Cas9 (CRISPR associated protein 9) is an RNA-guided DNA endonuclease enzyme associated with the CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) adaptive immunity system in Streptococcus pyogenes, among other bacteria. S. pyogenes utilizes Cas9 to memorize[1] and later interrogate and cleave foreign DNA,[2] such as invading bacteriophage DNA or plasmid DNA.

Cas9 (CRISPR associated protein 9) is an RNA-guided DNA endonuclease enzyme associated with the CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) adaptive immunity system in Streptococcus pyogenes, among other bacteria. S. pyogenes utilizes Cas9 to memorize[1] and later interrogate and cleave foreign DNA,[2] such as invading bacteriophage DNA or plasmid DNA. Cas9 performs this interrogation by unwinding foreign DNA and checking for sites complementary to the 20 basepair spacer region of the guide RNA. If the DNA substrate is complementary to the guide RNA, Cas9 cleaves the invading DNA. In this sense, the CRISPR-Cas9 mechanism has a number of parallels with the RNA interference (RNAi) mechanism in eukaryotes.

Apart from its original function in bacterial immunity, the Cas9 protein has been heavily utilized as a genome engineering tool to induce site-directed double strand breaks in DNA. These breaks can lead to gene inactivation or the introduction of heterologous genes through non-homologous end joining and homologous recombination respectively in many laboratory model organisms. Alongside zinc finger nucleases and TALEN proteins, Cas9 is becoming a prominent tool in the field of genome editing.

Cas9 has gained traction in recent years because it can cleave nearly any sequence complementary to the guide RNA.[2] Because the target specificity of Cas9 stems from the guide RNA:DNA complementarity and not modifications to the protein itself (like TALENs and Zinc-fingers), engineering Cas9 to target new DNA is straightforward.[3] Versions of Cas9 that bind but do not cleave cognate DNA can be used to localize transcriptional activator or repressors to specific DNA sequences in order to control transcriptional activation and repression.[4][5] While native Cas9 requires a guide RNA composed of two disparate RNAs that associate to make the guide – the CRISPR RNA (crRNA), and the trans-activating RNA (tracrRNA).,[2] Cas9 targeting has been simplified through the engineering of a chimeric single guide RNA (chiRNA). Scientists have suggested that Cas9-based gene drives may be capable of editing the genomes of entire populations of organisms.[6] In 2015, scientists used Cas9 to modify the genome of human embryos for the first time.[7]

To survive in a variety of challenging, inhospitable habitats that are filled with bacteriophage, bacteria have evolved methods to evade and fend off predatory viruses. This includes the recently appreciated CRISPR system. CRISPR loci are composed of short, palindromic repeats that occur at regular intervals composed of alternate CRISPR repeats and variable CRISPR spacers. These CRISPR loci are usually accompanied by adjacent CRISPR-associated (cas) genes. In 2005, it was discovered by three separate groups that the spacer regions were homologous to foreign DNA elements, including plasmids and viruses. These reports provided the first biological evidence that CRISPRs might function as an immune system.

CRISPR-Cas systems divided into three major types (type I, type II, and type III) and twelve subtypes, which are based on their genetic content and structural differences. However, the core defining features of all CRISPR-Cas systems are the cas genes and their proteins: cas1 and cas2 are universal across types and subtypes, cas3, cas9, cas10 are signature genes for type I, type II, and type III respectively.

The CRISPR-Cas defense can be described in three stages:

Stage 1: CRISPR spacer integration. Protospacers and protospacer-associated motifs (shown in red) are acquired at the “leader” end of a CRISPR array in the host DNA. The CRISPR array is composed of spacer sequences (shown in colored boxes) flanked by repeats (black diamonds). This process requires Cas1 and Cas2 (and Cas9 in type II[1]), which are encoded in the cas locus, which are usually located near the CRISPR array.

Stage 2: CRISPR expression. Pre-crRNA is transcribed starting at the leader region by the host RNA polymerase and then cleaved by Cas proteins into smaller crRNAs containing a single spacer and a partial repeat (shown as hairpin structure with colored spacers).

Stage 3: CRISPR interference. crRNA with a spacer that has strong complementarity to the incoming foreign DNA begins a cleavage event (depicted with scissors), which requires Cas proteins. DNA cleavage interferes with viral replication and provides immunity to the host. The interference stage can be functionally and temporarily distinct from CRISPR acquisition and expression (depicted by white line dividing the cell).

Cas9 features a bi-lobed architecture with the guide RNA nestled between the alpha-helical lobe (blue; Fig. 2) and the nuclease lobe (cyan, orange and gray). These two lobes are connected through a single bridge helix. There are two nuclease domains located in the multi-domain nuclease lobe, the RuvC (gray) which cleaves the non-target DNA strand, and the HNH nuclease domain (cyan) that cleaves the target strand of DNA. The RuvC domain is encoded by sequentially disparate sites that interact in the tertiary structure to form the RuvC cleavage domain (See Figure 3).

A key feature of the target DNA is that it must contain a protospacer adjacent motif (PAM) consisting of the three-nucleotide sequence- NGG. This PAM is recognized by the PAM-interacting domain (PI domain, orange) located near the C-terminal end of Cas9. Cas9 undergoes distinct conformational changes between the apo, guide RNA bound, and guide RNA:DNA bound states, which are detailed below.

