METHOD FOR THE GENERATION OF DUMBBELL-SHAPED DNA VECTORS

Abstract

Dumbbell-shaped DNA minimal vectors represent genetic vectors solely composed of the gene expression cassette of interest and terminal closing loop structures. Dumbbell vectors for small hairpin RNA or microRNA expression are extremely small-sized which is advantageous with regard to cellular delivery and nuclear diffusion. Conventional strategies for the generation of small RNA-expressing dumbbell vectors require cloning of a respective plasmid vector which is subsequently used for dumbbell protection. Here, we present a novel cloning-free method for the generation of small RNA-expressing dumbbell vectors which also does not require any restriction endonucleases. The method comprises the PCR amplification of a universal DNA template using primers containing the sense or antisense strand of the sequence of interest, the denaturing and refolding of the amplified product to form stem-loop structures, and the structures are covalently closed using DNA ligases to obtain dumbbell structures.

Claims

1. A cloning-free and endonuclease-free method to generate a dumbbell-shaped vector that includes a hairpin-structured expression cassette comprising: i) providing a preparation comprising a single stranded nucleic acid template comprising a target nucleic acid molecule comprising two complementary sequence segments wherein each segment comprises a transcription promoter sequence and further comprises a transcriptional terminator and wherein said two complementary sequence segments are separated by a third sequence segment; ii) providing a preparation wherein said single stranded nucleic acid template additionally comprises a stem; iii) contacting said single stranded nucleic acid template with a first oligonucleotide primer comprising a 3′-hydroxyl group that is complementary to at least part of the 3′ terminal nucleotide sequence of said single stranded nucleic acid template and further comprising a 5′ nucleotide sequence not complementary to the target nucleic acid molecule wherein said oligonucleotide primer comprises the sequence coding for the 3′ arm of a hairpin-structured RNA and further contacting said single stranded nucleic acid template with a first blocking oligonucleotide that is complementary to at least part of the 5′ terminal nucleotide sequence of said single stranded nucleic acid template; iv) providing polymerase chain reaction components to primer extend the 3′ annealed first oligonucleotide primer; v) contacting said extended oligonucleotide primer with a second oligonucleotide primer comprising a 3′-hydroxyl group that is complementary to at least part of the 3′ terminal nucleotide sequence of said extended oligonucleotide primer and further comprising a 5′ nucleotide sequence not complementary to the target nucleic acid molecule wherein said oligonucleotide primer comprises the sequence coding for the 3′ arm of a hairpin-structured RNA and further contacting said single stranded nucleic acid template with a second blocking oligonucleotide that is complementary to at least part of the 5′ terminal nucleotide sequence of said extended oligonucleotide primer; vi) polymerase chain amplify the template to synthesize a pool of template DNA and annealing said templates to create double stranded nucleic acid comprising a sequence coding for the 3′ arm of a hairpin-structured RNA at the 5′ nucleotide sequences of the plus strand and the minus strand and comprising a sequence coding for the 5′ arm of a hairpin-structured RNA at the 3′ nucleotide sequences of the plus strand and the minus strand; vii) heat denature said double stranded nucleic acid and then cool down to allow intra-molecular refolding of the resulting plus strand and minus strand DNA to create preformed oligomeric stem-loop structures comprised of plus or minus strand DNA; viii) contacting said preformed oligomeric loops structures of plus strand and minus strand DNA with a single strand-specific DNA ligase to link the terminal 5′-phosphorylated 5′ overhang to the terminal 3′-OH group of the 3′ overhang of the same DNA strand in an intramolecular ligation to create covalently closed plus strand-derived and minus strand-derived dumbbell-shaped vector DNA comprising hairpin-structured template DNA for the transcription of hairpin-structured RNA; ix) contacting said dumbbell-shaped vector DNA with a DNA exonuclease to remove all primers and non-covalently closed DNA that harbours 5′ and 3′ ends.

2. A cloning-free and restriction endonuclease-free method to generate a dumbbell-shaped vector that includes an expression cassette comprising: i) providing a preparation comprising a single stranded nucleic acid template comprising a target nucleic acid molecule comprising two complementary sequence segments wherein each segment comprises a transcription promoter sequence and further comprises a transcriptional terminator and wherein said two complementary sequence segments are separated by a third sequence segment; ii) providing a preparation wherein said single stranded nucleic acid template additionally comprises a stem; iii) contacting said single stranded nucleic acid template with a first oligonucleotide primer comprising a 3′-hydroxyl group that is complementary to at least part of the 3′ terminal nucleotide sequence of said single stranded nucleic acid template and further comprising a 5′ nucleotide sequence not complementary to the target nucleic acid molecule and further contacting said single-stranded nucleic acid with a first set of blocking oligonucleotides comprising one, two, three or more oligonucleotides that are complementary to at least part of the 5′ terminal nucleotide sequence of said single stranded nucleic acid template; iv) providing polymerase chain reaction components to primer extend the 3′ annealed first oligonucleotide primer; v) contacting said extended oligonucleotide primer with a second oligonucleotide primer comprising a 3′-hydroxyl group that is complementary to at least part of the 3′ terminal nucleotide sequence of said extended oligonucleotide primer and further comprising a 5′ nucleotide sequence not complementary to the target nucleic acid molecule and further contacting said single stranded nucleic acid template with a second set of blocking oligonucleotides comprising one, two, three or more oligonucleotides that are complementary to at least part of the 5′ terminal nucleotide sequence of said extended oligonucleotide primer; vi) polymerase chain amplify the template to synthesize a pool of extended oligonucleotide primers (and annealing said templates to create double stranded nucleic acid comprising a sequence coding for the 3′ arm of a hairpin-structured RNA at the 5′ nucleotide sequences of the plus strand and the minus strand and comprising a sequence coding for the 5′ arm of a hairpin-structured RNA at the 3′ nucleotide sequences of the plus strand and the minus strand;) vii) contacting said pool of extended oligonucleotide primers with a third oligonucleotide primer comprising a 3′ hydroxyl group that is complementary to at least part of the 3′ terminal nucleotide sequence of the oligonucleotide primer extended by said first oligonucleotide primer and further comprising a 5′ nucleotide sequence not complementary to said extended oligonucleotide primer and further contacting said pool of extended oligonucleotide primers with a fourth oligonucleotide primer comprising a 3′ hydroxyl group that is complementary to at least part of the 3′ terminal nucleotide sequence of the oligonucleotide primer extended by said second oligonucleotide primer and further comprising a 5′ nucleotide sequence not complementary to said extended oligonucleotide primer and further contacting said extended oligonucleotide primers with said first and second set of blocking oligonucleotides; viii) repeating steps iv) and vii) using a fifth and sixth, seventh and eight or more pairs of oligonucleotide primers wherein the last set of primers comprises 3′ hydroxyl groups and 5′ phosphate groups; ix) polymerase chain amplify the template to synthesize a pool of extended oligonucleotide primers and annealing said templates to create double stranded nucleic acid comprising a sequence coding for the 3′ arm of a hairpin-structured RNA at the 5′ nucleotide sequences of the plus strand and the minus strand and comprising a sequence coding for the 5′ arm of a hairpin-structured RNA at the 3′ nucleotide sequences of the plus strand and the minus strand; x) heat denature said double stranded nucleic acid and then cool down to allow intra-molecular refolding of the resulting plus strand and minus strand DNA to create preformed oligomeric stem-loop structures comprised of plus or minus strand DNA; xi) contacting said preformed oligomeric loops structures of plus strand and minus strand DNA with a single strand-specific DNA ligase to link the terminal 5′-phosphorylated 5′ overhang to the terminal 3′-OH group of the 3′ overhang of the same DNA strand in an intramolecular ligation to create covalently closed plus strand-derived and minus strand-derived dumbbell-shaped vector DNA comprising hairpin-structured template DNA for the transcription of hairpin-structured RNA; xii) contacting said dumbbell-shaped vector DNA with a DNA exonuclease to remove all primers and non-covalently closed DNA that harbours 5′ and 3′ ends.

