Electrostatic remodeling of adenine deaminase in adenine base editor and application of adenine deaminase in elimination of off-target effect in whole genome

Abstract

The invention provides electrostatic remodeling of adenine deaminase in an adenine base editor and application of the adenine deaminase in elimination of off-target effect in a whole genome. The invention reveals for the first time that by performing strategic charge remodeling on a DNA-protein interface of an adenine deaminase Tad8e * subunit (a non-catalytic domain adjacent to a catalytic domain Tad8e), the off-target activity can be effectively eliminated while the targeted editing efficiency is maintained. The Tad8e * variant carrying charge reversal mutation significantly reduces a catalytic window of ABE8e, and completely eliminates off-target editing in a whole genome range. According to the method, DNA-protein charge reconstruction is determined as a universal strategy for improving the accuracy of the base editing system, and the key bottleneck of clinical transformation is solved.

Description

Technical Field

[0001] This invention belongs to the field of gene editing research. More specifically, this invention relates to the electrostatic remodeling of adenine deaminase in an adenine base editor and its application in eliminating off-target effects across the entire genome. Background Technology

[0002] Adenine base editors (ABEs) are a class of gene editing tools developed based on the CRISPR / Cas system, which use engineered adenine deaminase to achieve precise A-to-G base conversion.

[0003] The adenine base editor is formed by fusing the Cas9 nickase of Streptococcus pyogenes (nCas9) with a directionally evolved tRNA adenosine deaminase of Escherichia coli (TadA*). It deaminates target adenine (A) to hypoxanthine (I) via a deoxyinosine intermediate, and then performs the A→G conversion after DNA replication.

[0004] Adenine base editors, with their precise single-base editing capabilities, have achieved groundbreaking applications in multiple fields, including genetic disease treatment (single-gene genetic diseases, neurodegenerative diseases), cancer treatment, agriculture and biotechnology (crop improvement, microbial engineering), gene function research, and epigenetic regulation. Currently, many ABE-related clinical trials have entered Phase I / II, mainly focusing on rare diseases and oncology. However, the clinical application of adenine base editors still faces limitations in off-target effects (such as chromosomal structural variations) and delivery efficiency.

[0005] The key factor determining the specificity of the adenine base editor lies in the electrostatic interaction between the Tad8e deaminase domain and its corresponding DNA substrate—a mechanism that is still not fully understood.

[0006] It is noteworthy that the enhanced catalytic efficiency of ABE8e stems from its homodimerization: the Tad8e subunit from the main ABE8e molecule and another Tad8e subunit from the secondary ABE8e molecule (named Tad8e*) exhibit a synergistic effect (Lapinaite, Audrone, et al. "DNA capture by a CRISPR-Cas9–guided adenine baseeditor." Science 369.6503(2020):566-571). This dimer structure accelerates deamination by enhancing DNA binding stability and optimizing catalytic conformation, but the potential sacrifice of editing specificity requires further systematic evaluation. Structural studies suggest that the positively charged DNA-binding interface of Tad8e / Tad8e* may unexpectedly stabilize the conformation of non-target DNA, thereby expanding the catalytic range and promoting off-target deamination.

[0007] How to decouple targeting efficiency from off-target activity through systematic regulation of these electrostatic interactions remains unsolved. To fill this knowledge gap, it is necessary to analyze at the mechanistic level how the charge dynamics of the DNA-protein interface affect editing accuracy—this is the core challenge driving ABE8e towards clinical application. Summary of the Invention

[0008] The purpose of this invention is to provide an electrostatic remodeling of adenine deaminase in an adenine base editor and its application in eliminating off-target effects across the entire genome.

[0009] In a first aspect of the present invention, a method is provided to reduce the gene editing window of the adenine base editor ABE8e and improve its specificity (inhibit / eliminate off-target effects), comprising modifying the adenine deaminase in the ABE8e homodimer, wherein one adenine deaminase subunit from one of the homodimers is TadA8e (located in the main ABE8e molecule) and the other adenine deaminase subunit is TadA8e* (located in the secondary ABE8e molecule);

[0010] The modification includes: recharging the amino acid residues at the interface of the TadA8e-TadA8e* spatial structure that are close to (including in contact with) the DNA (substrate), thereby reducing the gene editing window of the adenine base editor ABE8e and improving specificity (inhibiting / eliminating off-target effects).

[0011] In one or more embodiments, the improved specificity includes reducing off-target effects of the adenine base editor.

[0012] In one or more embodiments, the improved specificity includes reducing nonspecific binding between the base editor and DNA.

[0013] In one or more embodiments, one of the ABE8e homodimers is defined as a primary ABE8e molecule and the other is defined as a secondary ABE8e molecule.

[0014] In one or more embodiments, the TadA8e is derived from the main ABE8e molecule as the catalytic domain Tad8e.

[0015] In one or more embodiments, the TadA8e* comes from the sub-ABE8e molecule as a non-catalytic domain adjacent to the catalytic domain Tad8e.

[0016] In one or more embodiments, the charge remodeling includes: analyzing the charge type of amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure, introducing a charge reversal mutation, and modifying positively charged amino acid residues into negatively charged amino acid residues; preferably, the positively charged amino acid is R, H or K, and the negatively charged amino acid is D or E.

