JP4444564B2 - Method for purification and detection of target double-stranded DNA sequence by triple helix interaction - Google Patents

Description

  The present invention relates to a novel target DNA sequence capable of forming a triple helical structure and a novel DNA purification method. More specifically, the purification method of the present invention utilizes hybridization of a target DNA sequence with an oligonucleotide. The method of the present invention has been found to be particularly useful because pharmaceutical quality double-stranded DNA can be purified in high yield.

  The present invention also relates to a novel method for detecting, quantifying, isolating or selecting a DNA molecule comprising the specific target sequence.

  The purification method of the present invention is mainly based on triple helix interaction of an oligonucleotide composed of a specific target DNA sequence and a natural or modified base.

  Homopyrimidine oligonucleotides have been shown to interact specifically in the large groove of the DNA double helix and form locally a triple stranded structure called the triple helix (Moser et al., Science 238 (1987) 645; Povsiz et al., J. Am. Chem. 111 (1989) 3059). These oligonucleotides selectively recognize a DNA double helix in a region having an oligopurine-oligopyrimidine sequence, that is, an oligopurine sequence in one strand and an oligopyrimidine sequence in a complementary strand, and locally have a triple helix therein. Form. The base of the third homopyrimidine oligonucleotide chain forms a hydrogen bond (Hoogsteen-type bond) with the Watson-Crick base pair purine.

  Similarly, a triple helical structure can also be formed between a homopurine oligonucleotide and a homopurine-homopyrimidine double-stranded DNA. In this type of structure, the purine base of the oligonucleotide forms a reverse Hoogsteen-type bond with the purine base of the double-stranded DNA.

  These site-specific triple helix interactions are utilized in particular by Looney et al. (Science 241 (1988) 456) and Helene et al. (BBA 1049 (1990) 99; WO 95/18223), the former regulating the expression of certain genes The latter describes the formation of a triple helix structure between the target sequence present in the promoter or coding region and the oligonucleotide, and these are likely to be inhibited by RNA polymerase at the initiation and / or extension level. The gene expression profile can be regulated.

  The use of this type of triple helix interaction to purify plasmid DNA from a complex mixture of DNA molecules mixed with other components is also described in international application WO 96/18744 for the purification of plasmid DNA. More specifically, this application describes a method for purifying double-stranded DNA comprising contacting a complex mixture with a carrier covalently bound to an oligonucleotide capable of forming a triple helix by hybridization with a specific sequence of a target DNA. It is described.

In this triple helix specific interaction for purification purposes, the specificity is Hoogsteen type hydrogen between the third strand thymine (T) base composed of oligonucleotide and the AT base pair of double stranded DNA. It is due to pairing by binding to form a T ^* AT triad. Similarly, protonated cytosine located in the third strand pairs with the GC base pair of double-stranded DNA to form a ⁺ C ^* GC triad (Sun et al., Curr. Opin Struc Biol. 3 (1993) 345). ). These T ^* AT and ⁺ C ^* GC triads (referred to as canonical triads) have been conventionally accepted to ensure the maximum stability of the triple helix. However, stabilization of the triple helix also involves a number of other factors such as, for example, cytosine percentage, pH, medium salinity, and triple helix environment. It has also been widely described that the introduction of so-called non-canonical (ie, different from T ^* AT and ⁺ C ^* GC triads) triads results in considerable structural deformation at the triple helix level, which inevitably becomes quite unstable. Furthermore, the introduction of various non-canonical triads has also been examined within the scope of comparative studies (Roberts et al., Proc. Natl. Acad. Sci. USA 88, 9397; Fossella et al., (1993) Nucleic Acids Research 21, 4511; Govers et al. , Nucleic Acids Research (1997) 25, 3787), it has also been found that the instability of the triple helix varies depending on the type of non-canonical triad introduced.

  This method can rapidly and efficiently purify pharmaceutical quality target DNA, but it is necessary that a sufficiently long sequence, preferably a complete homopurine sequence, is present in one of the two strands of the DNA to be purified. In addition, this sequence must be complementary to the third DNA strand. This sequence may be naturally occurring in the target double-stranded DNA sequence desired to be purified, or may be artificially inserted.

