Increased B7-H3 expression in GBM and the correlation between AURKA and B7-H3. First, we investigated the differentially expressed genes (DEGs) between GBM (WHO-IV and isocitrate dehydrogenase [IDH] WT) and low-grade gliomas (LGG, WHO-II, and IDH mutant) via the TCGA and CGGA databases (Figure 1A). We identified 4,616 DEGs in the TCGA database and 3,659 DEGs in the CGGA database (P < 0.01 and |log [fold change]| > 1) (Figure 1A and Supplemental Figure 1, A and B; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.173700DS1). We performed an intersection analysis between the DEGs identified in both the TCGA and CGGA databases and the genes annotated under the Gene Ontology (GO) term for cell cycle (accession no. GO:0007049), resulting in the identification of 23 DEGs that were part of the cell cycle gene set (Figure 1B). Furthermore, these DEGs identified in both the TCGA and CGGA databases were intersected with immune checkpoint genes from the Sinobiological Database (https://www.sinobiological.com/category/immunecheckpoint-proteins-list). We identified 9 genes that were in the immune checkpoint gene set (Figure 1B). Moreover, we assessed the correlation between the 23 DEGs in the cell cycle gene set and the 9 DEGs in the immune checkpoint gene set in GBM (Figure 1C and Supplemental Figure 1C). Our data reveal that the inhibitory immune checkpoint CD276 was significantly correlated with most of the DEGs identified in the cell cycle gene set, including AURKA. Additionally, our findings revealed a fold change of approximately 3.03 for AURKA between GBM and LGG in the CGGA database (Figure 1D; P < 0.0001). In the TCGA database, the fold change in AURKA expression between patients with GBM and those with LGG was approximately 5.00 (Figure 1E; P < 0.0001). Furthermore, we conducted a comparative analysis using both the TCGA GBM database and Genotype-Tissue Expression Project (GTEx) data to examine differences in AURKA expression between GBM and normal brain tissues. Our results reveal a significant difference in AURKA expression between GBM and normal brain tissues (P < 0.0001, fold change of approximately 5.80; Figure 1F). Similarly, CD276 was more highly expressed in GBM than in LGG (Figure 1, G and H). Additionally, CD276 expression was elevated in tumor tissue compared with normal brain tissue (Figure 1I).

Correlation of AURKA with CD276 in GBM. (A) The filtering process of the DEGs between GBM and LGG in the TCGA and CGGA datasets. (B) Venn diagrams displaying the overlap of gene sets within the cell cycle and immune checkpoint pathways, respectively, on the basis of the DEGs in both the TCGA and CGGA cohorts. (C) Heatmaps displaying the correlations in the CGGA datasets between cell cycle genes among the DEGs in both the TCGA and CGGA cohorts and immune checkpoint genes among the DEGs in both the TCGA and CGGA cohorts. (D and E) AURKA mRNA expression in GBM (n = 190 for CGGA, n = 142 for TCGA) and LGG (n = 135 for CGGA, n = 419 for TCGA) samples from the CGGA dataset (D) and TCGA dataset (E). (F) AURKA mRNA expression in GBM (n = 163) and normal brain tissues (n = 207) from TCGA datasets and the GTEx database. (G and H) CD276 mRNA expression in GBM (n = 190 for CGGA, n = 142 for TCGA) and LGG (n = 135 for CGGA, n = 419 for TCGA) samples from the CGGA dataset (G) and TCGA dataset (H). (I) CD276 mRNA expression in GBM (n = 163) and normal brain tissues (n = 207) from TCGA datasets and the GTEx database. (J and K) Western blot analysis of AURKA and B7-H3 expression in peritumoral, LGG, and GBM samples (n = 4/group) with quantification (K); β-actin was used as the internal control. Statistical significance was assessed via 2-tailed unpaired Student’s t test (D, E, G, and H) and 1-way ANOVA followed by Tukey’s multiple-comparison test (K). The data are presented as the mean ± SD (D, E, G, H, and K). **P < 0.01, ****P < 0.0001.

