In their recent paper Chatterjee and colleagues (2018) describe a natural protospacer adjacent motif (PAM) plasticity of previously uncharacterized Cas9 endonuclease from Streptococcus canis (ScCas9). Chatterjee et al (2018) report ScCas9’s affinity to minimal 5’-NNG-3’ PAM sequences and describe its accurate editing capabilities, proposing ScCas9 as an alternative genome editing tool.
Different RNA-guided endonucleases of the CRISPR-Cas system are widely used as an essential component for genome editing and regulation. As all of them have constrains which limit the choice of sequences available for targeting and modification, researchers have been describing or engineering new Cas-variants toexpand targeting capabilities.
Cas9, the most widely used variant, has been derived from Streptococcus pyogenes (SpCas9) and for successful gene modification it requires a 5’-NGG-3’ protospacer adjacent motif (PAM) downstream from the target.
Chatterjee et al (2018) describe a Cas9 ortholog from Streptococcus canis (ScCas9) which prefers a minimal 5’-NNG-3’ PAM. The researchers validated the PAM binding sequence, as well as ScCas9 specificity and base editing efficiency in assays with human cells and suggested the newly described ortholog as an effective alternative for targets with 5’-NNG-3’ PAM sequences.
The authors proposed that two insertions in ScCas9 sequence, which differentiate it from SpCas9, cause the relaxed PAM specificity. To confirm this hypothesis, they engineered ScCas9 without one of the two or both insertions and showed that while variants without one of either insertions have a modified PAM specificity, removal of both insertions changed binding preference toward the canonical 5’-NGG-3’ PAM, indicating an intermediate specificity between that of SpCas9 and ScCas9.
However, in engineered SpCas9 with either one or both above mentioned insertions, PAM binding either failed or did not change significantly from the original SpCas9, which suggests that while insertions play an important role in PAM preference, other sequence features also have their contribution.
Chatterjee et al (2018) also conducted an off-target survey using ScCas9. Based on previous studies they chose three genomic sites in the human genome which are known to be SpCas9 off-target sites and a single off-target site identified by others while engineering of high-fidelity Cas9 variants. In the survey, SpCas9 confirmed high on- and off-targeting in all three sites. While ScCas9 showed comparable on-target activity, two out of three sites had negligible off-target rates and the remaining site showed 1.5-fold decrease in off-target ratio.
In a mismatch tolerance survey where ScCas9 accuracy was tested over a wider PAM targeting range, it generally tolerated single mismatches better than double ones, while SpCas9 showed weak or no tolerance. The researchers compared ScCas9 and SpCas9 modification ability and efficiency in 24 targets in human genes and confirmed their previous findings, however, detected some variation efficiency within each PAM group and across different genes, therefore, they conclude that careful choice of the target presents a crucial step for successful ScCas9-gene modification.
More development is needed for the broad application of ScCas9 since while it showed a better efficiency in non–5’-NGG-3’ targets than SpCas9, the latter still was more efficient in editing 5’-NGG-3’ genomic targets.
To facilitate future discoveries of new Cas9 proteins with novel PAM characteristics Chatterjee et al (2018) have established an open-access pipeline SPAMALOT. This resource intends to help expanding the targeting range of CRISPR, and therefore, enrich the current genome modification toolbox.