1. 2014-10-30 Seong-Eui Hong

2. Background 1987 2002 2005 2007 2012 2013

3. Identification of palindromic and repetitive sequence

4. Study on the repetitive DNA sequence

5. Suggestion of the role in cell immunity

6. Experimental confirmation of adaptive immunity

7. Mechanisms of CRISPR in adaptive immunity Figure from http://www.biochem.or.kr/ksbmb_webzine tracrRNA: trans-acting crRNA

8. Genome editing using CRISPR

9. Introduction Aim: Generation of cancer model using CRISPR in mice Target gene: 1)Pten 2)p53 3)beta-catenin

10. Generation of Pten mutations in adult animals using CRISPR Pten: negative regulator of PI3K/Akt Step1: cloning of pX330 vector co-expressing an sgRNA targeting Pten (sgPten) and Cas9 Step2: in vitro mutation in mouse 3T3 cell using sgPten Step3: in vivo delivery of luciferase plasmid DNA using hydrodynamic tail-vein injection, which can deliver DNA to 20% of hepatocytes Step4: in vivo delivery sgPten and sgGFP in FVB mice during 2 weeks

11. Result: delivery of sgPten 3.36% hepatocytes with (-) Pten staining 0.4% hepatocytes with intermediate Pten staining (=heterozygous mutation) Elevated staining of pAkt in sgPten mice Oil red O staining indicates lipid accumulation in sgPten mice, which is a known phenotype in Pten mutation in liver Conclusion: in vivo CRISPR-mediated genome editing was able to generate Pten (-) cells in the liver, mimicking liver-specific conditional deletion of Pten in mice

12. Intermediate loss of Pten

13. Question: Loss of Pten is due to CRISPR-mediated mutation? Method: Deep sequencing 2.66% of the reads had indels in Pten locus 0.56% indels in sgGFP Most of variants were predicted to cause frameshift mutations 1~2bp indels cause the disruption of the Pten reading frame These indels were at the predicted sgPten induced Cas9 cutting site Whereas the indels in sgGFP distributes randomly Pten loss (scored by IHC) strongly correlated with the Pten indels Conclusion: These data indicate that for most cells, expression of the sgPten vector results in mutation of Pten

14. Question: long-term phenotype following sgPten treatment? Method: Analysis the livers from 3 sgPten-treated mice at 4 months Lipid accumulation Pten loss Elevated pAkt No induction of p53 despite the activation of pAkt No tumor up to 4 months

15. Question: off-target? Recent studies identified that Cas9 can tolerate mismatches between sgRNA and genomic DNA depending on the sgRNA sequence and the position of the mismatches Question: potential off-target effects of sgPten in the liver? Method: 1)Prediction of sgPten off-target sites using a published prediction tool 2)Amplification of the Pten locus and the top four potential off-target sites using Surveyor assay Result: There were no Surveyor nuclease cutting, indicating that the frequency of off-target editing is below the limit of detection of this assay

16. Test of a nickase version of Cas9 Cas9 generates double strand breaks Cas9(D10A) makes single-strand DNA (ssDNA) breaks and was reported to have further reduced levels of off-target effects

17. Test of a nickase version of Cas9 By deep sequencing, 2.76% indels at the Pten locus compared to 0.26% in sgGFP-treated mice were observed Pten (-) cells were observed

18. Generation of p53 mutations in adult animals using CRISPR 6.061% of indels in p53 was observed No tumor up to 3 months

19. Co-injected sgPten and sgp53 4.06% for Pten and 6.46% for p53 At 3 months post-injection, all 5 mice developed liver tumors with bile duct differentiation features Tumors were positive for cytokeratin 19, a marker of bilinear lineage cells bi-allelic mutations of both genes

20. Question: Can CRISPR be used to directly introduce gain of-function mutations? Target: Ctnnb1 gene, which encodes  -catenin, a transcription factor in the Wnt signalling pathway that is frequently mutated in liver cancer Phosphorylation at four ser/thr results in the degradation of Ctnnb1 Method: co-injection of two pX330 plasmids carrying i) sgRNA targeting Ctnnb1 and ii) 200 nt ssDNA oligonucleotide containing four point mutation in five mice injected with sg  catenin and ssDNA, 0.5% of hepatocytes exhibited nuclear  -catenin Accumulation of  -catenin was associated with increased levels of glutamine synthetase, a  -catenin target gene

21. Targeting of  -catenin Reduced level of p-  -catenin Nuclear accumulation of  -catenin As a result of deep sequencing, small but detectable percentage of sequencing reads contained the four ‘G’ point mutations present in the ssDNA Conclusion: These data demonstrate that CRISPR system can be used to directly induce gain-of-function mutation or other substitutions via homologous recombination in vivo

22. Conclusion The authors illustrate the potential to directly disrupt tumor suppressor genes and generate point mutations in oncogenes in adult mouse liver using the CRISPR/Cas system This approach generated compound Pten and p53 indels at low frequency but was sufficient to induce multifocal tumors More efficient delivery techniques, such as adenovirus or adeno-associated virus, more potent sgRNAs, and longer homologous recombination templates might also improve the overall efficiency of this method and expand the range of tissue that could be targeted Consistent with recent studies showing that long-term Cas9/sgRNA expression is not toxic in cells