C1 - Base Editing and Prime Editing
681: Unconstrained Mitochondrial DNA Editing with Evolved TALE Proteins
Type: Poster Session
Poster Board Number: 681
Presentation Details
Session Title: Wednesday Posters: Base Editing and Prime Editing
The human mitochondrial genome is a 16,569 bp long, multicopy, circular double-stranded DNA (dsDNA) molecule that encodes 37 genes essential for cellular energy metabolism. Pathogenic mitochondrial DNA (mtDNA) variants are prevalent in ~1 in 8,000 people and are causal in incurable metabolic disorders. Current mtDNA editing strategies rely on all-protein systems, such as transcription activator-like effector (TALE)-based technologies. In general, these technologies are composed of a TALE, which acts as a programmable DNA-binding protein, fused to an effector domain. In particular, dsDNA deaminase A (DddA)-derived cytosine base editors (DdCBEs) consist of TALE-DddAtox fusions. Given its preference for dsDNA, DddAtox, the deaminase domain of DddA, is split into two inactive halves to avoid toxicity. Thus, a DdCBE monomer incorporates either the N- or C-terminus of split DddAtox downstream of a TALE. In the context of DdCBE pairs, binding of their respective DNA-binding domains to adjacent target sequences enables the reassembly of functional DddAtox, followed by targeted cytidine deamination. This process leads to programmed C•G-to-T•A conversions in mtDNA. Despite their robustness, the versatility of canonical DdCBEs is thought to be limited by the requirement of a thymine immediately upstream of their respective TALE target sequences. Following this conventionally accepted double 5’-T constraint (one per DdCBE monomer), about 12% of the human mitochondrial genome is theorized to remain inaccessible to canonical DdCBEs. Here, aiming to elucidate the relevance of the 5’-T constraint in the context of DdCBE-mediated mitochondrial base editing, and enable access to 100% of the editable motifs in the human mitochondrial genome, we generated DdCBEs containing TALE proteins engineered to bypass the 5’-T constraints, herein referred to as αDdCBEs. First, seeking to characterize the activity and specificity profiles of αDdCBEs relative to DdCBEs, we compared pairs of each type of editor in both 5’-T-compliant and 5’-T-noncompliant formats across multiple mtDNA loci accessible to both types of editors. Unexpectedly, we observed that DdCBEs are active even when they are 5’-T-noncompliant. Nonetheless, as hypothesized, 5’-T-noncompliant αDdCBEs tend to be significantly more active than 5’-T-noncompliant DdCBEs, reaching the same levels of activity as 5’-T-compliant DdCBEs. Subsequently, we compared αDdCBEs and DdCBEs at MT-TC and MT-TL1, two mitochondrial loci considered to be inaccessible to DdCBEs. Remarkably, DdCBEs were as active as αDdCBEs at MT-TC, reaching editing efficiencies of up to 50% in HEK293T cells 3 days post-transfection. Nevertheless, at MT-TL1, αDdCBEs were more active than DdCBEs. Finally, seeking to optimize a base editing strategy for the installation of the pathogenic m.3242G>A variant in MT-TL1, we screened an array of αDdCBEs at this locus. Notably, all editors in this approach contained DddA11, an engineered DddAtox variant known to be highly active. Interestingly, we observed that sliding either one or both arms of an αDdCBE pair even just a couple of nucleotides led to dramatically distinct editing profiles. This screening approach resulted in the installation of the m.3242G>A pathogenic variant at about 50% editing efficiency with low bystander editing in HEK293T cells. In summary, αDdCBEs are generally better editors than DdCBEs, successfully bypassing the 5’-T constraint, which rather than being an absolute barrier for DdCBE-mediated mtDNA editing, acts as a limiting factor in most target loci. These observations support unconstrained mitochondrial base editing as a strategy for disease modeling and potentially gene therapy applications.
Plain Language Summary
Mitochondrial base editing is a technology that can help us create disease models and gene therapies for mitochondrial diseases, for which there are few cellular and animal models and no cures. These technologies allow us to change one nucleotide or 'letter' in mitochondrial DNA, which is present in most cells within our bodies, into another letter. However, current mitochondrial base editing technologies are limited by specific rules known as context requirements, such as the presence of a specific letter in the vicinity of the DNA sequence that is being modified. Aiming to expand the mitochondrial base editing toolbox to facilitate the generation of disease models and gene therapies, we created and characterized mitochondrial base editors with fewer context requirements. These unrestricted mitochondrial base editors have allowed us to efficiently edit genes that were previously hard to edit, which will facilitate their application in research and, eventually, healthcare.
Santiago R. Castillo1, Karl J. Clark1, Stephen C. Ekker2
1Mayo Clinic, Rochester, MN,2University of Texas at Austin, Austin, TX"
Find This Session