gRNA design - complimentary to the target or the non target strand?

[Katherine Cummins]

Morning

Have gotten myself confused - when designing gRNA I had always picked a sequence (using Benchling) with the 17-20bp sequence, say 5' to 3' on the sense strand, followed by NGG, all on the same strand. 

Doing some reading however, and it seems that the gRNA sequence actually binds to the opposite strand to which the PAM is on; so there is the target sequence on the target strand, and the PAM sequence on the non-target strand (reference below)

Does this mean that gRNA should be a sequence that is complementary to the non-PAM strand?

I'm so confused. 

Katherine

Nature Biotech 2016, Enhancing homology-directed genome editing by catalytically active and inactive CRISPR-Cas9 using asymmetric donor DNA, Christopher D Richardson1,2, Graham J Ray1,2, Mark A DeWitt1,2, Gemma L Curie1,2 & Jacob E Corn1,2

[Gunjan Kumar]

Hi Katherine,

That is indeed correct! 

Your sgRNA should be designed to target the strand opposite that which contains the PAM. An easier way to think of this would be that your sgRNA sequence should be exactly the same as the 18-20 bp sequence that comes directly before the PAM on the non-target strand.

Here's a link from IDT that helps to better visulize this: https://www.idtdna.com/pages/support/faqs/what-is-a-pam-sequence-and-where-is-it-located

Hope this clears things up!

Cheers,

Gunjan

[Sverre Grødem]

Hi Gunjan,

If this is the case, then why do so many guide design tools give you guides on the PAM-containing strand? I see that Synthego does not, but both chopchop and Crispr-era do. 

Puzzled.

Sverre

[Gunjan Kumar]

Hi Sverre,

Reading through the Clontech site, it seems that they are suggesting that it doesn't matter which strand is targeted. However, all the target optimized libraries available consistently target the non-PAM containing strand. This falls in line with the conformational changes of the Cas9 after PAM recognition and the positioning of the RuvC and HNH endonuclease domains (http://sites.tufts.edu/crispr/crispr-mechanism/rna-guide/). So, unless I am misunderstanding something about Cas9 biology, to me it doesn't make sense that the targets would be against the PAM containing strand...

If there is anyone else who has a different idea about what's going on, please do feel free to chime in!

Best,

Gunjan