3. A cloning-free and restriction endonuclease-free method to generate a dumbbell-shaped dual expression vector comprising two expression cassettes wherein said vector comprises: i) providing a preparation comprising a first single stranded nucleic acid template comprising a target nucleic acid molecule comprising the reverse complement of a first transcriptional promoter at the 5′ terminal sequence and the sequence of a second transcriptional terminator at the 3′ terminal sequence; ii) contacting said first single stranded nucleic acid template with an first oligonucleotide primer comprising a 5′-phosphate and a 3′-hydroxyl group that is complementary to at least part of the 3′ terminal nucleotide sequence of said single stranded nucleic acid template and further comprising a 5′ nucleotide sequence not complementary to the target nucleic acid molecule wherein said oligonucleotide primer comprises a first sequence of interest to be transcribed and further comprising the reverse complement sequence of a first transcriptional terminator and further comprising a modified nucleotide sequence that prevents extension of the 5′ nucleotide sequence not complementary to the target nucleic acid molecule; iii) providing polymerase chain reaction components and primer extending the 3′ annealed oligonucleotide primer to form a second template; iv) contacting said second template with a second oligonucleotide primer comprising a 5′-phosphate and a 3′-hydroxyl group that is complementary to at least part of the 3′ terminal nucleotide sequence of said second template and further comprising a 5′ nucleotide sequence not complementary to the second template wherein said oligonucleotide primer comprises a second sequence of interest to be transcribed and further comprising the reverse complement sequence of a second transcriptional terminator and further comprising a modified nucleotide sequence that prevents extension of the 5′ nucleotide sequence not complementary to the second template; v) providing polymerase chain reaction components and primer extending the 3′ annealed oligonucleotide primer to form a double stranded nucleic acid; vi) polymerase chain amplify the double stranded nucleic acid to synthesize a pool of template DNA and annealing said templates to create double stranded nucleic acid comprising a 5′ nucleotide sequence not complementary to the target nucleic acid molecule; and vii) contacting the annealed template nucleic acid with a 5 DNA ligase to link the terminal 5′-phosphate of the non-complementary 5′ nucleotide sequence to the 3′-OH of said amplified template nucleic acid to create a terminal loop structure.

4. The method according to any of the claims 1 to 3 wherein said transcription promoter is derived from an RNA polymerase III promoter.

5. The method according to any of the claim 4 wherein said RNA polymerase III promoter is i) a U6 promoter; or ii) a H1 promoter; or iii) a minimal H1 promoter that includes an inverted polymerase III transcriptional terminator.

6. The method according to any of the claims 1 to 5 wherein said transcription promoter is derived from an RNA polymerase II promoter.

7. The method according to claim 6 wherein said RNA polymerase II promoter is a CMV promoter.

8. The method according to any one of claims 1 to 7 wherein said transcription terminator nucleotide sequence is a RNA polymerase II or RNA polymerase III termination sequence.

9. The method according to claim 8 wherein said RNA polymerase III termination sequence comprises one or more motifs comprising the nucleotide sequence TTTTT.

10. The method according to any of the claims 1 to 9 wherein said stem is derived from i) an artificial sequence; or ii) a microRNA stem; or iii) a microRNA-30 stem.

11. The method according to any of the claims 1 to 10 wherein said third sequence segment is an oligomeric DNA sequence derived from i) an oligomeric DNA sequence; or ii) a tetrameric DNA sequence; or iii) comprises the sequence TTTT.

12. The method according to any of the claims 1 to 11 wherein said vector comprises a nucleic acid sequence to be transcribed.

13. The method according to claim 12 wherein said nucleic acid sequence to be transcribed is a therapeutic nucleic acid molecule.

14. The method according to claim 13 wherein said therapeutic molecule is selected from the group: a siRNA or shRNA, a miRNA or pre-miRNA, an antisense RNA, an antisense miRNA, an aptamer, trans-splicing RNA, a guide RNA, a single-guide RNA, crRNA or tracrRNA, a mRNA or pre-mRNA.

15. The method according to claim 12 wherein said nucleic acid sequence to be transcribed encodes a therapeutic protein or peptide.

16. The method according to claim 13 wherein said therapeutic protein or peptide is selected from the group: Cas9, Cas9n, hSpCas9, hSpCas9n, HSVtk, a cell death trigger protein, cholera toxin, diphtheria toxin, alpha toxin, anthrax 5 toxin, exotoxin, pertussis toxin, shiga toxin, shiga-like toxin Fas, TNF, caspases, initiator caspases, caspase 2, 8, 9, 10, 11, 12, and effector caspases, caspase 3, 6, 7 and purine nucleoside phosphorylase.

17. The method according to any of the claims 1 to 16 wherein said first and said second oligonucleotide primers comprise a 5′ phosphate.

18. The method according to any of the claims 1 to 17 wherein said hairpin-structured RNA sequence is a small hairpin RNA or a microRNA precursor.

19. The method according to any of the claims 1 to 18 wherein said blocking oligonucleotides are complementary to the complete 5′ terminal nucleotide sequence half of said single stranded nucleic acid template or said extended oligonucleotide primer.

20. The method according to claim 19 wherein said blocking oligonucleotides comprise the nucleotide sequences set forth in SEQ ID NO: 9 and SEQ ID NO: 10.

21. The method according to any of the claims 1 to 20 wherein said heat denaturing comprises an incubation at 96° C. for 5 min.

22. The method according to any of the claims 1 to 21 wherein said cooling down comprises a gradual cooling down from 96° C. to room temperature.

23. The method according to any of the claims 1 to 22 wherein said DNA ligase is a single strand-specific or double strand-specific DNA ligase.

24. The method according to any of the claims 1 to 23 wherein said single strand-specific DNA ligase is the CircLigase.

25. The method according to any of the claims 1 to 24 wherein said exonuclease is the T7 DNA polymerase.

26. The method according to any of the claims 1 to 25 wherein said first and second oligonucleotide primers comprise the reverse complement sequence CAT of a translational start codon in their 5′ nucleotide sequence not complementary to the target nucleic acid molecule.

27. The method according to claims 1 to 26 wherein the primers of said last set of primers comprise the reverse complement sequence CTA, TTA or TCA of a translational stop codon TAG, TAA or TGA in their 5′ nucleotide sequence not complementary to said extended oligonucleotide primer.

28. The method according to any of the claims 1 to 27 wherein the primers of said last set of primers comprise the reverse complement sequence of a transcriptional terminator in their 5′ nucleotide sequence not complementary to said extended oligonucleotide primer.

29. The method according to claims 3 to 28 wherein said oligonucleotide primer comprises a nucleotide sequence that is non-complementary with said target nucleic acid molecule but includes a region of internal complementarity over part of its length that forms a stem loop structure.

30. The method according to claims 3 to 29 wherein said oligonucleotide primer includes a palindromic nucleotide sequence over part of its length.

31. The method according to claims 3 to 30 wherein said oligonucleotide primer modification is the inclusion of a site that is not recognised as template for base-pairing during primer extension by the DNA polymerase in said primer.

32. The method according to claims 3 to 31 wherein said oligonucleotide primer modification is the inclusion of an abasic site in said primer.

33. The method according to claims 3 to 32 wherein said abasic site is an apurinic/apyrimidinic site.

34. The method according to claims 3 to 33 wherein said apurinic/apyrimidinic sites comprise a tetrahydrofuran.

35. The method according to claims 3 to 34 wherein said abasic site comprises at least one or at least three apurinic/apyrimidinic sites.

36. The method according to claims 3 to 35 wherein said abasic site separates the region complementary to the 3′ terminal nucleotide sequence of said single stranded nucleic acid template and the 5′ nucleotide sequence not complementary to the target nucleic acid molecule.

37. The method according to claims 3 to 36 wherein said oligonucleotide primer comprises a non-complementary nucleotide sequence comprising a transcriptional terminator.

38. A method for the transfection of primary cells isolated from a human subject comprising: i) providing an isolated sample comprising cells to be transfected; ii) forming a preparation comprising said isolated cell sample and contacting said sample with a dumbbell-shaped vector prepared using the method of any one of claims 1 to 37; iii) providing transformation conditions that enable introduction of said dumbbell-shaped vector into said primary cell sample and sustained expression of a nucleic acid molecule included in said vector.

39. An ex vivo method to treat a patient suffering from a disease that would benefit from gene therapy comprising the steps: i) obtaining a sample from said subject comprising cells to be transfected; ii) forming a cell culture preparation comprising a dumbbell shaped vector prepared using the method of any one of claims 1 to 37 and providing conditions to transfect said vector into said cells; and iii) administering the transfected cells to said subject.

40. The nucleic acid sequence as set forth in SEQ ID NO: 23.

41. A DNA vector comprising the nucleotide sequence as set forth in SEQ ID NO: 23.

Description

[0229] An embodiment of the invention will now be described by example only and with reference to the following figures:

[0230] FIG. 1. Basic scheme for universal template (UT)-assisted cloning-free dumbbell production. The universal DNA template is a double-stranded DNA that comprises an inverted repeat including all sequences required to express a coding or non-coding gene of interest such as a promoter, expression cassette, and transcriptional terminator. The inverted repeats are separated by a stretch of nucleotides. The protocol for dumbbell generation and purification comprises 4 steps. Step 1: PCR amplification of the universal template using 5′phosphorylated forward (Fw) and/or reverse (Rv) primers, both introducing either the antisense or sense strand of the sequence of interest; blocking oligos b1 and b2 are added into the reaction to suppress template refolding and to facilitate primer binding. Dotted lines represent plasmid sequences beyond the UT. Step 2: For dumbbell structure pre-folding, the PCR product is diluted, heat-denatured, and slowly cooled down to room temperature. Step 3: Dumbbell structures are covalently closed using a single-strand DNA ligase. Step 4: Treatment with T7 DNA polymerase removes oligos and non-ligated dumbbell DNA yielding covalently closed dumbbell vector DNA. Dotted arrows indicate transcribed sequences. All steps: Cyan: (+) strand UT DNA; magenta: (−) minus strand UT DNA; grey: blocking oligos; green: sense (s) sequence; yellow: antisense (as) sequence.

