Mitochondrial base editor for realizing extremely low cell nucleus off-target editing and application of mitochondrial base editor

Abstract

The invention discloses a mitochondrial base editor with extremely low cell nucleus off-target. MTS in the editing system is selected from a 2xMTS form of any one of seven mitochondrial targeting positioning signals, UGI in the editing system is in an Inteins splitting form or a Splitts splitting form, a 2xMTS-UGI (Inteins) or 2xMTS-UGI (Splitts) skeleton carrier formed by the MTS and the UGI realizes C / G-T / A editing of a target position on mitochondrial DNA, and the off-target value of a cell nucleus is extremely low. On the basis of the 2xMTS-UGI (Inteins) or 2xMTS-UGI (Splitts) skeleton vector, the double-stranded deamination protein inhibitory enzyme is jointly expressed, C / G to T / A editing of the target position on the mitochondrial DNA is also realized, the off-target value of the cell nucleus is as low as the background level, and a safer tool is provided for the gene editing technology.

Description

Technical Field

[0001] The present invention belongs to the technical field of gene editing and is a mitochondrial base editor that realizes C / G to T / A editing on DNA, and its cell nucleus off-target editing is extremely low. Background Art

[0002] Mitochondrial dysfunction can cause mitochondrial diseases, and the organ defects and physical symptoms caused by these diseases are often different. However, what they have in common is that the mitochondrial DNA (mtDNA) of patients with mitochondrial diseases often carries mutations, and the mutations affect the main components encoding oxidative phosphorylation, thereby causing mitochondrial dysfunction in patients. There is currently a lack of treatment for mitochondrial diseases. Chemotherapy for mitochondrial diseases, such as supplementing patients with coenzyme Q10 and vitamins, cannot change the mutated mtDNA. Some non-chemotherapy methods, including mitochondrial replacement and mitochondrial enhancement, are committed to achieving permanent cure, but their recovery efficiency is difficult to grasp. Because the double-layer membrane structure of mitochondria hinders the input of sgRNA, the powerful CRISPR-Cas technology is also difficult to function in mitochondria.

[0003] Restriction endonucleases, mito-ZFNs, and mito-TALENs that fuse mitochondrial localization signal peptides can introduce DNA double-strand breaks (DSBs) into mutant mtDNA to reduce the proportion of mutant mtDNA. However, these methods cannot achieve precise editing of single bases in mtDNA, and therefore cannot be used to explore precise treatments for mitochondrial diseases or to construct animal disease models for mtDNA mutations.

[0004] The DddA domain of the bacterial toxin protein from Burkholderia cenocepacia, DddAtox, has double-stranded DNA deaminase activity. By connecting to TALE to form a cytosine base editor (DdCBE), it can achieve site-specific C / G to T / A mutations in mitochondria. However, the original version of DdCBE will produce a large number of extensive off-target edits in the nuclear genome. DddA mutants such as K1389A / T1391A obtained by alanine mutation and co-expression of bpNLS-DddI with DdCBE reduced nuclear off-targets, but still failed to reduce its nuclear off-targets to extremely low levels or introduce additional protein components into the cell. Summary of the invention

[0005] The first objective of the present invention is to provide a mitochondrial base editor with extremely low nuclear off-target effects in response to the above-mentioned deficiencies in the prior art.

[0006] The second object of the present invention is to provide a mitochondrial base editing system with extremely low nuclear off-target.

[0007] The third object of the present invention is to provide applications of the mitochondrial base editor and the mitochondrial base editing system.

[0008] The fourth object of the present invention is to provide a method for achieving C / G to T / A editing on mitochondrial DNA using this system.

[0009] The purpose of the present invention can be achieved through the following technical solutions:

[0010] An improved mitochondrial targeting signal 2xMTS, wherein the improved mitochondrial targeting signal is in the form of 2xMTS in which two MTSs are connected in series, and the first MTS and the second MTS are connected by GS amino acid as a linker; the MTS is selected from any one of the mitochondrial targeting signal MTSs from genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD.

[0011] As a preferred embodiment of the present invention, the nucleotide coding sequence of Linker is GGAAGC

[0012] As a preferred embodiment of the present invention, the amino acid sequences of the improved mitochondrial targeting signal, the mitochondrial targeting signal MTS of genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD are shown in SEQ ID NOs: 1-7, respectively.

[0013] In the present invention, the MTS derived from genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD replace the original MTS in the DdCBE system, so that the nuclear off-target of the replaced system is reduced. The above seven MTSs are connected in series to form 2xMTS, which are connected by a linker in the middle. At this time, the nuclear off-target of the modified 2xMTS-DdCBE is further reduced, and the mitochondrial editing efficiency is improved.

[0014] The mitochondrial targeting signal (MTS) has the ability to enhance the entry of DdCBE into mitochondria. It has not been disclosed as a mitochondrial targeting signal for DdCBE, but the relevant information of the gene in which it is located exists in databases such as MitoCarta3.0, and its specific functional properties have not been studied in depth. The above-mentioned modification of the DdCBE backbone carrier is completely independently developed in the present invention.

[0015] A mitochondrial localization skeleton vector, wherein the mitochondrial localization skeleton vector contains the improved mitochondrial targeting signal.

[0016] As a preferred embodiment of the present invention, the mitochondrial localization backbone vector is selected from a pair of four 2xMTS-UGI (Splits) -DdCBE backbone vectors or a pair of four 2xMTS-UGI (Inteins) -DdCBE backbone vectors; the backbone vectors are paired according to Half DddA G1333C and G1333N, and G1397C and G1397N. The 2xMTS-UGI (Splits) -DdCBE backbone vector is a double-stranded DNA cytosine deaminase DddA domain from Burkholderia cenocepacia divided into half The four split protein pairs of DddA G1333N / G1333C and G1397N / G1397C were fused with the glycosylase inhibitor UGI and then inserted between the two restriction sites NheⅠ and PmeⅠ of the MTS-TALE backbone vector. On this basis, UGI was split at S38 with the intein Intein, and the 2xMTS was added in front of the C-terminus of the intein. The split half components of UGI were connected with IRES, thereby obtaining four 2xMTS-UGI (Inteins)-DdCBE backbone vectors; based on the previous four 2xMTS-UGI (Inteins)-DdCBE backbone vectors, the intein Intein was removed in situ to obtain four 2xMTS-UGI (Splits)-DdCBE backbone vectors.

