Inserting DNA with CRISPR

 

 

Most prokaryotes rely on the CRISPR-Cas system for adaptive immunity against viruses and mobile elements (1, 2). Small RNAs produced from CRISPR direct Cas effector proteins to seek and destroy nucleic acids from invaders that have complementary target sites (3). There are multiple types of CRISPR, which are defined on the basis of their protein composition. Recently, RNA-guided nucleases from types II and V CRISPR systems, Cas9 and Cas12, have revolutionized genome editing by allowing programmed DNA sequence alterations (4). However, robust and targeted insertion of a large DNA segment into eukaryotic genomes has remained challenging. On page 48 of this issue, Strecker et al. (5) show that a CRISPR-associated transposase (CAST) mediates highly efficient, RNA-guided insertion of cargo DNA into the bacterial Escherichia coligenome. Moreover, Klompe et al. (6) report another CRISPR-guided DNA transposase system that operates similarly. These studies offer new tools that could transform genetic engineering and gene therapy research.

 

 

 

In a typical CRISPR application, Cas9 or Cas12 is directed by its guide RNA to the intended complementary genomic site flanked by a protospacer-adjacent motif (PAM), and creates a DNA double-strand break (DSB) (4). The outcome of gene editing largely depends on the repair of this DSB by the endogenous pathways in the host cell. Repair by end-joining DNA repair pathways often predominates and tends to introduce heterogeneous small DNA insertions or deletions (indels) that disrupt a gene’s function. However, many applications require site-specific knock-in of large DNA segments, such as a therapeutic gene or reporter genes. This can be achieved through accurate DSB repair by the homology-directed repair (HDR) pathway, which requires long sequence homology between a supplied DNA template and the genomic regions flanking the insertion site. HDR is inefficient and often restricted to certain stages of the cell cycle or by cell type (7, 8). Therefore, there is an urgent need for tools that enable robust and targeted DNA integration.

DNA transposons, also known as “jumping genes,” can move from one genomic location to another, and their relocations and insertions are catalyzed by an enzyme complex called transposase. Since 2017, the CAST systems have been discovered through bioinformatics mining of new CRISPR variants (9, 10). CAST comprises a miniature type I or V CRISPR-Cas encoded within a Tn7-like transposon. The mini-CRISPR systems encode protein machineries that either lack the Cas3 nuclease-helicase of the type I CRISPR system or carry inactivating mutations in the catalytic residues of Cas12k of the type V system. Accordingly, they could be competent for target sequence binding but not for DNA cleavage. It is speculated that these defective CRISPRs have been hijacked by Tn7-like transposons to serve a role other than prokaryotic adaptive defense. Instead, they might increase the evolutionary success of transposons by enabling their RNA-guided spread across bacterial genomes and other mobile DNA elements (9, 10). The putative RNA-programmable nature of CAST contrasts with the poor target site selectivity of canonical transposases, which insert their cargo DNA semirandomly.

Strecker et al. established the functionality of two CAST loci from cyanobacteria Scytonema hofmanni (Sh) and Anabaena cylindrica (Ac) in vivo and elucidated the underlying mechanism. Each CAST locus is ∼20 kilobases, comprising Tn7-like transposase genes (tnsB, tnsC, tniQ), additional cargo genes, and genes encoding a type V-K CRISPR and its effector Cas12k as well as trans-activating crRNA (tracrRNA), which is an invariant small RNA cofactor that is essential for Cas12k. The authors assayed potential transposition into plasmid DNA in the heterologous host E. coli. A helper plasmid encoding all CAST protein and RNA components was delivered into E. coli, together with a donor plasmid and a target plasmid (see the figure). Targeted donor integration occurred in one specific orientation, with an efficiency of 13 to 75%. Integration was guided by CRISPR-Cas12k, because it requires tracrRNA, a cognate target, and a 5′ flanking GTN PAM. Most notably, the insertion sites were clustered unidirectionally within a narrow window of DNA sequence, 60 to 66 base pairs for ShCAST and 49 to 56 base pairs for AcCAST, downstream from the PAM. Donor cargo DNA of up to 10 kilobases was integrated efficiently into the target plasmid.

 

•  Download high-res image
•  Open in new tab
•  Download Powerpoint

 

 

The PAM and target sequence for CRISPR-Cas12k remain intact after transposition, yet they cannot support additional rounds of effective CAST integration into the same downstream region that already received a copy of the transposon. This observation resembles the “target immunity” phenomenon where bacterial T7 transposon avoids repeated insertion into the same site (11). Notably, the 5–base pair sequence before each integration site is duplicated to flank the other end of the inserted DNA. This is consistent with staggered single-stranded DNA gaps generated by T7 transposons (12), which are presumably sealed by gap-filling host factors. Taken together, CAST is a bona fide CRISPR-guided, prokaryotic transposon.

Strecker et al. reconstituted CAST-catalyzed DNA insertion in vitro and also demonstrated that CAST could be repurposed as a targeted DNA insertion tool for bacterial genome engineering. ShCAST was reprogrammed against 48 different target sites in noncoding regions of the E. coligenome and achieved targeted insertion at 29 loci. For a few sites, the efficiencies are impressively high (50 to 80%), obviating the need for positive selection to detect transposition. Insertions were also profiled on a genome-wide scale using unbiased deep sequencing. Approximately half of the total insertions were on-target, whereas the off-target insertions were scattered along the chromosome. Many of the top off-target sites occurred repeatedly between E. coli samples for distinct guide RNAs and thus are likely caused by a CRISPR-independent mechanism. Limiting the amount of transposase produced in the cells may help reduce off-target effects.

Klompe et al. characterize a transposon from bacteria Vibrio cholerae that employs a mini–type I CRISPR-Cas to insert cargo DNA in two possible orientations. The integration is CRISPR-guided and robust, and displays low (less than 10%) off-target activity.

Can we leverage the CRISPR-guided DNA transposase systems reported by Strecker et al. and Klompe et al. for highly efficient, targeted genome integration in eukaryotes? Compared to the current viral- or transposon-based therapeutic gene delivery into human cells (13), CRISPR-guided DNA transposase systems could be used to insert custom genes into the desired site and therefore avoid oncogenic risks associated with random genomic integration. Unlike Cas9- or Cas12-based editing tools, CRISPR-guided DNA transposase systems do not require DSB repair by the host HDR pathways, and therefore may enable highly efficient gene knock-ins in a wider variety of cell types and tissues. However, a limitation is that the terminal ends of the transposon will be incorporated together with the intervening cargo DNA, making applications that require “scarless” insertion impossible. The discovery of the CRISPR-guided DNA transposase systems again illustrates the power of bioinformatic mining of microbial genomes in uncovering new CRISPR variants with functions beyond defense (9, 10). As more CRISPR-linked accessory genes are discovered, the noncanonical facets of CRISPR biology will be further revealed (14, 15).

 

 

(원문: 여기를 클릭하세요~)