  We have DNA molecules that have a target DNA sequence in one strand that is not composed primarily of purine bases, despite the presence of bases that are not complementary to the bases of the oligonucleotide that results in the formation of non-canonical triads. It has now been surprisingly discovered that a stable triple helix structure can be formed with three DNA strands.

  More specifically, the target double-stranded DNA sequence newly identified by the present inventors includes a homopurine sequence interrupted with a predetermined number of pyrimidine bases in one strand. We have also discovered that using these incomplete homopurine-homopyrimidine DNA sequences, DNA molecules containing these sequences can be efficiently purified by triple-helix interactions.

  The newly identified sequences are further particularly useful for detecting, quantifying, isolating or sorting DNA molecules containing these sequences.

Accordingly, the present invention provides a general formula:
5 '-(R) _n- (N) _t- (R') _m- 3 '
(In the formula, R and R ′ represent a nucleotide sequence composed only of purine bases, n and m are integers of 8 or less, the sum of n + m is 6 or more, and N is both a purine base and a pyrimidine base. A new target DNA sequence comprising one sequence with a sequence having a nucleotide sequence comprising t, wherein t is an integer of 7 or less), said DNA sequence interacting with a third DNA strand to form a triple helix It is possible.

Accordingly, the homopurine sequences R and R ′ located in the 5 ′ and 3 ′ portions of the target DNA sequence have a total length of 6 or more. These sequences contain adenine and guanine bases that can interact with the third strand to form a triple helical structure composed of T ^* AT and ⁺ C ^* GC canonical triads. The homopurine sequences R and R ′ preferably comprise at least a total of 2 guanines and at least 2 adenines. More preferably, these purine sequences contain an (AAG) type motif.

  The length t of the central sequence N is less than 7 purine and pyrimidine base pairs, and according to the present invention it is possible to interact with the third DNA strand to form a non-canonical triad. The length of the central array N is preferably 1 or more and 7 or less. The length of the central array N is more preferably 2 or more and 7 or less.

The term “canonical triad” refers to a two nucleotide triad in which the double-stranded DNA AT and GC doublets interact with T and ⁺ C bases to become T ^* AT and ⁺ C ^* GC triads, respectively. These two triads have the greatest stability of the 16 existing triads.

The term “non-canonical triad” refers to the other 14 nucleotide triads. These triads are formed by non-specific interaction of double stranded DNA and a third DNA strand and are less stable than T ^* AT and ⁺ C ^* GC canonical triads. T ^* CG and T ^* GC non-canonical triads formed by the interaction of CG or GC doublet of the target sequence and thymine (T) of the third strand, CG doublet of target sequence and guanine (G) of the third strand, respectively. G ^* CG non-canonical triad formed by interaction, C ^* AT and C ^* TA non-canonical triad formed by the interaction of AT and TA doublets of the target sequence and cytosine (C) located in the third strand, respectively. G ^* CG non-canonical triad formed by the interaction of CG doublet and third chain guanine (G), or T ^* TA formed by the interaction of TA doublet and third chain thymine (T) Can be mentioned.

Of course, as well as the 5 ′ and 3 ′ purine ends, the central sequence N also has a T due to the interaction of the thymine (T) and cytosine (C) bases located in the third DNA strand with the AT and GC doublets, respectively. ^* AT and ⁺ C ^* GC canonical triads can be formed.

The central sequence N preferably includes purine and pyrimidine bases that form up to six non-canonical triads. The non-canonical triad formed by the interaction of the central portion and the oligonucleotide is preferably selected from T ^* CG, T ^* GC, C ^* AT and C ^* TA non-canonical triad. Examples of suitable distributions of these triads are the formation of 6 non-canonical triads, including 1 C ^* AT, 1 C ^* TA, 2 T ^* CG and 2 T ^* GC, 2 Formation of 5 non-canonical triads comprising 2 C ^* ATs and 3 T ^* CG, or 3 non-canonical triads comprising 2 T ^* GCs and 1 C ^* AT. There may be several T ^* TA non-canonical triads, but in this case they are not arranged consecutively in a triple helix.

