Person: Mi, Xiaoli
Loading...
Email Address
AA Acceptance Date
Birth Date
Research Projects
Organizational Units
Job Title
Last Name
Mi
First Name
Xiaoli
Name
Mi, Xiaoli
1 results
Search Results
Now showing 1 - 1 of 1
Publication Lessons From Using Genome Editing to Create a Faithful Model of a Novel Congenital Anemia(2018-05-15) Mi, XiaoliThe ease of genome sequencing over the past decade has greatly facilitated the identification of genetic mutations causing a large number of human diseases. However, in the vast majority of cases, how these mutations lead to disease remains an enigma. Targeted genome engineering via the RNA-guided CRISPR/Cas9 nuclease system offers the possibility of faithfully mimicking molecular mutations in their endogenous context to study underlying mechanisms. In this thesis, I explored the question of whether genome editing is an effective approach to model a novel genetic disease. Recently, our laboratory has identified two patients with a distinct form of congenital dyserythropoietic anemia, characterized by erythroid and megakaryocytic dysplasia and mild elevations in fetal hemoglobin levels. Curiously, the coding sequence of DNA from the patients was unremarkable but further analysis revealed a common hemizygous point mutation in the last intron of GATA1 (chrX: 48,652,176 C>T in hg19). This gene encodes the hematopoietic master regulatory transcription factor GATA1 essential for the differentiation of erythrocytes and megakaryocytes. Genetic mapping and segregation within the families provided strong evidence to support a causal relationship between the mutation and disease. To model the disease in a tractable system and further understand the underlying basis by which the mutation results in disordered blood production, I performed genome editing with CRISPR/Cas9 in a human erythroid cell line to recreate this intronic GATA1 mutation. I designed and cloned nine distinct single guide RNAs (sgRNAs) and identified one that cleaved the target locus in an efficient manner. Using this sgRNA, I showed that an increased mass ratio of exogenous repair template to sgRNA to Cas9 dramatically improved GATA1 editing efficiency and diversity. The rate of genomic perturbation increased from 20% to 80% and incorporation of the intronic mutation via homology directed repair occurred at 3.8%. Using three computational methods, I characterized all allele-specific editing events in 133 isogenic cell lines and mapped them by genomic position relative to the Cas9 cleavage site. I found that most GATA1 changes occurred within 15 bp of the Cas9 cleavage site and the editing frequency is higher upstream than downstream at each position equidistant from the predicted double-strand breaks. Consistent with this observation, most GATA1 edits occurred within the intron and those that altered the downstream canonical splice acceptor site and exon were exclusively heterozygous. These findings suggest that the human erythroid cell line disfavors alteration of sequences essential for proper function of GATA1, via splicing regulation or protein coding. Notably, the desired homozygous C>T mutations co-occurred with additional modifications such as deletion of an alternative splice acceptor site and the protospacer adjacent motif (PAM) required for Cas9 binding. It is possible that concurrent deletion of the alternative splice acceptor site rescues the splicing defect from the mutation and genome editing recurred until PAM was rendered unrecognizable. Following these experiments, my colleagues demonstrated that the C>T mutation results in activation of the alternative splice acceptor site and a partial intron retention event using transient expression of a minigene. The observed splicing alteration was confirmed in patient samples and further experiments showed that the resultant GATA1 protein is inactive. While many recent proof-of-principle studies have described successful application of genome editing to generate models of common diseases with well-established genetic alterations, no study has yet modeled a novel disease with a novel mutation and provided insight into disease pathophysiology. I conclude with important limitations of using CRISPR/Cas9 for disease modeling and discuss possible solutions and efficient alternatives. These lessons have broad implications for future applications of genome editing and classical approaches to study disease in an effort to generate functionally and clinically relevant knowledge.