[0017] As a preferred embodiment of the present invention, the specific construction steps of the mitochondrial localization skeleton vector are as follows:

[0018] (1) Based on the principle that the highest splicing efficiency is achieved when cysteine, serine or threonine is connected to the C-terminus of the intein, the glycosylase inhibitor UGI protein cleavage site, i.e., UGI (S38), is selected, and the corresponding gene fragment is then constructed;

[0019] (2) Amplify the required MTS sequence by PCR;

[0020] (3) The original DdCBE backbone vector MTS-TALE was digested with NheⅠ and Pme I to obtain a DNA backbone fragment with homologous sequences to the halfDddA fragment and the UGI-C gene fragment at both ends;

[0021] (4) Design primers with appropriate homology arms and perform PCR amplification on the half DddA fragment G1333N / G1333C or G1397N / G1397C, the UGI fragments UGI-N, UGI-C, the intein Intein, IRES (SEQ ID NO: 28) and MTS, so that the two ends of the half DddA fragment have at least 15 bp of homology sequence with the MTS-TALE backbone after restriction digestion in (3) and with UGI-N, and the two ends of UGI-N have at least 15 bp of homology sequence with the half There is at least 15 bp of homology between the DddA fragment and the N-terminal Intein-N of the intein, there is at least 15 bp of homology between the N-terminal Intein-N of the intein and the IRES element, the amplified MTS-1 and the next MTS-2 are connected by amino acid GS, and at least 15 bp of homology between the previous MTS-1 and the IRES element, at least 15 bp of homology between the next MTS-2 and the C-terminus of the intein and between it and UGI-C, and at least 15 bp of homology between UGI-C and the MTS-TALE backbone after enzyme digestion in (3); wherein the amino acid sequence of half DddA fragment-G1333C is shown in SEQ ID NO: 29, the amino acid sequence of half DddA fragment-G1333N is shown in SEQ ID NO: 31, the amino acid sequence of DddA fragment-G1397C is shown in SEQ ID NO: 33, and the amino acid sequence of half DddA fragment-G1397N is shown in SEQ ID NO: NO: 35, the N-terminal Intein-N amino acid sequence of the intein is shown in SEQ ID NO: 15, and the C-terminal Intein-C amino acid sequence of the intein is shown in SEQ ID NO: 17; the UGI-N amino acid sequence is shown in SEQ ID NO: 19, and the UGI-C amino acid sequence is shown in SEQ ID NO: 21;

[0022] (5) homologously recombining the MTS-TALE backbone fragment, half DddA fragment, UGI-N, Intein-N, IRES, MTS-1, MTS-2, Intein-C and UGI-C fragments after enzyme digestion in (3);

[0023] (6) The products of multi-fragment recombination were transferred into DB3.1 competent cells with resistance to ccdb toxin and screened with ampicillin. Single clones were selected for Sanger sequencing verification to obtain the correct 4 transitional backbone vectors;

[0024] (7) Based on the above four transitional backbone vectors, the required CMV promoter fragment and MTS sequence are amplified by PCR, so that there is at least 15 bp of homologous sequence between the amplified CMV promoter fragment and the backbone, MTS-3 and the next MTS-4 are connected by amino acid GS, and at least 15 bp of homologous sequence exists between the previous MTS-3 and the CMV promoter fragment, and at least 15 bp of homologous sequence exists between the next MTS-4 and the backbone;

[0025] (8) using Not I and Nde I to digest the transitional backbone vector described in step (6) to obtain a DNA backbone fragment having homologous sequences at both ends to the CMV promoter fragment and the MTS-4 gene fragment;

[0026] (9) The backbone fragment, CMV promoter fragment, amplified MTS-3 and MTS-4 fragments are homologously recombined, and four correct 2xMTS-UGI (Inteins) -DdCBE backbone vectors are obtained according to the above step (6); according to the same method, the inteins in the four 2xMTS-UGI (Inteins) -DdCBE backbone vectors are removed in situ to obtain four 2xMTS-UGI (Splits) -DdCBE backbone vectors; wherein the MTS-1, MTS-2, MTS-3, and MTS-4 are selected from the MTS described in any one of claims 1 to 2.

[0027] As a preferred embodiment of the present invention, the nucleotide sequence of the DdCBE original backbone vector MTS-TALE is shown in SEQ ID NO: 23.

[0028] As a further preferred embodiment of the present invention, the four 2xMTS-UGI(Inteins)-DdCBE backbone vectors are: 2xMTS-ccdb-G1333N-(UGI-N)-(Intein-N)-IRES-2xMTS-(Intein-C)-(UGI-C), 2xMTS-ccdb-G1333C-(UGI-N)-(Intein-N)-IRES-2xMTS-(Intein-C)-(UGI-C),

[0029] 2xMTS-ccdb-G1397N-(UGI-N)-(Intein-N)-IRES-2xMTS-(Intein-C)-(UGI-C),

[0030] 2xMTS-ccdb-G1397C-(UGI-N)-(Intein-N)-IRES-2xMTS-(Intein-C)-(UGI-C);

[0031] The four 2xMTS-UGI(Splits)-DdCBE backbone vectors are: 2xMTS-ccdb-G1333N-(UGI-N)-IRES-2xMTS-(UGI-C), 2xMTS-ccdb-G1333C-(UGI-N)-IRES-2xMTS-(UGI-C),

[0032] 2xMTS-ccdb-G1397N-(UGI-N)-IRES-2xMTS-(UGI-C),

[0033] 2xMTS-ccdb-G1397C-(UGI-N)-IRES-2xMTS-(UGI-C).

[0034] A mitochondrial base editor with extremely low nuclear off-target, comprising an RVD library and the mitochondrial localization backbone vector described in the present invention. All backbone vectors of the present invention are matched with a TALE assembly system. The backbone vector toxin gene ccdb contains Bsa I restriction sites on both sides. After restriction cleavage, different sticky ends are generated for specific connection to TALE sequences that recognize specific DNA sequences. The assembly method is described in the authorized patent: A mitochondrial DNA editing system based on TALE assembly, patent number ZL202110688797.9.

[0035] A mitochondrial base editing system, comprising the mitochondrial base editor described in the present invention, or comprising the mitochondrial base editor described in the present invention and a DddI plasmid.

[0036] As a preferred embodiment of the present invention, the DddI plasmid construction method is as follows: the DdCBE original backbone vector MTS-TALE is digested with Nde I and Pme I, primers with suitable homology arms are designed, and the CMV promoter fragment and the DddI gene fragment (SEQ ID NO: 27) are amplified, so that there is at least 15bp of homology sequence between the CMV promoter fragment and the DddI protein fragment and the backbone after digestion, and there is at least 15bp of homology sequence between the CMV promoter fragment and the DddI protein fragment. The backbone fragment, CMV promoter fragment and DddI protein fragment are homologously recombined, and the product of multi-fragment recombination is transferred into DB3.1 competent cells with ccdb toxin resistance and screened with ampicillin, and a single clone is selected for Sanger sequencing verification to obtain the correct DddI plasmid.

[0037] The application of the mitochondrial localization backbone vector, the mitochondrial base editor or the mitochondrial base editing system described in the present invention in the C / G to T / A editing of mitochondrial DNA does not include a method for treating a disease.

