumors in both mouse models (Figure [109]S3A,B,
   Supporting Information). Interestingly, we found that the DEGs in NSG
   mice tumors overlapped with those in BALB/c mice tumors by only 17
   genes, despite the total number of DEGs exceeding 200. This finding
   suggests that the unique transcription profile in BALB/c mice may be
   responsible for the antitumor effect of BP‐PEG. Specifically, we
   identified 234 genes that were differentially expressed only in BALB/c
   mice (Figure [110]3B). To understand which pathways are primarily
   involved in the regulation of BP‐PEG's effects, we conducted pathway
   enrichment analysis (Figure [111]3C). The results indicated a
   significant upregulation of immune‐related pathways, particularly
   cytokine–cytokine receptor interactions (Figure [112]3C), which aligns
   with the notion that BP‐PEG's antitumor function depends on an intact
   immune system.

Figure 3.

   Figure 3
   [113]Open in a new tab

   BP‐PEG specifically activates immune regulatory pathways in
   immunocompetent mice. A) Schematic of bulk RNA sequencing performed on
   tumor tissues from BALB/c and NSG mice treated with BP‐PEG or saline.
   B) Venn diagram showing the number of differentially expressed genes
   (DEGs) in tumors from BALB/c and NSG mice. C) Pathway enrichment
   analysis of the 234 DEGs uniquely regulated in BALB/c mice tumors (B),
   highlighting immune‐related pathways such as cytokine–cytokine receptor
   interaction. D) Heatmap of the DEGs in BALB/c mice tumors involved in
   immune microenvironment remodeling after BP‐PEG treatment. E) Schematic
   diagram illustrating the DEGs’ main functions in tumor
   microenvironment. F) Quantitative PCR validation of the DEGs showed in
   (D). Data are presented as mean ± SD; p‐values were determined by
   two‐tailed unpaired t‐test; *p < 0.05, ***p < 0.001.

   Next, we sought to determine which genes were involved in these
   pathways and their levels of change. Therefore, we conducted a heatmap
   analysis of the DEGs involved in immune microenvironment remodeling
   (Figure [114]3D,E). The analysis revealed that most of these genes were
   upregulated in the BP‐PEG treatment group and contributed to reshape
   the tumor immune microenvironment, including genes such as IL1A and
   CXCL2 (Figure [115]3D,E). We further confirmed the upregulation or
   downregulation of these genes by BP‐PEG using qPCR (Figure [116]3F). In
   conclusion, our data suggest that BP‐PEG exerts its antitumor effects
   by specifically activating immune regulatory pathways, thereby
   reshaping the tumor immune microenvironment.

2.4. BP‐PEG Enhances Neutrophil Recruitment to the Tumor Microenvironment

   We next sought to determine whether BP‐PEG treatment alters the tumor
   immune microenvironment. Mass cytometry, an advanced technique capable
   of simultaneously detecting over 40 markers, was employed to provide a
   comprehensive analysis of the immune cell populations within the tumor
   microenvironment. Tumor samples were collected from BALB/C mice,
   dissociated into single cells, and subjected to mass cytometry (Figure
   [117]4A). This analysis identified 26 distinct cell clusters using
   t‐distributed stochastic neighbor embedding (t‐SNE) (Figure [118]4A).
   The clusters were characterized based on marker expression
   (Figures [119]4A and [120]S4A, Supporting Information), and key markers
   were visualized in t‐SNE plots (Figure [121]4B). We identified
   neutrophils, macrophages, dendritic cells, B cells, CD8^+ T cells,
   CD4^+ T cells, helper T cells, regulatory T cells (Tregs), and NK cell
   clusters (Figures [122]4A and [123]S4A, Supporting Information).

Figure 4.

   Figure 4
   [124]Open in a new tab

   BP‐PEG enhances neutrophil recruitment and activates antitumor immune
   cells. A) Mass cytometry (CyTOF) analysis of tumor‐infiltrating immune
   cells from BALB/c mice treated with BP‐PEG or saline. t‐SNE plots
   display 26 distinct cell clusters. B) t‐SNE visualization of key immune
   cell defined markers. C) Comparison of immune cell population
   distributions between saline and BP‐PEG‐treated groups in t‐SNE plots.
   D) Quantification of immune cell clusters, showing a significant
   increase in cluster 18 (neutrophils) after BP‐PEG treatment. E) Bar
   graph illustrating the percentage of different neutrophils clusters
   among CD45⁺ cells. F,G) The expression intensity of granzyme B (GZMB)
   in CD8⁺ T cells (F) and NK cells (G). H) Immunohistochemical staining
   of tumor sections for Ly6G and GZMB. Left: Representative images
   showing Ly6G and GZMB‐positive cells; Right: Quantification of the
   relative percentage of Ly6G⁺ and GZMB⁺ cells per field. Data are
   presented as mean ± SD; p‐values were determined by two‐tailed unpaired
   t‐test; ns, not significant; *p < 0.05, ***p < 0.001.

   We next compared the composition of immune cell populations between
   saline and BP‐PEG‐treated groups (Figure [125]4C). The t‐SNE analysis
   revealed that cell cluster 18, a subpopulation of neutrophils,
   exhibited the most significant increase in both number and percentage
   following BP‐PEG treatment (Figure [126]4C–E). Posttreatment, this
   neutrophil subpopulation accounted for over 20% of CD45^+ cells,
   emerging as the dominant cell population (Figure [127]4D,E).
   Additionally, we analyzed the proportions of other immune cell types
   and observed a reduction in immunosuppressive cells, specifically
   regulatory T cells (Tregs) and type M2 macrophages, following BP‐PEG
   treatment, while other cell populations showed no significant change
   (Figure [128]S4B–F, Supporting Information). Further, we assessed the
   antitumor activity of two key immune cell types‐cytotoxic T lymphocytes
   (CTLs) and natural killer (NK) cells‐by examining the expression of the
   cytotoxic marker granzyme B (GZMB). The data revealed a substantial
   upregulation of GZMB intensity in the predominant subpopulations of
   CD8^+ T cells and NK cells (Figure [129]4F,G), indicating that BP‐PEG
   treatment enhances the antitumor activity of these immune cells.

   To further confirm the increase in neutrophil numbers, we performed
   immunohistochemical staining of tumor tissues using the neutrophil
   marker Ly6G (Figure [130]4H). Quantification of neutrophil ratios in
   the tissue confirmed a significant increase in both the number and
   percentage of neutrophils following BP‐PEG treatment (Figure [131]4H).
   In addition, we assessed the antitumor activity of immune cells in
   continuous tissue sections, which had been stained for Ly6G. The
   results showed a significant increase in GZMB‐positive cells,
   indicating enhanced antitumor functionality of immune cell after BP‐PEG
   treatment (Figure [132]4H). In summary, our findings suggest that
   BP‐PEG promotes neutrophil recruitment, contributing to the enhanced
   antitumor activity of effector cells, such as CD8^+ T cells and NK
   cells.