[0231] FIG. 2. Scheme for universal template (UT)-assisted cloning-free production of shRNA and miRNA expressing dumbbell vectors. The universal DNA template is a 262 bp double-stranded DNA that comprises an inverted repeat of the 99 bp minimal H1 promoter (mH1), a polymerase III transcriptional terminator (T.sub.5), and the hsa-miR-30 precursor stem. The inverted repeats are separated by four T's (plus strand). The protocol for dumbbell generation and purification comprises 4 steps. Step 1: PCR amplification of the universal template using 5′phosphorylated forward (Fw) and/or reverse (Rv) primers, both introducing either the antisense or sense strand of a shRNA together with half of the shRNA loop sequence; blocking oligos b1 and b2 are added into the reaction to suppress template refolding and to facilitate primer binding. Dotted lines represent plasmid sequences beyond the UT. Step 2: For dumbbell structure pre-folding, the PCR product is diluted, heat-denatured, and slowly cooled down to room temperature. Step 3: Dumbbell structures are covalently closed using a single-strand DNA ligase. Step 4: Treatment with T7 DNA polymerase removes oligos and non-ligated dumbbell DNA yielding covalently closed dumbbell vector DNA. Dotted arrows indicate transcribed sequences. All steps: Cyan: (+) strand UT DNA; magenta: (−) minus strand UT DNA; grey: blocking oligos; green: shRNA or miRNA sense (s) sequence; yellow: shRNA or miRNA antisense (as) sequence.

[0232] FIG. 3. Scheme for universal template (UT)-assisted cloning-free production of longer dumbbell vectors. The universal DNA template is a double-stranded DNA that comprises an inverted repeat including all sequences required to express a coding or non-coding gene of interest such as a promoter, expression cassette, and transcriptional terminator. The inverted repeats are separated by a stretch of nucleotides. The protocol for dumbbell generation and purification comprises 4 steps. Step 1: PCR amplification of the universal template using each a set of forward (Fw) and/or reverse (Rv) primers in which the respective outermost primers are 5′phosphorylated, both sets stepwise introducing either the antisense or sense strand of the sequence of interest; two set of blocking oligos b1.sub.n+i and b2n.sub.+i are added into the reaction to suppress template refolding and to facilitate primer binding. Dotted lines represent plasmid sequences beyond the UT. Step 2: For dumbbell structure pre-folding, the PCR product is diluted, heat-denatured, and slowly cooled down to room temperature. Step 3: Dumbbell structures are covalently closed using a single-strand DNA ligase. Step 4: Treatment with T7 DNA polymerase removes oligos and non-ligated dumbbell DNA yielding covalently closed dumbbell vector DNA. Dotted arrows indicate transcribed sequences. All steps: Cyan: (+) strand UT DNA; magenta: (−) minus strand UT DNA; grey: blocking oligos; green: sense (s) sequence; yellow: antisense (as) sequence.

[0233] FIG. 4. Scheme for universal template (UT)-assisted cloning-free production of non-coding RNA expressing dumbbell vectors. The universal DNA template is a double-stranded DNA that comprises an inverted repeat including all sequences required to express a non-coding gene of interest such as a promoter, expression cassette, and transcriptional terminator. The inverted repeats are separated by a stretch of nucleotides. The protocol for dumbbell generation and purification comprises 4 steps. Step 1: PCR amplification of the universal template using each a set of forward (Fw) and/or reverse (Rv) primers in which the respective outermost primers are 5′phosphorylated and may introduce the reverse complement of a transcriptional terminator if non-palindromic sequences are to expressed, both sets stepwise introducing either the antisense or sense strand of the sequence of interest; two sets of blocking oligos are added into the reaction to suppress template refolding and to facilitate primer binding. Dotted lines represent plasmid sequences beyond the UT. Step 2: For dumbbell structure pre-folding, the PCR product is diluted, heat-denatured, and slowly cooled down to room temperature. Step 3: Dumbbell structures are covalently closed using a single-strand DNA ligase. Step 4: Treatment with T7 DNA polymerase removes oligos and non-ligated dumbbell DNA yielding covalently closed dumbbell vector DNA. Dotted arrows indicate transcribed sequences. All steps: Cyan: (+) strand UT DNA; magenta: (−) minus strand UT DNA; grey: blocking oligos; green: sense (s) sequence; yellow: antisense (as) sequence.

[0234] FIG. 5. Scheme for universal template (UT)-assisted cloning-free production of coding RNA expressing dumbbell vectors. The universal DNA template is a double-stranded DNA that comprises an inverted repeat including all sequences required to express a coding gene of interest such as a promoter, expression cassette, and transcriptional terminator. The inverted repeats are separated by a stretch of nucleotides. The protocol for dumbbell generation and purification comprises 4 steps. Step 1: PCR amplification of the universal template using each a set of forward (Fw) and/or reverse (Rv) primers in which the respective innermost primers introduce the reverse complement of a translational start codon (5′CAT3′) and the respective outermost primers are 5′phosphorylated introducing the reverse complement of a translational stop codon, both sets stepwise introducing either the antisense or sense strand of the sequence of interest; two sets of blocking oligos b1.sub.n+i and b2.sub.n+i are added into the reaction to suppress template refolding and to facilitate primer binding. Dotted lines represent plasmid sequences beyond the UT. Step 2: For dumbbell structure pre-folding, the PCR product is diluted, heat-denatured, and slowly cooled down to room temperature. Step 3: Dumbbell structures are covalently closed using a single-strand DNA ligase. Step 4: Treatment with T7 DNA polymerase removes oligos and non-ligated dumbbell DNA yielding covalently closed dumbbell vector DNA. Dotted arrows indicate transcribed sequences. All steps: Cyan: (+) strand UT DNA; magenta: (−) minus strand UT DNA; grey: blocking oligos; green: sense (s) sequence; yellow: antisense (as) sequence.

[0235] FIG. 6. Basic scheme for universal template (UT)-assisted cloning-free generation of dumbbell-shaped dual expression vectors. The universal DNA template is a double-stranded DNA that comprises two promoters, one in each strand and which are either identical or different, and potentially other regulatory sequences. The protocol for dumbbell generation and purification comprises 4 steps. Step 1: PCR amplification of the universal template using 5′phosphorylated forward (Fw) and/or reverse (Rv) primers, both introducing a sequence of interest (the same or different sequences), the reverse complement of a transcriptional terminator (same or different), one or multiple abasic positions (gap), and a 5′terminal sequence that is capable of forming a hairpin structure; blocking oligos b1 and b2 are added into the reaction to suppress template refolding and to facilitate primer binding. Dotted lines represent plasmid sequences beyond the UT. Step 2: For dumbbell structure pre-folding, the PCR reaction is cooled down to room temperature. Step 3: Dumbbell structures are covalently closed using a double-strand DNA ligase. Step 4: Treatment with T7 DNA polymerase removes oligos and non-ligated dumbbell DNA yielding covalently closed dumbbell vector DNA. Dotted arrows indicate transcribed sequences. All steps: cyan: UT DNA; yellow and green: Sequences of interest; magenta: abasic primer site (gap); grey: blocking oligos.

[0236] FIG. 7. Detailed scheme for universal template (UT)-assisted cloning-free generation of dumbbell-shaped dual expression vectors. The universal DNA template is a double-stranded DNA that comprises two promoters P1 and P2, one in each strand and which are either identical or different. The protocol for dumbbell generation and purification comprises 4 steps. Step 1: PCR amplification of the universal template using 5′phosphorylated forward (Fw) and/or reverse (Rv) primers, both introducing a sequence of interest SOI1 and SOI2 (the same or different sequences), the reverse complement of a transcriptional terminator T1(−) (same or different), one or multiple abasic positions (gap), and a 5′terminal sequence that is capable of forming a hairpin structure; blocking oligos b1 and b2 are added into the reaction to suppress template refolding and to facilitate primer binding. Dotted lines represent plasmid sequences beyond the UT. Step 2: For dumbbell structure pre-folding, the PCR reaction is cooled down to room temperature. Step 3: Dumbbell structures are covalently closed using a double-strand DNA ligase. Step 4: Treatment with T7 DNA polymerase removes oligos and non-ligated dumbbell DNA yielding covalently closed dumbbell vector DNA comprising two expression cassettes, P1(+)-SOI1-T1(+) and P2(+)-SOI2-T2(+), for expression of sequences of interest SOI1 and SOI2. Dotted arrows indicate transcribed sequences. All steps: cyan: UT DNA; yellow and green: Sequences of interest; magenta: abasic primer site (gap); grey: blocking oligos.

