Back

A virus co-opts vector's m⁶A machinery for adaptive compensation

Fu, B.; Hu, J.; Liu, Q.; Zhang, R.; Wu, X.; Guo, X.; Liu, S.; Huang, M.; Gong, P.; Wei, X.; Yang, J.; Tan, Q.; Liang, J.; He, C.; Zhou, X.; Nauen, R.; Shi, X.; Zhang, Y.; Bass, C.; Yang, X.

2026-08-27 microbiology
10.64898/2026.08.27.747426 bioRxiv
Show abstract

Insect vectors harbor diverse viruses that threaten human health and agricultural production, yet the genetic basis of mutualistic virus-vector interactions remain elusive. This underscores a long-standing paradox in pest management: viruses that cause destructive diseases also facilitate vector adaptation in ways that benefit both parties. Here we show that the globally devastating tomato yellow leaf curl virus (TYLCV) acts as a Cooperative Partner, co-opting its vector's m6A epitranscriptomic machinery to offset the reproductive fitness costs of neonicotinoid resistance in the whitefly. In resistant vectors, TYLCV infection alleviates reproductive deficits via the m6A-dependent pathway, where the methyltransferase METTL14 and demethylase ALKBH4 coordinately stabilize vitellogenin (Vg) transcripts to boost female fecundity. Mechanistically, the viral C2 and CP proteins directly interact with METTL14 and ALKBH4, respectively, remodeling their RNA binding affinity to facilitate mA modification on Vg. Disrupting either of these viral-vector protein interactions restores a reproductive cost phenotype in viruliferous resistant vectors, offering novel targets for sustainable pest control. Our findings uncover a fundamental role of epigenetic marks in adaptive trait, reshaping beneficial virus-vector relationships and providing new insights for innovative public health and crop protection strategies to address global challenges.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.