Background
Lung cancer is the most common type of cancer worldwide and is associated with the highest mortality among all common cancers. Non-small cell lung carcinoma (NSCLC) is a major subtype of epithelial lung cancer and accounts for a majority of cases [
1]. Metastatic spread of cancer is the reason for most NSCLC-related deaths [
2]. Lung cancer cells frequently metastasize to some major organs, such as bone, brain, lung and liver [
3]. Although most NSCLC patients receive multiple treatments, including surgery, radiation, chemotherapy, targeted therapy, NSCLC still exhibits low cure rate, high recurrence and mortality [
4]. Attempts have been made to develop new strategies for the treatment for NSCLC [
5]. To date, immunotherapy has shown promise in patients with metastatic NSCLC.
In recent years, immune checkpoint therapy becomes a breakthrough strategy to reactivate antitumor immune responses [
6]. Programmed cell death protein 1 (PD-1) and its ligand PD-L1 axis, as an inhibitory immune checkpoint signaling pathway, play a crucial role in the progression of tumor. In particular, inhibiting PD-1 and its ligand PD-L1 axis has shown remarkably results in NSCLC treatment. Currently, two antibodies (nivolumab and pembrolizumab) against PD-1 and two (atezolizumab and durvalumab) against PD-L1 have been approved for treating advanced stage NSCLC [
7]. However, anti-PD-1/PD-L1 therapy remains challenging because of the low response rates [
8]. Multiple factors, including T-cell infiltration, neoantigen burden and tumor metabolism, are involved in the immune checkpoint blockade [
9].
B7-H3, also known as CD276, is a type I transmembrane protein that shares up to 30% amino acid identity with PD-L1 [
10,
11]. The receptor of the B7-H3 protein remains unclear. B7-H3 contributes to a co-inhibitory immune signal during modulation of cytotoxic lymphocyte function in cancer immunity [
12]. While B7-H3 protein is expressed at low levels in most normal tissues, it is aberrantly expressed on differentiated malignant cells and cancer-initiating cells, with limited heterogeneity, and in multiple tumor types, including lung, colon, breast and ovarian cancers [
13]. In head and neck squamous cell cancer and pancreatic ductal adenocarcinoma, high B7-H3 expression is observed on cancer-initiating cells. Moreover, it is overexpressed in the tumor vasculature and stroma fibroblasts [
14,
15]. B7-H3 plays an important role in tumor immune evasion and metastasis [
16]. In NSCLC patients, B7-H3 overexpression is frequently associated with lower level of tumor-infiltrating lymphocytes [
17]. B7-H3 expression may be correlated with EGFR gene expression status and efficacy of anti-PD-1 therapy [
18]. Recent evidence indicated that B7-H3 expression potentially involves in resistance to anti-PD-1/PD-L1 blockade in NSCLC and ovarian cancer [
18,
19]. B7-H3 represents an attractive target for antibody-based immunotherapy. However, the molecular mechanisms of B7-H3-modulation of cancer immunity are not clear.
Due to its broad expression across a panel of tumor types, B7-H3 becomes an attractive therapeutic target. Recently, several monoclonal antibodies (mAbs) targeting B7-H3 have generated promising results against pancreatic adenocarcinoma [
20], glioblastoma [
21,
22], pediatric solid tumors [
23,
24], and lymphoma [
25]. Specifically, recently developed anti-B7-H3 mAb, 8H9 (omburtamab) and its humanized forms, inhibited the growth of different B7-H3-positive tumor cells through antibody-dependent cell-mediated cytotoxicity (ADCC) or as immunoconjugates in preclinical [
26‐
29].
124I could be safely delivered with 8H9 by direct injection into human pontine gliomas for both PET imaging and therapy [
30], while
131I-8H9 administered to the cerebrospinal fluid showed potential in improving survival among patients with metastasis to the central nervous system and the leptomeninges [
24].
