One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering (2013)

Haoyi Wang et al. One-Step Generation of Mice Carrying Mutations in Multiple Genes by CRISPR/Cas-Mediated Genome Engineering. Cell. May 9, 2013; 153(4): 910–918.

Genetically modified mice have proved over the years to be an indispensable model for many human disease, as well as good model for several types of human cancer. Therefore production of these mice represents an important process for biological sciences and medicine. There are two major technical approaches by which mice with specific mutations can be obtained. First is trough manipulation of mouse embryonic stem (ES) cells, which are then injected into wild-type blastocysts. Primarily mutations were introduced trough homologous recombination, which is by nature an extremely rare event in mammalian cells. After chimeric mice are obtained it is very costly and time consuming procedure, to generate pure single-knockout mice and even more so to generate double-knockout mice. Second technical approach is by direct embryo manipulation. This was possible to perform, after more efficient approaches for introducing site specific mutations based on site-specific nucleases were developed. These are meganucleases, zinc finger nucleases (ZNFs) and transcription activator-like effector nucleases (TALENs). However all these methods require complex designs and modifications on protein level and are therefore not exactly appropriate for multiplexed gene targeting. Recently a novel approach for gene editing based on type II bacterial CRISPR/Cas9 adaptive immune system has been reported. Cas9 nuclease uses guide RNA molecule to recognize specific cut site instead of protein DNA interactions and is therefore easier to manipulate and more suitable for multiple gene targeting.

Designing custom-made mouse models for specific disease

Genetically modified mice represent a good model organisms for studying relationships between gene mutations and disease phenotypes. For mice strain to be designated as a model of the human disease, the mutated gene needs to be orthologous to that in humans and as well cause the same disease. There are two main procedures by which mutations can be introduced. First is insertional mutagenesis [1]. This is mutagenesis of DNA by insertion of one or more bases which can occur naturally trough insertion of viruses or transposons into the organism’s preexisting DNA. Both mechanisms can be artificially used for research purposes but the downside is that they both lack site specific insertion. Because of random or semirandom fashion of insertion into the hoste genome this is not a systematic approach to identifying genes whose perturbation may have facilitated the promotion of tumor growth and thus does not allow creating of model organism for specific disease [2]. Second and approach are gene-targeting methods, where mutations are introduced trough homologous recombination (HR) in mouse embryonic stem (ES) cells [1]. This was a beaktrough approach in genome manipulation as it allowed targeted manipulation of genetic material. Key to this approach are exogenous repair templates that contain sequence homology to the donor site. Trough manipulation of these repair templates knockin and knockout animals can be made via manipulation of germline competent stem cells which are then further microinjected into wild-type (WT) embryos at the stage of blastocysts. These are then transferred to pseudopregnant recipient females which give birth to chimeric pups. Tissues of chimeric animals have therefore developed both from ES cells and recipient blastocysts. Based on the color of the coat, mice with 50 % or more ES cell contribution are selected and bred to make pure knockout mice [3]. However although manipulation trough HR allows for precise targeting it is limited by the nature of which the desired recombination events occur (1 in 106-109 cells) [4][5].Therefore producing single and multiple gene knockout animals is a costly and time consuming procedure. Moreover there are no established ES cell lines in many species [1]. To overcome the above mentioned problems, alternative methods have been developed. They base on direct injection of DNA or mRNA of site-specific nucleases into the one-cell embryo. These nucleases generate double-strand breaks at a specified locus on the DNA which tend to be repaired by error-prone Non-Homologous End Joining (NHEJ) mechanisms. Repairing often results in mutant animal carrying a deletion or insertion at the cut site [1]. After a series of later studies it became clear that site specific generation of double-stranded breaks could not only be used for disrupting genes via deletion or insertion at the cut site, but in fact this process as well greatly stimulate genome editing trough HR-mediated recombination events [4]. This means that if in accordance to site-specific nuclease a donor plasmid with regions that are homologous to the ends at the DSB is coinjected, it is more likely that animals with targeted integrations will be produced.

