Microenvironmental regulation of tumor-associated neutrophils in malignant glioma: from mechanism to therapy

Abstract

Glioma is the most common primary intracranial tumor in adults, with high incidence, recurrence, and mortality rates. Tumor-associated neutrophils (TANs) are essential components of the tumor microenvironment (TME) in glioma and play a crucial role in glioma cell proliferation, invasion and proneural-mesenchymal transition. Besides the interactions between TANs and tumor cells, the multi-dimensional crosstalk between TANs and other components within TME have been reported to participate in glioma progression. More importantly, several therapies targeting TANs have been developed and relevant preclinical and clinical studies have been conducted in cancer therapy. In this review, we introduce the origin of TANs and the functions of TANs in malignant behaviors of glioma, highlighting the microenvironmental regulation of TANs. Moreover, we focus on summarizing the TANs-targeted methods in cancer therapy, aiming to provide insights into the mechanisms and therapeutic opportunities of TANs in the malignant glioma microenvironment.

Introduction

Glioma is the most common primary intracranial tumor in adults, with high incidence, recurrence, and mortality rates [1]. Overall, the broad category of gliomas (ICD-O-3 histology codes 9380–9384 and 9391–9460) represents approximately 24.5% of all primary brain tumors and 80.9% of all malignant tumors in adults [2]. Glioblastoma (GB), classified as grade 4, adult-type diffuse gliomas, represents the most common malignant brain tumor in adults with a 5-year survival rate of only 6.8% [2,3,4]. Current treatments for newly diagnosed patients suggested by newest guidelines mainly involve maximum safe surgical resection, followed by radiotherapy (RT) and radiochemotherapy combined with temozolomide (TMZ), as well as other combination treatments such as adjuvant chemotherapy and Tumor Treating Fields [4, 5]. Over the past decade, treatment strategies for glioma have continued to evolve. However, due to the highly invasive and drug resistance of glioma, the overall survival (OS) of glioma patients has not significantly improved, novel therapeutic targets and precision treatment strategies are urgently needed [6].

The tumor microenvironment (TME) is a key factor in tumor progression and has been one of the research hotspots in recent years. It is generally believed that the TME plays a crucial role in various malignant behaviors such as proliferation, invasion, angiogenesis, and drug resistance. In glioma TME, non-tumor cells mainly include tumor-associated macrophages (TAMs), dendritic cells (DCs), tumor-associated neutrophils (TANs), lymphocyte, astrocyte, endothelial cells, and others [7]. Previous studies focused more on the role of TAMs and other cells in the glioma microenvironment, the function of TANs was not appreciated for a long time because it was not as abundant as TAMs. With the development of bioinformatic analysis tools and single-cell transcriptome sequencing technology in recent years, the role of TANs in the progression of various tumors, including glioma, has been gradually discovered and attracted attention [8, 9].

Neutrophils are the most abundant myeloid cells in the circulation of humans and mice, with a ratio to monocytes of approximately 7:1 to 10:1 [10]. During the pathophysiological process of tumors, neutrophils proliferate, extravasate, and enter the TME from the peripheral blood through specific mechanisms, becoming TANs. Studies over the past decade have shown that TANs are not merely inert “bystanders” as traditionally believed, but rather important components that influence tumor progression and exhibit both pro-tumor and anti-tumor effects [11,12,13,14,15,16]. Histologically, it has been confirmed that neutrophils infiltrate more in malignant glioma [17]. Substances produced by TANs, such as matrix metalloproteinase 9 (MMP9), promote angiogenesis in glioma [18]. Furthermore, the role of TANs in resisting anti-angiogenic therapy and immune checkpoint therapy has been confirmed [19]. Conversely, other studies have shown that TANs also play an anti-tumor role, such as direct or indirect cytotoxic effects [20, 21]. The functions of TANs in glioma microenvironment are context-dependent and still need deep investigation.

In this review, we mainly discuss how TANs participate in the malignant behavior of glioma, including the origin of TANs, the impacts of TANs on glioma cells, and a particular focus on the microenvironmental regulation of TANs. We also summarize the research progress of TANs-targeted therapies in glioma treatment and novel neutrophil-based drug delivery systems in cancer therapy, with an attempt to highlight the important roles of TANs in glioma microenvironments and tumor progression, which finally shed light on TANs-targeted glioma therapy.

From neutrophils to tumor-associated neutrophils

Neutrophils develop and mature in the bone marrow

Neutrophils, recognized as crucial participants in acute inflammation, are a subset of polymorphonuclear leukocytes. They are typically the first leukocytes to be recruited to inflammatory sites, including tumors, exerting a significant impact on tumor progression [11, 22]. Neutrophils continuously originate from myeloid stem cells in the bone marrow. Granulocyte colony-stimulating factor (G-CSF) plays a pivotal role in promoting this process. During maturation, neutrophils go through the following stages: myeloblast, promyelocyte, myelocyte, metamyelocyte, band granulocyte, and eventually become mature polymorphonuclear/segmented granulocyte [22]. Mature neutrophils express G-CSF receptor (G-CSF-R), Toll-like receptors (TLRs), and CXC chemokine receptor 2 (CXCR2) on their surfaces. These receptors are closely associated with the egress of neutrophils from the bone marrow [22] (Fig. 1).

Fig. 1

Neutrophil maturation and recruitment Neutrophils develop and mature in the bone marrow. Mature neutrophils travel out of the bone marrow into the peripheral blood in the presence of G-CSF, GM-CSF, TLR ligands and CXCLs. Subsequently, cytokines such as CXCLs, IL-1β and other cytokines recruit neutrophils into the TME. Neutrophils enter the TME and become TANs and differentiate into N1 and N2 types in response to IFN-γ, TGF-β signaling. New TANs classifications identified by dcTRAIL-R1 and CD101 statuses

Neutrophils are recruited to the TME and designated as TANs

In patients with solid tumors, mature neutrophils in the organism are classified into two major categories based on their localization—peripheral blood neutrophils (PBNs) and TANs infiltrating the tumor tissue [14]. The neutrophil-to-lymphocyte ratio (NLR), the ratio of neutrophil count to lymphocyte count in peripheral blood, is a biomarker linking the neutrophil-dominated innate immune response to the lymphocyte-supported adaptive immunity [23]. NLR can to some extent reflect the status of TME, serving as a commonly used indicator for clinical efficacy assessment and prognosis evaluation [24]. For example, patients with high NLR values are diagnosed with high-grade glioma and associated with poor OS [24].

