FUSION PROTEIN THAT IMPROVES GENE EDITING EFFICIENCY AND APPLICATION THEREOF

Abstract

Provided are a fusion protein that improves gene editing efficiency and an application thereof. The fusion protein comprises a single-stranded DNA binding protein functional domain, nucleoside deaminase and nuclease. According to CBEs, when carrying our base conversion from C-G to T-A, nucleoside deaminase such as cytosine deaminase carries out deamination by using single-stranded DNA as a substrate, and by re-fusing the single-stranded DNA binding protein functional domain on the fusion protein of the nucleoside deaminase and nuclease, the chance of single-stranded DNA being exposed to the nucleoside deaminase is greatly increased, thereby significantly improving base editing efficiency. The present disclosure provides a breakthrough improvement of single-base gene editing technology and can greatly promote the application thereof in aspects such as gene editing, gene therapy, cell therapy, animal model making, and crop genetic breeding.

Claims

1. A fusion protein for improving gene editing efficiency, comprising functional domain of a single-stranded DNA binding protein, nucleoside deaminase and nuclease.

2. The fusion protein according to claim 1, wherein the order for the connection of the fusion protein is: the nucleoside deaminase is located at N-terminal or C-terminal of the nuclease, and the functional domain of the single-stranded DNA binding protein is located at the N-terminal or C-terminal of the nucleoside deaminase and the nuclease, or between the nucleoside deaminase and the nuclease.

3. The fusion protein according to claim 2, wherein the nucleoside deaminase is located at the N-terminal of the nuclease.

4. The fusion protein according to claim 3, wherein the functional domain of the single-stranded DNA binding protein is located between the nucleoside deaminase and the nuclease.

5. The fusion protein according to claim 1, wherein the single-stranded DNA binding protein comprises sequence-specific single-stranded DNA binding protein or non-sequence-specific single-stranded DNA binding protein.

6. The fusion protein according to claim 5, wherein the non-sequence-specific single-stranded DNA binding protein is selected from one or more of RPA70, RPA32, BRCA2, hnRNPK, PUF60 and Rad51; and the sequence-specific single-stranded DNA binding protein is selected from one or more of TEBP, Teb 1 and POT1.

7. The fusion protein according to claim 5, wherein the functional domain of the single-stranded DNA binding protein comprises at least one of the following four domains, or partial polypeptide fragments of the following four domains having a function of binding to single-stranded DNA, and any combination thereof: OB fold, KH domain, RRMS and whirly domain of the single-stranded DNA binding protein.

8. The fusion protein according to claim 5, wherein the functional domain of the single-stranded DNA binding protein comprises DNA binding domain of Rad51 or DNA binding domain of RPA70.

9. The fusion protein according to claim 8, wherein amino acid sequence of the DNA binding domain of Rad51 comprises a sequence of SEQ ID NO: 1; or, coding sequence of the DNA binding domain of Rad51 comprises a sequence of SEQ ID NO: 2.

10. The fusion protein according to claim 8, wherein amino acid sequence of the DNA binding domain of RPA70 comprises a sequence of SEQ ID NO: 11; or, coding sequence of the DNA binding domain of RPA70 comprises a sequence of SEQ ID NO: 12.

11. The fusion protein according to claim 1, wherein the deaminase comprises cytosine deaminase or adenosine deaminase.

12. The fusion protein according to claim 11, wherein the cytosine deaminase comprises rat-derived cytosine deaminase.

13. The fusion protein according to claim 12, wherein amino acid sequence of the rat-derived cytosine deaminase comprises a sequence of SEQ ID NO: 3; or, coding sequence of the rat-derived cytosine deaminase comprises a sequence of SEQ ID NO: 4.

14. The fusion protein according to claim 11, wherein the cytosine deaminase comprises human-derived cytosine deaminase APOBEC3A.

15. The fusion protein according to claim 14, wherein amino acid sequence of the human-derived cytosine deaminase APOBEC3A comprises a sequence of SEQ ID NO: 13; or, coding sequence of the human-derived cytosine deaminase APOBEC3A comprises a sequence of SEQ ID NO: 14.

16. The fusion protein according to claim 11, wherein the cytosine deaminase comprises mutant of the cytosine deaminase APOBEC3A, and the mutant mutates asparagine at position 57 of the cytosine deaminase APOBEC3A into glycine.

17. The fusion protein according to claim 16, wherein the cytosine deaminase APOBEC3A is derived from human.

18. The fusion protein according to claim 17, wherein amino acid sequence of the cytosine deaminase APOBEC3A comprises the sequence of SEQ ID NO: 13; or, coding sequence of the cytosine deaminase APOBEC3A comprises the sequence of SEQ ID NO: 14.

19. The fusion protein according to claim 16, wherein amino acid sequence of the mutant of the cytosine deaminase APOBEC3A comprises a sequence of SEQ ID NO: 15; or, coding sequence of the cytosine deaminase APOBEC3A comprises a sequence of SEQ ID NO: 16.

20. The fusion protein according to claim 1, wherein the nuclease is selected from one or more of Cas9, Cas3, Cas8a, Cas8b, Cas10d, Cse1, Csy1, Csn2, Cas4, Cas10, Csm2, Cmr5, Fok1 and Cpf1.

21. The fusion protein according to claim 20, wherein the nuclease is Cas9.

22. The fusion protein according to claim 21, wherein the Cas9 is selected from Cas9 derived from Streptococcus pneumoniae, Staphylococcus aureus, Streptococcus pyogenes or Streptococcus thermophilus.

23. The fusion protein according to claim 21, wherein the Cas9 is selected from Cas9 mutants VQR-spCas9, VRER-spCas9 or spCas9n.

24. The fusion protein according to claim 23, wherein amino acid sequence of spCas9n comprises a sequence of SEQ ID NO: 5; or, coding sequence of the spCas9n comprises a sequence of SEQ ID NO: 6.

25. The fusion protein according to claim 1, wherein the fusion protein further comprises NLS.

26. The fusion protein according to claim 25, wherein the NLS is located at at least one terminal of the fusion protein.

27. The fusion protein according to claim 25, wherein amino acid sequence of the NLS comprises a sequence of SEQ ID NO: 7; or, coding sequence of the NLS comprises a sequence of SEQ ID NO: 8.

28. The fusion protein according to claim 1, wherein the fusion protein further comprises more than two copies of UGI.

29. The fusion protein according to claim 28, wherein the UGI is located at at least one terminal of the fusion protein.

30. The fusion protein according to claim 28, wherein amino acid sequence of the UGI comprises a sequence of SEQ ID NO: 9; or, coding sequence of the UGI comprises a sequence of SEQ ID NO: 10.

31. Any one of the following biomaterials of A)-C): A) a gene encoding the fusion protein according to claim 1; B) a recombinant vector containing the gene of A); C) a recombinant cell or recombinant bacterium containing the fusion protein according to claim 1 or containing the gene of A).

32. The biomaterial according to claim 31, wherein the cells are T cells, hematopoietic stem cells, bone marrow cells, red blood cells or red blood cell precursor cells.

33. An sgRNA for gene editing of a target gene in cells, wherein target sequence of the sgRNA comprises at least one of SEQ ID NO: 17-36.

34. The sgRNA according to claim 33, wherein the cells are T cells, hematopoietic stem cells, bone marrow cells, red blood cells or red blood cell precursor cells.

35. The sgRNA according to claim 33, wherein the target gene is at the promotor of HBG1 or HBG2.

36. A single-base gene editing system, wherein the system comprises the fusion protein according to claim 1, and the sgRNA, and the sgRNA guides the fusion protein to conduct single-base gene editing on target gene in target cell.

37. The single-base gene editing system according to claim 36, wherein target sequence of the sgRNA comprises at least one of SEQ ID NO: 17-36; or, the cells are T cells, hematopoietic stem cells, bone marrow cells, red blood cells or a red blood cell precursor cells, or, the target gene is at the promotor of HBG1 or HBG2.

38. A method for preparing a product for gene editing, treating or preventing disease, animal model or new plant variety, which comprises using the fusion protein according to claim 1.

39. The method according to claim 38, wherein the disease is beta-hemoglobinopathy, and the beta-hemoglobinopathy comprises beta-thalassemia or sickle cell anemia.

40. A method for improving the efficiency of single-base gene editing, which comprises the steps of introduction the fusion protein according to claim 1 and the sgRNA into cell to edit target gene, wherein the sgRNA guides the fusion protein to conduct single-base gene editing on target gene in target cell.

41. The method according to claim 40, wherein the target sequence of the sgRNA comprises at least one of SEQ ID NO: 17-36; or, the cells are T cells, hematopoietic stem cells, bone marrow cells, red blood cells or red blood cell precursor cells; or, the target gene is at the promotor of HBG1 or HBG2.

42. A method for constructing animal models of disease, which comprises the steps of introduction the fusion protein according to claim 1 and the sgRNA into animal cells to conduct gene editing of the target gene.

43. The method according to claim 42, wherein the target sequence of the sgRNA comprises at least one of SEQ ID NO: 17-36.

44. The method according to claim 43, wherein the target sequence of the sgRNA comprises the sequence of SEQ ID NO: 36, and the target gene comprises a DMD gene.

45. The method according to claim 42, wherein the animals are mammals, or, the cells are embryonic cells, or, the method of introduction is one or any combination of vector transformation, microinjection, transfection, lipid transfection, heat shock, electroporation, transduction, gene gun, and DEAE-dextran mediated transfer, or, the introduction is carried out using mRNA of the fusion protein according to claim 1 and the sgRNA.

46. The method according to claim 45, wherein the animals are rats or mice; or, when the method of introduction is microinjection, the concentration of the mRNA of the fusion protein according to claim 1 for introduction is 1-1,000 ng/μL.