The central sequence preferably comprises a maximum of three C or T pyrimidine bases that form T ^* CG and C ^* TA or G ^* TA non-canonical triads. The three pyrimidine bases are not contiguous and can interact with the third DNA strand to form T ^* CG and C ^* AT non-canonical triads and T ^* AT and ⁺ C ^* GC canonical triads A or G It is preferred that they are separated by a purine base.

  According to one particular embodiment of the invention, the target double-stranded DNA sequence is the sequence 5'-AA GAA GCA TGC AGA GAA GAA-3 '(SEQ ID NO: 1).

The third DNA strand capable of interacting with the double-stranded DNA sequence of the present invention can be, for example, an oligonucleotide or another double-stranded DNA strand in a locally unpaired state, with the following bases: Can be included.
A thymine capable of forming an AT doublet of the target double-stranded DNA sequence and a T ^* AT canonical triad, and a CG and GC doublet of the target DNA sequence, respectively, and a T ^* CG and T ^* GC non-canonical triad ( T) (Soyfer et al., In Triple Helical Nucleic Acids (1996) Springer, New York, pp. 151-193);
-Guanine (G) capable of forming a G ^* TA triad with a TA doublet of double-stranded DNA (Soyfer et al., In Triple Helical Nucleic Acids (1996) Springer, New York, pp. 151-193);
A cytosine (C) capable of forming a GC, AT and TA doublet of the target double-stranded DNA, respectively, and ⁺ C ^* GC (protonated cytosine ⁺ C) or C ^* AT and C ^* TA non-canonical triad; and -Uracil capable of forming triplets with AU or AT base pairs of the target sequence (Bates et al., Nucleic Acids Research 23 (1995) 3627).

  The third DNA strand used preferably contains a cytosine-rich homopyrimidine sequence, which is present in protonated state at acidic pH and stabilizes the triple helix. Such oligonucleotides can include, for example, (CCT) n sequences, (CT) n sequences, or (CTT) n sequences, where n is an integer from 1-20. It is particularly advantageous to use sequences of the (CT) n or (CTT) n type, or sequences with (CCT), (CT) or (CTT) motifs attached.

  When the third DNA strand exists as an oligonucleotide, it may be natural (that is, may be composed of an unmodified natural base) or may be chemically modified. In particular, it is advantageous if the oligonucleotide is chemically modified in place to make it resistant or protected against nucleases or to increase its affinity for a specific sequence.

  According to the present invention, the term “oligonucleotide” means any nucleoside chain whose main chain has been modified in order to increase nuclease resistance. Possible modifications include phosphorothioate oligonucleotides (Xodo et al., Nucleic Acids Research, 22 (1994) 3322) capable of forming a triple helix with DNA, and oligonucleotides with formacetal or methyl phosphonate backbone (Mattuccicci). Et al., J. Am. Chem. Soc., 113 (1991) 7767). Similarly, oligonucleotides synthesized with nucleotide alpha anomers capable of forming triple helices with DNA can be used (Le Doan et al., Nucleic Acids Research, 15 (1987) 7749). Another modification of the backbone is a phosphoramidite linkage. For example, the phosphoramidite N3′-P5 ′ internucleotide linkage described by Gryaznov et al. (J. Am. Chem. Soc., 116 (1994) 3143) can be mentioned, and an oligo that forms a particularly stable triple helix with DNA. Nucleotides are obtained. Other backbone modifications can also include the use of ribonucleotides, 2'-O-methyl ribose, or phosphodiesters (Sun et al., Curr. Opinion in Structure. Biol., 3 (1993) 3143). Furthermore, a triple helix can also be formed by replacing the phosphorus backbone with a polyamide backbone such as PNA (peptide nucleic acid) (Nielsen et al., Science, 254 (1991), 1497; Kim et al., J. Am. Chem. Soc., 115 (1993) 6477-6481).