Adoptive immunotherapy that utilizes effector lymphocytes expressing tumor-specific antibodies is a promising approach to treat cancer [
31‐
33]. Genetic modifications using B7-H3 chimeric antigen receptors (CARs) gave promising results in xenografts of pediatric tumors [
20,
23], glioblastoma [
21] and melenoma [
34]. To facilitate immune cell responses, two modalities, CARs and bispecific killer cell engagers (BiKE), use single chain variable fragments (scFvs) to redirect cytotoxic lymphocytes against specific antigens that are expressed on tumor cells [
35]. These novel strategies have emerged as potentially curative therapies in the treatment for leukemia and some solid tumors. Here, we aimed to systematically evaluate the value of B7-H3 as a target in NSCLC via T cells expressing B7-H3-specific CARs and BiKE-redirected natural killer (NK) cells.
Methods
Human tissue samples
Either formalin-fixed paraffin-embedded or surgical tissue samples were obtained from the Second Affiliated Hospital, Guangzhou Medical University, and Zhuhai People’s Hospital, Jinan University. Buffy coat samples collected from healthy adult donors were obtained from Macau Blood Center. All procedures were in accordance with the ethical standards approved by the human ethics committees.
RNA-sequencing analysis
The B7-H3 mRNA expression in 110 normal lung tissues, 488 lung adenocarcinoma (LUAD) samples and 509 lung squamous cell carcinoma (LUSC) samples with different stages was analyzed using R packages ggplot2 and ggbeeswarm. Data were downloaded from the Cancer Genome Atlas (TCGA) (
http://www.oncolnc.org).
Cell lines and cell culture conditions
The HEK293, Daudi, MDA-MB-231, NCI-H23, HCC827, A549, BT-474, OVCAR-3, SK-OV-3 cell lines were obtained from the Stem Cell Bank, Chinese Academy of Sciences. HCC 1954 cell line was obtained from MSKCC. DLD-1, HCT 116 and PANC-1 cell lines were provided by Prof. Hang-Fai Kwok. SF188 and U251 cell lines are provided by Prof Gang Li. OCI-AML-3 and MOLM-13 cell line was provided by Dr. Tong-Kam Leung. These cell lines were cultured in either RPMI-1640, DMEM or DMEM/F12 supplemented with 10% fetal bovine serum (FBS) (GIBCO),100 U/mL penicillin and 100 µg/mL streptomycin (GIBCO) at 37 °C with 5% CO2. The FreeStyle 293-F cells were obtained from Invitrogen and cultured in Freestyle 293 expression medium (GIBCO) in the incubator shaker set at 125 rpm, 8% CO2, and 37 °C.
Antibodies
Antibodies for APC-CD3, PE-CD4, PE/Cy7-CD8, Alexa Fluor647-CD56, PE-CD107a, PE-CCR7, PE/Cy7-CD62L, PE/Cy7-perforin, PE-granzyme B, PE-PD-L1 were purchased from BioLegend. Goat anti-B7-H3 antibody (MAB1027) was purchased from R&D System. Humanized anti-B7-H3 IgG (8H9) was expressed and purified in the laboratory (for details, see Additional file
1). Mouse anti-CD16 antibody (3G8) was purchased from BD Biosciences. PE-conjugated anti-human Fc antibody and HRP anti-human IgG antibody were purchased from Invitrogen. HRP-conjugated rabbit anti-goat IgG antibodies were purchased from Jackson ImmunoResearch. Rabbit antihuman PD-L1 antibody (13,684), HRP-conjugated anti-β-actin and anti-GAPDH antibodies were purchased from Cell Signaling.