Under the influence of several molecules (see below), neutrophils move out from the blood and are recruited into the TME. When neutrophils traffic into TME, they are referred to as TANs [14]. Studies indicate that CXCR2 is a critical molecular mediator of neutrophil recruitment. CXCR2, expressed on the surface of neutrophils, is functionally associated with mediating neutrophil activation [25, 26]. The level of CXCR2 is closely correlated with the prognosis of glioma. Yang et al. have demonstrated that elevated CXCR2 levels indicate a significantly higher recurrence and metastasis rate in glioma patients, serving as an adverse prognostic indicator [27]. Glioma cells produce G-CSF and granulocyte-macrophage colony-stimulating factor (GM-CSF), promoting the egress of neutrophils from the bone marrow and prolonging their lifespan [16, 28, 29]. This process is also closely associated with CXCR2 and its ligands CXC motif ligands (CXCLs) [29]. CXCR2 ligands include CXCL1, CXCL2, CXCL5, CXCL6, CXCL8 (Interleukin-8, IL-8), with IL-8 playing a crucial role in the recruitment of neutrophils to the microenvironment of glioma. Various components in the TME, such as IL-1β, TNF-α, can stimulate GB cells to produce IL-8, recruiting neutrophils (Fig. 2A) [30,31,32]. Fas in the TME, when bound to FasL on the surface of glioma cells, activates downstream signaling pathways, promoting the expression of IL-8 and recruiting neutrophils (Fig. 2B) [33]. Additionally, neutrophil extracellular traps (NETs) can promote the secretion of IL-8 by glioma cells through the HMGB1/RAGE/NF-κB axis, recruiting neutrophils [34]. These neutrophils, in turn, mediate the formation of NETs through the PI3K/AKT/ROS axis, creating a positive feedback loop (Fig. 2C) [34]. It has also been found that IL-8/CXCL1 interaction likewise plays an important role in TANs recruitment in the zebrafish glioma model [35]. Apart from IL-8, IL-1β is also an essential cytokine for promoting neutrophil recruitment. Wang et al. have suggested that the binding of LINC01116 to DDX5 promotes the expression of IL-1β in glioma cells, facilitating the recruitment of neutrophils (Fig. 2D) [36]. Lee et al. have indicated that ectopic expression of CD133 induces an increase in IL-1β expression, which promotes neutrophil recruitment (Fig. 2E) [37].

Fig. 2

Recruitment of TANs IL-8 and IL-1β can recruit neutrophils into TME. (A) IL-1β, TNF-α in TME improve IL-8 production in glioma cells. (B) FasL in TME improves IL-8 production in glioma cells. (C) Positive feedback loop in TANs and glioma cells. NETs produce HMGB1 and promote IL-8 secretion in glioma cells by HMGB1/RAGE/NF-κB axis. IL-8 binds to CXCR2 on TANs promoting NETs formation by PI3K/AKT/ROS axis. (D) LINC01116 combined with DDX5 promote IL-1β secretion in glioma cells. (E) Ectopic expression of CD133 in glioma stem cells stimulates glioma cells secreting IL-1β

TANs are “educated” to play dual roles in TME

Numerous studies indicate that TANs exhibit both pro- and anti-tumor effects. In a considerable portion of research, like the differentiation of TAMs into M1-like and M2-like phenotypes, TANs are also described as N1-type (with anti-tumor effects) and N2-type (with pro-tumor effects) [11, 38, 39]. Several molecules have been identified to distinguish between N1 and N2. N1 markers include intercellular cell adhesion molecule-1 (ICAM-1), inducible nitric oxide synthase (iNOS), C-C motif ligand 3 (CCL3), tumor necrosis factor-α (TNF-α), etc.; N2 markers include CCL17, CCL2, arginine (Arg), CCL5, vascular endothelial growth factor (VEGF), etc [35]. What is clearer is that transforming growth factor-β (TGF-β) signaling in the TME tends to drive neutrophils towards a pro-tumor phenotype (N2), while type I interferon (IFN) signaling or inhibition of TGF-β signaling can drive neutrophils towards an anti-tumor phenotype (N1) [38]. In addition, N1 and N2-type TANs may also vary with TME: in early tumor stages, TANs exhibit stronger cytotoxic effects on tumor cells, whereas in advanced tumor stages, TANs exhibit more of a pro-tumor phenotype [40, 41]. (Fig. 1)

Some scRNAseq data suggest that N1/N2 may not be the so-called ‘polarization’ in different directions, but rather that TANs exhibit different effects on tumor cells at different degrees of activation. Using scRNAseq technology, they identified TANs as different from N1/N2 in multiple differentiation states. Wu and colleagues identified TANs as 10 distinct transcriptional states, including inflammation, angiogenesis, and antigen presentation, from 225 samples from 17 cancer types [8]. A newly published article in Science presents a novel perspective on TANs recruitment and reprogramming [42]. In pancreatic tumor model, researchers classified infiltrating neutrophils into three different neutrophil states, T1, T2 and T3 [42]. Their states can be determined by dcTRAIL-R1 and CD101, i.e., T1 (dcTRAIL-R1⁻CD101⁻), T2 (dcTRAIL-R1⁻CD101⁺), and T3 (dcTRAIL-R1⁺) [42]. Of these, T1 and T2 neutrophils are transitional neutrophil populations, differentiated from immature and mature neutrophils, respectively, while T3 neutrophils are in a terminally differentiated state, formed by further reprogramming of T1 and T2 neutrophils [42]. This process is irreversible [42]. In addition, the researchers further found that TME stimuli are necessary for the upregulation of dcTRAIL-R1 i.e., the shift to T3 neutrophils, but this process does not require the intervention of any tumor-soluble factors [42]. This finding suggests a defined program within neutrophils that enables them to acquire the T3 phenotype independent of their maturation stage [42]. At the functional level, T3 neutrophils are located in hypoxic niche and highly express VEGFA, promoting angiogenesis and tumor growth [42]. (Fig. 1)

