A base editor acgbemax that simultaneously enables purine and pyrimidine replacement

Abstract

The application discloses a base editor ACGBEmax realizing simultaneous replacement of purine and pyrimidine, relates to the technical field of gene editing, and comprises, from N end to C end, an HMCES protein, a bifunctional deaminase TadDual, an nCas9(D10A) protein and an engineered N-methyl purine DNA glycosylase eMPG. The base editor can realize simultaneous replacement of purine and pyrimidine, significantly increases the diversity of mutants after targeted editing, has the advantages of high editing efficiency and low Indels rate, and has great application value in aspects of in vitro and in vivo protein mutation screening and identification of oncogenic amino acid mutations.

Description

Technical Field

[0001] This invention relates to the field of gene editing technology, and more specifically to ACGBEmax, a base editor that simultaneously achieves purine and pyrimidine substitutions. Background Technology

[0002] With the development of high-throughput genome sequencing technology, numerous human genetic diseases have been confirmed to be related to gene mutations. Similarly, productive traits in livestock and crops are also affected by gene mutations. However, the relationship between most mutations and their associated phenotypes remains poorly understood. Importantly, in situ mutation induction of target genes using gene editing technology has become a powerful method for elucidating these genotype-phenotype relationships. Furthermore, in directed protein evolution, introducing mutations into target proteins is a crucial step in creating a library of protein mutants from which superior variants can be screened. The use of gene editing tools allows for the efficient introduction of mutations into target proteins in mammalian cells, making directed protein evolution possible in mammalian cells.

[0003] CRISPR / Cas9-based gene editing technologies typically induce insertions or deletions, which often result in loss of gene function, making them unsuitable for studying the effects of specific nucleotide mutations. While homologous recombination (by binding Cas9 and donor DNA) can theoretically introduce desired mutations, its low efficiency limits its practical application in inducing precise nucleotide mutations. Recently, base editors (BEs) and prime editors (PEs) have emerged as powerful tools capable of inducing base substitutions at target genomic sites in mammalian cells, proving invaluable for generating saturated mutant libraries of SNVs and MNVs.

[0004] Cytosine base editors (CBEs) and adenine base editors (ABEs) are two major single-base editing technologies. CBEs convert CG to TA, while ABEs convert AT to GC. These editors induce base transitions directly through deamination reactions without causing double-strand breaks (DSBs). However, CBEs and ABEs are limited to single-base transitions, which restricts their application in site-directed saturation mutagenesis. To expand the diversity of base mutations, dual-base editors have been developed by fusing cytosine and adenine deaminases with nCas9 (D10A), enabling simultaneous C-to-T and A-to-G transitions via a single sgRNA. These editors, such as A&C-BEmax, SPACE, and STEME, broaden the scope of base editing. However, both single-base and dual-base editors remain limited to conversion mutations and cannot induce transversion mutations, further restricting mutation diversity.

[0005] To overcome this limitation, engineered glycosylation enzymes (such as uracil DNA glycosylation enzyme, UNG, and N-methylpurine DNA glycosylation enzyme, MPG) have been integrated into base editors, enabling the induction of transversion mutations such as CG or AY, thereby expanding the mutation spectrum. Furthermore, researchers have developed deaminase-independent base editors, such as fusing engineered UNG mutants with nCas9 (D10A) to induce substitution and transversion mutations at target T and C sites. Another deaminase-independent guanine base editor (gGBE) has also been developed by fusing nCas9 (D10A) with an engineered MPG variant to achieve editing at the target G site. Despite these advances, existing base editors still cannot meet the demands of saturation mutations. Summary of the Invention

[0006] To address the aforementioned technical problems, the present invention aims to provide a base editor, ACGBEmax, that simultaneously achieves purine and pyrimidine substitutions. This base editor can simultaneously achieve purine and pyrimidine substitutions, significantly increasing the diversity of mutants after targeted editing. It also has the advantages of high editing efficiency and low indels rate, and has significant application value in in vitro and in vivo protein mutation screening and identification of oncogenic amino acid mutations.

[0007] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: A base editor ACGBEmax that can simultaneously realize purine and pyrimidine substitution is provided. The base editor ACGBEmax includes, from the N-terminus to the C-terminus, HMCES protein, bifunctional deaminase TadDual, nCas9(D10A) protein and engineered N-methylpurine DNA glycosylase eMPG.

