Imaging CXCL12-CXCR4 signaling in ovarian cancer therapy

Abstract

Chemokine CXCL12 and receptor CXCR4 have emerged as promising therapeutic targets for ovarian cancer, a disease that continues to have a dismal prognosis. CXCL12-CXCR4 signaling drives proliferation, survival, and invasion of ovarian cancer cells, leading to tumor growth and metastasis. Pleiotropic effects of CXCR4 in multiple key steps in ovarian cancer suggest that blocking this pathway will improve outcomes for patients with this disease. To quantify CXCL12-CXCR4 signaling in cell-based assays and living mouse models of ovarian cancer, we developed a click beetle red luciferase complementation reporter that detects activation of CXCR4 based on recruitment of the cytosolic adapter protein β-arrestin 2. Both in two-dimensional and three-dimensional cell cultures, we established that bioluminescence from this reporter measures CXCL12-dependent activation of CXCR4 and inhibition of this pathway with AMD3100, a clinically-approved small molecule that blocks CXCL12-CXCR4 binding. We used this imaging system to quantify CXCL12-CXCR4 signaling in a mouse model of metastatic ovarian cancer and showed that treatment with AMD3100 interrupted this pathway in vivo. Combination therapy with AMD3100 and cisplatin significantly decreased tumor burden in mice, although differences in overall survival were not significantly greater than treatment with either agent as monotherapy. These studies establish a molecular imaging reporter system for analyzing CXCL12-CXCR4 signaling in ovarian cancer, which can be used to investigate biology and therapeutic targeting of this pathway in cell-based assays and living mice.

Figure 1. Image reporters, transduced cells, and click beetle complementation system.

A) Panel shows lentiviral vectors and stably transduced cells used for imaging studies. B) Schematic diagram of click beetle red complementation system. N- and C-terminal fragments of click beetle red luciferase are fused to the C-termini of CXCR4 and β-arrestin 2, respectively. CXCL12 binding to CXCR4 causes recruitment of β-arrestin 2 to the activated receptor, reconstituting click beetle red luciferase to produce light.

Figure 2. CXCL12-CXCR4 signaling pathway in ovarian cancer cells.

A) HeyA8-CXCR4-CBRN/Ar-CBC and parental HeyA8 cells were treated with increasing concentrations of CXCL12 for 10 minutes. Western blot of total cell lysates shows phosphorylated and total AKT, respectively. We used GAPDH as a loading control. Graph shows relative band intensities for phosphorylated AKT in each cell line normalized to total AKT and GAPDH. B) HeyA8-CXCR4-CBRN/Ar-CBC cells were treated with 100 ng/ml CXCL12-α for 0, 5, 15, or 30 minutes in the absence or presence of 1 µM AMD3100. Total cell lysates were probed for phosphorylated and total ERK1/2, respectively. Lysates also were analyzed by Western blot for expression of β-arrestin 2-CBC and endogenous β-arrestin 1 and 2. GAPDH is shown as a loading control. C) Flow cytometry of CXCR4 expression in various HeyA8 cell lines used in this study. Dark line and dashed lines in histogram plots denote isotype control and staining with CXCR4 antibody.

Figure 3. HeyA8-CXCR4-CBRN/Ar-CBC cells report on receptor activation.

A) HeyA8-CXCR4-CBRN/Ar-CBC cells were plated as two-dimensional cultures in 96 well plates and incubated with increasing concentrations of CXCL12-α in the presence of 1 µM AMD3100 or vehicle control. Data were graphed as mean values for luminescence relative to cells not treated with CXCL12 (n = 4 per condition). Error bars denote SEM. B) HeyA8-CXCR4-CBRN/Ar-CBC cells were treated with 100 ng/ml CXCL12 for increasing periods of time in the presence of 1 µM AMD3100 or vehicle control. Data were graphed as mean values ± SEM for luminescence relative to cells not incubated with CXCL12 (n = 4 per condition). *, p<0.05; **, p<0.01.

Figure 4. CXCR4 signaling and cell viability in spheroid cultures.

A) Spheroid cultures of HeyA8-CXCR4-CBRN/Ar-CBC and HeyA8-CXCL12-GL cells were treated for 24 hours with increasing concentrations of AMD3100. Data were graphed as mean values ± SEM for luminescence for click beetle complementation relative to spheroids treated only with vehicle control (n = 10 spheroids per condition). B) HeyA8-FL-GFP cells were cultured as spheroids with HeyA8-CXCL12-GL or HeyA8-GL cells and then treated with increasing concentrations of cisplatin for 24 hours. Data were graphed as mean values ± SEM for firefly luciferase luminescence relative to spheroids not treated with cisplatin (n = 10 spheroids per condition). *, p<0.05.

Figure 5. Imaging association of CXCR4 and β-arrestin 2 in living mice.

A) Representative images of mice with intraperitoneal implants of HeyA8-CXCR4-CBRN/Ar-CBC and HeyA8-CXCL12-GL cells. Images were obtained before and following 5 days of treatment with osmotic pumps containing AMD3100 or vehicle control. Scale bar depicts ranges of photon flux values displayed on pseudocolor images. B) Fluorescence from ovarian cancer cells was measured in living mice treated with AMD3100 or vehicle control. Graph shows mean values+SEM for fluorescence relative to values measured on day 0 before beginning treatment. Arrows show the period when osmotic pumps were in place. C) Quantified photon flux data for click beetle red complementation in mice treated with AMD3100 or vehicle control, respectively. Data were normalized to tumor fluorescence for each mouse and graphed as mean values+SEM (n = 7 mice per group). *, p<0.05.

Figure 6. Combination therapy with AMD3100 and cisplatin decreases tumor burden.

A) Area under the curve analysis of imaging data for ratios of click beetle red luciferase complementation for CXCR4 and β-arrestin 2 normalized to eqFP650 fluorescence (tumor burden) in mice implanted with HeyA8-CXCR4-CBRN/Ar-CBC and HeyA8-CXCL12-GL cells. Data are shown for groups treated with vehicle control, AMD3100, cisplatin, or both AMD3100 and cisplatin for two weeks beginning one week after implanting cells. Graph shows mean values+SEM (n = 7 mice per group). *, p<0.05. B) Area under the curve analysis of fluorescence from eqFP650 produced by implanted ovarian cancer cells over the course of the experiment. Graph shows mean values+SEM. *, p<0.05, **, p<0.01. C) Kaplan-Meier curves for survival of mice treated with vehicle, AMD3100, cisplatin, or AMD3100 and cisplatin. All treatment groups differ from vehicle control (p<0.05) but not from each other.