Immunoglobulin superfamily surface molecules are implicated in the propagation of the stimulatory and inhibitory signals in the immune cell lineages ( Figure 1 ). The most widespread and thoroughly described immune checkpoint involves the Programmed cell death 1 (PD-1) and its ligand, Programmed cell death 1 ligand 1 (PD-L1, B7-H1). The interaction between PD-1 on the T cell surface and its ligand mediates multiple immunosuppressive effects such as apoptosis and functional exhaustion of conventional T cells (Tconv), reduced cytokine secretion, and generation of Tregs and TAMs (12). PD-L1 is expressed in numerous neoplasms, including brain tumors. PD-L1 overexpression was observed in about 90% of GBM tumor cells and GBM-associated macrophages (13). Moreover, GBM-infiltrating CD4⁺ and CD8⁺ lymphocytes have been shown to express both PD-1 and PD-L1, which indicates induction of Tregs and reprogramming of Tconv to the self-inhibiting lymphocytes (14). Tumor PD-L1 can be exposed on the surface of GBM-derived extracellular vesicles and delivered to the distant sites by bloodstream (15). The cytoplasmic region of PD-1 contains the immunoreceptor tyrosine-based switch motif (ITSM), which recruits Src homology region 2 domain-containing phosphatase-2 (SHP-2). Recruited SHP-2 mediates dephosphorylation of TCR-associated CD3 and ZAP70 and inhibits CD28 co-stimulatory signals. It leads to the deactivation of pathways such as PI3K/Akt and NF-κB, which results in reduced transcriptional activity, inactivation of the pathways downstream of the T-cell receptor (TCR) and lower IL-2 production (16). Another mechanism of PD-1/PD-L1 axis involves shielding the costimulatory molecules CD80 by PD-L1. PD-L1 has been shown to interact with CD80, thereby preventing its binding to CD28, which is required for the stimulation of T cells by antigen-presenting cells (APCs) (17).

The effects of PD-L2 (CD273), the second PD-1 ligand, on the immune activation are similar to PD-L1, but still not the same. However, unlike PD-L1, PD-L2 has only one receptor – PD-1. PD-L2 is not as widespread in tumors, which indicates its secondary role in forming the inhibitory TME (18). Nevertheless, PD-L2 overexpression has been detected in HGG and is associated with the wild-type status of isocitrate dehydrogenase 1 (IDHwt) and a highly invasive mesenchymal GMB phenotype. To sum up, overexpression of PD-1 and both ligands correlate with a large count of GBM-induced Tregs and a bad prognosis for patients (19–22).

Cytotoxic T-lymphocyte associated protein 4 (CTLA-4) plays a key role in tumor immune reactions with PD-1. CTLA-4 is homologous to the costimulatory T cell receptor CD28 and binds to the same ligands CD80 and CD86, but with a significantly higher affinity (23). Thus, tumor-infiltrating lymphocytes (TILs) express CTLA-4 to disrupt the costimulatory signaling by shielding CD80 and CD86 from CD28 in a manner similar to PD-L1 (23). At a molecular level, CTLA-4 signaling inhibits AKT phosphorylation and activation of the transcription factors, such as NF-κB, AP-1, and NF-AT, induced by co-stimulatory CD28 (24). Blocking CTLA-4 by monoclonal antibodies protects T cells from negative regulation and restores antitumor immune reactions. This makes CTLA-4 inhibitors promising antitumor agents, since CTLA-4 is involved in cancer development, including brain tumors ( Table 1 ) (35). Elevated CTLA-4 expression was detected in HGG patients mostly characterized by the IDHwt status and mesenchymal cell type. CTLA-4 overexpression leads to a lower survival of HGG and LGG patients. There is a strong correlation of CTLA-4 levels and enhanced tumor infiltration with Treg and inhibitory macrophages. CTLA-4 expression also correlates with the expression of other immune checkpoints such as PD-1, CD40, ICOS, and TIGIT (36).