Methods for introducing DSBs at specific sites on DNA

Site specific DSBs which are required for effective editing of genome can be achieved via one of four major classes of DNA binding proteins. These are meganucleases, zinc finger nucleases (ZF), transcription activator-like effectors (TALEs), and most recently the RNA-guided DNA endonuclease Cas9 from the type II bacterial adaptive immune system CRISPR. The first three classes use protein-DNA interactions for recognizing specific DNA sequences. Meganucleases recognize recognition site of 12-40 bp in length and can be artificially modified to cut at specific sites [6]. ZF and TALE proteins recognize 1 or 3 nucleotides and don’t have nuclease activity of their own. Though they can be assembled in desired combinations and attached to the nuclease. Although the above mentioned ways can be effective, there are many challenges associated with engineering of modular DNA-binding proteins. The last and the newest class of DNA binding proteins is CRISPR nuclease Cas9, which guided to the specific site on the DNA via short guide RNA and thus does not require any complex protein modifications or laborious and time-consuming construct assembly as in the case of previously mentioned methods. DSB at a specific location can be achieved only by modifying short guide RNA which needs to be complementary to the region of interest [4].

CRISPR Cas9

CRISPR/Cas9 is a newly developed system for introducing precise mutations and targeted changes to the genome of a living cell. It is based on bacterial and archaeal adaptive immune defenses named clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated protein-9 nuclease (Cas9) [7]. Role of this immune system is to detect and destroy invading viruses and plasmids. CRISPR loci consist of several different Cas genes and an interspaced array. This array consists of series of repeat sequences (direct repeats) and is interspaced by variable sequences called spacers. Spacers are homologous to regions within sequences of foreign genetic elements called protospacers and have a role of guiding Cas enzymes to degrade target sequences [4]. Spacers and other CRISPR arrays are first transcribed as a single RNA, followed by adequate processing into shorter CRISPR RNAs (crRNAs). So far three types of CRISPR mechanisms have been identified (types I-III). Types I and III are more complex and consist of multiple proteins which form complexes with crRNA. Their loci codes for multiple Cas proteins which form complexes with crRNA in case of type I and with Cmr or Csm RAMP in case of type III [4]. On the contrary, the type II system comprises of only one protein component (Cas9) and two RNAs. First is a short RNA called crRNA and second is a so called tacrRNA. Both RNAs form a complex with Cas9, whereas crRNA is needed for complementary binding with foreign DNA and tracrRNA is needed for stabilizing the formed complex [8]. The above mentioned CRISPR/Cas system II has been further redesigned for laboratory gene-editing purposes. As it was taken from bacteria, firstly the codon-usage of Cas9 enzyme sequence needed to be optimized to achieve better expression in mammalian systems. Secondly crRNA and tracrRNA have been successfully combined into chimeric molecule called single guide RNA (sgRNA). Fusion of these two molecules simplifies the CRISPR/Cas9 system to only two components. A guide RNA that is subject to adjustment, depending on the DNA sequence we are targeting and Cas9 enzyme which is a common component independently of the targeted sequence. Therefore role of sgRNA is to bind to Cas9 and guide it to the target site, where the DNA is cleaved and the NHEJ repair mechanism is triggered [8].

Simultaneous Targeting of multiple genes

In this study authors decided to test weather is it possible to target several genes with the CRISPR/Cas9 method at the same time. As previously mentioned producing single gene knockout mice with homologous recombination is a very costly and time consuming procedure which can take from about 6 to 12 months [9]. In order to produce mice carrying mutations in several genes, additional intercrossing of single-mutant mice is required, which is very labor intensive. Therefore a method allowing for editing of multiple genes in a single step would be of great interest and value.

