Prospects of antibodies targeting CD47 or CD24 in the treatment of glioblastoma

Abstract Glioma is a malignant tumor with the highest incidence among all brain tumors (about 46% of intracranial tumors) and is the most common primary intracranial tumor. Among them, glioblastoma (GBM) is highly malignant and is one of the three refractory tumors with the highest mortality rate in the world. The survival time from glioblastoma diagnosis to death is only 14–16 months for patients with standard treatment such as surgery plus radiotherapy and chemotherapy. Due to its high malignancy and poor prognosis, in‐depth studies have been conducted to explore effective therapeutic strategies for glioblastoma. In addition to the conventional surgery, radiotherapy, and chemotherapy, the glioblastoma treatments also include targeted therapy, immunotherapy, and electric field treatment. However, current treatment methods provide limited benefits because of the heterogeneity of glioblastoma and the complexity of the immune microenvironment within a tumor. Therefore, seeking an effective treatment plan is imperative. In particular, developing an active immunotherapy for glioblastoma has become an essential objective in the field. This article reviews the feasibility of CD47/CD24 antibody treatment, either individually or in combination, to target the tumor stem cells and the antitumor immunity in glioblastoma. The potential mechanisms underlying the antitumor effects of CD47/CD24 antibodies are also discussed.

There have been several potential treatments for glioblastoma, such as immune checkpoint inhibitors and CAR-T. However, the enthusiasm toward immunotherapy has been dampened because not all patients with glioblastoma benefit from these treatments. Studies on the microenvironment of glioblastoma indicate that the number of microglial cells and macrophages in the tumor exceeds the infiltrating T cells. 3 The lack of T cells in the tumor microenvironment is different from other tumor types, such as melanoma or lung cancer. 4 Therefore, myeloid-derived cells may be the key to glioblastoma, and controlling their differentiation and polarization may bear the same importance as activating adaptive immunity.
It is recently discovered that the tumor conveys the "don't eat me" signal of innate immune surveillance through CD47-SIRPα 5 and CD24-Siglec-10 6 action. In preclinical studies, antibodies targeting CD47/CD24 yield encouraging results in various types of tumors. [7][8][9][10][11] A variety of anti-CD47 antibodies have entered clinical trials (Table 1). Further studies have found that CD47 can affect the polarization of tumor-associated macrophages, 12 while CD24 has no relevant reports. The M2-type polarization of tumor-associated macrophages can promote tumor growth, invasion, blood vessel formation, etc. 13,14 In addition, CD47/CD24 is expressed in tumor stem cells and cause the emergence of tumor resistance and promote tumor recurrence. 15,16 The primary target cells of anti-CD47/CD24 antibodies are microglia/macrophages. 6,9 Thus, we believe that these antibodies may trigger an antitumor immune response by activating myeloid innate immune cells. When used together with immune checkpoint inhibitors that activate systemic immunity, this treatment could offer surprising effects.

| G LIOB L A S TOMA
Glioblastoma (GBM) is a common malignant tumor that originates in the brain. According to CBTRUS (Central Brain Tumor Registry of the United States), glioblastoma accounts for 14.9% of all brain tumors in the United States. This tumor is characterized by its prominent invasiveness and poor prognosis. The 5-year survival rate for GBM is as low as 5.5%. 17 At present, the standard treatment of GBM is mainly the extensive surgical resection, supplemented by radiotherapy and temozolomide chemotherapy. 18 An array of new biomarkers for glioblastoma has been identified recently. [19][20][21] The mechanisms underlying tumor cell growth and invasion have been elucidated. 22,23 New treatment strategies for glioblastoma have been under active research, including the targeted therapy based on molecular biomarkers, [24][25][26] alternating electric field therapy that acts on mitosis of tumor cells, 27,28 and immunotherapy that targets different aspects of tumor immunity. [29][30][31][32][33] Based on current research results, we need to pay attention to two key factors that affect the effectiveness of glioblastoma treatment: tumor stem cells (TSC) and tumor immune microenvironment.