[0038] The present invention selects the corresponding two mitochondrial base editor vectors described in the system of the present invention for pairing according to the principle of pairing G1333N and G1333C, and G1397N and G1397C, transfects them into the cells to be edited, and screens the successfully transfected cells with puromycin to achieve DNA editing from C / G to T / A.

[0039] The use of the mitochondrial localization skeleton vector, the mitochondrial base editor or the mitochondrial base editing system described in the present invention in preparing a C / G to T / A editing kit for mitochondrial DNA.

[0040] A kit for C / G to T / A editing of mitochondrial DNA, comprising the mitochondrial base editor or the mitochondrial base editing system.

[0041] The method for using the mitochondrial base editor of the present invention is not used for the treatment of diseases, and is characterized in that it comprises the following steps:

[0042] Step 1. According to the DNA sequence of the site to be edited, the TALE sequence for recognizing the DNA on both sides of the editing site is designed, and the four backbone vectors in the mitochondrial base editor 2xMTS-UGI (Inteins) -DdCBE or 2xMTS-UGI (Splits) -DdCBE are selected, and the TALE sequences are respectively connected to the backbone vectors by the Golden Gate cloning method to assemble the mitochondrial base editor vector; after the assembled vector is transformed into DH5α competent state, monoclonal PCR verification, Sanger sequencing and other steps, a correctly assembled mitochondrial base editor vector can be obtained, and the mitochondrial base editor plasmid can be further obtained by extracting the plasmid;

[0043] Step 2. Transfect the correctly assembled mitochondrial base editor plasmid into the cells to be edited in a Half DddA pairing manner; or co-transfect the mitochondrial base editor plasmid and DddI plasmid into the cells to be edited in a molar ratio of 1:1 between a single mitochondrial base editor plasmid and a DddI plasmid, screen the successfully transfected cells with 1 μg / mL puromycin, verify the editing efficiency of the mitochondrial base editor by Sanger sequencing or high-throughput sequencing, further culture the edited cells to obtain cells with stable mitochondrial DNA mutations, and perform subsequent functional analysis; wherein the Half DddA pairing manner is: G1333C and G1333N pairing, G1397C and G1397N pairing.

[0044] The working principle of the present invention is as follows: after the mitochondrial base editing system is transfected into cells, the editing site can be edited from C / G to T / A, mainly by deaminating the target base C to form dU through double-stranded DNA cytosine deaminase, and further editing from C / G to T / A is achieved through cell replication.

[0045] Beneficial effects:

[0046] In the present invention, the MTS derived from genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD replace the original MTS in the DdCBE system, so that the nuclear off-target of the replaced system is reduced. The above seven MTSs are connected in series to 2xMTS, and the linker is connected in the middle. At this time, the nuclear off-target of the modified 2xMTS-DdCBE is further reduced, and the mitochondrial editing efficiency is improved. In the present invention, 8 positions of UGI are selected to insert inteins respectively, wherein the inteins are split at UGI (S38) to obtain the best inhibitory effect of nuclear off-target, and it is found that UGI may have self-assembly, thereby developing a direct split form (Splits) of UGI (S38). Therefore, we combined 2xMTS with UGI (Splits) / UGI (Inteins) to form a new mitochondrial base editor skeleton, and the MTS therein can be replaced with any one of the 7 MTSs.

[0047] The mitochondrial base editor described in the present invention is used for C / G to T / A editing of mitochondrial DNA, and in particular can achieve extremely low nuclear off-target editing.

[0048] The mitochondrial base editing system described in the present invention can be used to explore the function of mitochondrial DNA mutations. At the same time, the tools of the present invention can be used to repair mutated DNA in cells.

[0049] In order to achieve C / G to T / A editing of mitochondrial DNA, TALE-DddAtox reported by Mok BY et al. can achieve C / G to T / A editing in the "tC" sequence on mitochondrial DNA (Nature. 2020 Jul; 583 (7817): 631-637.). The mitochondrial base editor with extremely low nuclear off-target editing developed by the present invention can greatly reduce its nuclear off-target editing without affecting its mitochondrial DNA editing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 Schematic diagram of the 8 MTS backbone vectors of the mitochondrial base editing system with extremely low nuclear off-target editing.

[0051] Figure 1 It is a Burkholderia cenocepacia-DddAtox with different segmentation types and a mitochondrial localization signal (MTS). The vector carries the ccdb gene element, which is located between two Bsa I restriction sites and carries an ampicillin resistance element. Based on the DdCBE backbone, its MTS is changed to 2xMTS connected with a linker in the middle, and its UGI is cut at S38. The backbone vector of the UGI intein split (Inteins) form is UGI-N connected to Intein-N, and UGI-C is connected to Intein-C and 2xMTS in sequence. The split halves of UGI are connected by IRES. The backbone vector of the direct split (Splits) form of UGI is (UGI-N) connected to IRES-2xMTS-(UGI-C).

[0052] Figure 2 MTS screening process and tandem strategy.

[0053] Figure 2 In order to select 78 MTS from the MitoCarta3.0 database for the first round of efficiency screening and the second round of nuclear off-target value screening, 7 MTS signals with strong mitochondrial targeting ability and low nuclear off-target were finally selected, and the 1xMTS in the original DdCBE backbone vector was made into 2xMTS after concatenation.

[0054] Figure 3 MTS first and second round screening results.

[0055] Figure 3 a is to construct the DdCBE plasmid pair targeting the mitochondrial DNA-G3733 site. Figure 278 MTS signals selected from the database replaced the MTS in the G3733 site-DdCBE system. After the cells were transfected with the DdCBE system, a library was constructed to detect its editing efficiency at the G3733 site. 33 MTS signals with an efficiency greater than or equal to 50% of the WT-DdCBE efficiency were selected. The dotted line is the 50% efficiency line of the WT-DdCBE. Figure 3 b is to construct a DdCBE plasmid pair targeting the mitochondrial DNA-G3733 site and a T vector with DNA-G3733 site information. Figure 3 The 33 MTS signals screened in a replaced the MTS in the G3733 site-DdCBE system, and the DdCBE system was co-transfected with Tm.G3733 to build a library to detect its editing efficiency at the G3733 site, and the 7 MTSs with the highest efficiency on mitochondrial DNA and the lowest off-target value on the T vector were selected. Figure 4 2xMTS-DdCBE editing system edits the mitochondrial m.3733 site.

[0056] Figure 4 The 1xMTS-DdCBE system and the 2xMTS-DdCBE system for the 7 screened MTSs, namely, the genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD, were used to compare the mitochondrial editing, T vector off-target, and the ratio of the two at the m.3733 site. All of the MTSs were further able to improve the mitochondrial editing efficiency, reduce T vector off-target, and improve the ratio of the two after 2xMTS treatment. Among them, the improvement effect of SDHD was the most obvious, that is, the 2xSDHD-DdCBE system had a lower nuclear off-target value than the 1xSDHD-DdCBE system.