2.5. Neutrophils Are Essential for BP‐PEG's Antitumor Activity

   Given the significant infiltration of neutrophils into tumor tissues
   following BP‐PEG treatment, we sought to determine whether neutrophils
   are required for BP‐PEG's antitumor effects. To test this hypothesis,
   we conducted neutrophil depletion experiments using an anti‐Ly6G
   antibody in 4T1 tumor‐bearing mice, with a schematic diagram outlining
   the experimental design (Figure [133]5A). Starting three days after 4T1
   tumor cell inoculation, mice were injected daily with the anti‐Ly6G
   antibody. Once tumors reached a volume of approximately 100 mm^3, the
   mice were randomized into four treatment groups: 1) Saline, 2) BP‐PEG,
   3) anti‐Ly6G, or 4) a combination of BP‐PEG and anti‐Ly6G
   (Figure [134]5A,B). Consistent with previous studies,^[ [135]^29 ^]
   anti‐Ly6G treatment alone had minimal impact on tumor growth and weight
   (Figure [136]5B,C). While BP‐PEG effectively inhibited tumor growth and
   reduced tumor weight, this effect was largely reversed in the mice
   treated with the combination of BP‐PEG and anti‐Ly6G
   (Figure [137]5B,C). Additionally, while BP‐PEG treatment markedly
   suppressed lung metastasis of 4T1 cells, neutrophil depletion
   upregulated lung metastasis (Figure [138]S5A, Supporting Information).

Figure 5.

   Figure 5
   [139]Open in a new tab

   Neutrophils are essential for BP‐PEG's antitumor effects. A) Schematic
   of the experimental design for neutrophil depletion using anti‐Ly6G
   antibody in 4T1 tumor‐bearing BALB/c mice. B) Representative tumors
   from 4T1 cells‐bearing BALB/c mice treated with saline, BP‐PEG,
   anti‐Ly6G, or BP‐PEG plus anti‐Ly6G. C) Tumor growth curves and the
   tumor weights of the 4T1 tumors. D) Immunohistochemical staining for
   Ly6G and GZMB in tumor sections. Left: Representative images showing
   successful neutrophil depletion and GZMB staining, right:
   Quantification of the relative percentage of Ly6G⁺ and GZMB⁺ cells per
   fields. E) Flow cytometry was employed to analyze the percentage of
   neutrophils within the CD45^+ cell population and GZMB^+ cells among
   lymphocytes isolated from 4T1‐derived tumors. The results were
   subsequently quantified. Data are presented as mean ± SD; p‐values were
   determined by two‐tailed unpaired t‐test; ns, not significant; **p <
   0.01, ***p < 0.001.

   We confirmed the efficiency of neutrophil depletion using
   immunohistochemistry (IHC) staining and flow cytometry, which showed a
   near‐complete absence of Ly6G‐positive cells in the anti‐Ly6G‐treated
   group, confirming successful neutrophil depletion (Figure [140]5D,E).
   Furthermore, immunohistochemical staining and flow cytometry for
   granzyme B showed no substantial difference in the percentage of
   GZMB‐positive cells between the control and neutrophil‐depleted groups.
   However, the percentage of GZMB‐positive cells increased with BP‐PEG
   treatment but decreased following neutrophil depletion
   (Figure [141]5D,E), indicating a reduction in antitumor immune
   responses associated with neutrophil depletion. Altogether, these
   findings underscore the critical role of neutrophils in mediating
   BP‐PEG‐induced antitumor immunity.

2.6. BP‐PEG Enhances Neutrophil Recruitment Through Macrophage‐Mediated
Inflammatory Signaling Pathways

   To further understand how BP‐PEG treatment regulates neutrophil
   infiltration in tumor tissues, we collected tumors from BALB/c mice
   treated with either saline or BP‐PEG. CD45‐positive cells were isolated
   for single‐cell RNA sequencing to identify immune cell populations and
   analyze their gene expression profiles (Figure [142]6A). UMAP
   clustering of all sequenced cells revealed a notable enrichment of
   several immune cell subsets‐including neutrophils, macrophages, and NK
   cells‐following BP‐PEG treatment (Figures [143]6B,C and [144]S6A,
   Supporting Information). Then using the scRNA‐seq data, we explored the
   immune cell interactions that may contribute to enhanced neutrophil
   infiltration and a productive antitumor immune response. Cell–cell
   interaction analysis provides insights into interactions enriched in
   BP‐PEG‐treated mice, helping to elucidate the mechanism by which BP‐PEG
   promotes neutrophil infiltration and a productive antitumor immune
   response. Cells with high incoming interaction strength might be key
   responders, while those with high outgoing interaction strength could
   be major communicators in the system. The results showed that under
   saline treatment, CD8⁺ T cells and macrophage‐like cells (Mph‐C1)
   exhibited high incoming interaction strength (Figure [145]6D),
   suggesting they are primary targets or responders in the interaction
   network. B cells and macrophage‐like cells (Mph‐C2) had high outgoing
   interaction strength, indicating they could be major communicators in
   the system (Figure [146]6D). Under BP‐PEG treatment, although CD8⁺ T
   cells and Mph‐C1 still displayed high incoming interaction strength,
   Mph‐C2 and neutrophils became the cells with major outgoing interaction
   strength (Figure [147]6D), indicating they emerged as significant
   communicators in the system, especially macrophages.

Figure 6.

   Figure 6
   [148]Open in a new tab

   BP‐PEG promotes neutrophil recruitment via macrophage‐mediated
   inflammatory signaling. A) Workflow for single‐cell RNA sequencing
   (scRNA‐seq) of CD45⁺ immune cells isolated from tumors of BALB/c mice,
   comparing those treated with BP‐PEG and saline. B) UMAP plots showing
   the clustering of immune cells based on treatment conditions. C) Bar
   graph quantifying the proportions of each immune cell type in tumors
   treated with saline versus BP‐PEG. D) Analysis of cell–cell
   interactions, highlighting the strengths of incoming and outgoing
   interactions among immune cells under saline and BP‐PEG treatments. E)
   Quantification of mean fluorescence intensity of BP‐PEG‐FITC in myeloid
   cells within tumor tissues, Data are presented as mean ± SD; *p < 0.05,
   **p < 0.01. F) Gene Ontology (GO) pathway enrichment analysis of
   macrophages, based on upregulated genes observed in BP‐PEG‐treated
   samples compared to saline‐treated samples. G) Cell–cell communication
   networks through the CXCL–CCR signaling pathway between macrophages and
   neutrophils, using upregulated genes from BP‐PEG treatment relative to
   saline treatment. H) A schematic diagram of the experimental design is
   provided. Saline, SB225002 (10 mg kg^−1), and BP‐PEG (10 mg kg^−1) were
   administered to 4T1 tumor‐bearing BALB/c mice in separate treatment
   groups (n = 5 per group). Tumor images were captured, and tumor weights
   were analyzed at the end of the study. Data are presented as mean ± SD;
   p‐values were determined by two‐tailed unpaired t‐test; ns, not
   significant; ***p < 0.001.