Generation of B7-H3-specific CAR-T cells
The DNA sequence of the single chain variable fragment (scFv) of 8H9 antibody was synthesized and then introduced into the pLVX lentivirus backbone plasmid, containing E1α promoter, a CD8 leader sequence, CD8-α transmembrane domain, a 4-1BB and a CD3ζ intracellular signaling domains. Lentiviruses carrying the B7-H3-CAR were produced in HEK293 cells that were co-transfected with B7-H3-CAR lentiviral vector and packaging plasmids (PsPAX2 and pMD2.0G) as previously described [
36]. The lentiviral supernatants were collected at 72 h after the transfection and filtrated through a 0.45 µm filter (Millipore), followed by centrifugation for 2 h at 28,000 rpm. For titer calculation, HEK293T cells (5 × 10
5 per well) were seeded into 12-well plate and cultured to reach 80% confluence. The different dilutions of lentiviruses were mixed with polybrene at 5 μg/ml in 1 ml fresh medium. The culture medium in the wells was replaced by the mixture of lentiviruses and incubated for 2 days. After harvesting, the infection percentages of HEK293T cells were counted based on Zsgreeen detection by the flow cytometer. The positive ratio of each well was recorded. The titer can be calculated from cell counting using the following formula:
$${\text{titer}}\left( {\text{TU/ml}} \right) = \frac{{{\text{Total}}\;{\text{cell}}\;{\text{number }} \times {\text{positive}}\;{\text{rate}}}}{{{\text{added}}\;{\text{volume}}\;{\text{of }}\;{\text{virus}}\left( {{\mu l}} \right){ }}} \times 1000.$$
Human peripheral blood mononuclear cells (PBMCs) were isolated from buffy coats using Ficoll reagents (GE Healthcare) and cultured in complete RPMI-1640 supplemented with 2 mM L-glutamine, 100U/mL penicillin, 100 μg/mL streptomycin, 10% FBS and 100U/mL r-human IL-2 (rhIL-2) (CellGenix). PBMCs were stimulated by Dynabeads Human T-Activator CD3/CD28 (Gibco) at a 1:1 T cell: bead ratio in the growth medium supplemented with rhIL-2. One day after activation, the activated T cells were transduced with the lentivirus at a multiplicity of infection (MOI) of 5. Spin infection was performed with polybrene at 570 × g for 1 h at 32 °C, followed by incubation for 2 days. The anti-CD3/CD28 beads were magnetically removed on day 7. T cells were expanded in complete RPMI-1640 until used in vitro or in vivo assays. Cells were fed every 2 days and used within 20 days of expansion in all experiments. Vehicle control T cells were produced in the same conditions.
Production of the B7-H3/CD16 BiKE
The scFv of the humanized anti-B7-H3 antibody 8H9 and the scFv of anti-CD16 antibody 3G8 were generated by coupling of heavy chain variable region (VH) and light chain variable region (VL) via the (GGGGS)3 linker, respectively. The scFvs were cloned into pComb3x vector. To generate a BiKE targeting B7-H3 and CD16, the anti-B7-H3 scFv 8H9 and anti-CD16 scFv 3G8 were linked with an additional (GGGGS)3 linker and then cloned into the pSecTag B expression vector. Freestyle 293-F cells were used to express bispecific antibodies. Transfection into HEK293 cells was performed as described previously [
37]. The soluble scFv was expressed and purified as previously described [
26]. The anti-B7-H3 x CD16 bsAb, anti-B7-H3 scFv and anti-CD16 scFv were purified using Ni–NTA agarose beads (Qiagen).
Cytotoxicity assays
Cytotoxicity of CAR-T cells was measured using the Calcein-AM release method as previously described with modifications [
38]. Targeted cells at 1 × 10
6 cells/mL were incubated with 10 μM of Calcein-AM for 30 min at 37 °C. For CAR-T cells, targeted cells seeded at 1 × 10
4 cells/well in the 96-well plate were co-incubated with effector CAR-T cells at different effector-to-target (E:T) ratios from 5:1 to 40:1 in a total volume of 200 μL for 4 h.