[0057] Figure 52 shows the editing status of the mitochondrial m.13513 site by the xMTS-DdCBE editing system.

[0058] Figure 5 The 1xMTS-DdCBE system and the 2xMTS-DdCBE system for the 7 screened MTSs, namely, the genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD, were used to compare the mitochondrial editing of the m.13513 site, the off-target status of the T vector, and the ratio of the two. Among them, almost all MTSs can further improve the ratio of the two after 2xMTS treatment. Among them, the improvement effect of SDHD is the most obvious, that is, the nuclear off-target value of the 2xSDHD-DdCBE system is lower than that of the 1xSDHD-DdCBE system.

[0059] Figure 6 Schematic representation of intein cleavage of UGI.

[0060] Figure 6The DdCBE system for intein cleavage of UGI cannot produce good recognition, cleavage and hydrolysis of intein outside mitochondria due to the presence of MTS signal; inside mitochondria, due to the hydrolysis of MTS signal, the N-terminus and C-terminus of the intein recognize, bind and hydrolyze, and the UGI split halves on both sides are connected by peptide bonds, and the DdCBE system assembles into an active structure and exerts its function.

[0061] Figure 7 Schematic diagram of different carriers of DddA and UGI components in the intein-cleaved DdCBE system.

[0062] Figure 7 Vector schematic diagram of the intein splitting of the G1397N half, G1397C half or UGI of DddA. C1 is the intein splitting of the G1397N half, C2 is the intein splitting of the G1397C half, C3 is the intein splitting of the UGI of G1397N-UGI, C4 is the intein splitting of the UGI of G1397C-UGI, and C5 is the intein splitting of the UGI of G1397N-UGI and the UGI of G1397C-UGI at the same time. C6-C10 corresponds to C1-C5 one by one. The plasmids with IRES in C1-C5 are removed from IRES and divided into two plasmids, which is the C6-C10 combination. The combination of L-G1397N-UGI (Splits) + R-G1397C-UGI-(Splits) is a combination of directly splitting the UGI of G1397N-UGI and the UGI of G1397C-UGI at the same time, namely UGI (Splits).

[0063] Figure 8 Editing of intein splitting UGI and UGI direct splitting.

[0064] Figure 8 a is the mitochondrial editing, T vector off-target and the ratio of the different combinations of DddA and UGI in the intein-cutting DdCBE system at the mitochondrial m.3733 site, among which the C5 combination can achieve high target site editing and low nuclear off-target, with a high ratio effect; Figure 8 b is the amino acid sequence of UGI and the cleavage positions of the eight selected UGIs as the source of the segmentation positions in the figure below; Figure 8 c is the mitochondrial editing status, T vector off-target status and the ratio of the eight cutting positions of UGI at the mitochondrial m.3733 site, among which S38 can achieve the effect of high target site editing and low nuclear off-target, with a high ratio; Figure 8 d is the DdCBE system directly split by UGI. Due to the presence of MTS signal outside the mitochondria, UGI cannot produce good self-assembly; inside the mitochondria, due to the hydrolysis of MTS signal, the UGI split half undergoes self-assembly, and the DdCBE system assembles into an active structure and exerts its function.

[0065] Fig. 9 Editing of UGI intein split and direct split forms.

[0066] Fig. 9 The results show the editing status of UGI's inteins and split forms on mitochondria at the m.3733 and m.13513 sites, the off-target status of T vectors, and the ratio of the two. It is confirmed that at different sites and different DddA-half protein combinations, both the inteins and split forms of UGI can reduce nuclear off-targets and increase the ratio of the two while basically maintaining the editing of the target site.

[0067] Fig.10 Editing of the mitochondrial m.3733 site by the mitochondrial base editing system with extremely low nuclear off-target editing.

[0068] Fig.10 In order to transform the DdCBE system skeleton into 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins), and load the 7 selected MTS signals into the corresponding MTS positions, the editing of the mitochondrial m.3733 site, the off-target of the T vector and the ratio of the two were analyzed when the DddI plasmid was co-transfected with or without different combinations. When there was no DddI plasmid co-transfection, the combination of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) could basically reduce the nuclear off-target and increase the ratio of the two while basically maintaining the editing of the target site. Compared with the high nuclear off-target produced by the control, the optimized structure produced extremely low nuclear off-target. When the DddI plasmid was co-transfected, the nuclear off-target values ​​of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) were reduced to the background level.

[0069] Fig.11 Editing of the mitochondrial m.8363 site by the mitochondrial base editing system with extremely low nuclear off-target editing.

[0070] Fig.11In order to transform the DdCBE system skeleton into 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins), and load 2 of the 7 MTS signals screened into the corresponding MTS positions, the editing of the mitochondrial m.8363 site, the off-target of the T vector, the ratio of target editing to T vector off-target, the off-target of nuclear pseudogenes, and the ratio of target editing to nuclear pseudogene off-target were analyzed in different combinations with or without DddI plasmid co-transfection. When there is no DddI plasmid co-transfection, the combination of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) can basically reduce nuclear off-target and increase the ratio of the two while basically maintaining the editing of the target site. Compared with the high nuclear off-target produced by the control, the optimized structure produces extremely low nuclear off-target. When DddI plasmid was co-transfected, the nuclear off-target values ​​of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) were reduced to the background level.

[0071] Fig.12 Editing of the mitochondrial m.13513 site by the mitochondrial base editing system with extremely low nuclear off-target editing.

[0072] Fig.12 In order to transform the DdCBE system skeleton into 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins), and load 2 of the 7 MTS signals screened into the corresponding MTS positions, the editing situation at the mitochondrial m.13513 site, the off-target situation of the T vector and the ratio of the two were compared when the DddI plasmid was co-transfected with different combinations.

[0073] The (Inteins) combination can basically reduce nuclear off-target and improve the ratio of the two while maintaining the target site editing. Compared with the high nuclear off-target produced by the control, the optimized structure produces extremely low nuclear off-target.

[0074] The nuclear off-target value of (Inteins) was reduced to the background level. DETAILED DESCRIPTION

[0075] The present invention is described in detail below in conjunction with the examples, so as to fully understand and implement how to apply the present invention to achieve C / G to T / A editing of DNA while producing extremely low off-target editing in the nucleus.

[0076] The present invention is further described below with reference to the accompanying drawings. The present invention provides a novel DNA editing tool: a mitochondrial base editor that can achieve C / G to T / A editing with extremely low nuclear off-target editing, the system comprising: 8 mitochondrial-localized backbone vectors (such as Figure 1 ), DdCBE localized in mitochondria can be selected to edit mitochondrial DNA. The RVD library can select any library that is compatible with the editing system. The library used in the embodiment of the present invention is shown in CellDiscov.2021Sep 3; 7(1):78, and the assembly method is shown in the authorized patent: A mitochondrial DNA editing system based on TALE assembly, patent number ZL202110688797.9. The steps, reagents, modules, etc. not described in detail in the embodiments are well known to those skilled in the art and will not be repeated in this application.

