Biology subjects

Duve, T.

Publications and source records attributed to Duve, T..

3 recordsLinked to original sources

bioRxiv2025

Martini 3 coarse-grained models of azobenzene-based photolipids: Modulation of membranes with light

Photoswitchable lipids offer an attractive way to manipulate the biophysical properties of membranes by means of light. Their application to modulate membrane properties, manipulate membrane proteins, and photocontrol cargo release is gaining popularity. Here, we present coarse-grained Martini 3 models for azobenzene and azobenzene-based photoswitchable lipids. Our models show good agreement with atomistic reference simulations. Furthermore, we apply our coarse-grained photolipid models to study photocontrol of lateral phase separation, protein flexibility, and membrane permeability. The results agree well with experimental data from the literature and highlight the broad applicability of our Martini 3 photolipid models. They will enable studying the impact of photolipid switching on large membrane systems as well as on their (bio)molecular interaction partners.

biophysics↗

bioRxiv2025

Photopharmacology in Action: Conformational Landscape of a Photoswitchable Covalent Kinase Inhibitor

Photopharmacology is a rapidly evolving field that uses light to control drug activity with high spatial and temporal precision, offering innovative therapeutic strategies with reduced side effects. In this study, we investigate a photoswitchable covalent inhibitor for the MAP kinase JNK3, a target for the treatment of neurodegenerative diseases. The inhibitor, which is based on a diazocine photoswitch, can undergo reversible photoisomerization of a double bond, switching between two (meta-)stable isomers. Atomistic molecular dynamics simulations, comprising almost 40 {micro}s of total simulation time, reveal how the distinct conformational spaces of the two isomers modulate their interactions with JNK3 within the ATP-binding pocket. We show that only the metastable trans isomer can form a covalent bond with JNK3, thereby permanently inhibiting its function. In contrast, for the cis state the distance between the inhibitor and the targeted cysteine residue is too large to allow covalent bond formation. Furthermore, our simulations reveal that the covalent bond, in combination with the environment of the protein pocket, hinders the full back-relaxation of the trans isomer to the stable cis isomer. Finally, our data shows that the covalently bound stable cis isomer modulates the conformations of JNK3, mainly of its activation loop. Our findings provide molecular insights into the complex dynamics of photoswitchable inhibitors, guiding future drug design for light-controlled therapeutics.

biochemistry↗

bioRxiv2025

Martini 3 protein models - a practical introduction to different structure bias models and their comparison

Biophysical characterization of protein structure and dynamics is essential in many scientific fields, including molecular biology, drug discovery, and enzyme design. Molecular dynamics (MD) simulations have become an increasingly important tool for studying these properties. This chapter provides a hands-on introduction to protein modeling with the Martini 3 coarse-grained (CG) force field. We outline its two-layer framework, where the first layer defines molecular topology and interactions, while the second layer applies structural bias to maintain secondary, tertiary, and quaternary structures. Three structure bias approaches - Elastic Network (EN), G[o]Martini, and OLIVES - are discussed, highlighting their advantages and trade-offs. Using a protein kinase as a case study, we demonstrate the step-by-step setup of Martini 3 protein models for simulations with the program package GROMACS, including system preparation, fine-tuning, and validation against atomistic reference simulations. Additionally, we explore the combination of an intrinsically disordered region (IDR) with a structured protein domain. By the end of this chapter, readers will have the necessary expertise to apply Martini 3-based protein modeling to a wide range of research applications.

biophysics↗