De novo design of miniproteins as modulators for potassium ion channels Kv7
Abstract
Voltage-gated ion channels enable rapid and selective ion conduction across cell membranes and are essential for electrical signalling in excitable tissues. By shaping action potentials and resting membrane potential, they support core physiological processes including muscle contraction, cardiac rhythm and brain function. Dysfunction of these channels is associated with severe neurological and cardiovascular disease, making them important but challenging pharmacological targets. This project focuses on the Kv7, or KCNQ, family of voltage-gated potassium channels, which are implicated in epilepsy, sleep disorders and neurodevelopmental disease.
The overall goal is to discover and computationally validate new subtype- and state-selective modulators of Kv7 channels. Until now, the project was focused solely on identifying small-molecule activators and inhibitors that bind to selected Kv7 isoforms. However, small molecules often display off-target selectivity, as highlighted by the clinical history of retigabine/ezogabine, a Kv7.2/Kv7.3 activator that demonstrated the therapeutic potential of Kv7 modulation but was withdrawn after safety concerns and limited clinical use. Therefore, as parallel approach, we aim to design de novo mini-protein binders directed against extracellular or state-dependent structural features. The long-term motivation is to achieve higher selectivity of therapeutic agents toward Kv7 targets and thereby reduce adverse effects caused by broad or off-target channel modulation.
The project will combine structure-based modelling and machine-learning-assisted protein design. Available experimental structures will be used to construct representative channel conformations and subtype-specific target sites. For the protein-design, we will generate large libraries of mini-protein scaffolds using de novo design workflows, followed by sequence optimization, structure prediction, interface scoring and stability filtering. Top candidates will be refined by molecular dynamics simulations to assess binding stability, conformational compatibility and specificity across related Kv7 isoforms. Where possible, computational hits will be selected for available experimental assays, so that the project produces testable hypotheses rather than only in silico rankings.
The expected outcome of the project is a ranked and structurally justified set of candidates Kv7 modulators. The project will also deliver a reusable computational workflow for comparing subtype- and state-selective targeting strategies across Kv7 family members. This will provide mechanistic insight into which structural features are most suitable for selective modulation and will guide subsequent medicinal chemistry or protein-engineering optimization. Access to Berzelius is essential because the project requires high-throughput generation, prediction and filtering of large libraries of protein-design candidates, followed by GPU-accelerated structure prediction. The requested resources will therefore directly support the central scientific objective: to move from broad Kv7 modulation toward rationally designed, selective therapeutic candidates.