| Glioma tumor stem cells
In 2016, the World Health Organization classified glioblastoma into four different subtypes based on gene mutations and high expression of specific biomarkers: anterior, neurological, classic, and interstitial glioblastoma. 34,35 Glioblastoma is viewed as an aggressive tumor with substantial heterogeneities. 36 41 They found that the cloning, self-renewal, continuous xenotransplantation, and differentiation potential of TRTICs are surprisingly similar to tumor stem cells. Furthermore, they discovered that TRTICs could tolerate both radiotherapy and chemotherapy, and these cells are characterized by the expression of surface markers such as CD44 and CD24. 41 It is believed that the vast heterogeneities of tumor cells contribute to the current failures in treating glioblastoma in the clinic. 42

Intrinsic Factors in Tumor Cells External Factors of Tumor Cells
Signaling via mitogen-activated protein kinases 131

Expression of immune checkpoint molecules
Acquired mutations encoding the phosphatase PTEN 131 Infiltration by myeloid-derived suppressor cells 132 Activation of the WNTβ-catenin pathway 133 Desmoplastic tumor stroma (a barrier to lymphocyte infiltration) 134 Alterations signaling via the cytokine IFNγ 133 Loss of heterozygosity of loci containing genes encoding human leukocyte antigens 133 Downregulation of neoantigens 135

TA B L E 1 Regulatory factors of tumor immunity
Since these heterogeneous tumor cells may have been differentiated from TSCs, 43,44 a logical strategy for glioblastoma treatment would be targeting the TSCs. To do so, one must understand the mechanism by which TSCs escape from therapeutic targeting. An increasing body of evidence suggests that specific anti-apoptotic and pro-survival pathways are activated, and the drug effluxes increased in TSCs ( Figure 1A,B), which all contribute to the development of glioblastoma resistance to antitumor therapies. 45 Due to the functional differences between TSCs and their differentiated offspring cells at the transcriptome, epigenetics, and metabolic levels, 46 no single treatment is currently effective for glioblastoma. Therefore, the development of new therapies targeting TSCs may be the key to "eradicating" the tumor.

| Immunosuppression of glioblastoma
After the body recognizes a tumor antigen, antigen-presenting cells present it to effector cells. The effector cells then carry out immune responses to specifically kill tumor cells. This process, which is defined as tumor immunity, is quite complex and is regulated by multiple pathways. See Table 1 for details.
To achieve effective immunotherapy for glioblastoma, one must overcome two significant obstacles: 1. immune-privileged; and 2. immunosuppression.
It is previously believed that the central nervous system is an immune-privileged organ lacking antigen-presenting cells. Microglia cells present in brain tissue under noninflammatory conditions and play a limited antigen-presenting role but are less effective than the peripheral macrophages. 47 The expression of major histocompatibility complexes is reduced in microglia. 48 In addition, a single-cell sequencing research reported that tumor-associated macrophages overexpressed genes encoding MHCII components (H2-Aa, H2-Ab1, ANDH2-Eb1), as well as CD52, a costimulatory signal, which medi- ates T-cell activation and proliferation. There indirectly suggested that tumor-associated macrophages may also be involved in antigen presentation during tumor immunity. 49 Although activated microglia can present antigens to active lymphocytes, 13

| CD47 and tumor
CD47 is involved in regulating tumor invasion and metastasis, and the underlying mechanisms have been extensively studied. 78,79 Studies have shown that CD47 is overexpressed in almost all types of tumors and tumor stem cells, including gliomas, acute myeloid leukemia, non-Hodgkin's lymphoma, and breast cancer. This overexpression is positively correlated with poor prognosis. 5,15,80,81 Therefore, an anti-CD47 antibody therapy is warranted.
Numerous preclinical studies indicated that the monoclonal anti-CD47 antibody has an excellent antitumor effect. The underlying mechanism mainly includes the following: (1) CD47 antibody therefore inhibit tumor growth. 12 In a similar mechanism, CD47 antibody can promote the phagocytosis of glioma cells in the central nervous system by activating TA-MG. 86 Clinical trials testing the monoclonal anti-CD47 antibody have been carried out in a variety of solid cancers and hematopoietic cancers (see Table 2). The anti-CD47 antibody was used either alone or in combination with other antibodies (Rituximab, Cetuximab, Nivolumab) or adjuvant therapy. However, these combination treatment trials are primarily in phase I.