47. The method according to claim 46, wherein the concentration ratio of the mRNA of the fusion protein according to claim 1 to the sgRNA used for introduction is 1:(5-1).

48. A method for drug screening, evaluation of therapeutic effects of disease or research on treatment mechanism of disease, which comprises using the animal model obtained by the method according to claim 42.

49. A product for treating or preventing beta-hemoglobinopathy, which comprises: delivery vector of the gene in A) according to claim 31 and the sgRNA, the sgRNA guides the fusion protein to conduct single-base gene editing on the target gene in the target cell; and the target gene is at the promotor of HBG1 or HBG2.

50. The product according to claim 49, wherein the target sequence of the sgRNA comprises a sequence of SEQ ID NO: 35; or, the beta-hemoglobinopathy comprises beta thalassemia or sickle cell anemia; or the cells are T cells, hematopoietic stem cells, bone marrow cells, red blood cells or red blood cell precursor cells.

51. The product according to claim 49, wherein the delivery vector comprises a viral vector or a non-viral vector; the viral vector comprises adeno-associated viral vector, adenoviral vector, lentiviral vector, retroviral vector or oncolytic virus vector; and the non-viral vector comprises cationic high-molecular polymer or liposome.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0068] FIG. 1 shows a structural diagram of the fusion of the functional domains of different single-stranded DNA binding proteins with BE4max. Wherein NLS is a nuclear localization signal (its amino acid sequence is SEQ ID NO: 7 and coding sequence is SEQ ID NO: 8); rA1 is cytidine deaminase APOBEC1 (its amino acid sequence is SEQ ID NO: 3 and coding sequence is SEQ ID NO: 4); SpCas9n is a Cas9n derived from Streptococcus pyogenes (its amino acid sequence is SEQ ID NO: 5 and coding sequence is SEQ ID NO: 6); UGI is a uracil glycosidase inhibitor (its amino acid sequence is SEQ ID NO: 9 and coding sequence is SEQ ID NO: 10); and SSDBD is the functional domain of a single-stranded DNA binding protein.

[0069] FIG. 2 shows the comparison of C-to-T base editing efficiency (i.e., ordinate, in %) achieved by hyBE4max and BE4max at 8 targets on 293T.

[0070] FIG. 3 shows the comparison of average C-to-T base editing efficiency (i.e., ordinate, in %) generated by hyBE4max and BE4max at 8 targets on 293T.

[0071] FIG. 4 shows the comparison of base editing efficiency (i.e., ordinate, in %) of indels generated by hyBE4max and BE4max at 8 targets on 293T.

[0072] FIG. 5 shows a structural diagram of fusion proteins A3A-BE4max and hyA3A-BE4max, wherein hA3A is a human-derived cytidine deaminase APOBEC3A (its amino acid sequence is SEQ ID NO: 13 and coding sequence is SEQ ID NO: 14); and NLS, spCas9n and UGI were the same as those in FIG. 1.

[0073] FIG. 6 shows the comparison of C-to-T base editing efficiency (i.e., ordinate, in %) achieved by hyA3A-BE4max and A3A-BE4max at 8 endogenous targets on 293T.

[0074] FIG. 7 shows the comparison of average C-to-T base editing efficiency (i.e., ordinate, in %) achieved by hyA3A-BE4max and A3A-BE4max at 8 endogenous targets on 293T.

[0075] FIG. 8 shows the comparison of base editing efficiency (i.e., ordinate, in %) of indels generated by hyA3A-BE4max and A3A-BE4max at 8 endogenous targets on 293T.

[0076] FIG. 9 shows a structural diagram of fusion proteins eA3A-BE4max and hyeA3A-BE4max. Wherein A3A N57G is an N57G mutant of hA3A used in FIG. 5 (its amino acid sequence is SEQ ID NO: 15 and coding sequence is SEQ ID NO: 16); and NLS, spCas9n and UGI are the same as those in FIG. 1.

[0077] FIG. 10 shows the comparison of C-to-T base editing efficiency (i.e., ordinate, in %) achieved by hyeA3A-BE4max and eA3A-BE4max at 11 endogenous targets on 293T.

[0078] FIG. 11 shows the comparison of average C-to-T base editing efficiency (i.e., ordinate, in %) achieved by hyeA3A-BE4max and eA3A-BE4max at 11 endogenous targets on 293T.

[0079] FIG. 12 shows the comparison of base editing efficiency (i.e., ordinate, in %) of indels generated by hyeA3A-BE4max and eA3A-BE4max at 11 endogenous targets on 293T.

[0080] Wherein C in the abscissa of FIGS. 2, 3, 5, 6, and 11 and the numbers following it represent the position of C edited as T on the corresponding target sequence, for example, C5 represents the efficiency of editing the C at position 5 from the 5′ end of the corresponding target sequence as T.

[0081] FIG. 13 shows a schematic diagram of hyeA3A-BE4max targeting HBG1/2 promoter region −117G, wherein −117 G>A mutation is shown in red, the core sequence of binding site of transcription factor BCL11A is indicated by a box, the PAM sequence is in blue, and G>A transformation destroys the binding site of transcription inhibitor BCL11A and activates the expression of HBG1/2 in HUDEP-2(Δ.sup.Gγ).

[0082] FIG. 14 shows the comparison of C-to-T base editing efficiency (i.e., ordinate, in %) achieved by hyeA3A-BE4max, eA3A-BE4max, A3A-BE4max, and hyA3A-BE4max targeting HBG-117G in HEK293T cells.

[0083] FIG. 15 shows a schematic diagram of construction of lentiviral vectors Lenti-117G-hyA3A-BE4max-P2A-GFP and Lenti-117G-hyeA3A-BE4max-P2A-GFP. Wherein, the target sequence of the sgRNA is HBG-117G.

[0084] FIG. 16 shows the comparison of C-to-T base editing efficiency (i.e., ordinate, in %) achieved by hyeA3A-BE4max and hyA3A-BE4max targeting HBG-117G in HUDEP-2 (Δ.sup.Gγ) cells.

[0085] FIG. 17 shows the comparison of globin mRNA expression in HUDEP-2 (Δ.sup.Gγ) cells infected with lentiviruses Lenti-117G-hyeA3A-BE4max-P2A-GFP and Lenti-117G-hyA3A-BE4max-P2A-GFP after cell differentiation. Wherein **** represents the difference significance level P<0.0001.

[0086] FIG. 18 shows a schematic diagram of animal model construction of hyA3A-BE4max targeting Duchenne muscular dystrophy (DMD) gene.

[0087] FIG. 19 shows the comparison of F0 high-throughput sequencing results after microinjection of A3A-BE4max and hyA3A-BE4max.

[0088] FIG. 20 shows the average ratio of Reads containing TAA stop codons in F0 produced by injection of A3A-BE4max and hyA3A-BE4max.

[0089] FIG. 21 shows the expression of dystrophin in F0 mice detected by immunofluorescence staining.

[0090] FIG. 22 shows the germ line inheritance (F0.fwdarw.F1) of DMD mutant mice.

[0091] FIG. 23 shows the off-target analysis of the predicted off-target site combination of hyA3A-BE4max and DMD-sg3 on the F0 generation.

DETAILED DESCRIPTION

1. The editing efficiency of BE4max fused with the functional domain of Rad51DBD (1-114aa) single-stranded DNA binding protein is improved most obviously

[0092] 1.1 Plasmid Design and Construction

[0093] 1.1.1 According to the characteristics that Apobec1 of CBEs in single-base editing technology uses single-stranded DNA as substrate, we designed 10 different functional domains of human-derived non-sequence-biased single-stranded DNA binding proteins (mainly RPA70 (630aa)-A, RPA70-B, RPA70-AB, RPA70-C, RPA32-D, BRCA2-OB2, BRCA2-OB3, HNRNPK KH domain, PUF60 RRM, Rad51 DBD) (Table 1); since the reported fusion protein tended to be inactive at the C-terminal of BE4max (the first figure from top to bottom in FIG. 1), these functional domains were thus fused at the N-terminal of BE4max (the second figure from top to bottom in FIG. 1), and meanwhile, two endogenous targets EMX1 site1 and Tim3-sg1 from human were designed (the sequences are shown in Table 2).

[0094] 1.1.2 DNAs of 10 different functional domains of human-derived non-sequence-biased single-stranded DNA binding proteins shown in Table 1 were synthesized, and then seamlessly cloned and assembled to the N-terminal of BE4max in plasmid pCMV-BE4max (addgene, #112093), and 10 recombinant plasmids were constructed respectively (FIG. 1): pRPA70-A-BE4max, pRPA70-B-BE4max, pRPA70-AB-BE4max, pRPA70-C-BE4max, pRPA32-D-BE4max, pBRCA2-OB2-BE4max, pBRCA2-OB3-BE4max, pKH-BE4max, pRRM-BE4max, and pRad51DBD-BE4max.

[0095] The DNAs of the targets EMX1 site1 and Tim3-sg1 shown in Table 2 were synthesized, and connected to Bbs I site of sgRNA expression plasmid U6-sgRNA-EF1α-GFP (used to express the sgRNA of the corresponding targets), respectively, to obtain recombinant plasmids pE and pT.

[0096] 1.1.3 The plasmids constructed in 1.1.1 and 1.1.2 were sequenced by sanger to ensure that they are completely correct.

