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.

Genetic encoding of isobutyryl-, isovaleryl-, and ß-hydroxybutryl-lysine in E. coli.

Here we report the synthesis and genetic encoding of the lysine post translational modifications, ß-hydroxybutyryl-lysine, isobutyryl-lysine and isovaleryl-lysine. The ability to obtain a homogenous protein samples with site-specific incorporation of these acylated lysine residues can serve as a powerful tool to study the biological role of lysine post translational modifications.

Estimating Average Velocities of Particle Arrival Using the Time Duration of the Current Signal in Stochastic Blocking Electrochemistry.

In stochastic blocking electrochemistry, microparticles generate individual current steps when they adsorb on a microelectrode and decrease the current and flux of a redox mediator reacting at the surface. The amplitude of the current step informs on particle size and landing locus, while step frequency correlates with particle transport. Here, we report a new method to estimate the average arrival velocities of single rod-shaped bacteria (bacilli). The method relies on simulating the nearby threshold distance from the surface where the bacillus no longer perturbs mediator flux and the current step approaches zero. We estimated the average velocities of bacillus arrival by dividing the threshold distance over the current step duration, a parameter that here we detect for the first time and increases with bacillus length. By comparing diffusional fluctuations to bacillus average velocity, we estimated diffusion and migration contributions as a function of bacterium size. Average arrival velocities increase with bacillus length at the same time as migration intensifies and diffusion weakens. Our analysis is universal and more effective in determining transport mode contributions than the present approach of comparing theoretical and experimental step frequencies. Uncertainty in landing locus is inconsequential because the step duration used to calculate the average arrival speed already contains such information and knowing bacillus electrophoretic mobility or ζ-potential is not needed. Additionally, by simulating and assigning edge landings to the most repeated values of current steps in a recording, we obtain bacilli lengths and widths similar to scanning electron microscopy, from which we infer landing orientation.