[0008] Furthermore, the nucleotide sequence encoding the HMCES protein is shown in SEQ ID NO.1.

[0009] Furthermore, the nucleotide sequence encoding the bifunctional deaminase TadDual is shown in SEQ ID NO.2.

[0010] Furthermore, the nucleotide sequence encoding the nCas9(D10A) protein is shown in SEQ ID NO.3.

[0011] Furthermore, the nucleotide sequence encoding the engineered N-methylpurine DNA glycosylation enzyme eMPG is shown in SEQ ID NO.4.

[0012] The present invention also provides a gene encoding the above-mentioned base editor ACGBEmax, the nucleotide sequence of which is shown in SEQ ID NO.5.

[0013] The present invention also provides an expression vector comprising the above-mentioned genes.

[0014] The present invention also provides a host cell comprising the above-described expression vector.

[0015] This invention also provides the application of the above-mentioned base editor ACGBEmax, genes, expression vectors, or host cells in gene editing.

[0016] The present invention has the following beneficial effects:

[0017] 1. The base editor of this invention can efficiently induce A, C and G mutations in vitro and in vivo, and has the ability to generate diverse base substitution mutations at the editing site.

[0018] 2. ACGBEmax is a concept that can effectively mediate random mutations in endogenous genes, generate diverse protein mutant libraries in situ for functional or phenotypic screening, and has great potential in the analysis of oncogenic amino acid mutations. Attached Figure Description

[0019] Figure 1 A schematic diagram illustrating the design and optimization of the adenine / cytosine / guanine base editor (ACGBE);

[0020] Figure 2 Distribution of base editing results for genomic sites HPRT, PDCD1, CIITA, VEGFA, and DNMT3A;

[0021] Figure 3 Distribution of base editing results at genomic sites B2M, ACE2, and PCSK9;

[0022] Figure 4 The graph shows the editing efficiency statistics for ACGBEv1, ACGBEv2, and ACGBEv3.

[0023] Figure 5 The statistical results of the Indels rate for ACGBEv1, ACGBEv2, and ACGBEv3 are shown in the figure.

[0024] Figure 6 The graph shows the editing efficiency statistics for ACGBEv3, ACGBEv4, and ACGBEv5.

[0025] Figure 7 Graphs showing the Indels rate statistics for ACGBEv3, ACGBEv4, and ACGBEv5;

[0026] Figure 8 Figure showing the characterization results of the base editing activity of ACGBEmax;

[0027] Figure 9A graph showing the statistical results of the number of mutations at HPRT-site2 mutation sites at the DNA level;

[0028] Figure 10 This is a graph showing the statistical results of the number of mutations at HPRT-site2 mutation sites at the amino acid level.

[0029] Figure 11 The figure shows the results of the orthogonal R-loop analysis method for evaluating the Cas9-independent off-target effects.

[0030] Figure 12 Figure showing the off-target effects of ACGBEmax as assessed by deep RNA sequencing;

[0031] Figure 13 This is a graph showing the results of the fold change detection of amino acid mutations;

[0032] Figure 14 The graph shows the statistical results of the incidence of liver tumors in experimental mice induced by ACGBEmax.

[0033] Figure 15 The image shows the results of immunohistochemical (IHC) analysis.

[0034] Figure 16 This is a figure showing the results of next-generation sequencing (NGS) of mouse liver tumor tissue. Detailed Implementation

[0035] The principles and features of this invention are described below. The examples given are for illustrative purposes only and are not intended to limit the scope of the invention. Unless otherwise specified in the examples, conventional conditions or conditions recommended by the manufacturer should be followed. Reagents or instruments whose manufacturers are not specified are all commercially available products.

[0036] Example 1: Design and Optimization of ACGBEs

[0037] A schematic diagram of the design and optimization of the adenine / cytosine / guanine base editor (ACGBE) is shown below. Figure 1 As shown.

[0038] (1) To achieve simultaneous editing of A, C, and G bases, cytosine deaminase hA3A (Y130F) and adenine deaminase TadA8e were first fused to the N-terminus of nCas9 (D10A), while an engineered guanine glycosylation enzyme (eMPG) was linked to the C-terminus of nCas9 (D10A), thus constructing ACGBEv1. Subsequently, hA3A (Y130F) was replaced with TadA-derived TadCDd, thus constructing ACGBEv2. To obtain a more compact protein structure, cytosine and adenine deaminases were replaced with TadA variant-derived bifunctional deaminase TadDual, resulting in ACGBEv3.