TABLE-US-00001 TABLE 1 Different functional domain sequences of single-stranded DNA binding proteins used Name Sequence RPA70-A coding sequence (5′-3′): Cagtccaaagtggtgcccattgccagcctc actccttaccagtccaagtggaccatttgt gctcgtgttaccaacaaaagtcagatccgt acctggagcaactcccgaggggaagggaag cttttctccctagaactggttgacgaaagt ggtgaaatccgagctacagctttcaatgag caagtggacaagttctttcctcttattgaa gtgaacaaggtgtattatttctcgaaaggc accctgaagattgctaacaagcagttcaca gctgttaaaaatgactacgagatgaccttc aataacgagacttccgtcatgccctgtgag amino acid sequence: QSKVVPIASL TPYQSKWTICARVTNKSQIRTWSNSRGEGK LFSLELVDESGEIRATAFNEQVDKFFPLIE VNKVYYFSKGTLKIANKQFTAVKNDYEMTF NNETSVMPCE RPA70-B coding sequence (5′-3′): Gatttcacggggattgatgacctcgagaac aagtcgaaagactcacttgtagacatcatc gggatctgcaagagctatgaagacgccact aaaatcacagtgaggtctaacaacagagaa gttgccaagaggaatatctacttgatggac acatccgggaaggtggtgactgctacactg tggggggaagatgctgataaatttgatggt tctagacagcccgtgttggctatcaaagga gcccgagtctctgatttcggtggacggagc ctctccgtgctgtcttcaagcactatcatt gcgaatcctgacatcccagaggcctataag cttcgtggatggtttgacgcagaaggacaa gcctta amino acid sequence: DFTG IDDLENKSKDSLVDIIGICKSYEDATKITV RSNNREVAKRNIYLMDTSGKVVTATLWGED ADKFDGSRQPVLAIKGARVSDFGGRSLSVL SSSTIIANPDIPEAYKLRGWFDAEGQAL RPA70-AB encoding sequence (5′-3′): RPA 70-A + RPA70-B amino acid sequence: RPA70-A  RPA70-B RPA70-C coding sequence (5′-3′): Ggagggagtaacaccaactggaaaaccttg tatgaggtcaaatccgagaacctgggccaa ggcgacaagccggactactttagttctgtg gccacagtggtgtatcttcgcaaagagaac tgcatgtaccaagcctgcccgactcaggac tgcaataagaaagtgattgatcaacagaat ggattgtaccgctgtgagaagtgcgacacc gaatttcccaatttcaagtaccgcatgatc ctgtcagtaaatattgcagattttcaagag aatcagtgggtgacttgtttccaggagtct gctgaagctatccttggacaaaatgctgct tatcttggggaattaaaagacaagaatgaa caggcatttgaagaagttttccagaatgcc aacttccgatctttcatattcagagtcagg gtcaaagtggagacctacaacgacgagtct cgaattaaggccactgtgatggacgtgaag cccgtggactacagagagtatggccgaagg ctggtcatgagcatcaggagaagtgcattg atg (SEQ ID NO: 12) amino acid sequence: GGSNTNWKTLYEVKSENLGQGDKPDYFSSV ATVVYLRKENCMYQACPTQDCNKKVIDQQN GLYRCEKCDTEFPNFKYRMILSVNIADFQE NQWVTCFQESAEAILGQNAAYLGELKDKNE QAFEEVFQNANFRSFIFRVRVKVETYNDES RIKATVMDVKPVDYREYGRRLVMSIRRSAL M (SEQ ID NO: 11) RPA32-D coding sequence (5′-3′): Gcccagcacattgtgccctgtactatatct cagctgctttctgccactttggttgatgaa gtgttcagaattgggaatgttgagatttca caggtcactattgtggggatcatcagacat gcagagaaggctccaaccaacattgtttac aaaatagatgacatgacagctgcacccatg gacgttcgccagtgggttgacacagatgac accagcagtgaaaacactgtggttcctcca gaaacatatgtgaaagtggcaggccacctg agatcttttcagaacaaaaagagcctggta gcctttaagatcatgcccctggaggatatg aatgagttcaccacacatattctggaagtg atcaatgcacacatggtactaagcaaa amino acid sequence: AQHIVPCTISQLLSATLVDEVFRIGNVEIS QVTIVGIIRHAEKAPTNIVYKIDDMTAAPM DVRQWVDTDDTSSENTVVPPETYVKVAGHL RSFQNKKSLVAFKIMPLEDMNEFTTHILEV INAHMVLSK BRCA2- coding sequence (5′-3′): OB2 Ttatcatcgcttttcagtgatggaggaaat gttggttgtgttgatgtaattattcaaaga gcataccctatacagtggatggagaagaca tcatctggattatacatatttcgcaatgaa agagaggaagaaaaggaagcagcaaaatat gtggaggcccaacaaaagagactagaagcc ttattcactaaaattcaggaggaatttgaa gaacatgaagaaaacacaacaaaaccatat ttaccatcacgtgcactaacaagacagcaa gttcgtgctttgcaagatggtgcagagctt tatgaagcagtgaagaatgcagcagaccca gcttaccttgagggttatttcagtgaagag cagttaagagccttgaataatcacaggcaa atgttgaatgataagaaacaagctcagatc cagttggaaattaggaaggccatggaatct gctgaacaaaaggaacaaggtttatcaagg gatgtcacaaccgtgtggaagttgcgtatt gtaagctattcaaaaaaagaaaaagattca gttatactgagtatttggcgtccatcatca gatttatattctctgttaacagaaggaaag agatacagaatttatcatcttgcaacttca aaatctaaaagtaaatctgaaagagctaac atacagttagcagcgacaaaaaaaactcag tatcaacaactaccggtttcagatgaaatt ttatttcagatttaccagccacgggagccc amino acid sequence: LSSLFSDGGNVGCVDVIIQRAYPIQWMEKT SSGLYIFRNEREEEKEAAKYVEAQQKRLEA LFTKIQEEFEEHEENTTKPYLPSRALTRQQ VRALQDGAELYEAVKNAADPAYLEGYFSEE QLRALNNHRQMLNDKKQAQIQLEIRKAMES AEQKEQGLSRDVTTVWKLRIVSYSKKEKDS VILSIWRPSSDLYSLLTEGKRYRIYHLATS KSKSKSERANIQLAATKKTQYQQLPVSDEI LFQIYQPREP BRCA2- coding sequence (5′-3′): OB3 Gacctaataggatttgtcgtttctgttgtg aaaaaaacaggacttgcccctttcgtctat ttgtcagacgaatgttacaatttactggca ataaagttttggatagaccttaatgaggac attattaagcctcatatgttaattgctgca agcaacctccagtggcgaccagaatccaaa tcaggccttcttactttatttgctggagat ttttctgtgttttctgctagtccaaaagag ggccactttcaagagacattcaacaaaatg aaaaatactgttgag amino acid sequence: DLIGFVVSVVKKTGLAPFVYLSDECYNLLA IKFWIDLNEDIIKPHMLIAASNLQWRPESK SGLLTLFAGDFSVFSASPKEGHFQETFNKM KNTVE hnRNPK coding sequence (5′-3′): KH Aacactgatgagatggttgaattacgcatt domain ctgcttcagagcaagaatgctggggcagtg attggaaaaggaggcaagaatattaaggct ctccgtacagactacaatgccagtgtttca gtcccagacagcagtggccccgagcgcata ttgagtatcagtgctgatattgaaacaatt ggagaaattctgaagaaaatcatccctacc ttggaagag amino acid sequence: NTDEMVELRILLQSKNAGAVIGKGGKNIKA LRTDYNASVSVPDSSGPERILSISADIETI GEILKKIIPTLEE PUF60 coding sequence (5′-3′): RRM Tgccgcgtctacgtgggctctatctactat gagctgggggaggacaccatccgccaggcc tttgccccctttggccccatcaagagcatc gacatgtcctgggactccgtcaccatgaag cacaagggctttgccttcgtggagtatgag gtccccgaagctgcacagctggccttggag cagatgaactcggtgatgctggggggcagg aacatcaaggtgggcagacccagcaac amino acid sequence: CRVYVGSIYYELGEDTIRQAFAPFGPIKSI DMSWDSVTMKHKGFAFVEYEVPEAAQLALE QMNSVMLGGRNIKVGRPSN Rad51 encoding sequence (5′-3′): DBD Atggcaatgcagatgcagcttgaagcaaat gcagatacttcagtggaagaagaaagcttt ggcccacaacccatttcacggttagagcag tgtggcataaatgccaacgatgtgaagaaa ttggaagaagctggattccatactgtggag gctgttgcctatgcgccaaagaaggagcta ataaatattaagggaattagtgaagccaaa gctgataaaattctggctgaggcagctaaa ttagttccaatgggtttcaccactgcaact gaattccaccaaaggcggtcagagatcata cagattactactggctccaaagagcttgac aaactacttcaa (SEQ ID NO: 2) amino acid sequence: MAMQMQLEANADTSVEEESFGPQPISRLEQ CGINANDVKKLEEAGFHTVEAVAYAPKKEL INIKGISEAKADKILAEAAKLVPMGFTTAT EFHQRRSEIIQITTGSKELDKLLQ (SEQ  ID NO: 1)

TABLE-US-00002 TABLE 2 Targets and sequences used SEQ Name of ID targets Sequence (5′-3′) NO: EMXI site1 GAGTCCGAGCAGAAGAAGAAGGG 17 Tim3-sg1 TTCTACACCCCAGCCGCCCCAGG 18 VEGFA site2 GACCCCCTCCACCCCGCCTCCGG 19 Lag3-sg2 CGCTACACGGTGCTGAGCGTGGG 20 HEK3 GGCCCAGACTGAGCACGTGATGG 21 HEK4 GGCACTGCGGCTGGAGGTGGGGG 22 EMX1-sg2p GACATCGATGTCCTCCCCATTGG 23 Nme1-sg1 AGGGATCGTCTTTCAAGGCGAGG 24

[0097] 1.2. Cell Transfection

[0098] 5×10.sup.5 HEK293T cells were plated into a 24-well plate. When the cells grew to 70%-80%, the plasmid combinations were transfected according to pssDBD-BE4max:pE (or pT)=750 ng:250 ng. 3 replicate wells were set for transfection of each plasmid combination, with 2×10.sup.5 cells per well. Simultaneously, a blank control without transfection of any plasmid was set.

