Glioblastoma, also known as glioblastoma multiforme (GBM), is one of the most malignant diseases that is thought to originate within the central nervous system (CNS) (1). Although significant progress has been made in various treatments, such as chemotherapy, radiotherapy and surgery, the survival rate among glioblastoma patients has not shown progressive improvement using these conventional treatments (2–4). In the past few decades, there have been significant new therapeutic advances in oncology, wherein small molecules have shown promising curative effects among many types of malignant tumors. However, given the special anatomic structure of the CNS, many therapeutic molecules are not able to reach the target zone because of the blood–brain barrier (BBB), which is a highly selective barrier in the CNS (5), leading to poor survival and prognosis of patients with malignant CNS tumors.

Nonetheless, recently, immune therapy has shown promising outcomes (6–9) by overcoming the selectivity of BBB as immune cells are highly infiltrative in the tumor microenvironment of glioblastomas (10). The immune system in the CNS plays a significant role in GBM development (11). The tumor microenvironment of GBM is infiltrated with various types of immune cells and cytokines. CD8-positive (CD8+) T cells are the most important type of immune cells in immune therapy as they function as tumor cell killers. Furthermore, macrophages are the most infiltrated immune cells (12, 13) in GBM. Hence, tumor-associated macrophages (TAMs) account for the majority of macrophages within GBM, and are recruited from circulation. Most TAMs are considered to be immunosuppressive agents that are associated with tumor immune escape, as well as angiogenesis and invasion. Microglia are resident immune cells of the CNS that interact as housekeepers of the CNS and participate in the innate immune responses and antigen presentation. However, microglia are silenced in the GBM microenvironment and support immunosuppression. Given their common biological functions, microglia and macrophages in GBM tumors (as in other tumors) are known as tumor-associated microglia/macrophages (TAM/Ms) (14, 15). The crosslink between GBM-associated immune cells and TAM/Ms and CD8+ T cells, which show great potential given their direct and indirect contact with GBM cells in the immune response, can serve as a potential therapeutic target for GBM.

Herein, we outline the physiological, pathological, and micro-environment features of TAM/Ms and CD8+ T cells with regard to GBM development, as well as the impact of their crosstalk on tumor malignancy, and a potential translational scheme for glioblastoma treatment.

Macrophages and microglia have a significant function in the innate immune response of a healthy brain tissue, such as in anti-inflammatory and scavenger processes (14). Microglia and macrophages are classified as neuronal support cells that are located between nerve fibers, and function to carry out the primary immune responses of the nervous system as killer cells and phagocytes. Microglia originate from the embryonic mesoderm and are located in the nerve tissue; likewise, some macrophages develop in a similar manner. However, macrophages can also be derived from monocytes that migrate from circulation to become adult macrophages (14).

In the physiological innate response of the nervous system, macrophages and microglia recognize pathogens or eliminate cells through the pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs). These activate a cascade of immune response reactions that function against infection, ischemia, trauma, or other threats. As oligomeric protein complexes are activated in response to PAMPs and DAMPs, inflammasomes activate and release inflammatory cytokines, including IL-1β and IL-18. Various inflammasomes, inflammatory cytokines, and chemokines play contrasting roles with regard to cancer development and progression. Macrophages and microglia are classified into M1 and M2 clusters that represent pro-inflammatory and anti-inflammatory components, respectively. Pro-inflammatory cytokines are commonly categorized in relation to the M1 phenotype, and include interferon-gamma (IFN-γ), tumor necrosis factor-α (TNF-α), IL-2, and IL-12. In contrast, immunosuppressive molecules, such as IL-4, IL-10, IL-13, and tumor growth factor-β (TGF-β), are classified in relation to the M2 phenotype (16–19). Notably, within the tumor microenvironment, TGF-β plays crucial roles with regard to tumor immunity, and targeted TGF-β signaling blockade in helper T cells elicits an effective tissue-level cancer defense response that can provide a basis for therapies that are directed toward the cancer microenvironment (20); TGF-β is also the most potent immunosuppressor against cancer cells, and this significant effect is mediated in multiple ways, including polarization of macrophages to M2 cells. Additionally, several previous studies have indicated that anti-TGF-β exerts a potential and promising effect on tumor immunity (21–24). The relative literature on TGF-β indicates that TGF-β likely also plays an important role in the regulation of pathological processes in GBM patients. Moreover, a therapy targeting TGF-β can serve as a potential strategy in GBM treatment in the future. Recently, additional biomarkers have been described to play a role in macrophage polarization. TNF-α, IL-12, and IL-23 have all been confirmed to be related to the M1 phenotype (14, 25). Inducible NO synthase (iNOs) has been shown to be associated with M2 clusters in GBM cells (26). Together with a certain polarization proportion, M1 and M2 clusters of macrophages and microglia help maintain CNS homeostasis. Although M1 and M2 phenotypes are different owing to levels of differentiation, they have plasticity to a certain extent, and TAM/Ms show overlapped proportion between M1 and M2 clusters (27, 28).