  The third strand thymine may be substituted with 5-bromouracil, thereby increasing the affinity of the oligonucleotide for DNA (Povsic et al., J. Am. Chem. Soc., 111 (1989) 3059). . The third strand may further contain an unnatural base, such as 7-deaza-2′-deoxyxanthosine (Milligan et al., Nucleic Acids Res., 21 (1993) 327), 1- (2-deoxy-α -D-ribofuranosyl) -3-methyl-5-amino-1H-pyrazolo [4,3-d] pyrimidin-7-one (Koh et al., J. Am. Chem. Soc., 114 (1992) 1470), 8 -Oxoadenine, 2-aminopurine, 2'-O-methyl pseudoisocytidine or other optional modifications known to those skilled in the art (Sun et al., Curr. Opinion in Struct. Biol., 3 (1993) 345). It is done.

  The purpose of other modifications of the third strand is in particular to improve the interaction and / or affinity of the third strand with the specific sequence. In particular, a completely advantageous modification of the invention is the methylation of the 5-position cytosine of the oligonucleotide. The thus methylated oligonucleotide has excellent properties of forming a stable triple helix with a specific sequence in a near neutral pH range (≧ 5; Xdo et al., Nucleic Acids Research 19 (1991) 5625). It is therefore possible to operate at a higher pH than prior art oligonucleotides, ie at a very low risk of plasmid DNA degradation.

  The length can be individually adjusted by those skilled in the art depending on the desired selectivity and stability of the interaction.

  The third DNA strand of the present invention can be synthesized by any known method. In particular, it can be produced by a nucleic acid synthesizer. Other optional methods known to those skilled in the art can of course be used.

  These third DNA strands or these oligonucleotides contain a specific internal region N containing a mixed (pyrimidine-purine) internal region N of 7 nucleotides or less sandwiched between two homopurine regions R and R ′ on both sides as described above. A triple helix can be formed with a double-stranded DNA sequence. Such a homopurine region can contain, for example, a GAA type motif.

  For example, the following sequence:

A target double-stranded DNA sequence corresponding to the sequence 5′-AA GAA GCA TGC AGA GAA GAA-3 ′ (SEQ ID NO: 1) capable of forming a triple helix with an oligonucleotide comprising a sequence selected from Can do.

Triple helix formation can be obtained in the presence of Mg ²⁺ ions, which can optionally help stabilize this structure (Vasquez et al., Biochemistry 34 (1995) 7243; Beal et al., Science 251 (1991) 1360).

According to one preferred embodiment, the target DNA sequence of the present invention can be naturally present in double-stranded DNA, in which case, for example, in the sequence of a gene of interest such as a therapeutic or experimental gene or a marker gene. It is particularly advantageous to use oligonucleotides capable of forming a triple helix with such sequences present. In this regard, the present inventors analyzed the nucleotide sequences of various target genes and tested the stability of triple-helix interactions with (CTT) n-type oligonucleotides. It could be shown that a stable triple helix was formed despite the presence of non-canonical triads such as ^* CG, T ^* GC, C ^* AT, C ^* TA and T ^* TA.

  The sequence naturally present in the double-stranded DNA is the sequence 5′-AA GAA GCA TGC AGA GAA GAA-3 ′ (ID1) present in the sequence of the human FGF1 gene (Jaye et al., Science 233 (1986) 541). ) (SEQ ID NO: 1), sequence 5′-GAA GAA GCA of a human gene encoding factor IX involved in coagulation (Kurachi et al., Proc. Natl. Acad. Sci. USA 79 (1982) 6461) CGA GAA G-3 ′ (SEQ ID NO: 6), sequence 5′-AAA GAA AGC AGG GAA G-3 ′ (sequence) of secreted alkaline phosphatase SeAP gene (Millan et al., J. Biol. Chem., 261 (1986) 3112) No. 7) and 5'-GAA GAG GAA GAA G- '(SEQ ID NO: 8), sequence 5'-AAG GAG AAG AAG AA-3' (SEQ ID NO: 9) of human α-fetoprotein hαFP gene (Gibbs et al., Biochemistry 26 (1987) 1332), human GAX suppressing restenosis The sequence 5′-AA GAT GAG GAA GAA G-3 ′ (SEQ ID NO: 10) of the gene (Gorski et al., Mol. Cell. Biol., 13 (1993) 3722), and the human VEGFB-167 gene (Olfsson et al., J. Biol. Biol.Chem., 271 (1986) 19310), sequence 5′-GGC AAC GGA GGA A-3 ′ (SEQ ID NO: 13). When a triple helix is formed with a sequence present in a therapeutic or experimental gene, the target sequence is naturally present in the double-stranded DNA molecule, and an artificial specific sequence is incorporated into the target double-stranded DNA molecule or a plasmid having this gene. Is particularly advantageous since it does not need to be modified. Alternatively, the target sequence may be artificially introduced into double-stranded DNA.

