METHODS AND MEANS FOR GENETIC ALTERATION OF GENOMES UTILIZING DESIGNER DNA RECOMBINING ENZYMES

Abstract

The invention provides methods and means for specifically altering the DNA sequence in a genome, in particular for genome editing by deleting or replacing a sequence of interest. Advantageously, the invention uses two non-identical sequences naturally occurring in a genome as target sites two which DNA-recombining enzymes are generated. The invention is in particular useful for medicine, in particular to repair a mutation in a genome or to delete predefined genetic material from cells or tissue and to cure diseases. An advantage of the invention is that it allows precise site directed altering of DNA without engaging host DNA repair pathways and thereby works without inducing random insertions and deletions (in-dels).

Claims

1-26. (canceled)

27. A method for site-specific DNA recombination, comprising introducing into a host cell a designer DNA-recombining enzyme, wherein the designer DNA-recombining enzyme comprises at least two different monomers, wherein the at least two different monomers recognize non-identical naturally occurring half sites in a genome of a host cell.

28. The method of claim 27, wherein the recombination causes a deletion, an insertion, an inversion, or a replacement of genetic content.

29. The method of claim 27, wherein the genome is a human genome.

30. The method of claim 27, wherein the enzyme is capable of inducing a site-specific DNA recombination in a genome of a host cell by recombining two target sequences naturally occurring in the genome.

31. The method of claim 27, wherein the target sequences are not identical or wherein the target sequences are identical.

32. The method of claim 27, wherein the designer DNA-recombining enzyme is capable of causing a deletion, an insertion, an inversion, or a replacement of a nucleotide sequence in the genome of the host cell.

33. The method of claim 27, wherein each target sequence comprises two half sites with an intervening spacer sequence.

34. The method of claim 27, wherein the half sites are between 10-20 nucleotides in length.

35. The method of claim 27, wherein the half sites are 13 nucleotides in length.

36. The method of claim 27, wherein the half sites differ from each other by not more than 7 nucleotides.

37. The method of claim 27, wherein the half sites differ from each other by not more than 5 nucleotides.

38. The method of claim 27, wherein the half sites differ from each other by not more than 4 nucleotides.

39. The method of claim 27, wherein the half sites differ in at least two nucleotides, at least three nucleotides, at least four nucleotides, or at least eight nucleotides.

40. The method of claim 33, wherein the spacer sequence is between 5 and 12 nucleotides in length.

41. The method of claim 33, wherein the spacer sequence is 8 nucleotides in length.

42. The method of claim 33, wherein the spacer sequence comprises a non-palindromic sequence having at least 2 positions of asymmetry.

43. The method of claim 33, wherein the spacer sequence comprises a non-palindromic sequence having at least 4 positions of asymmetry.

44. The method of claim 33, wherein the spacer sequence comprises a non-palindromic sequence having at least 6 positions of asymmetry.

45. The method of claim 33, wherein each target sequence comprises an identical spacer sequence.

46. The method of claim 33, wherein each target sequence comprises a non-identical spacer sequence.

Description

[0230] The invention is illustrated by the following figures and non-limiting examples:

[0231] FIG. 1 shows a general scheme of a preferred Designer DNA Recombining enzyme induced Gene Replacement (DRiGR) according to the invention. Different monomers R1-4 of DNA recombining enzymes and their respective DNA binding sequences loxR1-loxR4 (half sites of target sites) are shown. The site of interest in the genome (frowny face) is replaced by a desired sequence (smiley).

[0232] FIG. 2 shows a general scheme of a Designer DNA Recombining enzyme induced Gene Deletion (DRiGD). Again, different monomers R1-4 of DNA recombining enzymes binding at their respective DNA binding sequences loxR1-loxR4 are shown in the upper scheme. Here the target sites bear each identical 8 bp spacers. The site of interest in the genome (frowny face) is here cut out.

[0233] FIG. 3 illustrates that if the donor vector is in excess to the genomic DNA, Gene Replacement induced by the DNA recombining enzyme is driven towards the replacement of the genomic DNA with the donor vector DNA. Preferably, the donor vector is present in multiple copies in the nucleus of the target cell, such as in AAV vectors, which results in efficient repair of cellular mutations.

