PAM in Crispr-cas9: The Adjacent Motif the Nuclease Needs

The protospacer adjacent motif in CRISPR-Cas9 is a short DNA sequence the nuclease must see next to its target before it cuts that site. For Streptococcus pyogenes Cas9 the usual class is NGG on the target DNA, immediately downstream of the protospacer. A perfect spacer that sits beside the wrong motif is not a near-miss. For that Cas9 class, it is not a site.

What the Adjacent Motif Is

PAM is a property of the target DNA, not a codon inside the guide. The guide spacer matches the protospacer. The PAM sits beside that stretch, on the DNA being inspected. Addgene’s Cas9-component note is blunt: Cas9 binds and cuts a dsDNA locus only if the target lies upstream of a PAM. Remove the guide RNA or remove the PAM, and Cas9 neither binds nor cuts.

The name is literal. Protospacer is the foreign sequence the CRISPR system once sampled. Adjacent motif is the short nearby word the nuclease uses as a license. In editing work the same geometry holds: the spacer you designed still needs that neighboring DNA word, or the protein will not open the duplex at that coordinate.

The audience is anyone reading a guide-design table for the first time. The purpose of the definition is to stop treating PAM as optional decoration next to a 20-base match. It is the first filter, not a footnote after seed scoring.

Why Cas9 Will Not Cut Without It

Jinek et al., 2012 put the requirement on the bench. In the S. pyogenes type II system the PAM conforms to an NGG consensus, with two G:C base pairs one base downstream of the crRNA-binding sequence. Transformation assays and in vitro cleavage both needed that GG motif. Mutating the PAM on the non-complementary strand of a dsDNA target blocked cleavage. Single-stranded targets did not show the same PAM dependence. The paper’s reading is that PAM licenses duplex unwinding and R-loop formation, not a later decorative check.

That is the information gain you can cite: missing PAM is a hard site filter for that nuclease on dsDNA, not a low off-target score. A design tool that still lists the coordinate is showing a spacer match. It is not showing a licensed Cas9 cut. Do not convert Jinek’s assay into a genome-wide NGG frequency law. The paper proves necessity at the sites it tested. It does not count how often NGG appears in a human chromosome.

NGG Is the SpCas9 Class, Not Every Cas Enzyme

Addgene’s CRISPR guide lists 3′ NGG for SpCas9 and different PAMs for other Cas proteins and engineered variants. Staphylococcus aureus Cas9 is commonly listed as NNGRRT-class. Cas12a enzymes in that table use a TTTV-class motif on the opposite side of the spacer. Those are different proteins with different licenses. Applying NGG to every CRISPR enzyme is a category error.

Engineered “PAM-flexible” Cas9s exist and are listed on the same Addgene table. They change which adjacent words count. They do not delete the idea of an adjacent motif. Do not write a PAM-free Cas9 product claim from a variant name. And do not turn Cas12a’s TTTV into a scaffold lecture here. Direct-repeat versus Cas9 sgRNA architecture is a different page, planned later in this series. This page only refuses to treat NGG as a universal CRISPR law.

Not Seed Matching and Not an Off-Target Score

Once a PAM is present, the nuclease still has to pair the spacer to the protospacer. The PAM-proximal stretch of that pairing is the seed. Seed mismatches are a different failure mode, and they already live on the guide seed-region page. A site can have a perfect NGG and still fail if the seed does not pair. That is not a missing-PAM problem.

Off-target cleavage is a third job: the nuclease cut somewhere you did not intend. That definition lives on the off-target page. A missing PAM is not an off-target score of zero. It is a site that this Cas9 class should not open. Mixing the three vocabularies — motif license, seed pairing, off-target — makes a guide table look more quantitative than it is. RNA architecture (tracrRNA versus an sgRNA cassette) is yet another object, already covered on the scaffold page.

A Site Without the Motif Is Off-Limits for That Nuclease

If the only available coordinate lacks the required PAM class, you do not “try it anyway” with wild-type SpCas9. You pick another coordinate or another nuclease whose motif matches the DNA that is actually there. A spacer that looks beautiful in a viewer is still unused sequence if the adjacent word is wrong.

After the motif rule is stated, a design surface such as ZettaCRISPR is one place to inspect remaining PAM-bearing candidates. Seeing a listed guide is not proof the nuclease will cut, and the tool does not invent a PAM-free cut. The takeaway is the license: no matching adjacent motif, no Cas9 cut at that dsDNA site.

Frequently Asked Questions

What happens if a candidate Cas9 site has no NGG-class PAM?

That SpCas9 class will not cut that dsDNA site. Pick another site or another nuclease with a matching PAM class.

Is the PAM part of the guide RNA spacer?

No. The PAM sits on the target DNA next to the protospacer. The spacer matches the protospacer, not the PAM.