TANs involving in malignant behaviors of tumor

TANs promote tumor invasion

Several studies across different tumors have confirmed the correlation of TANs with the invasiveness, malignancy, and adverse prognosis of tumors, and that neutrophil depletion therapy contributes to promoting survival [43]. The invasion of tumor cells into normal brain tissue is a characteristic of malignant glioma [44]. Isocitrate dehydrogenase (IDH), a key enzyme in the Krebs cycle, catalyzes the oxidative decarboxylation of isocitrate. IDH has two isoforms, IDH1 and IDH2, with most mutations involving the substitution of arginine at position 132 by histidine in IDH1 (R132H) and arginine at position 172 by lysine in IDH2 (R172K) [45]. Based on the status of IDH, glioma can be classified into wild-type IDH (wt-IDH) and mutant IDH (mu-IDH) [45]. Compared with wt-IDH, mu-IDH exhibits weaker invasiveness and higher rates of progression-free survival (PFS) [45, 46]. Amankulor et al. compared the immune cell composition in wt-IDH1 and mu-IDH1 glioma and found that mu-IDH1 glioma had fewer immune cell infiltration, especially neutrophils [17]. Also, they noted a lower presence of neutrophil chemokines in the homogenates of mu-IDH1 glioma [17]. Further, neutrophil depletion using an anti-Ly6g antibody had a significant survival benefit in wt-IDH1 glioma mice and no significant effect in mu-IDH1 glioma mice [17, 47].

The awakening of dormant tumor cells is a common cause of tumor invasion, metastasis, and recurrence. In experimental tumor models, NETs formed during the inflammation can awaken tumor cells [48]. NETs, generated by neutrophils, are extracellular network structures constructed from DNA fiber (nuclear and/or mitochondrial DNA) and modified with histones and enzymes [48, 49]. Neutrophil elastase (NE) and MMP9 are pivotal among these. Using a model of lung cancer caused by chronic inflammation, Albrengues et al. have found that NETs DNA connects to the extracellular matrix (ECM), transports NE and MMP9 released by neutrophils to the vicinity of the substrate, facilitating the breakdown and reorganization of laminin [50]. This recombined laminin awakens dormant cancer cells via integrin activation and the FAK/ERK/MLCK/YAP signaling pathway, promoting aggressive tumor growth and metastasis [50]. Studies on various intracranial tumors similarly support the promoting role of these two enzymes in tumor invasion. NE is absent in normal brain tissue but is present in the border zones of high-grade glioma [51, 52]. Compared to low-grade glioma, the borders of GB exhibit more NE, correlating with stronger invasiveness [51]. A study including samples from 62 glioma patients (surgical or biopsy) has suggested that higher levels of MMP9 in glioma mean stronger invasiveness, higher grades, and poorer prognosis [53]. Additionally, MMP9 promotes the proliferation of glioma cells [53]. (Fig. 3A)

TANs promote tumor proliferation

TANs have been shown to directly promote tumor cell proliferation. Researchers co-cultured U87MG glioma cells with human neutrophils and observed neutrophil promotion of glioma cell proliferation in both 2D and 3D co-culture experiments [54]. Interestingly, soluble factors mediating glioma-neutrophil communication vary over time. After 24 h and 72 h of co-cultivation, the production of IL-8 in glioma-neutrophil cultures increased significantly and decreased sharply after 120 h. The production pattern of TNF-α is similar to IL-8. The opposite was observed for IL-1β. It was virtually unproduced in the first few hours, but its release rose after 120 h [54]. Correspondingly, glioma cells were attacked by neutrophils in the first 24 h of co-cultivation, and after 120 h, tumors responded to the presence of neutrophils with increased viability [54]. The crosstalk between neutrophils and glioma cells potentiates the differential release of immune cell recruiting and suppression molecules promoting tumor progression [54]. One study based on zebrafish glioblastoma model indicated that neutrophils are recruited in early stage during oncogenesis and increases tumor cell proliferation [35]. The possible explanation is that ROS released by TANs can cause DNA damage and induce tumorigenesis in peri-tumor cells [35, 54]. Both lung adenocarcinoma model and non-small cell lung cancer (NSCLC) model confirm that TANs produce NE, which promotes tumor cell proliferation by upregulating the PDGFR/PI3K pathway and PGE2 synthesis [55, 56]. CCK8 assay demonstrated that NETs significantly contributed to the proliferation rate of glioma cells, which was reduced by DNase I, hydrolase can degrade NETs [34]. Although NE is an important component of NETs, it is not clear whether the role of NETs in promoting tumor proliferation is wholly or partially dependent on NE. TANs also promote tumor proliferation indirectly through the secretion of soluble factors. Bone morphogenetic protein 2 (BMP2) and TGF-β2 secreted by TANs promote the expression of miR-301b-3p in hepatocellular carcinoma (HCC), which has been shown to promote hepatocellular carcinoma (HCC) proliferation and stem cell characteristics [57]. Meanwhile, miR-301b-3p can suppress CYLD expression in HCC cells, consequently leads to hyperactive NF-κB signaling, higher level of CXCL5 secretion, and more TANs recruitment, forming a positive feedback loop [57]. (Fig. 3B)

Many studies focus on the role of ferroptosis of tumor cells on tumor invasion and proliferation, but contradictions still exist. TANs are closely associated with ferroptosis of tumor cells. TANs transfer myeloperoxidase (MPO)-containing particles into GB cells, and MPO mediates ferroptosis of GB cells by increasing lipid-based ROS, clarifying the mechanism of necrosis in GB cells [58]. In triple-negative breast cancer (TNBC), TANs produce NETs, which mediate ferroptosis resistance of TNBC cells through the TLR9/Merlin signaling axis and promote TNBC proliferation [59]. Both TANs-mediated ferroptosis resistance and ferroptosis of tumor cells enhance the invasiveness of tumor and suggest a poor prognosis [58, 59]. (Fig. 3B)

TANs promote tumor angiogenesis

Like normal organs, tumor tissues obtain oxygen and nutrients through blood vessels, which is mainly achieved through angiogenesis. As a critical manifestation of glioma progression, angiogenesis mainly involves the following steps: (1) vascular rupture, (2) degradation of the vascular basement membrane and surrounding extracellular matrix (ECM), and (3) migration of vascular endothelial cells (ECs) and new blood vessel formation [60]. TANs predominantly influence steps (2) and (3). Various evidence indicates that neutrophils play a significant role in angiogenesis under normal physiological conditions [61]. In the context of cancer, TANs serve as a significant source of pro-angiogenic factors in the TME, no less than TAMs [18]. TANs promote tumor angiogenesis mainly through interaction with ECs.