[0077] Example 1

[0078] DdCBE was used to perform the first round of efficiency screening and the second round of nuclear off-target value screening on 78 MTS selected from the MitoCarta3.0 database, and finally 7 MTS signals with strong mitochondrial targeting ability and low nuclear off-target were selected, and the 1xMTS in the original DdCBE backbone vector was connected in series to become 2xMTS. The 2xMTS modification can inhibit nuclear off-target to a greater extent here. The UGI (Inteins) and UGI (Splits) structures that can maximally inhibit the nuclear off-target of DdCBE were obtained by splitting the inteins of the UGI protein in DdCBE (i.e., UGI-Inteins) and directly splitting UGI (i.e., UGI-Splits). The 2xMTS obtained in the above two modifications were assembled and combined with UGI (Inteins) / UGI (Splits) to form a new DdCBE structure of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins). The novel DdCBE efficiently edited mitochondrial DNA in human 293FT cells and could greatly inhibit the nuclear off-target of DdCBE when there was no co-expression of DddI plasmid; when co-transfected with DddI plasmid at a 1:1 molar ratio, the nuclear off-target was reduced to the background level.

[0079] (I) Design of TALE sequence and DdCBE assembly for MTS signal screening

[0080] Select the m.3733 and m.13513 sites on human mitochondrial DNA (reference sequence version: NC_012920.1) for editing (such as Figure 3-5As shown in the figure, the TALE sequence to be recognized was designed according to the DNA sequence near the site, where the left TALE recognition sequence of m.3733 was AGCCCAAACAATCT and the right TALE recognition sequence was AGTAGAATGATGGCT; the left TALE recognition sequence of m.13513 was CCTCACAGGTTTCTACT and the right TALE recognition sequence was GTTTGCGGTTTCGAT. The RVD modules of different recognized DNA bases (Cell Discov. 2021Sep 3; 7(1):78) were assembled into four mitochondrial-localized DdCBE backbone vectors by the Golden Gate method (the backbone was obtained by digesting the original DdCBE backbone vector with NheⅠ and PmeⅠ, and then homologously recombined with G1333C, G1333N, G1397C, and G1397N fragments in sequence). The assembly method is shown in the authorized patent: A mitochondrial DNA editing system based on TALE assembly, patent number ZL202110688797.9.

[0081] The assembled product was transformed into DH5α competent cells and screened on a solid LB culture plate containing ampicillin. After overnight culture, a single clone was picked and identified by PCR using the following primers. The primer sequences are shown in Seq ID NO: 24 and 25. According to the number of RVDs, the positive clones were judged, such as 16 RVDs should be 1759 bp in size, 15 RVDs should be 1657 bp in size, etc. Then the above primers were used for forward and reverse sequencing, and the positive clones with correct sequences were amplified and plasmids were extracted. According to the experimental requirements of the site, the L-G1397N+R-G1397C combination with the highest editing efficiency at the m.3733 site was selected for subsequent screening.

[0082] The 78 MTSs selected from the MitoCarta3.0 database were used to replace the MTSs in the L-G1397N+R-G1397C combination with the highest editing efficiency at the m.3733 site obtained above, mainly by digesting the original plasmid with NdeⅠ and NotⅠ and then homologously recombining with the fragments of each MTS and the CMV promoter fragment to obtain a DdCBE plasmid pair targeting the m.3733 site loaded with 78 MTSs. The T vector skeleton (shown in SEQ ID NO: 37) was digested with XcmⅠ, and the skeleton was connected to the synthesized DNA base sequence of the m.3733 site by T4 to obtain the T vector of the m.3733 site, namely the Tm.G3733 plasmid, and the Tm.G13513 plasmid was obtained in the same way.

[0083] (II) Transfect cells to detect the editing efficiency of different combinations of DdCBE and select the best MTS signal

[0084] Select 293FT cells as the gene editing object. When the cell density reaches 80-90%, replace the culture medium 2 hours in advance. After counting the cells, take 2x10 5 The cells were used for cell electroporation (instrument: Lonza 4D-nucleofector).

[0085] In the first round of screening, the L-G1397N+R-G1397C plasmid pair with 78 MTS targeting the m.3733 site and the control L-G1397N+R-G1397C plasmid pair (MTS is ATP5F1B) were combined and electroporated into 293FT cells in the form of plasmid pairs. 400ng of the vectors at the left and right ends of each repeat were transfected, and the cells were electroporated using the SF-Cell Line 4D-nucleofector X-kit, with 3 replicates per group. After electroporation, the cells were inoculated into 12-well plates and cultured for 24 hours, followed by cell screening using 1μg / mL of puromycin. After puromycin selection, the cells were cultured for 72 hours, the cells were digested and centrifuged, and the cells were collected for testing the DNA mutation efficiency.

[0086] The DNA extraction method of the collected cell samples is as follows: 50 μL of QuickExtract TM The cells were resuspended in DNA Extraction Solution (Lucigen), then heated at 65°C for 60 min, vortexed and centrifuged, and heated at 98°C for 5 min. Green Taq Mix (Vazyme) was used to amplify the target fragment containing the editing site on the mitochondrial DNA. The amplified product can be used for Sanger sequencing; or Phanta Super-Fidelity DNA polymerase (Vazyme) and high-throughput sequencing library primers were used for high-throughput sequencing to analyze the editing efficiency. The analysis results showed that there were 33 MTS signals with an efficiency greater than or equal to 50% of the WT-DdCBE efficiency. The dotted line is the 50% efficiency line of WT-DdCBE ( Figure 3 a).

[0087] In the second round of screening, the L-G1397N+R-G1397C plasmid pair targeting the m.3733 site with the above 33 MTSs and the control L-G1397N+R-G1397C plasmid pair (MTS is ATP5F1B) were combined and electroporated into 293FT cells in the form of DdCBE plasmid pair plus Tm.G3733 plasmid. The electroporation process and sample library construction process are the same as Figure 3The process described in a, note that 50ng of T vector (here Tm.G3733 plasmid) was transfected in each replicate. Analyze the results and select the 7 MTSs with the highest efficiency on mitochondrial DNA and the lowest off-target value on T vector ( Figure 3 b) are mitochondrial targeting signals MTS from genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD, respectively, with amino acid sequences as shown in SEQ ID NOs: 1-7, and nucleotide sequences as shown in SEQ ID NOs: 8-14. Figure 3 The DdCBE ligation method of the m.3733 site in a was used to connect the 7 MTSs at the m.13513 site and the DdCBE plasmid pair of the control MTS. These 7 MTSs and the control MTS were connected in series to form 2xMTS. The mitochondrial editing, T vector off-target and ratio of the 1xMTS-DdCBE system and the 2xMTS-DdCBE system were compared at the m.3733 site and the m.13513 site. The electroporation process and sample library construction process are the same Figure 3 b. The analysis results showed that almost all MTSs could further improve the ratio of the two after 2xMTS treatment, among which SDHD had the most obvious improvement ( Figure 4-5 ).