ADCC assay was performed using PBMC as effectors as previously described with modifications [
39]. Targeted cells were labeled with Calcein-AM. Different concentrations of antibodies were incubated with the mixture of tumor cells and PBMC at a 10:1 E:T ratio in a total volume of 200 μL for 4 h. The spontaneous release control wells and maximum release target control wells were set up in all experiments. Mean fluorescence intensity (MFI) was measured using PerkinElmer Multimode Reader at 495/515 nm. The specific lysis ratios were calculated according to the formula:
$${\text{lysis}}\;{\text{ratio }}\left( \% \right) = \frac{{{\text{MFI }}\left( {{\text{sample }}\;{\text{lysis}}} \right) - {\text{MFI }}\left( {{\text{spontaneous}}\;{\text{ release}}} \right)}}{{{\text{MFI }}\left({{\text{maximum}} \;{\text{lysis}}} \right) - {\text{MFI }}\left( {{\text{spontaneous}}\;{\text{release}}} \right)}} \times 100\% .$$
Xenografted mouse models
All animal studies were conducted using protocols approved by the Animal Ethics Committees, University of Macau. Six- to eight-week-old female NOD/SCID mice were bred in the Animal Facility at Faculty of Health Sciences. Six- to eight-week-old female NOD/SCID mice were randomly divided into groups (n = 5 per group). Each mouse was subcutaneously inoculated with 100 μL PBS containing 1.5 × 106 tumor cells. To investigate the therapeutic effect of B7-H3-specific CAR-T cells, 5 days after tumor inoculation, mice in each group were intravenously treated with PBS, PBS containing 1 × 107 CAR-T cells or vehicle T cells per mouse on day 0 and day 7, respectively. To investigate the therapeutic effect of anti-B7-H3 × CD16 BiKE, antibodies (BiKE, B7-H3 scFv, CD16 scFv, 100 μg per mouse) were administered intravenously twice a week for a total of four doses. In animals that also received human effector cells with or without antibodies, PBMCs from healthy donors were injected intravenously into mice (1 × 107 cells per mouse). Tumor growth was measured using digital calipers twice a week, and tumor volumes were calculated using the formula V = ½(length × width2).
Flow cytometry
Human B7-H3 expression on cell surface of tumor cells was detected using anti-B7-H3 IgG 8H9, followed by a fluorescently-labeled anti-human Fc antibody. For cytokine measurement of CAR-T cells, target cells (A549 and HCT 116) were seeded at 1 × 105 cells/well in a 96-well plate and co-incubated with effector B7-H3-CAR T cells at an E:T ratio of 10:1 for 6 h or 24 h. Cytokine (IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNF-α) secretion in the culture supernatants was measured using BD™ Cytometric Bead Array Human Th1/Th2 Cytokine Kit II (BD Biosciences) following the manufacture’s instruction. To determine T cell differentiation, CAR-T cells were co-incubated with tumor cells (A549 and HCT 116) in a E:T of 10:1 for 6–24 h, followed by antibody staining for CD3, CD4, CD8, CD107a, CCR7, CD62L, and apoptosis dyes (7-AAD and Annexin V). For measurement for intracellular perforin and granzyme B, T cells were fixed and permeabilized, followed by stained with an anti-perforin antibody and an anti-granzyme B antibody, respectively. All samples were analyzed using the CytoFLEX S (Beckman Coulter) or Accuri™ C6 (BD Biosciences) flow cytometry. Data were processed using FlowJo software.
Dynamic apoptotic detection by FRET
A fluorescence resonance energy transfer (FRET)-based measurement of caspase-3 was used to determine the real-time apoptosis in tumor cells [
40]. Briefly, A549 cells carrying caspase-3 biosensor (A549-C3) were seeded 1 × 10
5 cells/well in the 24-well plate and incubated at 37 °C, 5% CO
2 to grow until confluence. After 1 day, 1 × 10
6 PBMCs with or without 5 μg/mL BiKE were added to the plate. Then, cells were incubated for different periods to observe the time-dependent FRET changes. The fluorescent images of living cells were acquired using a fluorescence microscope (Zeiss Axio Observer 7) with an excitation at 436 ± 10 nm, and the emission for YFP detection at 535 ± 12.5 nm and CFP at 480 ± 15 nm. Imaging data were analyzed using ZEN software (Zeiss).
