Date: 10th December 2019
The CRISPR-Cas systems are currently the hottest family of gene editing tools to hit our news and offer hope for new gene therapy treatments. With Cas9 being the figurehead of the family and currently showing promising signs in clinical trials other members of the Cas family are starting to emerge from the shadows.
Cas9, 12 and 13 all belong to the CRISPR/Cas Class 2 system and we have reported the use of a high fidelity Cas12a which exhibited reduced off-target effects , and a Cas13-derived strategy as an anti-viral agent.
However, more recently the Class 1, CRISPR-Cascade system has come to our attention. Whereas Class 2 utilise a single effector, Class 1 are comprised of multi-subunit complex effectors.
In October, Rodolphe Barrangou and his team, transcriptionally activated endogenous genes in plants using the CRISPR-Cascade system (mediated by Cas3) after previously showing its successful use in human cells (http://eha.3a7.mywebsitetransfer.com/new-crispr-tool-eukaryotic-gene-regulation/). However, this system still remains relatively unexplored, and its potential for use in gene therapy is very much in its infancy.
Now scientists from Japan have used the Cas3-based Class 1 CRISPR system to successful repair a gene mutation responsible for Duchenne muscular dystrophy in patient-derived cells.
The team wanted to assess the DNA cleavage activity of the Cascade system in human cells by characterising efficiency and off-target events.
A luciferase-based single-strand annealing (SSA) recombination assay was set up as an initial assessment tool. This consisted of a luciferase reporter plasmid whereby the luciferase cassette was split and could recombine into a translationally active form following CRISPR-Cas-induced double-strand break and SSA. Further plasmids containing six individual Cascade genes, including Cas3 (an endonuclease), were also synthesised, together with a plasmid containing the CRISPR RNA (crRNA) sequence (complementary to the target DNA). The plasmids were lipofected into a human cell line and luciferase activity was determined as a measure of DNA cleavage and repair.
The group found that the CRISPR-Cas3 system showed robust reporter activity, and required all components of the Cascade machinery. They were able to assess efficiencies of targeting various PAM sequences and further characterisation of the catalytic activities of Cas3 revealed that DNA cleavage was through both its nuclease and helicase domains.
The next question to address was whether the Cascade system could target endogenous genes. In order to assess this, crRNA target sites for the EMX1 and CCR5 genes were selected and designed in human cells and next generation sequencing (of PCR amplicons) was used to determine edits.
Whilst small insertion/deletion (indel) mutations were induced by Cas9 (used as a comparison), Cas3-mediated DNA targeting revealed bulk deletions of several kb and mutations of less than 100 bp were not observed.
To further assess the general efficiency of the CRISPR-Cas3 system, the authors then chose to target eight endogenous genes and again compared the efficiency with the Cas9 system at the same loci. Cas3-induced large deletions with up to 60.3% editing efficiency were observed whereas Cas9 induced small indels with up to 78.2% editing efficiency.
Whole genome sequencing was then utilised to provide an unbiased method of assessing the deletions and these showed that whereas the CRISPR-Cas3 system mediated a broad range of DNA indels upstream of the target site in human cells, Cas9 facilitated small indels.
Whilst the data presented here clearly showed the capacity of Cas3 to mediate gene knock-outs, it was unclear whether gene knock-in could be performed.
To explore this the group utilised a mutated reporter system, and demonstrated that the restoration of fluorescence using a donor ‘replacement’ vector could be achieved at a higher efficiency using the Cas3 Cascade compared to the Cas9 system.
Cas3 and Cas9 systems were then both thoroughly assessed for off-target effects. The data supported a high level of ‘on-target’ specificity with the CRISPR-Cas3 system, which was comparable or lower than the off-target activity observed for the CRISPR-Cas9 system.
The ability of any gene editing system to translate into a therapeutic setting is a giant leap. In order to test this, the group wanted to introduce relevant genome edits in patient-induced pluripotent stem cells (iPSCs).
As a proof-of-concept- experiment to assess long-range deletion by the Cas3 system they induced exon skipping in the dystrophin gene which, when mutated in humans, causes Duchenne Muscular Dystrophy (DMD). By targeted skipping of a specific exon, a DMD phenotype can be converted into a milder phenotype thus alleviating symptoms. This previously published work showed that restoration of the dystrophin protein reading frame in iPSCs from a DMD patient who lacked exon 44, could be achieved by skipping exon 45.
Initial experiments were performed in human cell lines, and single-cut CRISPR-Cas3 showed significantly higher levels of exon skipping than CRISPR-Cas9 with two sgRNAs (cas9 requires two sgRNAs for long-range deletions).
The team then wished to test the CRISPR-Cas3 system further and DMD-iPSCs were isolated from a DMD patient harbouring a deletion of exon 44, and these were targeted for exon 45 skipping by the CRISPR-Cas3.
The system was able to cause exon skipping, and thus ‘corrected’ the genome with an efficiency of up to 14%. Two iPSC clones containing the edits were subsequently differentiated into skeletal muscle cells and showed successful restoration of the DMD protein.
The study here demonstrates that CRISPR-Cas3 can be used successfully in human cells and is able to edit the genome in a variety of ways.
Importantly it offers advances over Cas9 which is currently most widely used in a clinical setting. Whilst it is true that Cas9 more efficiently edits the genome at the target sites, Cas3 is more efficient at both knock-outs and –ins at distances of a few dozen to a few hundred bp upstream of the targeted site. This may offer significant advantages in regions that are difficult to target by Cas9, such as those far away from PAM sequences, repetitive sequences or those near transposon elements.
Furthermore, Cas3 can induce large deletions with a single crRNA, whereas Cas9 requires two sgRNAs. Not only is Cas3 more proficient at these types of edits, it is likely this is a contributory factor to less off-targets effects.
One potential worry is unpredictable, uncontrollable large genomic deletions by Cas3. These may induce genome instability and toxicity in mammalian cells although the authors did not witness such events in this study. However, this will be an important area for future investigation and may involve other factors such as anti-CRISPR proteins. However, adding complexity to an already multifaceted system will also increase safety concerns and will have to be addressed later down the line.
The complexity of the system due to its multi-effectors also brings with it inherent problems of delivery. The encasing and targeting of cargo in nanoparticles, or with shuttle peptides for example has been successfully demonstrated for Cas9, but has yet to be tested for the CRISPR-Cas3 system. However, with technology accelerating at such speed this is not likely to be inhibitory for long.
However, long-range deletions lend themselves well for fighting infectious disease, bacteria pathogens or, for example, editing non-coding RNAs. These are often involved in a wide range of diseases such as neurological, cardiovascular, developmental and cancer and hold great therapeutic interest. As such it is likely this new CRISPR tool will be adopted readily, with potential applications in drug discovery, disease prevention, and crop improvement this may be the new rising star of the family.
Morisaka, H., K. Yoshimi, Y. Okuzaki, P. Gee, Y. Kunihiro, E. Sonpho, H. Xu, N. Sasakawa, Y. Naito, S. Nakada, T. Yamamoto, S. Sano, A. Hotta, J. Takeda and T. Mashimo (2019). “CRISPR-Cas3 induces broad and unidirectional genome editing in human cells.” Nature Communications 10(1): 5302.