# Structural Modelling of Krüppel‐Like Factor 15 Zinc Finger Binding Domain to DNA Using AlphaFold 3.0: Potential Therapeutic Target for Type 2 Diabetes

## TL;DR

This study models the DNA binding of KLF15 protein using AlphaFold and identifies key residues that could be targeted for treating type 2 diabetes.

## Contribution

The study identifies specific residues in KLF15's zinc finger domain critical for DNA binding, offering new therapeutic targets for type 2 diabetes.

## Key findings

- Residues K334A, R334A, Y332A, and R392A significantly affect KLF15's DNA binding affinity.
- Wild-type KLF15 ZnF domain shows stronger DNA binding energy than alanine-substituted variants.
- Unstable dynamics of substituted residues reduce binding free energy compared to wild-type.

## Abstract

Krüppel‐like factor 15 (KLF15) is a transcription factor contributing to the pathophysiology of multiple diseases, including metabolic syndromes. It is 416 residues long, with a C2H2‐type zinc finger (ZnF) domain that binds to GC‐rich regions regulating transcription. The role of KLF15 in glucogenesis and glucose level maintenance is well established. However, the DNA interaction mechanism at the atomic level remains unresolved. Here, we utilised computational structural biology tools to address this knowledge gap. The KLF15 ZnF–domain interacting with DNA was modelled with AlphaFold 3.0. Alanine substitution of the KLF15 ZnF domain–DNA complex revealed that residues K334A, R334A, Y332A and R392A significantly affect the binding affinities (ΔΔG) to DNA. To understand the conformational stability and dynamics of the KLF15 ZnF–domain complexes, 100‐ns molecular dynamics simulations were performed. Additionally, molecular mechanics‐generalised Born (MM/GBSA) surface area was utilised to calculate total binding energies. The binding energies of the wild‐type KLF15 ZnF domain (−94.0 ± 0.17 kcal/mol) demonstrated a more robust binding affinity to DNA than K334A (−30.4 ± 0.35 kcal/mol), R344A (−42.8 ± 0.37 kcal/mol), Y332A (−47.7 ± 0.42 kcal/mol) and R392A (−30.8 ± 0.30 kcal/mol). The findings highlighted the unstable dynamics of the alinine substituted resdiues that consequently reduce the binding free energy compared to the wild‐type KLF15 ZnF domain. In conclusion, the four identified residues are essential to recognise KLF15 ZnF DNA binding and can be considered potential hotspots for the therapeutics development for type 2 diabetes.