[0017] In one or more embodiments, the amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure include: residues H128, R129, H96 or R98 located in Tad8e*, or residues H52, R111, R150, R23, R26, R47, H57, K110 or R153 located in Tad8e.

[0018] In one or more embodiments, the amino acid residue is H128, H96, R98, or R129 located at Tad8e*.

[0019] In one or more embodiments, the amino acid residue is H128 located at Tad8e*.

[0020] In one or more embodiments, when a charge reversal mutation is performed on one of the amino acid residues, the other amino acids in TadA8e and TadA8e* of the dimer remain unchanged (wild type).

[0021] In one or more embodiments, charge reversal mutations are performed on one, two or more amino acid residues.

[0022] In one or more embodiments, the mutation at the amino acid site is calculated based on the full-length amino acid sequence of the TadA8e protein (a single subunit).

[0023] In one or more embodiments, the TadA8e protein includes a protein selected from the group consisting of: (a1) a protein with the amino acid sequence of SEQ ID NO:1; (b1) a protein derived from (a1) having the function of the (a1) protein, formed by substituting, deleting, or adding one or more (e.g., 1-20; preferably 1-15; more preferably 1-10, such as 5, 3) amino acid residues of the amino acid sequence of SEQ ID NO:1; (c1) a protein derived from (a1) having the function of the (a1) protein, having at least 80% (preferably 85%; more preferably 90%; more preferably 95%, such as 98%, 99%) homology to the protein sequence defined by (a1); or (d1) a protein active fragment defined by (a1), or a protein formed by adding a tag sequence, an enzyme digestion sequence, or a reporter protein to both ends thereof. The TadA8e also includes the aforementioned biologically active fragment.

[0024] In another aspect of the invention, a variant of adenine deaminase TadA8e is provided, comprising a mutation selected from the group consisting of amino acid residues: H128, R129, H96, R98, H52, R111, R150, R23, R26, R47, H57, K110, or R153; preferably, it comprises a mutation selected from the group consisting of residues located at Tad8e*: H128, R129, H96, or R98; wherein the amino acid residue mutation is D or E.

[0025] In another aspect of the invention, an adenine base editor is provided, which is an ABE8e homodimer, one of which (the main ABE8e molecule) includes the adenine deaminase subunit TadA8e, and the other includes the adenine deaminase subunit TadA8e* (located in the secondary ABE8e molecule); the amino acid residues at the interface of the TadA8e-TadA8e* spatial structure that are close to the DNA are remodeled by charge and have charge reversal mutations, the charge reversal mutations including: positively charged amino acid residues mutating into negatively charged amino acid residues; preferably, the positively charged amino acid is R, H or K, and the negatively charged amino acid is D or E.

[0026] In one or more embodiments, the amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure include: residues H128, R129, H96, or R98 located in Tad8e*, and residues H52, R111, R150, R23, R26, R47, H57, K110, or R153 located in Tad8e; more preferably, the amino acid residues are H128 located in Tad8e*.

[0027] In one or more embodiments, the adenine base editor further includes an element operatively linked to the adenine deaminase for adenine editing; preferably, the element includes a Cas nuclease.

[0028] In one or more embodiments, the adenine base editor further includes an NLS operatively linked to the adenine deaminase.

[0029] In one or more embodiments, the Cas nuclease includes (but is not limited to): Cas9 nuclease or its homologs, the homologs including (but not limited to): SpRY nuclease, SaKKH nuclease, IscB nuclease.

[0030] In one or more embodiments, the NLS is a BPNLS.

[0031] In one or more embodiments, the TadA8e variant also includes a linker peptide between itself and the Cas nuclease.

[0032] In another aspect of the invention, isolated polynucleotides are provided, said polynucleotides encoding: any of the adenine deaminase TadA8e variants described above; or any of the adenine base editors described above.

[0033] In another aspect of the invention, a construct, or an expression vector containing the construct, or a recombinant cell containing the construct is provided; the construct or expression vector comprises the polynucleotide encoding the adenine deaminase TadA8e variant, or the adenine base editor.

[0034] In one or more embodiments, the construct includes the expression cassette of the adenine base editor and an operatively linked sgRNA expression cassette for a target gene.

[0035] In one or more embodiments, the construct further includes a reporter protein expression cassette.

[0036] In one or more embodiments, the recombinant cells contain the aforementioned construct or expression vector, or the aforementioned polynucleotide (encoding the corresponding amino acid sequence) is integrated into the genome.

[0037] In another aspect of the invention, the use of the adenine base editor is provided for gene editing or for preparing reagents for gene editing; wherein the gene editing has a reduced gene editing window of the adenine base editor ABE8e and improved specificity (inhibition / elimination of off-target effects); preferably, the gene editing is adenine base editor-mediated gene editing; preferably, the gene editing is an A·T to G·C conversion (A / T>G / C bases).

[0038] In one or more embodiments, the adenine base editor is used for gene editing as a cellular-level method, including single-cell, two-cell, or multi-cell methods.

[0039] In one or more embodiments, the described use is applied to cells or cell cultures isolated in vitro.

[0040] In one or more embodiments, the use is for purposes other than disease diagnosis or treatment.