Simultaneous Targeting up to Five Genes in ES Cells

First goal of study was to design efficient CRISPR/Cas9 system, that could efficiently target several genes simultaneously in ES cells. Functionally redundant genes were taken as targets, as mutations in these genes will have smaller effect on the fitness of the organism. These genes are Tet1, Tet2, and Tet3 from ten-eleven translocation (Tet) family of genes. In order to measure cleavage efficiency at each target site, each gene has first been targeted separately and measured for cleavage efficiency by the Surveyor assay which is based on the ability of Surveyor nuclease to selectively cleave at the sites where deformed DNA duplexes are formed by hybridization of mutated and wild-type strands. ES cells were therefore transfected with plasmids expressing Cas9 and sgRNA. After transfection single cell colonies were obtained for genotypization. Results showed that all three sites were cleaved successfully. Tet1 was cleaved with the efficiency of 36 %, Tet2 with 48 % and Tet3 with 36 %. Next restriction fragment length polymorphism (RFLP) assay was performed to determine how many of the cells that were efficiently cleaved at target site have both alleles mutated. This could be done due to each target site containing restriction enzyme recognition site. Targeted sites were amplified by PCR and digested with the respective enzyme. Alleles that have been successfully targeted will therefore lose their restriction site and will appear at the expected size after separation on PAGE. The results of RFLP showed that between 65 % and 81 % ES cell clones were mutated at Tet genes and that approximately 77 % of them carried mutations in both alleles. High efficiency of single-gene modification led to further questioning whether all three genes could be targeted at a time. In order to test this, ES cells were cotransfected with the plasmids Cas9 and sgRNAs targeting Tet1, Tet2 and Tet3. As with single-gene modification RFLP assay was performed. Results indicated that approximately 20 % of clones had mutations in all six alleles of the three genes. To verify results the PCR products of Tet1, Tet2 and Tet3 were subcloned and sequenced. Finally 5-hydroxymethylcytosine (5hmC) levels of targeted clones were compared to WT ES cells in order to test whether mutations in all alleles would affect the function of Tet proteins. Depletion of 5hmC levels has previously been reported in Tet1/Tet2 double knockout ES cells produced by traditional gene targeting methods [10]. In accordance to these results, significant reduction of 5hmC levels was found in all clones carrying biallelic mutations in all three genes. To further test the potential of CRISPR/Cas9 system for multiple-gene targeting, two additional targets (Sry and Uty) were added to the preexisting Tet1, Tet2 and Tet3. After analysis 10 % of the screened clones were identified to carry mutations in all alleles of the five genes. This results demonstrate the efficiency of the method for multiple-gene targeting [11].

One-Step Generation of Single-Gene Mutant Mice by Zygote Injection

Although single and multiple gene manipulation in ES cells with CRISPR/Cas9 has proved as efficient, deriving pure knockout mice from chimeric animals, produced by injection of ES cells in embryos and further in pseudopregnant females, is still an extensive procedure. In order to simplify this procedure, it was tested weather mutant mice could be generated with this technique by direct one cell embryo manipulation. To do so, mRNA of Cas9 and sgRNA were injected directly, instead of cotransfection of DNA plasmids coding for Cas9 and sgRNA, as done previously. Cas9 mRNA was produced by in vitro transcription and was further capped and polyadenylated to protect against early degradation. Different concentrations of Cas9 mRNA were used to determine the optimal ratio between efficiency and toxicity. Surprisingly non of the injected concentrations showed toxicity as all embryos showed normal blastocyst development. At his stage gene editing efficiency was measured using RFLP assay, showing that editing was more efficient at higher concentrations of the injected RNA. Next the embryos were transplanted in to pseudopregnant mice to test weather pups carrying desired mutations could be obtained. Results showed that about 10 % of mice gave birth. To test weather mutation occurred in both of the targeted alleles, RFLP assay and sequencing analysis were performed on the postnatal mice. Results demonstrated, that between 50 and 90 % carried biallelic mutation [11].

