RNA sequencing data from retinoblastoma tumors and normal retinal tissues was accessed via the GEO database accession numbers “type”:”entrez-geo”,”attrs”:”text”:”GSE125903″,”term_id”:”125903″GSE125903, “type”:”entrez-geo”,”attrs”:”text”:”GSE111168″,”term_id”:”111168″GSE111168, and “type”:”entrez-geo”,”attrs”:”text”:”GSE87042″,”term_id”:”87042″GSE87042?21, 22, 23 and reprocessed at CHOP as described below. significantly enriched expression in both the SCLC and neuroblastoma stem cell compartment.By solving the crystal structure of the D3-GPC2-Fab/GPC2 complex at 3.3 ? resolution, we further illustrate that this GPC2-directed antibody-drug conjugate (ADC; D3-GPC2-PBD), that links a human GPC2 antibody (D3) to DNA-damaging pyrrolobenzodiazepine (PBD) dimers, binds a tumor-specific, conformation-dependent epitope of the core GPC2 extracellular domain name. We then show that this ADC Pivmecillinam hydrochloride induces durable neuroblastoma and SCLC tumor regression via induction of DNA damage, apoptosis, and bystander cell killing, notably with no signs of ADC-induced toxicity. These studies provide preclinical data to support the clinical translation of ADCs targeting GPC2. and robust efficacy in a proof-of-concept neuroblastoma study.12 However, it remains unclear how broadly acting GPC2-targeting ADCs would be across different cancer histotypes and furthermore, what is the molecular basis of the D3-GPC2-PBD ADCs specificity, potent efficacy, and favorable tolerability profile, all critical elements to the clinical translation of GPC2 ADCs. We now report high levels of GPC2 across several other cancers, most notably small-cell lung cancers (SCLCs), and fully validate GPC2 as a robust immunotherapeutic target by showing enrichment of GPC2 expression in the cancer stem cell compartment. In turn, we provide a detailed molecular characterization of the D3-GPC2 antibody that explains its high specific affinity for tumor-associated GPC2 and its favorable safety profile. Finally, we demonstrate potent and sustained anti-tumor efficacy for this GPC2 ADC across a panel of genomically and clinically diverse neuroblastoma and SCLC preclinical models via induction of DNA damage, apoptosis, and robust bystander cytotoxicity. Results GPC2 is usually a MYCN transcriptionally activated cell-surface oncoprotein in SCLCs We have previously validated high levels of GPC2 around the neural-derived pediatric cancers neuroblastoma, medulloblastoma, and retinoblastoma.12 Now to more comprehensively define the full spectrum of GPC2-expressing pediatric and adult malignancies, we queried expression across a large number of cancers using multiple RNA-sequencing datasets, including data from the St. Jude PeCAN portal (https://pecan.stjude.cloud/home), the PedcBioPortal for Integrated Childhood Cancer Genomics (https://pedcbioportal.kidsfirstdrc.org), the R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl), the Cancer?Genome Atlas?(TCGA; https://www.cancer.gov/about-nci/organization/ccg/research/structural-genomics/tcga), and the Broad Institute Cancer Cell Line Encyclopedia (CCLE; https://depmap.org/portal/; Physique?S1A).20, 21, 22, 23, 24, 25, 26, 27, 28, Pivmecillinam hydrochloride 29, 30, 31 These data validated our prior findings?showing that a majority of neuroblastomas, medulloblastomas, and retinoblastomas express high levels of expression in a number of?non-medulloblastoma pediatric brain tumors, as well as subsets?of pediatric acute lymphocytic leukemias (ALLs; Physique?S1B) and other pediatric solid tumors. Importantly, these data also revealed that several adult cancers have overexpression of gene family member, and continued poor outcomes despite multimodal toxic therapies.32 First, comparing expression in SCLCs with normal lung tissues revealed that this tumor-normal differential expression pattern was the most significant among the glypican family and also that high levels of were restricted to Pivmecillinam hydrochloride malignant lung cells (Determine?1A), similar to the expression pattern in embryonal cancers such as retinoblastoma (Physique?S1C). Next, we subjected five genomically diverse SCLC cell lines with varied expression, genomic drivers, and MYC family expression profiles (Table S1) to GPC2-directed flow cytometry and western blot, confirming neuroblastoma-comparable GPC2 and MYC family expression (Figures 1BC1D; Figures S1D and S1E). To confirm high levels of GPC2 expression on SCLCs in SCLCs, we next quantified MYCN binding to the canonical E-Box motif present upstream of the transcription start site with chromatin immunoprecipitation (ChIP) PCR in the promoter (Physique?1H). Furthermore, MYCN depletion in H526 cells resulted in concurrent GPC2 downregulation (Figures 1I and 1J). Finally, we also investigated the phenotypic consequences of depleting GPC2 in this SCLC cell line panel, which revealed robust apoptosis induction and significant growth inhibition in GPC2-depleted cells comparable to what we have observed in neuroblastoma (Figures 1K and Ntf5 1L).12 These data, taken together with our previous findings.
RNA sequencing data from retinoblastoma tumors and normal retinal tissues was accessed via the GEO database accession numbers “type”:”entrez-geo”,”attrs”:”text”:”GSE125903″,”term_id”:”125903″GSE125903, “type”:”entrez-geo”,”attrs”:”text”:”GSE111168″,”term_id”:”111168″GSE111168, and “type”:”entrez-geo”,”attrs”:”text”:”GSE87042″,”term_id”:”87042″GSE87042?21, 22, 23 and reprocessed at CHOP as described below