CRISPR Base Editing

Base editing relies on the fusion of a deaminase enzyme with a catalytically impaired variant of a Cas nuclease, typically Cas9 or Cas12. The catalytically impaired Cas protein is engineered to retain its DNA-binding but is rendered incapable of introducing double-strand breaks. Instead, it precisely positions the deaminase to act on a target base within a specific DNA sequence. The Cas protein’s function in this system is purely to facilitate sequence targeting and spatial positioning, ensuring that the deaminase can act with high precision at the intended site within the genome.

Dead Cas9 (dCas9) and Cas9 nickase (nCas9) are widely used in base editing. dCas9 is a catalytically inactive Cas9 variant that binds DNA with high precision, guided by a base editing gRNA molecule, but lacks nuclease activity. This allows it to act solely as a DNA-binding scaffold, positioning the fused deaminase enzyme directly at the target site. Cas9 nickase, on the other hand, retains one active nuclease domain, introducing a single-strand nick instead of a cut through both strands. While this nick can enhance the DNA repair process, increasing the efficiency of the intended base conversion, it can also result in unintended insertions or deletions (indels) at the repair site, which may compromise editing precision. Both variants, however, remain crucial components of base editors for achieving precise and controlled genetic modifications.

Once bound to the target site, the linked deaminase chemically modifies the target base. A deaminase is a type of enzyme that removes an amino group (-NH2) from a molecule, a process known as deamination. These enzymes play a critical role in various biological pathways by modifying specific nucleotides in DNA or RNA, effectively altering their chemical structure. For instance, certain deaminases, like cytidine deaminases, can convert a cytosine (C) base into uracil (U) in RNA or into a uracil-like intermediate in DNA. This targeted chemical change creates opportunities to rewrite genetic information without physically cutting the DNA, making deaminases powerful molecular biology tools.

The deaminase enzyme is physically incorporated into the base editor complex by fusing it to the catalytically impaired Cas protein, typically at specific termini such as the N- or C-terminal regions. This strategic placement is crucial to ensure proper spatial alignment between the enzyme and the target DNA, allowing the deaminase to access and modify the desired nucleotide. However, this integration poses structural challenges, as the fusion must maintain the stability, activity, and targeting precision of both the catalytically impaired Cas protein and the deaminase. Improper configuration can hinder the complex's functionality or reduce editing efficiency. On the other hand, when effectively designed, this fusion creates a precise and compact system that combines the targeting power of the Cas protein with the enzymatic activity of the deaminase, enabling highly localized and efficient base modifications.

A guide RNA (gRNA) for base editing requires specific design criteria to ensure precision and efficiency, distinguishing it from gRNAs used for Cas9-mediated knockouts. With base editing, the gRNA must precisely target the site where the desired nucleotide modification occurs, positioning it within the editing window of the deaminase-Cas fusion complex. Unlike gRNAs for knockouts, where the primary goal is to induce a double-strand break at or near a functional gene region, base editing gRNAs prioritize accurate placement of the catalytically impaired Cas protein to ensure efficient and specific catalytic activity by the deaminase enzyme.

The editing window typically spans a narrow range of bases in the protospacer region, with the exact position depending on the molecular architecture of the base editor. The precise positioning of the target base within this window is critical for efficient editing, as bases outside this range cannot be converted effectively. Additionally, strict PAM requirements, dictated by the Cas nuclease component, further constrain the targetable regions and influence the initial binding and targeting of the base editor complex. Off-target potential and the risk of unintended bystander edits within the editing window add further complexity, requiring careful design to minimize such events. Aligning these considerations demands thorough in silico testing and iterative refinement of base editor gRNAs, ensuring they are both accurately positioned, and are compatible with the structural and functional parameters of the base editing system.

CBEs convert a cytosine (C) to a thymine (T), resulting in a C•G to T•A substitution. This base conversion is typically produced by the common cytidine deaminase, APOBEC1, fused to a Cas9 nickase. The base editor gRNA directs the CBE to the target site, unwinding the double-stranded DNA and exposing a single-stranded DNA region where the target cytosine is located. In this exposed region, the APOBEC1 deaminase catalyzes the conversion of cytosine to uracil through deamination. This intermediate uracil is structurally similar to thymine and pairs with adenine during DNA replication. In subsequent rounds of DNA synthesis, the uracil is interpreted as thymine by the replication machinery, effectively completing the C•G to T•A base substitution.

The use of uracil as an intermediate is critical to the editing process because direct enzymatic conversion of cytosine to thymine is chemically unfavorable and not feasible with current enzyme systems. By leveraging the natural deamination pathway, the CBE efficiently induces a transition mutation without creating double-stranded breaks. To enhance editing efficiency and prevent cellular repair mechanisms from reversing the change, cytosine base editors also include uracil glycosylase inhibitors (UGI). These inhibit the base excision repair enzyme uracil N-glycosylase (UNG), which would normally recognize and remove uracil from DNA, restoring the original cytosine. The presence of UGI ensures the uracil remains in place long enough to be replicated as thymine in subsequent cell divisions, preventing the cells from repairing the edit back to a cytosine. The original cytosine base editor, called BE3, was published by Komor et al. in 2016 and demonstrated efficient C-to-T conversions in mammalian cells.