To further validate this observation, we performed Western blotting analysis on 4 LGG tissue samples and GBM tissue samples in comparison with 4 peritumoral samples. We confirmed increased B7-H3 and AURKA protein expression in tissue samples obtained from 4 patients with GBM compared with those obtained from 4 patients with LGG and 4 peritumoral samples via Western blotting (Figure 1, J and K).

AURKA regulates B7-H3 expression through EGFR activation. Since our results demonstrate that AURKA expression was positively correlated with CD276 expression, we explored whether AURKA regulates B7-H3 expression. First, AURKA expression was detected in the human glioma cell lines LN18, U87-MG, U373, LN229, TJ906, M059K, and U251 (Supplemental Figure 2A), which were used as in vitro models. LN229 cells, which express low levels of AURKA, were used to overexpress AURKA with AURKA_cDNA_Flag (Supplemental Figure 2B). LN18 and U87-MG cells that expressed high levels of AURKA were used to silence AURKA via specific shRNA#1, shRNA#2 or shRNA#3, with shRNA#3 significantly decreasing AURKA expression in both LN18 and U87-MG cells (Supplemental Figure 2, C and D). Furthermore, we analyzed the proteomics profiling of GBM cell lines U87_MG treated with or without shRNA#3 targeting AURKA. Proteomic analysis revealed that B7-H3 expression was lower in the shAURKA#3 groups than in the shNC groups in the U87-MG cell line (Figure 2A). And we confirmed that silencing of AURKA (shAURKA#3) led to a significant decrease in the B7-H3 protein level in LN18 and U87-MG cells (Figure 2, B and C) and in the CD276 mRNA expression level in LN18 cells (Supplemental Figure 2E). Additionally, B7-H3 expression on the cell membrane decreased after AURKA was knocked down in U87-MG cells (Figure 2D). Similarly, knockdown of AURKA (siAURKA) significantly decreased B7-H3 protein expression in LHG cell lines (Supplemental Figure 2F). In contrast, overexpression of AURKA significantly increased CD276 protein and mRNA expression levels in LN229 cells (Figure 2, E and F, and Supplemental Figure 2G). Several studies have demonstrated the association between the upregulation of B7-H3 and EGFR signaling (20, 21). Thus, we speculated that AURKA regulates B7-H3 expression through EGFR signaling. The phosphorylation of EGFR at Try1068 (pEGFR/EGFR) changed significantly after the knockdown or overexpression of AURKA in U87-MG, LN18, and LN229 cells, whereas the total EGFR expression level did not change significantly (Figure 2, B, C, and E).

AURKA regulates B7-H3 expression. (A) Volcano plot of the proteomics data showing reduced expression of B7-H3 and SDCBP (red dots) in U87-MG cells treated with shRNA targeting AURKA. (B and C) The protein expression levels of B7-H3, total AURKA, SDCBP, p-EGFR (Y1068), and total EGFR in LN18 cells (B) and U87-MG cells (C) expressing shNC or AURKA-shRNA#3 were detected via Western blotting, and the quantifications are shown on the right. β-Actin was used as the internal control. (D) B7-H3 expression on the cell surface in U87-MG cells expressing shNC or AURKA-shRNA#3 was detected by flow cytometry, and the quantification of the results is shown on the right. Cells that were only stained with isotype control antibodies were used as the negative control (NC). (E) The protein expression levels of B7-H3, total AURKA, SDCBP, p-EGFR (Y1068), and total EGFR in LN229 cells expressing vector or AURKA_cDNA_Flag were detected via Western blotting, and the quantifications are shown on the right. β-Actin was used as the internal control. (F) B7-H3 expression on the cell surface in LN229 cells expressing vector or AURKA_cDNA_Flag was detected by flow cytometry, and the quantification of the results is shown on the right. Cells that were only stained with isotype control antibodies were used as the negative control (NC). All samples were biologically independent, and 3 independent experiments were performed. Comparisons were performed via 1-tailed unpaired Student’s t test (A) or 2-tailed unpaired Student’s t test (B–F). The data are presented as the mean ± SD (B–F). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Previously, AURKA was shown to promote EGFR phosphorylation by inhibiting the degradation of SDCBP (16). Our proteomics analysis also demonstrated that inhibition of AURKA by shRNA decreases SDCBP protein levels (Figure 2A). We confirmed that silencing AURKA (shAURKA#3) resulted in a significant reduction in the SDCBP protein in LN18 and U87-MG cells (Figure 2, B and C). Consistently, overexpression of AURKA significantly increased the level of the SDCBP protein in LN229 cells (Figure 2E).