   Given that macrophages have high phagocytic function, we propose that
   when BP‐PEG is injected into mice, it is first engulfed by macrophages,
   which increases the inflammatory response in these cells. This response
   leads to the production of various cytokines that promote neutrophil
   infiltration and reshape the antitumor environment. To test this
   hypothesis, we detected the distribution of BP‐PEG‐FITC in tumors using
   flow cytometry (Figure [149]S6B, Supporting Information). The results
   showed that macrophages indeed contained the highest level of
   BP‐PEG‐FITC among myeloid cells (Figure [150]6E), which is consistent
   with their primary role in phagocytosis. Altogether, we identified
   macrophages as likely initiators of interactions among immune cells in
   the tumor microenvironment (TME), based on the high predicted outgoing
   interaction strength under BP‐PEG treatment.

   We then analyzed functional changes in immune cells using pathway
   enrichment analysis. In macrophages, genes involved in neutrophil
   migration were upregulated under BP‐PEG treatment, along with genes
   related to the regulation of neutrophil degranulation, innate immune
   response, and positive regulation of the inflammatory response
   (Figure [151]6F). These results demonstrate that BP‐PEG uptake by
   macrophages indeed promotes inflammatory signaling pathways, which in
   turn promote neutrophil migration and function. In bulk tissue RNA
   sequencing, we found that cytokines related to neutrophils‐such as
   IL1A, CXCL1, CXCL2, and CSF3‐were upregulated in BP‐PEG‐treated tumors.
   According to scRNA‐seq data, the inflammatory related pathway was also
   upregulated, and CXCL2 was also found to be upregulated in mixed CD45
   positive cells and was predominantly expressed by macrophages and
   neutrophils, indicating that CXCL2 may contribute to neutrophil
   infiltration under BP‐PEG treatment (Figure [152]S6C, Supporting
   Information). Additionally, using ELISA to detect CXCL2 expression in
   THP‐1 cells, we confirmed that CXCL2 is produced by macrophages and
   upregulated by BP‐PEG (Figure [153]S6D, Supporting Information).

   We next analyzed the CXCL‐CCR signaling pathways between macrophages
   and neutrophils. The results showed that CXCL2‐CXCR2, CCL9‐CCR1,
   CCL6‐CCR1, and CCL3‐CCR1 signaling pathways were enhanced under BP‐PEG
   treatment (Figure [154]6G). Notably, the communication probability
   within the CXCL2–CXCR2 signaling pathway was significantly upregulated,
   with CXCL2 expression elevated under BP‐PEG treatment. To further
   investigate the role of this pathway, we analyzed the expression of
   CXCL2 receptors, CXCR1 and CXCR2, in the tumor microenvironment using
   scRNA‐seq data. Our analysis revealed that CXCR2 was highly expressed,
   predominantly on neutrophils (Figure [155]S6E, Supporting Information),
   suggesting that CXCL2 primarily exerts its function through interaction
   with CXCR2 on neutrophils. To further investigate whether the
   CXCL2–CXCR2 signaling pathway regulates the antitumor effects of
   BP‐PEG, we treated mice with the CXCR2 inhibitor SB225002
   (Figure [156]6H). The results showed that SB225002 alone had no impact
   on tumor growth; however, when mice were treated with both BP‐PEG and
   SB225002, the antitumor effect of BP‐PEG was almost completely
   abolished (Figure [157]6H). These findings demonstrate that the
   antitumor effects of BP‐PEG are primarily dependent on the CXCL2–CXCR2
   signaling pathway. Altogether, our data suggest that macrophages may
   promote neutrophil infiltration under BP‐PEG treatment through the
   CXCL–CCR signaling pathway, with a particularly critical role for
   CXCL2–CXCR2.

2.7. BP‐PEG Enhances Tumor‐Killing Ability of Neutrophils and NK Cells

   We then sought to understand how neutrophils regulate antitumor
   activity. Therefore, we analyzed the differential number of
   interactions and interaction strengths between neutrophils and key
   antitumor immune cells, specifically CD8⁺ T cells and NK cells (Figure
   [158]7A). The data revealed that the number of interactions between
   neutrophils and NK cells was higher than that between neutrophils and
   CD8⁺ T cells or other cells (Figure [159]7A). Moreover, the interaction
   strength between neutrophils and NK cells was also notably stronger
   (Figure [160]7A). Although interactions between neutrophils and CD8⁺ T
   cells were also upregulated under BP‐PEG treatment (Figure [161]7A),
   the potential ligand–receptor interactions between these cell types
   were fewer and weaker compared to those between neutrophils and NK
   cells (Figure [162]S7A, Supporting Information). Immunofluorescence
   analysis further demonstrated close spatial localization of neutrophils
   and NK cells under BP‐PEG treatment (Figure [163]S7B, Supporting
   Information). These findings suggest that neutrophils primarily
   regulate the antitumor functions of NK cells.

Figure 7.

   Figure 7
   [164]Open in a new tab

   BP‐PEG enhances the tumor‐killing ability of neutrophils and NK Cells.
   A) Differential number of interactions or interaction strength,
   visualized as a heatmap, where red increased and blue represents
   decreased signaling in BP‐PEG compared to saline treated tumors. The
   top‐colored bar plot represents the sum of a column of values displayed
   (incoming signaling). The right‐ colored bar plot represents the sum of
   a row of values (outgoing signaling). B) Gene Ontology (GO) pathway
   enrichment analysis of neutrophils, based on upregulated genes observed
   in BP‐PEG‐treated samples compared to saline‐treated samples. C)
   Analysis of the T2 gene signature expression score in neutrophils. D)
   Gene Ontology (GO) pathway enrichment analysis of NK cells, based on
   upregulated genes observed in BP‐PEG‐treated samples compared to
   saline‐treated samples. E) Experimental setup for assessing the
   tumor‐killing ability of neutrophils and NK cells isolated from tumors.
   Immune cells were co‐cultured with GFP‐labeled 4T1 cells. F)
   Quantification of GFP‐positive tumor cells after co‐culture. Data are
   presented as mean ± SD; p‐values were determined by two‐tailed unpaired
   t‐test; ns, not significant; **p < 0.01, ***p < 0.001.

