|dc.description.abstract||Variation of DNA sequence in the human genome has relatively predictable effects in protein-coding sequences, but is not well understood in the remaining 98% of the genome. A current focus of research is on the sites that regulate transcription, where DNA sequence variation can have effects to increase or decrease the expression of nearby genes. These functional variants are believed to exert their effects by altering the binding of transcription factors to the genome. While this is a widely-held assumption, and supported by many observations, the model still lacks formal proof. In order to test this model, I performed a comprehensive analysis of a 17-member family that has been intensively studied as part of the 1000 Genomes project as well as developed a method to efficiently edit lymphoblastoid cell lines (LCLs) using CRISPR/Cas9.
First in Chapter 1, 1 review the genetic polymorphisms found within human populations and discuss the difficulty of ascribing disease risk to variants. l then introduce the various components of transcription regulation and how they are influenced by genotype. I further explore the use of lymphoblastoid cell lines as a model system and how this confers the manifestation of their molecular phenotypes. I then detail the application of CRISPR in both genome editing and epigenetic editing.
In Chapter 2, 1 describe our rigorous approach to extensively characterize the molecule phenotypes of the LCLs derived from the 17-member family. The data revealed novel miRNAs, expression and chromatin accessibility quantitative trait loci (eQTLs, caQTLs) and allelic patterns of molecular organization of transcriptional regulatory processes. Through the combination of data generated in previous studies and by us, we prioritized an intronic enhancer, demonstrating allele specific occupancy of NF&kgr;B, at the TBC1 D4 locus for further studies. To assess the impact of the variant on regulatory phenotypes, I performed genome editing and validated molecular changes. Using a CRISPR/Cas9 system tethered to NF&kgr;B subunit p65, I retargeted NF&kgr;B to the locus to confirm the hypothesis that it is disruption of TF binding that mediates these altered molecular phenotypes.
In Chapter 3, I discuss the development of a strategy to perform high efficiency genome editing in LCLs, a difficult cell-type to transfect and clone. I remark on its potential in future applications.
Finally, in my Conclusions I discuss the limitations of and gained perspective from the studies. The technical difficulties of utilizing lymphoblastoid cell lines in CRISPR assays hindered further investigation of the mechanisms mediating genetic variant effects as well as increased sample sizes. The study serves as a foundation upon which questions regarding distal cis-regulatory sites could be explored. While we establish that the retargeting of a TF to a genetically-"silenced" cis-regulatory site can partially reconstitute the expression of the associated gene, the downstream mechanisms of the observed reactivation require follow-up investigation, particularly the roles of chromatin modulators. I also review CRISPR/Cas9 principles corroborated through the development of our high efficiency editing method in LCLs. The well-documented and thorough molecular characterization of the 17-member kindred studied in this thesis provides a wealth of data to act as a springboard for future studies.||en_US