Development of ADAM10 antibodies to target oncogenic cell surface receptor signalling

A Disintegrin and metalloproteinase 10 (ADAM10) is a transmembrane protease responsible for cleaving many cell surface proteins essential in development, including receptors of the Eph, EGFR, and Notch families. De-regulation of ADAM10 cleavage has been reported in many diseases, including cancer, where its expression is associated with defective receptor signalling. Unlike matrix metalloproteases, substrate specificity of ADAM10 lies in its non-catalytic domains: A negatively charged substrate recognition pocket in the Cysteine-rich domain is known to confer ADAM10 cleavage specificity, for example in EphA/ephrinA signalling. Therefore, we hypothesize that antibodies against this region of ADAM10 would prevent protease activity. Anti-ADAM10 monoclonal antibodies were successfully generated and characterised using newly optimized assays. One anti-ADAM10 monoclonal antibody, 8C7, showed conformation-dependent specificity and reduced binding to ADAM10 substrate recognition domain mutants, suggesting that its specificity towards this region. Interestingly, 8C7 inhibited ephrin cleavage and ephrin-induced Eph receptor internalisation, phosphorylation and Eph-mediated cell segregation, confirming its ability to counteract ADAM10 mediated signalling. Thus, we attempted to define the mechanism of action of 8C7 and its role in tumour suppression. We resolved the crystal structure of 8C7 in complex with ADAM10 and discovered that it binds adjacent to the substrate recognition pocket, including a cysteine residue that forms a disulphide bond with the conserved CxxC motif of ADAM10. The same motif regulates structural conformation of the related ADAM17, by disulphide switching catalysed by protein disulphide isomerases (PDIs). Accordingly, mutation of this motif in ADAM10 reduced 8C7 binding, as did modulation of redox conditions, previously shown to modulate ADAM17 conformation and activity. Importantly, 8C7 selectively binds an active form of ADAM10, which we find to be prevalent in mouse tumour models and in human solid tumours compared to normal tissues. Additionally, we demonstrate that 8C7 treatment inhibits growth of human colon tumour xenografts coincident with inhibition of Notch signalling and down-regulation of Notch, EGFR and Eph receptor levels. In line with the critical role of Notch in maintenance of an undifferentiated cell phenotype, including that of tumour initiating colon cancer cells, we find 8C7 specifically targets tumour cells expressing the putative tumour stem cell marker CD133, and displaying active Notch-signalling within tumours. CD133+ cell population, also known to mediate chemoresistance, is effectively targeted by 8C7 in combination with chemotherapy. In summary, we have successfully developed anti-ADAM10 mAbs that inhibit tumour growth and blocks Notch receptor signalling. We have also identified a highly active novel conformation of ADAM10, predominantly expressed in cancers and selectively recognized by the 8C7 monoclonal antibody. Together, we propose that 8C7 has promising potential as a novel therapeutic and diagnostic tool.