Cas9 recognizes the stem-loop architecture inherent in the CRISPR locus, which mediates the maturation of crRNA-tracrRNA ribonucleoprotein complex.[11] Cas9 in complex with CRISPR RNA (crRNA) and trans-activating crRNA (tracrRNA) further recognizes and degrades the target dsDNA.[12] In the co-crystal structure shown here (Fig. 4), the crRNA-tracrRNA complex is replaced by a chimeric single-guide RNA (sgRNA, in red) which has been proved to have the same function as the natural RNA complex.[2] The sgRNA base paired with target ssDNA is anchored by Cas9 as a T-shaped architecture. This crystal structure of the DNA-bound Cas9 enzyme reveals distinct conformational changes in the alpha-helical lobe with respect to the nuclease lobe, as well as the location go the HNH domain. protein consists of a recognition lobe (REC) and a nuclease lobe (NUC). It should be noted that all regions except the HNH form tight interactions with each other and sgRNA-ssDNA complex, while the HNH domain forms few contacts with the rest of the protein. In another conformation of Cas9 complex observed in the crystal, the HNH domain is not visible. These structures suggest the conformational flexibility of HNH domain.

So far, three crystal structures have been studied and published. One representing a conformation of Cas9 in the apo state, and two representing Cas9 in the DNA bound state.

In sgRNA-Cas9 complex, based on the crystal structure, REC1, BH and PI domains have important contacts with backbone or bases in both repeat and spacer region.[15][14] Several Cas9 mutants including REC1 or REC2 domains deletion and residues mutations in BH have been tested. REC1 and BH related mutants show lower or none activity compared with wild type, which indicate these two domains are crucial for the sgRNA recognition at repeat sequence and stabilization of the whole complex. Although the interactions between spacer sequence and Cas9 as well as PI domain and repeat region need further studies, the co-crystal demonstrates clear interface between Cas9 and sgRNA.

Previous sequence analysis and biochemical studies have suggested Cas9 contain RNase H and HNH endonuclease homologous domains which are responsible for cleavages of two target DNA strands, respectively. These results are finally proved in the structure. Although the low sequence similarity, the sequence similar to RNase H has a RuvC fold (one member of RNase H family) and the HNH region folds as T4 Endo VII (one member of HNH endonuclease family). Previous works on Cas9 have demonstrated that HNH domain is responsible for complementary sequence cleavage of target DNA and RuvC is responsible for the non-complementary sequence.[16]

Due to the unique ability of Cas9 to bind to essentially any complement sequence in any genome, researchers wanted to use this enzyme to repress transcription of various genomic loci. To accomplish this, the two crucial catalytic residues of the RuvC and HNH domain can be mutated to alanine abolishing all endonuclease activity of Cas9. The resulting protein coined ‘dead’ Cas9 or ‘dCas9’ for short, can still tightly bind to dsDNA. This catalytically inactive Cas9 variant has been used for both mechanistic studies into Cas9 DNA interrogative binding and as a general programmable DNA binding RNA-Protein complex.

The interaction of dCas9 with target dsDNA is so tight that high molarity urea protein denaturant can not fully dissociate the dCas9 RNA-protein complex from dsDNA target.[17] dCas9 has been targeted with engineered single guide RNAs to transcription initiation sites of any loci where dCas9 can compete with RNA polymerase at promoters to halt transcription.[18] Also, dCas9 can be targeted to the coding region of loci such that inhibition of RNA Polymerase occurs during the elongation phase of transcription.[18] In Eukaryotes, silencing of gene expression can be extented by targeting dCas9 to enhancer sequences, where dCas9 can block assembly of transcription factors leading to silencing of specific gene expression.[5] Moreover, the guide RNAs provided to dCas9 can be designed to include specific mismatches to its complementary cognate sequence that will quantitatively weaken the interaction of dCas9 for its programmed cognate sequence allowing a researcher to tune the extent of gene silencing applied to a gene of interest.[18]This technology is similar in principle to RNAi such that gene expression is being modulated at the RNA level. However, the dCas9 approach has gained much traction as there exist less off-target effects and in general larger and more reproducible silencing effects through the use of dCas9 compared to RNAi screens.[19] Furthermore, because the dCas9 approach to gene silencing can be quantitatively controlled, a researcher can now precisely control the extent to which a gene of interest is repressed allowing more questions about gene regulation and gene stoichiometry to be answered.

Beyond direct binding of dCas9 to transcriptionally sensitive positions of loci, dCas9 can be fused to a variety of modulatory protein domains to carry out a myriad of functions. Recently, dCas9 has been fused to chromatin remodeling proteins(HDACs/HATs) to reorganize the chromatin structure around various loci.[18] This is an important in targeting various eukaryotic genes of interest as heterochromatin structures hinder Cas9 binding. Furthermore, because Cas9 can react to heterochromatin, it is theorized that this enzyme can be further applied to studying the chromatin structure of various loci.[18] Additionally, dCas9 has been employed in genome wide screens of gene repression. By employing large libraries of guide RNAs capable of targeting thousands of genes, genome wide genetic screens using dCas9 have been conducted.[20]

Another method for silencing transcription with Cas9 is to directly cleave mRNA products with the catalytically active Cas9 enzyme.[21] This approach is made possible by hybridizing ssDNA with a PAM complement sequence to ssRNA allowing for a dsDNA-RNA PAM site for Cas9 binding. This technology makes available the ability to isolate endogenous RNA transcripts in cells without the need to induce chemical modifications to RNA or RNA tagging methods.

In contrast to silencing genes, dCas9 can also be used to activate genes when fused to transcription activating factors.[18] These factors include subunits of bacterial RNA Polymerase II and traditional transcription factors in eukaryotes. Recently, genome-wide screens of transcription activation have also been accomplished using dCas9 fusions named ‘CRISPRa’ for activation.[20]