To determine whether the phosphorylation of EGFR is the true effector of AURKA-mediated B7-H3 expression, the EGFR activator NSC228155 or epidermal growth factor (EGF) was used in AURKA-knockdown cells. Treatment with 10 μM NSC228155 for 8 hours significantly increased the phosphorylation of EGFR at Tyr1068 (pEGFR/EGFR) (Figure 3, A and D), which was followed by an increase in CD276 protein and mRNA levels after AURKA knockdown in the LN18 and U87-MG cell lines (Figure 3, A–F). Similarly, treatment with 500 ng/mL EGF for 24 hours also increased the phosphorylation of EGFR at Tyr1068 (pEGFR/EGFR) and increased CD276 mRNA and protein expression levels after AURKA knockdown in U87-MG cell lines (Figure 3, G and H).

AURKA regulates B7-H3 expression through EGFR phosphorylation. (A–F) The expression levels of various markers were analyzed in LN18 cells (A–C) and U87-MG cells (D–F) expressing shNC or AURKA-shRNA#3 treated with NSC228155 (10 μM, 8 hours) or DMSO. (A and D) Protein expression of B7-H3, total AURKA, p-EGFR (Y1068), and total EGFR was assessed by Western blotting. (B and E) The mRNA level of CD276 was measured by qPCR. (C and F) B7-H3 expression on the cell surface was analyzed using flow cytometry. (G and H) The expression levels of various markers were analyzed in U87-MG cells expressing shNC or AURKA-shRNA#3 with or without EGF (24 hours, 500 ng/mL). (G) Protein expression of B7-H3, total AURKA, p-EGFR (Y1068), and total EGFR was assessed by Western blotting. (H) The mRNA level of CD276 was measured by qPCR. (I–K) The expression levels of various markers were analyzed in LN229 cells expressing vector or AURKA_cDNA_Flag with erlotinib (24 hours, 60 μM) or DMSO. (I) Protein expression of B7-H3, total AURKA, p-EGFR (Y1068), and total EGFR was assessed by Western blotting. (J) The mRNA level of CD276 was measured by qPCR. (K) B7-H3 expression on the cell surface was analyzed using flow cytometry. The quantifications are shown on the right (A, C, D, F, G, I, and K). β-Actin was used as the internal control (A, D, G, and I). Cells that were only stained with isotype control antibodies were used as the negative control (NC) (C, F, and K). Statistical significance was assessed by 1-way ANOVA followed by Tukey’s multiple-comparison test (A–K). The data are presented as the mean ± SD (A–K). All samples were biologically independent, and 3 independent experiments were performed. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

In contrast, treatment with the EGFR inhibitor erlotinib (60 μM) for 24 hours suppressed the phosphorylation of EGFR at Tyr1068 (pEGFR/EGFR) and was accompanied by decreases in CD276 protein and mRNA expression levels in AURKA-overexpressing LN229 cells (Figure 3, I–K).

These findings confirm that EGFR phosphorylation plays a functional role in AURKA-mediated B7-H3 expression. The effect of AURKA on the activation of EGFR may result from maintaining the stability of SDCBP.