[0234] FIG. 4 demonstrates a general method to obtain a DNA Recombining enzyme for DRiGR by applying a novel directed evolution strategy (Duo-SLIDE DRiGR) that delivers a pair of monomers of DNA recombining enzymes that work in conjunction to recombine two different target sites (white triangle lox 1 and black triangle lox 2) present in the genome starting with a source vector with two libraries of DNA recombining enzymes. Expression of single recombinases in this system is also feasible (s. FIG. 15). R1, R2 and R3 stand for three different restriction enzymes or the respective restriction sites. P1 and P2 stand for two different PCR primers or the respective primer binding sites.

[0235] When these monomers of DNA recombining enzymes are expressed in conjunction with a DNA template also carrying the two target sites (here the pDonor plasmid), recombination at both target sites leads to the exchange of the DNA fragments. Because the incoming DNA fragment removes a restriction site from the pDuoSLiDE vector, effective recombinases that promoted the replacement can be amplified by PCR after digestion of the DNA with the restriction enzyme. Cycling through this system in combination with random mutagenesis and DNA shuffling uncovers the most efficient recombinases promoting the reaction. By using an origin of replication (R6K-ori) on pDonor that is inactive in the host, the process of integration through one target site (e.g. lox1) and excision through the other target site (e.g. lox2) can be uncoupled. Furthermore, integration can be identified through antibiotic selection on kanamycin containing plates.

[0236] FIG. 5 demonstrates a similar method to obtain a DNA Recombining enzyme for DRiGR by applying a novel directed evolution strategy (QuSLiDE DRiGR) that delivers two pairs of monomers of DNA recombining enzymes that work in conjunction to recombine two different target sites. Here each monomer is directed to one half site of one target site (white triangle lox 1 and black triangle lox 2) present in the genome starting with a source vector with four libraries of DNA recombining enzymes. (R1, R2 and R3 as well as P1 and P2 have the meaning as in FIG. 4.)

[0237] If the DNA on the donor vector contains the wild-type allele and the genome contains a disease causing mutation, the gene defect is repaired in cells (s. FIG. 1).

[0238] FIG. 6 demonstrates a general method to obtain a DNA Recombining enzyme for DRiGD by applying a novel directed evolution strategy (Duo-SLIDE DRiGD) that delivers a pair of monomers of DNA recombining enzymes that work in conjunction to recombine two different target sites (white triangle lox 1 and black triangle lox 2) present in the genome starting with a source vector with two libraries of DNA recombining enzymes. (R1, R2 and R3 as well as P1 and P2 have the meaning as in FIG. 4.)

[0239] When these monomers of DNA recombining enzymes are expressed, recombination between two copies of the pDuoSLiDE vector removes a restriction site from the pDuoSLiDE vector, effective recombinases that promoted the deletion can be amplified by PCR after digestion of the DNA with the restriction enzyme. Cycling through this system in combination with random mutagenesis and DNA shuffling uncovers the most efficient recombinases promoting the reaction.

[0240] FIG. 7 demonstrates a similar method to obtain a DNA Recombining enzyme for DRiGD by applying a novel directed evolution strategy (QuSLiDE DRiGD)) that delivers two pairs of monomers of DNA recombining enzymes that work in conjunction to recombine two different target sites. Here each monomer is directed to one half site of one target site (white triangle lox 1 and black triangle lox 2) present in the genome starting with a source vector with four libraries of DNA recombining enzymes. (R1, R2 and R3 as well as P1 and P2 have the meaning as in FIG. 4.)

[0241] This set of monomers allows deleting a site of interest in a Genome (s. FIG. 2).

[0242] FIG. 8 shows a method know in the state of the art to obtain a tailored recombinase that works one target sites by substrate-linked protein evolution (SLiPE) (Buchholz F and Stewart AF, 2001). Here the vector only contains one library and two identical target sites (dashed triangles).

[0243] FIGS. 9A-9C show the application of the invention to replace a mutation in the exon 8 of the F9 gene by Duo-SLIDE DRiGR. FIG. 9A shows the schematic presentation of exon 8 and flanking sequence of the F9 gene. The two selected target sequences (loxF9a—SEQ ID No. 3 and loxF9b—SEQ ID No. 4) are indicated by triangles (black and white) and the nucleotide sequence and the chromosomal positions are shown. FIG. 9B illustrates the method to obtain the two recombinase monomers F9-1 (SEQ ID No. 1, 3, 5 and 28) and F9-2 (SEQ ID No. 2, 4, 6 and 29) and the replacement reaction. Confirmation that the recombination reactions have happened as predicted was confirmed by DNA sequencing. FIG. 9C shows an agarose gel demonstrating the restriction pattern for pDuoF9 (source) (lane 1) and pDuoF9 (product) (lane 2) after digestion with NdeI and SacI. M=molecular marker.