[0041] In another aspect of the invention, a method for gene editing (including gene editing whose primary purpose is the conversion of A·T to G·C) is provided, comprising mediating gene editing with the adenine base editor or the construct or an expression vector containing the construct.

[0042] In one or more embodiments, gene editing is performed by transfecting or injecting a receptor with a nucleic acid sequence encoding the adenine base editor.

[0043] In one or more embodiments, the receptor includes somatic cells or germ cells; preferably, the germ cells include embryonic cells or fertilized eggs.

[0044] In one or more embodiments, the gene editing method is an in vitro method for non-living organisms.

[0045] In one or more embodiments, the gene-editing method is targeted at objects that do not develop into living organisms.

[0046] In one or more embodiments, the gene editing method is a cellular-level method, including single-cell, two-cell, or multi-cell methods.

[0047] In another aspect of the invention, a reagent or kit for gene editing is provided, comprising: the TadA8e variant, or a polynucleotide encoding thereof; or the adenine base editor, or a polynucleotide encoding thereof; or the construct or an expression vector containing the construct.

[0048] Other aspects of the invention will be apparent to those skilled in the art from the disclosure herein. Attached Figure Description

[0049] Figure 1 Electrostatic remodeling of the Tad8e-DNA interface enhances ABE8e editing specificity.

[0050] (a) Cryo-electron microscopy analysis of the Tad8e-Tad8e* dimer-DNA interface revealed 14 positively charged residues that are crucial for DNA interactions (9 in Tad8e and 5 in Tad8e).

[0051] (b) Schematic diagram of charge reversal mutation strategy (positive charge → aspartic acid), showing the modification scheme of Tad8e / Tad8e* residues and the evaluation process in homologous 293T cells containing 102 pre-defined DNA target sites.

[0052] (c) Comparison of editing efficiency between ABE8e variant and wild type (WT), with R74D variant showing a significant decrease in activity (P<0.05, Student T test).

[0053] (d) Narrowing of the editing window observed in the charge-reversed variants of Tad8e (7 / 9 variant) and Tad8e* (5 / 5 variant).

[0054] Figure 2 Charge reversal engineering at the Tad8e*-DNA interface eliminates the off-target effects of ABE8e across the entire genome.

[0055] (a) Flowchart of GOTI technology (two-cell embryo injection whole genome off-target analysis);

[0056] (b) The targeted editing efficiency of the ABE8e variant in tdTomato+ cells (92.7±1.2% to 99.1±0.8%) compared with the baseline level of the tdTomato- control group (<5±1%);

[0057] (c) Comparison of genome-wide insertion / deletion mutation frequencies between wild-type ABE8e (350±50 events / embryo), BE3 control group (300±20 events / embryo) and ABE8e-H128D (no significant difference from Cre control group, P>0.05);

[0058] (d) Mutation characteristics analysis: A comparison between ABE8e-WT-induced A / T>G / C switching (accounting for 92% of off-target sites) and the lack of sequence preference in ABE8e-H128D.

[0059] Figure 3 DNA interface Tad8e* charge reversal engineering eliminates the off-target effects of ABE8e throughout the transcriptome;

[0060] (a) Off-target analysis of whole transcriptome: HEK293T cells transfected with the ABE8e variant were subjected to RNA sequencing (RNA-seq) by flow cytometry sorting (FACS) 48 hours after transfection;

[0061] (b) Quantitative analysis of adenosine to inosine (A-to-I) editing events showed that the off-target level of ABE8e-H128D was close to the background value (1.2 ± 0.3 times vs EGFP control, P = 0.27); ABE8e-H52D (reduced by 34%) and R100D (reduced by 41%) were significantly lower than those of wild type (ABE8e-WT) (*P < 0.01, two-tailed t test).

[0062] Figure 4 ABE8e-H128D can achieve powerful PCSK9 gene editing in mice.

[0063] Figure 5 After intravenous injection of low-dose LNP (0.5-4 mg / kg) via the tail vein, liver editing efficiency and changes in systemic lipid profiles were monitored during a 6-month observation period. Detailed Implementation

[0064] Base editing technologies, particularly adenine base editors (ABEs), have revolutionized efficient and precise genomic medicine by enabling targeted switching of A / T>G / C base pairs without inducing DNA double-strand breaks. Despite progress, the clinical translation of ABEs is hampered by pervasive off-target effects, manifested as unintended edits across the genome and transcriptome. These off-target activities pose significant safety risks, limiting the feasibility of treatments, including those using state-of-the-art ABE8e editors. While previous efforts to improve specificity have focused primarily on modifying deaminase activity or altering delivery systems, they have struggled to overcome the problem of impaired targeting efficiency or completely eliminate transcriptome-wide off-target effects.

[0065] The adenine base editor ABE8e holds revolutionary potential in the field of precise genome editing, but its therapeutic applications remain limited by genome-wide off-target effects and transcriptome-wide side effects. This invention reveals for the first time that strategic charge remodeling of the DNA-protein interface of Tad8e* (a non-catalytic domain adjacent to the catalytic domain Tad8e) can effectively eliminate off-target activity while maintaining targeted editing efficiency. Structural-functional analysis shows that the Tad8e* variant carrying charge reversal mutations significantly reduces the catalytic window of ABE8e and completely eliminates genome-wide off-target editing (verified by the GOTI method). The engineered ABE8e-H128D variant exhibits strong inhibition of transcriptome-wide off-target effects in both in vitro and in vivo models, achieving unprecedented specificity. This invention establishes DNA-protein charge remodeling as a universal strategy for improving the accuracy of base editing systems, solving a key bottleneck in clinical translation and expanding its medical application prospects.

