Mechanism Study Of Positional Susceptibility Effect Of Chemically Modified SiRNA

Owing to its ability to specifically silence target genes, siRNA is widely used in thefunctional investigation of genes as well as therapeutic drug development. In theclinical application, siRNA has to be chemically modified to enchance its serumstability and decrease the off-target effect. Although the chemical modificationremarkably improves some properties of the native siRNA, the silencing activities ofsome chemically modified siRNAs were found to be greatly compromised. It is a bigchallenge to obtain optimal stabilizing effect while maintaining full silencing abilitydue to a lack of understanding of how different chemical modifications wouldinfluence the efficacy of siRNA. Much effort has been devoted into solving thisproblem, yet no systematic and comprehensive guideline has been proposed. Based onthis, we started current work on exploring the relationship between the site-specificchemical modification of siRNA and its corresponding silencing activity.To explore the effect of siRNA modification on gene silencing, a2’-OMemodification was individually made at every position in two siRNA antisense strands,except for the first nucleotide. The gene silencing potency of these siRNAs wasquantified to evaluate the influence of the modification on RNAi, using a previouslyreported assay system. The result that different positions of antisense showed differenttolerance effects on2’-OMe modification, hinted the positional susceptibility effect of chemically modified siRNA. Surprisingly, we found an unreported position (position14of antisense strand) showed great intolerance to2’-OMe modification in bothsiRNAs. To assess if position14is a general modification susceptibility position insiRNA,19more siRNA were synthesized and investigated. Instead of profiling theeffect across the whole siRNA strand, the four positions13,14,15and16were tested.For each position and nucleotide identity, three siRNAs were examined. Compared tothe other positions, an unique modification intolerance was identified at position14for all the tested siRNAs. Furthermore, our double modification assay demonstratedthat the silencing activity of multiple modifications resulted from the synergic effectsof each single modified position.We further studiesd the possible mechanism of site specific effect of modification onsilencing activity. We first examined and ruled out the possible contribution of Tmvalue change. Then we measured the amount of the14thposition-modified antisensewhich was loaded into the AGO2protein. Compared with the native antisense strand,the loading amount of modified antisense strand decreased substantially, which was inagreement with the decreased silencing activity. This result indicated that thedecreased loading amount of modified antisense was responsible for the diminishedsiRNA silencing activity. Furthermore, the dilution transfection assay demonstratedthat the decreased loading amount was not the only reason of the susceptibility ofposition14. Besides the RISC loading, modification could also compromise thefunctionality of RISC. To this end, we superimposed the human AGO2structure tothe structure of bacterial Thermus thermphilus AGO. The structural modelingdemonstrated that2’-modification at position14would affect the normal functionalityof the highly conserved L2loop in PIWI domain, and further disturb the RISC activity.Besides, we successfully decreased the sense-strand-mediated off-target effect byintroducing2’-OMe modification at the position14of sense strand.In this study, we have identified the position14of antisense strand as the mostintolerant position to2’-modification in siRNA in terms of maintaining silencing activity, and explored the underlying mechanism of this phenomenon. This studyprovided a guideline to optimize the strategy for siRNA modification.