Cell surface antigen discovery is of great interest for biomedical research both for isolation of rare cell populations and therapeutic targeting. We developed a rapid, cost-effective, fully in vitro technology which facilities the simultaneous target discovery and human antibody generation on the surface of virtually any cell population of interest. We apply our technique to human colorectal cancer-initiating cells (CICs) and identify hundreds of unique human antibodies. We characterized the top three antibody candidates targeting these CICs and identify their protein targets as integrin α7 (ITGA7), HLA-A1 and integrin β6 (ITGB6). We demonstrate that these antibodies can be used to isolate self-renewing colorectal CICs, and that the integrin α7 antibody can prospectively identify glioblastoma brain tumor initiating cells as well as human muscle stem cells. We also demonstrate that genetic ablation of integrin β6 impedes colorectal CIC function. The methodology can be readily applied to other cell populations including stem cells, cancer, or immune cells to facilitate the rapid identification of novel targets and simultaneous generation of potent and specific antibodies with therapeutic potential.

No abstract

Background: Glioblastoma (GBM) is the most common malignant primary adult brain tumor, characterized by extensive cellular and genetic heterogeneity. Even with surgery, temozolomide chemotherapy and radiation, tumor re-growth and patient relapse are inevitable, with a median survivorship of just 15 months. Genomic profiling studies have shown that clonal evolution within GBM may be driven by cancer treatment, such that the recurrence may no longer resemble the genetic landscape of the original primary tumor. Furthermore, intratumoral heterogeneity associated with clonal evolution complicates biomarker discovery and treatment personalization and underlies treatment failure. Thus, modeling clonal heterogeneity and evolution to understand cancer progression is critical for the development of effective therapeutic approaches. We aim to identify new therapeutic targets that drive clonal evolution in treatment-refractory GBM and develop novel and empirical therapeutic paradigms targeting recurrent GBM. Experimental Procedure: We employed a transcriptomic, proteomic and functional genomics approach to discover and validate genes that drive GBM recurrence. Using a therapy-adapted patient-derived xenograft (PDX) model of treatment-refractory GBM, we profiled the transcriptomic and proteomic landscape of treatment-naïve primary GBM through conventional chemotherapy and radiation therapy, and into recurrence. To complement the transcriptomic data, we used an unbiased genome-wide CRISPR-Cas9 screening platform to identify genes essential for self-renewal in recurrent GBM, as well as to identify novel sensitizers and suppressors of conventional therapy. Furthermore, we coupled cellular DNA barcoding technology with our PDX model to profile the clonal evolution of tumor cells through therapy. Results: Integrative analysis of deep sequencing and surface proteomics of tumor cells harvested at tumor formation, minimal residual disease after chemoradiotherapy, and tumor recurrence from the PDX model resulted in the identification of novel therapeutic targets in treatment-refractory GBM. Using CRISPR, potential targets were knocked out in patient-derived GBMs in order to characterize the effect on self-renewal and tumor formation. We report the successful barcoding of patient-derived primary, treatment-naïve GSCs at a single cell resolution that were expanded into clonal populations, intracranially engrafted in immunodeficient mice and treated with SoC therapy. Conclusion: We have generated a translational pipeline from initial target discovery, through target validation, to building new biotherapeutics against novel targets, and preclinical testing in our PDX model of treatment-resistant GBM. A promising lead panel of biotherapeutic modalities is being translated into early clinical development, generating targeted therapies and hope for future GBM patients. Note: This abstract was not presented at the meeting. Citation Format: Parvez Vora, Chitra Venugopal, Chirayu Chokshi, Maleeha Qazi, Nazanin Tatari, Kevin Brown, Nicholas Yelle, Jarrett Adams, David Tieu, Mathieu Seyfrid, Mohini Singh, Neil Savage, Minomi Subapanditha, David Bakhshinyan, Laura Kuhlmann, Thomas Kislinger, Sachdev Sidhu, Jason Moffat, Sheila Kumari Singh. A glioblastoma translational pipeline: discovery of novel tumor antigens that drive GBM recurrence [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 570.

Human glioblastoma (hGBM) carries a dismal prognosis and inevitably relapses despite aggressive therapy. Many of the 14 members of the Eph receptor tyrosine kinase family are expressed in hGBM initiating cells (GICs) and constitute potential molecular targets. We hypothesize that multiple members of the EphR family play a critical role in hGBM recurrence. Using a highly specific human EphR antibody panel, we identified differential expression of EphRs in recurrent hGBM (rGBM). We further characterized EphR co-expression along with multiple GIC markers using mass cytometry (CyTOF). Here we show that EphA2 and EphA3 co-expression marks a highly tumorigenic cell population in rGBM that is enriched in GIC marker expression, and exhibits higher in vitro and in vivo self-renewal and proliferation capacity as compared to EphA2+/EphA3-, EphA2-/EphA3+ or EphA2-/EphA3-cells. Knockdown of EphA2 and EphA3 blocks this self-renewal and proliferation capacity, and is marked by increase in the expression of differentiation marker GFAP. Next, we generated and tested a bispecific antibody (BsAb) that co-targets EphA2 and EphA3. In vitro treatment of rGBM with BsAb led to phosphorylation of EphA2 and EphA3, eventually leading to receptor internalization and degradation. The cellular effect of EphA2/A3 blockade was mediated through the down regulation of Akt and MAPK. Intracranial treatment of immune-deficient mice harboring hGBM with BsAb resulted in non-invasive and significantly smaller tumors. Hence, EphA2 and EphA3 co-expression marks an even more potent GIC population in rGBM, and targeting either single EphA2+ or EphA3+ populations alone will allow the remaining GICs to drive tumor recurrence. For the first time, we show that strategic co-targeting of both EphA2 and EphA3 with a BsAb presents a novel and rational therapeutic approach to recurrent GBM, where multiple GIC populations may be driving the intra-tumoral heterogeneity underlying disease progression. IMMU-04. IMMUNOPHENOTYPING OF PEDIATRIC BRAIN

No abstract