  The second aspect of the present invention is a method for purifying double-stranded DNA. According to this method, a solution containing DNA as a mixture with other components is brought into contact with the third DNA strand as described above. In this case, the third DNA strand is preferably an oligonucleotide capable of forming a triple helix by hybridization with a specific sequence present in the double-stranded DNA as described above. It is preferable to bring double-stranded DNA into contact with the oligonucleotide immobilized on the carrier in a solution. More preferably, the oligonucleotide is stably or covalently bound to the carrier. Therefore, the step of contacting the solution containing the double-stranded DNA is performed in such a manner that the DNA mixed with the other components on the oligonucleotide-bound carrier is obtained in order to efficiently and rapidly obtain the double-stranded DNA desired to be purified. It is advantageous to pass the solution.

  Such carriers are well known to those skilled in the art and comprise, for example, beads or microparticles (eg latex particles) or any other suspension carrier. Oligonucleotides may be grafted onto polymer-type molecules of natural or synthetic origin. The polymer to which the oligonucleotide is bound preferably has a property that it can be easily separated from the solution after forming a triple helix with double-stranded DNA. Natural polymers can include proteins, lipids, sugars and polyols. Synthetic polymers include polyacrylamide, polyethylene glycol, styrene derivatives and thermosensitive polymers (eg, poly (N-isopropylacrylamide) type compounds that are soluble at low temperatures but become insoluble above the phase transition temperature) (T. Mori). Et al., Biotechnology and Bioengineering, 72 (2001) 261).

  The purification method of the present invention is not only for i) DNA molecules that do not contain a homopurine sequence long enough to form a stable triple helix with a homopyrimidine oligonucleotide, but also ii) several pyrimidine bases were interrupted in the homopurine sequence. This is particularly advantageous because DNA molecules can also be purified. In addition to being able to purify more types of DNA molecules, this method is rapid and results in particularly high yields and purity.

  In addition, DNA molecules can be purified from complex mixtures containing other nucleic acids, proteins, endotoxins (eg, lipopolysaccharides), nucleases, etc., and pharmaceutical quality purified DNA is obtained.

  Oligonucleotides are generally functionalized for covalent attachment to a support. That is, the 5 'or 3' position may be modified with a thiol, amine or carboxyl end group. In particular, the addition of a thiol, amine or carboxyl group makes it possible to bind the oligonucleotide to a carrier having for example a disulfide, maleimide, amine, carboxyl, ester, epoxide, cyanogen bromide or aldehyde group. These bonds are formed by setting the disulfide, thioester, ester, amide or amine bond between the oligonucleotide and the carrier. Other optional methods known to those skilled in the art, such as bifunctional coupling agents, can also be used.

Furthermore, in order to improve hybridization with the bound oligonucleotide, it is advantageous if the oligonucleotide comprises an “arm” and a “spacer” base sequence. In fact, when the arm is used, the oligonucleotide can be bound at a selected distance from the carrier, and the interaction condition with DNA can be improved. The arm is advantageously composed of a linear carbon chain containing 1-18, preferably 6-12 CH ₂ type groups, and an amine attached to the column. The arm binds to a “spacer” phosphate composed of oligonucleotides or bases not involved in hybridization. Thus, a “spacer” can include a purine base. For example, a “spacer” can include a GAGG sequence.

The oligonucleotide bound to the purification support can have, for example, the sequence 5′-GAGG CTT CTT CTT CTT CTT CTT CTT CTT-3 ′ (GAGG (CTT) ₇ ; SEQ ID NO: 11), where the GAGG base is involved in the triple helix structure Although not, a space can be formed between the oligonucleotide and the coupling arm.