[0237] FIG. 8. Sequence, cloning and validation of universal template (UT). (A), UT plus strand sequence. Cyan: miR-30 5′ and 3′ stem; magenta: transcriptional terminator (sense and antisense); orange: loop-forming tetranucleotide sequence; black: mH1 sequence (sense and antisense). (B), UT cloning scheme. 5′ phosphorylatyed oligos ODN2 and ODN3 were annealed with oligos ODN1 and ODN4, respectively. The resulting duplexes were ligated and the double-stranded UT sequence was cloned into pVax1 using the HindIII and BamHI restriction sites. Cyan: UT plus strand; magenta: UT minus strand. (C), Analytical 1% agarose gel electrophoresis showing the annealed oligo pairs ODN1/2 and ODN3/4 as well as the ligated hybrids. (D) Analytical HindIII and/or BamHI digestion and 1% agarose gel electrophoresis of the cloned plasmid pVax1-UT. HindIII/BamHI double digestion indicates the 262 bp UT band. (E), Sequencing of the HindIII and BamHI cloning sites of pVax1-UT.

[0238] FIG. 9. PCR-based universal template (UT)-assisted dumbbell production. (A) PCR amplification of the UT depends on the addition of blocking oligos. Without adding blocking oligos (−), no PCR amplification was observed and only 126 bp refolded UT single-strands were detected. When adding blocking oligos (+), both the PCR-amplified double-stranded dumbbell DNA (303 bp) and refolded dumbbell single strands (146 bp) were detected. (B) Optimization of PCR conditions. PCR yield were virtually independent of the annealing temperature in the range between 52° C. and 65° C. Addition of 5% DMSO elevated the yields of double-stranded dumbbell DNA. (C) Assessment of DNA products referring to Steps 1 to 4 defined in FIG. 1. Step 1: PCR UT amplification yields double-stranded dumbbell DNA (303 bp) and re-folded dumbbell single strands (146 bp); Step 2: Heat-denaturation and refolding converts double-stranded dumbbell DNA into dumbbell single strands; Step 3: single-stranded DNA ligation covalently closes dumbbell vector DNA if 5′phosphorylated primers were used for PCR; Step 4: Exonuclease treatment removes un-ligated DNA and yields covalently closed dumbbell vectors (146 bp).

[0239] FIG. 10. Sequences and structures of the complete plus and minus strand-derived luciferase-targeting dumbbell vectors. The lamin NC-targeting vectors harbour the same dumbbell cores.

[0240] FIG. 11. Dumbbell vector generation. Generation of 146 bp dumbbell vector DNA by 5′ phosphorylation of the PCR product (left side) or alternatively by using 5′ phosphorylated primers for the PCR reaction (right side). Steps are referring to Steps 1 to 4 defined in FIG. 1.

[0241] FIG. 12. Knock-down of lamin A/C in HEK 293T cells by plus (+) and minus (−) strand-derived lamin-targeting (Lam) dumbbell (db) vectors monitored using intracellular FACS. (A) Sequences and structures of dumbbell vectors and transcribed lamin A/C-targeting shRNAs. shRNA secondary structures were drawn according to predictions by mfold and RNAfold. (B-D) Representative histogram overlays of one experiment. (B) Stained (primary anti-lamin A+C antibody plus secondary donkey anti-rabbit IgG H&L) vs. unstained (primary anti-lamin A+C antibody only) non-transfected live cells. (C) Knockdown triggered by 0.1, 0.5 or 2.5 pmol plus strand-derived anti-lamin A/C shRNA expressing db vectors [Lam(+)db] or 3 pmol anti-lamin NC positive control siRNA (Lam-siRNA). (D) Knockdown triggered by 0.1, 0.5 or 2.5 pmol minus strand-derived anti-lamin NC shRNA expressing db vectors [Lam(−)db] or 3 pmol Lam-siRNA. (E-G) Knock-down of lamin NC in stained HEK293T cells relative to the non-transfected cells (100%) represented by the fraction of lamin NC-stained cells (E), the geometric mean fluorescence intensity (gMFI) of lamin NC-stained cells (F), and the median fluorescence intensity of lamin NC-stained cells (G). The control dumbbell (control db) was a 1:1 mix of plus and minus strand-derived luciferase targeting dumbbell DNA. Values are mean values±SEM of three independent experiments. The statistical analysis was performed using Student's t-test. P values indicate significance relative to the stained no transfection control. (B-G) NTC: No transfection control; No DNA control: Buffer transfected cells.

[0242] FIG. 13. Determination of conversion yields for the plus (A) and minus (B) strand-derived luciferase-targeting dumbbells. PCR was performed using a non-phosphorylated forward and a 5′-phosphorylated reverse primer. 6 □g PCR products were denatured, refolded, and ligated in 50 □l reaction volume using 100 U CircLigase. Ligated refolded dumbbell DNA was treated with exonuclease and directly comparable samples before and after exonuclease were analysed using 1% agarose gel electrophoresis. Only successfully ligated covalently closed dumbbell DNA resists exonuclease treatment. Yields of converting non-ligated into ligated overall dumbbell DNA where determined by quantifying band intensities using the software ImageJ v1.48. Since either only the plus or the minus strand-derived dumbbell DNA, i.e. only 50% of the total DNA, can theoretically be ligated in each reaction, the actual conversion yield of plus or minus strand-derived dumbbell DNA is twice as high as the detected overall conversion yield.

[0243] FIG. 14. Knock-down of firefly luciferase in HEK293T cells by plus (+) and minus (−) strand-derived luciferase-targeting (Luc) dumbbell (db) vectors. (A) Sequences and structures of dumbbell vectors and transcribed luciferase-targeting shRNAs. shRNA secondary structures were drawn according to predictions by mfold and RNAfold. (B) Selective generation of (+) or (−) strand-derived db vectors using either phosphorylated forward (5′P-Fw) or reverse (5′P-Rv) primers. Steps refer to Steps 1 to 4 defined in FIG. 1. (C,D) Functional validation of plus (+) and/or minus (−) strand-derived luciferase targeting db vectors. Cells were co-transfected with firefly luciferase reporter vector pGL3 and 0.5 or 1.5 pmol dumbbell vector DNA. NTC: no transfection control. Firefly luciferase mRNA (C) or expression (D) levels relative to the uninhibited negative control were measured 48 h post transfection using RT-qPCR or luciferase reporter assays. Relative RNA levels were calculated in terms of fold change (2.sup.−ΔΔCt) where ΔCt=C.sub.t luciferase−C.sub.t Δ-Actin. Values are mean values±SEM of three independent experiments. The statistical analysis was performed using Student's t-test (C) or repeated one-way ANOVA with Tukey's post hoc multiple comparison test (D).

[0244] FIG. 15. Purity control of dumbbell vectors after exonuclease treatment. (A), 1% agarose gel electrophoresis of plus (+) and minus (−) strand-derived luciferase (Luc) and lamin A/C (Lam)-targeting dumbbell (db) vectors. (B), Capillary gel electrophoresis (Bioanalyzer, Agilent) of Lam(−)db DNA. The identified 136 bp peak refers to the 146 bp dumbbell vector DNA. The difference in detected and expected vector size refers to the fact that the gel retardation of dumbbell vector DNA slightly differs from that of double-stranded DNA fragments of the DNA ladder marker.

[0245] FIG. 16. Monitoring lamin A/C expression by intracellular fluorescence-activated cell sorting (FACS). Cellular lamin A/C expression n non-transfected cells (NTC) was labelled by anti-lamin A+C primary antibody (left panels: unstained) and then stained with donkey anti-rabbit IgG H&L Alexa Fluor 647-A secondary antibody (right panels: stained). Gating of live HEK293T cells (A), Alexa Fluor 647-A scatter plots (B), and histograms (C).