Alisertib inhibits glioma cell proliferation and induces B7-H3 expression via EGFR activation. Previous studies have demonstrated that alisertib, a clinically validated specific AURKA inhibitor, suppresses the growth of various glioma cell lines (22). To assess the cytotoxic effects of different concentrations of alisertib on glioma cells, we measured the viability of U87-MG and LN18 cells after 72 hours of alisertib treatment. Our findings revealed that alisertib treatment substantially inhibited the proliferation of U87-MG and LN18 cells in a dose-dependent manner. The 50% inhibitory concentration (IC50) of alisertib for U87-MG cells was 7.71 μM; the IC50 for LN18 cells was 1.64 μM (Supplemental Figure 3A).

Our results revealed that alisertib treatment upregulated B7-H3 protein expression by activating EGFR signaling in glioma cells. First, we confirmed that alisertib treatment increased B7-H3, SDCBP, and phosphorylated EGFR expression in glioma cells (Figure 4, A–F, and Supplemental Figure 3, B–D). Treatment with alisertib eliminated autophosphorylation of AURKA at Thr288 (T288), accompanied by a dramatic increase in AURKA protein and mRNA levels (Figure 4, A–D, and Supplemental Figure 3, C–F). The same results were observed in 293T cells, where upregulated B7-H3 expression was observed after treatment with alisertib (Supplemental Figure 3, G and H). However, there was no significant change in B7-H3 expression in normal human astrocyte (NHA) cells (Supplemental Figure 3, I and J), suggesting that alisertib treatment upregulated B7-H3 expression in tumor cells but not NHA cells.

The AURKA inhibitor upregulates B7-H3 expression. (A–D) The protein expression levels of B7-H3, SDCBP, total AURKA, p-AURKA (T288), and p-EGFR (Y1068) and the total EGFR in LN18 (A), U87-MG (B), LHG (C) and LS (D) cells treated with increasing concentrations of alisertib for 24 hours were detected via Western blotting, and the quantifications are shown on the right. β-Actin was used as the internal control. (E and F) B7-H3 on the cell surface of LHG (E) and LS (F) cells treated with alisertib (24 hours, 0.1 μM) or DMSO was detected via flow cytometry, and the quantification of the results is shown on the right. Cells that were only stained with isotype control antibodies were used as the negative control (NC). Statistical significance was assessed by 1-way ANOVA followed by Tukey’s multiple-comparison test (A–D) and by a 2-tailed unpaired Student’s t test (E and F). The data are presented as the mean ± SD (A–F). All samples were biologically independent, and 3 independent experiments were performed. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Next, we investigated how alisertib treatment increases B7-H3 protein levels. On the basis of these previous findings, we hypothesized that alisertib affects EGFR activation, which regulates B7-H3 protein levels. To test this hypothesis, we disrupted EGFR activation by administering the EGFR inhibitor erlotinib. We treated LN18 and U87-MG cells with alisertib in the presence or absence of erlotinib. Our results showed that alisertib increased AURKA expression and EGFR phosphorylation at Tyr1068 without erlotinib (Figure 5, A and D). Following alisertib treatment, the cells were treated with 60 μM erlotinib for 24 hours. As expected, erlotinib treatment reduced the phosphorylation of EGFR at Tyr1068 and decreased CD276 protein and mRNA levels in LN18 and U87-MG cells (Figure 5, A–F).

The AURKA inhibitor regulates B7-H3 expression through EGFR phosphorylation. (A) The protein expression of B7-H3, p-AURKA (T288), total AURKA, p-EGFR (Y1068), and total EGFR in LN18 cells treated with alisertib (24 hours, 5 μM), erlotinib (24 hours, 60 μM), or both was detected by Western blotting. (B) B7-H3 expression on the cell surface of LN18 cells treated with alisertib (24 hours, 5 μM), erlotinib (24 hours, 60 μM), or both was detected via flow cytometry, and the quantifications are shown on the right. Cells that were only stained with isotype control antibodies were used as the negative control (NC). (C) The mRNA level of CD276 in LN18 cells treated with alisertib (24 hours, 5 μM), erlotinib (24 hours, 60 μM), or both. (D) The protein expression levels of B7-H3, p-AURKA (T288), total AURKA, p-EGFR (Y1068), and total EGFR in U87-MG cells treated with alisertib (24 hours, 5 μM), erlotinib (24 hours, 60 μM), or both were detected by Western blotting. (E) B7-H3 expression on the surface of U87-MG cells treated with alisertib (24 hours, 5 μM), erlotinib (24 hours, 60 μM), or both was detected by flow cytometry, and quantifications are shown on the right. Cells that were only stained with isotype control antibodies were used as the negative control (NC). (F) The mRNA level of CD276 in U87-MG cells treated with alisertib (24 hours, 5 μM), erlotinib (24 hours, 60 μM), or their combination. The quantifications are shown on the right (A and D). β-Actin was used as the internal control (A and D). Statistical significance was assessed by 1-way ANOVA followed by Tukey’s multiple-comparison test (A–F). The data are presented as the mean ± SD (A–F). All samples were biologically independent, and 3 independent experiments were performed. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