[0066] As used in this invention, the term "animal" is not particularly limited, as long as its cells have a genome in the general sense and the gene-editing system is active within its cells. For example, the animal can be a mammal, including humans, non-human primates (monkeys, orangutans), livestock and agricultural animals (e.g., pigs, sheep, cattle), mice (mice), and rodents (e.g., mice, rats, rabbits), etc.

[0067] As used in this invention, the term "cell" includes, but is not limited to: somatic (tissue) cells, induced pluripotent stem cells, and germ cells (such as fertilized egg cells and oocytes).

[0068] As used in this invention, "mutation" refers to the substitution of a residue in a sequence (e.g., a nucleic acid or amino acid sequence) by another residue, or the change of one or more residues in a sequence to another residue, or the occurrence of deletion or insertion. In the editor of this invention, it is desirable to reduce the gene editing window of the adenine base editor ABE8e and improve specificity (suppress / eliminate off-target effects).

[0069] As used herein, unless otherwise stated, the terms “TadA8e variant”, “TadA8e mutant”, “TadA8e variant”, and “mutant TadA8e” are used interchangeably.

[0070] If it is necessary to represent the unmutated TadA8e (TadA8e* sequence is the same), it can be an enzyme with an amino acid sequence such as SEQ ID NO:1. Unless otherwise stated, the mutation sites of the variants in this invention are based on the sequence shown in SEQ ID NO:1.

[0071] In this invention, unless otherwise stated, the variant is identified by “the amino acid that was replaced at the original amino acid position” to indicate the mutated amino acid in the variant, such as H128D, which means that the amino acid at position 128 is replaced by D by the H of the starting enzyme.

[0072] In this invention, the "target gene" can also be called the "purpose gene," which refers to a gene of interest that, after being introduced into a cell, is useful for observing changes in the cell, regulating cell performance, or improving cell-related diseases.

[0073] As used in this invention, the term "off-target effect" refers to the failure of a modification to a specific location in the genome to achieve the pre-set target, resulting in a deviation from the intended modification or failure to perform the modification. Causes of "off-target effects" include, but are not limited to: inaccurate binding to the target site, inaccurate cleavage operations after sequence recognition, and insufficient precision in editing the cleavage site.

[0074] As used in this invention, "operationally linked" or "operably coupled to" refers to a situation where certain portions of a linear DNA sequence can regulate or control the activity of other portions of the same linear DNA sequence. For example, if a promoter controls transcription of a sequence, then it is operably coupled to the coding sequence. "Operationally sequential coupling" refers to the coupling of elements in a specific order, such as from amino acid to carboxyl terminus.

[0075] As used in this invention, an "element" refers to a series of functional nucleic acid / protein sequences useful for protein expression, which are systematically constructed to form an expression construct. The sequences of the "elements" can be those provided in this invention, as well as variants thereof, provided that these variants substantially retain the function of the "elements," obtained by inserting or deleting bases (e.g., 1-50 bp; preferably 1-30 bp, more preferably 1-20 bp, even more preferably 1-10 bp), or by random or site-directed mutagenesis.

[0076] As used in this invention, the term "construction" refers to a single-stranded or double-stranded DNA molecule that has been artificially modified to contain DNA fragments arranged and combined according to sequences not found in nature. The "construction" may include a "plasmid," an "in vitro transcription product," or a viral vector; or, the "construction" may be contained within an expression vector as part of the expression vector.

[0077] As used in this invention, the "sgRNA" refers to "Single-guide RNA (sgRNA)," which is designed based on a "target site on a target gene." Its sequence is sufficient to synergize with the endonuclease Cas to guide a Cas-mediated DNA double-strand break at the target site. In this invention, the "sgRNA" includes sgRNA in RNA form (such as mRNA), as well as sequences in DNA form corresponding to the sgRNA sequence or constructs containing said sequences, as long as they can be processed or converted into active "sgRNA" within the cell.

[0078] Developing base editors with minimized off-target activity remains a key challenge in therapeutic genome editing. This invention discloses a charge remodeling strategy targeting the Tad8e* DNA-binding interface, which effectively eliminates the off-target effects of ABE8e. By introducing charge reversal mutations, the inventors constructed a Tad8e* variant that narrows the catalytic window, significantly inhibiting genome-wide and transcriptome-wide off-target activity without affecting editing efficiency. This invention not only overcomes a key obstacle in the clinical translation of ABE, but also establishes DNA-protein charge remodeling as a universal paradigm for optimizing the accuracy of genome editing systems.