   We then analyzed functional changes in neutrophils and NK cells using
   pathway enrichment analysis. In neutrophils, pathway enrichment
   analysis revealed upregulation of pathways involved in the innate
   immune response, neutrophil degranulation, myeloid leukocyte migration,
   and oxidative stress and redox pathways under BP‐PEG treatment
   (Figure [165]7B). These results indicate that neutrophils are involved
   in regulating cell migration—including their own—and that their
   antitumor function was enhanced under BP‐PEG treatment, as evidenced by
   the upregulation of genes involved in neutrophil degranulation and
   oxidative stress and redox pathways. Additionally, recent studies^[
   [166]^30 ^] have identified three types of neutrophils in the tumor
   microenvironment—T1, T2, and T3—where T1 and T2 exert antitumor
   functions and T3 exhibits protumor activity. To further explore this in
   our model, we analyzed the neutrophil subtypes and found that the gene
   signature profiles defining these neutrophil types were also expressed
   in the neutrophils observed in our study (Figure [167]S7C, Supporting
   Information). Notably, T1 and T2 neutrophils were upregulated under
   BP‐PEG treatment, while T3 neutrophils showed no significant change
   (Figures [168]7C and [169]S7C, Supporting Information). Altogether,
   these findings demonstrate that BP‐PEG treatment can reprogram
   neutrophils to enhance their antitumor function.

   Finally, we analyzed pathway enrichment in NK cells. Functions related
   to granzyme‐mediated programmed cell death were enhanced under BP‐PEG
   treatment, indicating that antitumor functions were indeed increased
   under BP‐PEG treatment (Figure [170]7D). To confirm that BP‐PEG could
   enhance the tumor‐killing ability of neutrophils and NK cells, we
   isolated these cells from tumors treated with saline and BP‐PEG using
   MACS and then co‐cultured them with GFP‐labeled 4T1 cells
   (Figure [171]7E). After 24 h, we counted the number of GFP‐positive
   cells under a microscope. The number of cells per field was analyzed,
   and the data showed that BP‐PEG indeed enhanced the tumor‐killing
   ability of neutrophils and NK cells (Figure [172]7F).

   To further demonstrate that BP‐PEG's antitumor effects are dependent on
   NK cells in vivo, we performed an NK cell depletion assay (Figure
   [173]S7D,E, Supporting Information). The results showed that depletion
   of NK cells partially reversed the antitumor effects of BP‐PEG, as
   tumor weights increased in BP‐PEG‐treated mice lacking NK cells.
   However, the tumor weights in these mice remained lower than those in
   the saline‐treated control group (Figure [174]S7D,E, Supporting
   Information). Altogether, these findings demonstrate that BP‐PEG
   enhances the tumor‐killing abilities of NK cells and contributes to its
   antitumor efficacy.

3. Discussion

   This study demonstrates that black phosphorus‐polyethylene glycol
   (BP‐PEG) nanosheets exert significant antitumor effects in breast
   cancer models by modulating the tumor immune microenvironment (TME),
   specifically by promoting neutrophil recruitment and activation. Unlike
   conventional chemotherapeutics that directly target tumor cells, BP‐PEG
   enhances innate immune responses, leading to inhibition of tumor growth
   and suppression of metastasis. Importantly, BP‐PEG does not directly
   induce cytotoxicity in tumor cells at physiologically relevant
   concentrations; instead, its antitumor efficacy is mediated through
   immune modulation. This is evidenced by the lack of therapeutic effect
   in immunodeficient NSG mice and the reversal of antitumor activity upon
   neutrophil depletion in immunocompetent mice, underscoring the pivotal
   role of neutrophils in BP‐PEG‐mediated tumor suppression.
   Transcriptomic analyses revealed upregulation of cytokines and
   chemokines such as IL‐1α and CXCL2, which are crucial for neutrophil
   recruitment and activation. Furthermore, mass cytometry and single‐cell
   RNA sequencing confirmed enhanced neutrophil infiltration and
   functional activation. Thus, BP‐PEG reprograms the TME from
   immunosuppressive to immunostimulatory, promoting antitumor immunity.
   Overall, our study demonstrates that neutrophils exhibit remarkable
   plasticity and can acquire an antitumor phenotype in response to BP‐PEG
   therapy.

   Neutrophils have a dual function in tumor immunity that depends on both
   context and specific subsets. Historically, they have been considered
   protumorigenic because they promote angiogenesis, facilitate
   metastasis, and suppress adaptive immunity.^[ [175]^31 ^] Specifically,
   type 2 tumor‐associated neutrophils have been linked to oncogenesis,
   tumor progression, metastasis, and the suppression of adaptive immune
   responses.^[ [176]^23 ^] An increased neutrophil‐to‐lymphocyte ratio is
   viewed as a strong predictor of poor prognosis in various cancer types
   and is associated with reduced effectiveness of immunotherapies.^[
   [177]^32 ^]

   However, neutrophils are a heterogeneous population, and under certain
   conditions, they can exert antitumor effects.^[ [178]^33 , [179]^34 ^]
   Under specific conditions of acute inflammation or in response to
   targeted therapeutic interventions, neutrophils can adopt
   antitumorigenic phenotypes. For instance, they can enhance antigen
   presentation and co‐stimulation of T cells, thereby initiating adaptive
   immune responses.^[ [180]^35 ^] Studies have demonstrated that
   alleviating hypoxia can increase neutrophil infiltration into tumors,
   leading to tumor cell killing through mechanisms involving NADPH
   oxidase, reactive oxygen species (ROS), and matrix metalloproteinase‐9
   (MMP‐9).^[ [181]^30 ^] And neutrophil elastase, which induces apoptosis
   in tumor cells by cleaving the death domain of CD95.^[ [182]^36 ^]
   Therefore, the role of neutrophils in cancer is complex and highly
   dependent on the context.

   Our study demonstrated that BP‐PEG nanosheets appear to polarize
   neutrophils toward an antitumor phenotype, characterized by increased
   degranulation, production of ROS, and expression of cytotoxic
   mediators. Upon UMAP and clustering of scRNA‐seq data of tumors from
   treated mice, we found the neutrophils showed increased in the innate
   immune response, neutrophil degranulation, myeloid leukocyte migration,
   and oxidative stress and redox pathways, suggesting that highly
   differentiated neutrophils adopt an antitumor phenotype. And the
   interaction between macrophages and neutrophils is critical for the
   immunomodulatory effects of BP‐PEG (Figure [183]8 ). Uptake of BP‐PEG
   by macrophages activates inflammatory signaling pathways, leading to
   the secretion of chemokines that recruit neutrophils to the tumor site
   (Figure [184]8). This creates a positive feedback loop that amplifies
   the antitumor immune response. Enhanced interactions among macrophages,
   neutrophils, natural killer (NK) cells, and CD8⁺ T cells emphasize the
   importance of immune cell cross‐talk in BP‐PEG‐mediated tumor
   suppression (Figure [185]8).