[0246]

TABLE-US-00001 Sequence Listing SEQ ID NO 1: 5′- AGCTTCGCGCTCACTGAGAAGATTTTTCTGTGCTCTCATACAGAACTTATA AGATTCCCAAATCCAAAGACATTTCACGTTTATGGTGATTTCCCAGAACAC ATAGCGACATGCAAATATGAATTGTCCAGTT-3′ SEQ ID NO 2: 5′P- GGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCATAAAC GTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGAGCACAG AAAAATCTTCTCAGTGAGCGCGA-3′ SEQ ID NO 3: 5′P- TTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCAT AAACGTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGAGC ACAGAAAAATCTTCTCAGTAGGCAAAG-3′ SEQ ID NO 4: 5′- GATCCTTTGCCTACTGAGAAGATTTTTCTGTGCTCTCATACAGAACTTATA AGATTCCCAAATCCAAAGACATTTCACGTTTATGGTGATTTCCCAGAACAC ATAGCGACATGCAAATATGAATTGTCCAGAAAACT-3′ SEQ ID NO 5: 5′-tgaaggctcctcagaaacagctcCGCGCTCACTGAGAAGATTT-3′ SEQ ID NO 6: 5′-tgaaagcccagatcgtcaccacccgcCGCGCTCACTGAGAAGATTT- 3′ SEQ ID NO 7: 5′-agagaggctcctcagaaacagctcTTTGCCTACTGAGAAGATTTTTCT GT-3′ SEQ ID NO 8: 5′-agagaagcccagatcgtcaccaccttTTTGCCTACTGAGAAGATTTTT CTGT-3′ SEQ ID NO 9: 5′- GGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCATAAAC GTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGAGCACAG AAAAATCTTCTCAGTGAGCGCGA-3′ SEQ ID NO 10: 5′- TTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCAT AAACGTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGAGC ACAGAAAAATCTTCTCAGTAGGCAAAG-3′ SEQ ID NO 11: 5′-CGCTGGGCGTTAATCAAAGA-3′ SEQ ID NO 12: 5′-CTGGCACCCAGCACAATG-3′ SEQ ID NO 13: 5′-GTGTTCGTCTTCGTCCCAGT-3′ SEQ ID NO 14: 5′-GCCGATCCACACGGAGTACT-3′ SEQ ID NO 15: 5′- TTTTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACC ATAAACGTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGA GCACAGAAAAATCTTCTCAGTAGGCAAAGAGCTGTTTCTGAGGAGCCTCTC TTGAAGGCTCCTCAGAAACAGCTCGCGGCTCACTGAGAAGATTTTTCTGTG CTCTCATACAGAACTTATAAGATTCCCAAATCCAAAGACATTTCACGTTTA TGGTGATTTCCCAGAACACATAGCGACATGCAAATATGAATTGTCCAG-3′ SEQ ID NO 16: 5′- CUUCUCAGUGAGCGCGGAGCUGUUUCUGAGGAGCCUUCAAGAGAGGCUCCU CAGAAACAGCUCUUUGCCUACUGAGAAGAUU-3′ SEQ ID NO 17: 5′- TTTTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACC ATAAACGTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGA GCACAGAAAAATCTTCTCAGTGAGCGCGGAGCTGTTTCTGAGGAGCCTTCA AGAGAGGCTCCTCAGAAACAGCTCTTTGCCTACTGAGAAGATTTTTCTGTG CTCTCATACAGAACTTATAAGATTCCCAAATCCAAAGACATTTCACGTTTA TGGTGATTTCCCAGAACACATAGCGACATGCAAATATGAATTGTCCAG-3′ SEQ ID NO 18: 5′- CUUCUCAGUGAGGCAAAGAGCUGUUUCUGAGGAGCCUCUCUUGAAGGCUCC UCAGAAACAGCUCCGCGCUCUACUGAGAAGAUU-3′ SEQ ID NO 19: 5′- TTTTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACC ATAAACGTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGA GCACAGAAAAATCTTCTCAGTAGGCAAAAGCCCAGATCGTCACCACCTCTC TTGAAGGTGGTGACGATCTGGGCTCGCGCTCACTGAGAAGATTTTTCTGTG CTCTCATACAGAACTTATAAGATTCCCAAATCCAAAGACATTTCACGTTTA TGGTGATTTCCCAGAACACATAGCGACATGCAAATATGAATTGTCCAG-3′ SEQ ID NO 20: 5′- CUUCUCAGUGAGCGCGAGCCCAGAUCGUCACCACCUUCAAGAGAGGUGGUG ACGAUCUGGGCUUUUGCCUACUGAGAAGAUU-3′ SEQ ID NO 21: 5′- TTTTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACC ATAAACGTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGA GCACAGAAAAATCTTCTCAGTGAGCGCGAGCCCAGATCGTCACCACCTTCA AGAGAGGTGGTGACGATCTGGGCTTTTGCCTACTGAGAAGATTTTTCTGTG CTCTCATACAGAACTTATAAGATTCCCAAATCCAAAGACATTTCACGTTTA TGGTGATTTCCCAGAACACATAGCGACATGCAAATATGAATTGTCCAG-3′ SEQ ID NO 22: 5′- CUUCUCAGUGAGGCAAAAGCCCAGAUCGUCACCACCCUCUUGAAGGUGGUG ACGAUCUGGGCUCGCGCUCACUGAGAAGAUU-3′ SEQ ID NO 23: 5′- CGCGCTCACTGAGAAGATTTTTCTGTGCTCTCATACAGAACTTATAAGATT CCCAAATCCAAAGACATTTCACGTTTATGGTGATTTCCCAGAACACATAGC GACATGCAAATATGAATTGTCCAGTTTTCTGGACAATTCATATTTGCATGT CGCTATGTGTTCTGGGAAATCACCATAAACGTGAAATGTCTTTGGATTTGG GAATCTTATAAGTTCTGTATGAGAGCACAGAAAAATCTTCTCAGTAGGCAA A-3′ SEQ ID NO 24: 5′- TTTTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACC ATAAACGTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGA GCACAGAAAAATCTTCTCAGTAGGCAAAGAGCTGTTTCTGAGGAGCCTCTC TTGAAGGCTCCTCAGAAACAGCTCGCGGCTCACTGAGAAGATTTTTCTGTG CTCTCATACAGAACTTATAAGATTCCCAAATCCAAAGACATTTCACGTTTA TGGTGATTTCCCAGAACACATAGCGACATGCAAATATGAATTGTCCAG-3′ SEQ ID NO 25: 5′- TTTTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACC ATAAACGTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGA GCACAGAAAAATCTTCTCAGTGAGCGCGGAGCTGTTTCTGAGGAGCCTTCA AGAGAGGCTCCTCAGAAACAGCTCTTTGCCTACTGAGAAGATTTTTCTGTG CTCTCATACAGAACTTATAAGATTCCCAAATCCAAAGACATTTCACGTTTA TGGTGATTTCCCAGAACACATAGCGACATGCAAATATGAATTGTCCAG-3′

Materials & Methods

[0247] Oligodeoxyribonucleotides (ODNs) and primers Universal Template. Due to internal self-complementarity, the universal template could not be generated by gene synthesis and instead was assembled from two pairs of complementary oligodeoxyribonucleotides (IDT, Skokie, Ill.) (FIG. S1A,B): Oligo UT1: 5′-AGCTTCGCGCTCACTGAGAAGATTTTTCTGTGCTCTCATACAGAACTTATAAGATTCCCA AATCCAAAGACATTTCACGTTTATGGTGATTTCCCAGAACACATAGCGACATGCAAATAT GAATTGTCCAGTT-3′; Oligo UT2: 5′P-GGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCATAAACGTGAAATGT CTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGAGCACAGAAAAATCTTCTCAGTGAG CGCGA-3′; Oligo UT3: 5′P-TTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCATAAACGTGAA ATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGAGCACAGAAAAATCTTCTCAGT AGGCAAAG-3′; Oligo UT4: 5′-GATCCTTTGCCTACTGAGAAGATTTTTCTGTGCTCTCATACAGAACTTATAAGATTCCCAA ATCCAAAGACATTTCACGTTTATGGTGATTTCCCAGAACACATAGCGACATGCAAATATG AATTGTCCAGAAAACT-3′. Bold: HindIII and BamHI compatible overhangs; underlined: dumbbell loop-forming tetranucleotide. The 5′ phosphorylated oligos UT2 and UT3 were hybridized with the complementary oligos UT1 and UT4, respectively. The resulting UT1/UT2 and UT3/UT4 duplexes were ligated to form the universal template sequence bearing HindIII and BamHI-compatible 5′-overhangs which was subsequently cloned into pVax1 (Thermo Fisher Scientific, Waltham, Mass.) yielding the universal template vector pVax1-UT (FIG. S1B). For cloning we used the recA-deficient E. coli strain Top10. Cloning of the universal template was confirmed by PCR and by FastDigest BamHI/HindIII (Thermo Fisher Scientific, Waltham, Mass.) endonucleolytic cleavage followed by analytical agarose gel electrophoresis which yielded the expected insert size of 262 bp and by sequencing of the ligation sites (FIG. S1C,D). Sequencing of the complete universal template was unsuccessful due to the high degree of self-complementarity.