VEGF/VEGFR interactions, one of the most classical pathways in ECs, activate downstream pathways including PLC/ERK pathway, PI3K/AKT/mTOR pathway, small GTPases, and others [62]. They are widely involved in vascular development, cell survival, regulation of vasomotion, cell shape, cell migration and polarization, and many other processes [62]. TANs can produce large amounts of VEGFA, which stimulates ECs and promotes tumor angiogenesis through the VEGFA/VEGFR2 pathway [18]. In addition, since a large amount of VEGF is sequestered in the ECM in an inactivated form, MMP9, produced by TANs, degrades ECM and releases a large amount of VEGF, which effects on the nearby ECs and promotes tumor angiogenesis [18, 61]. The activity of MMP9 is inhibited by tissue inhibitors of metalloproteinase-1 (TIMP-1). This inhibitor is often released in conjunction of MMP9 to form an MMP9/TIMP-1 complex thereby preventing its protein hydrolysis activity [61]. However, TANs do not express tissue inhibitors of metalloproteinase-1 (TIMP-1), which forms an MMP-9/TIMP-1 complex with MMP9 to inhibit its protein hydrolysis activity, thus requiring only a small amount of TANs to produce significant pro-tumor angiogenic effects [61, 63]. In addition, IL-8 secreted by tumor cells recruits TANs and facilitates the rapid release of MMP9 from TANs. Meanwhile, MMP9 can potentiate IL-8 via amino-terminal processing, resulting in a 10-fold higher potency in TANs activation, consequently promoting MMP9 release and IL-8 mediated recruitment of TANs. This creates a positive feedback loop, recruiting more TANs and promoting more angiogenesis [61, 64,65,66]. (Fig. 3C)

Bv8 (also known as prokineticin 2) is important in promoting tumor angiogenesis [67]. As a homolog of endocrine gland-derived VEGF (EG-VEGF), Bv8 plays the role of mitogen or survival factor of ECs in angiogenesis [68]. Studies have shown that G-CSF activates the JAK signaling and STAT3 pathways, promoting neutrophil proliferation [18]. Meanwhile, G-CSF induces the upregulation of Bv8 expression on neutrophils in a STAT3-dependent manner [69, 70]. Bv8 can activate STAT3 in a JAK2-dependent manner, establishing a positive feedback loop that generates a large amount of Bv8 in the TME and promotes tumor angiogenesis [69, 70]. Blocking Bv8, in turn, can inhibit glioma angiogenesis [71]. Unlike VEGFR (a kind of receptor tyrosine kinases, RTKs), the receptors for Bv8 and EG-VEGF belong to the GPCR family and are not drug targets for tyrosine kinase inhibitors (TKIs) [68, 69]. This may also be one of the reasons why tumors get TKI therapy-resistant. (Fig. 3C)

Recent studies have found that TANs produce Osteopontin (OPN) and MMP14promoting EC chemotaxis and accelerating vascular network formation, respectively [72]. OPN maintains EC survival through activation of the PI3K-AKT pathway, followed by Bcl-xL upregulation and NF-κB activation [72, 73]. Meanwhile, OPN also enhances VEGF expression through the AKT and ERK pathways, which in turn acts as a positive feedback signal activating PI3K/AKT and ERK pathways, enhancing angiogenesis directly [73]. The pro-angiogenic effect of MMP14 is associated with VEGFR1 [74]. In ECs, VEGFR1 is considered as a negative regulator of VEGFR2 [62]. Compared to VEGFA/VEGFR2 interaction, VEGFA has a stronger affinity for VEGFR1 but a weaker ability to induce VEGFR1 phosphorylation [62]. MMP14 can bind selectively and then degrade VEGFR1 on the surface of ECs, thereby enhancing the sensitivity of VEGFR2 to VEGFA, resulting a strong proliferation effect on ECs at low concentrations of VEGFA [74]. (Fig. 3C)

In addition to their pro-angiogenic role, ECs can also affect neutrophil adhesion in TME. Co-culture experiments showed that crosstalk between neuroblastoma cells and ECs downregulates adhesion molecule CD44R expression impairing neutrophil adhesion and invasion, blocking neutrophil recruitment to ECs [75]. Researchers suggested that this might be a tumor escape strategy [75]. Interestingly, the addition of isolated tumor cell membrane fragments to ECs rather than soluble cell culture supernatant caused this effect, suggesting that the interaction of ECs with tumor cells is cell-cell direct contact [75]. This process may be related to the PKC signaling pathway, and blocking PKC can prevent the reduction in neutrophil binding capacity [75]. (Fig. 3C)

TANs promote PN-MES transition

GB has been classified into 4 molecular subtypes—Proneural (PN), Mesenchymal (MES), Classical (CL), and Neural (NL). The NL type has been recognized as a contamination of specimens by normal neurons and has been excluded from the most recent classifications [76,77,78]. Compared to the PN-type, MES-type GB is more aggressive and has a worse prognosis [79]. Presently, most studies have suggested that the critical signaling pathways driving the PN-MES transition (PMT) include STAT3, JAK, and NF-κB pathways, involving various activated molecules such as CXCLs, growth factors, and neurotransmitters [77, 79]. The prevailing view is that the crosstalk between TAMs and GB cells plays a pivotal role in PMT [77, 79, 80]. However, some recent evidence has suggested that TANs also play an indispensable and crucial role in PMT.