[0099] pssDBD-BE4max represents: any one of plasmids pRPA70-A-BE4max, pRPA70-B-BE4max, pRPA70-AB-BE4max, pRPA70-C-BE4max, pRPA32-D-BE4max, pBRCA2-OB2-BE4max, pBRCA2-OB3-BE4max, pKH-BE4max, pRRM-BE4max, and pRad51DBD-BE4max; and the plasmid pCMV-BE4max was used as a negative control.

[0100] 1.3. Genome Extraction and Preparation of Amplicon Library.

[0101] 72 h after transfection, genomic DNA of the cells was extracted using Tiangen Cell Genome Extraction Kit (DP304). Afterwards, the corresponding identification primers (Table 3) were designed according to the operation process of Hitom kit, i.e., a bridging sequence 5′-ggagtgagtacggtgtgc-3′ was added to the 5′ terminal of the forward identification primer, and a bridging sequence 5′-gagttggatgctggatgg-3′ was added to the 5′ terminal of the reverse identification primer to obtain one round of PCR products; then, the first round of PCR products was used as a template to conduct a second round of PCR, followed by the PCR products were mixed together for gel-cutting, recovering, purification, and then sent to a company for deep sequencing.

TABLE-US-00003 TABLE 3 Identification primers of the targets used Name of Sequence targets (5′-3′) EMX1 site1 F: ggagtgagtacggtgtgcGTGG TTCCAGAACCGGAGGACAAAG R: gagttggatgctggatggGTTT GTGGTTGCCCACCCTAGTCAT Tim3-sg1 F: ggagtgagtacggtgtgcCGCT TGAGTCTTGGCTCTCCTTCTC R: gagttggatgctggatggCACC ACGTTGCCACATTCAAACACA VEGFA F: ggagtgagtacggtgtgcGACA site2 GACAGACAGACACCGCCC R: gagttggatgctggatggACAG CCCAGAAGTTGGACGAAAAGT Lag3-sg2 F: ggagtgagtacggtgtgcTTCC TACCCCTGGAGCTTCTCAACT R: gagttggatgctggatggCCTC CGGGACCCACGCTCAG HEK3 F: ggagtgagtacggtgtgcAGGG AAACGCCCATGCAATTAGTCT R: gagttggatgctggatggCCCT GTCTAGGAAAAGCTGTCCTGC HEK4 F: ggagtgagtacggtgtgcCAGA GGGTCCAAAGCAGGATGACAG R: gagttggatgctggatggCTTT CAACCCGAACGGAGACACACA EMX1-sg2p F: ggagtgagtacggtgtgcGTGG TTCCAGAACCGGAGGACAAAG R: gagttggatgctggatggGTTG TGGTTGCCCACCCTAGTCAT Nme1-sg1 F: ggagtgagtacggtgtgcGGGG AGGCAGACACACAAACAGAAA R: gagttggatgctggatggGCGC TCATGACCTACCCTGTATCAC

[0102] 1.4. Analysis and Statistics of Deep Sequencing Results

[0103] The ratios of C to T and Indels were calculated by using the deep sequencing results of step 1.3 using a BE-analyzer website. The results were shown in Table 4 and Table 5.

[0104] The results show that, compared with BE4max, the BE4max fused with the functional domain of Rad51 single-stranded DNA binding protein (Rad51DBD-N-BE4max or Rad51DBD-BE4max) most obviously improved the C-to-T editing efficiency on the target, followed by the BE4max fused with the functional domain of a RPA70-C single-stranded DNA binding protein.

[0105] 2. Best Editing Efficiency of hyBE4max

[0106] In order to further test the fusion position of the functional domain of Rad51 single-stranded DNA binding protein with the highest improvement on the C-to-T editing efficiency on the targets in step 1, the Rad51 DBD was fused to two other different positions of BE4max, and three recombinant plasmids of BE4max fused with Rad51 DBD (the third to the fifth figures from top to bottom in FIG. 1) were transfected into cells along with recombinant plasmids pE or pT, respectively, according to the method of 1.2 in step 1, and the results of editing efficiency were obtained according to the methods of 1.3 and 1.4 in step 1 (Table 4 and Table 5).

[0107] Three types of BE4max fused with Rad51 DBD shown in the third to the fifth figures from top to bottom in FIG. 1 are as follows: [0108] Rad51DBD-N-BE4max: Rad51 DBD was fused between NLS and rA1 in BE4max, i.e., Rad51 DBD was located at the N-terminal of rA1 and spCas9n; [0109] Rad51DBD-C-BE4max: Rad51 DBD was fused between spCas9n and UGI in BE4max, i.e., Rad51 DBD was located at the C-terminal of rA1 and spCas9n; [0110] hyBE4max: Rad51 DBD was fused between rA1 and spCas9n in BE4max.

TABLE-US-00004 TABLE 4 Results of editing efficiency of target EMX1 site1 (in %) Base editor Repeat C5 C6 BE4max Rep. 1 46.7 45.2 Rep. 2 44.9 42.9 Rep. 3 62.5 60.7 RPA70-A-BE4max Rep. 1 41 36.9 Rep. 2 41 36.9 Rep. 3 41.9 37.2 RPA70-B-BE4max Rep. 1 42.7 36.2 Rep. 2 42.6 36.3 Rep. 3 41 33.3 RPA70-AB-BE4max Rep. 1 48.4 47 Rep. 2 47.1 45.5 Rep. 3 45.7 44.1 RPA70-C-BE4max Rep. 1 65.2 63 Rep. 2 64.9 62.8 Rep. 3 64.7 62 RPA32-D-BE4max Rep. 1 41.8 37.3 Rep. 2 41.3 36.9 Rep. 3 41 36.7 BRCA2-OB2-BE4max Rep. 1 30 20.6 Rep. 2 30 20.5 Rep. 3 31 20.8 BRCA2-OB3-BE4max Rep. 1 32.5 25.8 Rep. 2 32.4 25.8 Rep. 3 32.1 25 KH-BE4max Rep. 1 34.8 21.7 Rep. 2 34.6 21.6 Rep. 3 31.9 19.3 RRM-BE4max Rep. 1 34.8 25.2 Rep. 2 35.2 25.3 Rep. 3 34.7 24.3 Rad51DBD-BE4max Rep. 1 68 66.4 Rep. 2 67.7 66.2 Rep. 3 65.9 60.6 hyBE4max Rep. 1 80.8 75.1 Rep. 2 81.3 75.5 Rep. 3 70 64.5 BE4max-C-Rad51 Rep. 1 41.5 38.9 Rep. 2 42.3 39.8 Rep. 3 45.2 42.9 Blank control Rep. 1 0.1 0 Rep. 2 0 0 Rep. 3 0 0

TABLE-US-00005 TABLE 5 Results of editing efficiency of target Tim3-sg1 (in %) Base editor Repeat C3 C6 C8 C9 C10 C11 BE4max Rep. 1 24.1 25.2 21.3 17.9 5.6 2.5 Rep. 2 23.7 24.7 20.8 17.4 5.6 2.5 Rep. 3 24.3 25.9 21.4 17.9 5.2 2.5 RPA72-A-BE4max Rep. 1 17.7 17.7 11.9 10.5 7.1 4.7 Rep. 2 18 17.8 12 10.6 7.3 4.8 Rep. 3 17.9 17.6 12.3 10.7 7.5 5 RPA72-B-BE4max Rep. 1 15.4 14.2 9.9 8.3 5 2.9 Rep. 2 15.9 14.6 10.1 8.6 5.1 2.9 Rep. 3 15.1 14.3 9.4 7.8 4.7 2.7 RPA72-AB-BE4max Rep. 1 17.2 11.7 5.8 3.7 1.4 0.7 Rep. 2 17.3 11.6 5.8 3.7 1.4 0.6 Rep. 3 16.6 9.8 4.3 2.6 0.9 0.4 RPA70-C-BE4max Rep. 1 27.8 29.5 20.4 15.4 3.5 1.8 Rep. 2 28.1 29.4 20.9 15.7 3.8 1.9 Rep. 3 30.1 31.7 21.4 17 4.2 1.5 RPA32-D-BE4max Rep. 1 17.5 15.8 9.7 8.6 5.6 3.2 Rep. 2 17.9 16 9.9 8.8 5.8 3.3 Rep. 3 16.1 13.4 8.7 7.3 4.7 2.9 BRCA2-OB2-BE4max Rep. 1 9.2 9.3 7.9 7.3 6.2 5.6 Rep. 2 9.2 9.3 7.9 7.3 6.2 5.6 Rep. 3 8.8 9 7.4 7 5.6 4.7 BRCA2-OB3-BE4max Rep. 1 10.7 10.4 7.4 5.9 3.7 2.8 Rep. 2 13.1 10.1 8.2 7.7 4 2.6 Rep. 3 10.4 10 7.5 6.9 4.8 3.3 KH-BE4max Rep. 1 6.6 5.9 3.9 3.1 2.1 1.5 Rep. 2 6.5 5.9 4.1 3.1 2 1.5 Rep. 3 6.7 6.7 4.6 3.3 2.3 1.5 RRM-BE4max Rep. 1 8.1 6.9 4.8 3.8 2.5 1.6 Rep. 2 8 6.6 4.8 3.7 2.4 1.5 Rep. 3 7.7 7.1 5.2 4.5 3 2 Rad51DBD-BE4max Rep. 1 39.4 40.6 33.1 31.2 24.4 16.8 Rep. 2 39.2 40.6 32.7 30.7 23.6 16.6 Rep. 3 34.3 34.7 27.8 25.9 19.7 13.2 hyBE4max Rep. 1 49.5 49.2 49 50.1 49.1 47 Rep. 2 48.8 48.6 48.1 49.7 48.6 46.8 Rep. 3 43.3 42.6 43.4 44.5 44.1 41.8 BE4max-C-Rad51 Rep. 1 18.8 20.7 15.6 12.6 3.3 1.8 Rep. 2 18.9 20.4 15.6 12.7 3.5 1.9 Rep. 3 21.1 22.5 16.4 12.9 3.1 1.7 Blank control Rep. 1 0.1 0 0 0 0 Rep. 2 0 0 0 0 0 0 Rep. 3 0 0 0 0 0 0

[0111] The results in Table 4 and Table 5 shown that, compared with Rad51 DBD fused between NLS and rA1 in BE4max (i.e., Rad51DBD-N-BE4max), Rad51 DBD fused between rA1 and spCas9n in BE4Max (i.e., hyBE4max) most obviously improves the C-to-T editing efficiency on targets.

