Check out how we performed in vivo cytosine base editing using LNP mediated delivery and how we analyzed on- and off-target effects of this approach! It was a pleasure to work on this piece together with @lukas_villiger @schwanklab and all our amazing collaborators!! https://twitter.com/schwanklab/status/1353762353226113027
1/8 Cytosine base editors (CBEs) have previously been shown to elicit guide-independent off-target deamination on RNA and DNA. Here, we use different approaches for short- and long-term expression of CBEs and evaluate their off-target effects in the mouse liver in vivo.
2/8 Lipid nanoparticles allow transient expression of CBEs, long enough for therapeutic editing to occur (we achieve 21% editing in a phenylketonuria mouse model). This transient expression already reduces the risk for unwanted side effects of BEs.
3/8 We performed whole-transcriptome sequencing of edited livers directly and 1m after LNP-injection and and also compare these to AAV-treated livers. Surprisingly, in vivo expression of CBEs does not lead to excessive off-target RNA deamination as was shown in previously.
4/8 A reason for this could be that CBE expression in the mouse liver using LNP- or AAV-delivery is much lower (<1000tpm) compared to CBE expression in HEK293T cells after plasmid delivery (150 000 tpm!), where significant off-target deamination was found.
5/8 Exposure to CBE, still poses the risk of unguided DNA deamination. We therefore isolate primary hepatocytes from treated animals, clonally expand them, confirm CBE exposure by on-target sequencing to then perform whole-genome sequencing on 11 AAV- and 24 LNP-treated clones.
6/8 This makes it possible to get rid of the high noise generated by single-cell DNA amplification or low sensitivity of bulk sequencing.
7/8 We further assess our detection limit by computationally spiking in APOBEC signatures. This allows us to estimate that off-target DNA deamination from CBE expression in vivo by LNP or AAV does not generate more than 50 SNVs per genome.
8/8 Considering that the genomes of human liver cells naturally accumulate thousands of mutations over a lifetime, the therapeutic benefit of CBEs and their engineered variants with even lower off-target deamination may outweigh the potential risk of a few additional SNVs.