[0244] FIG. 10 shows an exemplary evolution process of adapting the libraries of DNA recombining enzymes to a given target site—here applied to the loxF9a and loxF9b target sites.

[0245] Here DNA recombining enzymes that are active on the loxF9a site (F9A) were obtained by designing two symmetric intermediate sites AL and AR and performing 20 or 17 cycles of SLiPE (s. FIG. 8). After combining and shuffling the libraries of DNA recombining enzymes obtained 108 cycles of SLiPE were performed on the asymmetric loxF9a site (F9A). Agarose gels of indicated generation cycles are shown, with the line with two triangles indicating the non-recombined band and the line with one triangle showing the size of the recombined band. The grey triangle below the gel pictures indicates the reduced amount of L-arabinose added to the growth medium. To obtain DNA recombining enzymes that are active on the loxF9b site (F9B) a first set of two symmetric intermediate sites BLS and BRS were designed and 9 cycles of SLiPE (s. FIG. 8) were performed. After shuffling the libraries of DNA recombining enzymes again SLiPE was performed on a second set of symmetric intermediate sites BL (30 cycles) and BR (22 cycles) was performed. Finally, after combing and shuffling the libraries of DNA recombining enzymes obtained 17 cycles of SLIPE were performed on the asymmetric loxF9b site (F9B). Finally (not shown) the two libraries of DNA recombining enzymes obtained on the loxF9a site and the loxF9b site were cloned into pDuoF9 (source)—s. FIGS. 9A-9C.

[0246] FIG. 11 illustrates the use the F9 designer DNA recombining enzymes in a therapeutic setting, both F9-1 and F9-2 coding sequences are cloned into a delivery vector (such as an adeno-associated viral vector) together with the donor sequence that contains loxF9a and loxF9b flanking the wild-type, or Padua mutation (R338L) of exon 8 of the F9 gene. Delivery of such a vector into target cells in multiple copies replaces the inactivating mutation in the genome.

[0247] FIG. 12 shows the result of a screen of the human PCSK9 Gene on Chromosome 1 for 8 bp repeats. The exons of the PCSK9 Gene are indicated by bold boxes. The 8 bp repeats are separated by at least 150 bp and no more than 2 kb apart and lie in potential target sites (indicated by vertical lines) with a length of 34 bp, that were screened not occurring elsewhere in the genome. The sequences of interests between the potential target sites are indicated by horizontal lines.

[0248] FIG. 13 shows the result of a similar screen of the genome of the human papilloma virus 16 (HPV16) for 8 bp repeats that are separated by at least 150 bp and no more than 2 kb apart and lie in potential target sites of 34 bp, that were screened not to occur in the human genome. The sequences encoding for the viral proteins are indicated in bold boxes with arrows (greater-than signs).

[0249] FIG. 14 demonstrates a different general method to obtain a DNA Recombining enzyme for DRiGR used in example 3 step 3 by applying a novel directed evolution strategy (SLIDE DRiGR 2.0) that delivers a library of DNA recombining enzymes that recombines two different target sites (white triangle lox 1 and black triangle lox 2) present in the genome starting. R1, R2, R3, R4 and R5 stand for five different restriction enzymes or the respective restriction sites. P3 and P4 stand for two different PCR primers or the respective primer binding sites. When this library of DNA recombining enzymes is expressed in the presence of a DNA template also carrying the two target sites (here the pDonor-ex8 plasmid), recombination at both target sites leads to the exchange of the DNA fragments. Because the incoming DNA fragment removes restriction sites from the pF9 vector, effective recombinases that promoted the replacement can be amplified by PCR after digestion of the DNA with the restriction enzymes R1-3 and RecBCD. Cycling through this system in combination with random mutagenesis and DNA shuffling uncovers the most efficient recombinases promoting the DRIGR reaction.