Figure 8.

   Figure 8
   [186]Open in a new tab

   Proposed working model (Created in BioRender. Yu, F. (2024), Agreement
   number: TE27PIVUES).

   Much of our current understanding of neutrophils' impact on tumor
   growth and therapy response derives from animal studies where
   neutrophils are broadly depleted during treatment.^[ [187]^37 ^]
   However, given the emerging evidence of diverse neutrophil subsets with
   varying functions within tumors, indiscriminate depletion strategies
   may eliminate both pro‐tumorigenic and antitumorigenic neutrophil
   populations.^[ [188]^25 ^] From a therapeutic standpoint, reprogramming
   tumor‐associated neutrophils toward an antitumor phenotype may be more
   effective for tumor control than complete neutrophil depletion.^[
   [189]^25 ^] Therefore, gaining a deeper understanding of how BP‐PEG
   influences the acquisition of distinct neutrophil states would allow
   for selective manipulation of neutrophil subpopulations. This, in turn,
   would enable us to develop a more nuanced understanding of the role of
   neutrophils in immunotherapy.

   The immunomodulatory properties of BP‐PEG nanosheets offer potential
   for integration into existing cancer immunotherapies. For example,
   combining BP‐PEG with immune checkpoint inhibitors may potentiate
   T‐cell‐mediated antitumor responses by alleviating immunosuppression
   within the TME.^[ [190]^38 , [191]^39 ^] Furthermore, BP‐PEG's ability
   to harness innate immune responses, particularly through
   neutrophil‐mediated mechanisms, offers a complementary strategy to
   current therapies. This approach may address limitations of
   conventional chemotherapies, which often exacerbate neutrophil
   infiltration and contribute to chemoresistance.^[ [192]^28 ^] To
   facilitate clinical translation, comprehensive investigations into the
   long‐term biodistribution, biodegradation, and clearance of BP‐PEG are
   essential to establish its safety profile. Additionally, elucidating
   the molecular mechanisms underlying its immune modulation will be
   crucial for therapeutic optimization.^[ [193]^40 ^] Additionally,
   exploring the effects of BP‐PEG on other immune cell subsets, such as
   regulatory T cells and myeloid‐derived suppressor cells, could provide
   deeper insights into its immunomodulatory capacity.^[ [194]^40 ^]
   Future research should evaluate BP‐PEG in other tumor models and
   metastatic contexts will help validate its broad applicability. This
   research enhances our understanding of BP's interactions with the
   immune system and highlights BP‐PEG's potential as a novel
   immunotherapeutic agent.

   Additionally, BP has emerged as promising nanomaterials for enhancing
   wound healing processes due to their unique structural, chemical, and
   functional properties^[ [195]^41 , [196]^42 , [197]^43 ^] The 2D
   structure of BP‐PEG offers a high surface area‐to‐volume ratio,
   facilitating extensive interactions with biological molecules and cells
   at the wound site.^[ [198]^44 ^] This high surface area enhances the
   adsorption of proteins, cytokines, and growth factors essential for
   wound healing.^[ [199]^41 , [200]^44 ^] Additionally, BP‐PEG's ability
   to generate reactive oxygen species (ROS) provides antimicrobial
   protection while promoting angiogenesis and tissue regeneration,
   thereby accelerating the wound healing process.^[ [201]^41 , [202]^45
   ^] Our study on BP's modulation of neutrophils and macrophages provides
   key insights into its potential for enhancing immune responses and
   promoting tissue regeneration. We believe these findings could be
   leveraged for a range of biomedical applications, especially for
   treating infected wounds, where BP's dual actions can significantly
   improve healing outcomes

   There are still some limitations of our studies, First, the
   concentration range of BP‐PEG used in vivo requires further detailed
   investigation. It is crucial to identify an optimal concentration
   window where BP‐PEG can effectively elicit an antitumor immune response
   without causing toxicity to normal tissues. Determining this balance
   will help ensure the therapeutic efficacy of BP‐PEG while minimizing
   potential adverse effects on healthy cells. Second, the precise
   mechanisms by which BP‐PEG is internalized by macrophages and
   neutrophils, induces the expression of inflammatory factors within the
   tumor microenvironment, and reprograms neutrophils toward antitumor
   functions need to be elucidated. Specifically, understanding how BP‐PEG
   influences neutrophil reprogramming will provide deeper insights into
   its immunomodulatory effects and enhance the development of targeted
   cancer therapies. Addressing these limitations in future studies will
   strengthen the understanding and application of BP‐PEG in cancer
   immunotherapy.

4. Conclusion

   In conclusion, BP‐PEG nanosheets effectively modulate the tumor immune
   microenvironment to inhibit breast cancer growth and metastasis by
   promoting neutrophil‐dependent antitumor immunity. These findings
   advance our understanding of BP‐PEG's mechanisms of action and
   highlight the potential of nanomaterial‐based strategies in
   reprogramming tumor‐associated neutrophils toward an antitumor
   phenotype to favor antitumor responses.

5. Experimental Section

Synthesis of BP‐PEG

   BP crystals were exfoliated in N‐methyl‐2‐pyrrolidone (NMP) through
   ultrasonic treatment for 6 h. Following this, 1 mg of BPNS precipitate
   and 10 mg of PEG‐NH2 were dispersed in deionized water. The resulting
   mixture underwent sonication for 15 min and was then stirred for 4 h.
   To remove any unreacted PEG‐NH2, the solution was centrifuged at 12 000
   rpm for 10 min. The obtained BP‐PEG dispersion was subsequently washed
   with deionized water to eliminate residual PEG and solvents, and it was
   stored in the dark at 4 °C. The characteristics of the BP‐PEG
   nanosheets were then analyzed using several techniques, including
   transmission electron microscopy (TEM), zeta potential measurement,
   energy‐dispersive X‐ray spectroscopy (EDS), Fourier‐transform infrared
   spectroscopy (FTIR), X‐ray diffraction (XRD), Raman spectroscopy, and
   UV–vis spectroscopy.

Cell Line and Cell Culture

   4T1 cells were cultured in RPMI 1640 medium (Gibco) supplemented with
   10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. And
   MDA‐MB‐231 cells were maintained in DMEM (Gibco) with the same
   supplements of 10% FBS and 1% penicillin–streptomycin. All cell lines
   were incubated in a humidified environment at 37 °C with 5% CO₂.