  Various carriers can be used to practice the present invention. It may be a functionalized chromatographic support, a bulk or prepacked column, a functionalized plastic surface or a magnetic or non-magnetic functionalized latex bead. A gel permeation chromatography carrier is preferred. For example, available chromatographic supports include polymers such as agarose, acrylamide or dextran and derivatives thereof (eg, Sephadex®, Sepharose®, Superose®, etc.), poly (styrenedivinylbenzene), or Grafted or non-grafted silica. The chromatography column uses a chromatographic carrier of a density suitable for this particular embodiment and may be diffusion or perfusion, so-called “fluidized bed” or “foaming”.

  The method of the present invention can be used to purify any kind of double-stranded DNA. For example, circular circles such as minicircles (Darquet et al., Gene Therapy 6 (1999) 209), linear fragments, or plasmids that generally have one or more therapeutic or experimental genes. This plasmid can also contain, for example, a conditional origin of replication (eg, the pCOR plasmid described in Sobrier et al., Gene Therapy 6 (1999) 1482), a marker gene, and the like. The method of the present invention can be applied directly to cell lysates. In this embodiment, the plasmid amplified by cell culture after transformation is directly purified after cell lysis. The method of the present invention can also be applied to a transparent lysate (ie, a supernatant obtained after neutralization of cell lysate and centrifugation). Of course, it can also be applied to pre-purified solutions using known methods. This method can also purify linear or circular DNA having a target sequence from a mixture containing DNA of various sequences. The method of the present invention can also be used for purification of double-stranded RNA.

  The cell lysate may be a prokaryotic cell lysate or a eukaryotic cell lysate. Examples of prokaryotic cells include Escherichia coli, Bacillus subtilis, Salmonella typhimurium, Staphylococcus aureus, and Streptomyces bacteria. Examples of eukaryotic cells include animal cells, yeasts, fungi, and the like, and specific examples include Kluyveromyces or Saccharomyces yeasts, or cells such as COS, CHO, C127, NIH3T3, MRC5, and 293.

  The method of the present invention is particularly advantageous because very high purity plasmid DNA can be obtained quickly and easily. In particular, as illustrated in the Examples, the method can efficiently separate plasmid DNA from contaminating components such as fragmented chromosomal DNA, RNA, endotoxin, protein or nuclease.

  The methods of the present invention are also useful for purifying and purifying DNA molecules, particularly therapeutic genes such as the FGF1 gene, which are produced and purified on a commercial scale and require a purity suitable for pharmaceutical use.

  According to a third aspect, the present invention provides a method for detecting, quantifying and selecting a double-stranded DNA molecule comprising at least one target sequence as described above, a) a solution suspected of containing said molecule is a third DNA strand, For example, it relates to a method comprising contacting a labeled oligonucleotide to form a stable triple helix, and b) detecting an optionally formed complex between a double stranded DNA and a third DNA strand.

  The method is particularly useful in the scope of genomic analysis, for example by allowing the detection of specific DNA sequences in the genome or the selection of specific sequences.

  According to this aspect of the invention, the third DNA strand or oligonucleotide can be labeled with a marker detectable by spectroscopy, photochemistry, biochemistry, immunochemistry or chemical means.

For example, such markers may be composed of a radioisotope ^{(32 P, 33 P, 3} H, 35 S) or a fluorescent molecule (5-bromodeoxyuridine, fluorescein, acetylaminofluorene, digoxigenin).

  The marker is preferably performed by adding a marker molecule to the polynucleotide by primer extension or addition of 5 'or 3' ends.

  Examples of non-radioactive markers are in particular FR FR1010975 or Urdea et al. (1988, Nucleic Acids Research, 11: 4937-4957) and Sanchez-pescador et al. (1988; J. Clin. Microbiol., 26 (10): 1934. -1938).

  According to this particular aspect of the invention, a third DNA strand or oligonucleotide may also be immobilized on the carrier as described above.

  The fourth aspect of the present invention relates to a pack or kit for purifying and / or detecting the presence of the double stranded DNA of the present invention in a complex mixture, comprising one or more of the above oligonucleotides. The oligonucleotide may be immobilized on a carrier and / or a detectable marker may be added.

  According to this aspect of the invention, the detection kit comprises a plurality of oligonucleotides of the invention that can be used to detect a target target double-stranded DNA sequence.