Dysregulated expression of the S100 protein family is a common feature in human cancers [81]. S100A4, an important member of the S100 protein family, plays a significant role in tumor progression, metastasis, and mesenchymal transition [82]. Liang et al. have indicated that MES-type GB significantly overexpresses S100A4, whereas the PN type downregulates S100A4 expression [83]. Meanwhile, consistent with the finding that neutrophils can increase S100A4 expression in vitro, S100A4 was significantly co-expressed with CD64 in samples from GB patients, which collectively suggests that TANs promote PMT in GB by facilitating the expression of S100A4 [83]. Chow et al. further identified S100A4 as an upstream regulator of other PMT modulators, such as SNAIL2 and ZEB, which have been explicitly identified to promote the invasion and mesenchymal transition of GB [84]. (Fig. 3D)

Another strong evidence comes from Chen et al. They used CRISPR/Cas9 technology to knock out monocyte chemoattractant protein (MCP) to abolish TME recruitment to monocytes [85]. In a platelet-derived growth factor B (PDGFB)-driven GB model, monocyte depletion was accompanied by a massive influx of neutrophils into the TME and elevation of the neutrophil chemokines CXCL1 and CXCL2. In contrast, these phenomena were not present in the Nf1-silenced GB model [85]. Further experiments demonstrated that neutrophil-derived TNF-α directly drives the mesenchymal transition of PDGFB-driven primary GB cells [85]. (Fig. 3D)

Fig. 3

TANs participate in malignant behaviors of glioma cells (A) NETs assist NE and MMP9 in the catabolism of Laminin, which ultimately promotes the arousal of dormant glioma cells and promotes glioma invasion. (B) TANs produce mtROS, NE, BMP2, TGF-β2, regulate ferroptosis resistance or ferroptosis of tumor cells promoting tumor proliferation. (C) TANs secrete BV8, IL-8, VEGFA, MMP14 and OPN, promoting tumor angiogenesis via influencing EC. Tumor cell-EC interaction reduces neutrophil recruitment. (D) TANs promote PMT of glioma

Interaction of TANs with TME niches and components

Hypoxia

Hypoxia is a common niche in solid tumors and is no exception in glioma. The hypoxia-inducible factor (HIF) family, as crucial transcription factors, become activated under hypoxic conditions, regulating various hypoxia-responsive genes to support cell survival in hypoxic environments [86,87,88]. Other transcription factors such as NF-κB also play vital roles and crosstalk with HIF [89]. Studies have found that in highly heterogeneous tumors, low oxygen signaling, and hypoxia-related genes are more prevalent, correlating with poorer prognoses [90]. Therefore, elucidating the role of the hypoxic niche in tumor progression is both important and necessary.

Hypoxia typically arises due to rapid cancer cell proliferation and the lack of sufficient oxygen-supplying blood vessels [91]. Correspondingly, hypoxia can induce angiogenesis through various signaling pathways, including directly inducing angiogenic factors within tumor cells and indirectly recruiting pro-angiogenic cells [86]. Studies have suggested that hypoxia can induce tumor cells to produce IL-8, one of the key cytokines involved in TANs recruitment [92]. It has also been found that hypoxia induces tumor cells to produce many extracellular vesicles (EVs), which transport amounts of IL-8 and miR-451 [93]. Although it has not been described that hypoxia-derived EV directly affects neutrophil function, the EVs intrinsic IL-8 and miR-451 have been described to play opposite roles in neutrophil recruitment [93]. As mentioned above, IL-8/CXCR1/2 signaling promotes neutrophil recruitment to the TME. Moreover, IL-8 acts as a neutrophil activator that activates the MAPK signaling pathway, inhibits neutrophil apoptosis and enhances neutrophil proliferation [94]. On the contrary, miR-451 inhibits the phosphorylation of p38 MAPK and suppresses neutrophil chemotaxis [95]. Further studies on the specific effects of EVs are needed. Triner et al., in addition, have found that the CXCL1 gene is upregulated in colon cancer in a HIF2α-dependent manner, recruiting TANs through the CXCL1/CXCR2 signaling to promote colorectal cancer progression [96]. Recent studies have discovered that GB, under hypoxic conditions, can produce acrolein, inhibiting neutrophil activation and promoting N2 TANs phenotype [97]. Acrolein can produce TGF-β-like effects, promoting neutrophil expression of Arg1, inhibiting MPO expression, superoxide production and ROS generation, inhibiting tumor-killing activity, and inducing N2 phenotype polarization [97]. Dimercaprol, which is a kind of acrolein scavenger, can inhibit tumor growth significantly [97]. Further studies have shown that acrolein modifies the Cys310 residue in AKT, which is a key residue in the catalytic structural domain of AKT [97]. This modification by acrolein inhibits AKT phosphorylation, blocks the downstream signaling pathway of PI3K/AKT and inhibits neutrophil activation [97]. Overall, the interaction between hypoxia and neutrophils results in pro-tumor genesis. (Fig. 4A).

TAMs

Concerning brain tumors, the concept of TAMs is generally broadened into two cell types: circulating monocyte/macrophage and resident microglia within the central nervous system [98, 99]. TAMs represent most immune cells within the brain tumor microenvironment, constituting up to 30% of the tumor mass [7]. This process may be related to the PKC signaling pathway, and blocking PKC can prevent the reduction in neutrophil binding capacity [100]. The conventional view is that TAMs exist in two polarized states, M1-like (classically activated), which is usually anti-tumor, and M2-like (alternatively activated), which is usually pro-tumor [99]. A novel view, on the other hand, suggests that the M1/M2-like classification inadequately reflects the actual states of TAMs in vivo, thus classifying them into immunosuppressive cells and proliferative macrophages by molecular subtype and functional state [101]. This can help us better understand the role of TAMs in TME.