[0250] FIG. 15 highlights the steps of the novel directed evolution strategy (SLIDE DRiGR) used in example 3 step 3. In a sub step 1, pF9 (source) and pDonor are transformed into bacteria. Only if the expression of the recombinases is induced, growth of colonies resistant to the antibiotic can be observed. In sub step 2, the plasmid DNA of the obtained colonies is digested using restriction enzymes R1-3. This will enrich for clones which have undergone successful DRIGR and result in a pure preparation of pF9 (product). The purity of the plasmid preparation is verified with a restriction digest using the enzymes XhoI and XmaI, which shows the unique restriction pattern of pDuoF9 (product). Additionally, DNA sequencing reveals the correct sequence of pF9 (product).

[0251] FIG. 16 demonstrates a different general method to obtain a DNA Recombining enzyme for DRiGR by applying a novel directed evolution strategy (Duo-SLIDE DRiGR 2.0) that delivers one or a pair of monomers of DNA recombining enzymes that work in conjunction to recombine two different target sites (white triangle lox 1 and black triangle lox 2) present in the genome starting with a source vector with one or two libraries of DNA recombining enzymes. R1, R2, R3, R4 and R5 stand for five different restriction enzymes or the respective restriction sites. P3 and P4 stand for two different PCR primers or the respective primer binding sites. When these monomers of DNA recombining enzymes are expressed in conjunction with a DNA template also carrying the two target sites (here the high-copy number plasmid pDonor-ex8), recombination at both target sites leads to the exchange of the DNA fragments. Because the incoming DNA fragment removes a restriction site from the pDuoSLiDE vector, effective recombinases that promoted the replacement can be amplified by PCR after digestion of the DNA with the restriction enzyme. Importantly, no antibiotic selection marker is used in this assay. Cycling through this system in combination with random mutagenesis and DNA shuffling uncovers the most efficient recombinases promoting the reaction.

[0252] FIG. 17 shows alignment between Cre (state of the art), a recombinase obtained after step 1 of example 3 (R #1), a recombinase obtained after step 2 of example (3R #7-B5), and a recombinase obtained after step 3 of example 3 (F9-3). The alignment shows how additional mutations have arisen during later stages of evolution.

[0253] FIG. 18 shows a general scheme of the method of the invention, exemplified by the DRiGD method and steps of the invention.

[0254] FIG. 19 and FIG. 20 show a more detailed scheme of the method of the invention, exemplified by the DRiGD method and applied to the human genome.

[0255] FIGS. 21A-21C illustrate the DRIGD method of the invention of applied to a target site on the human chromosome 7—further described in example 5. FIG. 21A shows the target sites: Hex 1 and Hex 2 are the target sites used for excision—HexL, HexR, HexR1 and HexR2 are intermediate target sites. FIG. 21B shows the evolution of specific recombinases for the Hex1 and Hex2 target sites. FIG. 21C shows recombination efficiency of two clones (Clone #7 and Clone #30) obtained.

[0256] FIG. 22 shows the precise excision of a sequence from a plasmid substrate using the recombinases obtained in example 5.

EXAMPLE 1

[0257] To demonstrate the utility of DRiGD, the inventors have screened the human genome for all 8 bp repeats, separated by at least 150 bp: The potential target sites are indicated sequences sequences 13 bp left and 13 bp right of the 8 bp repeat not occurring elsewhere in the human genome in a window to delete a maximum of 2 kb from the genome. FIG. 12 shows the results for all 309 possible target sites for DNA recombining enzymes in the human gene PCSK9. FIG. 13 shows the results for all 176 possible target sites for DNA recombining enzymes in the human papilloma virus 16 (HPV16).

EXAMPLE 2

[0258] To demonstrate the feasibility of the DRiGR invention, the inventors have generated recombinases that when applied in concert can replace exon eight of the human factor 9 (F9) gene. This exon frequently carries mutations in patients suffering from hemophilia B. Our results demonstrate that this exon can be efficiently replaced for a different sequence providing an appropriate donor vector and expression of two recombinases working in concert (FIG. 6, FIGS. 9A-9C and FIG. 11).

[0259] Materials and methods as described in WO 2008/083931 A1, WO 2011/147590 A1, WO2016034553 A1 and in the publication Buchholz F and Stewart AF, 2001 are used, if not specified otherwise.