[0039] To compare these three ACGBE versions, their editing activity at multiple endogenous genomic sites was evaluated in HEK293T cells. Plasmids encoding ACGBEs and sgRNA expression vectors were transiently transfected into cells, and base mutations at target sites were analyzed 48 hours post-transfection. Results are as follows: Figure 2 and 3 As shown, in the eight genomic sites tested (HPRT, PDCD1, CIITA, VEGFA, DNMT3A, B2M, ACE2, and PCSK9), all three base editors efficiently edited A, C, and G bases within the editing window. Notably, since none of these three ACGBEs contain a uracil DNA glycosylase inhibitor (UGI), the uracil generated from cytosine deamination was subsequently removed by endogenous UNG, forming base-free sites (AP sites). These AP sites were repaired via the base excision repair (BER) pathway, resulting in cytosine-to-thymine (C to T) mutations, and occasionally cytosine-to-guanine (CG) and a small number of cytosine-to-adenine (CA) conversions. Similarly, eMPG-mediated guanine (G) removal activated the BER pathway, inducing G to other base mutations. Unexpectedly, eMPG also exhibited some hypoxanthine (I) glycosylase activity, leading to A to other base mutations.

[0040] Furthermore, all three versions of ACGBEs produced a certain proportion of insertions and deletions (indels). This is likely because glycosylation enzymes and Cas9 (D10A) cleavage enzymes inevitably generate both AP sites and cleavage sites simultaneously in both the non-target and target DNA strands, leading to the formation of double-strand breaks (DSBs). It can be observed that these three ACGBEs exhibit different editing efficiencies and patterns for different bases within the editing window.

[0041] The editing efficiency and indels rate statistics for the three ACGBEs are as follows: Figure 4 and 5 As shown, when considering the overall editing of eight endogenous genomic sites, ACGBEv1, ACGBEv2, and ACGBEv3 have similar editing window sizes, but the total base editing efficiency of ACGBEv1 (average 70.81%) is significantly lower than that of ACGBEv2 (average 75.07%) and ACGBEv3 (average 73.26%). Furthermore, the indel rate of ACGBEv1 (average 11.76%) is significantly higher than that of ACGBEv2 (average 6.93%) and ACGBEv3 (7.76%). The base editing efficiency and indel rate of ACGBEv2 and ACGBEv3 are comparable. Given its more compact structure, we chose ACGBEv3 for further optimization.

[0042] (2) Although the above-mentioned ACGBEs can achieve various base substitution mutations, the indels they generate may lead to frameshift mutations, hindering the generation of effective variants and thus affecting the screening effect. In order to reduce indels, nCas9 (D10A) in ACGBEv3 was replaced with inactivated Cas9 (dCas9), generating ACGBEv4. The AP site protection protein HMCES was also introduced into ACGBEv3, generating ACGBEv5 (i.e., ACGBEmax).

[0043] Then, the base editing capabilities of ACGBEv4 and ACGBEv5 at the same eight endogenous genomic sites were further tested in HEK293T cells. The editing efficiency and indel rate statistics are as follows: Figure 6 and 7 As shown in the figure, ACGBEv4 effectively induced A, C, and G mutations, but significantly reduced the indel rate (1.81%) at the cost of reduced editing efficiency (45.16%). In contrast, ACGBEv5 had editing efficiency comparable to ACGBEv3 (70.06%), while significantly reducing the indel rate (5.81%). Therefore, ACGBEmax was chosen for further characterization.

[0044] (3) The base editing activity of ACGBEmax was characterized in HEK293T cells using ten additional endogenous genomic sites, and the results are as follows: Figure 8 As shown, the results demonstrate its efficient ability to generate diverse base substitution mutations at different sites. For A bases, A to G mutations were the most frequent (0.60% to 69.53%), followed by A to C (0.17% to 22.92%) and A to T mutations (0.11% to 7.84%). For C bases, C to T mutations were dominant (1.85% to 36.35%), followed by C to G (0.26% to 19.03%), while C to A mutations were less common (0.11% to 6.02%). For G bases, G to C (0.20% to 32.98%) and G to T (0.35% to 15.05%) were the most common mutations, while G to A mutations were less frequent (0.21% to 6.22%).

