Four years after the CRISPR/Cas9 technology took the field of mouse genetics by storm, it is time to draw some initial conclusions about how the technology has performed and discuss some future developments already in sight.
CRISPR/Cas9 is a gene-editing technology based on the ability of introducing a double-strand break in selected regions of the genome and exploiting the DNA repair systems to modify the targeted sequence. By introducing the genetic modification directly in embryos and bypassing the embryonic stem (ES) cell manipulation steps, CRISPR/Cas9 represents a faster and cheaper way to provide researchers with the in-vivo models they need. Compared to other pre-existing gene editing tools such as Zinc Fingers and TALENs, CRISPR/Cas9 owes its success to its simplicity and robustness, as thousands of scientists around the world can testify. Simplicity and robustness are, however, not enough to make a successful genome-editing tool on its own and, as every technology used to modify the mammalian genome, CRISPR/Cas9 needs to be specific, efficient, reproducible and flexible.
In the initial days of CRISPR/Cas9, a major concern was that the system might not have been specific enough to be of much use to geneticists, due to its tendency of modifying unrelated sequences in the genome. Initial reports claimed that CRISPR/Cas9 could induce double-strand breaks not only in the specific target sequence, but also in many different genomic locations. These lesions were referred to as “off-target events.”
Because of these observations, mouse geneticists were concerned that the modification of the off-target sites would result in secondary phenotypes in their models, and that the effort to segregate the unwanted mutations once the model was ready would nullify the advantages of using the technology in the first place.
It turned out that the worries of off-target modifications had been exaggerated and off-target events were exceedingly rare when the technology was used to modify the mouse embryo. Although the scientific community is still collecting data on CRISPR/Cas9 and it is premature to make a definitive conclusion about its specificity, it is possible to state that off-target events do not represent a major concern when using this technology to modify the mouse genome.
The use of CRISPR/Cas9 to modify the mouse genome directly in the fertilized oocyte is today the standard approach for the use of this technology. Opposite to genomic manipulation of ES cells, gene editing in embryos does not allow any kind of selection of the desired event, implying that it must occur with very high frequency. In other words, if the desired modification event occurs in 10 percent of cases, one will need to modify 10 oocytes to obtain one single founder. On the other hand, if the frequency of the event is 1 percent, one must modify 100 embryos to obtain one correct founder.
A major effort has therefore been devoted to increase the efficiency of the on-target modification, while paying close attention at avoiding a rise in off-target events. The use of Cas9/RNA ribonucleoprotein complexes in the injection procedure, for example, has been very effective in providing a high enough nuclease activity to efficiently trigger the DNA repair mechanism on the on-target site and favor the modification of the targeted sequence at high frequency.
A different issue that affects the efficiency of CRISPR/Cas9 is mosaicism of founder animals. Mosaicism is a phenomenon occurring when CRISPR/Cas9 introduces a DNA double-strand break on the target sequence in different chromosomes. Since the repair will happen independently on the sister chromosomes, different alleles might be introduced in parallel. To add complexity to the system, the CRISPR/Cas9 system is usually active throughout the first embryonic cell divisions, inducing independent mutagenesis events on the same locus in different chromosomes. The end result might be founder animals carrying as many as five different allelic variants of the same gene that need to be segregated by breeding to the next generation. In most cases, researchers are interested in only one specific modified allele and mosaicism reduces the number of useful offspring and, therefore, the efficiency of the entire process. Many different attempts have been done to reduce the level of mosaicism with mixed results and, as of today, the best way to deal with mosaicism is performing an in-depth and exhaustive characterization of founder animals.
Flexibility is the ability of a specific technology to introduce different kinds of modifications in the genome. Whether gene targeting in ES cells has proven an extremely flexible technology, CRISPR/Cas9 still suffers from many limitations. As of today, CRISPR/Cas9 cannot efficiently introduce in the embryonic genome complex modifications requiring homologous recombination over long stretches of DNA (e.g. conditional knockout alleles or gene replacements).