Mouse Models

   6 to 8‐week‐old female BALB/c and NSG mice were purchased from Shanghai
   SLAC Laboratory Animal and maintained under specific pathogen‐free
   conditions with ad libitum access to food and water. All animal
   experiments were conducted in accordance with the experimental animal
   guidelines of the University of Science and Technology of China
   (USTCACUC27120124035). For the orthotopic transplantation experiment, 5
   × 10^5 4T1‐luciferase cells were resuspended in 15 µL PBS and mixed
   with an equal volume (15 µL) of Matrigel (Corning, 354 234). The
   mixture was then injected into the fourth and ninth mammary glands of
   the mice. When the tumors reached a volume of 100–150 mm^3, mice were
   randomly divided into two or four groups. BP‐PEG (10 mg kg^−1) was
   administered intravenously every other day. Anti‐Ly6G antibody (Bio X
   Cell, # BE0075‐1) was administered intraperitoneally at 12.5 µg in
   100 µL PBS per mouse daily, starting three days post‐tumor inoculation.
   Ultra‐LEAF Purified anti‐Asialo‐GM1 Antibody (Biolegend, #146 002) was
   administered intraperitoneally at 25 µL per mouse every 3 days.
   SB225002 (MedChemExpress, #HY‐16711) (10 mg kg^−1) was administered
   intraperitoneally every 3 days. Tumor growth was monitored using
   calipers every 3 days, and tumor volume was calculated using the
   formula: 0.5 × length × width × width.

   FVB‐MMTV‐PyMT(no.T004993) transgenic mice were purchased from
   GemPharmatech. When the tumor volume reached 100–150 mm^3, mice were
   randomly assigned to two groups and received either BP‐PEG (10
   mg kg^−1) or vehicle via tail vein injection. Tumor growth was
   monitored as described above. All mice were euthanized using CO₂
   inhalation.

   For in vivo imaging, mice were intravenously injected with fluorescein
   isothiocyanate (FITC)‐labeled BP‐PEG at a dose of 10 mg kg^−1 24 h
   before sacrifice. Organs were collected, and imaging was performed
   using an Animal Imaging System (IVIS Spectrum).

Bulk RNA‐seq

   Fresh mammary tumors were collected 2 days after treatment with BP‐PEG
   or vehicle. RNA was extracted from the tumors using TRIzol Reagent
   (Invitrogen, Carlsbad, CA, USA) following the manufacturer's protocol.
   Library preparation was conducted using the VAHTS Universal V6 RNA‐seq
   Library Prep Kit for Illumina, according to Illumina's standard
   instructions. An Agilent 4200 Bioanalyzer was employed to assess the
   concentration and size distribution of the cDNA library before
   sequencing on an Illumina NovaSeq 6000. High‐throughput sequencing was
   performed according to the manufacturer's instructions (Illumina). Raw
   reads were filtered using seqtk before being mapped to the genome with
   HISAT2 (version 2.0.4). Gene fragments were counted using StringTie
   (version 1.3.3b), followed by TMM (trimmed mean of M‐values)
   normalization. Differential expression analysis was conducted using the
   edgeR package. Significantly differentially expressed genes (DEGs) were
   defined as those with a false discovery rate (FDR) q‐value less than
   0.05 and a fold‐change greater than 2. KEGG (Kyoto Encyclopedia of
   Genes and Genomes) pathway enrichment analysis was performed using the
   appropriate R package.

Quantitative Reverse Transcription PCR (qRT‐PCR)

   qRT‐PCR was performed as previously described. Briefly, total RNA was
   extracted from cells using Trizol Reagent (Invitrogen, Carlsbad, CA,
   USA) according to the manufacturer's protocol. The RNA concentration
   was determined using a NanoDrop spectrophotometer (Life Technologies,
   Carlsbad, CA, USA). cDNA was synthesized from 1 µg of RNA using
   Oligo(dT) primers and Superscript III reverse transcriptase, following
   the manufacturer's protocol (Vazyme, China). Quantitative PCR was
   carried out using SYBR Premix (Vazyme, China). Data were collected
   using a PIKOREAL 96 real‐time PCR system (Thermo Scientific, MA, USA).
   All reactions were performed in triplicate. The primers used in this
   study are listed in the key resources table (Table [203]S1, Supporting
   Information). Primers were designed using PrimerBank or OriGene
   (OriGene Technologies, Inc.) and verified for specificity using the
   NCBI Primer‐BLAST tool. The primers were synthesized by SANGON
   (Shanghai, China).

Flow Cytometry Analysis (FACS)

   Tumors were collected immediately after euthanizing the mice. Tumor
   tissues were dissociated into single‐cell suspensions using collagenase
   treatment. The resulting single‐cell suspension was incubated with Red
   Blood Cell Lysis Buffer (Beyotime Biotechnology) for 10–15 min at 4 °C
   to lyse red blood cells. Cells were then washed, counted using an
   automated cell counter (Countstar), and 1 × 10⁶ cells were blocked with
   FcR Blocking Reagent (BioLegend) or mouse serum. The cells were
   subsequently stained for cell surface markers using antibodies
   purchased from BioLegend. Flow cytometry analysis was performed using a
   CytoFLEX flow cytometer (BD Biosciences), and the data were analyzed
   with CytExpert software (version 2.4.0.28) and FlowJo software (Tree
   Star, Inc., Ashland, OR).

Mass Cytometry (CyTOF)

   Tumor tissues harvested from the mice were processed into single‐cell
   suspensions using the same methods as those used for flow cytometry. A
   total of 40 metal‐conjugated antibodies were pre‐prepared using the
   Maxpar Antibody Labeling Kit, following the manufacturer's instructions
   (Table [204]S1, Supporting Information). The qualified cell samples
   were then incubated with these antibodies for surface marker staining.
   After washing, the cells were analyzed using the Helios3 CyTOF system
   (Fluidigm, USA) for signaling detection. Data acquisition and analysis
   were performed by PLTTech Inc. (Hangzhou, China). Briefly, after
   normalization with the EQ Four Element Calibration Beads, CyTOF data
   were analyzed on the Cytobank platform
   ([205]https://www.cytobank.org/). Nonlinear t‐distributed stochastic
   neighbor embedding (t‐SNE) dimensionality reduction and unsupervised
   PhenoGraph clustering were conducted using the cytofkit package in R to
   identify distinct immune cell populations.

Single‐Cell RNA Sequencing

   Tumor tissues harvested from mice were processed into single‐cell
   suspensions using the same methods employed for flow cytometry. The
   single‐cell suspensions were stained with a fluorochrome‐labeled CD45
   antibody for sorting and labeled with DAPI to identify live cells.
   CD45⁺ cells were isolated using a Beckman cell sorter.