[0112] 3. Working Characteristics of hyBE4max

[0113] In order to further describe the working characteristics of hyBE4max fairly, 6 additional targets VEGFA site2, Lag3-sg2, HEK3, HEK4, EMX1-sg2p, and Nme1-sg1 (the sequences shown in Table 2) were designed and connected to BbsI site of plasmid U6-sgRNA-EF1α-GFP to obtain recombinant plasmids pV, pL, pH3, pH4, pEP and pN. Theses plasmids were sequenced by sanger to ensure that they are completely correct.

[0114] The recombinant plasmid containing hyBE4max in step 2 along with recombinant plasmids pE, pT, pV, pL, pH3, pH4, pEP or pN were respectively transfected into cells according to the method of 1.2 in step 1; the results of editing efficiency were obtained according to the methods of 1.3 and 1.4 in step 1; and statistical mapping was performed using GraphPad Prism 8.0.

[0115] The results were shown in FIGS. 2 and 3, in the editing windows C3-C8, the C-to-T editing efficiency of hyBE4max was 19-71%, and that of the corresponding BE4max was 13-47%; in the editing windows C9-C12, the C-to-T editing efficiency of hyBE4max was 19-55%, and that of the corresponding BE4max was 1.4-17%. Compared with BE4max, in the editing windows C3-C8, the average C-to-T editing efficiency of hyBE4max was 1.6-2.2 times that of BE4max; in the editing windows C9-C12, the average C-to-T editing efficiency of hyBE4max was 3.3-17 times that of BE4max. The production of indels maintained by hyBE4max was relatively low (FIG. 4).

[0116] 4. Effects of Fusion Proteins Containing Different Cytosine Deaminase

[0117] (1) Working Characteristics of Fusion Protein hyA3A-BE4max

[0118] 4.1.1. Rad51-DBD was synthesized according to the coding sequences in Table 1, and then seamlessly cloned and assembled between hA3A and spCas9n in the plasmid pCMV-A3A-BE4max expressing protein A3A-BE4max (FIG. 5); and the recombinant plasmid pA expressing fusion protein hyA3A-BE4max (FIG. 5) was constructed.

[0119] 4.1.2. 8 human endogenous targets were synthesized sequentially: the target sequences of EMX1 site1, Tim3-sg1, VEGFA site2, EMX1-sg2p, and Nme1-sg1 were shown in Table 2; the target sequences of FANCF site1, EGFR-sg5, and EGFR-sg21 were shown in Table 6; and recombinant plasmids pB1, pB2, . . . , pB8 expressing the corresponding targets of sgRNA were obtained by connecting them to Bbs I site of sgRNA-expression plasmid pU6-sgRNA-EF1α-GFP, respectively.

[0120] 4.1.3. The plasmids constructed in 4.1.1 and 4.1.2 were sequenced by sanger to ensure that they are completely correct.

TABLE-US-00006 TABLE 6 Targets and sequences used SEQ Name of ID targets Sequence (5′-3′) NO: FANCFsitel GGAATCCCTTCTGCAGCACCTGG 25 EGFR-sg5 GTGCTGGGCTCCGGTGCGTTCGG 26 EGFR-sg21 CAAAGCAGAAACTCACATCGAGG 27

[0121] 4.1.4. Cell Transfection

[0122] 5×10.sup.5 HEK293T cells were plated into a 24-well plate. When the cells grew to 70%-80%, the plasmid combination were transfected according to pA (or plasmid pCMV-A3A-BE4max):pB1 (or pB2, pB3, . . . , pB8)=750 ng:250 ng. 3 replicate wells were set for transfection of each plasmid combination, with 2×10.sup.5 cells per well. Simultaneously, a blank control without transfection of any plasmid was set.

[0123] 4.1.5. Genome Extraction and Preparation of Amplicon Library

[0124] The procedure of step 1.3 was followed, wherein the identification primers for the targets of FANCF site1, EGFR-sg5 and EGFR-sg21 were shown in Table 7, and the other identification primers for expression were shown in Table 3.

TABLE-US-00007 TABLE 7 Identification primers for targets used Name of targets Sequence (5′-3′) FANCFsite1 F: ggagtgagtacggtgtgc CAGAGAGGCGTATCATTTCGCGGAT R: gagttggatgctggatgg CCAGGTGCTGACGTAGGTAGTGCTT EGFR-sg5 F: ggagtgagtacggtgtgc CTTGTGGAGCCTCTTACACCCAGTG R: gagttggatgctggatgg CTCCCCACCAGACCATGAGAGGC EGFR-sg21 F: ggagtgagtacggtgtgc GCAGCATGTGGCACCATCTCA R: gagttggatgctggatgg TGGACCCCCACACAGCAA

[0125] 4.1.6. Analysis and Statistics of Deep Sequencing Results

[0126] The procedure of step 1.4 was followed.

[0127] The results show that, compared with the protein A3A-BE4max, the editing efficiency of single-base C-to-T at different positions (C3-C15) of individual targets was significantly improved by the fusion protein hyA3A-BE4max (FIG. 6). Compared with A3A-BE4max, the high activity windows of hyA3A-BE4max were expanded from the original C3-C11 to C3-C15; wherein the editing efficiency of single-base C-to-T by hyA3A-BE4max was 1.1-2.3 times that of A3A-BE4max at C3-C11 far away from PAM region, while the editing efficiency of single-base C-to-T by hyA3A-BE4max was 3.1-4.1 times that of A3A-BE4max at C12-C15 near PAM region; the editing efficiency of single-base C-to-T at C12-C15 near PAM region was more obviously improved by hyA3A-BE4max. And the indels maintained by hyA3A-BE4max were relatively low (FIG. 8).

[0128] (2) Working Characteristics of Fusion Protein hyeA3A-BE4max

[0129] 4.2.1. Construction of Working System Plasmid

[0130] Rad51-DBD was synthesized according to the coding sequences in Table 1, and then seamlessly cloned and assembled between eA3A and spCas9n in the plasmid pCMV-eA3A-BE4max expressing the protein eA3A-BE4max (FIG. 9), and the recombinant plasmid pAe expressing the fusion protein hyeA3A-BE4max (FIG. 9) was constructed.

[0131] 4.2.2. Construction of Target Plasmid

[0132] Meanwhile, 11 human endogenous targets were designed and synthesized: the target sequences of EMX1-sg2p, EMX1 site1, and Nme1-sg1 were shown in Table 2; the target sequence of EGFR-sg21 was shown in Table 6; and the other target sequences were shown in Table 8; which were respectively connected to BbsI site of sgRNA-expression plasmid U6-sgRNA-EF1α-GFP to express sgRNA of corresponding target, thus obtaining recombinant plasmids pC1, pC2, . . . , pC11.

[0133] 4.2.3. The plasmids constructed in 4.2.1 and 4.2.2 were sequenced by sanger to ensure that they are completely correct.

TABLE-US-00008 TABLE 8 Targets and sequences used SEQ Name of ID targets Sequence (5′-3′) NO: CTLA-sg1 CTCCCTCAAGCAGGCCCCGCTGG 28 EGFR-sg5 GTGCTGGGCTCCGGTGCGTTCGG 29 CDK10-sg1 TTCTCGGAGGCTCAGGTGCGTGG 30 EMX1-sg1 GCTCCCATCACATCAACCGGTGG 31 HPRT1-sg6 GCCCTCTGTGTGCTCAAGGGGGG 32 EGFR-sg26 CATGCCCTTCGGCTGCCTCCTGG 33 CCR5-sg1 TAATAATTGATGTCATAGATTGG 34

[0134] 4.2.4. Cell Transfection-Verification of hyeA3A-BE4max Working System

[0135] 5×10.sup.5 HEK293T cells were plated into a 24-well plate. When the cells grew to 70%-80%, the plasmid combination were transfected according to pA (or plasmid pCMV-eA3A-BE4max):pC1 (or pC2, pC3, . . . , pC11)=750 ng:250 ng. 3 replicate wells were set for transfection of each plasmid combination, with 2×10.sup.5 cells in each well. Simultaneously, a blank control without transfection of any plasmid was set.

[0136] 4.2.5. Genome Extraction and Preparation of Amplicon Library

[0137] The procedure of step 1.3 was followed, wherein the identification primers for EMX1-sg2p, EMX1 site1 and Nme1-sg1 were shown in Table 3, the identification primers for EGFR-sg21 were shown in Table 7, and the other target sequences were shown in Table 9.