ICOSLG (B7-H2, or CD275), the ligand of the inducible T cell costimulatory protein (ICOS), is expressed on the surface of somatic cells and APCs. Despite the positive costimulatory role of the ligand, the outcome of ICOS/ICOSLG binding depends on TME. During tumor development, ICOSLG can both promote and suppress tumor progression, since it activates Tconv cells and, at the same time, induces Tregs through NF-κB signaling (37). The TME is characterized by increased expansion and infiltration of Tregs and suppression of Teff functions (11), therefore, ICOSLG has a greater effect on Treg. ICOSLG is expressed on GBM tumor cells, its upregulation being associated with the presence of glioblastoma stem cells and IL-10-producing T cells as well as the mesenchymal phenotype. As a result, patients with ICOSLG overexpression have a lower overall survival (37). These data are in line with evidence that ICOS is overexpressed in glioma-infiltrating Tregs, as well as with the established link between high ICOS levels and a bad prognosis (38). Nevertheless, protein distribution in the body is also important. For instance, low ICOS levels in the blood plasma of GBM patients was associated with a global immunosuppression and the lower overall survival (39). ICOS/ICOSLG was shown to be co-expressed with a number of inhibitory immune checkpoints such as PD-1/PD-L1/PD-L2 (38).

B7 homolog 3 (B7-H3, CD276) is a type I transmembrane protein which exerts immunosuppressive activity by triggering T cell exhaustion. In healthy tissues, B7-H3 can be found on resting fibroblasts and osteoclasts, endothelial cells, activated T cells, natural killer cells (NK) and APCs. The inhibitory function of B7-H3 is widely used by tumors, including gliomas. Patients with IDHwt HGG have high levels of B7-H3 (40, 41). B7-H3 overexpression also correlates with a lower survival in LGG patients (42). Duerinck et al. studied the mutually exclusive expression profiles of B7-H3 and PD-L1 and suggested B7-H3 to be the major factor responsible for the failure of anti-PD-1 and anti-CTLA-4 HGG therapy (35).

The B7-H3 signaling cascade involves the activation of JAK2/STAT3 survival pathway leading to tumor growth and epithelial–mesenchymal transition in glioma cells. The exosomal transport of B7-H3 can also enhance tumor aggressiveness and facilitate immune escape in medulloblastoma (43) and neuroblastoma (44). Nevertheless, the role of B7-H3 in the immune response against GBM remains controversial. The receptor for B7-H3 has not been identified yet, but it is mostly likely present on the surface of activated CD4⁺ and CD8⁺ cells (45, 46). B7-H3 can have several candidate receptors, since B7-H3 was shown to act not only as an inhibitory molecule, but also as a stimulatory one (47).

B7 homolog 4 (B7-H4, VTCN1, B7x, B7S1) is a type I transmembrane protein of B7 family. Normally, B7-H4 is expressed by dendritic cells (DCs) and APCs. Although the B7-H4 overexpression was detected in several types of cancer, it is not considered as a typical tumor marker (48). B7-H4 levels have been shown to correlate positively with a tumor grade and a poor prognosis in glioma patients (49). B7-H4 production in GBMs has been shown to depend on IL-6 signaling via IL-6/JAK/STAT pathway activation and is associated with an elevated number of TAMs (50). B7-H4 levels does not correlate with expression of other immune checkpoints. For instance, B7-H4 and B7-H3 co-expression was observed only in 10% of GBM cases. Interestingly, B7-H4 and PD-L1 were co-expressed only in 2% of gliomas, making it most reasonable to assume a possible functional redundancy of these molecules (51). Similar, to B7-H3, the receptor for B7-H4 is still unknown.

VISTA (V-domain Ig suppressor of T cell activation), also known as B7-H5, is highly expressed in myeloid cells and TILs. VISTA acts as an activating ligand for APCs and an inactivating one for T cells (41). VISTA suppresses T cell proliferation and cytokine production by inhibiting NF-κB pathway (52). VISTA is commonly expressed in tumors and positively correlates with the WHO tumor grade and a poor prognosis for glioma patients (53). Moreover, VISTA is often co-expressed along with other inhibitor immune checkpoints such as B7-H3, PD-1, PD-L1, LAG-3, TIM-3 (54).

