Using AlphaFold2 to Predict the Conformations of Side Chains in Folded Proteins

Maisuradze, G.; Thakur, A.; Khatri, K.; Haldane, A.; Levy, R. M.

doi:10.1101/2025.02.10.637534

Home/Archive/bioRxiv

bioRxiv · 10.1101/2025.02.10.637534

Using AlphaFold2 to Predict the Conformations of Side Chains in Folded Proteins

Maisuradze, G.·Thakur, A.·Khatri, K.·Haldane, A.·Levy, R. M.

View original source ↗Export metadata ↓

Abstract

AlphaFold has revolutionized protein structure prediction by accurately creating 3D structures from just the amino acid sequence. However, even with extensive research validating its overall accuracy, a key question remains: Can AlphaFold predict the conformation of individual amino acid residue side chains within a folded protein? This is important for the field of molecular modeling, particularly when predicting the effects of mutations on protein stability and ligand binding. AlphaFold generates a set of atomic coordinates not just for the mutated side chain but also for potential rearrangements across the entire protein structure. In this study we investigate the ability of ColabFold, an online implementation of AlphaFold2 (AF2), to predict the conformations of residue side chains in folded proteins. We find that over a set of 10 benchmark proteins, the side chain conformation prediction error of ColabFold is on average [~]14% for {chi}1 dihedral angles, and increases to [~]48% for {chi}3 dihedral angles. The prediction error is smaller for non-polar side chains and is somewhat improved using structural templates. ColabFold demonstrates a bias towards the most prevalent rotamer states in the protein data bank (PDB), potentially limiting its ability to capture rare side chain conformations effectively. As an application of AlphaFold to explore the structural consequences of strongly cooperative mutations on side chain rearrangements, we employ a Potts sequence-based statistical energy model to perform large scale mutational scans of two proteins ABL1 and PIM1 kinase, searching for the most strongly cooperative mutational pairs, and then use ColabFold to predict the structural signatures of this cooperativity on the interacting side chains. Our results demonstrate that integration of the sequence-based Potts model with AlphaFold into a single pipeline provides a new tool that can be used to explore the fundamental relationship between protein mutations, cooperative changes in structure, and fitness.

Explore related subjects

biophysics Preprint

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Maisuradze, G., Thakur, A., Khatri, K., Haldane, A., Levy, R. M.. 2025-02-14. Using AlphaFold2 to Predict the Conformations of Side Chains in Folded Proteins. https://doi.org/10.1101/2025.02.10.637534

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Design of colipids with controlled lateral distance of head group moieties for targeting strategies in the field of lipid-based nanoparticles

Lipid nanoparticles and liposomes are nanoparticle drug formulations which are firmly established in clinics. In order to expand their application and improve efficacy, surface modifications can be a key strategy, especially for cell specific targeting. Targeting can be achieved by integrating ligand functionalized lipids or lipidoids on the nanoparticle surface. In this case, not solely the targeting ligand itself, but also the structure of the whole targeting lipid/lipidoid can affect the interaction of the targeting ligand with the binding motive on cells. Especially, distal spacer structures between lipid anchor and ligand unit, often composed of polyethylene glycol chains, provide a certain distance of the ligand from the nanoparticle surface. A concept which is currently neglected is the control of the lateral distance between ligand structures on the nanoparticle. Herein we investigated the possibility to design lipidoids with a lateral spacer composed of polyethylene glycol units. We synthesized and characterized different lipidoids which provide a primary amine in the head group for possible ligand coupling. In our model lipidoids the head groups were functionalized with a paramagnetic probe to apply electron paramagnetic resonance (EPR) methods for the characterization of the lipidoid interaction with model nanoparticles, in our case liposomes. Continuous wave EPR was used to study the efficiency of the novel lipidoids to integrate into lipid bilayers of model liposomes, which was shown to depend on lipidoid structure and liposome composition. Furthermore, using double electron-electron resonance (DEER), we observed lipidoid conformational changes occurring upon integration into liposomes.

biophysics↗

bioRxiv2026

Binding mechanisms of therapeutic antibodies to human claudin-18.2

Claudin-18.2 (CLDN18.2) is a tight junction component with highly restricted expression in the paracellular space of nonmalignant gastric epithelial cells. It becomes accessible on tumor cell surface during malignant transformation, making CLDN18.2 an attractive target for targeted therapies against gastric cancer. Recently, zolbetuximab has become the first FDA-approved monoclonal antibody that targets CLDN18.2 in advanced gastric cancer, and others like osemitamab have entered phase 3 clinical trials. However, the high-resolution structure of CLDN18.2 is lacking, and how CLDN18.2 is recognized by therapeutic antibodies remains undefined. Here, we report cryo-EM structures of human claudin-18.2 (hCLDN18.2) in complex with the Fab fragments of zolbetuximab and osemitamab, revealing the structure of hCLDN18.2 and its binding mechanism to the therapeutic antibodies. Further analyses reveal that three residues in the first extracellular loop of hCLDN18.2 that are different from human claudin-18.1 (hCLDN18.1) play a critical role in antibody binding. Our data elucidate the structural basis for selective binding of therapeutic antibodies to hCLDN18.2, which should aid rational design of next-generation immunotherapies targeting hCLDN18.2.

biophysics↗

bioRxiv2026

Structural relationships between the SLC12 cotransporter C-terminal domain, universal stress proteins and adenylating enzymes

Several SLC12 cotransporters bind ATP in a cytoplasmic C-terminal domain (CTD). We analysed 16,192 CTD-containing SLC12 sequences and compared 42,845 representative structures to examine the relationship between this nucleotide site, universal stress proteins (USPs) and adenylating enzymes. In phosphorylated NKCC1, D891 and D927 lie near the ATP phosphate chain, while R801 and K889 form a basic phosphate-binding clamp. NKCC1 retains this arrangement, whereas KCC paralogs replace or lose several corresponding groups. In structural comparisons with USP and adenylating-enzyme references, both CTD lobes score between universal stress proteins and EC-annotated domains, and the two lobes form separate lineages that include bacterial transporters. Lobe-specific and nucleotide-frame comparisons place ATP in the corresponding structural region used for -phosphate chemistry by catalytic members of this fold. Published work on UspG provides a precedent for protein-bound AMP in a natural USP. In the examined SLC12 structures and USP models, candidate protein hydroxyls lack an in-line geometry for attack at ATP P. Protein adenylation would require a change in protein or nucleotide geometry, or a different acceptor. These results define a structural basis for testing conditional protein adenylation at SLC12 nucleotide sites. The required evidence is a covalent protein-AMP adduct, identification of its acceptor and measurement of the expected pyrophosphate coproduct, with soluble nucleotide analysis to distinguish accompanying ATP reactions.

biophysics↗