[0248] Primers for the production of firefly luciferase- or lamin A/C-targeting shRNA expressing dumbbells. Luciferase- or lamin NC-specific primers were synthesized by AITbiotech (Singapore) or IDT (Singapore). Upper case letters indicate the universal template binding sites and lower case letters indicate the shRNA coding sequences in which the loop-forming nucleotides are underlined: Forward primers: FP_Luciferase 5′-tgaaggctcctcagaaacagctcCGCGCTCACTGAGAAGATTT-3′; FP_Lamin 5′-tgaaagcccagatcgtcaccacccgcCGCGCTCACTGAGAAGATTT-3′. Reverse primers: RP_Luciferase 5′-agagaggctcctcagaaacagctcTTTGCCTACTGAGAAGATTTTTCTGT-3′. RP_Lamin 5′-agagaagcccagatcgtcaccaccttTTTGCCTACTGAGAAGATTTTTCTGT-3′. Blocking ODNs: Two blocking ODNs (IDT, Skokie, Ill.) were added to the PCR to suppress refolding and self-priming of the universal template strands. Block_1:

TABLE-US-00002 5′- GGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCATAAAC GTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGAGCACAG AAAAATCTTCTCAGTGAGCGCGA-3′; Block_2: 5′- TTCTGGACAATTCATATTTGCATGTCGCTATGTGTTCTGGGAAATCACCAT AAACGTGAAATGTCTTTGGATTTGGGAATCTTATAAGTTCTGTATGAGAGC ACAGAAAAATCTTCTCAGTAGGCAAAG-3′.

[0249] Primers for quantitative reverse transcription PCR (RT-qPCR). Primers for the quantification of luciferase and β-actin mRNA levels were synthesized by AITbiotech (Singapore). PCR forward primers: qPCR_FP_Luciferase 5′-CGCTGGGCGTTAATCAAAGA-3′; qPCR_RP_β-actin 5′-CTGGCACCCAGCACAATG-3′. Reverse transcription and PCR reverse primers: qPCR_RP_Luciferase 5′-GTGTTCGTCTTCGTCCCAGT-3′; qPCR_RP_β-actin 5′-GCCGATCCACACGGAGTACT-3′.

[0250] Primer phosphorylation. For the generation of strand-specific dumbbell vectors, either the forward or the reverse primers were 5′-phosphorylated. Each 50 pmol primer were incubated with 10 U T4 polynucleotide kinase (Thermo Fisher Scientific, Waltham, Mass.) in the presence of 1 mM ATP at 37° C. for 20 mins followed by heat inactivation of the enzyme at 75° C. for 10 mins.

Dumbbell Vector Generation

[0251] PCR amplification of dumbbell vector DNA. PCR amplification of the universal template and appendage of shRNA encoding DNA was carried out using 1 U Taq DNA Polymerase (Invitrogen), 1.0 μM of each primer and blocking ODNs, 0.2 mM of each dNTP (Invitrogen), 100 ng of HindIII/BamHI cleaved pVax1-UT, 5% v/v DMSO (Thermo Fisher Scientific, Waltham, Mass.) in a reaction volume of 30-50 μl in 1×Taq DNA Polymerase buffer (Invitrogen). Linearisation of pVax1-UT usually improves the PCR yields but is not essential. Thermal cycling was carried out as follows: Initial denaturation at 96° C. for 5 mins; 27 cycles of denaturation (95° C., 30 sec), annealing (59° C., 30 sec), and extension (72° C., 1 min); and final extension at 72° C. for 10 mins. A 50 μl PCR reaction yielded about 10 μg DNA.

[0252] Strand separation and annealing. PCR products were purified through silica-membrane based spin columns (QIAquick PCR purification kit, Qiagen, Hilden, Germany). Purified products were diluted to 400 μl in 1× hybridisation buffer (1 M NaCl, 100 mM MgCl.sub.2, and 200 mM Tris-HCl, pH 7.4), heat-denatured at 96° C. for 5 mins followed by gradual cooling to room temperature to allow for intramolecular folding of plus and/or minus strand dumbbell vectors. The resulting DNA was concentrated using ethanol precipitation, pelleted by centrifugation, and resuspended in nuclease-free water.

[0253] Ligation of single-stranded loop DNA. 1 to 6 μg (˜10 to 60 pmol) of DNA was incubated with 2.5 mM MnCl.sub.2, 1 M Betaine (Sigma, St. Louis, Mo.) and 50 to 100 U CircLigaseII (Epicenter, Madison, Wis.) in 1× CircLigaseII reaction buffer at 60° C. for 16 hours, followed by heat inactivation of the ligase at 80° C. for 10 mins. Highest conversion yields were observed when ligating 6 μg DNA with 100 U CircLigase.

[0254] Exonuclease treatment. After ligation, products were treated with 10 U of T7 DNA polymerase (Thermo Fisher Scientific, Waltham, Mass.) at 37° C. for 1 h followed by heat inactivation at 80° C. for 10 mins. Products were assessed on 10% native polyacrylamide gels or 1% agarose gels, stained with ethidium bromide post electrophoresis and/or purified using phenol-chloroform-isoamylalcohol (25:24:1) extraction (1×), chloform-isoamylalcohol (24:1) re-extraction (3×), and ethanol precipitation.

Target Gene Knockdown Assays

[0255] Luciferase knockdown assays. HEK293T cells were maintained in Dulbecco's Modified Eagle's Medium (Hyclone, South Logan, Utah) supplemented with 10% (v/v) fetal bovine serum (Hyclone, South Logan, Utah) and 1% penicillin-streptomycin antibiotic solution (Thermo Fisher Scientific, Waltham, Mass.). 24 hours prior to transfection, 2×10.sup.4 cells/well were seeded in a 96-well plate. Cells were co-transfected with 100 ng of luciferase expression plasmid pGL3 (Promega, Madison, Wis.) and 1.5 pmol or 0.5 pmol of either plus- or minus-strand dumbbell vector DNA using Lipofectamin 2000 (Thermo Fisher Scientific, Waltham, Mass.) and a reagent:DNA ratio of 1:2.5. For the positive control (pGL3 only) empty pVax1 (Thermo Fisher Scientific, Waltham, Mass.) was used as feeder DNA to ensure all cells received the same quantity of DNA. 48 hours post transfection, cells were washed with sterile PBS and lysed in 20 μl passive lysis buffer (Promega, Madison, Wis.) for 20 mins employing gentle shaking. 10 μl of lysate was treated with 50 μl of LARII reagent (Promega, Madison, Wis.) and luminescence was quantified on the Biotek Reader (Biotek Instruments, Winooski, Vt.).

[0256] Monitoring lamin A/C knockdown by intracellular fluorescence-activated cell sorting (FACS). HEK293T cells were cultivated and seeded in 96-well plates 24 hours prior to transfection as described above. Cells were transfected with 0.1, 0.5 or 2.5 pmol dumbbell vector DNA or 3 pmol siGENOMELamin A/C control siRNA (Dharmacon, Lafayette, Colo.) using Lipofectamine 3000 (Thermo Fisher Scientific, Waltham, Mass.) according to the manufacturer's protocol, Medium was changed 24 hours post transfection and cells were harvested after 48 hours. For FAGS analyses, the media was aspirated and the cells were rinsed once with PBS before trypsinisation with 50 μl of 1× trypsin-EDTA (Gibco), Trypsinised cells were collected by centrifugation at 4200 rpm for 6 min in 200 μl media. Pelleted cells were resuspended in 100 μL media, fixed and permeabilized with Intracellular Fixation and Permeabilization Buffer Set (eBioscience, San Diego, Calif.) according to manufacturer's protocol prior to intracellular staining. To assess lamin A/C knockdown, cellular lamin A/C was stained by anti-lamin A+C antibody (ab133256) (1/200) and donkey anti-rabbit IgG AF647 (ab150075) (1/200) (Abeam, Cambridge, UK), FACS was performed on LSRFortessa cell analyser, and FACSDiva software v6.1.3 (BD Biosciences, San Jose, Calif.) was used for the acquisition of the samples. Flow Jo software V10.5.2 (Tree Star, Ashland, Oreg.) was used for data analyses.

Computational Secondary Structure Prediction

[0257] Minimum free energy secondary structures of DNA and RNA were folded using the algorithms mfold and/or RNAfold..sup.17,18

Statistical Analysis

[0258] Diagrams represent mean values±SEM of three independent experiments. The statistical analysis was performed using repeated one-way ANOVA with Tukey's post hoc multiple comparison's test (luciferase knock-down data) or using Student's t-test (lamin NC knock-down data). The GraphPad Prism version 7 software (GraphPad, La Jolla, Calif.) was used for the statistical analysis. P values are as indicated.