[0079] Based on the inventors' new discovery, a method is provided to reduce the gene editing window of the adenine base editor ABE8e and improve its specificity. This method includes modifying the adenine deaminase in the ABE8e homodimer, wherein one adenine deaminase subunit from the homodimer is TadA8e (located in the main ABE8e molecule), and the other adenine deaminase subunit is TadA8e* (located in the secondary ABE8e molecule). The modification includes recharging the amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure, thereby reducing the gene editing window of the adenine base editor ABE8e and improving its specificity (inhibiting / eliminating off-target effects).

[0080] In a preferred embodiment of the present invention, the charge remodeling includes: analyzing the charge type of amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure, introducing a charge reversal mutation, and modifying positively charged amino acid residues into negatively charged amino acid residues; preferably, the positively charged amino acid is R, H or K, and the negatively charged amino acid is D or E.

[0081] This invention also provides an adenine base editor, which is an ABE8e homodimer. One of the homodimers (the main ABE8e molecule) includes the adenine deaminase subunit TadA8e, and the other includes the adenine deaminase subunit TadA8e* (located in the secondary ABE8e molecule). The amino acid residues at the interface of the TadA8e-TadA8e* spatial structure, which are close to the DNA, undergo charge remodeling and possess charge reversal mutations. These charge reversal mutations include: positively charged amino acid residues mutating into negatively charged amino acid residues. Preferably, the positively charged amino acid is R, H, or K. The negatively charged amino acid is D or E; more preferably, the amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure include: residues H128, R129, H96 or R98 located in Tad8e*, or residues H52, R111, R150, R23, R26, R47, H57, K110 or R153 located in Tad8e*; more preferably, the amino acid residue is H128 located in Tad8e*.

[0082] This invention can also be applied to fragments, derivatives, and analogs of the TadA8e variants described herein. As used herein, the terms “fragment,” “derivative,” and “analyte” refer to proteins that substantially retain the same biological function or enzymatic activity as the TadA8e variants of this invention. Fragments, derivatives, or analogs of TadA8e can be (i) proteins with one or more conserved or non-conserved amino acid residues (preferably conserved amino acid residues) substituted, which may or may not be encoded by the genetic code; or (ii) proteins having substituents in one or more (e.g., 1-20, more preferably 1-10, even more preferably 1-8, 1-5, 1-3, or 1-2) amino acid residues; or (iii) proteins formed by fusing additional amino acid sequences to this protein sequence (e.g., leader sequences, secretory sequences, sequences used to purify this protein, or proteoprotein sequences, or fusion proteins). These fragments, derivatives, and analogs are within the scope well known to those skilled in the art as defined herein. However, the condition that must be met is that the amino acid sequence of the TadA8e variant and its fragments, derivatives and analogs must contain at least one mutation specifically pointed out above in this invention.

[0083] For example, in this art, substitution with amino acids of similar or identical properties generally does not alter protein function. Similarly, adding or deleting one or more amino acids at the C-terminus and / or N-terminus generally does not change protein function. This term also includes enzyme-active fragments and enzyme-active derivatives of the TadA8e variant. However, in these variant forms, there is certainly a mutation at at least one of the key positions described above in this invention.

[0084] In the adenine base editor of the present invention, each element can be a recombinant protein or a synthetic protein, preferably a recombinant protein.

[0085] The present invention also includes variants formed by changing the functionality of each element, i.e., functionally conservative variants (including fragments, derivatives and analogues, etc.).

[0086] The present invention also provides a multinucleotide sequence encoding a selected or modified enzyme variant or a conserved variant thereof.

[0087] The polynucleotide encoding the mature protein of the aforementioned variant includes: a coding sequence that encodes only the mature protein; a coding sequence of the mature protein and various additional coding sequences; a coding sequence of the mature protein (and optional additional coding sequences) and a non-coding sequence. "Polynucleotide encoding a protein" can be a polynucleotide that includes the protein itself, or it can include additional coding and / or non-coding sequences.

[0088] The selected, optimized, or modified full-length enzyme nucleotide sequences or fragments thereof of the present invention can generally be obtained by PCR amplification, recombinant methods, or artificial synthesis. For PCR amplification, primers can be designed based on the nucleotide sequences disclosed in the present invention, especially the open reading frame sequences, and the relevant sequences can be amplified using commercially available cDNA libraries or cDNA libraries prepared according to conventional methods known to those skilled in the art as templates. When the sequence is long, it is often necessary to perform two or more PCR amplifications, and then splice the fragments amplified from each amplification in the correct order.

[0089] Once the relevant sequence is obtained, it can be obtained in large quantities using recombination methods. This typically involves cloning it into a vector, transferring it into cells, and then isolating the sequence from the proliferated host cells using conventional methods.

[0090] The present invention also relates to vectors containing the polynucleotides of the present invention, host cells generated by genetic engineering using the vectors of the present invention or selected, optimized or modified enzyme-coding sequences, and methods for generating the proteins of the present invention via recombinant technology.

[0091] Vectors containing the appropriate DNA sequence and appropriate promoter or control sequence can be used to transform appropriate host cells or recipient cells.

[0092] This invention also provides a method for gene editing, including gene editing mediated by the adenine base editor described in this invention. Besides using the adenine base editor described in this invention for gene editing, other gene editing reagents known in the art can be used. In the embodiments of this invention, preferred constructs and implementation methods are provided.

