(504g) Now You See Me, Now You Don’T—Building Better Biologics through Immune Evasion

Biotherapeutics are helping to reshape modern medicine, but these powerful drugs bear unique risks and limitations compared to chemotherapeutics. For example, protein drugs are subject to immune surveillance in the body, and ensuing anti-drug immune responses can manifest clinical complications including altered pharmacokinetics, loss of efficacy, formation of toxic immune complexes, and allergic reactions of varying severity. Thus, understanding and mitigating biotherapeutic immunogenicity is becoming an integral part of biopharma’s drug development process.

T cell epitopes are immunogenic peptides proteolytically processed from intact proteins and displayed on the surface of antigen presenting cells via class II MHC. Subsequent binding of peptide-MHC complexes by receptors on CD4⁺ helper T cells initiates a series of signaling events that culminate with B cell maturation and production of high affinity antibodies able to bind the offending, intact biotherapeutic. These detrimental anti-drug immune reactions can be short-circuited by reengineering a biotherapeutic such that its constituent peptides evade molecular recognition by class II MHC, however “deimmunizing” mutations should not undermine the stability or activity of the folded and functional drug. Deimmunization via T cell epitope deletion must therefore balance two competing objectives: reduce immunogenicity yet maintain function.

To address this molecular engineering challenge, our collaborative groups have developed deimmunization algorithms that simultaneously optimize proteins for both low immunogenic potential and high molecular fitness. These computational tools map the dual-objective protein design space, identifying the Pareto optimal frontier, or the set of variants whose predicted immunogenicities and functionalities are not simultaneously dominated by any other design. A preemptive understanding of the tradeoffs between extent of deimmunization and maintenance of function can provide valuable guidance during the biotherapeutic engineering and optimization process, and we have validated the algorithms in extensive proof-of-concept studies.

Most recently, we have employed these algorithms to design next-generation antibiotics based on lysostaphin, a peptidoglycan hydrolase with potent bactericidal activity towards drug-resistant Staphylococcus aureus, commonly called MRSA. Lysostaphin’s anti-MRSA activity is well documented, but the enzyme is highly immunogenic due to its own microbial origins. In this presentation we describe detailed characterization of our lead candidate’s immunogenicity and antibacterial activity. Our experimental systems include ex vivo human cellular immunoassays and in vivo model studies in humanized HLA-transgenic mice. The data from these studies provide a unique opportunity to connect the dots between mutagenic deletion of putative T cell epitopes, silencing of T cell activation, suppression of the anti-drug antibody response, and improving in vivo efficacy. This deimmunized antibacterial drug candidate is now undergoing IND-enabling studies with an eye towards first-in-human trials. More broadly, our protein deimmunization algorithms represent a platform technology with the potential to accelerate and de-risk biotherapeutic development.