Immunohistochemistry
For immunohistochemistry (IHC), the samples were fixed with 10% formalin and processed for paraffin embedding. Sectioned slices were deparaffinized in xylene and rehydrated in graded alcohol, and placed in Tris-buffered saline (TBS) for 15 min. After the antigen retrieval and inactivation of endogenous peroxidase, sections were blocked with animal nonimmune serum (Maxvision) and incubated with goat anti-human B7-H3 primary antibodies MAB1027 and 8H9 overnight at 4 °C, respectively. After the incubation with corresponding secondary antibodies (Maxvision) for 15 min, the sections were stained using the detection kit (Maxvision). Cell nuclei were stained with hematoxylin (Sigma). Finally, sections were dehydrated with absolute ethanol. The pictures of samples were captured in the Olympus TH4-200 microscope.
ECAR and OCR measurements
XF96 glycolysis stress test and mito stress test were performed using Seahorse XFe96 Extracellular Flux Analyzer (Agilent) to measure the extracellular acidification rate (ECAR) and oxygen consumption rate (OCR). One day prior to the assay, A549 cells were seeded 7000 cells/well in a 96-well Seahorse plate. A549 cells were treated with 100 nmol/L anti-B7-H3 antibody 8H9 or control antibody for 24 h. Before measurements, the growth medium was replaced with XF assay medium and the cells were incubated in a CO2-free incubator for 1 h. After the tests, cells were lysed with lysis buffer (0.1% triton, 10 mM Tris–HCl) and Bradford reagent was utilized to determine the total amount of protein in each well. Final concentrations of reagents utilized in the glycolysis stress were as follows: 10 mmol/L glucose, 2 μmol/L of oligomycin, and 50 mmol/L of 2-deoxy-D-glucose (2-DG) in each group. Final concentrations of reagents utilized in the mito stress test were as follows: 2 μmol/L of oligomycin, 2 μmol/L of FCCP, and 0.5 μmol/L of antimycin A/rotenone in each group.
Intracellular ROS determination
Intracellular reactive oxygen species (ROS) was measured using 2,7-dichlorofluorescin diacetate as a probe. A549 cells were harvested and incubated with BiKE and control antibodies for 24 h. Then, cells were placed in growth medium, loaded with the probe for 15 min and then washed once with PBS. Fluorescence intensity was measured using the Accuri™ C6 flow cytometry.
Statistical analysis
All statistical analyses were performed using GraphPad Prism software. Data are represented as mean \(\pm\) standard deviation or individual values. Significant differences were calculated using the two-way ANOVA, Student’s t tests, nonparametric Mann–Whitney test, or log-rank test. P values are represented as: *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001.
Discussion
Despite great progress in the development of immune checkpoint inhibitors, particularly the encouraging therapy of PD-L1/PD-1 blockade, a portion of NSCLC patients remain generally resistant to these therapies [
42]. There is an ongoing research aimed to discover novel immune checkpoint biomarkers and develop novel immunotherapy approaches. Overexpression of B7-H3 in tumors and its low levels in normal tissues have been described [
14,
43]. Recent studies have demonstrated that B7-H3 overexpression was associated with poor prognosis in NSCLC patients [
16,
17]. Consistently, this study showed that B7-H3 overexpression was significantly associated with 5yr-OS of NSCLC patients. We demonstrated that B7-H3 was tightly expressed in human NSCLC. The expression of B7-H3 in tumor and normal tissues was analyzed based on the data from the TCGA and IHC. According to the TCGA data, the mRNA expression levels of B7-H3 were highly associated with the malignancy grade of NSCLC. The IHC results further showed that 76% of NSCLC tissues were positive for B7-H3, which was consistent with the previous studies [
18]. B7-H3 expression may be correlated with tumor metastasis, because moderate expression of B7-H3 was also observed in paracancerous tissues, although the intensity was weaker than that in tumor tissues. Thus, this finding has further confirmed utility of B7-H3 as an oncoimmunology target.