The interaction between TANs and TAMs has been established in various solid tumors. It promotes oncostatin M and IL-11 associated tumor progression in intrahepatic cholangiocarcinoma (ICC) and post-treatment progression in NSCLC [102, 103]. Researchers have also identified antitumor outcomes resulting from the TANs-TAMs interaction. In dermal sarcoma models, TANs promote TAMs to produce IL-12, subsequently activating the IFN-γ pathway in αβT cells, generating an antitumor effect [104]. But unfortunately, in glioma, studies on TANs-TAMs interactions remain largely unexplored. Researchers tend to focus more on the interactions of TANs or TAMs, separately, with glioma cells and other cells, even though there is overlap in the cytokines expressed or received by TANs and TAMs (such as CCR2, TGF-β, IFN-γ, VEGF, GM-CSF, etc.) [98, 99, 105]. It has been shown that OPN produced by glioma stem cells recruits both TANs and TAMs to glioma TME [106]. It has also been suggested that TANs can promote biopsy-induced glioma cell metastasis by recruiting TAMs [28]. Additionally, an interesting and yet unresolved issue is that, apart from the MES subtype of GBM, TANs recruitment in glioma appears to be repulsed by TAMs. This pattern differs from what is observed in infections or other tumors, where monocytes/macrophages produce neutrophil chemokines to recruit neutrophils [10]. In short, the TANs-TAMs interaction in glioma is a worth exploring research direction. (Fig. 4B)

T-cell

The dysfunction of T-cell is the most important mechanism behind the immune suppression in glioma [99]. TANs produce a high level of ROS that can enter T-cell, damaging telomeres in DNA, accelerating T-cell aging, consequently leading to T-cell dysfunction [107]. Arginine is a crucial factor for sustaining T-cell activation, and Arginase 1 (Arg1) released by TANs extracellularly significantly depletes extracellular L-arginine, causing T-cell dysfunction [108]. This mechanism often results in systemic T-cell suppression in GB patients [108].

The interaction between the programmed cell death receptor-1 (PD-1) and its ligand (PD-L1) represents the classical T-cell immune checkpoint that acts as a “brake” on immune function [109]. In glioma, this pathway serves as a significant mechanism causing T-cell suppression [110]. Sequential IF studies have shown that CD4⁺ T helper cell, Treg, and CD8⁺ T-cell are closer to TANs than to other TME cells [111]. Meanwhile, PD-L1⁺ TANs were closest to PD-1⁺CD8⁺ T-cell and the most abundant PD-L1⁺ cell type within the PD-1⁺CD8⁺ T-cell niche was TANs [111]. This stands as critical evidence of TANs-mediated immune suppression in glioma TME. Furthermore, TANs can induce PD-1 expression on CD4⁺ effector memory T-cell (TEM) within the GB microenvironment. Simultaneously, TANs exhibit high levels of PD-L1, depleting CD4⁺ TEM function through the PD-1/PD-L1 axis, promoting immune suppression [112]. In a certain GB model with defective type I IFN receptor, tumor-infiltrating CD11b⁺Ly6G⁺ cell increase, while CD4⁺FoxP3⁺ cell (immunosuppressive Treg) increase and CD8⁺ cytotoxic T cell (CTL) decrease [113]. Ferroptosis also plays an important role in T-cell related immunosuppression. Ferroptosis is a non-apoptotic, regulated cell death triggered by dysregulation of redox mechanisms. New studies have shown that TANs are highly susceptible to ferroptosis and even spontaneously undergoes ferroptosis [114]. Meanwhile, ferroptosis accounts for most of the death program in TANs [114]. Ferroptosis in TANs can generate a large amount of PGE2 and oxidized phosphatidylethanolamine (ox-PE), which collectively inhibit CD8 + T cell activity, producing immunosuppression [114] (Fig. 4C).

Interestingly, the interaction between TAN and T-cell also exhibits some antitumor effects. In lung cancer, TANs with an antigen-presenting cell (APC)-like phenotype trigger antitumor T-cell responses and support anticancer activity [21]. Lad et al. have reported a specific population of TANs with DC characteristics [20]. These particular TANs, originating from the local cranial bone marrow, accumulate in the GB microenvironment, can activate T-cell cytotoxicity and memory, inhibiting tumor growth in vivo [20]. Recent findings provide strong support for this anti-tumor effect and have made great strides. Wu et al. analyzed the transcriptional profile of TANs and classified them into 10 different states, of which HLA-DR⁺ TANs express MHC class II molecules and exhibit antigen-presenting properties [8]. Co-culture experiments have shown that HLA-DR⁺ TANs can present tumor neoantigens and activate T-cell directly through ligand-receptor interactions, then stimulate T-cell to express TNFα and generate reactive T-cell responses [8]. According to metabolic analysis, leucine catabolism leads to high HLA-DR expression and initiates antigen presentation mechanisms [8]. Leucine diet combined with anti-PD-1 therapy significantly reduces tumor volume and stabilizes the disease [8]. (Fig. 4C)

Fig. 4

Crosstalk between TANs and glioma microenvironment niches and components (A) Hypoxia stimulates glioma cells to produce IL-8, CXCL1 and Acrolein, promoting neutrophil recruitment and N2-type transformation. (B) TANs and TAMs interaction upregulate STAT3 signaling in tumor cells resulting in protumor effect; CCL3⁺ TANs and SPP1⁺ TAMs interaction promotes post-therapy tumor progression; TANs promote TAMs IL-12 secretion activating IFN-γ signaling in αβ T-cell, leading to antitumor effect. (C) TANs led to immunosuppression via T-cell dysfunction, ICB and ferroptosis; DC-like TANs and HLA-DR⁺ TANs exert an antitumor effect

TANs promote glioma therapy resistance

TANs promote RT resistance in glioma

As the infiltration of tumor cells extends far beyond the lesion, postoperative RT is an important part of glioma treatment, helping to preserve patient function and improve survival [5]. Several studies have demonstrated that TANs are involved in promoting glioma RT resistance [115]. Radiation-induced senescent GB cells recruit TANs and promote angiogenesis via NF-κB signaling [115, 116]. Meanwhile, TANs support GB to glioma stem cells transformation through NOS2-NO-ID4 signaling [115, 116]. In addition, TANs secrete S100A4, which promotes glioma stem cells proliferation and stemness and facilitates RT resistance formation [83, 115]. In lung cancer model, researchers have found that lung cancer cells upregulate GLUT1 expression in TAN, which promotes TANs survival and RT resistance [117]. Increased sensitivity of lung cancer to RT was observed in the TANs GLUT1 knock-out lung cancer mouse model [117].