[0260] The target sequences (loxF9a—SEQ ID No. 7 and loxF9b—SEQ ID No. 8) were selected by comparing sequences in the human genome with half sites of previously identified target sites (LoxP, LoxH, LoxM7, LoxM5, loxLTR, LoxBTR) of DNA recombining enzymes:

TABLE-US-00008 Name Sequence SEQ ID No. loxF9a CTCATTACATTTA ACCAAAAT TATCACAATATAA  7 loxF9b CCATCTTTTGTTA GATTTGAA TATATACATTCTA  8 loxP ATAACTTCGTATA ATGTATGC TATACGAAGTTAT  9 loxH ATATATACGTATA TAGACATA TATACGTATATAT 10 loxM7 ATAACTCTATATA ATGTATGC TATATAGAGTTAT 11 LoxM5 ATAACTTCGTGCA ATGTATGC TGCACGAAGTTAT 12 loxLTR ACAACATCCTATT ACACCCTA TATGCCAACATGG 13 loxBTR AACCCACTGCTTA AGCCTCAA TAAAGCTTGCCTT 14

[0261] In the table the first (left) and second (right) half sites in the target sites are underlined.

[0262] In a first step the two recombinase libraries were evolved separately using the SLIDE approach as described before (Buchholz and Stewart, 2001 and WO 2002044409 A2—s. FIG. 8). In short, each library was evolved towards recombining intermediate target sites (AL, AR, BLS, BLRS, BL, BR) and finally to recombining one of the two target sites, loxF9a or loxF9b. New mutations were introduced through error-prone PCR and DNA-shuffling and selective pressure was further regulated via the expression level of the recombinases. Active recombinases were selected by their ability to recombine the respective target site to delete the sequence flanked by the target sites from the vector (s. FIG. 8).

[0263] The following intermediate sequences were used:

TABLE-US-00009 Name Sequence SEQ ID No. loxF9-AL CTCATTACATTTA ACCAAAAT TAAATGTAATGAG 15 loxF9-AR TTATATTGTGATA ACCAAAAT TATCACAATATAA 16 loxF9-BLS CCAACTTTTGATA GATTTGAA TATCAAAAGTTGG 17 loxF9-BRS TAGACTTTATATA GATTTGAA TATATAAAGTCTA 18 loxF9-BL CCATCTTTTGTTA GATTTGAA TAACAAAAGATGG 19 LOXF9-BR TAGAATGTATATA GATTTGAA TATATACATTCTA 20

[0264] Recombinase coding sequences from these two libraries were then cloned in the vector backbone pDuoF9 (source) (SEQ ID No. 21, s. FIGS. 9A-9C). In this vector the two recombinase libraries are expressed from a shared inducible promoter in an operon-like structure, which transcriptionally links both enzymes together. The size of this library exceeded 100.000 clones.

[0265] To carry out DRiGR, 50 μl of electrocompetent E. coli XL1-Blue cells were transformed with 1 ng pDuoF9 (source) library and grown overnight in 200 ml LB medium containing 15 μg/ml chloramphenicol. On the next day, 1 ml of the overnight culture was used to inoculate 100 ml of fresh medium and grown for 2 h. Then the culture was split into 2×50 ml and L-arabinose was added to a final concentration of 50 μg/ml to induce recombinase expression in one of the cultures. After 2.5 h of incubation the cells were put on ice and prepared for electroporation as described before (Sambrook and Russell, 2001). The electrocompetent cells were resuspended in 200 μl water, and 50 μl of the cell suspension was immediately used for transformation of 0.4 ng pDonor at 1700 V. The bacteria were then allowed to recover for 2 h in SOC medium before the entire suspension was plated on LB agar plates containing 15 μg/ml kanamycin.

[0266] No colonies had grown overnight on the agar plates where uninduced cells had been plated, validating that recombinase expression was required for the integration of the pDonor vector (SEQ ID No. 22). All colonies growing from the induced samples were pooled and cultured in 220 ml LB medium containing 15 μg/ml kanamycin and 50 μg/ml L-arabinose to allow possible recombination through the second pair of target sites.

[0267] On the next day the plasmid DNA was isolated from the culture. To enrich for clones that had undergone successful DRiGR, the DNA was digested with enzymes cleaving any non-pDuoF9 (product)—negative selection—and re-transformed. To this end, 500 ng of plasmid preparation were digested with 1 μl of each NdeI, AvrII and PspXI. After 3 h incubation at 37° C. DNA was precipitated and resuspended in 50 μl water. 1 μl of this was used for transformation of 50 μl electrocompetent E. coli XL1-Blue cells. The transformed cells were grown in 220 ml LB medium containing 15 μg/ml kanamycin—positive selection-.

