New CRISPR method allows stable knock-in of large genes.
Major progress on the CRISPR front. With the help of CRISPR/Cas, IPB scientists and French partners have succeeded in integrating larger DNA segments of several kilobases into the genome of higher plants in a stable, scar-free and very efficient manner. Until now, CRISPR has only been effective in creating deletions and point mutations, which usually result in a loss of function, i.e. the knock-out of the target gene. Efficient knock-in of (foreign) genes into the plant or mammalian genome usually fails due to the organism-internal repair of double-strand breaks, which is mainly based on imprecise and error-prone non-homologous end-joining (NHEJ) of the cleaved DNA strands. On the other hand, homology-directed repair (HDR) of DNA double-strand breaks allows precise and scar-free integration or replacement of large DNA fragments. However, since HDR, in contrast to NHEJ, requires 3'-overhanging ends and Cas9 predominantly produces blunt ends, Cas-mediated double-strand breaks tend to be repaired by the non-homologous repair machinery. Error-free insertion of larger gene segments has therefore only been achieved in a few transformation events to date.
Increasing the HDR efficiency of CRISPR knock-in approaches can resolve this dilemma and has been pursued for some time albeit with moderate success. With their study, the Halle scientists may now have achieved a breakthrough in this matter. According to their hypothesis, the Cas-generated blunt double-strand breaks must be turned into overhanging 3'-ends to increase the number of homology-directed repair events. This can be done by fusing the CRISPR endonucleases with 5' exonucleases, which then act directly at the restriction site to convert blunt ends into sticky 3’-ends. For this approach, the scientists combined Cas9 and Cas12a endonucleases with various 5' exonucleases of viral, plant, bacterial, and human origin. These exo-endonuclease fusion constructs - and, for comparison, the conventional CRISPR endonucleases - were analyzed in a transient reporter assay based on the tobacco mosaic virus in Nicotiana benthamiana.
Result: Using this approach, the scientists could identify two exonucleases from the herpesvirus family and the T7 bacteriophages, that (in fusion with Cas9) increased the HDR frequency up to 38-fold and thus enabled the scar-free insertion of a large 3,8 kb gene segment. In Arabidopsis, the fusion of Cas9 with an exonuclease of herpes viruses also led to a 10-fold increase in the frequency of stably transformed and therefore heritable knock-ins in the first generation of transformants. In wheat plants, the scientists were able to produce permanent knock-ins, also in the first generation, in over one percent of the transformants, which are stably passed on to future generations.
How does the reporter assay work? The reporter assay developed by the Halle-based team utilizes a transgenic N. benthamiana line containing a modified tobacco mosaic virus (TMV) vector in which the gene for the RNA-dependent RNA polymerase (RdRP) is compromised by a deletion of 3.8 kb. Without intact RdPR, both the replication of the viral RNA and the formation of the encoded downstream genes for a viral envelope protein and a movement protein are prevented. The Halle scientists replaced the envelope protein with a green fluorescent protein (GFP), making the TMV no longer infectious, but allowing viral replication to be detected using GFP fluorescence. The deleted 3.8 kb sequences in the RdPR gene were replaced by a recognition sequence for the Cas endonuclease.
The TMV reporter in the transgenic line is constitutively expressed from the Arabidopsis actin2 promoter, but replication and also GFP expression cannot take place unless functional RNA polymerase is present. Accordingly, GFP expression only occurs when the defective RdRP coding sequence is precisely repaired by HDR using the missing DNA sequence. The Halle scientists generated the necessary 3,8 kb repair segment and fitted it with additional homology arms that are identical to the remaining viral RdRP sequence and required for homology-directed repair. After transformation of the repair construct into the cells and its integration at the site of the planned insertion, the reconstitution of the viral RNA polymerase was measurable directly by detecting GFP expression. The also expressed viral movement protein promoted the local spread of the vector, so that only a few days after inoculation the individual repair events could be easily counted as macroscopically visible GFP spots. The HDR events induced by each specific exonuclease could thus be quantified and compared. The precise repair of the TMV RdRP gene was also confirmed by PCR and sequencing of genomic DNA from infiltrated leaf discs.
Obviously, the performance of the exonucleases used plays a role in the generation of free 3'-ends, as in vivo these enzymes compete with the endogenous repair factors of non-homologous repair for free DNA ends. The processing time and activity of the exonucleases can be optimized by various factors such as the spatial position relative to the fused endonuclease or the exonuclease's affinity for DNA double-strand breaks. Accordingly, monomeric 5'-exonucleases with a high affinity for blunt ends are best suited - in duet with the endonuclease - to improve the HDR rate when larger DNA segments are to be knocked in.
Conclusion: These new Cas-Exo fusion proteins are promising tools for gene targeting in higher plants and possibly other organisms. The Cas-Exo fusion can save considerable breeding time, especially in the generation of stable, i.e. heritable, traits. For example, the scientists achieved error-free repair rates of 1.1% in the T0 generation of the stable Cas9-Exo-mediated transformation of wheat. A screening of 50-100 transformants should therefore be sufficient to isolate individual plants with the desired properties. In future, this method could be used, to insert advantageous alleles from related wild species directly at the homologous locus into elite varieties or breeding lines. For scientific studies, this approach offers great opportunities to simply replace certain plant genes with modified copies. Until now, this could only be achieved by first knocking out the desired gene and then complementing it with the modified gene. In the case of essential genes, this blunt procedure would be lethal for the plant and is therefore not applicable to many genes. With CRISPR/Cas-Exo, however, genes of interest can be elegantly modified in one step to study their function.