Adenine base editors convert adenine (A) to guanine (G), effectively performing an A•T to G•C substitution. Different from the original CBE reported in 2016, there was no known natural enzyme that deaminates adenine in DNA at the time, they only acted on RNA. To overcome this, researchers needed to engineer the E. coli tRNA adenosine deaminase, TadA. To function on DNA, this engineered TadA variant forms a heterodimer with wild-type TadA and is fused to the Cas9 nickase in the ABE complex. Once the ABE complex binds to DNA, the DNA is unwound exposing the targeted adenine. The ABE deaminase then converts adenine into inosine (I), which the cell interprets as guanine during replication, producing an A•T to G•C substitution. The first ABE, known as ABE7.10, was reported by Gaudelli et al. in 2017, which showed high-efficiency editing at multiple genomic sites.

Each base editor is guided to the target site by a base editor gRNA, which results in the base conversion by the deaminase enzyme. By choosing the appropriate base editor and base editing guide RNA, cytosine and adenine base editors could address a large number of disease-causing point mutations. This has led researchers to incorporate base editors in applications such as cell and gene therapies.

Bystander edits represent a critical challenge in the field of precision gene editing, particularly in applications requiring high accuracy, such as therapeutic use. Bystander edits are unintended nucleotide changes that occur within the editing window of a base editor, often due to the enzyme’s activity on nearby bases rather than solely at the intended target site. The consequences of bystander edits can vary significantly depending on the genomic context and the coding frame involved. For instance, they may result in silent mutations that do not alter the encoded protein, missense mutations that substitute one amino acid for another, or more severe deleterious mutations that could disrupt protein function. For example, a bystander edit in a coding region could destabilize a protein or, in regulatory regions, alter gene expression unpredictably - compromising both research outcomes and therapeutic safety.

Bystander edits are particularly problematic in therapeutic gene editing - correcting a pathogenic mutation in a critical gene must avoid introducing new mutations that could trigger harmful side effects or unintended consequences in the patient. Several factors contribute to the likelihood of bystander edits, including the sequence composition around the target site, the specificity of the base editing enzyme, and the design of the base editing guide RNA. Additionally, different base editor variants have unique activity profiles, with some displaying broader editing windows that elevate the risk of such unintended changes.

Advances in base editing are helping address these challenges. Enhanced variants such as BE4max and ABE8e exhibit improved specificity and narrower editing windows, significantly reducing the risk of bystander edits. Additionally, computational tools and refined enzyme engineering approaches are enabling more precise targeting and minimizing off-target activity. These innovations are essential for ensuring the safety and reliability of base editing technologies in both research and therapeutic contexts.

Base editors are significantly larger than traditional CRISPR systems due to the fusion of a deaminase enzyme with a catalytically impaired Cas protein. This increased size presents a major challenge for delivery, particularly with adeno-associated viruses (AAVs), which have a limited packaging capacity of approximately 4.7 kilobases. The size of base editors often exceeds this limitation, making AAV-mediated delivery less feasible for therapeutic applications.

To overcome this obstacle, researchers have developed innovative strategies to enable the efficient delivery of base editors. One such approach involves the use of dual-vector systems, which split the base editor components between two separate AAVs that reassemble in target cells. Additionally, non-viral lipid nanoparticles (LNPs) have gained traction as an alternative delivery method, offering the ability to encapsulate larger payloads and reduce the risk of immune responses. Another promising solution is the engineering of compact base editors, such as leveraging Cas variants like Cas9-ms, which are specifically designed to reduce size while maintaining editing efficiency.

These advancements are crucial for tailoring delivery strategies to address therapeutic needs. Compact base editors and innovative delivery approaches like LNPs not only improve delivery efficiency but also minimize immune reactions and enhance targeting precision at specific genomic sites. These solutions are actively expanding the potential of base editing for clinical applications, enabling researchers to tackle genetic disorders with greater flexibility and efficacy.

Base editing efficiency is impacted by factors such as chromatin accessibility and cell-type-specific DNA repair pathways. Tightly packed heterochromatin regions can limit the ability of base editors to access certain genomic sites, significantly reducing editing efficiency. This poses a particular challenge when targeting genes in these inaccessible areas for research or therapeutic purposes.

Cell-specific DNA repair mechanisms, such as base excision repair (BER) and mismatch repair (MMR), further complicate base editing. These pathways can interact unpredictably with editing intermediates, leading to variability in outcomes or, in some cases, unintended insertions or deletions (indels). These effects highlight the importance of refining base editor tools to achieve more consistent and precise results, especially for therapeutic applications.

To address these challenges, researchers are developing strategies to enhance accessibility and manage DNA repair pathway interactions. Chromatin-modifying compounds, for instance, are being explored to open tightly packed heterochromatin and provide base editors better access to their target sites. Additionally, approaches to modulate DNA repair pathways, such as using small molecules to inhibit BER or MMR activity, are being tested to create more predictable and controlled editing outcomes. These advancements are critical for expanding the effectiveness and precision of base editing tools in both research and clinical settings.