  That is, oligonucleotides immobilized on a carrier can be aligned on a matrix such as a “DNA chip”. Such alignment matrices are described in particular in US Pat. No. 5,143,854 and PCT applications WO 90/15070 and 92/10092.

  A carrier matrix in which oligonucleotides are immobilized at a high density is described, for example, in US Pat. No. 5,412,087 and PCT application WO95 / 11995.

  EXAMPLES Hereinafter, although an Example demonstrates this invention in detail, the following Examples are only illustrations and do not limit invention.

General cloning and molecular biology techniques Conventional molecular biology techniques such as restriction enzyme digestion, gel electrophoresis, DNA fragment ligation, transformation in E. coli, nucleic acid precipitation, sequencing, etc. are described in the literature (Maniatis et al., ( 1989) Molecular cloning: a laboratory manual, 2nd edition, Cold Spring Harbor Laboratory, Cold Spring Harbor Laboratory Press, New York, New York; Auber et al., (1987) Crow harbor Laboratory Press, New York; Restriction enzymes were obtained from New-England Biolabs, Beverly, MA (Biolabs).

  Oligonucleotides are synthesized using Applied Biosystem's automated DNA synthesizer 394 according to the manufacturer's instructions using phosphoramidite chemistry with the α-position protected with a cyanoethyl group (Sinha et al., Nucleic Acids Research, 12 (1984) 4539; Giles (1985)).

  The oligonucleotides used to synthesize the affinity gel are used without modification to the Amersham Pharmacia Biotech (Uppsala, Sweden) or Eurogentec (Seraing, Belgium) products.

  Strains capable of replicating pCOR plasmids and the growth and selection conditions for these plasmids have been described in the literature (Sobrier et al., Gene Therapy 6 (1999) 1482).

Construction of plasmid 1.1 Plasmid pXL3179 (pCOR-FGF1)
The plasmid pXL3179 shown in FIG. 1 is a vector derived from the plasmid pXL2774 (WO97 / 10343; Sobrier et al., Gene Therapy 6 (1999) 1482), and contains the early region of human cytomegalovirus (hCMV IE E / P) and the SV40 virus. Fusion of human fibroblast interferon signal peptide and FGF1 (fibroblast growth factor 1) cDNA under the control of a promoter derived from the polyadenylation signal of late region (SV40 late poly A; Genbank SC4CG) (sp-FGF1, Joanneau) Et al., A gene encoding PNAS 88 (1991) 2893) was introduced.

1.2 Plasmid pXL3296 (pCOR)
Plasmid pXL3296 was derived from plasmid pXL3179, and the sequence of the sp-FGF1 gene was replaced with the multiple cloning buoy of plasmid pUC28 (Benes et al., Gene 130 (1993) 151). The plasmid pXL3296 is shown in FIG.

1.3 Plasmid pXL3426 (pCOR-ID1)
Plasmid pXL3426 was derived from plasmid pXL3296 and inserted the sequence 5′-GATCCAAGAAGCATGCAGAAGAATTC-3 ′ between the BglII and XhoI sites. The plasmid pXL3426 is shown in FIG.

Construction of other plasmids with target sequences Plasmid pXL3675 is derived from plasmid pXL3296, the sequence 5'-GAAGAAGGGAAAAGAATCTG-3 'is inserted between the HpaI and XbaI sites, plasmid pXL3676 is also derived from plasmid pXL3296, and between HpaI and XbaI sites. The sequence 5′-GAAGAAAGGAGAGAGAAGTCTG-3 ′ was inserted into the plasmid pXL3713, and the DNA sequence 5′-GAAGAAGTTTAAGAAGATCTG-3 ′ was inserted between the HpaI and XbaI sites of pXL3296. The thus constructed plasmid was purified with a CsCl chloride gradient, and the sequence of the insert was confirmed by sequencing. These preparations were used in the following examples.