[0093] In this invention, there are no particular limitations on the applicable gene editing targets; they can be somatic cells or germ cells, animal cells or human cells.

[0094] In this invention, sgRNA can be introduced into cells separately or together with the adenine base editor described above.

[0095] The sgRNA and the adenine base editor described herein may be included in the pharmaceutical composition.

[0096] The pharmaceutical composition may also contain a pharmaceutically acceptable carrier. Such carriers include (but are not limited to): saline, buffer solutions, glucose, water, glycerol, ethanol, and combinations thereof. Generally, the pharmaceutical formulation should be matched to the route of administration. The pharmaceutical compositions of the present invention can be formulated as injections, for example, prepared using conventional methods with physiological saline or an aqueous solution containing glucose and other excipients. The pharmaceutical compositions are preferably manufactured under aseptic conditions. The dosage of the active ingredient is a therapeutically effective amount.

[0097] In this invention, by strategically modifying the DNA-protein charge interactions at the Tad8e* interface, unprecedented editing precision can be achieved while maintaining targeting efficiency.

[0098] According to the demonstrable data of this invention, structure-based Tad8e* charge reversal engineering reduces the editing window by 62% compared to ABE8e-WT, effectively inhibiting deaminase slippage—a major source of genome-wide off-target effects. Unlike traditional strategies that disrupt deaminase-DNA interactions through steric hindrance or reduced affinity, the H128D mutation in ABE8e-H128D selectively inhibits off-target binding while maintaining canonical site catalytic activity by introducing local negative charge repulsion. This mechanism is robust in various experimental models: compared to ABE8e-WT, ABE8e-H128D reduces transcriptome off-target editing by 89% and 76% in primary hepatocytes (in vitro) and mouse livers (in vivo), respectively, outperforming existing high-fidelity variants such as ABE8e-SGC8 (which only reduces it by 54%).

[0099] In summary, this invention establishes DNA-protein charge remodeling as an effective strategy for precise base editing. ABE8e-H128D, by decoupling specificity enhancement and catalytic activity, provides a superior solution for the treatment of chronic diseases requiring long-term editing and strict control of off-target risks.

[0100] The invention will be better understood from the following examples. However, those skilled in the art will understand that the specific methods and results are merely for illustrating the invention and not for limiting it. Experimental methods in the following examples that do not specify specific conditions are generally performed according to conventional conditions such as those described in J. Sambrook et al., Molecular Cloning: A Laboratory Manual, Science Press, or according to the manufacturer's recommendations.

[0101] TadA8e (adenosine deaminase) sequence:

[0102]

[0103] Example 1: The effect of electrostatic remodeling of the Tad8e-DNA interface on ABE8e editing performance

[0104] Through in-depth research and analysis, the inventors discovered that by regulating the electrostatic dynamics of Tad8e-DNA interactions, the off-target effects of ABE8e-mediated genome-wide sgRNA-independent processes can be reduced. Based on the previously established mechanism—that ABE8e editing efficiency depends on Tad8e-Tad8e* dimerization during deaminase activation—14 positively charged residues at the Tad8e dimer-DNA interface were identified using cryo-electron microscopy structural analysis. Figure 1 a) Of these, 9 are located in Tad8e and 5 are located in Tad8e*.

[0105] The nine residues located in Tad8e include: R23, R26, R47, H52, H57, K110, R111, R150, and R153.

[0106] The five residues located at Tad8e* include: R74, H96, R98, H128, and R129.

[0107] Using electrostatic guidance theory, each of the 14 positively charged residues was individually mutated to aspartic acid (Asp), and each variant was introduced into ABE8e. The resulting variants include: 9 Tad8e variants: R23D, R26D, R47D, H52D, H57D, K110D, R111D, R150D, and R153D; and 5 Tad8e* variants: R74D, H96D, R98D, H128D, and R129D.

[0108] When performing variant analysis, unless otherwise stated, only mutations at one position of Tad8e or Tad8e* are considered; when applying dimers, at least one of Tad8e or Tad8e* in the variant must be wild-type and the other must have a mutation at one position.

[0109] Evaluation was performed in homologous 293T cells carrying 102 pre-defined DNA target sites (i.e., target sequences, amplified using uniform primer pairs). Variants modified with Tad8e / Tad8e* residues were analyzed in homologous 293T cells containing 102 pre-defined DNA target sites. Figure 1 b).

[0110] Specifically, the U6 promoter-sgRNA-target sequence-CMV promoter-EGFP-PA (PolyA) was introduced into a lentiviral backbone vector, packaged into a virus, and then transfected into 293T cells for culture. Table 1 provides examples of the sgRNA sequences targeting 102 pre-defined DNA target sites.

[0111] Table 1

[0112]

[0113] ABE8e setup: Connect the following elements in sequence: BPNLS, TadA8e (wild-type or mutant dimer), linker, nCas9, BPNLS, polyA.

[0114] CMV-BPNLS-Tad8e (carrying wild-type or mutant Tad8e dimer)-nCas9-PA-CMV-PuroR-mCherry-PA was introduced into the plasmid backbone and transfected into the aforementioned cells. After 24 hours of culture, cells were selected with puromycin, and positive cells were collected after 3 days. DNA was extracted and NGS sequencing was performed to analyze the editing efficiency.

