Slaw, K. | bioRxiv Science

bioRxiv2026

Deep-learning-based design of an orthogonal self-labeling protein from K-Ras(G12C)

Self-labeling protein (SLP) tags enable versatile labeling of proteins in live cells, yet only two SLPs are commonly used, limiting multiplexing. Here, we repurposed the oncoprotein K-Ras(G12C) and its covalent inhibitors to create a third, orthogonal SLP system. We used a deep-learning-based approach to radically redesign the sequence of K-Ras(G12C) while preserving its covalent-inhibitor binding pocket. Our top design, LUCI-tag, is a 19 kDa, monomeric, thermostable SLP that rapidly and covalently reacts with commercially available K-Ras(G12C) inhibitors bearing diverse payloads. Unlike existing SLPs, LUCI-tag exhibited payload-agnostic labeling kinetics, outperforming HaloTag7 and SNAP-tag for a negatively charged payload. X-ray crystal structures of drug-bound and drug-free LUCI-tag showed the design was structurally accurate and contained a preorganized inhibitor-binding pocket that could explain its rapid labeling kinetics. Proteomics and cell-signaling experiments confirmed LUCI-tag is biologically inert. LUCI-tag enabled rapid, wash-free, live-cell imaging and simultaneous three-color multiplexed experiments with HaloTag7 and SNAP-tag. This work establishes LUCI-tag as an immediately useful orthogonal SLP and demonstrates that covalent drug-target pairs can be repurposed into a broadly applicable platform for protein labeling.

bioengineering↗

bioRxiv2025

Zero-shot design of drug-binding proteins via neural selection-expansion

Computational design of molecular recognition remains challenging despite advances in deep learning1-3. The design of proteins that bind to small molecules has been particularly difficult because it requires simultaneous optimization of protein sequence, protein structure, and ligand conformation1-7. Despite their promise, current deep-learning algorithms have struggled to navigate this landscape, precluding the zero- or few-shot design of binders. Here we show that the combination of two neural networks in an iterative design algorithm can create small-molecule binding proteins from scratch with high accuracy. To optimize a design in the joint distribution of sequence, structure, and ligand conformation, we use a pair of neural networks that were trained on reciprocal tasks. We train and use a graph neural network, LASErMPNN, to design protein sequence given protein-ligand co-structure, and we use RoseTTAFold-All Atom8 (RFAA) to predict protein-ligand co-structure given protein sequence. We iteratively apply these two networks to design proteins that bind the drug, exatecan, a topoisomerase I inhibitor that is prone to inactivation by hydrolysis9. Each of four experimentally tested designs bound the drug, with the lowest dissociation constant (Kd) near 100 nM. The hit rate and highest affinity design each surpassed the current state-of-the-art method by 5- and 70-fold, respectively. We further show that LASErMPNN can improve upon its own designs in a manner resembling chain-of-thought reasoning. Without experimental input, LASErMPNN suggested two mutations that increased affinity by over two orders of magnitude (Kd = 1.2 {+/-} 0.2 nM). Designs were selective, structurally accurate, and achieved their intended purpose to protect the drug from hydrolysis. Our work describes a recipe for using neural networks to automate the design of high affinity small-molecule binding proteins, which should have wide application in the creation of novel drug-delivery vehicles, antidotes, sensors, and enzymes.

bioengineering↗