We present a de novo antibody design platform that generates high-affinity, drug-like leads against traditionally challenging targets including GPCRs, transporters, and pMHCs. Our approach leverages high hit rates to enable direct screening in a native cellular context, allowing us to address proteins that are difficult to recombinantly express and purify.

Intrinsically disordered proteins (IDPs) and regions (IDRs) underpin many of the most “undruggable” processes in biology, from transcriptional control to phase separation and toxic aggregation. In this talk, I will present a deep-learning–enabled framework for designing de novo binders and proteases that recognize short motifs (>=8 aa) and PTM-defined epitopes in IDRs with high affinity and specificity. I will highlight applications to Myc, tau, TDP-43, huntingtin, and viral or oncogenic IDRs, cancer receptors, and show how these reagents enable targeted inhibition, relocalization, and catalytic cleavage; and how this general modality can be applied to broadly unexplored biological and therapeutic ideas.

We have been working to improve BEVS by experimenting with different insertion sites in the baculovirus genome and testing alternative baculovirus promoters to improve quality and yield for more complex target proteins. We have also demonstrated significant improvement in stability of the gene of interest over multiple passages.

Cell-free protein synthesis (CFPS) is emerging as a transformative platform for recombinant therapeutic production. Here, we present advances in bacterial CFPS systems that improve yield, scalability, and product quality while accommodating complex modalities, including post-translational modifications like glycosylation. These innovations position CFPS to accelerate therapeutic discovery, streamline development, and enable flexible, distributed biomanufacturing for next-generation biologics.

Three chemically defined media for CHO-K1 production of the VRC01 mAb were compared in this study regarding growth, titer, and N- glycosylation. ActiCHO P achieved high productivity and the most consistent glycan profiles, closely matching ActiPro, whereas EX-CELL 325 showed lower performance. Because media changes can alter critical quality attributes, comparability is essential. Results indicate ActiCHO P is a reliable alternative medium without compromising product quality.

Production of a wide variety of clinically relevant G-protein coupled receptors at quantities sufficient for a variety of downstream applications is challenging. Therefore, we established and improved screening methods to evaluate membrane protein construct expression and solubilization. Furthermore, we show translatability of production from milliliter to 10-liter scale. Using several examples, we show how our purification workflow can produce GPCRs, including for immunization, structural biology, and biophysical characterization.

GPCR targets are recombinantly expressed in soluble form using machine learning–designed engineered scaffolds. These engineered GPCR surrogates express well, support post-translational modifications by production in human cells, bind specifically to their native ligands, and are structurally validated using experimental methods. This strategy enables high-yield, soluble expression of previously intractable GPCRs in a functionally and structurally validated format to support drug discovery efforts.

GPCRs are a class of highly druggable targets that are difficult to access in recombinant systems in amounts and quality sufficient for binding-first methods, including DNA Encoded Library (DEL) screening for hit finding and then also for follow-up of hits. In this talk, I will review some of our successful strategies and tactics for accessing these targets using membrane mimetics including SMALPs and other detergent-free formulations.

Access to recombinant proteins is vital for biotechnology. Cell-free protein synthesis systems could address this need, but widespread utilization remains limited by cost and complexity. To address these limitations, we carried out a multi-dimensional optimization, testing more than 1,200 reagent formulations, to discover a simple and reproducible system based on 12 components that can produce up to 3.7 ± 0.2 g/L at a ~99% reduction in cost. We also demonstrate cell-free production of fifteen therapeutically relevant products using this formulation. We anticipate that our work will further democratize the use of cell-free systems for protein manufacturing and biotechnology.

We present a systems engineering framework for microbial protein production that enables plug-and-play integration of advanced tools to unlock new manufacturing capabilities. By combining scalable two-stage expression, redox reprogramming for efficient disulfide bond formation, and cell-programmed downstream processing, we establish a unified platform for nanobody production.

Antibodies have broad utility in imaging, targeted gene delivery, and disease therapy, and many of these applications require conjugation to secondary molecules. Unfortunately, conventional conjugation approaches are limited by destabilization of structure, heterogeneity, and technically demanding multi-step reactions. To overcome these challenges, we developed a straightforward and highly general platform for site-specific antibody conjugation that blends metabolic glycoengineering with protein design, presenting a highly efficient strategy to produce antibody conjugates.