[0088] (III) Design of TALE sequence and DdCBE assembly during UGI splitting

[0089] Select the m.3733 and m.13513 sites on human mitochondrial DNA (reference sequence version: NC_012920.1) for editing (such as Figure 8-9As shown), according to the DNA sequence near the site, the recognition TALE sequence was designed, among which the left TALE recognition sequence of m.3733 is AGCCCAAACAATCT, and the right TALE recognition sequence is AGTAGAATGATGGCT; the left TALE recognition sequence of m.13513 is CCTCACAGGTTTCTACT, and the right TALE recognition sequence is GTTTGCGGTTTCGAT. The RVD modules that recognize different DNA bases (Cell Discov. 2021 Sep 3; 7(1): 78) were assembled onto four mitochondrial-localized DdCBE backbone vectors by the Golden Gate method (the backbone was obtained by digesting the original DdCBE backbone vector with NheⅠ and PmeⅠ, and then homologously recombined with G1333C (SEQ ID NO: 30), G1333N (SEQ ID NO: 32), G1397C (SEQ ID NO: 34), and G1397N (SEQ ID NO: 36) fragments in sequence). The assembly method can be found in the authorized patent: A mitochondrial DNA editing system based on TALE assembly, patent number ZL202110688797.9.

[0090] The assembled product was transformed into DH5α competent cells and screened on a solid LB culture plate containing ampicillin. After overnight culture, a single clone was picked and identified by PCR using the following primers. The primer sequences are shown in Seq ID NO: 24 and 25. According to the number of RVDs, the positive clones were judged, such as 16 RVDs should be 1759 bp, 15 RVDs should be 1657 bp, etc. Then, the above primers were used for forward and reverse sequencing, and the positive clones with correct sequences were amplified to extract plasmids.

[0091] right Figure 3-5The DddA protein and UGI protein in the DdCBE (MTS is 1x control MTS) plasmid at the m.3733 site obtained were split using intein peptides, resulting in ten splitting combinations C1-C10. C1 was for splitting the G1397N half by intein peptides, C2 was for splitting the G1397C half by intein peptides, C3 was for splitting the UGI of G1397N-UGI by intein peptides, C4 was for splitting the UGI of G1397C-UGI by intein peptides, and C5 was for splitting both the UGI of G1397N-UGI and the UGI of G1397C-UGI by intein peptides. The C1-C5 split combination is obtained by homologous recombination connection of the backbone of the DdCBE plasmid at the m.3733 site after digestion with NheⅠ and Pme I, and then (add half DddA when there is no half DddA) split protein-N, Intein-N, IRES, MTS, Intein-C, split protein-C (add UGI when there is no UGI). C6-C10 corresponds to C1-C5 one by one. The C1-C5 plasmid with IRES is removed from the IRES and divided into two plasmids to form the C6-C10 combination ( Figure 7 ). The UGI cleavage position test plasmid is obtained by replacing the UGI part with the C5 split combination plasmid pair. The C5 split combination plasmid pair (i.e., UGI (Inteins)) removes Intein-N and Intein-C in situ to obtain the UGI direct split combination (i.e., UGI (Splits)). The plasmid related to the m.13513 site can be obtained by the connection method of the m.3733 site.

[0092] (IV) Transfect cells to test the editing efficiency of different combinations of DdCBE and select the best split method

[0093] according to Figure 3-5 The cell transfection method and sample library construction process in the article were used to transfect and build the library for the C1-C10 plasmid pair (co-transfected with Tm.G3733 plasmid). The analysis results showed that the C5 combination can achieve high target site editing with low nuclear off-target and high ratio effect ( Figure 8 a) According to Figure 3-5 The cell transfection method and sample library construction process in the paper were used to transfect and construct libraries for plasmid pairs with different cutting positions of UGI. The analysis results showed that the S38 site splitting can achieve high target site editing with low nuclear off-target and high ratio effect ( Figure 8 c) According to Figure 3-5The cell transfection method and sample library construction process in the m.3733 site and the m.13513 site were transfected and library constructed with plasmid pairs of UGI's inteins and direct split forms. The results showed that at different sites and different DddA-half protein combinations, both the inteins and direct split forms of UGI could reduce nuclear off-target and increase the ratio of the two while basically maintaining target site editing ( Fig. 9 ).

[0094] (V) Design of TALE sequences of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) and assembly of DdCBE

[0095] Select the m.3733, m.8363, and m.13513 sites on human mitochondrial DNA (reference sequence version: NC_012920.1) for editing (e.g. Figure 10-12 As shown), according to the DNA sequence near the site, the recognition TALE sequence was designed, among which the left TALE recognition sequence of m.3733 is AGCCCAAACAATCT, and the right TALE recognition sequence is AGTAGAATGATGGCT; the left TALE recognition sequence of m.8363 is AAGAGAACCAACACCTCT, and the right TALE recognition sequence is ACGGTAGTATTTAGT; the left TALE recognition sequence of m.13513 is CCTCACAGGTTTCTACT, and the right TALE recognition sequence is GTTTGCGGTTTCGAT. The RVD modules that recognize different DNA bases (Cell Discov. 2021 Sep 3; 7(1): 78) were assembled onto 8 mitochondrial-localized backbone vectors (4 2xMTS-UGI (Splits) and 4 2xMTS-UGI (Inteins)) using the Golden Gate method. The assembly method can be found in the authorized patent: A mitochondrial DNA editing system based on TALE assembly, patent number ZL202110688797.9.

[0096] The assembled product was transformed into DH5α competent cells and screened on a solid LB culture plate containing ampicillin. After overnight culture, a single clone was picked and identified by PCR using the following primers. The primer sequences are shown in Seq ID NO: 24 and 25. According to the number of RVDs, the positive clones were judged, such as 16 RVDs should be 1759 bp, 15 RVDs should be 1657 bp, etc. Then, the above primers were used for forward and reverse sequencing, and the positive clones with correct sequences were amplified to extract plasmids.

[0097] The MTS in the 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) plasmid pairs targeting the m.3733 site obtained above was replaced from the control MTS (ATP5F1B) to the MTS of the genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD, respectively. The MTS in the 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) plasmid pairs targeting the m.8363 and m.13513 sites obtained above was replaced from the control MTS (ATP5F1B) to the MTS of the genes NDUFAB1 and SDHD, respectively. The T vector backbone (shown in SEQ ID NO: 37) was digested with XcmⅠ, and the backbone was ligated with the synthesized DNA base sequence of the m.8363 site by T4 to obtain the T vector of the m.8363 site, namely, the Tm.G8363 plasmid.