TANs promote anti-angiogenesis therapy resistance in glioma

TKIs, whose main targets are RTKs, such as EGFR, PDGFR, VEGFR, and are commonly used anti-angiogenic agents that effectively inhibit glioma proliferation and invasion [118]. TANs can produce substances such as VEGFA, MMP9, MMP14, OPN, etc., which enhance VEGFA/VEGFR interaction that promotes glioma angiogenesis. This pro-angiogenic effect largely cuts down the therapeutic efficacy of TKIs or requires larger doses of TKIs to achieve the same efficacy [19]. In addition to VEGF, TANs also produce Bv8 to promote tumor angiogenesis. Unlike VEGFR, the receptor for Bv8 belongs to the GPCR family, therefore, TKIs cannot inhibit the pro-angiogenic effect produced by Bv8, resulting in TKIs resistance in tumors [68, 69]. Several studies based on different solid tumors corroborate this and find that inhibiting TANs can enhance the efficacy of TKIs [119,120,121]. Additional studies have demonstrated that TANs mediate glioma resistance to anti-angiogenic therapy by promoting glioma mesenchymal transition. de Groot JF and colleagues compared acquired bevacizumab-resistant versus non-resistant GB subtypes and found a highly significant correlation between acquired resistance to anti-VEGF therapy and mesenchymal features in GB [122]. Subsequently, they determined that TANs promote GB mesenchymal transition through S100A4 and discovered that S100A4 knockdown neutrophils attenuated mesenchymal transformation and angiogenesis of GB by not affecting neutrophil proliferation [83]. Hence, a combination therapy involving S100A4 inhibitors may be a viable option to counteract resistance to anti-angiogenic therapy in glioma. No relevant clinical/preclinical trials were retrieved, and the specific effects of this regimen need to be further investigated.

TANs promote immune checkpoint blockade therapy resistance in glioma

Immune checkpoint blockade (ICB) has shown remarkable success in various solid tumors, but good efficacy has not been observed in glioma, an important reason being the presence of significant immunosuppression in glioma [123, 124]. Blocking the PD-1/PD-L1 axis represents one of the classic ICB therapies. PD-1 is primarily expressed in activated T-cells and its interaction with PD-L1 reduces the immune response of T-cells [109, 124]. Commonly used drugs include PD-1 antibodies such as Nivolumab, Pembrolizumab, and emerging PD-L1 antibodies such as Atezolizumab [124, 125]. However, a prerequisite for these therapies to be effective is that T-cells possess (at least to some extent) normal functionality. As mentioned earlier, TANs in the glioma microenvironment can cause T-cell dysfunction through various pathways, particularly depleting the functionality of CD4⁺TEM via the PD-1/PD-L1 axis [107, 108, 112, 113]. These can promote glioma treatment resistance to ICB.

Therapies targeting TANs

Anti-angiogenesis therapy

Presently, the standard therapy for newly diagnosed glioma involves surgical resection followed by radiotherapy plus TMZ combined chemotherapeutic treatment [5, 126]. However, almost all patients experience recurrence [5, 126]. Bevacizumab, an anti-VEGF antibody, has been approved for treating recurrent glioma in several countries, including the United States [5]. Nonetheless, current clinical trials have not shown distinct advantages of bevacizumab. Although the use of bevacizumab improves PFS, the OS of patients was not improved compared to standard therapy [127, 128]. Apart from bevacizumab, regorafenib has shown promising progress in recent studies. Regorafenib is an orally available multikinase inhibitor used against angiogenesis, stromal, and RTK [129]. In a phase II clinical trial, regorafenib, when compared to lomustine (TMZ analog), increased the median OS from 5.6 to 7.4 months, with a comparable rate of severe adverse events [129]. Regorafenib might be a potential novel therapy. (Table 1)

Table 1 Clinical trials targeting TANs and/or its correlating agents

PD-1/PD-L1 blockade

PD-1/PD-L1 blockade has shown significant efficacy in many tumors but remains limited effective in glioma. The classic PD-1 monoclonal antibody nivolumab has efficacy comparable to bevacizumab [130]. Combining nivolumab with standard therapy does not influence patient survival [126]. Single-agent use of the PD-1 antibody pembrolizumab is effective for only a small subset of recurrent GB patients [131]. The combination of the PD-L1 antibody durvalumab with bevacizumab has proven ineffective [132]. These less optimistic results encourage researchers to explore new directions. Several studies have shown that adjuvant or neoadjuvant PD-1 monoclonal antibodies, such as nivolumab and pembrolizumab, can enhance the anti-tumor immune response of glioma patients, resulting in a positive effect on glioma treatment [133, 134]. Pembrolizumab combined with DNX-2401, an oncolytic virus capable of inducing an anti-tumor immune response in patients, has shown significant survival benefits for some glioma patients [135]. Furthermore, innovative delivery methods are potential avenues of development. A phase I clinical trial has confirmed that intracerebral administration of nivolumab and CTLA-4 inhibitor ipilimumab after maximal safe resection is feasible, safe, and has a positive effect on OS [136]. Recent studies have found that leucine stimulates TANs to highly express HLA-DR+, present tumor neoantigens and stimulate T cells to produce reactive T cell responses [8]. Animal studies have shown that a leucine diet combined with anti-PD-1 therapy significantly reduces tumor volume and stabilizes the disease [8]. (Table 1)

Targeting CXCLs-CXCR1/2 signaling

IL-8 (CXCL8) plays an important role in neutrophil recruitment to TME and is involved in tumor malignant behavior. Pre-clinical studies have shown that the IL-8 antagonist BMS-986253 significantly reduces neutrophil recruitment and attenuates mesenchymal transition in triple-negative breast cancer [137]. Phase I studies have found that BMS-986253 significantly reduces IL-8 levels and is being evaluated in combination therapy with immune checkpoint inhibitors (NCT02536469) [138]. CXCR2 is an essential neutrophil chemotactic factor receptor, recruiting neutrophils to the TME and regulating TANs participation in tumor angiogenesis [25]. CXCR1 also plays an important role in neutrophil recruitment [35]. In addition, the efficacy of CXCR1/2 antagonists in the treatment of non-neoplastic diseases has been demonstrated. For example, a phase II study demonstrated that MK-7123, a CXCR2 antagonist, significantly reduced neutrophil counts and significantly improved lung function in COPD patients (NCT01006616) [139]. Using CXCR2 antagonists for tumor therapy is a promising strategy, and this has shown good progress in clinical trials of Reparixin for breast cancer (NCT02001974, NCT02370238) [140]. A study using a CXCR2 antagonist (SB225002) in high-grade glioma demonstrated its potential to inhibit tumor growth during tumor development and progression [141]. (Table 1)