With regard to GBM, TAM/Ms have common biomarkers, e.g., cluster of differentiation 68 (CD68), which is also the most commonly expressed protein in various malignant diseases (29). Nevertheless, CD163, which is highly specific for M2 macrophages, is also highly expressed in GBM cells (29–31), which suggests a dominance of the M2 cluster of TAM/Ms in the GBM tissue. In an early study on GBM-related macrophages, M1 and M2 clusters were classified together as being histologically similar to each other. Studies have indicated that the inclusion of TAM/Ms into the M2 family is responsible for their non-phagocytic function and anti-immune responses in GBM; TAM/Ms are further classified into M2a, M2b, and M2c subtypes based on their different roles in GBM response. However, specific exclusive characteristics of the M2 family between categories remains to be further elucidated (15). The M1 and M2 polarity classification is defined in the inflammatory response. However, TAM/Ms in GBM play a role in the carcinogenesis process. Differences in the origin of macrophages and activation of the innate immune response induces different responses, dependent on the microenvironment, helping the formation of heterogeneous communities and subgroups with appropriate functions. In the GBM microenvironment, the predominant activity of TAM/Ms takes place through the M2 phenotype, whereas the immune status of TAM/Ms favors immune suppression. Moreover, the enrichment of the M2 cluster TAM/Ms is associated with the prognosis and survival of GBM patients (32–34). From a biochemistry perspective, GBM cells produce immunosuppressive cytokines, including IL-10 (35), macrophage-inducing chemokine monocyte chemotactic protein-1 (CCL2) (16), and cell-surface protein colony-stimulating factor 1 receptor (CSF1R) (36). Transcriptomic analysis of GBM cells has demonstrated a positive relationship between invasion-related genes and immunosuppressive genes (37, 38). Hence, polarized TAM/Ms have been reported to stimulate carcinogenesis via their immune-escape effect and immunosuppressive secretions. Further, cancer growth factors have also been identified in GBM cells co-cultured with TAM/Ms (25). With the help of these growth factors, GBM cells are able to suppress TAM/M polarization into the M1 phenotype and induce polarization into the M2 phenotype by recruiting and re-differentiating macrophages into M2-dominant clusters, thus creating a pro-carcinogenesis immunosuppressive environment.