[0098] (VI) Transfected cells were used to test the editing efficiency of different combinations of DdCBE, confirming the extremely low nuclear off-target level of the optimized DdCBE

[0099] according to Figure 3-5 The cell transfection method and sample library construction process in were used to transfect and construct libraries for the 7 screened MTS signals and the control MTS targeting m.3733 site 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) plasmid pairs (co-transfected with Tm.G3733 plasmid) with or without DddI plasmid co-transfection (the transfection molar ratio with a single DdCBE plasmid in the plasmid pair was 1:1). The analysis results showed that when there was no DddI plasmid co-transfection, the combination of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) could basically reduce nuclear off-target and increase the ratio of the two while basically maintaining target site editing. Compared with the high nuclear off-target produced by the control, the optimized structure produced extremely low nuclear off-target. When DddI plasmid was co-transfected, the nuclear off-target values ​​of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) were reduced to the background level ( Fig.10 ).

[0100] according to Figure 3-5 The cell transfection method and sample library construction process in the above Fig.10The two MTS signals NDUFAB1 and SDHD with the best comprehensive effects and the control MTS targeting m.8363 and m.13513 sites, 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) plasmid pairs (co-transfected with Tm.G3733 plasmid), were transfected and library constructed with or without DddI plasmid co-transfection (the transfection molar ratio with the single DdCBE plasmid in the plasmid pair was 1:1). The analysis results showed that when there was no DddI plasmid co-transfection, the combination of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) could basically reduce nuclear off-target and increase the ratio of the two while basically maintaining target site editing. Compared with the high nuclear off-target produced by the control, the optimized structure produced extremely low nuclear off-target. When DddI plasmid was co-transfected, the nuclear off-target values ​​of 2xMTS-UGI (Splits) and 2xMTS-UGI (Inteins) were reduced to the background level ( Figure 11-12 ).

[0101] Figure 10-12 The results showed that the optimized mitochondrial base editor with extremely low nuclear off-target editing can efficiently edit endogenous mitochondrial sites, and its nuclear off-target value is extremely low.

[0102] The N-terminal Intein-N nucleotide sequence of the intein involved in the above embodiments is shown in SEQ ID NO: 16, the C-terminal Intein-C nucleotide sequence of the intein is shown in SEQ ID NO: 18; the UGI-N nucleotide sequence is shown in SEQ ID NO: 20, and the UGI-C nucleotide sequence is shown in SEQ ID NO: 22; the DddI nucleotide sequence is shown in SEQ ID NO: 27; the IRES nucleotide sequence is shown in SEQ ID NO: 28, the half DddA fragment-G1333C nucleotide sequence is shown in SEQ ID NO: 30, the DddA fragment-G1333N nucleotide sequence is shown in SEQ ID NO: 32, the half DddA fragment-G1397C nucleotide sequence is shown in SEQ ID NO: 30, and the half DddA fragment-G1397N nucleotide sequence is shown in SEQ ID NO: 30.

[0103] The above description shows and describes the embodiments of the invention. In summary, the present invention provides a mitochondrial base editor with extremely low nuclear off-target editing. Experiments have confirmed that a modified mitochondrial DNA editor of the present invention: a mitochondrial base editor with extremely low nuclear off-target editing can successfully achieve G / C to A / T editing on human mitochondria and greatly reduce its nuclear off-target.

Claims

1. Improved mitochondrial targeting signal 2xMTS, characterized in that: The improved mitochondrial targeting signal is in the form of 2xMTS in which two MTS are connected in series, and the first MTS and the second MTS are connected by GS amino acid as a linker; the MTS is selected from any one of the mitochondrial targeting signal MTSs from genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD.

2. The improved mitochondrial targeting signal according to claim 1, characterized in that The amino acid sequences of the mitochondrial targeting signals MTS of genes COQ3, DBT, NDUFAB1, PCCA, PRDX3, SCO1, and SDHD are shown in SEQ ID NOs: 1-7, respectively.

3. A mitochondrial targeting backbone vector, characterized in that: The mitochondrial targeting skeleton vector contains the improved mitochondrial targeting signal according to claim 1 or 2.

4. The mitochondrial targeting skeleton vector according to claim 4, characterized in that: The mitochondrial localization backbone vector is selected from a pair of four 2xMTS-UGI (Splits) -DdCBE backbone vectors or a pair of four 2xMTS-UGI (Inteins) -DdCBE backbone vectors; the 2xMTS-UGI (Splits) -DdCBE backbone vector is a double-stranded DNA cytosine deaminase DddA domain from Burkholderia cenocepacia divided into half The four split protein pairs of DddA G1333N / G1333C and G1397N / G1397C were fused with the glycosylase inhibitor UGI and then inserted between the two restriction sites NheⅠ and PmeⅠ of the MTS-TALE backbone vector. On this basis, UGI was split at S38 with the intein Intein, and the 2xMTS was added in front of the C-terminus of the intein. The split half components of UGI were connected with IRES, thereby obtaining four 2xMTS-UGI (Inteins)-DdCBE backbone vectors; based on the previous four 2xMTS-UGI (Inteins)-DdCBE backbone vectors, the intein Intein was removed in situ to obtain four 2xMTS-UGI (Splits)-DdCBE backbone vectors.

