eGFP is the largest published sequence inserted. Lots of us are working on this and its still a nascent field, but I suspect it all boils down to the guide RNA used and the locus targeted. I would try something like lacZ (3.5-kb) versus eGFP or perhaps luciferase (at ~1-kb) as controls in a cell line first before join in vivo.
In theory it should not be different from other nuclease mediated double strand breaks, here we and other succeeded in integrating over 5 kb (e.g. with ZFNs). Strange enough this seems to be more difficult using the CRISPR/Cas system and I find in the literature only integrations of about 1-2 kb integrating e.g. Cre or GFP (sometimes using 2 cuts at the same time). However on posters, I saw a study where they also managed to integrate larger sequences, but with a low efficiency. I would also highly appreciate an links for publications where people managed to integrate large fragments w/o.
Thank you for your info and suggestion. We were able to insert a 7.2 kb fragment into the genome of ES cells confirmed by 5' end PCR and real time PCR, and we are conducting Southern blot to further confirm the results. However, still wondering if we could get similar results in mouse fertilized eggs.
We did it normally but with a traditional target vector with 3kb and 9kb homologous arms and screened with G418. We didn't do a GFP/similar insertion, so there is no comparison but the efficiency is high compared to a tradition method, which was 10% once we got. The efficiency was calculated by dividing the correctly targeted numbers with the total G418 positive colony numbers. Just wondering if somebody tried similar and would like share?
Thanks! We're just getting started on doing this, so nothing to share just yet, but looking for pointers/tips. We're seeing those kinds of numbers (10% or higher) in HEK293 with smaller inserts, but haven't tested any large inserts.