Previous studies have validated that TAM/Ms are related to the prognosis/survival outcomes of GBM patients. One previous study (39) used automated quantitative immunofluorescence to identify the prognostic impact of TAMs in GBM. The results of their study suggested that M2-like TAMs hold an unfavorable prognostic value in high-grade gliomas. Similarly, Pimenta et al. (40) reported that untreated GBM patients with TAM infiltration demonstrated a shorter overall survival compared with the patients without TAM infiltration. Additionally, the number of microglia/macrophages with positive staining for CD163 and CD204, which are thought to be markers for M2 macrophages, was found to be correlated to the histological grade of gliomas, whereas the ratio of M2 macrophages in the TAM/Ms was related to the histological grade (41). This suggests that evaluation of the proportion of M2 microglia/macrophages in GBM is useful for the assessment of the prognosis of patients with gliomas, including GBM. Kaffes et al. (42) determined that high expression of the TAM-related gene AIF1, which encodes the TAM-specific protein IBA1, is correlated with a worse prognosis in pro-neural GBM, but confers a survival benefit in mesenchymal tumors. Recently, numerous studies have reported that TAM/Ms regulation may be a potential therapy strategy in cases with GBM. A new report demonstrated that PI3Kγ inhibition could suppress TAM/M accumulation in glioblastoma microenvironment, and enhance the anti-neoplastic effects of temozolomide in glioblastoma cells (43). In addition, the deletion of HuR (an RNA regulator) in TAM/Ms could attenuate glioma growth (44). Further, Martins et al. provide a comprehensive overview of microglia-centered combinatorial strategies against glioblastoma, and conclude that MG modulation, as a central paradigm in GBM immune tumor microenvironment, may lead to additional, long-lasting, and effective tumoricidal responses (45).

Altogether, TAM/Ms appear to be involved in the development of GBM via atypical polarization and by creating an immunosuppressive microenvironment, and are tightly associated with the outcome of GBM patients. Thus, we can infer that TAM/M polarization, recruitment, and function of immunosuppressive cytokines may be potential targets in GBM treatment.

CD8+ T cells are a subtype of T cells that develop within the thymus and have a cytotoxic effect on cancer cells through antigen recognition using major histocompatibility complex class I (MHC-1) (46). In cases of carcinoma, CD8+ T cells are characterized to be able to kill cancer cells directly at the end of their cancer immune response at any phase of malignancy.

There are three steps involved in the differentiation of naïve T cells into cytotoxic CD8+ T cells, which leads to a cytotoxic effect on GBM and other cancers. First, the immunogenic antigens that are aberrantly expressed are processed through antigen-presenting cells, such as dendritic cells (DC), and presented to naïve T cells at peripheral lymphatic sites, such as lymph nodes. Activated CD8+ T cells then go on to proliferate clonally and migrate to the GBM site, where CD8+ T cells recognize tumor cells through MHC-1 and release cytotoxic signals that initiate the anti-tumor effect. Traditionally, the central nervous system is considered to be an immune-exempted organ. Recently, many studies have identified that under pathological conditions, the integrity of the BBB and the blood–cerebrospinal fluid barrier are compromised, causing inflammatory cells to enter the brain and activate the immune system (47, 48). However, CD8+ T cell infiltration varies widely between different types of GBMs. An in vivo study demonstrated a substantially increased infiltration of CD8+ T cells in mesenchymal GBM, whereas immune infiltrates were rarely found in pro-neural GBM. Furthermore, the molecular subtype of cancer stem cells (CSCs) TGF-β-dependently contributed to the degree of immune infiltration in patients with GBM (49). Cancer vaccination has been developed to amplify this anti-cancer cytotoxic effect by CD8+ cells (9). However, glioblastoma can escape from natural immune response by immune inhibition. To overcome this challenge, immune checkpoint therapy has been developed that can help boost anti-cancer immune response (6). However, in some cases, GBM may undergo immune escape by not presenting tumor antigens or MHC-1, thus preventing recognition by CD8+ T cells. Genetic modification is another promising method that involves adaptively transferring target antigens to CD8+ T cells and cultivating them clonally. These CD8+ T cells, known as chimeric antigen receptor T cells (CAR-T), have the ability to recognize immune-escaped tumor cells (50, 51).

Various studies have assessed the importance of CD8+ T cells with regard to outcomes of GBM patients. Audencel is a dendritic cell (DC)-based cellular cancer immunotherapy against GBM. Although the recent phase II “GBM-Vax” trial shows no clinical efficacy of Audencel, as assessed through progression-free and overall survival of GBM patients, post-vaccination levels of CD8+ T cells in the blood were indicative of a significantly better survival among GBM patients (52). A report by Yang et al. (53), which included 519 GBM patients, suggested that tumors from long-term survivors, more likely than those form short-term survivors, have either intermediate or extensive CD8+ T cells infiltrates compared with focal or rare infiltrates. This indicates that prolific CD8+ T cell infiltrates appear to correlate with partitioned long-term survival among newly diagnosed GBM patients. Also, increased CD8+ to FoxP3+ regulatory T cell ratios showed a positive correlation with survival outcomes in primary GBM, indicating that absolute numbers of CD8+ T-cells and effective balance of CD8+ T cells to FoxP3+ regulatory cells are both informative for predicting clinical outcomes in patients with GBM (54).