   Cell suspensions were prepared using the Shbio Tissue Dissociation Kit
   (21517‐10) for tissue samples. The prepared cell suspensions, barcoded
   gel beads, and oil were loaded into separate wells of the Chromium Chip
   K. Using the 10x Genomics Chromium system, Gel Beads‐in‐Emulsion (GEMs)
   were generated. The GEMs were then transferred to a PCR instrument for
   reverse transcription. The barcoded gel beads contained 30‐nucleotide
   oligo‐dT reverse transcription primers, enabling poly‐A RNA from the
   cells to be reverse‐transcribed into first‐strand cDNA tagged with
   barcodes and Unique Molecular Identifiers (UMIs). Following reverse
   transcription, magnetic beads were used to purify the cDNA. The
   purified cDNA was then amplified by PCR. The concentration of cDNA was
   measured using a Qubit fluorometer, and fragment size was assessed
   using an Agilent 2100 Bioanalyzer.

   After cDNA amplification, the products were enzymatically fragmented
   and size‐selected using magnetic beads to obtain optimal fragment
   sizes. End repair, A‐tailing, and adapter ligation were performed to
   introduce the Read 2 sequencing primer. PCR was then used to construct
   cDNA libraries containing P5 and P7 adapters. The libraries were
   purified using magnetic beads. Library concentration was measured using
   a Qubit fluorometer, and fragment size was assessed using an Agilent
   4200 Bioanalyzer. Cluster generation and hybridization of the first
   read sequencing primer were performed according to the Illumina User
   Guide. The flow cell containing the clusters was loaded onto the
   sequencing machine. Paired‐end sequencing was performed, controlled by
   the data collection software provided by Illumina, with real‐time data
   analysis during the sequencing process.

Single‐Cell RNA Sequencing Data Analysis

   Data Processing and Initial Analysis: Raw single‐cell RNA sequencing
   (scRNA‐seq) data were processed using CellRanger software (10×
   Genomics)^[ [206]^46 ^] to perform alignment, gene quantification, and
   cell identification, converting FASTQ files into cell expression
   matrices. Based on the cell types and quantification results identified
   by CellRanger, the Seurat package^[ [207]^47 ^] was used for secondary
   filtering and analysis of the data, including subcluster analysis, Gene
   Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG)
   enrichment analysis, and cell–cell communication analysis.

   Quality Control and Filtering: For each sample, a tailored scRNA‐seq
   data analysis pipeline was used to cluster detected cells and identify
   cell types, enabling comparisons of cell composition and abundance
   across different samples. Data preprocessing and normalization can
   affect subsequent cell classification results; therefore, different
   datasets require specific filtering methods to avoid inaccuracies due
   to excessive abnormal cells or genes. Considering batch differences
   caused by experimental or sequencing factors (such as sequencing depth,
   UMI counts, gene expression levels, mitochondrial, and ribosomal
   content), applying the same filtering thresholds across all samples is
   unreasonable. The distributions of UMI counts, gene counts,
   mitochondrial gene percentages, and ribosomal gene percentages were
   statistically analyzed for each sample to calculate appropriate
   filtering thresholds. Cells with high mitochondrial gene content
   indicate potential cell death; thus, an upper mitochondrial gene
   percentage threshold was set.^[ [208]^48 ^] Cells with erythrocyte gene
   percentages exceeding 1% were filtered out. If the mitochondrial gene
   percentage filtering threshold for an individual sample exceeded 25%,
   it was re‐filtered using a threshold of 25%. As scRNA‐seq typically
   encapsulates cells into droplets, with each droplet ideally containing
   a single cell associated with a barcode, increasing cell loading can
   lead to a higher probability of droplets containing two or more cells
   (doublets or multiplets). These doublets shared the same barcode and
   were treated as single cells, potentially introducing erroneous
   information into subsequent analyses. Quality control based on total
   reads per cell, genes per cell, and mitochondrial gene percentage could
   filter out most doublets; however, some might remain. Therefore, the
   Scrublet algorithm^[ [209]^49 ^] was used to further filter doublets.

   Doublet Detection Using Scrublet: The Scrublet algorithm simulated
   doublets by randomly pairing barcodes and adding them to the original
   expression matrix. All cells (including simulated doublets) were then
   clustered, and each cell was assigned a doublet score based on
   clustering results. The doublet score was proportional to the number of
   associated simulated doublets; higher scores indicated a higher
   likelihood of being a true doublet.^[ [210]^49 ^]

   Cell Cycle Analysis: Cell cycle scores were calculated for each cell
   based on the expression of cell cycle genes,^[ [211]^50 ^] determining
   the cell cycle stage of each cell. Principal component analysis (PCA)
   was used to visualize and evaluate cell cycle effects. If the effect
   was significant, linear regression was employed to remove cell cycle
   effects. In this experiment, cell cycle effects were minimal and no
   adjustment was made.

   Highly Variable Gene Identification and Clustering: The top 2000
   variable genes were identified for each sample based on the mean
   expression and dispersion (variance/mean) for subsequent integration
   analysis.^[ [212]^51 ^] Specifically, genes were grouped into 20 bins
   based on average expression. The normalized dispersion for each gene
   was calculated as the absolute difference between the variance and the
   median variance of genes within the same bin. Integrated data were
   subjected to PCA for dimensionality reduction. The Louvain algorithm,
   also known as graph‐based clustering, was applied to the normalized
   data for clustering analysis. This algorithm is widely used in
   scRNA‐seq data analysis published in high‐impact journals and is
   effective in identifying hierarchical community structures by
   optimizing the modularity of the community network.

   Dimensionality Reduction and Visualization: Uniform Manifold
   Approximation and Projection (UMAP), a manifold learning technique for
   dimensionality reduction, was used for data visualization. UMAP offered
   visualization quality comparable to t‐SNE, retains more global
   structure, and had excellent runtime performance, making it suitable as
   a general‐purpose dimensionality reduction technique for machine
   learning.

   Differential Expression and Functional Enrichment Analysis: Clustering
   analysis divided the cells into distinct clusters. Based on sample
   grouping information, inter‐group differential analysis was performed
   for each cluster to identify differentially expressed genes (DEGs).
   DEGs were screened based on fold change (FC > 1.5 or FC < 0.66) and
   p‐value < 0.05. GO and KEGG enrichment analyses were conducted on the
   DEGs to identify significantly enriched functional pathways, aiming to
   explore functional heterogeneity among sample groups.

   Marker Gene Identification and Cluster Characterization: Marker genes
   were analyzed for all clusters using the Wilcoxon algorithm, employing
   a group one vs rest approach to score marker genes. Genes that were
   specifically highly expressed (logFC > 0.25) in each cluster and
   expressed in at least 20% of cells were selected as significant marker
   genes. While marker genes represent the characteristic genes expressed
   by each cluster, individual genes may not fully reflect the cluster's
   properties. Therefore, GO and KEGG enrichment analyses were performed
   on the marker genes of each cluster to further infer the cellular
   functional characteristics through the pathways and GO terms associated
   with the marker genes.