Example

Universal Template-Assisted, Cloning- and Endonuclease-Free Method for the Generation of Dumbbell-Shaped Vectors

[0259] In current state-of-the-art protocols, the generation of every new dumbbell vector starts with individual cloning the sequence of interest to be implemented into the dumbbell into a plasmid vector. The method of this invention requires to prepare only one universal template, comprising regulatory sequences such as promoter, enhancer, DNA nuclear localisation signal, intron, transcriptional terminator, RNA nuclear export signal, WPRE or others. The sequences of interest are then introduced via chemically synthesised PCR primers and no further cloning and no endonucleases are required (FIGS. 1-7). Such generated dumbbells can function as molecular decoys or expression vectors. The expression vectors can express non-coding RNA including shRNA, pre-miRNA, miRNA, aptamers, antisense RNA, and antisense miRNA (FIGS. 1-4, 6, 7) and/or express peptide or protein coding RNA (FIGS. 5-7). The expression cassettes for hairpin-structured RNA such as shRNA and pre-miRNA can be designed as hairpin template-transcribing expression cassettes in which the expression cassette resembles the hairpin structure of the transcribed RNA. In these vectors, redundant sequences are eliminated and transcription goes around one of the dumbbell loops. The hairpin template-transcribing dumbbell vectors can be generated using the method of this invention (FIGS. 1-5).

[0260] The new method is based on PCR amplification of a universal DNA template which, for shRNA expression, comprises an inverted repeat of (i) the minimal H1 promoter,.sup.15 (ii) a polymerase III transcriptional terminator (T.sub.5), and (iii) the hsa-miR-30 precursor stem (FIG. 8A). The hsa-mir-30 stem was reported to facilitate shRNA processing and has been successfully implemented in dumbbell vector design..sup.13,16 Once generated, the universal template can be used for cloning-free generation of any shRNA expressing dumbbell vectors. Sequences coding for the expression of the respective small RNA are introduced during the PCR by the PCR primers (Step 1). Irrespective of the small RNA-specific 5′ portion of the PCR primers, they all harbour the same 3′ terminal target binding sites which facilitates parallelised PCR amplifications. Both strands of the universal DNA template have a high degree of self-complementarity and to improve its amplification, blocking oligos are added to the PCR reaction to suppress intramolecular refolding of the denatured DNA and to support primer binding. Each of the two DNA strands (+ and −) of the resulting double-stranded PCR product yields, after dilution, heat denaturation and intramolecular refolding, an open dumbbell scaffold with dangling 5′ and 3′ ends (Step 2). These ends are then ligated using a single-stranded DNA (ssDNA) ligase (Step 3). All DNA molecules harbouring 5′ or 3′ ends are removed by exonuclease digestion yielding clean, covalently closed dumbbell vectors (Step 4). With the decision of using either one or two 5′-phosphorylated PCR primers, the plus strand, the minus strand or both strands will produce dumbbell vectors.

[0261] Generation of the universal template was challenging due to the high degree of self-complementarity and all attempts to generate the universal template by gene synthesis failed. Instead the universal template was assembled from two pairs of complementary oligodeoxyribonucleotides (oligos) in which the self-complementary sequence portions were separated from each other (FIG. 8A-C). Pairs of complementary oligos were annealed each forming one complementary 3′-overhang and either a HindIII or BamHI 5′-overhang. Pairs of annealed oligos were then first ligated using the adhesive 3′ ends, gel purified, and inserted into the cloning vector pVax1 using the HindIII and BamHI cloning sites yielding the universal template vector pVax1-UT. Successful cloning of the universal template was proven by analytical restriction endonuclease cleavage and subsequent gel electrophoresis of the fragments as well as by sequencing: The HindIII/BamHI double digestion yielded the expected insert size of 262 bp; sequencing of the complete insert was unsuccessful due to insert self-complementarity but the cloning sites could be sequenced (FIG. 8D,E).

[0262] Next we aimed to PCR-amplify the universal template using primers that introduced the sequence coding for a published firefly luciferase-targeting shRNA..sup.9 However, intrinsic self-complementarity of the universal template was impeding conventional PCR amplification which didn't yield any product of the expected size. Products were observed after adding two long blocking oligos into the PCR reaction (FIG. 9A). These blocking oligos were designed such that they were complementary to the respective 5′ half of the plus or the minus strand of the universal template, thus suppressing intramolecular strand refolding and facilitating primer binding to the 3′ ends of the template DNA (FIG. 2). Because the blocking oligos bind to the universal template sequence, they represent a constant, target- and shRNA-independent component of this dumbbell generation protocol. The obtained PCR products corresponded in size with the double-stranded universal temple (303 bp) and the refolded single strands (146 bp). Addition of 5% (v/v) DMSO into the PCR reaction yielded more of the larger product, indicating a more efficient amplification as primer binding and extension competed more successfully with single-strand refolding (FIG. 9B). Heat denaturation and refolding of the purified PCR products then yielded more of the hairpin structured single-strands (FIG. 9C, lanes 2). As expected, the ssDNA ligation (lanes 3) and subsequent exonuclease digestion (lanes 4) yielded exonuclease-resistant dumbbell vector DNA only if 5′-phorsphorylated primers were used for the PCR (FIG. 9C).

[0263] With the decision to use either a 5′-phosphorylated forward primer, a 5′-phoshphorylated reverse primer or two phosphorylated primers for PCR, only (i) plus strand-derived dumbbells, or (ii) minus strand-derived dumbbells, or (iii) a mix of both can be generated (FIGS. 2 and 10). In order to obtain a mix of plus and minus strand-derived dumbbells, it doesn't make a difference if the 5′ ends of the PCR primers or alternatively of the PCR product are phosphorylated (FIG. 11). In this example, plus and minus strand-derived dumbbells and the expressed shRNAs are very similar but not identical as they differ with regard to sequence and structure in the loops and in the hsa-miR-30 stem (FIGS. 12A and 10). The asymmetry in the mi RNA stem region is owed to the fact that correct transcription of a partly mismatched miRNA precursor RNA can only be achieved if the hairpin template-transcribing dumbbell harbours corresponding mismatches as well. Consequently, only the plus strand-derived dumbbell expresses the shRNA extended with the original miR-30 stem (FIG. 12A). The shRNA expressed from the minus strand-derived dumbbell is extended with a miR-like stem formed by the antisense sequences of miR-30 and carries a loop that represents the reverse complement of the loop in the plus strand-derived shRNA. The observed conversion yield, i.e. the fraction of refolded 146 bp dumbbell vector DNA that was successfully ligated and resisted subsequent exonuclease treatment, was measured to be 34% or 28% for the production of the plus or minus strand-derived luciferase-targeting dumbbells (FIG. 13). Considering, that only one PCR primer was phosphorylated for the generation of these dumbbells and that consequently only half of the refolded DNA could theoretically be ligated, then the actual conversion yield of ligatable plus or minus strand-derived dumbbell DNA is 68% or 56%. In these reactions we ligated 6 μg of DNA using 100 U of CircLigase.

[0264] Employing the above protocol using either phosphorylated forward or reverse primers, we generated both plus and minus strand-derived luciferase- or lamin NC-targeting dumbbells in separate reactions (FIGS. 12 and 14). The purity of the vectors after exonuclease treatment was controlled using agarose gel electrophoresis (FIG. 15A). Additional capillary gel electrophoresis determined the purity of the minus strand-derived lamin NC-targeting dumbbell to be 83% (FIG. 15B). To measure dumbbell vector-triggered luciferase knockdown, HEK293T cells were co-transfected with the luciferase expression vector pGL3-Control and 0.5 or 1.5 pmol of plus or minus strand-derived dumbbell vector DNA using Lipofectamine 2000. 48 hours post transfection, firefly luciferase mRNA and activity levels were quantified relative to the pGL3-Control vector (FIG. 12C,D). Both dumbbells triggered a significant, dose-dependent luciferase knockdown which surprisingly was more pronounced in case of the minus strand-derived dumbbell vector indicating the non-natural miR-like stem was functional. The knockdown triggered by the plus strand-derived dumbbell was 85% (p<0.001) or 50% (p<0.001) at 1.5 or 0.5 pmol vector DNA, and the minus strand-derived dumbbell triggered 97% (p<0.001) or 75% (p<0.001) knockdown at 1.5 or 0.5 pmol DNA, respectively relative to the pGL3 positive control. To investigate the knockdown of lamin NC, HEK293T cells were transfected with 0.1, 0.5 or 2.5 pmol of plus or minus strand-derived dumbbell vector DNA or alternatively with 3 pmol siGENOMELamin A/C positive control siRNA or 0.5 pmol luciferase-targeting dumbbell control vector DNA (1:1 mix of plus and minus strand-derived dumbbells) using Lipofectamine 3000. 48 hours post transfection, intra-cellular lamin NC was stained using rabbit anti-lamin A+C primary antibody and donkey anti-rabbit IgG H&L AF647 secondary antibody, and lamin A/C knock-down was monitored by flow cytometry analyses (FIGS. 14,16). While the plus strand-derived dumbbell triggered a significant, dose-dependent lamin A/C knock-down at 2.5 or 0.5 pmol DNA, the knock-down observed with the minus strand-derived dumbbell was less pronounced.