TABLE-US-00009 TABLE 9 Identification primers for targets used Identification of primers Sequence (5′-3′) CTLA-sg1 F: GGAGTGAGTACGGTGTG CACTTTAACCCCAAGTCTAG CAAGC R: GAGTTGGATGCTGGATG GATAATTAATCCAATGACTG TGAAGAGC EGFR-sg5 F: GGAGTGAGTACGGTGTG CCTTGTGGAGCCTCTTACAC CCAGTG R: GAGTTGGATGCTGGATG GCTCCCCACCAGACCATGAG AGGC CDK10-sg1 F: GGAGTGAGTACGGTGTG CGGTGATGGGTTACTGTGAG CAGGAC R: GAGTTGGATGCTGGATG GGTCCCCAAGTGCCCTGTTT CGTTAT EMX1-sg1 F: GGAGTGAGTACGGTGTG CGTGGTTCCAGAACCGGAGG ACAAAG R: GAGTTGGATGCTGGATG GGTTTGTGGTTGCCCACCCT AGTCAT HPRT1-sg6 F: GGAGTGAGTACGGTGTG CTTCCTGATTTTATTTCTGT AGGACTG R: GAGTTGGATGCTGGATG GTCTACAGTCATAGGAATGG ATCTATCA EGFR-sg26 F: GGAGTGAGTACGGTGTG CACCTCCACCGTGCAGCTC R: GAGTTGGATGCTGGATG GAGCGCAGACCGCATGTGAG CCR5-sg1 F: GGAGTGAGTACGGTGTG CTGCACAGGGTGGAACAAGA TGGATT R: GAGTTGGATGCTGGATG GATGACCAGCATGTTGCCCA CAAAAC

[0138] 4.2.6 Analysis and Statistics of Deep Sequencing Results

[0139] The procedure of step 1.4 was followed.

[0140] The results show that, compared with the protein eA3A-BE4max, the editing efficiency of single-base C-to-T at different positions (C3-C15) of individual targets was significantly improved by the fusion protein hyeA3A-BE4max, and the high activity windows were expanded from the original C3-C11 to C3-C15, which can specifically target the single base C in TC motif to achieve the C-to-T conversion (FIG. 10); wherein, in C3-C11 far away from PAM region, the editing efficiency of single-base C-to-T by hyeA3A-BE4max was 1.6-2.8 times that of eA3A-BE4max, and at C12-C15 near PAM region, the editing efficiency of single-base C-to-T by hyeA3A-BE4max was 4.5-31.9 times that of eA3A-BE4max, i.e., at C12-C15 near PAM region, the editing efficiency of single-base C-to-T was more obviously improved by hyeA3A-BE4max (FIG. 11). While the indels maintained by hyeA3A-BE4max were relatively low (FIG. 12).

[0141] 5. Gene Therapy Using Fusion Protein hyeA3A-BE4max

[0142] 5.1. Test of Editing Efficiency of hyeA3A-BE4max Targeting HBG-117G Sites on HEK293T Cells

[0143] Transfected HEK293T cells with the plasmid combination according to the cell transfection method of 4.2.4, pA (or plasmid pCMV-A3A-BE4max, or pAe, or pCMV-eA3A-BE4max):pC12=750 ng:250 ng; and deep sequencing and statistical analysis were carried out according to the method of 4.2.6.

[0144] The construction method of the recombinant plasmid pC12 is as follows: the sgRNA target sequence of HBG-117G (GGCTATTGGTCAAGGCAAGGCTGG, SEQ ID NO: 35) was connected to the BbsI site of the sgRNA-expression plasmid U6-sgRNA-EF1α-GFP to express the sgRNA of corresponding target to obtain the recombinant plasmid pC12.

[0145] The identification primers used for the above deep sequencing target HBG-117G are as follows:

TABLE-US-00010 HBG-117G F: AGTGAGTACGGTGTGCTGGAATGACTGAATCGGAACAAGGC; HBG-117G R: GTTGGATGCTGGATGGCTGGCCTCACTGGATACTCTAAGACT.

[0146] Results: as shown in FIG. 14, A3A-BE4max not only targeted the −117G>A (C11) mutation in the HBG1/2 promoter region, but also produced the −109G>A (C3), −122G>A (C16) “bystander” mutation; for eA3A-BE4max and hyeA3A-BE4max, eA3A-BE4max accurately edited the G-to-A conversion on −117 (i.e., the C-to-T conversion of complementary chain) without causing “bystander” mutation, but the efficiency was very low; and compared with eA3A-BE4max, hyeA3A-BE4max improved the G-to-A conversion efficiency by 6.6 times, and detectable “bystander” mutation was not produced at both −109 and −122. The results suggest that hyeA3A-BE4max exhibited the characteristics of targeting HBG1/2 promoter region −117G accurately and efficiently (the mechanism was summarized in FIG. 13).

[0147] 5.2. Construction of Lentiviral Vector and Virus Packaging

[0148] 5.2.1 Construction of Lentiviral Vector

[0149] pLenti-BE3-P2A-Puro (Addgene, #110838) was used as skeletal vector, the coding sequence of hyA3A-BE4max was cloned seamlessly to replace BE3 on the skeletal vector to obtain lentiviral vector Lenti-hyA3A-BE4max-P2A-GFP.

[0150] pLenti-BE3-P2A-Puro (Addgene, #110838) was used as skeletal vector, the coding sequence of hyeA3A-BE4max was cloned seamlessly to replace BE3 on the skeletal vector to obtain a lentiviral vector Lenti-hyeA3A-BE4max-P2A-GFP.

[0151] The target sequence of HBG-117G in 5.2 was connected to the upstream of the above two lentiviral vectors hyA3A-BE4max or hyeA3A-BE4max (top figure of FIG. 15) to obtain a recombinant plasmid Lenti-117G-hyA3A-BE4max-P2A-GFP and a recombinant plasmid Lenti-117G hyeA3A-BE4max-P2A-GFP (bottom figure of FIG. 15).

[0152] 5.2.2. Lentiviral Packaging

[0153] 5.2.2.1. Transfection

[0154] On the 1.sup.st day, HEK293T cells in good growth condition were digested and placed in 10 cm dishes, with about 30 dishes for each virus. On the 2.sup.nd day, when the confluence reached 80%-90%, the plasmid was transfected with following amount: Lenti-117G-hyA3A-BE4max-P2A-GFP (or Lenti-117G-hyeA3A-BE4max-P2A-GFP):PSPAX2:PMD2.G=10 μg:10 μg:10 μg.

[0155] 5.2.2.2. Collection and Purification of Virus

[0156] The virus supernatant was collected from HEK293T cell supernatant at 48 h (i.e., 0 h was recorded from transfection) and 72 h after transfection. The supernatant was centrifuged at 4° C. under 4000 g for 10 min, the cell debris was removed, then the supernatant was filtered through a 0.45 μm filter into a 40 mL ultrafiltration centrifuge tube, the lentiviral crude extract was added into a filter cup and centrifuged at 4° C. under 25000 g for 2.5 hours. After the centrifugation, the centrifugal device was taken out, and filter cup was separated from the filtrate collection cup. The liquid in the sample collection cup is concentrated virus liquid (containing lentivirus Lenti-117G-hyA3A-BE4max-P2A-GFP or lentivirus Lenti-117G-hyeA3A-BE4max-P2A-GFP). The concentrated virus liquid was removed, subpackaged and stored in a virus tube, and preserved at −80° C. for long-term storage.

[0157] 5.3. Gene Therapy

[0158] 5.3.1. HUDEP-2 (Δ.sup.Gγ) Cells Infected With Virus

[0159] 5×10.sup.4 HUDEP-2 (Δ.sup.Gγ) cells were plated in culture medium with a total volume of 100 μl in 3 wells in a 96-well plate, and then infected with lentivirus Lenti-117G-hyA3A-BE4max-P2A-GFP and Lenti-117G-hyeA3A-BE4max-P2A-GFP in equal titer, respectively. The infection system is as follows:

TABLE-US-00011 Lenti-117G- Lenti-117G- hyA3A- hyeA3A- BE4max- BE4max- Blank Reagents P2A-GFP P2A-GFP control virus 10 MOI 10 MOI 0 Polybrene (100×) 1 μl 1 μl 1 μl Culture medium Added up to 100 μl

[0160] 5.3.2. Detection of Editing Efficiency

[0161] The HUDEP-2 (ΔGγ) cells infected with lentivirus Lenti-117G-hyA3A-BE4max-P2A-GFP or Lenti-117G-hyeA3A-BE4max-P2A-GFP were sorted by flow cytometry to obtain GFP-positive cells, the GFP-positive cells were cultured, and then collected after the number of cells was more than 5×10.sup.4, the genomic DNA was extracted, and the deep sequencing and analysis were carried out according to the method of 5.1.

[0162] Results: compared with hyA3A-BE4max, hyeA3A-BE4max efficiently targeted the precise −117G>A mutation in HUDEP-2 (ΔGγ) cells and shown higher activity (FIG. 16).