To date, two receptors for VISTA have been identified. VSIG-3 (IgSF11) is a member of the immunoglobulin superfamily which is highly expressed in gliomas. VSIG-3 is usually associated with high-grade malignancies and a worse outcome (55). The interaction between VSIG-3 and VISTA inhibits T cell proliferation and production of proinflammatory cytokines and chemokines (56).

PSGL1 (selectin P ligand), also known as SELPLG or CD162, is another receptor for VISTA. It has been suggested that PSGL1 stimulation may inhibit AKT and ERK signaling induced by TCR stimulation in some tumors. PSGL1 was detected on HGG and is co-expressed with VISTA (55). However, PSGL1 has not been studied thoroughly in the context of brain tumors.

B7 homolog 6 (B7-H6), or NCR3LG1, is a B7 family immune checkpoint protein which acts as an endogenous costimulatory ligand. The extracellular domain of NKp30 on the surface of NK cells is a receptor for the extracellular part of B7-H6 (57). B7-H6 binding to NKp30 induces NK cells activation. This protein is almost absent in normal tissues and mononuclear cells from the peripheral blood of healthy donors; however, it can be detected on the surface of neutrophils and proinflammatory macrophages in the presence of proinflammatory cytokines such as TNF- α, IL-1β, or TLR ligands (58). B7-H6 is also selectively expressed on a range of brain tumor cells such as human neuroblastoma (59), astrocytoma (60), and glioma (61). B7-H6 overproduction positively correlates with tumor aggressiveness and a poor prognosis. In gliomas, B7-H6 regulates a spectrum of biological processes such as proliferation, migration, invasion, survival, and cell cycle control by activating the PI3K/Akt, ERK/MAPK, and c-Myc/RNMT signaling pathways (62).

B7 homolog 7, also known as HHLA2, is not expressed in healthy tissues, except for the placenta, gut, kidney, breast tissues, and macrophages. HHLA2 is absent in the brain, even in glial cells and neurons; however, it was detected in endothelial cells. HHLA2 is highly expressed in tumors, and at low levels, it was found in LGGs and, less frequently, in HGGs (63). HHLA2 expression is downregulated with tumor progression. Moreover, HHLA2 overexpression is associated with the prolonged overall survival in GBM patients (63). HHLA2 was shown to interact with CD28H and stimulate T cell proliferation and cytokine production via AKT phosphorylation (64). Nevertheless, there is evidence indicating that high B7-H7 expression in other cancer types is associated with a poor prognosis. For example, HHLA2 was found to be highly expressed in osteosarcoma and colorectal carcinomas and positively correlated with metastasis and a poor prognosis (65, 66). It is assumed that HHLA2 has at least two ligands with opposing functions, making it in a way similar to B7-H3. CD28H or TMIGD2 is the confirmed HHLA2 ligand with stimulatory activity, while the second ligand with an inhibitory activity has not been identified yet (67).

LAG-3 (Lymphocyte-activation Gene-3, or CD223) is expressed on microglial cells (68). LAG-3 is closely related to CD4 and can bind to MHC II (69). LAG-3 triggers CD4⁺ T cell exhaustion and limits T cell proliferation by competing for Zn ions with Lck causing its dissociation from complex with TCR. Cleavable by ADAM10/17 proteases cytoplasmic C-terminus of LAG-3 contains domains rich in glutamic acid which are responsible for acidification and withdrawal of Zn. LAG-3 presence in TME was associated with an enhanced CD8⁺ T cells infiltration, PD-1⁺ TILs and PD-L1⁺ IDHwt glioma cells (70). LAG-3 overexpression correlated with a poor prognosis in LGG patients (71). However, the role of LAG-3 in HGG remains controversial. TILs in GBM TME were shown to express higher LAG-3 levels compared to lymphocytes from healthy donors (72). LAG-3 co-expression with CTLA-4, PD-1, and TIM-3 (73) is considered to be a risk factor in GBM patients based on bioinformatics studies (74). The role of LAG-3 alone on survival and prognosis in GBM patients is not clear.

