From AI modelling to validation: A 10-day pipeline for potential lead compounds validation

Area of Science:

• Agricultural Science
• Biochemistry
• Mycology

Background:

• Fusarium graminearum causes significant crop losses and produces deoxynivalenol (DON) mycotoxins.
• Coproporphyrinogen III oxidase (Cpo) is a crucial enzyme in haem biosynthesis and a potential fungicide target.
• No inhibitors targeting FgCpo have been previously identified or screened.

Purpose of the Study:

• To develop and validate a rapid pipeline for screening FgCpo-targeted antifungal inhibitors.
• To identify novel lead compounds for combating Fusarium head blight (FHB).
• To explore the potential of FgCpo as a fungicide target.

Main Methods:

• An integrated 10-day pipeline involving protein modelling (AlphaFold3), virtual screening (1.6 million compounds), molecular dynamics simulations, and bioassays.
• Dual-docking strategies were employed for efficient virtual screening.
• EC50 and MIC values were determined, alongside deoxynivalenol (DON) inhibition assays and gene expression analysis.

Main Results:

• A high-precision FgCpo 3D model was generated.
• Eight high-affinity compounds were identified, with Chemdiv_2925 showing competitive binding.
• Chemdiv_2925 demonstrated broad-spectrum antifungal activity (EC50: 23.55–198.83 μg/mL), inhibited F. graminearum spore germination (MIC: 41 μg/mL), and reduced DON production by 69.6%.

Conclusions:

• The 10-day pipeline enables rapid screening and validation of potential antifungal compounds.
• Chemdiv_2925 is a promising lead compound targeting FgCpo for pesticide development.
• This approach accelerates the discovery of efficient, low-toxicity fungicides for crop protection.