In conclusion, we demonstrated that alisertib treatment elevated B7-H3 expression in glioma cells by increasing EGFR activity. Therefore, we propose that the on-target effect of alisertib on the activation of EGFR may be attributed to its ability to increase AURKA expression, thereby maintaining SDCBP stability and subsequently activating the EGFR pathway.

AURKA inhibitors modulate B7-H3 and the immune microenvironment. We then assessed whether our observations had translational relevance. First, we investigated whether alisertib inhibited proliferation and increased B7-H3 expression in G261 mouse glioma cells at different concentrations and time points. The results showed that cell viability was substantially suppressed by alisertib treatment in a dose-dependent manner at 72 hours (Supplemental Figure 4A). B7-H3 expression in G261 cells was upregulated in a time- and dose-dependent manner (Supplemental Figure 4, B and C). We subsequently used a mouse syngeneic glioma model with G261-Luc cells to assess the efficacy of AURKA inhibition in vivo. Tumor-bearing mice were randomized into (a) the control and (b) alisertib groups. The tumor volume of C57BL/6 mice was detected via an in vivo imaging system (IVIS) at different time points (days 12, 19, 26, and 33 after inoculation with tumor cells). Alisertib was administered orally daily for 2 weeks. The time points of administration are shown in Figure 6A. We detected reduced tumor growth in the alisertib group compared with the control group after 2 weeks of alisertib treatment (Figure 6, B–D). Kaplan-Meier analysis of animal survival revealed that, compared with the control group, the alisertib group presented significantly longer survival times (P = 0.0487; Figure 6E).

AURKA inhibitors suppress tumor growth and upregulate B7-H3 expression. (A) Schematic diagram of the in vivo medication studies in G261-bearing mice. Alisertib or vehicle was administered daily on day 13 after tumor inoculation for 2 weeks. (B and C) Changes in the volume of orthotopic G261 tumors treated with or without alisertib at various time points were measured via the IVIS system (n = 5/group) and quantified. (D) The volume of orthotopic G261 tumors treated with or without alisertib on day 26 was measured via the IVIS system. (n = 5 for the alisertib group, n = 4 for the control group). (E) Kaplan-Meier survival plots of mice bearing orthotopic G261 tumors treated with or without alisertib (n = 5/group). (F) Schematic diagram of the in vivo medication studies in U87-MG–bearing mice. Alisertib or vehicle was administered daily on day 13 after tumor inoculation for 2 weeks. (G and H) Changes in the volume of orthotopic U87-MG tumors treated with or without alisertib at various time points were measured via the IVIS system (n = 6/group) and quantified. (I) The volume of orthotopic U87-MG tumors treated with or without alisertib on day 26 was measured via the IVIS system (n = 6 for the alisertib group, n = 5 for the control group). (J) IHC analysis of B7-H3 in orthotopic U87-MG tumors treated with or without alisertib; the quantifications are shown on the right (n = 6 for the alisertib group, n = 5 for the control group). Statistical significance was assessed by using a 2-tailed unpaired Student’s t test (D, I, and J) and the log-rank (Mantel-Cox) test (E). The data are presented as the mean ± SEM (C, D, H, I, and J). *P < 0.05, **P < 0.01. Scale bars: 50 µm.

