Novel anti-CD3 heavy chain-only antibodies for use in T-cell engaging therapeutics

Background: Multispecific T cell-engager (TCEs) antibodies have shown great promise as immunotherapeutic molecules. These therapies redirect T cells to recognize and kill cancer cells through co-engagement of tumor associated antigens and a T cell antigen, often CD3. Despite their clinical efficacy, the molecular complexity of current TCEs poses challenges for drug manufacturability, developability, and achieving desirable pharmacokinetic/pharmacodynamics. Currently, all approved TCE multispecifics rely on alphaCD3 binders derived from conventional IgG antibodies.Heavy chain-only antibodies (HCAbs) featuring a single-domain (VHH) binding moiety are attractive alternatives to conventional IgGs in immunotherapies. The absence of a light chain in HCAbs notably simplifies multispecific development. To-date, no alphaCD3 HCAbs with properties comparable to best-in-class clinically approved IgGs have been reported. Methods: Functional alphaCD3 HCAbs were discovered by combining llama immunization and a novel yeast-based HCAb discovery technology. Following rounds of yeast-based HCAb selection, T cell stimulating alphaCD3 HCAbs were identified, humanized, and optimized to exhibit desirable drug profiles. We demonstrated the potency of these lead molecules in two bispecific formats: 1+1 alphaCD3 X alphaCD20 and as a VHH-TCR fusion using a TCR specific for the gp100 peptide HLA complex. CD3ε binding was confirmed by x-ray crystallography using one of the lead HCAbs. Results: This study resulted in a panel of six functional alphaCD3 TCEs. These molecules predominantly competed with a known CD3 epitope, recognizing the CD3ε subunit of the human T cell receptor complex with weak to strong affinities and robust drug developability profiles. Crystallographic data and subsequent structural modelling show a unique angle of approach for our lead stimulating HCAb suggesting the interaction with the TCR complex is likely limited to only the CD3deltaε heterodimer, as CD3gammaε docking is modeled to produce significant steric clashes. The panel displayed a promising functional potency range as multispecifics; with two molecules displaying comparable in vitro tumor cell killing efficacy to clinically validated molecules such as the first-in-class TCR-alphaCD3 fusion TCE, tebentafusp. Conclusions: We showcase the discovery and engineering of novel alphaCD3 HCAbs with functional potency comparable to other clinically validated molecules. This panel of highly-developable HCAbs provide new building blocks to expand the design space for TCE multispecifics and enable the next generation of multispecific immunotherapeutics.