Although it has been shown as a proof of principle that these alleles can be introduced in the mouse genome using the CRISPR/Cas9 technology, the extent of the effort and the unpredictability of the timelines negate the reasons for selecting this approach in the first place. These limitations restrict the type of alleles that can be generated by using this technology to constitutive knock outs, knock in of point mutations and insertion of sequences coding for peptides recognized by specific antibodies (tags). It is currently the focus of a score of laboratories worldwide to improve the flexibility of the system and allow the introduction of complex alleles without losing the advantages of the technology.
Reproducibility is a key feature in every technology that is applied on a large scale by the scientific community. The inability to reproduce described results is often the main obstacle toward the transfer of a new technology from a single research laboratory to a wider audience. Lack of reproducibility is often caused by the intrinsic fragility of an experimental system (e.g. extreme sensitivity to pH or temperature conditions) making it impossible to apply the new process to a pipeline requiring a consistent performance.
Arguably, the major strength of CRISPR/Cas9 is its reproducibility, resulting in an impressive record of reports of its successful use by many different laboratories just a few months after the original publication appeared. Robustness of CRISPR/Cas9 is now widely recognized as one of its principal characteristics and, when properly designed, it is quite rare that a genome modification project based on this technology is not successful.
Modifying the mouse genome: Where are we now?
The advent of the CRISPR/Cas9 technology has changed mouse genetics. By decreasing prices and reducing timelines for the generation of new models, CRISPR/Cas9 has allowed many researchers to exploit the experimental power of in-vivo systems without investing the extensive resources previously required for obtaining the necessary tools.
The possibility to modify directly the embryo genome without the requirement of using ES cells has opened new model generation approaches such as model refitting, namely the introduction of additional genetic modifications into well-established multi-allelic existing models or the modification of transgenic alleles in already characterized mouse lines.
Moreover, the ability of introducing in an inexpensive and rapid way specific point mutations in the mouse genome has already become a widely used tool to validate human genetic data obtained by clinical studies, providing an invaluable contribution to the field of human genetics. A new and exciting field is the in-vivo screening of target genes by introduction of CRISPR/Cas9 elements via viral vectors, an application previously not conceivable with the standard tools of mouse genetics.
Modifying the mouse genome: What’s next?
CRISPR/Cas9 success in manipulating the mouse genome has raised very high expectations around this technology. In the past four years, however, it has become increasingly clear that CRISPR/Cas9 is very well suited for the introduction of only relatively simple modifications. A way to overcome these limitations is to take a step back and use the technology to modify ES cells rather than embryos.
The possibility of selecting thousands of independent ES cell clones allows the identification of rare events that might not be possible to identify when modifying directly the embryo genome. Although some of the intrinsic advantages of CRISPR/Cas9 are lost by bringing back ES cell culture in the equation, the possibility of introducing complex alleles such as genomic replacements can lead to the generation of specialized mouse models relevant for drug development and testing, making this approach worth keeping into consideration. In parallel, many laboratories are actively optimizing the use of CRISPR/Cas9 in embryos with the goal of increasing the flexibility of this technology without compromising its many advantages.
At the end, we must recognize that CRISPR/Cas9 is still in its infancy and that there is enormous room for further development. Due to the furious pace at which researchers around the globe are pursuing its optimization, it is not improbable that in a few years all genomic modifications will be introduced using CRISPR/Cas9 and that what is now considered not feasible might just become routine work in the near future. The first four years of CRISPR/Cas9 have been exciting, but the best is yet to come.
Dr. Adriano Flora joined Taconic Biosciences in 2011 and is currently associate director in the Product Management group. Flora obtained a Ph.D. degree in pharmacology and toxicology in 2001 at the University of Milano, Italy, before moving to Baylor College of Medicine in Houston as a postdoctoral fellow first and later as assistant professor.