During cancer development, tumor tissue is infiltrated by various immune response cells that can be classified into either immune-effector cells or immune-supporting cells. Among tumor-associated immune cells in GBM, CD8+ T cells are known to be major cytotoxic lymphocytes (CTLs) and TAM/Ms are the most infiltrated immune-supporting cells (55).

The dominant effect of TAM/Ms in GBM is immunosuppression, which is manifested through the expression of PD-L1 and PD-L2, as well as the CTL-associated protein 4 (CTLA-4) ligands CD80 and CD86. The main target of these immune checkpoints is CD8+ T cells, which leads to the inhibition of its cytotoxic effects. To expand into the therapeutic scenario, as in the case of immune checkpoint therapy, there are still several patients who do not respond to immune checkpoint inhibitors (8, 56). An ideal theoretical explanation for this may be that CD8+ T cells need to overcome an intrinsic negative regulation of the immune system. In cases of GBM, TAM/M-induced immunosuppression is one of the major reasons for passive response toward immune checkpoint therapy. In vivo studies have determined that TGF-β secreted by TAM/Ms inhibits T cell-mediated tumor clearance (57). Next, TGF-β targets proteins that are responsible for CTL-mediated tumor cytotoxicity, including perforin, granzyme A, granzyme B, first apoptosis signal ligand (FASL), and IFN-γ. While IFN-γ is the main trigger of cytotoxic activity, FASL is the key element that activates apoptosis of target cells. Neutralization of TGF-β through an antagonist in a mouse model has been shown to upregulate cytotoxic gene expression among CD8+ T cells (57). In contrast, TAM/M-originated TGF-β also causes differentiation of naïve T cells to regulatory T (Treg) cells, which is another CD8+ suppressor in the tumor microenvironment (58). Moreover, IL-10 secretion by TAM/Ms also stimulates Treg differentiation (59). A study by Takenaka et al. (60) reported that kynurenine produced by GBM cells activates aryl hydrocarbon receptor (AHR) in TAMs, which helps modulate their function and T-cell immunity. AHR drives the expression of the ectonucleotidase CD39 in TAMs, which promotes CD8+ T cell dysfunction by producing adenosine in conjunction with CD73, AHR, and CD39 expressed in TAMs. This participates in the regulation of the immune response, including CD8+ T cell regulation in GBM, and constitutes potential targets for immunotherapy.

TAM/Ms also play a significant role with regard to infiltration of CD8+ T cells in malignant tumors. Fibrosis is a major target for TAM/Ms to regulate CD8+ T cell migration (61, 62), led by extracellular matrix stiffness and collagen deposition. CCL2, produced in the glioma microenvironment, is essential for the recruitment of regulatory T cells and myeloid-derived suppressor cells (63), as well as macrophages that are dependent on colony stimulating factor-1 (CSF-1) for both differentiation and survival (64). Therefore, TAM depletion via CCL2 or CSF-1 inhibitors have been reported to increase CD8+ T cell migration and infiltration (65) as it helps overcome immunosuppression.

In addition, TAM/Ms are able to regulate T cell metabolism. Arginase belongs to a family of L-arginine enzymes and regulates arginine metabolism. Arginase overexpression by TAM/Ms leads to metabolic starvation of T cells through indoleamine 2,3-dioxygenase (IDO) pathway (59, 66).

Moreover, in melanoma and pancreatic cancer, TAM/Ms are able to intervene in CD8+ T cell function by expressing V-domain Ig suppressor of T cell activation (VISTA), which causes a suppressive CD8+ T cell response (67).

