Abstract

Abstract Chimeric Antigen Receptor-redirected T cells (CAR-Ts) remain challenging for the treatment of glioblastoma (GBM) due to the heterogeneous expression of targetable antigens, which leads to antigen-loss variants. In addition, the emerging role of the GBM-derived neurospheres (GBM-NS) as a critical cell subset in causing GBM recurrence highlights the needs for targeting these cells to achieve sustained responses. By exploiting a well-established culture system, we generated and expanded GBM-NS from 23 surgical samples, and tested using flow cytometry the expression of CSPG4, an antigen found to be overexpressed in GBM by mRNA profiling. We observed that 70% of GBM-NS displayed high expression of CSPG4 (71 to 99%), 17% moderate-high expression (51-70%), and 13% moderate-low expression (less than 50%). Based on these results, we hypothesized that CAR-Ts specific for the CSPG4 antigen would represent a broadly applicable strategy for the treatment of GBM. We generated CSPG4.CAR-Ts, encoding the 4-1BB endodomain, from six healthy donors. CSPG4.CAR-Ts efficiently eliminated 19 different GBM-NS, showing high to moderate-low expression of CSPG4, in co-culture experiments at the E:T ratios ranging from 2:5 to 1:5 (0.2±0.5% and 0.6±0.9% residual GBM-NS, respectively). By contrast, GBM-NS continued to grow in the presence of control T cells (60.7±17.6% residual GBM-NS). CSPG4.CAR-Ts, but not control T cells, also rapidly proliferated in response to GBM-NS as evaluated by the CFSE assay. CSPG4.CAR-Ts showed a Th1 cytokine profile in response to GBM-NS, releasing significantly more IFN-γ (3593.8±1718.1 pg/ml/2×10⁁5 cells) and IL-2 (258.8±153.3 pg/ml/2×10⁁5 cells) than control T cells (1.8±2.5 and 0.9±1.2 pg/ml/2×10⁁5 cells, respectively). For the in vivo experiments we compared CSPG4.CAR-Ts encoding three different co-stimulatory domains, specifically CD28, 4-1BB, and CD28-4-1BB and used CAR.CD19-Ts as negative control. Two different GBM-NS with moderate-low and high expression of CSPG4 were selected and transduced to express the FFluciferase gene to monitor the tumor growth by in vivo bioluminescence imaging. Both GBM-NS and CAR-Ts were intracranially injected in 5 wks old female nude mice. CSPG4.CAR-Ts were efficient in controlling tumor growth of both moderate and high CSPG4-expressing GBM-NS. In all CAR treated mice, we observed an early eradication of the tumor mass from high-CSPG4 expressing GBM-NS, and a significant improved survival in mice bearing moderate CSPG4-expressing GBM-NS. CAR-Ts encoding the 4-1BB were significantly more efficient than those encoding CD28 or CD28-4-1BB in prolonging survival (p = 0.04). Our data suggest that CSPG4 is a promising target for CAR-Ts in GBM and opens the path for pursuing this approach in the clinical setting. Citation Format: Serena Pellegatta, Barbara Savoldo, Chuang Su, Ferrone Soldano, Gaetano Finocchiaro, Dotti Gianpietro. Condroitin sulfate proteoglycan 4 (CSPG4)- redirected T cells eliminate glioblastoma-derived neurospheres. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr LB-249.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call