[0045] Furthermore, ACGBEmax was evaluated in two human cancer cell lines, HeLa and Hu7, and the results also showed that it could effectively induce A, C, and G mutations. These results indicate that ACGBEmax can generate diverse base substitution mutations at multiple endogenous genomic sites and has a certain preference for specific mutation types.

[0046] Example 2: ACGBEmax produces a greater variety of mutations than other base editors.

[0047] HPRT-site2 mutation sites were selected in HEK293T cells, and the base editing results of ACGBEmax were compared with those of traditional CBE, ABE, and ACBE. The statistical results of the number of mutations at the DNA and amino acid levels are as follows: Figure 9 and 10 As shown.

[0048] The results showed that at the DNA level, ACGBEmax induced a higher number of mutation types (25-42 mutation types per sgRNA) than ABE (10-18), CBE (7-25), and ACBE (7-32). Similarly, at the amino acid level, ACGBEmax generated 10-35 mutation types per sgRNA, far exceeding ABE (5-18), CBE (2-24), and ACBE (2-12). This indicates that ACGBEmax can generate a richer variety of amino acid mutations in endogenous protein-coding genes, which is advantageous for protein mutation screening.

[0049] Example 3: ACGBEmax induces low levels of Cas9-dependent and non-Cas9-dependent off-target editing.

[0050] To assess the potential off-target effects of ACGBEmax, we first used Cas-OFFinder software to predict possible Cas9-dependent off-target sites for three sgRNAs: PDCD1, VEGFA, and HPRT. Next-generation sequencing (NGS) analysis revealed no off-target effects for ACGBEmax except for the HPRT-OT4 site, and its off-target effects were significantly lower than those for CBE, ABE, and ACBE. We also used orthogonal R-loop analysis to assess Cas9-independent off-target effects, with results as follows: Figure 11 As shown, among the six R-loops generated by inactivated SaCas9, ACGBEmax exhibited significantly lower off-target effects than ABE, CBE, and ACBE. Furthermore, we assessed the off-target effects of ACGBEmax at the RNA level using deep RNA sequencing, with results as follows: Figure 12 As shown in the figure. The results showed that the number of A to I mutations induced by ACGBEmax was slightly higher than that in the control group, but the number of C to U mutations was comparable to that in the control group.

[0051] These results demonstrate that ACGBEmax achieves highly efficient targeted editing while minimizing off-target effects at both the DNA and RNA levels.

[0052] Example 4: ACGBEmax-mediated random mutations in endogenous genes for phenotypic screening

[0053] To validate the potential of ACGBEmax in generating saturated protein mutant libraries for functional screening studies, we introduced and screened for in situ mutations at the HPRT gene locus in human HHL-5 cells using ACGBEmax. Some single-amino acid mutations in the HPRT protein can confer resistance to 6-thioguanine (6TG). Therefore, we established a proof-of-concept mutation screening experiment using 6TG resistance as a readout metric to evaluate the HPRT mutant library generated by ACGBEmax.

[0054] First, we assessed the sensitivity of HHL-5 cells to 6TG, demonstrating that 100 μM 6TG was sufficient to induce cell death. Subsequently, we designed 33 sgRNAs targeting exons 3-8 of the HPRT gene for evaluation. Due to read length limitations in next-generation sequencing and to eliminate potential linkage effects between different exon mutations, we co-transfected cells with sgRNAs targeting different exons (forward and reverse) and then mixed the cells. After selecting successfully transfected cells with puromycin, 100 μM 6TG was added for further selection. Monoclonal formation was observed in the HPRT mutant cell population compared to the control group. After 6TG selection, next-generation sequencing (NGS) analysis of exons 3, 4, 6, 7, and 8 of the HPRT gene in surviving cells identified the corresponding HPRT mutants. These 6TG-selected cells were compared with cells treated with DMSO to calculate the fold change in HPRT protein mutants generated by ACGBEmax. Results are as follows: Figure 13 As shown, compared with the untreated group, cells screened with 6TG showed mutation enrichment in Exon3 and Exon8, among which p.C66R, p.G70K, p.G70W, p.G190R and p.G190A were recurring high-frequency mutations.