Although the regulatory mechanisms of B7-H3 have not been completely elucidated, B7-H3 function closely correlates with that of a cytotoxic lymphocyte. As described in previous studies, anti-B7-H3 antibody blockade augmented antitumor immunity of CD8
+ T cells and NK cells [
12,
18]. The anti-B7-H3 mAb 8H9 (omburtamab) has showed clinical potential in promoting Fc-dependent NK cell through ADCC [
26]. NK cells are the key component in ADCC and play a role in immunosurveillance and the prevention of tumor metastasis [
44,
45]. Unlike T cell, NK cells are devoid of receptors for tumor-associated antigens. Solid tumors are relatively resistant to NK cytotoxicity due to lack of NK activation ligands [
46]. Tumor metastasis is often associated with a low NK cell activity [
47,
48]. NK dysfunctions promote metastasis in several human malignancies [
49,
50]. In this study, to overcome resistance, we examined the ability of the B7-H3/CD16 BiKE to redirect NK cells to destroy tumor cells. In fact, BiKE triggers activation of resting NK cells through CD16a signaling and thereby induces the secretion of cytokines and degranulation against B7-H3-positive tumor cells. Importantly, due to the direct binding of antibody to CD16a, NK cell cytotoxicity induced by the BiKE was superior to the Fc-mediated ADCC of anti-B7-H3 8H9 IgG. In two xenografts involving A549 (high level of B7-H3) and NCI-H23 (low level of B7-H3), our results showed that BiKE was effective to suppress NSCLC growth in both models as was demonstrated with reduced tumor volume (75–80%), The in vivo activity of BiKE was not limited by B7-H3 antigen density of NCI-H23 at the low level. We assume that a threshold of B7-H3 density is required to induce NK activity. The NK cell effector mechanisms involve caspase-dependent apoptosis [
51]. Dynamics of tumor cell death induced by NK cells was observed using FRET-based quantitative live cell imaging system in which activation of caspase-3 was used as indication of apoptosis. As NK stimulation by BiKE leads to significant increases in apoptotic target cell death, this further confirmed the ability of BiKE to actively propagate signals for NK cell activation.
In order to exploit the potential of 8H9 as CAR, we described the effects of the B7-H3 CAR-based 8H9 in cytotoxicity, cytokine production and inhibition of tumor growth. Our results demonstrated that key cytokine levels and T cell degranulation were enhanced. B7-H3 blockade promoted T cells differentiation into the naïve and central memory T cells, reflecting the increase of T cell persistence. Indeed, the B7-H3 CAR of 8H9 showed significant antitumor activity in xenografts of NSCLC and colon cancer cells. The on-target off-tumor toxicity is a major safety concern in CAR-T therapy [
52]. Efforts have been made to limit the toxicity, including infusing low dose of CAR-T cells [
53] and optimizing antibody affinity for the target [
54]. Our results showed that the B7-H3 CARs are expected to have minimal off-target toxicity. Because B7-H3 expression is limited in normal tissues, B7-H3 CAR T cells mainly extravasated into solid tumors but not into normal tissues. Consistent with this observation, we did not find evidence of damage in normal tissues. It suggests that tumor regional infiltration of B7-H3 CAR-T cells may limit systemic toxicity by reducing the cross-reactivity of CAR-T cells in other organs.