[0115] The results showed that, except for the R74D variant (P<0.05, Student t-test), the other variants (13 in total) maintained similar editing efficiency to ABE8e-WT. Figure 1 c).

[0116] The edit window was significantly narrowed in the charge-reversal variants of Tad8e and Tad8e*. Figure 1 d) shows a significant improvement in the editing precision of electrostatically modulated variants, reducing / eliminating modifications to non-target sites.

[0117] according to Figure 1d, among which the improved editing precision of the charge-reversed variant of Tad8e* (5 / 5 variant) is more significant and superior, indicating that strategic charge reshaping of the DNA-protein interface of Tad8e* is of great significance.

[0118] Example 2: Off-target analysis of genome editing efficiency during two-cell embryo injection

[0119] To rigorously evaluate whether and how charge-reversed variants can reduce genome-wide off-target effects, the inventors employed the ultrasensitive GOTI technique (two-cell embryo injection genome-wide off-target analysis, see 201910153546.3).

[0120] In the GOTI experiment, mRNA, sgRNA, and Cre recombinase mRNA (CremRNA) selected from one of 14 variants of ABE8e were used. Figure 2 A single blastomer was co-injected into a two-cell stage embryo of an Ai9 reporter mouse. After Cre recombinase-mediated excision of the upstream 3X termination cassette of the tdTomato reporter gene, specific tdTomato markers were observed in the injected blastomer-derived cell lines. On day 14.5 (E14.5), tdTomato+ and tdTomato- cells from the digested embryos were separated by flow cytometry sorting (FACS). Genomic DNA was extracted for whole-genome sequencing (WGS) and off-target analysis of tdTomato-expressing cells. Unlabeled cells from uninjected sister blastomeres served as a control. Figure 2 a).

[0121] Table 2 shows examples of the sgRNA sequences involved.

[0122] Table 2

[0123]

[0124] Editing efficiency analysis showed that in tdTomato+ cells, the target site modification rate ranged from 89±2.52% to 99.1±0.8%, while the unedited tdTomato- control group maintained a baseline editing level of <5±1%. Figure 2 (b) Of the 14 candidate ABE8e variants, 13 significantly reduced genome-wide off-target effects, including: 9 Tad8e variants: R23D, R26D, R47D, H52D, H57D, K110D, R111D, R150D, and R153D; and 4 Tad8e* variants: H96D, R98D, H128D, and R129D.

[0125] Whole-genome sequencing (WGS) of matched cell populations revealed significant off-target features: the frequency of insertion / deletion mutations associated with ABE8e wild-type (WT) cells was approximately 350 ± 50. Figure 2 c). Notably, the ABE8e-H128D variant with charge reversal introduced into Tad8e* exhibited off-target activity close to background levels, showing no significant difference from the Cre control group (P>0.05), indicating that the genome-wide off-target effect of ABE8e-H128D has been reduced to the background level of the GOTI detection protocol.

[0126] Further analysis revealed that the ABE8e-WT modified sgRNA-dependent off-target sites were not detected in ABE8e-H128D edited embryos (P = 0.002, Fisher's exact test).

[0127] These analyses confirm that strategic charge reversal modifications at certain sites can significantly improve the whole-genome accuracy of the adenine base editor, indicating that electrostatic potential regulation is an effective strategy to improve ABE specificity.

[0128] Example 3: Analysis of the impact of ABE8e structural modifications on the transcriptome.

[0129] To assess whether structural modifications of ABE8e affect transcriptome-level specificity, the inventors performed RNA-seq analysis on wild-type (WT) ABE8e and candidate charge-reversal variants.

[0130] Plasmids encoding each editor were transfected into HEK293T cells. 48 hours after transfection, cells were separated by flow cytometry (FACS) for RNA extraction and sequencing. Figure 3 a).

[0131] Quantitative analysis of whole transcriptome adenosine to inosine (A-to-I) editing events revealed significant differences in off-target characteristics: compared with ABE8e-WT, the transcriptome off-target activities of ABE8e-H52D, -R100D and -H128D were reduced by 62-89% (P<0.01, two-tailed t-test). Figure 3 b).

[0132] Notably, the A-to-I frequency of ABE8e-H128D was close to background levels, showing no statistically significant difference compared to the EGFP control group (fold change = 1.2 ± 0.3, P = 0.27). While the other variants showed moderately reduced A-to-I frequencies (H52D: 34%; R100D: 41%), these differences were not statistically significant compared to ABE8e-WT. Figure 3 b).

[0133] The results indicate that charge reversal of DNA-interacting residues can significantly enhance the RNA specificity of the adenine base editor, with the ABE8e-H128D variant exhibiting high specificity, thus serving as the basis for subsequent analysis.

[0134] Example 4: Lipid Nanoparticles (LNP)

[0135] Establishment of lipid nanoparticles (LNPs):

[0136] 1) sgRNA was designed targeting the mouse PCSK9 gene and prepared using a chemical synthesis method; the sequence of the sgRNA is: CCCATACCTTGGAGCAACGG (SEQ ID NO:3);

[0137] 2) Preparation of ABE8e-H128D mRNA using in vitro transcription technology;

[0138] 3) Using a microfluidic device, encapsulate sgRNA and mRNA at a 1:1 mass ratio using LNP (SM102 formulation).