5. The mitochondrial targeting skeleton vector according to claim 4, characterized in that: The specific construction steps of the mitochondrial localization skeleton vector are as follows: (1) Based on the principle that the highest splicing efficiency is achieved when cysteine, serine or threonine is connected to the C-terminus of the intein, the glycosylase inhibitor UGI protein cleavage site, i.e., UGI (S38), is selected, and the corresponding gene fragment is then constructed; (2) Amplify the required MTS sequence by PCR; (3) The DdCBE original backbone vector MTS-TALE was digested with NheⅠ and Pme I to obtain a DNA backbone fragment with homologous sequences to the half DddA fragment and the UGI-C gene fragment at both ends; (4) Design primers with appropriate homology arms and perform PCR amplification on the half DddA fragment G1333N / G1333C or G1397N / G1397C, the UGI fragments UGI-N, UGI-C, the intein Intein, IRES and MTS, so that the two ends of the half DddA fragment have at least 15 bp of homology sequence with the MTS-TALE backbone after restriction digestion in (3) and with UGI-N, and the two ends of UGI-N have at least 15 bp of homology sequence with the half There is at least 15 bp of homology between the DddA fragment and the N-terminal Intein-N of the intein, there is at least 15 bp of homology between the N-terminal Intein-N of the intein and the IRES element, the amplified MTS-1 and the next MTS-2 are connected by amino acid GS, and at least 15 bp of homology between the previous MTS-1 and the IRES element, at least 15 bp of homology between the next MTS-2 and the C-terminus of the intein and between it and UGI-C, and at least 15 bp of homology between UGI-C and the MTS-TALE backbone after enzyme digestion in (3); wherein the amino acid sequence of half DddA fragment-G1333C is shown in SEQ ID NO: 29, the amino acid sequence of half DddA fragment-G1333N is shown in SEQ ID NO: 31, the amino acid sequence of DddA fragment-G1397C is shown in SEQ ID NO: 33, and the amino acid sequence of half DddA fragment-G1397N is shown in SEQ ID NO: NO: 35, the N-terminal Intein-N amino acid sequence of the intein is shown in SEQ ID NO: 15, and the C-terminal Intein-C amino acid sequence of the intein is shown in SEQ ID NO: 17; the UGI-N amino acid sequence is shown in SEQ ID NO: 19, and the UGI-C amino acid sequence is shown in SEQ ID NO: 21; (5) homologously recombining the MTS-TALE backbone fragment, half DddA fragment, UGI-N, Intein-N, IRES, MTS-1, MTS-2, Intein-C and UGI-C fragments after enzyme digestion in (3); (6) The products of multi-fragment recombination were transferred into DB3.1 competent cells with resistance to ccdb toxin and screened with ampicillin. Single clones were selected for Sanger sequencing verification to obtain the correct 4 transitional backbone vectors; (7) Based on the above four transitional backbone vectors, the required CMV promoter fragment and MTS sequence are amplified by PCR, so that there is at least 15 bp of homologous sequence between the amplified CMV promoter fragment and the backbone, MTS-3 and the next MTS-4 are connected by amino acid GS, and at least 15 bp of homologous sequence exists between the previous MTS-3 and the CMV promoter fragment, and at least 15 bp of homologous sequence exists between the next MTS-4 and the backbone; (8) using Not I and Nde I to digest the transitional backbone vector described in step (6) to obtain a DNA backbone fragment having homologous sequences at both ends to the CMV promoter fragment and the MTS-4 gene fragment; (9) The backbone fragment, CMV promoter fragment, amplified MTS-3 and MTS-4 fragments are homologously recombined, and four correct 2xMTS-UGI (Inteins) -DdCBE backbone vectors are obtained according to the above step (6); according to the same method, the inteins in the four 2xMTS-UGI (Inteins) -DdCBE backbone vectors are removed in situ to obtain four 2xMTS-UGI (Splits) -DdCBE backbone vectors; wherein the MTS-1, MTS-2, MTS-3, and MTS-4 are selected from the MTS described in any one of claims 1 to 2.

6. The mitochondrial targeting skeleton vector according to claim 5, characterized in that: The nucleotide sequence of the DdCBE original backbone vector MTS-TALE is shown in SEQ ID NO:

23.

7. The mitochondrial targeting skeleton vector according to claim 6, characterized in that: The four 2xMTS-UGI(Inteins)-DdCBE backbone vectors are: 2xMTS-ccdb-G1333N-(UGI-N)-(Intein-N)-IRES-2xMTS-(Intein-C)-(UGI-C), 2xMTS-ccdb-G1333C-(UGI-N)-(Intein-N)-IRES-2xMTS-(Intein-C)-(UGI-C), 2xMTS-ccdb-G1397N-(UGI-N)-(Intein-N)-IRES-2xMTS-(Intein-C)-(UGI-C), 2xMTS-ccdb-G1397C-(UGI-N)-(Intein-N)-IRES-2xMTS-(Intein-C)-(UGI-C); The four 2xMTS-UGI(Splits)-DdCBE backbone vectors are: 2xMTS-ccdb-G1333N-(UGI-N)-IRES-2xMTS-(UGI-C), 2xMTS-ccdb-G1333C-(UGI-N)-IRES-2xMTS-(UGI-C), 2xMTS-ccdb-G1397N-(UGI-N)-IRES-2xMTS-(UGI-C), 2xMTS-ccdb-G1397C-(UGI-N)-IRES-2xMTS-(UGI-C).

8. A mitochondrial base editor with extremely low nuclear off-target, characterized in that: It includes an RVD library and a mitochondrial localization backbone vector according to any one of claims 3 to 7, all backbone vectors are matched to a TALE assembly system, and the backbone vector toxin gene ccdb contains Bsa I restriction sites on both sides, and different sticky ends are generated after restriction digestion for specific connection to a TALE sequence that recognizes a specific DNA sequence.

9. A mitochondrial base editing system, characterized in that Comprising the mitochondrial base editor of claim 8, or comprising the mitochondrial base editor of claim 8 and a DddI plasmid.

10. Use of the mitochondrial localization backbone vector described in any one of claims 3 to 7, the mitochondrial base editor described in claim 8, or the mitochondrial base editing system described in claim 9 in C / G to T / A editing of mitochondrial DNA, wherein the use does not include a method for treating a disease.

11. Use of the mitochondrial localization backbone vector described in any one of claims 3 to 7, the mitochondrial base editor described in claim 8, or the mitochondrial base editing system described in claim 9 in preparing a C / G to T / A editing kit for mitochondrial DNA.

12. A kit for editing mitochondrial DNA from C / G to T / A, characterized in that: The kit contains the mitochondrial base editor described in claim 8 or the mitochondrial base editing system described in claim 9.

13. The method for using the mitochondrial base editor of claim 8, wherein the method is not used for the treatment of a disease, and is characterized in that: The following steps are involved: Step 1. According to the DNA sequence of the site to be edited, the TALE sequence for recognizing the DNA on both sides of the editing site is designed, and the four backbone vectors in the mitochondrial base editor 2xMTS-UGI (Inteins) -DdCBE or 2xMTS-UGI (Splits) -DdCBE are selected, and the TALE sequences are respectively connected to the backbone vectors by the Golden Gate cloning method to assemble the mitochondrial base editor vector; after the assembled vector is transformed into DH5α competent state, monoclonal PCR verification, Sanger sequencing and other steps, a correctly assembled mitochondrial base editor vector can be obtained, and the mitochondrial base editor plasmid can be further obtained by extracting the plasmid; Step 2. Transfect the correctly assembled mitochondrial base editor plasmid into the cells to be edited in a Half DddA pairing manner; or co-transfect the mitochondrial base editor plasmid and DddI plasmid into the cells to be edited in a molar ratio of 1:1 between a single mitochondrial base editor plasmid and a DddI plasmid, screen the successfully transfected cells with 1 μg / mL puromycin, verify the editing efficiency of the mitochondrial base editor by Sanger sequencing or high-throughput sequencing, further culture the edited cells to obtain cells with stable mitochondrial DNA mutations, and perform subsequent functional analysis; wherein the Half DddA pairing manner is: G1333C and G1333N pairing, G1397C and G1397N pairing.

Images