CRISPR/Cas9-mediated epigenetic repression of an enhancer linked to visceral obesity

Projects like the Human Genome Project, ENCODE, and RoadMap Epigenomics have enabled genome-wide association studies (GWAS) to map associations between genetic variants and human disease. However, most variants detected in GWAS are found in non-coding regions, which limits mechanistic and biological understanding of the associations. Many of these variants are thought to be situated in enhancers, affecting gene expression rather than gene structure. However, both enhancers themselves, as well as their target genes are incompletely mapped in most cells and tissues. Therefore, there is a great need for developing methods that can both detect enhancers in disease-associated loci and map their target genes. Active enhancers are characterized by the specific histone marks H3K4me1 and H3K27ac. The former mark can be removed by the lysine-specific demethylase 1 (LSD1) enzyme, which leads to repression of the enhancer. By fusing LSD1 to a catalytically dead Cas9 (dCas9) protein and introduce it to a relevant cell type, targeted epigenetic repression of an enhancer can be achieved. Moreover, by providing a transcriptional repressor alongside LSD1, an even more potent repression of the enhancer’s target gene(s) can be achieved. Visceral obesity is associated with increased risk of metabolic disease and premature death, but the genes contributing to disease risk are largely unknown. A recent GWAS identified the 11q23.3 locus to be associated with visceral obesity, but the associated variants are located in non-coding regions. Previous work by our lab has identified a predicted causal variant, rs1799993, and found it to be situated in an enhancer that is active in adipose-derived mesenchymal stem cells (AdMSC). The overall aim of this thesis is therefore to establish a method of epigenetic repression in AdMSC to identify the target genes of the 11q23.3 enhancer associated with visceral obesity. To this end, a dox-inducible lentiviral LSD1-dCas9 system coupled with constitutive lentiviral sgRNA(MS2)-MCP-KRAB expression was used in this thesis. Three different lentiviral constructs harboring Tet-on-3G-BSD, LSD1-dCas9-mCherry and sgRNA(MS2)-MCP-KRAB-zsGreen, respectively were generated and successively introduced into the AdMSC cell line ASC52telo after validation of the plasmids and viruses. Cells transduced with viruses encoding Tet-on-3G-BSD were selected using blasticidin, and cells transduced with LSD1-dCas9 and the sgRNA system were selected using FACS. Integration of lentiviruses into the genome of transduced cells and expression of the encoded constructs was assessed by qPCR of gDNA and cDNA, respectively. Protein levels of transduced constructs were analyzed by WB. The lentiviral plasmids were sequenced and found to contain the correct insert, except for a previously used Tet-On-3G construct that was found to be incorrect. Only correct plasmids were used for further experiments in this thesis. The plasmids were further functionally validated by transient overexpression in HT1080 cells with and without doxycycline, and found to express Tet-On-3G and LSD1-dCas9. A clear Tet-On-3G dependent dox-inducible effect was demonstrated for dCas9. Lentiviruses harboring the respective constructs were next successfully generated and found to contain viral concentrations of 4-9 x 1010 physical particles pr ml. A functional titer test revealed that about 1/1000 physical particles were infectious. ASC52telo cells were transduced with virus harboring Tet-On-3G-BSD using different multiplicity of infections (MOIs) and after BSD selection, the best MOI was found to be 15. The transduced cells were not able to grow as single colonies, thus a heterogenous population of Tet-On-3G expressing cells was expanded and successively transduced with lentivirus harboring LSD1-dCas9-mCherry. Only 2-8% of the cells were mCherry positive, but these populations were sorted, expanded and used in the subsequent transduction with virus containing sgRNA and KRAB. The latter transduction was more successful with 16-18% efficiency as determined by FACS. Triple-transduced cells, as well as single- and double-transduced controls, were sorted and successfully expanded. Intriguingly, mCherry expression was detected independently of Tet-On-3G. While expression of the Tet-On-3G and sgRNA-KRAB constructs were validated by qPCR, no clear expression of LSD1-dCAs9 could be demonstrated by qPCR or WB in the transduced cells, despite the positive mCherry signal in these cells. In conclusion, the establishment of an inducible CRISPR-Cas9 enhancer repression system in AdMSCs was partially established, but it remains to be determined whether the essential component LSD1-dCas9 is correctly expressed or not. Thus, repression of the 11q23.3 enhancer could not be performed, and consequently the target genes of the enhancer were not identified.

Masteroppgave i molekyl�rbiologi

MOL399

MAMN-MOL