[0163] 5.3.3 Differentiation and Detection of γ Globin Expression

[0164] The HUDEP-2 (ΔGγ) cells infected with lentivirus Lenti-117G-hyA3A-BE4max-P2A-GFP or Lenti −117G-hyeA3A-BE4max-P2A-GFP were sorted by flow cytometry to obtain GFP-positive cells, the GFP-positive cells were cultured until the number of cells was more than 5×10.sup.4, and HUDEP-2 (Δ.sup.Gγ) cells were collected after about 5-7 days for differentiation and expression. The differentiation process is as follows: [0165] 1×10.sup.5 HUDEP-2 (Δ.sup.Gγ) cells after being counted were differentiated in erythroid cell differentiation medium (IMDM), supplemented with 2% human blood AB-type plasma (serum) (Gemini, 100-512), 1% L-glutamine, 2 IU/mL of heparin, and erythropoietin (EPO, 3 IU/mL, PeproTech), 330 μg/mL Holo-human transferrin (Sigma-Aldrich), human stem cell factor (SCF, 50 ng/mL, PeproTech), 2% Pen/Strep (Gibco), and 10 μg/mL recombinant human insulin; the cells were then differentiated for 8 days.

[0166] Detection of γ globin expression: the cells were collected after 8 days of differentiation and total mRNA was extracted by phenol-chloroform extraction method. HiScript II Q RT SuperMix (Vazyme) was used to reversely transcribe the isolated mRNA; qPCR was performed on QuantiStudio 3 real-time PCR system (ABI), and mRNA levels of HBG and HBB were quantified by SYBRGreen qPCR. The primers (5′-3′) are as follows:

TABLE-US-00012 HBG-QPCR-F: GGTTATCAATAAGCTCCTAGTCC; HBG-QPCR-R: ACAACCAGGAGCCTTCCCA; HBB-QPCR-F: TGAGGAGAAGTCTGCCGTTAC; HBB-QPCR-F: ACCACCAGCAGCCTGCCCA.

[0167] Results: compared with WT cells of HUDEP-2 (Δ.sup.Gγ), hyA3A-BE4max and hyeA3A-BE4max could significantly increase the mRNA level of γ-globin in HUDEP-2 (ΔGγ) cells; and hyeA3A-BE4max increased the mRNA level of γ-globin in HUDEP-2 (ΔGγ) cells by 3 times as much as that of hyA3A-BE4max (FIG. 17).

[0168] 6. Using Fusion Protein hyA3A-BE4max to Construct Animal Model of DMD Disease

[0169] The mice used below are C57/BL6 mice.

[0170] 6.1. Construction of Transcription Template For Working System mRNA and Target sgRNA

[0171] A mouse-related gene sequence was download at NCBI, as shown in FIG. 18, a sgRNA (DMD-sg3 target sequence: ACATCTCATCAAGGACTTGTTGG, SEQ ID NO: 36) was designed at the target site (dystrophin gene, i.e., the site in the rectangle frame of exon 12 of DMD gene), and Oligo primers were ordered; the sgRNA formed by annealing was cloned into T7 vector selekton, and the DMD-sg3 template containing T7 promoter was amplified via in vitro transcription (IVT) using primer pairs IVT-PCR-F and IVT-PCR-R (Table 10), and primer pairs IVT-T7-hyA3A-BE4max-F and IVT-T7-hyA3A-BE4max-R (Table 10) were used to introduce T7 promoter into the mRNA template of hyA3A-BE4max or A3A-BE4max by PCR.

TABLE-US-00013 TABLE 10 PCR primers used in IVT Primer name Primer sequence (5′-3′) IVT-PCF-F GCGGCTTTGTTGAATAAATCGCATTCG IVT-PCF-R AAAAGCACCGACTCGGTGCC IVT-T7- TTAATACGACTCACTATAGGGAGAATG hyA3A-BE4 AAGAGGACCGCCGATGGCTC max-F IVT-T7- CTAGTCACCTCCCAGCTGAGACAGGTC hyA3A-BE4 max-R

[0172] 6.2. In Vitro Transcription of sgRNA (DMD-sg3)

[0173] A common DNA product purification kit was used to purify the PCR product in 6.1, the purified PCR product was then used as a linearized DNA template and T7 in vitro transcription kit (MEGAshortscript™ Kit) was used to carry out in vitro transcription. The transcribed sgRNA was purified by using lithium chloride precipitation method.

[0174] 6.3. Transcription of Working System mRNA (A3A-BE4max and HyA3A-BE4max)

[0175] T7 templates of A3A-BE4max and hyA3A-BE4max were transcribed in vitro using the in vitro RNA transcription kit (mMESSAGE mMACHINE®T7 Ultra Kit) to obtain the working system mRNA, which was then purified.

[0176] 6.4. Preparation of Microinjection Mixture

[0177] An injection mixture was prepared with nuclease-free water to obtain a mixture with a total volume of 20 μL. The injection mixture contains working system mRNA (mRNA containing A3A-BE4max or hyA3A-BE4max) with a final concentration of 100 ng/μL and sgRNA (DMD-sg3) with a final concentration of 200 ng/μL.

[0178] 6.5. Collection of One-Cell Stage Embryos

[0179] (1) The 1.sup.st day: donor female mice aged 6-8 weeks were intraperitoneally injected with 100 μL (5 IU) of PMSG working solution between 1 p.m. and 2 p.m.

[0180] (2) The 3.sup.rd day: 100 μL (5 IU) of hCG working solution was intraperitoneally injected into the female mice between 2 p.m. and 4 p.m. that had been injected with PMSG. After the injection, the hormone-treated female mice were co-caged one-to-one with male mice aged 10-14 weeks. Meanwhile, the estrous hormone-free female mice were mated with the tubal ligated male mice at around 4 p.m. for the preparation of pseudopregnant female mice.

[0181] (3) The 4.sup.th day: before 9 a.m., the recipient female mice that were co-caged with the ligated male mice were examined for the presence of pregnancy plugs, and the female mice having pregnancy plugs were collected in new cages for embryo transfer experiment in the afternoon.

[0182] (4) The superovulatory female donor mice were sacrificed by carbon dioxide asphyxiation method, and their oviducts were taken out and placed in dishes in which preheated M2 medium was added.

[0183] (5) The oviducts were placed in another new dish, and preheated M2 medium and hyaluronic acid were added into the dish with the volume ratio of M2 medium to hyaluronic acid being 9:1. The ampullae of the oviducts were pulled with tweezers under a stereomicroscope to release the embryos from the oviducts into the dish. Embryos were incubated in M2 medium containing hyaluronic acid until cumulus cells dropped. After the cumulus cells were removed, the embryos were transferred to a new dish, and M2 medium without hyaluronic acid was added into the dish, and the embryos were repeatedly rinsed with M2 medium to wash out both hyaluronic acid and cumulus cells.

[0184] (6) The washed embryos were transferred to a new dish, a few drops of KSOM medium was firstly added dispersedly into the dish, and then mineral oil was slowly added into the dish to separate and cover the KSOM medium with the mineral oil. In general, 6 droplets of KSOM medium can be added into a 35 mm dish with each droplet being 50 μL. 50 embryos as a group were placed firstly onto the middle KSOM medium droplets and rinsed, and then transferred to a new medium droplet. The embryos before microinjection were taken out and incubated in M2 medium in a cell incubator.

[0185] 6.6. Microinjection and Embryo Transfer

[0186] (1) Fixing needle, injection needle and silicified glass slide were prepared, and a drop of M2 medium covered with mineral oil was dropped into the middle of the slide.

[0187] (2) The injection needle was allowed to automatically suck and filled with the microinjection mixture prepared in step 2.4 by capillary action, and the injection needle was loaded onto the fixed handle of the microinjection apparatus.

[0188] (3) 50 embryos were transferred into M2 medium on the glass slide, and the fixing needle was moved close to the embryos, so that the embryos could be fixed onto the fixing needle under negative pressure. After the embryos were fixed, the cytoplasm was found under a high-power microscope, and the tip of the injection needle was pushed through the zona pellucida and cell membrane, and the mixture was injected into the cytoplasm of the embryo.

[0189] (4) The injected embryonic cells were transferred to a new droplet of M2 medium. The steps (3) and (4) were repeated until all embryos were injected. After injection of a group of experimental groups, the embryos were transferred to new KSOM medium and incubated in a cell incubator for 1-2 hours or overnight. After all the embryos were injected, the embryos killed by mechanical damage were excluded, and the healthy embryos were transferred to new KSOM medium.

[0190] (5) The pseudopregnant rats were anesthetized by intraperitoneal injection of 600 μL avertin. A shaver was used to remove the hair from the back of the mice. The skin after being shaved was wiped with 70% ethanol.

[0191] (6) A small opening was formed by cutting at the position of the ovary, blunt forceps were used to pull the fat pad of the ovary to pull the ovary out, and at the same time the ovary was fixed to the outside with hemostatic forceps, and blunt forceps were used to find the funnel-shaped orifice of the oviduct at the lower side of the ovarian bursa.

[0192] (7) Transfer needle was used to sequentially suck M2 medium with two small bubbles, and about 15 embryos, and the bubbles were for observing the position of embryos in the transfer needle.

[0193] (8) Ovarian bursa was gently opened, the funnel-shaped orifice of the oviduct was positioned using a shaft, the transfer needle was extended to the opening of the ovary, then the embryo in the transfer needle was punched out, and the transfer needle was gently withdrawn.

[0194] (9) The hemostatic forceps fixing the fat pad of the ovary was released, the ovary was put back into the original cavity, and the muscle opening and the skin opening were sewn respectively with sutures.

[0195] (10) The mice after surgery were placed on a heat-preserving table with a constant temperature of 37 degrees; the mice were transferred to a feeding cage for feeding after the mice regained consciousness, followed by waiting for embryo development until delivery. Typically, the successfully transplanted female mice gave birth after 3 weeks.

