Stable recombinant proteins that resemble the native target structure and function are essential for the advancement of pipeline projects. The complexity of membrane protein targets poses a unique series of challenges, and we have developed methods to produce these targets in formats useful for biologics discovery. I will present our evolving platform utilizing virus-like particles and other methods to enable targeting complex transmembrane targets.

Successful therapeutic antibody production requires that your antigen sample contains proteins that exist in therapeutically relevant conformations. For membrane protein antigens, this remains a significant challenge, as the membrane which stabilizes active conformation is often absent from the sample. Our development of SMALP solubilization resolved this by including the intact local membrane environment. In this talk, I discuss new enhanced polymers that can be used for conformational trapping of proteins.

Virus-like particles (VLPs) have emerged as a viable format for displaying complex membrane protein antigens for lead discovery and screening. However, VLPs present off-target epitopes that may convolute successful lead discovery. Herein, we will present a case study that deployed an optimized VLP-based immunization, as well as a VLP engineering approach that yielded 59 leads to a GPCR of therapeutic interest.

Multipass membrane proteins are an important class of therapeutic targets. However, generating diverse and effective antibody repertoires against these targets poses a significant challenge due to their complex structure and limited surface accessibility. In this perspective, we adopt various approaches such as leveraging transgenic animal platforms that produce fully human antibody repertoires and implementing a comprehensive immunization and screening strategy tailored specifically for the multipass membrane targets.

Multi-pass transmembrane proteins are difficult targets for antibody generation due to the challenge of making a protein with native conformation. I will present an mRNA-based immunization strategy that resulted in the identification of antibodies that bind to extracellular epitopes of challenging transmembrane proteins. This strategy incorporates mRNA-based antigens together with boosting, sorting, and screening strategies to yield antibodies against difficult target proteins.

Multi-pass membrane proteins represent a pivotal challenge for antibody discovery platforms. Leveraging a yeast-based, single B cell discovery platform, we have developed a high-throughput methodology for isolating full-length antibodies from both wildtype and humanized transgenic murine diversities against membrane-obligate targets. The discovery process results in large panels of high-affinity, clonally-diverse antibodies without the aid of a soluble recombinant antigen.

Post-translational modifications (PTMs) play a key role in human health and diseases. However, analysis of PTMs for diagnostic and therapeutic purposes is hindered by the difficulty of creating PTM-specific reagents using conventional methods. By combining rational engineering, structural analysis, and in vitro evolution we developed powerful probes for proteomics, genome-wide binding analysis, and imaging, allowing atomic resolution of the post-translationally modified proteome.

Adhesion G Protein-Coupled Receptors (aGPCRs) contain 33 family members that influence development and tissue homeostasis in a wide range of tissues. Signaling is thought to be induced when ligand binding liberates or unmasks a tethered agonist (TA) normally sequestered within a membrane proximal GAIN domain. We describe here a robust platform for evaluating TA-dependent and independent outputs and apply it to all 33 human aGPCRs. Our dataset is a rich source of signaling information with relevance to both biological investigation of different family members and development of therapeutics.

The isolation of binding proteins from combinatorial libraries has typically relied on the use of a soluble, recombinantly-expressed form of the target protein when performing magnetic selections or fluorescence-activated cell sorting. Appropriate target protein expression and subsequent purification represents a significant bottleneck for both library screening and binder characterization. As an alternative, the use of target proteins expressed on the surface of magnetized yeast cells in combinatorial library screening and quantitative assessment of binding affinities is discussed.

The Merus Biclonics platform utilizes proprietary common light chain technologies to rapidly generate panels of multi-specific antibodies. Using unbiased screening approaches, large numbers of multi-specific antibodies can be tested in relevant biological assays to identify those with the strongest biological responses. We will showcase the application of an organoid-based unbiased screening that led to the discovery of our anti-EGFR×Lgr5 Biclonics, MCLA-158.

The voltage-gated sodium (Nav) channels play a key role as a mediator of action potential propagation in C-fiber nociceptors and are established molecular targets for pain therapy. We used available structural and experimental data to guide Rosetta design of potent and selective natural peptide-based inhibitors of human Nav1.7 channels. Our lead peptides are highly potent, selective, and stable, and inhibit NaV1.7 currents in human sensory neurons. Rationally-designed peptide inhibitors of human NaV1.7 channels have transformative potential to define a new class of biologics to treat pain.

Outer membrane proteins (OMPs) control all interactions between Gram-negative bacteria and their environments including uptake and efflux of antibiotics. We created an algorithm that identifies bacterial OMPs from sequence. The quality of our algorithm allows us to identify ~1.8 million OMPs from prokaryotic genomes including 5 classes of structurally-unresolved OMPs. Our web-accessible database allows for further exploration of outer membrane proteins to uncover new targets for controlling antibiotic resistance.

Using computational methods that score physical features of protein structures or apply artificial intelligence approaches, homologs and de novo proteins are screened in silico to replace human signaling proteins and their receptors. Predicted HLA-II epitopes that pose immunogenicity risks are removed through targeted mutations without damaging structure or activity. The proteins have a range of affinities, specificities, and signaling properties, which are tuned through experimental deep mutagenesis and affinity optimization.