[0055] To validate truly enriched mutations, we constructed vectors expressing mutant and wild-type HPRT and transfected them into HPRT knockout (KO) HHL-5 cell lines for 6TG sensitivity analysis. The results showed that HPRT-KO cells and cells supplemented with the corresponding mutants were fully tolerant to 6TG treatment, while cells expressing wild-type HPRT died in 6TG medium. These findings demonstrate the concept of ACGBEmax as an effective tool for in situ generation of diverse protein mutant libraries for functional screening.

[0056] Example 5: Achieving in vivo mutations of the endogenous gene ACGBEmax in mouse liver to identify oncogenic amino acid mutations in the CTNNB1 protein.

[0057] Next, to verify the effectiveness of ACGBEmax in in vivo protein mutation and screening, CTNNB1 was selected as the target gene for in situ mutation of ACGBEmax at the single-base level, with the aim of identifying novel oncogenic mutations in CTNNB1 that may promote liver tumor formation in mice. To this end, nine forward and eleven reverse sgRNAs were designed to target key functional domains in exon 3 of CTNNB1. To induce liver tumor formation, ACGBEmax and the CTNNB1-targeting sgRNA plasmids were hydrodynamically injected into wild-type mice via the tail vein, along with a c-Myc overexpression plasmid. To avoid double-strand breaks (DSBs), the CTNNB1-targeting sgRNAs were divided into two groups: one group containing forward sgRNAs and the other containing reverse sgRNAs. Furthermore, to control for potential interference from background c-Myc expression, a non-targeting sgRNA was included as a control group.

[0058] Thirty days after injection, mice underwent necropsy. No liver tumors were observed in the control group. In contrast, 2 out of 10 mice injected with positive sgRNA developed liver tumors, and 3 out of 10 mice injected with reverse sgRNA also developed tumors. Extended observation (up to 60 days) showed no tumor formation in the control group, indicating that overexpression of c-Myc alone was insufficient to induce liver tumor formation within this timeframe. In contrast, as... Figure 14 As shown, ACGBEmax-induced CTNNB1 mutations significantly increased the incidence of liver tumors in experimental mice.

[0059] Fluorescence microscopy analysis revealed strong and uniform green fluorescence in the tumor tissue of the experimental mice, while only weak fluorescence was observed in the liver of the control mice. This indicates that ACGBEmax-induced CTNNB1 mutations are crucial for tumor formation. Furthermore, immunohistochemical (IHC) analysis results are as follows: Figure 15 As shown, CTNNB1 expression was significantly increased in tumor tissue, and CTNNB1 accumulated in the cell nucleus, which is characteristic of activating CTNNB1 mutations.

[0060] Furthermore, the next-generation sequencing (NGS) results of mouse liver tumor tissue are as follows: Figure 16As shown in the figure, the results revealed that several mutations in CTNNB1 were enriched in both the forward and reverse sgRNA genomes. The two most prevalent single-amino acid mutations were p.D32G (caused by an A-to-G mutation) and p.L63V (caused by a G-to-C mutation). CTNNB1 p.D32G is a known oncogenic mutation, while p.L63V has not been previously reported. Furthermore, complex mutant alleles involving point mutations and deletions were also enriched in the tumor. These results highlight the effectiveness of ACGBEmax in inducing in vivo mutations in endogenous genes and further confirm its significant application value in identifying oncogenic mutations.

[0061] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions or improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A base editor, ACGBEmax, that simultaneously performs purine and pyrimidine substitutions, characterized in that, The base editor ACGBEmax comprises, from N-terminus to C-terminus, HMCES protein, bifunctional deaminase TadDual, nCas9(D10A) protein, and engineered N-methylpurine DNA glycosylase eMPG. The nucleotide sequence encoding the HMCES protein is shown in SEQ ID NO.1; The nucleotide sequence encoding the bifunctional deaminase TadDual is shown in SEQ ID NO.2; The nucleotide sequence encoding the nCas9(D10A) protein is shown in SEQ ID NO.3; The nucleotide sequence encoding the engineered N-methylpurine DNA glycosylation enzyme eMPG is shown in SEQ ID NO.

4.

2. A gene encoding the base editor ACGBEmax as described in claim 1, characterized in that, Its nucleotide sequence is shown in SEQ ID NO.

5.

3. An expression vector comprising the gene of claim 2.

4. A host cell comprising the expression vector of claim 3.

5. The use of the base editor ACGBEmax of claim 1, the gene of claim 2, the expression vector of claim 3, or the host cell of claim 4 in gene editing for non-diagnostic or therapeutic purposes.

Images