It is clear that even the fully functional SaCas9 did not work properly, therefore none of the subsequent split SaCas9 worked either. Although we were not able to draw any conclusions from this data in the future we would like to debug our initial SaCas9 design and then continue to validate splits in the same workflow as the integrases where we will look for patterns between split site locations, different dimerization domains, and their orientations. It would be very exciting to find viable split sites and create a system for genomic editing using split SaCas9.
Towards the end of our project, the Feng Zhang group at MIT published the crystal structure of saCas9 in Cell1. The paper demonstrated the development of a conditionally dimerizable saCas9 system using the FKBP-FRB domain pairs and PYL-ABI domain pairs, which is what we wanted to accomplish.
One facet worth noting is that the Zhang group used the 3D structure of the saCas9 protein to target flexible linker regions that would serve as potential split sites. Out of the three split sites that they chose, the one that worked best was almost identical (one amino acid away) to one of the split sites that we identified using our model. Furthermore, the other two split sites that they chose which were not as successful happened to be located in regions with secondary structural elements.
This important result gave us some experimental validation that our model could offer scientists insights into choosing promising viable split sites for proteins, without only relying on 3D structures of proteins.
Citations
- Nureki, Osamu et al., “Crystal structure of staphylococcus aureus Cas9”, Cell, 2015.