Multifaceted C-X-C Chemokine Receptor 4 (CXCR4) Inhibition Interferes with Anti-Vascular Endothelial Growth Factor Therapy-Induced Glioma Dissemination

Abstract

Resistance to antiangiogenic therapy in glioblastoma (GBM) patients may involve hypoxia-induced expression of C-X-C motif chemokine receptor 4 (CXCR4) on invading tumor cells, macrophage/microglial cells (MGCs), and glioma stem cells (GSCs). We determined whether antagonizing CXCR4 with POL5551 disrupts anti-vascular endothelial growth factor (VEGF) therapy-induced glioma growth and dissemination. Mice bearing orthotopic CT-2A or GL261 gliomas received POL5551 and/or anti-VEGF antibody B20-4.1.1. Brain tissue was analyzed for tumor volume, invasiveness, hypoxia, vascular density, proliferation, apoptosis, GSCs, and MGCs. Glioma cells were evaluated for CXCR4 expression and polymorphism and POL5551's effects on CXCR4 ligand binding, cell viability, and migration. No CXCR4 mutations were identified. POL5551 inhibited CXCR4 binding to its ligand, stromal cell-derived factor-1α, and reduced hypoxia- and stromal cell-derived factor-1α-mediated migration dose-dependently but minimally affected cell viability. In vivo, B20-4.1.1 increased hypoxic foci and invasiveness, as seen in GBM patients receiving anti-VEGF therapy. Combination of POL5551 and B20-4.1.1 reduced both glioma invasiveness by 16% to 39% and vascular density compared to B20-4.1.1 alone in both glioma models. Reduced populations of GSCs and MGCs were also seen in CT-2A tumors. POL5551 concentrations, evaluated by mass spectrometry, were higher in tumors than in neighboring brain tissues, likely accounting for the results. Inhibition of CXCR4-regulated tumoral, stem cell, and immune mechanisms by adjunctive CXCR4 antagonists may help overcome antiangiogenic therapy resistance, benefiting GBM patients.

Figure 1

Tumor volume, invasiveness, and hypoxic foci in glioma-bearing mice treated with CXCR4 antagonist POL5551 (POL) and/or anti-murine VEGF antibody B20-4.1.1 (B20). Beginning on day 14 after tumor implantation, mice bearing CT-2A or GL261 glioma were administered POL5551, B20-4.1.1, and/or vehicles. On day 26, the brains were harvested and tissue sections were stained with hematoxylin and eosin to measure the tumor volume (A and B) and invasiveness (C and D) or immunostained to count the number of hypoxic foci (E and F). Tumor invasiveness (INV) was calculated using the equation, $\text{INV}\mathrm{=}\text{P}/\sqrt{4\text{πS}}$, where P is the perimeter and S is the surface area of the tumor; a noninvasive spherical tumor would have INV = 1.0, whereas INV is >1.0 for a tumor with increasingly irregular/invasive borders. The results were analyzed by analysis of variance. Data are expressed as means ± SEM. n = 9 to 10 per group (A–F). ^∗P < 0.05, ^∗∗P < 0.01, and ^∗∗∗P < 0.001.

Figure 2

Representative immunohistopathologic images of tumor invasion, hypoxic foci, vascular density, proliferation, and apoptosis in CT-2A glioma tumors. Tissue sections from CT-2A glioma-bearing mice treated with vehicles (first column), POL5551 alone (second column), POL5551 and B20-4.1.1 (third column), or B20-4.1.1 alone (fourth column) were stained with hematoxylin and eosin to show tumor invasive borders (A–D) or immunostained to demonstrate the presence of hypoxic foci (E–H), vascular density (I–L), cellular proliferation (M–P), and apoptosis (Q–T). Original magnification: ×10 (A–D and I–L); ×2 (E–H); ×20 (M–T).

Figure 3

Tumor vascular density, cell proliferation, and apoptosis in glioma-bearing mice treated with POL5551 (POL), B20-4.1.1 (B20), and/or vehicles. Brain tissues from the same glioma-bearing mice were immunostained to measure vascular density (CD105; A and B), tumor cell proliferation (pHH3; C and D), and apoptosis (caspase 3; E and F), and positively stained cells were enumerated per section. The results were analyzed by analysis of variance. Data are expressed as means ± SEM. ^∗P < 0.05.