The first described LAG-3 ligand is galectin-3 (Gal-3), a β-galactose-binding lectin involved in proliferation, cell adhesion, and apoptosis. Although galectin-3 is a proven immunomodulator, it is also considered as a glioma-related marker. Gal-3 expression was reported to correlate with the WHO grade of gliomas (75). Fibrinogen-like protein 1 (FGL1) is another functional LAG-3 ligand. Soluble FGL1 from the blood stream induces the surface LAG-3 and transmits an inhibiting signal to T cells (76). However, the role of FGL1 in glioma development is very complex and poorly understood (77).

CD155, also known as the poliovirus receptor (PVR), is a glycoprotein which belongs to the immunoglobulin superfamily. Its expression is inherent in malignant cells and is rarely found in normal tissues, except epithelial or endothelial cells. HGGs, including GBMs, are typically associated with CD155 overexpression (78) and with a lower survival rate. The same tendency was detected for LGG (79). CD155 has emerged as a tumor promoting antigen, upregulated on GBM and related to increased GBM aggressiveness and metastasis (80). The functions of this receptor were shown to depend on engaging ligands. It can activate NK cells by binding CD226 (T lineage specific activation antigen 1, TLisA1) and CD96 (Tactile) and, on the contrary, inhibit them by triggering TIGIT (81). In GBM, CD155 promotes TIGIT⁺ immune cell infiltration and the transition of the circulating NK cells to TIGIT⁺/CD226^- phenotype, while normally TIGIT⁺ NK cells are absent in the CNS and peripheral blood (82).

TIGIT is a co-inhibitory receptor which could bind both CD155 (with high affinity) and CD112 (nectin-2, with low affinity) (83). It is expressed on the surface of immune cells such as memory and activated T cells, Tregs, NK, and NKT cells. TIGIT binds CD155 with higher affinity compared to CD226, preventing NK cell stimulation and function via the CD155/CD226 pathway (84). CD112 is another co-inhibitory NK cell receptor, and its binding to TIGIT also contributes to inhibiting NK cells. This prevents IFN-γ secretion and cytolytic granule release by NK-cells (85). TIGIT is overexpressed in GBM TILs and peripheral blood T cells of patients with GBM as compared to lymphocytes from healthy donors. Nevertheless, in most patients, co-stimulating factor CD226 was also overexpressed in GBM-infiltrating immune cells along with TIGIT (82). It implies the possible competition for the ligand with prevalent CD155/TIGIT binding and subsequent inhibition of NK cell function (82). The co-expression of CD155 and PD-L1 was confirmed for tumor cells and TAMs (85), while TIGIT and PD-1 were upregulated on TILs and associated with poor overall survival (82, 86).

CD200, another member of the immunoglobulin superfamily, has recently been recognized as an immune checkpoint. This protein is expressed on various immune and stromal cells as well as tumor cells. In gliomas, CD200 facilitates tumor growth both in vivo and in vitro. Furthermore, the soluble form of this protein is carried to the cervical lymph nodes through the cerebral spinal fluid contributing to the suppression of lymphocytes (87). Soluble CD200 in the patient’s bloodstream contributes to systemic immunosuppression. The main mechanism of CD200-mediated immunosuppression is likely to be the switching of macrophage polarization toward the M2 phenotype and inducing myeloid-derived suppressor cells (MDSC) in TME (88). In GBM patients, a high plasma level of CD200 was associated with an increased accumulation of MDSCs (88). The role of this protein in GBM development is actively investigated, and to date, there are no studies reporting the link between CD200 expression and the prognosis in patients with brain tumors, further studies are required for better understanding of diagnostic/therapeutic potential of this molecule.

CD48 is an immune checkpoint, also known as the B-lymphocyte activation marker (BLAST-1) or signaling lymphocytic activation molecule 2 (SLAMF2). CD48 is expressed on cells of hematopoietic origin, especially on APCs. CD48 is a key molecule in immunological synapses and is essential for co-stimulation. It binds to CD2 and promotes T cell activation, as well as the function of granulocytes and NK cells (89). Despite the ability to activate immune cells, CD48 binding with the high-affinity receptor 2B4 (CD244, SLAMF4) results in NK cell dysfunction. CD48 expression was shown for several oncologic pathologies, particularly glioblastoma (90). CD48 upregulation in gliomas was associated with enhanced macrophage and T cell infiltration, the IDHwt status of mesenchymal subtype gliomas and a worse outcome. CD48 has a strong association with most checkpoints such as TIGIT, ICOS, TIM-3, but not with CTLA-4 and PD-L1 (91).

