CD24 offers a therapeutic target for control of bladder cancer metastasis based on a requirement for lung colonization

Abstract

Metastasis is lethal in most bladder cancer patients. Expression of CD24, a glycosyl phosphatidylinositol (GPI)-linked sialoglycoprotein and cancer stem cell marker, is associated with metastatic progression in multiple cancer types, yet the role of CD24 in this process remains unclear. While developing a murine model of human metastatic bladder cancer, we observed that tumor cell CD24 expression correlated with a propensity to metastasize to the lung. Our immunohistochemical evaluation of 60 paired primary and metastatic human bladder cancer samples revealed increased intensity (P < 0.001) and frequency (P < 0.001) of CD24 expression in metastases. To directly evaluate the role of CD24 in metastatic colonization, we manipulated CD24 expression in human bladder cancer cell lines using short hairpin RNA depletion, cDNA overexpression, and fluorescence-activated cell sorting selection. Although suppression of CD24 reduced acute tumor cell retention in the lungs of mice inoculated intravenously with cancer cells, this differential retention was no longer apparent after 24 hours, prompting us to evaluate the role of CD24 in lung colonization. Here, CD24 was found necessary for subsequent development of lung metastases. We next treated clinically detectable lung metastases in mice with anti-CD24 antibody and observed reduced tumor growth and prolonged survival. These findings suggest that CD24 is a lynchpin of metastatic progression and a promising therapeutic target for antimetastatic therapy.

Figure 1

Isolation and characterization of metastatic variants and their relevance to human cancer. A, monolayer growth of Luc and LuL-2 cells for more than 96 hours. *, P < 0.05. B, anchorage-independent growth of LuL-1 and LuL-2 at 21 days. **, P < 0.01. C, s.c. injection of 5 × 10⁵ cells into flanks of nude mice (n = 4; 2 sites per mouse per line). *, P < 0.05. D, assessment of lung metastatic capacity of 2.5 × 10⁶ Luc (n = 8) or LuL-2 (n = 8) tumor cells by inoculating mice via tail vein injection. Representative bioluminescent images show lung metastasis development at 4 weeks with 2-minute exposure and total thoracic photon flux for each mouse for more than 13 weeks. **, P < 0.01. E, comparison of overall survival curves. Data are mean ±SEM of at least 2 independent experiments. ***, P < 0.001. F, model of 59 differentially expressed genes used to stratify outcomes in 91 bladder cancer patients postcystectomy based on similarity to Luc or LuL-2 gene expression levels. **, P < 0.01.

Figure 2

CD24 protein expression in primary and metastatic human bladder cancer. IHC staining of 60 patient-matched samples of primary and metastatic urothelial tumors. A, representative samples. Original magnification is ×100; scale bar, 200 μm. B, percentage of primary and metastatic tumor specimens that express CD24. **, P = 0.006. C, percentage of evaluable samples with specific changes in IHC intensity and proportion staining scores when comparing matched primary and metastatic tumors. ***, P < 0.001.

Figure 3

CD24 cellular localization and effect on tumor cell retention and colonization kinetics in murine lungs. A, cohorts of mice (n = 12) were inoculated via tail vein with 1 × 10⁶ tumor cells and monitored with BLI and human-specific qPCR for presence in the lungs. Quantification of human DNA in the lungs of mice at indicated times after injection of tumor cells. **, P < 0.01; ***, P < 0.001. B, comparison of CD24-dependent metastatic burden, as detectable by BLI. Representative images were obtained at 8 weeks for mice injected with Luc-based cells and at 5 weeks for mice injected with Lul-2–based cells. C, comparison of lung metastatic burden for all cell lines, as measured by human-specific qPCR for total human DNA at 8 weeks postinoculation for Luc-based cell lines and 5 weeks for LuL-2–based cell lines. Data plotted on log scale are DNA quantity means for each animal.

Figure 4

Treatment of s.c. or lung metastatic bladder cancer with CD24 antibody. A, i.v. treatment of bolus saline or 5 mg/kg/d (0.1 mg in 100 μL) ALB9 mAb every 4 days for 3 treatments in mice with established and palpable s.c. tumors induced by LuL-2 cells. Triangles indicate treatment delivery time points. Subcutaneous tumors were measured biweekly and volumes were calculated as defined for a spheroid (4/3πL²W). B, ALB9 treatment of BLI detectable metastases; representative images show metastatic burden as detectable by BLI in ALB9 and isotype control (IgG1) mAb at 4 weeks. Average total photon flux is for a representative group of both ALB9 treatment (n = 8) and isotype control (n = 8) surveyed for more than 5 weeks by BLI. C, each treatment cohort included 22 mice inoculated with 2.5 × 10⁶ LuL-2 tumor cells via tail vein injections. Mice were treated i.v. with either isotype IgG1 mAb as a control or 5 mg/kg/d (0.1 mg in 100 μL) ALB9 mAb every 4 days for 4 treatments after detection of metastasis by BLI. A subset of these mice (n = 8) was followed weekly with BLI. **, P < 0.01; ***, P < 0.001. D, comparison of overall survival of mice with metastatic bladder cancer treated with ALB9 or IgG1 mAb. *, P = 0.032, log-rank, 1-tailed.