| CD47 and the treatment of glioblastoma
At present, the application of CD47 antibodies glioblastoma remains at the preclinical stage. The main reasons for the delayed testing of CD47 antibodies in clinical gliomas are related to the concerns about the antibody itself as described below and the possible resistance of gliomas to such therapy.

| CD47 antibody: concerns about safety and reliability
First, the off-target effect of the CD47 antibody includes the blockage of CD47 signaling in normal erythrocytes, thus increasing the expression of calreticulin, an "eating me" signal. 87 Meanwhile,

| Treatment resistance: Immune privilege and immunosuppression
Compared with other organs, brain parenchyma is separated from blood circulation by the blood-brain barrier. The lack of professional antigen-presenting cells in the brain makes antigen rec-  while inhibits cancer cell apoptosis. 122 In addition, CD24 has been proposed as a biomarker for the active proliferation of several types of cancer stem cells. Therefore, CD24 has attracted much attention as a potential molecule to target cancer cells/cancer stem cells.
Although CD24 has been extensively studied for its role in tumor growth, there are few reports identify CD24 as a therapeutic target.

| CD24 and the treatment of glioblastoma
The expression of CD24 is up-regulated in glioblastoma stem cells and functionally involved in the migration, infiltration, and metastasis of glioblastoma cells. 41,124 An overexpression of CD24 by more than two fold has been associated with poor overall survival in GBM, the poor survival may be related to increased "stemness" of tumor cells, 16 which provides a potential therapeutic target for glioblastoma.
From the translation perspective, the combined application of CD24 antibody with the CD47 antibody offers an additive effect against glioblastoma compared to either treatment alone. Further clinical evaluation on this combined treatment on glioblastoma is warranted.
The mechanism by which the anti-CD24 antibody promotes Another question is whether CD24 blockage with the anti-CD24 antibody leads to the over-activation of TLR-mediated proinflammatory reactions, which in turn triggers a cytokine storm, attacks normal cells, and results in autoimmune diseases. The latter is a particularly relevant concern for glioma therapy, as CD24 is widely expressed in brain parenchyma cells. The immunotherapy resistance of glioblastoma to CD24 antibody treatment needs to be further investigated ( Figure 2B).

| Perspectives
Immunotherapy has been developed as a novel treatment for glioblastoma. Such treatment needs the participation of the systemic immune system that orchestrates a strong and persistent cytotoxic response to brain glioma. However, due to the unique immune microenvironment within the central nervous system and in the glioblastoma, the development of clinical feasible immunotherapy for glioblastoma has been relatively slow.
With the increased understanding of the mechanisms underlying cancer immunology in the glioblastoma, especially the role of the "don't eat me" CD47-SIRPα signaling and CD24-Siglec10 signaling, we propose here that the combined blockage of these two pathways may effectively activate the innate immune responses toward glioblastoma. CD47 and CD24 are highly expressed in cancer stem cells.

ACK N OWLED G EM ENTS
This project has been supported by National Natural Science Foundation of China (NO. 81672824).

CO N FLI C T S O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
Hao Wu, Jialin Liu, Wen Yuan, and Zhifei Wang were responsible for literature review and investigation; Hao Wu and Jialin Liu were responsible for draft preparation; Wen Yuan was responsible for drawing; Ling Chen was responsible for manuscript review and editing. All authors have read and agreed to the published version of manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analyzed in this study.