CD47, also known as integrin-associated protein (IAP), is a transmembrane protein from the immunoglobulin superfamily. Normally, CD47 regulates phagocytosis through the interaction with SIRP-α receptors on macrophages (92). Several malignancies including gliomas express CD47. CD47 signaling was associated with AKT phosphorylation and PI3K/Akt pathway which resulted in tumor maintenance and survival (93). CD47 was associated with glioma stem-like cells and predicts a worse prognosis for patients (94).

CD277 or BTN3A1 (Butyrophilin subfamily 3 member A1) is a member of the immunoglobulin superfamily typically expressed on T cells, B cells, NK, DCs, and tumor cells (95). The role of CD277 in tumor progression is still poorly understood, but it can bind particular variants of γδ TCR causing their activation and cytotoxicity. There is a study indicating a carcinogenic role of CD277 in gliomas. According to (96), IDHwt glioblastomas expressed higher CD277 levels compared to WHO grade II and III astrocytomas and oligodendrogliomas. CD277 upregulation was associated with multiple effects on immune system including increased macrophage, B cell, and T cell infiltration as well as CD8⁺ T cell exhaustion (96). Moreover, CD277 was co-expressed with TIM-3, IL-10, and FoxP3 which correlated with a poor prognosis. CD277-dependent activation of the IL-6/JAK/STAT3 pathway may explain its co-expression with TIM-3, which activates the same pathway in cancer cells (96). However, unexpectedly, CD277 in GBMs was reported to trigger the anti-tumor immune responses in γδ T cells (97).

HVEM, or TNFRSF14, belongs to the tumor necrosis factor receptor (TNFR) family. HVEM is expressed on epithelial and mesenchymal cells and on resting lymphocytes, Tregs, NK and myeloid cells. HVEM was shown to activate as well as inactivate immune responses depending on the ligand (98). HVEM has several ligands: BTLA, CD160, gD, LIGHT, and LTα3 (98). The interaction between HVEM and BTLA or CD160 inhibits T cell functions, whereas its binding to LTα or LIGHT results in T cell activation (99). The expression of HVEM, LIGHT and BTLA was detected in gliomas (100). Aggressive subtypes of gliomas were shown to upregulate HVEM. Using immunohistochemistry, HVEM in gliomas was found predominantly in the microvascular proliferation region and at the edges of the necrotic zone (100). High HVEM levels predict a poor outcome (100). HVEM^high GBM tumors tend to contain larger numbers of immune and stromal cells in glioma microenvironment compared to the tumors with a low HVEM level. In glioblastoma samples, HVEM expression was shown to coincide with TIM-3, PD-1, PD-L1, CTLA-4, LAG-3, and VISTA (100).

HVEM is the only reliably identified BTLA receptor. BTLA (also known as CD272) is a transmembrane glycoprotein and the main inhibitory receptor on T cells. BTLA is found on the surface of immune cells, such as B and T lymphocytes, NK and NKT cells, myeloid cells. The interaction between BTLA and HVEM induces a branching signal, which promotes a proinflammatory signal by activating NF-κB (100), and, simultaneously, passes an inhibitory signal by recruiting tyrosine phosphatases SHP-1 or SHP-2, similarly to PD-1 (101). BTLA and HVEM molecules interact when expressed in cis on the membrane of the same cell or in trans on different cells (102). Interestingly, upon the cis-interaction of BTLA and HVEM, the inhibitory function of BTLA prevails over the activating function of HVEM (102). Moreover, BTLA/HVEM cis-binding shields HVEM molecules from BTLA molecules in trans as well as from the stimulation by other activating ligands such as LIGHT (103). Currently, the role of BTLA in glioma development is not well studied and requires future research.

