AAV vector-derived elements integrate into Cas9-generated double strand breaks and disrupt gene transcription

We previously found that an adeno-associated virus (AAV) vector containing S. aureus Cas9 and a multi-target guide (g)RNA could integrate into the genome and prematurely terminate transcription of Ube3a-ATS, a long non-coding RNA. Here, we assessed the performance of three additional AAV vectors containing S. aureus Cas9 and twenty-five vectors containing N. meningococcus Cas9, all targeting single sites within Ube3a-ATS. We found that none of these single-target gRNA vectors were as effective as multi-target gRNA vectors at reducing Ube3a-ATS expression in neurons. We also developed a new anchored multiplex PCR sequencing (AMP-seq) method and analysis pipeline to quantify the relative frequency of all possible editing events at target sites, including unresolved double-stranded breaks (DSBs) and capture of foreign DNA. We found that integration of AAV was the most frequent editing event (67-89% of all edits) at three different single target sites, far surpassing insertions and deletions (indels). None of the most frequently observed indels was capable of blocking transcription when incorporated into a Ube3a-ATS minigene reporter, whereas two vector derived elements—the polyA and reverse promoter—reduced downstream transcription by up to 50%. Since not all editing events disrupt gene transcription, our findings suggest that the probability that a gene trapping AAV integration event occurs is influenced by which vector-derived element(s) are integrated and by the number of target sites. AMP-seq on DNA from neurons treated with various NmCas9-gRNA AAVs: NmCas9-AAV gRNA23 (0-7 day treatment), NmCas9-AAV gRNA25 (7 day treatment), NmCas9-AAV gRNA21 (7 day treatment), NmCas9-AAV non-targeting gRNA (7 day treatment), or untreated DNA. AMP-seq on DNA from neurons of P1 (post-natal-day 1) C57BL/6 mice treated with SaCas9-gRNA AAVs (SaCas9-AAV-G3) or SaCas9-AAV non-targeting gRNA.