Targeting TGF-β signaling

TGF-β signaling plays a crucial role in glioma, inducing TANs to polarize toward the N2 type (pro-tumor), and several downstream signaling pathways are involved in tumor progression [38]. The use of TGF-β antagonists for the treatment of glioma is theoretically a viable option. Galunisertib (LY2157299 monohydrate) is a small molecule TGF-β-R1 kinase inhibitor that displays anti-tumor effects in GB animal models [142,143,144]. A phase I/II clinical trial has indicated that combining 300 mg/d of Galunisertib with TMZ chemotherapy can improve OS for GB patients [143]. Another phase II clinical trial has demonstrated that in recurrent GB patients, the treatment with Galunisertib plus lomustine did not improve OS compared to placebo plus lomustine treatment [145]. The specific efficacy of TGF-β blockade in glioma requires further investigation. (Table 1)

Targeting the S100 protein family

Multiple members of the S100 protein family exhibit dysregulated expression in cancer [81]. As mentioned above, TANs promote glioma metastasis through S100A4 [83]. Single-cell analysis has suggested that S100A4 could be a target for glioma immunotherapy [146]. Feng et al. have found that Allopregnanolone, a physiological neural modulator, significantly downregulates several proteins in GB, including S100A4 and S100A11, inhibiting GB cell survival [147]. An in vitro study has indicated that downregulating S100A12 (primarily expressed by neutrophils) significantly inhibits glioma cell growth [148]. These studies imply that the S100 protein family might be key molecular targets to explore new directions in glioma treatment.

Targeting NETs

NETs play a crucial role in the progression of various cancers, and targeting NETs is a potential new treatment direction. However, most of these studies have remained at the mechanistic level and have not achieved breakthrough progress [48]. In one study, NETs have led to endothelial dysfunction in advanced glioma patients, promoting a hypercoagulable state, which is necessary to cause the complication of venous thromboembolism (VTE) [149]. Researchers subsequently found that in vitro, the combination of DNase I and activated protein C significantly reduced endothelial dysfunction [149]. Furthermore, there is evidence suggesting that NETs might exhibit an anti-tumor effect [150]. Overall, although there is substantial evidence supporting the therapeutic targeting of NETs, we recommend consulting in-depth reviews on neutrophils, NETs, and cancer for more information [48, 49, 150,151,152].

Neutrophil-based drug delivery systems

Neutrophil-based drug delivery systems are a promising targeted cancer therapy strategy, primarily divided into two types: those utilizing neutrophils as carriers and neutrophil membrane-driven nanovesicles [153]. The most significant advantage of this system is its natural ability to traverse the blood-brain barrier (BBB) as neutrophils naturally recruit to the tumor microenvironment. In recent years, studies utilizing this system for glioma treatment have made significant progress. Wang et al. have reported a neutrophil-exosome system targeting the inflamed tumor microenvironment, showing success in crossing the BBB and infiltrating tumor tissues in a GB animal model loaded with doxorubicin in response to inflammatory stimuli [154]. Zhang et al. have reported a neutrophil-based micro-robot (neutrobot) capable of actively delivering drugs to malignant glioma [155]. Using magnetic drug-loaded particles phagocytosed by neutrophils and their chemotactic movement along inflammatory gradients, precise drug delivery significantly suppressed tumor cell proliferation [155]. Although emerging chimeric antigen receptor (CAR)-T cell therapy has been successful in many cancers, its effectiveness in gliomas is limited due to the difficulty of T cells crossing the BBB [156]. Recently, Chang et al. have engineered CAR-neutrophils using CRISPR/Cas9 in human pluripotent stem cells, targeting GB, delivering and releasing nanomedicines to kill tumor cells, prolonging the lifespan of mouse models, contributing to the application of CAR therapy in intracranial tumors [156].

Conclusions and perspectives

In this review, we provide a comprehensive description of the functions and mechanisms of TANs in glioma progression. In most cases, TANs exhibit a pro-tumoral N2 phenotype, promoting glioma cell proliferation, invasion, PN-MES transition, and resistance to radiotherapy and chemotherapy through paracrine and other forms [54]. It is important to note that the role of TANs in tumor progression is context-dependent, TANs can also adopt an anti-tumoral N1 phenotype that inhibits tumor growth [11]. In addition to tumor heterogeneity, the different phenotypes of TANs may be related to the time at which they are recruited into the tumor microenvironment, where they are more likely to exhibit an N1 phenotype at early stages and are gradually induced by the tumor microenvironment to an N2 phenotype over time. Furthermore, a large number of bioinformatic analysis studies have reported a series of new genes or genesets associated with the proportion of TANs infiltration in gliomas and the prognosis of glioma patients, which we have summarized in this paper. However, most of these reports are phenotypic studies, the mechanisms of the above genes and genesets on the infiltration and function of TANs still need in-depth elucidation.

Gliomas are hallmarked by their complex tumor microenvironments [157]. The extensive crosstalk between tumor cells and microenvironmental components within the glioma microenvironment and the formation of positive feedbacks that promote tumor growth and malignant progression, which are important causes of recurrence and treatment resistance in gliomas [157]. In this review, we focus on elucidating the interactions between glioma microenvironment components and TANs and their effects on tumor progression, including the hypoxic microenvironment, vascular endothelial cells, TAMs and T cells. TANs have extensive interactions with multiple microenvironmental components and are essential for maintaining their pro-tumoral N2 phenotype. Therefore, further investigation into the role of TANs in tumor progression and the development of targeted therapies, whether for gliomas or other malignancies, should take full account of the microenvironment in which TANs and tumor cells reside in the tumor, in conjunction with existing tools such as spatial transcriptomics.

Finally, we summarize current advances in TANs-based anti-tumor therapies. Although TANs are highly promising anti-glioma targets, because the mechanism TANs in glioma progression is still largely unclear, there are currently fewer targets that can specifically target TANs within tumors, and even fewer corresponding clinical trials. As the mechanism TANs continues to be elucidated, more therapies targeting TANs will become available for the clinical treatment of gliomas and other tumors, both alone and in combination with therapies including immune checkpoint inhibitors. Additionally, the treatment of glioma with drug delivery systems based on neutrophils represents a promising research direction that requires further refinement through more clinical experiments.