LIGHT, also known as TNFSF14, is another TNF superfamily member and a ligand for HVEM. As mentioned above, there is a strong correlation between HVEM and LIGHT expression in glioma microenvironment. LIGHT expression directly correlates with the glioma grade (104). Moreover, LIGHT overexpression has been associated with a highly aggressive tumor phenotype (IDHwt and mesenchymal subtype) (104). In GBM patients, it correlates positively with poor survival. There is evidence that LIGHT promotes tumor growth in gliomas in a HVEM-dependent manner (105). The LIGHT/HVEM pathway, similarly to the canonical TNF/TNFR pathway, activates NF-κB and PI3K through the TNF receptor-associated factors, triggering survival signaling and transcription of the inflammatory genes ( Figure 2A ) (98). LIGHT was found to be co-expressed with PD-1/PD-L1, TIM-3, B7-H3, and other inhibitory immune checkpoints (104).

CD70 is a well-studied TNF-like ligand that functions as a co-stimulatory molecule for T and B cells. CD70 is usually not expressed in a healthy tissue, except for peripheral blood leukocytes. It contributes to homeostatic signaling supporting lymphocyte survival in the absence of signals from TCR (106). CD70 levels are increased in several malignancies including 10% of primary LGGs and 35% of GBMs. GBMs and LGGs expressing CD70 have similar molecular characteristics and patient survival rates (107). Elevated CD70 has been linked to a poor prognosis in LGG patients with IDHwt. Most GBMs expressing CD70 have a mesenchymal phenotype, which negatively correlates with patients’ survival (107).

CD70 binds to the receptor CD27. CD27 is also a member of the TNFR family and is often presented on naïve and memory lymphocytes, NK cells, and mature DCs (108). As a co-stimulatory immune checkpoint, CD27 plays an essential role in survival and activation of T cells. CD70 expression stimulates tumor infiltration with immune cells, but it has no effect on CD27 expression. The possible explanation of this phenomenon lies in the CD70-dependent infiltration with macrophages instead of T cells (109) or the activation and generation of tumor Tregs triggered by CD27/CD70 (110). It is also possible that CD70 signaling can be mediated by an unknown inhibitory receptor on the T cell surface (107).

The CD40 glycoprotein and its ligand CD40L (CD154) are members of the TNF and TNFR superfamilies. CD40L is primarily expressed by the activated CD4⁺ T cells. CD40 expression is triggered by CD40L binding and is typical of B cells, macrophages, and DCs. Still, the roles of both receptor and ligand in tumor progression remain disputable. CD40 and its ligand were shown to be co-expressed on the surface of GBM cells (111). Interestingly, WHO grade III gliomas express higher CD40 and CD40L levels compared to GBMs. It was noted that the overexpression of both proteins could be associated with better overall and progression-free survival of GBM patients after tumor resection (112). However, high CD40 expression was detected in glioma biopsy samples and correlated with lesions and an increased vascularization (112). In another study, lower overall and progression-free survival rates were detected in LGG patients as well as patients with GBM expressing IDHwt and high levels of CD40. CD40 was upregulated in secondary gliomas as contrasted with primary gliomas. Werner and colleagues (113) did not find any correlation between CD40L expression and the overall survival rate; however, recent studies pointed out the negative correlation between CD40L levels and the disease outcome (114).

T cell immunoglobulin domain and mucin domain protein 3 or TIM-3, also known as Hepatitis A virus cellular receptor 2 (HAVCR2), is a surface receptor found in most lymphocytes and cells of myeloid origin. TIM-3 promotes CD8⁺ exhaustion and apoptosis, reduces IL-2 and IFN-ɣ production (115). TIM-3 is one of the most upregulated co-inhibitory immune checkpoints in cancer, especially in glioma. There is a proven positive correlation between TIM-3 and the WHO grade, the mesenchymal phenotype, and a worse prognosis (116). TIM-3 promotes tumor progression by inducing the macrophage migration and tumor-promoting M2 polarization via the IL-6/NF-κB pathway (116). TIM-3 is co-expressed with PD-1, LAG-3 (117), VISTA, PSGL1, and Galectin-9 (Gal-9) (55).

