Turning soluble TCRs into viable therapeutics means engineering heterodimers/multimers that express well and remain stable. This talk covers strategies to achieve this across TCR bispecific formats, tracing how these approaches have evolved. It highlights the interplay between affinity enhancement, format choices, expressibility, product quality, and yield, drawing on real lessons learned while taking complex TCR-based molecules from discovery through to developability.

Orphan GPCRs present unique challenges for protein expression and functional characterization due to limited tools and poorly understood biology. Using GPR75 as a model, we applied G protein-based approaches to enable functional characterization of this challenging receptor. This presentation will highlight strategies for expressing and evaluating functional GPCR systems and demonstrate how protein expression and G protein coupling can support characterization of therapeutically relevant orphan receptors.

Virus-like particles (VLPs) provide a lipid membrane environment for challenging targets, including GPCRs and tetraspanins. We present strategies to increase expression of these multipass membrane protein targets and enhance their incorporation into VLPs for antigen display, along with methods for subsequent particle purification. We also discuss adaptations of VLP-based antigen display for antibody discovery workflows. We identify optimized conditions that improve results for downstream antibody discovery and validation.

Membrane proteins are among the hardest targets to express and study, and conventional extraction requires detergents, which are laborious to optimize and often destabilize the protein. We use deep learning-based computational protein design to create WRAPs, de novo proteins that wrap around a membrane protein's hydrophobic surface. Fused to the target, WRAPs enable expression in E. coli without detergents or changes to the native sequence. WRAPed proteins are thermostable and retain enzymatic and ligand-binding function, and a WRAPed porin yielded a 2.95 Ã… cryo-EM structure; we are currently exploring potential applications in drug discovery and vaccine design.

With the rapid advancement of AI-driven drug discovery, the production of quality antigens for screening and selection has emerged as a critical bottleneck. By engineering antigens for enhanced expression, solubility, and stability, we unlock previously intractable targets for antibody discovery. We highlight applications to GPCRs and structurally complex proteins, demonstrating how generative AI coupled with high-throughput experimental validation can expand the landscape of therapeutically accessible targets.

Precision bioconjugation starts with engineering proteins that carry reliable, site-specific reactive handles. We will introduce UniTet, an evolved genetic code expansion platform that enables efficient tetrazine incorporation at low substrate concentrations across bacterial and mammalian expression systems, supporting antibody fragments, full-length antibodies, and functional ADC production. A complementary antibody–IL-2 study uses rapid tetrazine ligation to assemble conjugates with varied attachment sites and linker lengths, revealing topology-dependent changes in regulatory T cell expansion, systemic exposure, and inflammatory cytokine responses in vivo. Together, these advances connect efficient production of conjugation-ready proteins with programmable therapeutic function.