A multi-omic surfaceome study identifies DLK1 as an epigenetically regulated protein and immunotherapeutic target in neuroblastoma

Abstract

Meeting Abstract LB-B04
Background: Neuroblastoma (NB) is an embryonal tumor of the sympathetic nervous system that accounts for 12% of childhood cancer deaths. While the introduction of GD2 immunotherapy provides an improvement in time to progression, the therapy is toxic and impact on overall survival is minimal, supporting an urgent need for novel immunotherapies. To date, the cell surface landscape (surfaceome) of NB remains undefined, hindering the identification of immunotherapeutic targets.
Methods: To identify NB surfaceome proteins, we performed plasma membrane protein extraction using sucrose gradient ultracentrifugation coupled to mass spectrometry (nLC-MS/MS) in NB cell lines (n=12) and patient derived xenografts (PDX; n=10). These data were integrated with existing RNA-sequencing (NB=153; Normal=7859) and H3K27ac chromatin immunoprecipitation (ChIP)-sequencing data (from overlapping NB cell lines) to evaluate extracellular proteins differentially expressed in NB compared to normal tissues. Candidate targets were validated by immunohistochemistry on NB tumor and normal tissue microarrays (TMAs), flow cytometry and immunofluorescence. In-vitro functional studies were performed following genetic manipulation of candidate targets to assess cell proliferation, differentiation and viability. Finally, we tested ADCT-701 (a DLK1-directed antibody drug conjugate [ADC] with a pyrrolobenzodiazepine [PBD] warhead) in eight PDX models (study ongoing, total of 12 models initiated) with varying levels of DLK1 expression. At enrollment, two mice were each treated with a single dose of saline or 1mg/kg of B12-PL1601 (non-targeting PBD-conjugated ADC) or 1mg/kg ADCT-701 and mice were evaluated for 100 days or until tumor reached 2.0cm3. Results: We yielded on average 66% (range:60-68%) membrane protein enrichment with high reproducibility between biological replicates (80%; range:78-84%) and identified 4826 unique membrane proteins. Our approach confirmed known cell surface proteins in development as immunotherapeutic targets in NB (ALK, GPC2, NCAM1, DLL3 and CD276). Here, we prioritized DLK1 for further evaluation due to it being the only candidate with expression directly associated with a super enhancer element (P=6.09X10-5). RNA-sequencing and tissue microarray analysis of NB and normal tissues showed DLK1 to be overexpressed in a large subset of high-risk NB with minimal expression in normal tissues, excepting adrenal medulla and pituitary. Flow cytometry and immunofluorescence confirmed cell surface expression of DLK1 in a panel of NB cell lines. Genetic depletion of DLK1 using shRNA resulted in neurite outgrowth (P=7.26X10-5) and terminal differentiation. Full proteome analysis of DLK1 knockdown and control cell lines using MS showed regulation of proteins that control outgrowth of neurites (P=3.37X10-3) and development of neurons (P=3.76X10-3). To date, ADCT-701 treatment resulted in maintained complete response (N=2), complete response (N=3) and stable disease (N=1) in models with high DLK1 expression, while those with low/no expression showed disease progression (N=2).
Conclusion: DLK1 is an epigenetically regulated immunotherapeutic target in neuroblastoma. ADCT-701 shows potent activity in preclinical models of NB and should be prioritized for clinical development.