Targeting axonal guidance dependencies in glioblastoma with ROBO1 CAR T cells.

Resistance to genotoxic therapies and tumor recurrence are hallmarks of glioblastoma (GBM), an aggressive brain tumor. In this study, we investigated functional drivers of post-treatment recurrent GBM through integrative genomic analyses, genome-wide genetic perturbation screens in patient-derived GBM models and independent lines of validation. Specific genetic dependencies were found consistent across recurrent tumor models, accompanied by increased mutational burden and differential transcript and protein expression compared to its primary GBM predecessor. Our observations suggest a multi-layered genetic response to drive tumor recurrence and implicate PTP4A2 (protein tyrosine phosphatase 4A2) as a modulator of self-renewal, proliferation and tumorigenicity in recurrent GBM. Genetic perturbation or small-molecule inhibition of PTP4A2 acts through a dephosphorylation axis with roundabout guidance receptor 1 (ROBO1) and its downstream molecular players, exploiting a functional dependency on ROBO signaling. Because a pan-PTP4A inhibitor was limited by poor penetrance across the blood-brain barrier in vivo, we engineered a second-generation chimeric antigen receptor (CAR) T cell therapy against ROBO1, a cell surface receptor enriched across recurrent GBM specimens. A single dose of ROBO1-targeted CAR T cells doubled median survival in cell-line-derived xenograft (CDX) models of recurrent GBM. Moreover, in CDX models of adult lung-to-brain metastases and pediatric relapsed medulloblastoma, ROBO1 CAR T cells eradicated tumors in 50-100% of mice. Our study identifies a promising multi-targetable PTP4A-ROBO1 signaling axis that drives tumorigenicity in recurrent GBM, with potential in other malignant brain tumors.

Intratumoral Delivery of Chimeric Antigen Receptor T Cells Targeting CD133 Effectively Treats Brain Metastases.

PURPOSE: Brain metastases (BM) are mainly treated palliatively with an expected survival of less than 12 months after diagnosis. In many solid tumors, the human neural stem cell marker glycoprotein CD133 is a marker of a tumor-initiating cell population that contributes to therapy resistance, relapse, and metastasis. EXPERIMENTAL DESIGN: Here, we use a variant of our previously described CD133 binder to generate second-generation CD133-specific chimeric antigen receptor T cells (CAR-T) to demonstrate its specificity and efficacy against multiple patient-derived BM cell lines with variable CD133 antigen expression. RESULTS: Using both lung- and colon-BM patient-derived xenograft models, we show that a CD133-targeting CAR-T cell therapy can evoke significant tumor reduction and survival advantage after a single dose, with complete remission observed in the colon-BM model. CONCLUSIONS: In summary, these data suggest that CD133 plays a critical role in fueling the growth of BM, and immunotherapeutic targeting of this cell population is a feasible strategy to control the outgrowth of BM tumors that are otherwise limited to palliative care. See related commentary by Sloan et al., p. 477.

Comprehensive prediction of secondary metabolite structure and biological activity from microbial genome sequences.

Novel antibiotics are urgently needed to address the looming global crisis of antibiotic resistance. Historically, the primary source of clinically used antibiotics has been microbial secondary metabolism. Microbial genome sequencing has revealed a plethora of uncharacterized natural antibiotics that remain to be discovered. However, the isolation of these molecules is hindered by the challenge of linking sequence information to the chemical structures of the encoded molecules. Here, we present PRISM 4, a comprehensive platform for prediction of the chemical structures of genomically encoded antibiotics, including all classes of bacterial antibiotics currently in clinical use. The accuracy of chemical structure prediction enables the development of machine-learning methods to predict the likely biological activity of encoded molecules. We apply PRISM 4 to chart secondary metabolite biosynthesis in a collection of over 10,000 bacterial genomes from both cultured isolates and metagenomic datasets, revealing thousands of encoded antibiotics. PRISM 4 is freely available as an interactive web application at http://prism.adapsyn.com .