Combination strategies of immune checkpoint inhibitors provide significant benefit to some cancer patients [
55,
56]. In NSCLC patients, B7-H3 was associated with adaptive resistance to anti-PD-1 therapy [
18]. An anti–B7-H3 mAb, MGA271, has been assessed in a Phase I trial (NCT02475213) in combination with anti-PD-1 antibodies (pembrolizumab) in patients with melanoma. Combining blockades of B7-H3 and PD-1/PD-L1 resulted in the strength of therapeutic effects mediated by CD8 + T cells in the mouse ovarian cancer and pancreatic cancer models [
12,
18,
57]. In this study, IHC staining showed that approximately 30% of patients with NSCLC had co-expression of B7-H3 and PD-L1. On the other hand, a recent study demonstrated that PD-L1 expression by T cells weakened antitumor immunity through suppressing effector T cells and macrophages in tumor microenvironment [
58]. In our studies, PD-L1 expression was found in B7-H3 CAR-T cells irrespective of PD-L1 expression in tumor cells (Additional file
1: Fig. S13). We speculate that T-cell-expressed PD-L1 is involved in modulation of B7-H3 CAR-T cells. In the preliminary experiments, one anti-PD-L1 antibody improved cytotoxicity of B7-H3 CAR T cells at the high E:T ratio of 40:1 (Additional file
1: Fig. S14), suggesting anti-PD-L1 blockade should be able to prevent the immune-suppressive signaling of PD-L1 expressed by either tumor cells or T cells. In other experiments, we observed the increased productions of IL-2 and TNF-α by CAR T cells that were treated with anti-PD-L1 antibody (unpublished data). However, the role of PD-L1 has not yet been defined for B7-H3 CAR T cells. Future studies are needed to determine whether there is a correlation between T cell-associated PD-L1 and antitumor effects of B7-H3 CAR T cells in vivo. Unlike in T cells, PD-1 is not highly expressed in NK cells [
59]. In contrast, BiKE effectively controlled tumor growth by modulating NK cells probably due to the limited PD-1 levels. Therefore, it may be interesting to investigate further the therapeutic efficacies of dual blockade of B7-H3 and PD-L1 in NSCLC.
Elicitation of cellular mechanism behind the anti-B7-H3 blockade is important to reveal the immune-independent functions of B7-H3. The Warburg effect is a metabolic hallmark of tumor cells [
60]. Although recent studies suggested that B7-H3 modulated glucose metabolism in breast cancer and colorectal cancer [
61], the effects of B7-H3 on aerobic glycolysis in NSCLC remain unknown. Unlike other anti-B7-H3 antibodies, 8H9 can recognize the FG loop of B7-H3 [
26], which may be related to the regulatory functions of B7-H3. Here, we demonstrated that anti-B7-H3 blockade by 8H9 antibody switched cell metabolism from glycolysis to oxidative phosphorylation. The alteration of glucose metabolism probably was associated with cellular ROS-mediated pathway, consistent with the previous reports [
61]. These results clearly showed that, in addition to immune-regulatory function, B7-H3 possessed immune-independent functions in NSCLC.
In summary, our study shows that B7-H3 is an attractive target in NSCLC since it is highly expressed in 76% of the tumor specimens. Recent advances and positive clinical results from bispecific antibodies and CAR T cells have shown hope and excitement. Two treatment modalities, namely the B7-H3 CAR and BiKE derived from the anti-B7-H3 mAb 8H9, provided an effectively control of the tumor growth in NSCLC through promoting immune cell response. Safety and immunotoxicity are considered for these immunomodulatory therapeutics. Although CAR T-cell therapy has demonstrated significant antitumor activity, it can be frequently associated with cytokine release syndrome (CRS). A significant advantage of BiKE is to present low risks of systemic toxicity [
62]. In addition, another consideration is off-target toxicity that can cause dangerous side effects and is a major cause of clinical trial failure. Therefore, further research should explore the potential of anti-B7-H3 CAR and BiKE against NSCLC in the future preclinical and clinical studies.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.