Probably one of the fastest races of current times is the one protagonised by CRISPR/Cas. The discovery that a sophisticated defensive weapon against invasions by phages and naked plasmids, naturally present in bacterial and archaea genomes, could be used to engineer genomes has started a whirlwind of research and development. CRISPR/Cas is now widely recognised and used as the most versatile, easy-to-use, multiplexed tool that we have in our gene editing repertoire to date. The ability to create a precise double strand break (DSB) wherever we want in the genome/DNA of any organism, including humans, has opened an immense set of opportunities of modifying genes with additions, deletions and many other DNA alterations. Repairing mutated human genes is one of the most exciting applications of this new technology, with the potential to facilitate and advance three decades of gene therapy. Since its development in 2013, CRISPR/Cas has been tested preclinically to treat an increasing variety of human pathologies, from infectious diseases (HIV, hepatitis B) to genetic diseases such as hereditary tyrosinaemia, muscular dystrophy, sickle cell anaemia, Huntington's disease, or cystic fibrosis [1, 2]. In less than 5 years, the first clinical trial has been just launched at the West China Hospital in Chengdu for the delivery of engineered T cells to patients with aggressive lung cancer. And now, a lot more trials are waiting for final approval and expecting to start this year. However, although this race is very fast, is not without hurdles.
Class 2, Type II CRISPR/Cas9 System from Streptococcus thermophilus and S. pyogenes. The CRISPR locus contains a CRISPR array with repeat regions (black diamonds) separated by spacer regions (coloured rectangles) derived from invaders. Then four protein coding genes with cas9 among them and a tracrRNA (trans acting CRISPR RNA). The CRISPR array and the tracrRNA are transcribed, giving rise to a long pre-crRNA and a tracrRNA. These two RNAs hybridize via complementary sequences and are processed to shorter forms by Cas9 and RNase III. The resulting complex (Cas9+ tracrRNA + crRNA) then begins searching for the DNA sequences that match the spacer sequence (shown in red). Binding to the target site also requires the presence of the protospacer adjacent motif (PAM), which functions as a molecular handle for Cas9 to grab on to. Once Cas9 binds to a target site with a match between the crRNA and the target DNA, it cleaves the DNA three bases upstream of the PAM site. Cas9 contains two endonuclease domains, creating a double strand break (DSB) with blunt ends. The original configuration had 2 RNA segments: crRNA and tracrRNA while the engineered one has a unique chimeric combination of both.
For instance, this new and powerful editing tool is still used mainly in applications with a therapeutic benefit coming from indels (insertions and deletions) induced by the non-homologous end-joining (NHEJ) DNA repair pathway which is error-prone, as it is preferentially used in somatic adult cells. High fidelity homology-directed repair (HDR) uses either the wild type non-mutated homologous chromosome or an exogenous DNA as a template to rebuild the damaged genomic sequence. The efficiency of the system in gene repair and gene addition that is mediated by HDR is one of the biggest challenges that CRISPR/Cas technology has to face and amend to get widely used in repairing disease-causing mutations. Some of the improvements include rational design of sgRNAs (modifying the scaffold or reducing the length of the complementarity) sequence), template DNA donors (the design of optimal length single-stranded DNA donors or the exploitation of the inverse relationship between a mutation’s incorporation rate and its distance to the DSB achieving to achieve predictable control of zygosity), small inhibitor molecules of the NHEJ pathway or chemical modifications of the sgRNAs (which increase the stability and functionality of the sgRNA/Cas9 complex by decreasing the RNAse susceptibility of CRISPR guides) . Thanks to crystallographic and electron-microscopy data of the Cas9 protein structure and its interaction with the sgRNA, new Cas9 variants with enhanced affinities and activities could be engineered.
CRISPR/Cas9 mechanism. The Cas9 nuclease (purple rounded rectangle with white arrows signalling the DSB) binds to a single-guide RNA (sgRNA), which guided it to a specific region of the target DNA called the protospacer. The sgRNA is a hybrid of two components: CRISPR RNA (crRNA; blue) and trans-activating crRNA (tracrRNA; red). By NHEJ repair pathway, the cleavage of the target DNA by the Cas9 nuclease results in mutations that can knock out the target gene. If an appropriate donor DNA molecule is available, through HDR, correction of the sequence or genetic information (in green) can be added to the targeted DNA in a precise manner creating a knock in. By targeting two Cas9 nucleases to different regions of the target DNA, it is possible to delete the genetic information between the two regions or get an inversion of the sequence. If this occurs in two different chromosomes, we can get a translocation. CRISPR: clustered regularly interspaced short palindromic repeats; Cas: CRISPR-associated.