Gal-9 is the member of the galectin protein family. Gal-9 expression is typical of glioma and depends on the WHO grade and TIM-3 levels (118). Gal-9/TIM-3 interaction has been shown to induce exhaustion and apoptosis of Th1, but not Th2 cells ( Figure 2B ). Gal-9 also binds to PD-1 and shields it from PD-L1, promoting the resistance of TIM-3-positive T cells to cell death (119). Gal-9 co-expression with PD-L1 and their co-localization in some GBM cases serve as further evidence for the link between TIM-3/Gal-9 and PD-1/PD-L1 pathways as supported by published data (118). Gal-9 also was shown to correlate with the expansion of M2 macrophages and MDSCs in GBM tissues (120, 121).

A2AR, the immunosuppressive adenosine 2A receptor, is a member of the G protein-coupled receptor family. A2AR is expressed in the majority of immune cells, such as lymphocytes and cells of myeloid origin. This receptor tightly regulates adaptive immune responses via high affinity binding to adenosine. Their interaction triggers the cAMP/PKA/CREB pathway resulting in a reduced immune response (122). The adenosine/A2aR pathway was hijacked by tumor cells to evade the immune system. Currently, there is no evidence on the correlation between A2AR and a poor prognosis for GBM patients. Nevertheless, A2AR expression was reported to be a high-risk factor in the bioinformatics analysis of glioma samples (72). A2AR was upregulated in CD4⁺ and CD8⁺ glioma-infiltrating cells, its high levels being linked to the PD-1 and CD39/CD73 axis (72).

Although CD39 and CD73 are not fully recognized as immune checkpoints, they significantly contribute to tumorigenesis through ecto-5′-nucleotidase activity, which metabolize ATP to adenosine, and are commonly expressed on most B cells and monocytes and on some T cells. Under normal conditions, ATP is localized in the intracellular space, and its extracellular concentration grows during neuron release or in response to ischemia or hypoxia, which induces local inflammation (123). CD39 converts ATP to ADP and AMP, while CD73 converts AMP to adenosine, thereby switching the proinflammatory status of the microenvironment to the anti-inflammatory state (124). These events contribute to tumor growth, migration, and T cell function. CD73 expression was typically observed in tumor macrophages and Tregs, which directly inhibit the cytokine release and cytotoxic functions of CD8⁺ T cells (125). Importantly, according to the recent studies, CD39 and CD73 are usually co-expressed in tumor cells and their simultaneous action causes adenosine-dependent pro-tumor immune suppression (126). Downregulation of both CD39 and CD73 in TME correlated with a better prognosis for patients (127).

Indoleamine-2,3-dioxygenase (IDO1) is an immune checkpoint secreted molecule involved in tryptophane metabolism. IDO1 is widely expressed in various healthy tissues, including lung and gastrointestinal tract tissues, placenta, and immune cells. IDO1 can suppress T cell function and help to maintain the immune privileged status of some tissues like placenta and fetus (128). The immunosuppressive IDO1 pathway is involved in converting tryptophan into kynurenine. Tryptophan starvation activates general control nonderepressible 2 (GCN2), a serine/threonine kinase that phosphorylates eukaryotic initiation factor 2α kinase (eIF2α). These changes lead to decreased transcriptional activity and reduced fatty acid production (129). Kynurenine, in turn, activates the aryl hydrocarbon receptor and induces DC immune tolerance (129). IDO1 upregulation is commonly observed in cancer. In glioma, IDO1 levels positively correlate with the WHO tumor grade, IDHwt status, the mesenchymal subtype, and Treg expansion (130). IDO1 expression depends on IFN-ɣ release creating a trap for tumor-infiltrating effector and cytotoxic T cells (131). IDO1 is co-expressed with PD-L1, PD-L2, PD-1, CTLA-4, CD39, BTLA, and LAG-3. Taken together, these characteristics of IDO1 indicate a strong correlation between IDO1 expression and lower overall survival of GBM patients (130).