Subsequently, we assessed the effect of AURKA inhibitors on the expression of B7-H3 in vivo through an orthotopic xenograft model employing U87-MG–Luc cells. Tumor-bearing mice were randomized into (a) the control and (b) alisertib groups. The tumor volume of NSG mice was detected via IVIS at different time points (12, 19, and 26 days after inoculation with tumor cells). Alisertib was administered orally daily to the mice in the alisertib group for 2 weeks, and the time points of administration are shown in Figure 6F (the schematics are drawn in the MedPeer platform; https://image.medpeer.cn/). We detected reduced tumor growth in the alisertib group compared with the control group after 2 weeks of alisertib treatment (Figure 6, G–I). We used IHC to analyze B7-H3 expression in control and alisertib-treated tumors. Tumors in the alisertib group presented significantly higher levels of B7-H3 expression than did those in the control group (Figure 6J).

To analyze tumor-infiltrating T lymphocytes and myeloid cells, we performed immunofluorescence and IHC staining on control and alisertib-treated tumors. Notably, the infiltration of CD3+, CD4+, and CD8+ T cells was considerably greater in alisertib-treated tumors than in control tumors (Figure 7A). In contrast, the number of Foxp3+ T cells, which are Tregs, was significantly lower in the alisertib-treated tumors than in the control tumors (Figure 7A). Furthermore, we observed increased numbers of CD68+ and iNOS+ cells, indicative of M1 (inflammatory) phenotype macrophages, after 2 weeks of alisertib treatment (Figure 7B). Conversely, the number of ARG1+ cells, which are M2 antiinflammatory macrophages, was notably lower in the alisertib-treated tumors than in the control tumors (Figure 7B). In addition, tumors in the alisertib group also had significantly more perforin+ and granzyme B+ cells than did those in the control group (Figure 7C). These findings suggest that AURKA inhibition enhances the antitumor immune response in glioma orthotopic syngeneic models.

AURKA inhibitors affect immune cell infiltration and activation in glioma. (A and B) A panel of immune markers in orthotopic G261 tumors treated with or without alisertib were detected via immunofluorescence analysis, and the quantifications are presented below (n = 3/group). The T cell markers CD3, CD4, CD8, and Foxp3 (A). The macrophage markers CD68, iNOS, and ARG1 (B). (C) Perforin and granzyme B in orthotopic G261 tumors treated with or without alisertib were detected by IHC analysis, and the quantifications are shown on the right (n = 3/group). Statistical significance was assessed by using a 2-tailed unpaired Student’s t test (A–C). The data are presented as the mean ± SD (A–C). *P < 0.05, **P < 0.01. Scale bars: 50 µm.