Figure 4

Tumor macrophage/microglial cells (MGCs) and glioma stem cells (GSCs) in glioma-bearing mice treated with POL5551 (POL), B20-4.1.1 (B20), and/or vehicles. Brain tissues from the same glioma-bearing animals were immunostained to measure tumor MGCs (Iba1; A and B) and GSCs (nestin; C and D). Positively stained cells per section were scored on a 1 to 4 scale (1 if <1%, 2 if 1% to 10%, 3 if 11% to 50%, and 4 if 51% to 100%). The results were analyzed by Kruskal-Wallis test. Data are expressed as mean score ranks ± 95% confidence intervals. ^∗P < 0.05, ^∗∗∗P < 0.001.

Figure 5

Representative immunohistopathologic images of macrophage/microglial cells (MGCs) and glioma stem cells (GSCs) in CT-2A glioma tumors. Tissue sections from CT-2A glioma-bearing mice treated with vehicles (first column), POL5551 alone (second column), POL5551 and B20-4.1.1 (third column), or B20-4.1.1 alone (fourth column) were immunostained to demonstrate the presence of Iba1-positive MGCs [shown at two magnifications, low (A–D) and high (E–H), the latter with two subpanels separated by a black line showing the tumor center (left sides) and tumor periphery (right sides)] and nestin-positive GSCs (I–L). Original magnification: ×3.5 (A–D); ×20 (E–L).

Figure 6

Schematic diagram of the mouse CXCR4 gene locus on chromosome 1 showing the location of the two exons, coding regions, 3′- and 5′-unstranslated regions (3′UTR and 5′UTR, respectively), and 5′ regulatory region. Also indicated is the alignment of the Sanger sequencing reads for CT-2A and GL261 genomic DNA with mouse genome GRCm38/mm10. Each pair of primers for exon 2 overlapped by 100 bp to enhance the alignment of DNA sequences.

Figure 7

Plasma and brain tissue concentrations of POL5551 (POL) after a single dose of POL5551 or in combination with repeated B20-4.1.1 (B20) administration in CT-2A glioma-bearing mice. Beginning on day 14 after implantation, mice were administered B20-4.1.1 or vehicles for 12 days. On day 26 after a single dose of POL5551 (5 mg/kg s.c.), 0.5, 2, 6, or 24 hours before sacrifice, plasma was collected and the brains were dissected into tumor tissue (A), brain adjacent to tumor (B), and contralateral brain (C) to determine the concentration of POL5551 in plasma (ng/mL) and brain tissue (ng/g). The results were analyzed by analysis of variance. Data are expressed as means ± SEM. n = 4 per group (A–C). ^∗P < 0.05, ^∗∗P < 0.01, and ^∗∗∗P < 0.001.

Figure 8

Expression of CXCR4, and effect of POL5551 on CXCR4 ligand binding, cell viability, and migration in glioma cells in vitro. A: Using the Tag-lite CXCR4 ligand binding assay, CXCR4-expressing HEK293 cells were incubated with increasing concentrations of POL5551 or AMD3100 in the presence of a fixed concentration of fluorescent SDF-1α; the normalized responses were plotted, and the half-maximal inhibitory concentration was calculated. B: CT-2A and GL261 cells were both incubated under normoxic (N) and hypoxic (H; 1% O₂) conditions for up to 24 hours and for 16 hours, respectively, and the protein expression of CXCR4, HIF-1α, and β-actin was determined by Western blot analysis. C: CT-2A, GL261, and U87 glioma cells were incubated in quadruplicate with POL5551 (1 nmol/L to 1 μmol/L) for 16 hours, and cell viability was assessed by MTT assay. D and E: In a Boyden chamber assay, CT-2A and GL261 cells were incubated in triplicate with POL5551 (0 to 500 nmol/L) under normoxic and hypoxic conditions and allowed to migrate in the absence or presence of an SDF-1α gradient (50 nmol/L) for 16 hours, after which the number of migrated cells was counted. F: To assess whether cellular proliferation contributed to the observed changes, cells were allowed to migrate in the presence of 10 mmol/L hydroxyurea for 16 hours. The results were analyzed by analysis of variance. Data are expressed as means ± SEM (A and C–F). ^∗∗∗P < 0.001, ^∗∗∗∗P < 0.0001.