Abstract
Haseki tick virus (HSTV) is a recently discovered virus detected in human serum following tick bites, yet its protein repertoire remains uncharacterized. In this study, we applied an integrative approach based first on membrane topology prediction, followed by AI-based structural prediction and experimental validation to annotate the structural part of the HSTV polyprotein. For the first time, we recombinantly expressed one of the putative HSTV structural protein (SP1) and determined its overall architecture using small-angle X-ray scattering (SAXS). Structural comparisons of the AI-predicted HSTV SP1 models revealed only a vague resemblance to the pestiviral Erns and Npro. The strong agreement between experimental SAXS data and the AI-predicted HSTV SP1 model supported the conclusion that HSTV SP1 adopts a distinct spatial architecture in solution, one that is not captured by existing pestiviral structures but is reliably represented by modern AI-based prediction. Our findings indicate that HSTV SP1 adopts a fold not previously observed among characterized members of the Flaviviridae family. This work establishes a methodological pipeline for characterizing highly divergent viral proteins and provides the first insights into HSTV SP1, a virus with emerging zoonotic potential. These results lay the foundation for future functional and structural studies, diagnostic development, and evolutionary analyses of atypical Flaviviridae family members.