SaLT&PepPr is an interface-predicting language model for designing peptide-guided protein degraders.

Protein-protein interactions (PPIs) are critical for biological processes and predicting the sites of these interactions is useful for both computational and experimental applications. We present a Structure-agnostic Language Transformer and Peptide Prioritization (SaLT&PepPr) pipeline to predict interaction interfaces from a protein sequence alone for the subsequent generation of peptidic binding motifs. Our model fine-tunes the ESM-2 protein language model (pLM) with a per-position prediction task to identify PPI sites using data from the PDB, and prioritizes motifs which are most likely to be involved within inter-chain binding. By only using amino acid sequence as input, our model is competitive with structural homology-based methods, but exhibits reduced performance compared with deep learning models that input both structural and sequence features. Inspired by our previous results using co-crystals to engineer target-binding "guide" peptides, we curate PPI databases to identify partners for subsequent peptide derivation. Fusing guide peptides to an E3 ubiquitin ligase domain, we demonstrate degradation of endogenous ß-catenin, 4E-BP2, and TRIM8, and highlight the nanomolar binding affinity, low off-targeting propensity, and function-altering capability of our best-performing degraders in cancer cells. In total, our study suggests that prioritizing binders from natural interactions via pLMs can enable programmable protein targeting and modulation.

Engineering Single Pan-Specific Ubiquibodies for Targeted Degradation of All Forms of Endogenous ERK Protein Kinase.

Ubiquibodies (uAbs) are a customizable proteome editing technology that utilizes E3 ubiquitin ligases genetically fused to synthetic binding proteins to steer otherwise stable proteins of interest (POIs) to the 26S proteasome for degradation. The ability of engineered uAbs to accelerate the turnover of exogenous or endogenous POIs in a post-translational manner offers a simple yet robust tool for dissecting diverse functional properties of cellular proteins as well as for expanding the druggable proteome to include tumorigenic protein families that have yet-to-be successfully drugged by conventional inhibitors. Here, we describe the engineering of uAbs composed of human carboxyl-terminus of Hsc70-interacting protein (CHIP), a highly modular human E3 ubiquitin ligase, tethered to differently designed ankyrin repeat proteins (DARPins) that bind to nonphosphorylated (inactive) and/or doubly phosphorylated (active) forms of extracellular signal-regulated kinase 1 and 2 (ERK1/2). Two of the resulting uAbs were found to be global ERK degraders, pan-specifically capturing all endogenous ERK1/2 protein forms and redirecting them to the proteasome for degradation in different cell lines, including MCF7 breast cancer cells. Taken together, these results demonstrate how the substrate specificity of an E3 ubiquitin ligase can be reprogrammed to generate designer uAbs against difficult-to-drug targets, enabling a modular platform for remodeling the mammalian proteome.

DNA nanostructures coordinate gene silencing in mature plants.

Delivery of biomolecules to plants relies on Agrobacterium infection or biolistic particle delivery, the former of which is amenable only to DNA delivery. The difficulty in delivering functional biomolecules such as RNA to plant cells is due to the plant cell wall, which is absent in mammalian cells and poses the dominant physical barrier to biomolecule delivery in plants. DNA nanostructure-mediated biomolecule delivery is an effective strategy to deliver cargoes across the lipid bilayer of mammalian cells; however, nanoparticle-mediated delivery without external mechanical aid remains unexplored for biomolecule delivery across the cell wall in plants. Herein, we report a systematic assessment of different DNA nanostructures for their ability to internalize into cells of mature plants, deliver siRNAs, and effectively silence a constitutively expressed gene in Nicotiana benthamiana leaves. We show that nanostructure internalization into plant cells and corresponding gene silencing efficiency depends on the DNA nanostructure size, shape, compactness, stiffness, and location of the siRNA attachment locus on the nanostructure. We further confirm that the internalization efficiency of DNA nanostructures correlates with their respective gene silencing efficiencies but that the endogenous gene silencing pathway depends on the siRNA attachment locus. Our work establishes the feasibility of biomolecule delivery to plants with DNA nanostructures and both details the design parameters of importance for plant cell internalization and also assesses the impact of DNA nanostructure geometry for gene silencing mechanisms.