One-Step Generation of Double-Gene Mutant Mice by Zygote injection

Previous results demonstrated that single-gene mutant mice could be efficiently generated by zygote injection and that multiple-gene mutant mice could be generated by editing ES cells with CRISPR/Cas9. Next level was to test weather multiple-mutant mice could be generated, by zygote injection. For this purpose mouse single-cell embryos were coinjected with Tet1 and Tet2 sgRNAs and Cas9 mRNA. 21 % of these embryos developed to birth and 79 % of these neonatal mice were identified carrying targeted mutation in all four alleles. Lower concentrations of injected RNA resulted in higher birth rate (25 %), but at the cost of fewer pups (50 %) carrying mutations in all four alleles of the genes Tet1 and Tet2. These results demonstrate, double mutant mice could not only be efficiently generated using CRISPR/Cas9, but also in much less time, with respect to one month instead of 6 to 12 months that were required normally [11].

Multiplexed Precise HDR-mediated Genome Editing in Vivo

To take one step generation of mutant mice by zygote injection to a further level, it was tested weather precise homology directed repair could be applied to embryos. As already mentioned, double-stranded breaks could not only be used for disrupting genes via deletion or insertion at the cut site, but as well for precise homology direct repair (HDR)-mediated genome editing as this process greatly stimulates HR-mediated recombination events [4]. To explore weather this can be achieved directly in embryos, single-stranded DNA oligos were used instead of plasmids and were coinjected with Cas9 mRNA and sgRNAs. To simply test weather homologous recombination event occurred, oligos were designed in a way that restriction sites at the targeted regions would be changed to another restriction site and success of homologous recombination event could be tested with restriction enzymes after PCR amplification. Restriction site in Tet1, would be changed from SacI to EcoRI and in Tet2, from EcoRV to EcoRI. To detect oligo-mediated HDR events, blastocysts were isolated from embryos, followed by DNA isolation, amplification and digestion by EcoRI. Results showed, that for Tet1, HDR events occurred in 6 of 9 embryos and 9 out of 15 for Tet2, with several embryos having both alleles modified. To test weather multiple genes could be edited in this way, oligos for Tet1 and Tet2 were coinjected into zygotes. Now only one embryo out of 14 was identified to have both genes modified, but in only one allele of each gene, while other two were disrupted by NHEJ-mediated deletion and insertion. To test weather pups carrying desired mutations could be obtained, embryos were transplanted in to pseudopregnant mice. A total of 10 pups were born from 48 embryos indicating 21 % live-birth rate. Out of 10, 6 mice were found to contain EcoRI sites at both Tet1 and Tet2 loci. These results demonstrate that HDR-mediated CRISPR/Cas9 system can be used for generating precise mutations in multiple genes in a single step [11].

Toxicity of CRISPR/Cas9 System

To test toxicity of CRISPR/Cas9, several different amounts of mRNA encoding for Cas9 were microinjected in to one-cell mouse embryos, varying from 20-200 ng/µl at constant concentration of sgRNA being 20 ng/µl. As already mentioned higher concentrations of Cas9 mRNA resulted in more efficient gene disruption, but did not show to be significantly more toxicity. All embryos, even with highest concentrations normally developed to blastocysts. Small effect of toxicity could be seen from the survival rate of embryos after implantation, as higher concentrations yielded smaller survival rate in exchange for better efficiency. To further study toxicity of CRISPR/Cas9 authors looked for off-target effects in vivo. Off targets were identified based on previous studies suggesting that protospacer-adjacent motif (PAM) sequence NGG and sequence of 8 to 12 bases from the 3’ end of the sgRNA are most important for specificity of DNA cleavage [12]. Three of such targets exist on the genome for Tet1 and four for Tet2. All off targets were amplified with PCR, followed Surveyor assay, which reported no mutations in these regions. These results suggest high specificity and low toxicity of CRISPR/Cas9 system [11].