[0268] On the next day the plasmid DNA was isolated from the culture. The recombinase libraries were amplified by PCR using primers P1 and P2 (SEQ ID No.23 and 24) and the PCR products were purified on a column and subjected to a digest with the restriction enzymes SacI and SbfI. The isolated recombinase libraries were then ligated back into pDuoF9 (source) to start another cycle of DRiGR. Three rounds of Duo-SLIDE DRIGR (s. FIG. 10) were performed to enrich for recombinases that can efficiently carry out DRiGR on the target sites loxF9a and LoxF9b.

[0269] Primer Used:

TABLE-US-00010 Name Sequence SEQ ID No. P1 CTCTACTGTTTCTCCATAC 23 P2 AGGGAATAAGGGCGACA 24

[0270] The sequences of the plasmid pDuoF9 (product) is given in SEQ ID No.25

[0271] By this method the sequences of the following pairs of recombinase monomers were obtained: [0272] Rec F9-1a: SEQ ID No. 3 and Rec F9-2a: SEQ ID No. 4 [0273] Rec F9-1b: SEQ ID No. 5 and Rec F9-2b: SEQ ID No. 6

EXAMPLE 3

[0274] Step 1:

[0275] In a first step the two recombinase libraries were evolved separately using the SLIDE approach as described before (Buchholz and Stewart, 2001 and WO 2002044409 A2—s. FIG. 8). In short, each library was evolved as described in example 2 towards recombining intermediate target sites (loxF9-AL, AR, BLS, BLRS, BL, BR—s. example 2) and finally to recombining one of the two target sites, loxF9a or loxF9b. New mutations were introduced through error-prone PCR and DNA-shuffling and selective pressure was further regulated via the expression level of the recombinases. Active recombinases were selected by their ability to recombine the respective target site to delete the sequence flanked by the target sites from the vector (s. FIG. 8).

[0276] Step 2:

[0277] In a second step a single recombinase library was evolved to carry out DRiGR following a selection scheme with the replication deficient pDonor (s. FIGS. 9A-9C). This involved cloning the recombinase coding sequences from the libraries generated in step 1 in the vector backbone pF9 (source) (=pDuoF9 SEQ ID No. 21). In this vector one to four recombinase libraries can be expressed from a shared inducible promoter in an operon-like structure, which transcriptionally links both enzymes together. The size of this library exceeded 100.000 clones.

[0278] To carry out DRiGR, 50 μl of electrocompetent E. coli XL1-Blue cells were transformed with 1 ng pDuoF9 (source) library and grown overnight in 200 ml LB medium containing 15 μg/ml chloramphenicol. On the next day, 1 ml of the overnight culture was used to inoculate 100 ml of fresh medium and grown for 2 h. Then the culture was split into 2×50 ml and L-arabinose was added to a final concentration between 1-200 μg/ml to induce recombinase expression in one of the cultures. After 2.5 h of incubation the cells were put on ice and prepared for electroporation as described before (Sambrook and Russell, 2001). The electrocompetent cells were resuspended in 200 μl water, and 200 μl of the cell suspension was immediately used for transformation of 80 ng pDonor at 1700 V. The bacteria were then allowed to recover for 2 h in SOC medium before the entire suspension was used to incoculate 200 ml LB medium containing 15 μg/ml chloramphenicol, 5 μg/ml kanamycin and 1-200 μg/ml L-arabinose. A small part of the suspension was plated on agar plates to estimate the library size.

[0279] On the next day it was verified that no colonies had grown on the agar plates where uninduced cells had been plated, validating that recombinase expression was required for the integration of the pDonor vector (SEQ ID No. 22). The plasmid DNA was isolated from the liquid culture. To enrich for clones that had undergone successful DRiGR, the DNA was digested with enzymes cleaving any non-pDuoF9 (product)—negative selection—and re-transformed. To this end, 500 ng of plasmid preparation were digested with 1 μl of each NdeI, AvrII, PspXI and FspI. After 3 h incubation at 37 t the DNA was cleaned-up via microdialysis on a membrane filter. 3 μl of this digest was used for transformation of 50 μl electrocompetent E. coli XL1-Blue cells. The transformed cells were grown in 100 ml LB medium containing 15 μg/ml chloramphenicol and 15 μg/ml kanamycin—positive selection-.