   Cell Type Annotation: Cell types were annotated using SingleR^[
   [213]^52 ^] and marker genes. SingleR selects RNA‐seq data as a
   reference, identifies genes with high variability between different
   cell types in the reference database, and calculates the correlation
   between the predicted cells and the reference database. By iteratively
   removing the cell type with the lowest correlation, SingleR ultimately
   annotates the predicted cell types.

   Additionally, based on literature, marker gene lists related to all
   expected cell types were collected. SCINA (Semi‐supervised Category
   Identification and Assignment)^[ [214]^53 ^] assigned cell type labels
   to individual cells based on the expression levels of marker genes.
   SCINA is an automated algorithm for cell type detection and assignment
   in scRNA‐seq data, capable of assigning cell identities to all cells
   based on prior knowledge of features (marker genes) highly (or lowly)
   expressed in each cell group.

   Based on the marker genes provided in SCINA and the semi‐supervised
   annotation results, various visualization methods (heatmaps, bubble
   plots, UMAP/t‐SNE plots, violin plots) were used to display the
   expression characteristics of marker genes in various cell groups.

   Cell–Cell Communication Analysis: Cell communication occurs throughout
   organisms, forming complex regulatory networks. Signal interactions
   mainly include adjacency‐type and diffusion‐type transmission methods.
   Adjacency‐type transmission includes nerve signal transmission and
   antigen presentation, characterized by signal transmission between two
   closely associated cells, such as direct binding of membrane‐bound
   ligand‐receptor pairs. Diffusion‐type transmission primarily involves
   secreted proteins diffusing through the extracellular matrix or body
   fluids to other cells, with endocrine, paracrine, and autocrine being
   typical examples.

   The CellChat^[ [215]^54 ^] analysis method was used to infer,
   visualize, and analyze cell–cell communication networks. CellChat
   models communication probabilities and identifies important
   communication networks by subdividing detected ligand–receptor pairs
   into signaling pathways. CellChat constructed a cell communication
   reference database‐CellChatDB‐by selecting 2021 validated cell
   communication relationships (applicable to humans and mice),
   considering multi‐subunit receptors and other important signaling
   auxiliary factors, including soluble agonists, antagonists,
   co‐stimulatory, and co‐inhibitory membrane‐bound receptors.

   In Vitro Cytotoxicity Assay: Approximately 3000 cells per well were
   seeded in 96‐well plates 24 h before the addition of BP‐PEG. The cells
   were then treated with varying concentrations of BP‐PEG (0–50 µg mL^−1)
   for 24–72 h. Cell viability was assessed using the Cell Counting Kit‐8
   (CCK‐8) assay, following the manufacturer's instructions. All samples
   were analyzed in triplicate, and data are presented as the mean ±
   standard deviation from two to three independent experiments.

Co‐Culture Assay

   Peripheral blood samples were collected from breast cancer patients
   prior to surgery, in accordance with the experimental guidelines of the
   University of Science and Technology of China (2022KY256). To establish
   an in vitro co‐culture system, PBMCs were isolated from the blood
   samples using Ficoll–Paque density gradient centrifugation. The
   isolated PBMCs were then co‐cultured with MDA‐MB‐231 cells at a 10:1
   ratio for 48 h, with or without BP‐PEG. Viable MDA‐MB‐231 cells were
   captured and counted before processing for crystal violet staining. The
   PBMC‐containing suspension was also sent for flow cytometry analysis.

   For the co‐culture of neutrophils with 4T1 cells, single cells from
   tumors were isolated 48 h after BP‐PEG treatment, as described for flow
   cytometry. Neutrophils were isolated from the resulting cell suspension
   using magnetic selection with anti‐Ly‐6G microbeads (Miltenyi Biotec).
   Isolated neutrophils were co‐cultured with 4T1‐H2B‐GFP tumor cells at a
   40:1 neutrophil‐to‐cancer cell ratio. GFP^+ tumor cells were quantified
   24 h later by microscopy.

Histology and Immunohistochemistry (IHC)

   Tissue specimens from mice were fixed in 10% buffered formalin for 24 h
   and then stored in 70% ethanol until paraffin embedding. Sections of 5
   µm were cut and stained with hematoxylin and eosin (H&E) or used for
   immunohistochemical analysis. Immunohistochemistry was performed on
   formalin‐fixed, paraffin‐embedded tumor tissue sections using the
   biotin‐avidin method, as previously described. Sections were stained
   with antibodies against Ly6G (neutrophil marker) and granzyme B (GZMB).
   The immunoreactivity was developed using DAB, followed by
   counterstaining with hematoxylin. Images were captured using the
   3DHISTECH Pannoramic MiDi slide scanner equipped with a 20× objective
   lens.

Enzyme‐Linked Immunosorbent Assay (ELISA)

   Cells were treated according to their respective groups. Subsequently,
   the culture medium from each group was collected. After centrifugation,
   the supernatant was analyzed for CXCL2 concentrations using an ELISA
   assay (Lianke Biotechnology, Hangzhou, China). Following the addition
   of the stop solution, absorbance was measured at 450 and 630 nm using a
   microplate reader.

Drawing Schematic Diagram

   All schematic diagrams in this work were created in BioRender. Yu, F.
   (2024), Agreement number: KP27PKUH9C). Including images in
   Figures [216]1, [217]2, [218]3, [219]5, [220]6, [221]7, and [222]8, and
   [223]S1H and [224]S7D (Supporting Information).

Statistical Analysis

   The quantification of data analyses was mainly performed with GraphPad
   Prism statistical software, except bulk RNAseq and scRNAseq. The
   detailed information on the processing of bulk RNA‐seq and scRNA‐seq is
   described in the relevant sections. For in vivo assay, at least five
   mice were used for each group, and at least three samples or
   independent experiments were performed for other experiments. p‐Value
   was determined by unpaired t‐tests. A value of p < 0.05 was considered
   statistically significant.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   J.W. conducted the experiments and supported data acquisition. W.Q.Y.
   performed the analysis of all sequencing data. Y.X.S., H.J.Z., and
   B.Y.Z. prepared and characterized the various forms of Black Phosphorus
   used in this study. H.S. assisted to data analysis and interpretation.
   X.P. assisted to ELISA assay. Z.Y.Y., M.X.P., and Y.H. supervised
   manuscript revisions. F.Z.Y. conceptualized the project, conducted
   experiments, analyzed and interpreted the data, and authored the
   manuscript. Z.Y.Y., F.Z.Y., provided project oversight and supervised
   manuscript revisions. All authors reviewed and approved the final
   version of the manuscript for submission.

Supporting information

   Supporting Information
   [225]ADVS-12-2414779-s001.docx^ (6.4MB, docx)

Acknowledgements