Besides increasing the efficiency, one key question that has always concerned researchers is how specifically the endonuclease Cas9 is modifying the intended DNA sequence of interest. This means that we would like to keep control of the possibility of unintendedly modify similar sequences to the gRNAs in the genome (=off target effects) and avoid them as this would be of utmost importance in clinical settings. Besides bioinformatics tools, of limited utility, screening of CRISPR-treated cells with a genome-wide method of detection of DSBs (GUIDE-Seq and high-throughput gene translocation sequencing, HTGTS) can provide an answer. Currently, however, these techniques have relatively low efficiency in some primary cells and they need high sequence coverage. Two very recently published methods, SITE-Seq  and CIRCLE-seq , from the groups of Andrew May and Keith Joung, respectively, could represent an advantage. Both methods rely on an in vitro very sensitive biochemical phase, in which the naked isolated genomic DNA is cleaved with the CRISPR/Cas components and any target –either intended (on) or not (off)- gets specifically detected, enriched and sequenced. Some important conclusions were inferred: the recovery of off-target sites is directly proportional to the concentration of Cas9 ribonucleoprotein and it is also affected by delivery, cell type and time of exposure to the nuclease. Importantly, the predictive biochemical step can be applied to genomic DNA from a patient, making it feasible to get a personalised specificity profile that takes patient's nucleotide and structural intrinsic variants into account. In addition to the screening methods to predict and assess off-target sites, new high fidelity nucleases like SpCas9-HF1 or Cpf1, have successfully been developed . These new nucleases were developed after deeply screening into the Cas9 structure (SpCas9-HF1) trying to decrease non-specific interactions with its target DNA site, or screening other classes of type 2 CRISPR systems (Cpf1 = CRISPR from Prevotella and Francisella 1) for different cleavage patterns from the Cas9 nucleases.
The biggest hurdles of the CRISPR/Cas system: improving the efficiency and the specificity (avoiding off-target effects).
Focusing on in vivo applications of this technology, the specificity of the delivery system used, fitness of the edited cells and the possible immunogenicity of the administered components will affect the outcome. Adeno-associated viruses (AAV) are currently some of the most used delivery vectors for gene therapy due to their non-pathogenic nature, extrachromosomal persistence, broad infection pattern, low innate immunity and translatability from animal to human applications. The limited cargo size of AAVs (around 5Kb) could be overcome by using smaller Cas9 orthologues like SaCas9 from Staphylococcus aureus. Their wide tropism could be specified through tissue specific promoters to better control the expression of Cas9 only in certain type of cells, tissues or organs. The usefulness of in vivo gene editing, based on efficiency and specificity will depend also on the cell type targeted as only somatic stem cells, once edited, will transmit this modification to their progeny, thus ensuring a lasting benefit. Moreover, without a selective growth advantage of edited cells, which would allow targeted cells to out-compete non-modified ones, large number of cells need to be targeted in order to obtain a therapeutic benefit. Another potential caveat would be the immune responses provoked by repeated administration of the CRISPR/Cas components in clinical protocols that could either diminish its potency or cause unwanted side effects.
So, there is a further scope for improvement including the tool itself, vectors used for its delivery and translatability of gene editing to the clinic. Given that CRISPR/Cas9 is already a very powerful tool, we would need to increase its specificity, lowering the possibilities of targeting any other unwanted site in the genome. In parallel, new methodologies to more sensitively measure off-target mutations in cells below the error rate of current high throughput sequencing techniques would ideally have to be created. More efficient and specific delivery vectors for targeting our cell/tissue/organ of interest together with the development of new animal models that would match better the situation in human trials in terms of safety and efficacy are also areas of special interest for researchers.
Nevertheless, witnessing the speed of this field since the appearance of the CRISPR/Cas tools, we can assume that some of these improvements will be achieved in the not-so-distant future.
1 Komor et al. Enhanced proofreading governs CRISPR–Cas9 targeting accuracy. Cell. 169(3):559.
2 Maeder and Gersbach. Genome-editing Technologies for Gene and Cell Therapy. Mol Ther. 2016: 24(3):430-46
3 Barrangou R, Doudna JA. Applications of CRISPR technologies in research and beyond. Nat Biotechnol. 2016 ; 34: 933-941.
4 Chen et al. Enhanced proofreading governs CRISPR–Cas9 targeting accuracy. Nature, 2017.
5 Cameron et al. Mapping the genomic landscape of CRISPR-Cas9 cleavage. Nat Methods 2017. 14: 601-606
6 Tsai et al. CIRCLE-seq: a highly sensitive in vitro screen for genome-wide CRISPR-Cas9 nuclease off-targets. Nat Methods, 2017. 14: 607-614