[0280] On the next day the plasmid DNA was isolated from the culture, and 500 ng of DNA was digested with 1 μl of each NdeI and AvrII. The recombinase libraries were amplified by error-prone PCR using primers P1 # and P2 #(SEQ ID No. 30 and 31) and the PCR products were purified on a column and subjected to a digest with the restriction enzymes SacI and SbfI. The isolated recombinase libraries were then ligated back into pDuoF9 (source) to start another cycle of DRiGR. After cycles 3 and 8 DNA shuffling was carried out as described before (Buchholz and Stewart, 2001). Eleven rounds of Duo-SLIDE DRIGR were performed to enrich for recombinases that can efficiently carry out DRiGR on the target sites loxF9a and loxF9b.

[0281] Step 3:

[0282] In a third step a single recombinase library was evolved to carry out DRiGR 2.0 following a different selection scheme (s. FIG. 14) with the high-copy number plasmid pDonor-ex8 (s. FIG. 14, SEQ ID 37) without antibiotic selection and without separating integration from resolution of the donor vector. This involved cloning the recombinase coding sequences from the library generated in step 2 in the vector backbone pF9 (source) (=pDuoF9 (source) SEQ ID No. 21). The library was then transformed into electrocompetent E. coli XL1-Blue cells which were already carrying pDonor-ex8. The transformed cells were grown in 100 ml LB medium containing 15 μg/ml chloramphenicol, 5 μg/ml kanamycin and 1-200 μg/ml L-arabinose. The size of this library exceeded 100.000 clones.

[0283] On the next day the plasmid DNA was isolated from the culture, and 1000 ng of DNA was digested with 1 μl of each NdeI, AwlI, PspXI (restriction enzymes R1, R2 and R3—s. FIGS. 14 and 15) and the exonuclease RecBCD. The recombinase libraries were amplified by error-prone PCR using primers P3 and P4 (SEQ ID No. 32 and 33) and the PCR products were purified on a column and subjected to a digest with the restriction enzymes SacI and SbfI. The isolated recombinase libraries were then ligated back into pF9 (source) and transformed into electrocompetent E. coli XL1-Blue cells carrying pDonor-ex8. The transformed cells were grown in 100 ml LB medium containing 15 μg/ml chloramphenicol, 5 μg/ml kanamycin and 1-200 μg/ml L-arabinose and the next cycle thus began. Every three cycles DNA shuffling was carried out as described before (Buchholz and Stewart, 2001). Eleven rounds of SLIDE DRIGR (s. FIG. 14) were performed to enrich for recombinases that can efficiently carry out DRIGR on the target sites loxF9a and LoxF9b without antibiotic selection and without separating integration from resolution of the donor vector.

[0284] Primers Used:

TABLE-US-00011 Name Sequence SEQ ID No. P1# CGGCGTCACACTTTGCTATG 30 P2# CCCTTAAACGCCTGGTGCTA 31 P3 AAGATTAGCGGATCCTACCT 32 P4 GTGATTAGTTAGTGAGAGGC 33

[0285] By this method the sequences of the following recombinase monomers were obtained: R#1 (SEQ ID No. 34) after step 1: R#7-B5 (SEQ ID No. 35) after step 2 and Rec F9-3 (SEQ ID No. 36) after the final step 3.

[0286] The recombinase efficiency (DRGR efficiency) of the recombinases obtained after step 1, 2 and 3 are compared in the following table:

TABLE-US-00012 Example Selection Recombinase DRIGR Step scheme obtained SEQ ID No. efficiency 1 SLIDE R#1 34  0.5% (s. FIG. 8) 2 Duo-SLIDE R#7-B5: 35 12.3% DRIGR (s. FIG. 9A-9C) 3 SLIDE-DIGR F9-3 36 .sup. 90% 2.0 (s. FIG. 14)

EXAMPLE 4

[0287] To utilize the F9 recombinases in a therapeutic setting, both F9-1 and F9-2 coding sequences are cloned into a delivery vector (such as an adeno-associated viral vector) together with the donor sequence that contains loxF9a and loxF9a flanking the wild-type, or Padua mutation (R338L) of exon 8 of the F9 gene. Delivery of such a vector into target cells in multiple copies replaces the inactivating mutation in the genome (s. FIG. 11).

EXAMPLE 5

[0288] To demonstrate the utility of DRiGD recombinases for the excision of a specific DNA sequence on chromosome 7 of the human genome were evolved.