[0196] 6.7. Identification of Mouse Genome

[0197] The mice (in step 6.6) born about 10-15 days was taken and their toes were cut for genome identification. The specific steps are as follows:

[0198] 6.7.1 Genomic Extraction

[0199] (1) Their toes were cut and put into 1.5 mL centrifuge tubes, into which 500 μL of toe digestion liquid was added. The digestion liquid was prepared according to the ratio of protease K:tissue lysate=1:500, and then the tubes were placed in a water bath at 55° C. overnight;

[0200] (2) The toes digested overnight were taken out and placed at room temperature for 10-15 minutes, mixed thoroughly upside down, and centrifuged at 13000 rpm for 15 minutes.

[0201] (3) 400 μL supernatant was sucked out from each tube, an equal volume of chloroform was added, fully mixed. After DNA precipitation, the mixture was centrifuged at 12000 rpm for 10 minutes.

[0202] (4) 200 μL of 75% alcohol pre-cooled in a refrigerator at −20° C. was added into each tube and mixed gently, then centrifuged at 12,000 rpm for 5 min, and the supernatant was discarded, followed by air-drying in a clean workbench.

[0203] (5) 50-100 μL of deionized ultra-pure water was added according to the content of DNA, and the DNA was dissolved at 55° C. for 2 hours to obtain PCR template.

[0204] 6.7.2 Identification of Genome

[0205] Primer pair F/R of target DMA-sg3 (Table 11) was used to obtain a DNA fragment containing the target; firstly, the occurrence of double peaks was confirmed by first-generation sequencing, and then the editing efficiency was obtained by high-throughput deep sequencing. According to the high-throughput results, among the 10 F0s producing mutations in the hyA3A-BE4max treatment group, there were 6 homozygous terminating nonsense mutations from CAA to TAA (the mice numbered #BD03, #BD05, #BD07, #BD12, #BD15, and #BD16 in the lower figure of FIG. 19), while no terminating nonsense mutation was found in the F0 in the A3A-BE4max treatment group (upper figure of FIG. 19). The HYA3A-BE4max treatment group had a higher number of Reads (high-throughput sequencing fragments) containing the stop codon TAA as a percentage of total Reads in homozygous mutations compared to the A3A-BE4MAX treatment group (FIG. 20).

TABLE-US-00014 TABLE 11 Name of targets Sequence (5′-3′) DMD-sg3 F: ggagtgagtacggtgtgcTCA AACTCCCAATGATTTCCTCAAT R: gagttggatgctggatggTGC ACTTCAGCTTCTTCATCTTCTG

[0206] 6.8. Phenotype Identification of DMD Gene

[0207] Wild-type mice of 5 weeks old (blank control) and mice with DMD gene mutation identified in 2.7.2 were taken, and immunohistochemical detection was carried out according to the following methods:

[0208] The tibialis anterior muscles of mice was taken and rinsed with alcohol and PBS. They were placed into a small cube box covered with OTC gel and placed into an isopentane beaker, followed by freezing in liquid nitrogen for about 30 s, taking out and storing at −20° C., and then freezing and sectioning. The prepared sections were washed with PBST for 3 times/5 min, oil circles were drawn at the aligned positions of tissues. The sections were blocked by adding blocking solution, and after 1 h of blocking, the sections were incubated overnight with Laminin primary antibody or Dystrophin primary antibody (Abcam, ab11575 or Abcam, ab15277 diluted at 1:500), respectively. The sections were washed with PBST for 3 times/5 min, followed by incubation with 1:1000 anti-rabbit diluted secondary antibody for 2 h, and then washed with PBST for 3 times/5 min. DAPI diluted at 1:100 was added and incubated for 10 min, and then washed with PBST for 3 times/5 min. Followed by dropwise adding anti-quenching agent, adding a cover glass, sealing with nail polish, and finally observing the fluorescence of the sections with a fluorescence microscope.

[0209] The results show that, compared with the mice of WT (+/+) and A3A-BE4max-treated mice (such as #AD26), only six F0-generation mice (such as #BD03 in FIG. 21) in the hyA3A-BE4max-treated mice that caused C-to-T homozygous nonsense mutation at position 10 of DMA-sg3 target sequence shown no protein expression of DMD (FIG. 21), which also proved that the animal model of DMD disease was constructed successfully.

[0210] According to the fluorescence observation results, a summary of F0 generation mice produced by editing hyA3A-BE4max and A3A-BE4max was obtained (as shown in Table 12).

TABLE-US-00015 TABLE 12 Comparison of F0 mutation results after editing hyA3A-BE4max and A3A-BE4max Mutation frequency Proportion of Proportion of homozygous mutated nonsense treatment mice mutation A3A-BE4max 10/11 (91%) 0/11 (0) hyA3A-BE4max 10/10 (100%) 6/10 (60%)

[0211] 6.9. Genetic Analysis of Germ Line of DMD Mutant Mice (F0.fwdarw.F1)

[0212] The homozygous female mice #BD12 (F0) with DMD phenotype were mated with male wild-type mice, 8 homozygous F1 mice were obtained, and the genotypes of the born F1 were identified; and sanger sequencing results show that the Reads frequency of nonsense mutation produced in each F1 exceeded 96%, i.e., this nonsense mutation could be stably inherited to F1 generation (FIG. 22).

[0213] 6.10. Off-Target Detection

[0214] Off-target primers were designed by using a Cas-OFFinder function on a CRISPR RGEN Tools website (http://www.rgenome.net) (Table 13): firstly, the PAM type of the tested tool and the species type to be tested (for example, for a mouse, being Mus musculus (mm10)-Mouse) were selected, then, the designed sgRNA sequence with a PAM portion removed was filled in the box of Query Sequences, the mismatched bases were selected to be within 3, and the resulting DNA Bulge Size was within 1, and then the corresponding off-target primers were obtained after submission.

[0215] The off-target primers described above were used to implement PCR respectively on the genomic DNA of F0 mice that have been edited by hyA3A-BE4max (with WT as blank control), and then the products were subjected to high-throughput deep sequencing to obtain the results for the off-target efficiency (FIG. 23).

TABLE-US-00016 TABLE 13 Off-target sites of DMD-sg3 and PCR primers Name of off- Target sequence Primer sequence target sites (5′-3′) (5′-3′) DMD-sg3- acatttaatcaaggc F: ggagtgagtacg OT1 cttgttgg gtgtgcaaaagccta caacacaatgagaaa c R: gagttggatgct ggatggttgataagg ctctaccaatgtgaa c DMD-sg3- acctgtcatcaagca F: ggagtgagtacg OT2 cttgttgg gtgtgcaaggaatga ccttgcagagactgc c R: gagttggatgct ggatggtgctcagtc ccatggatgactgtg a DMD-sg3- tcctgtcatcaagga F: ggagtgagtacg OT3 cttgttgg gtgtgcggtgcagag actgaagaaaaggcc a R: gagttggatgct ggatggtctatctct gagtatgtcaggcac tgg DMD-sg3- tcctctcatcaagca F: ggagtgagtacg OT4 cttgttgg gtgtgcggaatgacc atccagagacttccc c R: gagttggatgct ggatggttggggacc ctatgctcagtccaa t DMD-sg3- acatctcacccagga F: ggagtgagtacg OT5 cctgtggg gtgtgcggcttcctt ttctgtcactgtggg a R: gagttggatgct ggatggggggactgc cactcctaactttca t DMD-sg- acttctcatcaagga F: ggagtgagtacg OT6 cctctggg gtgtgctccctatcc gtcttagtctggagc c R: gagttggatgct ggatggacaagaaac catgcatcagcttgt ga DMD-sg3- acttctcatcaagga F: ggagtgagtacg OT7 cctctggg gtgtgcgcagaaaga aagtacagacactta ca R: gagttggatgct ggatggaatgggtaa atagttaaaagccat ga DMD-sg3- acaactcaaaaagga F: ggagtgagtacg OT8 cttgttgg gtgtgcggctaccat aggccattttctgca t R: gagttggatgct ggatggtgagaatga atgttaggtcaggtg ga DMD-sg3- acatctcatctatga F: ggagtgagtacg OT9 tttgtagg gtgtgcggcttgccc aagtttattagggga gt R: gagttggatgct ggatggccagcccct tggcctcctttaaaa t DMD-sg3- acaactcctcatgga F: ggagtgagtacg OT10 cttgtggg gtgtgcacatcccag agaatcctgagttta aga R: gagttggatgct ggatggtgctaggag agatggttgtatcaa gga DMD-sg3- gcatctcatctagga F: ggagtgagtacg OT11 cttggtgg gtgtgctgtgatagg tgatttctcatgcac ca R: gagttggatgct ggatggatcacattc acttcccttcacccc g DMD-sg3- aaatctcatcaatta F: ggagtgagtacg OT12 cttgtagg gtgtgccagccttat ccatttacttgtgaa ttt R: gagttggatgct ggatggtcaagaaca ataaaacaaatgaca gc DMD-sg3- tcatgtcagcaagga F: ggagtgagtacg OT13 cttgttgg gtgtgcatgaccatc cagagactgacccac t R: gagttggatgct ggatggccacttctg tattagtctggcact ggc DMD-sg3- acatctgaccaaggg F: ggagtgagtacg OT14 cttgtagg gtgtgccagttcctg tcggcagttcctcaa t R: gagttggatgct ggatggtgtgccctg gaaaagatcacacag a DMD-sg3- acttctcagcaagca F: ggagtgagtacg OT15 cttgtagg gtgtgccaacggtag cctaaggactgtcca c R: gagttggatgct ggatggcaggcttcc tgactttcaaaaccg c

[0216] The results in FIG. 23 show that, compared with the blank control, the fusion proteins hyA3A-BE4max and DMD-sg3 substantially did not produce C-to-T editing on the 15 sites similar to DMD-sg3 targets in Table 13, i.e., the fusion proteins hyA3A-BE4max and DMD-sg3 in the present application substantially did not produce off-target effect.