Toxins, Vol. 17, Pages 600: Ai-Assisted Discovery of a Direct Physical Interaction Between a Venom Serpin from the Parasitoid Wasp Liragathis javana and a Host Serine Carboxypeptidase

Toxins doi: 10.3390/toxins17120600

Authors:
Jiale Wang
Xunyuan Jiang
Zemiao Xiao
Xuemei Tang
Kai Wan

Parasitoid wasp venoms represent highly specialized biochemical arsenals that evolved to manipulate host physiology and ensure successful development of the parasitoid offspring. However, the molecular targets and mechanisms underlying this complex host modulation remain poorly understood. To address this, we employed an AI-driven discovery pipeline, integrating the sequence-based predictor D-SCRIPT with the structural modeler AlphaFold3, to characterize LjSPI-1, a venom serpin from Liragathis javana. This computational workflow highlighted a previously unreported candidate partner—a host serine carboxypeptidase (Chr09G02510). Crucially, we detected a direct physical interaction between these two proteins through both in vitro pull-down and in vivo yeast two-hybrid assays, supporting this AI-prioritized interaction under experimental conditions. Our study identifies a high-priority molecular pairing and demonstrates the utility of an AI-guided strategy for uncovering candidate targets of venom proteins. In addition, guided by the predicted biochemical role of Chr09G02510, we propose several plausible physiological hypotheses linking this interaction to host peptide metabolism and immune modulation. These hypotheses serve as a conceptual basis for future mechanistic and toxicological investigations.

Source link

Jiale Wang www.mdpi.com