Combining alisertib with an anti–B7-H3 mAb reduces tumor size and increases CD8+ T cell infiltration. Since the AURKA inhibitor alisertib can upregulate B7-H3 expression and modulate the immune microenvironment in glioma cells, thereby increasing response rate and clinical efficacy of B7-H3–specific blocking mAbs, we hypothesized that combining an AURKA inhibitor with an anti–B7-H3 mAb could improve GBM patient prognosis. We created an orthotopic syngeneic model to investigate this hypothesis by injecting 2 × 105 G261-Luc cells into the mouse frontal lobe. Twelve days after inoculation, the mice were randomized into 4 groups: (a) anti-IgG, (b) anti–B7-H3 mAbs, (c) alisertib and anti-IgG, and (d) alisertib and anti–B7-H3 mAbs. For groups c and d, alisertib was administered for 2 days before being combined with an i.p. injection of either an anti–B7-H3 mAb or an isotype control antibody (Figure 8A). Compared with alisertib or anti–B7-H3 mAb monotherapy, alisertib combined with anti–B7-H3 mAbs had a synergistic inhibitory effect on tumor growth (Figure 8, B–D, and Supplemental Figure 4D). On day 33 after inoculation with tumor cells, the tumors in the alisertib and anti–B7-H3 mAbs group were significantly smaller than those in the alisertib and anti-IgG group (Figure 8D). However, this result may be biased as it was measured on smaller tumors due to some mice dying from tumor burden. Moreover, our data reveal that the combination treatment of alisertib and anti–B7-H3 mAbs significantly extended animal survival compared with the anti-IgG (P = 0.0026), alisertib and anti-IgG (P = 0.0197), and anti–B7-H3 mAbs (P = 0.0090) groups, as demonstrated by Kaplan-Meier analysis (Figure 8E). Owing to the lack of significant difference in survival between anti-IgG and anti–B7-H3 mAbs, we analyzed tumor-infiltrating immune cells among (a) anti-IgG, (c) alisertib and anti-IgG, and (d) alisertib and anti–B7-H3 mAbs. Compared with combination therapy with alisertib and anti-IgG, combination therapy with alisertib and anti–B7-H3 mAbs significantly increased the abundance of total CD3+ T cells, tumor-infiltrating CD8+ T cells, perforin+ cells and granzyme B+ cells but not total CD4+ T cells (Figure 9, A and B). Additionally, the number of Foxp3+ T cells in the combination therapy with alisertib and anti–B7-H3 mAbs group was lower than that in the combination therapy with alisertib and anti-IgG group, but there was no statistically significant difference between the 2 groups (Figure 9A). We did not find a significant difference in the number of CD68+, iNOS+, or ARG1+ cells between the combination therapy with alisertib and anti–B7-H3 mAbs and the combination therapy with alisertib and anti-IgG groups (Figure 9, C and D), indicating that anti–B7-H3 mAbs treatment did not substantially alter the level of macrophage infiltration induced by alisertib treatment.

The combination of alisertib with an anti–B7-H3 mAb reduces tumor size. (A) Schematic diagram of the in vivo combination treatment in G261-bearing mice. The mice were treated with alisertib (daily × 14 days), B7-H3 mAbs (300 μg/injection × 4) or an isotype control antibody (300 μg/injection × 4) in combination. (B and C) Changes in the volume of orthotopic G261 tumors in the anti-IgG, alisertib + anti-IgG, anti–B7-H3, and alisertib + anti–B7-H3 groups were measured via the IVIS system (n = 5/group) and quantified. (D) The volume of orthotopic G261 tumors in the alisertib + anti-IgG and alisertib + anti–B7-H3 groups on day 33 was measured via the IVIS system (n = 3 for the alisertib + anti-IgG group, n = 4 for the alisertib + anti–B7-H3 group). (E) Kaplan-Meier survival plots of mice bearing orthotopic G261 tumors in the anti-IgG, alisertib + anti-IgG, anti–B7-H3, and alisertib + anti–B7-H3 groups (n = 5/group). Statistical significance was assessed via 2-tailed unpaired Student’s t test (D) and the log-rank (Mantel-Cox) test (E). The data are presented as the mean ± SEM (C and D). *P < 0.05; **P < 0.01.

The combination of alisertib with an anti–B7-H3 mAb increases CD8+ T cell infiltration and activation. (A) A panel of immune markers in orthotopic G261 tumors from the anti-IgG, alisertib + anti-IgG, and alisertib + anti–B7-H3 groups were detected via immunofluorescence analysis, and the quantifications are shown on the below (n = 3/group). The T cell markers are CD3, CD4, CD8, and Foxp3. (B) Perforin and granzyme B in orthotopic G261 tumors from the anti-IgG, alisertib + anti-IgG, and alisertib + anti–B7-H3 groups were detected via IHC analysis, and the quantifications are shown on the below (n = 3/group). (C and D) A panel of immune markers in orthotopic G261 tumors from the anti-IgG, alisertib + anti-IgG, and alisertib + anti–B7-H3 groups were detected via immunofluorescence analysis, and the data are presented on the below (n = 3/group). The following markers of macrophages were used: CD68, iNOS, and ARG1. Statistical significance was assessed via 1-way ANOVA followed by Tukey’s multiple-comparison test (A–D). The data are presented as the mean ± SD (A–D). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Scale bars: 50 µm.