[0139] To evaluate the therapeutic potential of LNP delivery of ABE8e-H128D, the PCSK9-targeted editing effect was analyzed through a longitudinal in vivo mouse study.

[0140] Following intravenous injection of low-dose LNP (0.5-4 mg / kg) via the tail vein, liver editing efficiency and changes in systemic lipid profiles were monitored over a 6-month observation period. Figure 4 , Figure 5 ).

[0141] Liver tissue analysis showed that both ABE8e wild-type (WT) and ABE8e-H128D reached peak editing efficiency at a dose of 2 mg / kg (WT: 52±2.8% vs. H128D: 50±3.1%; P=0.42), and higher doses did not significantly improve efficiency. Follow-up studies on low-density lipoprotein cholesterol (LDL-C) showed that at a dose of 0.1 mg / kg, mice treated with ABE8e-H128D already exhibited a significant reduction in lipoprotein levels. At a concentration of 0.5 mg / kg, the LDL-C level in mice treated with ABE8e-H128D showed a more significant reduction, and higher concentrations did not significantly further promote a decrease in LDL-C.

[0142] The results showed that a single dose of LNP-ABE8e-H128D could achieve stable and precise editing of the PCSK9 gene in hepatocytes and achieve long-term regulation of LDL-C in mouse blood, demonstrating that ABE8e-H128D is as effective as ABE8e-WT in clinical use.

[0143] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims. Furthermore, all documents mentioned in this invention are incorporated herein by reference as if each document were individually incorporated by reference.

Claims

1. A method for narrowing the gene editing window of the adenine base editor ABE8e and improving its specificity, comprising modifying the adenine deaminase in the ABE8e homodimer, wherein one adenine deaminase subunit from the homodimer is TadA8e and the other adenine deaminase subunit is TadA8e*; The modifications include: By re-charge-modifying amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure, the gene editing window of the adenine base editor ABE8e is narrowed and its specificity is improved.

2. The method as described in claim 1, characterized in that, The charge remodeling includes: analyzing the charge type of amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure, introducing charge reversal mutations to modify positively charged amino acid residues into negatively charged amino acid residues; preferably, the positively charged amino acid is R, H or K, and the negatively charged amino acid is D or E.

3. The method as described in claim 1 or 2, characterized in that, The amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure include: residues H128, R129, H96 or R98 located in Tad8e*, or residues H52, R111, R150, R23, R26, R47, H57, K110 or R153 located in Tad8e. Preferably, the amino acid residue is H128, H96, R98 or R129 located at Tad8e*; More preferably, the amino acid residue is H128 located at Tad8e*.

4. A variant of adenine deaminase TadA8e, comprising mutations selected from the following group of amino acid residues: H128, R129, H96, R98, H52, R111, R150, R23, R26, R47, H57, K110, or R153; preferably, comprising mutations selected from the following group of amino acid residues: residues located at Tad8e*: H128, R129, H96, or R98; said amino acid residue mutation being D or E.

5. An adenine base editor, which is an ABE8e homodimer, one of which includes an adenine deaminase subunit TadA8e and the other includes an adenine deaminase subunit TadA8e*; the amino acid residues at the interface of the TadA8e-TadA8e* spatial structure close to the DNA are recharge-remodeled and have charge-reversal mutations, said charge-reversal mutations including: A positively charged amino acid residue is mutated into a negatively charged amino acid residue; preferably, the positively charged amino acid is R, H or K, and the negatively charged amino acid is D or E; More preferably, the amino acid residues at the interface close to DNA in the TadA8e-TadA8e* spatial structure include: residues H128, R129, H96, or R98 located in Tad8e*, and residues H52, R111, R150, R23, R26, R47, H57, K110, or R153 located in Tad8e; more preferably, the amino acid residues are H128 located in Tad8e*; more preferably, the amino acid residues are H128 located in Tad8e*.

6. The adenine base editor as described in claim 5, characterized in that, The adenine base editor further includes an element that is operatively linked to the adenine deaminase to perform adenine editing; preferably, the element includes a Cas nuclease.

7. An isolated polynucleotide, wherein the polynucleotide encodes: The adenine deaminase TadA8e variant of claim 4; or The adenine base editor according to any one of claims 5-6.

8. A construct, or an expression vector containing said construct, or a recombinant cell containing the like; said construct or expression vector comprising the polynucleotide of claim 7, encoding the adenine deaminase TadA8e variant, or the adenine base editor.

9. Use of the adenine base editor according to any one of claims 5-6 for gene editing or for preparing reagents for gene editing; wherein the gene editing has a reduced gene editing window of the adenine base editor ABE8e and improved specificity; preferably, the gene editing is adenine base editor-mediated gene editing; preferably, the gene editing is an A·T to G·C conversion.

10. A method of gene editing, comprising mediating gene editing using the adenine base editor or the construct or an expression vector containing the construct; Preferably, gene editing is performed by transfecting or injecting a receptor with a nucleic acid sequence encoding the adenine base editor; preferably, the receptor comprises: Somatic cells or germ cells; Preferably, the germ cells include embryonic cells or fertilized eggs.

Images