[0289] First, two different recombinases libraries were evolved using the described protocol of substrate linked directed evolution (Buchholz and Stewart, 2001 and WO 2002044409 A2—s. FIG. 8). The libraries were evolved against the left or the right half-site (HexL or HexR) of the final target sites Hex1 and Hex2 (s. FIG. 21A). To evolve the recombinase libraries for HexR, two intermediated target sites (HexR1 and HexR2) were used (s. FIG. 21B). Throughout the evolution process new mutations were introduced with error-prone PCR (using Primers P1 # and P2 #) and DNA-shuffling and selective pressure was further regulated via the expression level of the recombinases. Active recombinases were selected by their ability to recombine the respective target site to delete the sequence flanked by the target sites from the vector (s. FIG. 8—Standard SLiPE and FIG. 21B).

[0290] Next, both libraries (obtained on HexR and HexL) were combined in order to recombine the final target sites Hex1 and Hex2. Therefore, the libraries (library size was bigger than 100.000 recombinases) were cloned into the same expression plasmid and were co-expressed from a shared inducible promoter in an operon-like structure, which transcriptionally links both enzymes together (s. FIG. 6 pDUO-SLiDE DRiGD). The plasmids carrying both libraries were then transformed into XL-1 blue E. coli cells. After recovering the cells in 1 ml of SOC medium for one hour, 10 μl was plated on a LB agar plate with 15 μg/ml chloramphenicol. The next day 32 clones were picked and cultured in 500 μl of LB with 25 μg/ml chloramphenicol for 8 h. Next, 250 μl pre-culture was used to inoculate 2×5 ml of LB with 25 μg/ml chloramphenicol. In one of the 5 ml cultures 10 μg/ml L-arabinose was added to induce the expression of the recombinase dimer. The next day plasmid DNA was isolated and digested with SacI (R2) and SbfI (R3) to estimate the recombination efficiency in the induced and uninduced sample.

[0291] With this method two recombinase dimers (Clone #7 and Clone #30) were obtained that can carry out the excision reaction at high efficiency. The precise excision was confirmed by sequencing the recombined pDUO-SLiDE DRiGD—s. FIG. 22.

[0292] The target sites and the intermediate target sites and primers used are listed in the following table:

TABLE-US-00013 Hex target sites SEQ-ID Hex1 TACACAGTGTATATTGATTTTTATCAAATGCCTT 40 Hex2 TACACAATGTATATTGATTTTTATCAAATGCCTT 41 HexL TACACAGTGTATATTGATTTTTATACATTGTGTA 42 HexR AAGGCATTTGATATTGATTTTTATCAAATGCCTT 43 HexR1 AAGACATTTTATATTGATTTTTATAAAATGTCTT 44 HexR2 AACGCATTGGATATTGATTTTTATCCAATGCGTT 45 P1# CGGCGTCACACTTTGCTATG 30 P2# AAGGGAATAAGGGCGACACG 31

[0293] By this method the sequences of the following pairs of recombinase monomers were obtained: [0294] Hex-R-#7: SEQ ID No. 46 and Hex-L-#7: SEQ ID No. 47 [0295] Hex-R-#30: SEQ ID No. 48 and Hex-L-#30: SEQ ID No. 49

CITED NON-PATENT LITERATURE

[0296] Buchholz, F., & Stewart, A. F. (2001). Alteration of Cre recombinase site specificity by substrate-linked protein evolution. Nature Biotechnology, 19(11), 1047-1052. [0297] Karpinski J, Hauber I, Chemnitz J, et al. (2016). Directed evolution of a recombinase that excises the provirus of most HIV-1 primary isolates with high specificity. Nat Biotechnol 34(4):401-9. [0298] Sambrook, J., & Russell, D. W. (2001). Molecular Cloning: A Laboratory Manual, (3rd edition). [0299] Surendranath, V. et al. (2010). SeLOX—a locus of recombination site search tool for the detection and directed evolution of site-specific recombination systems. Nucleic Acids Res. 2010, 38(W293-W298) [0300] Wang M et al. (2016). Efficient delivery of genome-editing proteins using bioreducible lipid nanoparticles. PNAS 113:2868-2873. [0301] Zhang, C et al. (2015) Redesign of the monomer—monomer interface of Cre recombinase yields an obligate heterotetrameric complex. Nucleic Acids Res. 2015, 43 (18): 9076-9085.