Interactions between TAM/Ms and CD8+ T cells are known to be involved across all stages of GBM development, including infiltration, differentiation, and tumor cell interaction. This complexity of the immune microenvironment interplay causes the heterogeneity of GBM tissue across different patients. Based on this, there are many promising treatment targets for GBM that are likely helpful for patients who cannot benefit from the current standard treatment of GBM ( Figure 1 ).

Given their physiology, TAM/Ms do not have the characteristic of eliminating tumor cells directly. Individually, the ideal role of CD8+ T cells in tumor tissue is to carry out cytotoxic activity against malignant cells. Likewise, there are a few novel therapies that have tried to reverse the process of innate and adaptive immune responses that are suppressed by the tumor microenvironment, either through direct or indirect schemes. This includes immune checkpoint inhibitor (ICI) or chimeric antigen receptor T cell (CAR-T) therapy. Several ongoing clinical trials have depicted great progress in GBM treatment with ICI (56, 68) and CAR-T therapy (69, 70). However, a large number of patients still do not respond to these treatments (71), largely because of the failure of CD8+ T cells to overcome the immune inhibitory microenvironment. TAM/Ms are key cells that generate and maintain this immunosuppressive environment in GBM. There are potential treatments for GBM that underlie this crosstalk between the immune effector CD8+ T cells and the immune supporter TAM/Ms. Finally, a majority of therapies that target the regulation of TAM/Ms for enhancing effector T-cell function in GBM are based on three methods, i.e., TAM/M depletion, TAM/M reprogramming, and other novel effective biomarkers in TAM functioning.

There has been significant discussion regarding the immune system in the CNS over the past decade. However, the real situation has not been clearly described. The crosstalk between TAM/Ms and CD8+ T cells is still very important in the GBM immune response. Many potential targets have been studied for improving the prognosis of GBM patients and for disease prevention. Several studies have been successfully translated into clinical practice with promising results. Nevertheless, GBM, a malignant disease of the CNS, is still hard to treat. More studies are required to uncover the mechanisms associated with the symphonic immune interactions in GBM.

In this review, we discussed the role of TAM/Ms and CD8+ T cells in GBM, as well as the interactions between them. TAM/Ms in GBM are dominant in their anti-inflammatory phenotype, which is the main effector responsible for creating an immunosuppressive GBM microenvironment. TAM/Ms are able to inhibit the cytotoxic effect of CD8+ T cells through a multitude of biological pathways and cytokines. This suppressive process not only appears amidst development of GBM but also appears during the application of GBM treatment.

There are potential therapeutic targets that underlie the crosstalk between TAM/Ms and CD8+ T cells. Sabotaging TAM/M recruitment and migration can cause the depletion of TAM/Ms, which helps revive the cytotoxic effect of CD8+ T cells that can help lead the fight against GBM cells. CCL2 and CSF-1 inhibitor have shown significant potential for further clinical evaluation to improve CTL-related immune therapies, such as immune checkpoint inhibitor therapy, and CAR-T therapy. To reverse the immunosuppressive situation in the GBM tissue, reprograming TAM/Ms can help shift TAM/Ms from an anti-inflammatory phenotype, toward a pro-inflammatory phenotype, enhancing the CD8+ function of CTLs. IL-10, IL-12, and PI3Kγ have a significant function in this process. Furthermore, biological agents can be used to boost suitable therapies. Moreover, novel TAM factors have been tested with optimal results, as anti-tumor targets. Interactions between TAM/Ms and CD8+ T cells were considered to have a decisive power in GBM physiology, as well as immune therapy for GBM. Finally, the underlying targets for GBM therapy need to be further studied to clinically improve the outcomes of GBM patients.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

YZ and AS conceptualized the research project. ST, XL, AS, and JZ wrote the paper and made the original figures. ST, HW, YY, JQ, SH, and YW critically revised the texts and figures. AS, YD, and YZ supervised the research and led the discussion. All authors contributed to the article and approved the submitted version.

This work was funded by National Natural Science Foundation of China (81701144) and Zhejiang Provincial Natural Science Foundation of China (LQ19H160045).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.