[0265] The protocol described here combines all the advantages of previously reported protocols for dumbbell vector production. It represents (i) a cloning-free protocol which (ii) does not involve any restriction or nicking endonucleases, it (iii) employs an efficient intra-molecular ligation reaction, and it (iv) allows production of extremely small hairpin template-transcribing dumbbell vectors. The previously described gap-primer PCR protocol also involves an intra-molecular ligation but requires a cloning step for the generation of every new vector and it is not suitable to generate hairpin template-transcribing vectors due to the presence of abasic sequence positions..sup.14 Conversely, the method described by Jiang et al. is suitable to produce hairpin template-transcribing dumbbells but requires restriction and nicking endonucleases and involves a less efficient inter-molecular ligation reaction..sup.9,13 The PCR primers used for the protocol reported here always harbour the same 3′ terminal template binding sites as well as a 5′ terminal sequence that depends on and changes with the respective small RNA but which is to a great extent identical within each respective primer pair. Hence, the primer annealing temperatures are always the same and primer dimer formation can widely be excluded which both facilitates parallelised PCR reactions using a single cycling programme. The subsequent ligation reaction represents an intramolecular ligation which is generally more efficient compared with alternative protocols involving intermolecular loop ligation. As a corollary, the conversion yields observed for this method are higher than those reported for protocols employing inter-molecular ligation reactions. For the gap-primer PCR method, higher conversion yields of up to 92% were observed when ligating double-stranded nicked dumbbell DNA using the T4 DNA ligase; however, only slightly higher conversion yields of 75% were observed with the gap-primer PCR method when ligating dangling single-stranded 5′ ends with base-paired 3′ ends using the CircLigase. The purity of dumbbell DNA produced with the method described here was within the purity range of 82% to 94% of vectors produced with the gap-primer PCR method. Additional purification steps will be required for future pre-clinical and clinical applications.

[0266] We demonstrate the proof-of-principle that this new method can generate partly mismatched shRNA-expressing dumbbell vectors indicating the technology might also be explored for the generation of miRNA-expressing dumbbells. Mismatches in dumbbell vectors were reported earlier and demonstrated not to impair vector activity..sup.13 On the contrary, terminal single-nucleotide mismatches were found to improve nuclear targeting and activity of dumbbell-shaped expression vectors..sup.14

[0267] We observed that among the luciferase- or lamin NC-targeting dumbbells, the minus or plus strand-derived dumbbell exhibited a stronger target gene knockdown activity, respectively. This difference might be assigned to differences with regard to the efficiency and accuracy of endogenous shRNA processing by Dicer which depends on the sequence and structure of shRNA loops and stems. Here we employed the hsa-miR-30 stem, as miRNA stems were reported to facilitate shRNA processing and knockdown activities in most of the cases..sup.16 Consequently, though the respective plus and minus strand-derived dumbbells code for identical guide sequences, the transcribed small hairpin RNAs comprise different microRNA stems and different loops as emphasized above. Hence, differences in Dicer processing might lead to different guide RNA levels and/or differences with regard to the exact 5′ and 3′ termination of guide RNA sequences. These differences can account for the observation that, depending on the targeted sequences and the corresponding guide RNA sequences and structures, either the plus or the minus strand-derived dumbbell triggers stronger target gene knockdown. However, when forgoing the inclusion of a miRNA stem and when concurrently considering palindromic loop sequences, both plus and minus strand-derived dumbbells would be identical and a single reaction would generate a single vector only.

[0268] In conclusion, this novel method efficiently generates size-minimised hairpin template-transcribing dumbbells in a short period of time and at low costs and can be explored for the parallelised production of shRNA or miRNA expression vectors for functional genomics screens or drug development.

REFERENCES

[0269] 1. Biasco, L., Baricordi, C. and Aiuti, A. (2012). Retroviral Integrations in Gene Therapy Trials. Mol. Ther. 20, 709-716. [0270] 2. Yin, H., Kanasty, R. L., Eltoukhy, A. A., Vegas, A. J., Dorkin, J. R. and Anderson, D. G. (2014). Non-viral vectors for gene-based therapy. Nat. Rev. Genet. 15, 541-555. [0271] 3. Mok, P. L., Cheong, S. K., Leong, C. F., Chua, K. H. and Ainoon, 0. (2012). Extended and stable gene expression via nucleofection of MIDGE construct into adult human marrow mesenchymal stromal cells. Cytotechnology. 64, 203-216. [0272] 4. Kaur, T., Slavcev, R. A. and Wettig, S. D. (2009). Addressing the Challenge: Current and Future Directions in Ovarian Cancer Therapy. Curr. Gene Ther. 9, 434-458. [0273] 5. Lopez-Fuertes, L., Perez-Jiménez, E., Vila-Coro, A. J., Sack, F., Moreno, S., Konig, S. A., et al. (2002). DNA vaccination with linear minimalistic (MIDGE) vectors confers protection against Leishmania major infection in mice. Vaccine. 21, 247-257. [0274] 6. Schakowski, F., Gorschlüter, M., Junghans, C., Schroff, M., Buttgereit, P., Ziske, C., et al. (2001). A Novel Minimal-Size Vector (MIDGE) Improves Transgene Expression in Colon Carcinoma Cells and Avoids Transfection of Undesired DNA. Mol. Ther. 3, 793-800. [0275] 7. Zanta, M. A., Belguise-Valladier, P. and Behr, J.-P. (1999). Gene delivery: A single nuclear localization signal peptide is sufficient to carry DNA to the cell nucleus. Proc. Natl. Acad. Sci. USA 96, 91-96. [0276] 8. Fogg, J. M., Kolmakova, N., Rees, I., Magonov, S., Hansma, H., Perona, J. J. et al. (2006). Exploring writhe in supercoiled minicircle DNA. J Phys Condens Matter 18: S145-S159. [0277] 9. Jiang, X., Yu, H., Teo, C. R., Tan, G., Goh, S. C., Patel, P., et al. (2016). Advanced design of dumbbell-shaped genetic minimal vectors improves non-coding and coding RNA expression. Mol. Ther. 24(9), 1581-1591. [0278] 10. Cost, G. J. (2007). Enzymatic ligation assisted by nucleases: simultaneous ligation and digestion promote the ordered assembly of DNA. Nat. Protoc. 2, 2198-2202. [0279] 11. Taki, M., Kato, Y., Miyagishi, M., Takagi, Y. and Taira, K. (2004). Small-Interfering-RNA Expression in Cells Based on an Efficiently Constructed Dumbbell-Shaped DNA. Angew. Chem., Int. Ed. 43, 3160-3163. [0280] 12. Taki, M., Kato, Y., Miyagishi, M., Takagi, Y., Sano, M. and Taira, K. (2003). A Direct and efficient synthesis method for dumbell-shaped linear DNA using PCR in vitro. Nucleic Acids Symp. Ser. 3, 191-192. [0281] 13. Jiang, X. and Patzel, V. (2017). Formation of Minimised Hairpin Template-transcribing Dumbbell Vectors for Small RNA Expression. Bio-Protocol. 7(11), 1-10. [0282] 14. Yu, H., Jiang, X., Hang, L., Tan, K. T. and Patzel, V. (2015). Efficient Production of Superior Dumbbell-Shaped DNA Minimal Vectors for Small Hairpin RNA Expression. Nucleic Acids Research. 43(18), e120. [0283] 15. Myslinski, E., Ame, J. C., Krol, A. and Carbon, P. (2001). An unusually compact external promoter for RNA polymerase III transcription of the human H1 RNA gene. Nucleic Acids Res. 29, 2502-2509. [0284] 16. Zeng, Y., Wagner, E. J. and Cullen, B. R. (2002). Both natural and designed micro RNAs can inhibit the expression of cognate mRNAs when expressed in human cells. Mol. Cell. 9, 1327-1333. [0285] 17. Zuker, M. (2003). Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res. 31, 3406-3415. [0286] 18. Hofacker, I. L. (2003). Vienna RNA secondary structure server. Nucleic Acids Res. 13, 3429-3431.