Conclusion

As already mentioned, genetic manipulation of mice is an important approach for studying development as well as disease. Classic procedures for obtaining knockout animals, such as gene targeting by homologous recombination in ES cells, followed by transplantation and inbreeding of chimeric animals are very labor intensive and time consuming. To address the aforementioned problem, authors in this article have established three different approaches for generation of mice carrying not only one but multiple genetic alterations, based on the CRISPR/Cas9 system. They demonstrated, that CRISPR/Cas9-mediated genome editing in ES cells could simultaneously generate mutations in up to five genes with high efficiency. Secondly they showed that it is possible to directly modify mouse embryo by injection of Cas9 mRNA and sgRNA into fertilized egg. To take these findings to even a further level they demonstrated that multiplexed gene editing by precise homology directed repair could be efficiently applied to embryos by coinjecting single-stranded DNA oligos with Cas9 mRNA and sgRNAs into fertilized egg.

References

[1] H. Wang, H. Yang, C. S. Shivalila, M. M. Dawlaty, A. W. Cheng, F. Zhang, and R. Jaenisch, “One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering.,” Cell, vol. 153, no. 4, pp. 910–8, May 2013.

[2] M. Ranzani, S. Annunziato, D. J. Adams, and E. Montini, “Cancer gene discovery: exploiting insertional mutagenesis.,” Mol. Cancer Res., vol. 11, no. 10, pp. 1141–58, Oct. 2013.

[3] “KI Swanson Biotechnology Center: ES Cell and Transgenics: Services: Generation of Mice from ES Cells at MIT.” [Online]. Available: http://ki.mit.edu/sbc/escell/services/details. [Accessed: 28-Dec-2014].

[4] P. D. Hsu, E. S. Lander, and F. Zhang, “Development and Applications of CRISPR-Cas9 for Genome Engineering.,” Cell, vol. 157, no. 6, pp. 1262–1278, Jun. 2014.

[5] T. Gaj, C. a Gersbach, and C. F. Barbas, “ZFN, TALEN, and CRISPR/Cas-based methods for genome engineering.,” Trends Biotechnol., vol. 31, no. 7, pp. 397–405, Jul. 2013.

[6] L. M. Seligman, K. M. Chisholm, B. S. Chevalier, M. S. Chadsey, S. T. Edwards, J. H. Savage, and A. L. Veillet, “Mutations altering the cleavage specificity of a homing endonuclease.,” Nucleic Acids Res., vol. 30, no. 17, pp. 3870–9, Sep. 2002.

[7] “Addgene: Addgene’s CRISPR Guide.” [Online]. Available: https://www.addgene.org/CRISPR/guide/. [Accessed: 28-Dec-2014].

[8] D. W. Harms, R. M. Quadros, D. Seruggia, M. Ohtsuka, G. Takahashi, L. Montoliu, and C. B. Gurumurthy, “Mouse Genome Editing Using the CRISPR/Cas System.,” Curr. Protoc. Hum. Genet., vol. 83, pp. 15.7.1–15.7.27, Jan. 2014.

[9] M. R. Capecchi, “Gene targeting in mice: functional analysis of the mammalian genome for the twenty-first century.,” Nat. Rev. Genet., vol. 6, no. 6, pp. 507–12, Jun. 2005.

[10] M. M. Dawlaty, A. Breiling, T. Le, G. Raddatz, M. I. Barrasa, A. W. Cheng, Q. Gao, B. E. Powell, Z. Li, M. Xu, K. F. Faull, F. Lyko, and R. Jaenisch, “Combined deficiency of Tet1 and Tet2 causes epigenetic abnormalities but is compatible with postnatal development.,” Dev. Cell, vol. 24, no. 3, pp. 310–23, Feb. 2013.

[11] H. Wang, H. Yang, C. S. Shivalila, M. M. Dawlaty, A. W. Cheng, F. Zhang, and R. Jaenisch, “One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering.,” Cell, vol. 153, no. 4, pp. 910–8, May 2013.

[12] L. Cong, F. A. Ran, D. Cox, S. Lin, R. Barretto, N. Habib, P. D. Hsu, X. Wu, W. Jiang, L. a Marraffini, and F. Zhang, “Multiplex genome engineering using CRISPR/Cas systems.,” Science, vol. 339, no. 6121, pp. 819–23, Feb. 2013.