Identification of the 20mer internal sequence of the FGF1 gene forming a stable triple helix Various plasmids as described in the following examples were subjected to triple helix interaction affinity chromatography under standardized conditions. The affinity carrier was synthesized using the chromatographic carrier Sephacryl (registered trademark) S-1000 SF (Amersham Pharmacia Biotech) as follows. In the first step, Sephacryl® S-1000 gel dispersed in 0.2 M sodium acetate buffer was activated with sodium metaperiodate (3 mM, 20 ° C., 1 hour), followed by protein in the second step. The 5′-NH ₂ terminal portion was removed by reductive amination in the presence of ascorbic acid (5 mM) following a procedure similar to that described for the coupling of (Hornsey et al., J. Immunol. Methods 93 (1986) 83). To the aldehyde group of the activation matrix. That is, in all experiments reported to the present invention, oligonucleotides were conjugated according to this general procedure. All oligonucleotides _NH 2 at the 5 'end of the oligonucleotide - _{(CH 2) 6} - has the function of arm.

  All experiments to verify the formation of triple helix structure of oligonucleotide and double-stranded DNA and to measure its stability were conducted under the following conditions. An HR5 / 5 column (Amersham Pharmacia Biotech) in which 300 μg of purified plasmid was dissolved in 6 ml of 50 mM sodium acetate buffer (pH 4.5) to which 2 mM NaCl was added in each experiment and 1 ml of an affinity gel functionalized with the oligonucleotide of the present invention was packed. Was injected at a flow rate of 30 cm / h. After washing the column with 5 ml of the same buffer, the plasmid was eluted with 3 ml of 100 mM Tris / HCl column buffer (pH 9.0) supplemented with 0.5 mM EDTA, and the amount of plasmid eluted with pH 9.0 buffer was i). Measurement of 260 nm absorbance of the solution and ii) quantification by anion exchange chromatography on a Millipore GenPak-Fax column (Marquet et al., BioPharm, 8 (1995) 26).

When a column functionalized with the oligonucleotide 5′-NH _2- (CH ₂ ) ₆ -TT (CTT) ₆ -3 ′ (SEQ ID NO: 2) is used, as is clear from the purification results shown in Table 1 below, human In contrast to the control plasmid (pXL3296), which does not contain the sequence of the FGF1 gene and is not fixed to the corresponding column, a stable triple in the plasmid containing the complete FGF1 gene (pXL3179) or the internal ID1 sequence of the human FGF1 gene (pXL3426). A spiral is formed.

The sequence of plasmid pXL3426 was identified by subcloning various fragments of the FGF1 gene with reduced size. Therefore, 5′-AA GAA GCA TGC AGA GAA GAA-3 ′ (SEQ ID NO: 1), which is the internal sequence ID1 of the FGF1 gene, forms a stable triple helical structure with the oligonucleotide of the sequence of SEQ ID NO: 2. The resulting triple helical structure is 7 triads (6 of which are non-canonical, more specifically 2 T ^* GC triads, 2 T ^* CG triads, 1 C ^* AT triad and 1 C ^* TA triad) T ^* AT and ⁺ C ^* GC canonical triads with a length of 6 units (R, 5 'side) and 7 units (R', 3 'side) separated by an internal region (T) It contains two regions of the pyrimidine-purine-pyrimidine (Py ^* PuPy) type that form.

Identification of bases required for the stability of the triple helix structure in the 20-mer internal sequence ID1 of the FGF1 gene Four oligonucleotides were prepared based on the internal sequence ID1. Two of them lack 7 or 13 nucleotides from the 5 ′ side of ID1, and the other two lack 7 or 14 nucleotides from the 3 ′ side. Multiplied by plasmid pXL3426 is chromatographed on oligonucleotide _{5'-TT (CTT) 6 -3} '( SEQ ID NO: 2) or oligonucleotide FRB36, FRB38, triple helix interactions column functionalized using FRB39 or FRB40. Subsequently, the stability of the triple helix structure formed with the various cleaved internal sequences ID1 was tested by measuring the amount of each plasmid retained on the column.

As it is clear from the results shown in Table 2, whole ID1 sequence of _{5'-TT (CTT) 6 -3} ' type pXL3426 (20 mer) in order to form the oligonucleotide and stable triple helical structure of is necessary. In particular, the two Py ^* PuPy parts and the central part at the 5 ′ and 3 ′ ends contribute very greatly and cooperatively to the stability of the final structure.

Effect of the number of canonical triads and non-canonical triads on the stability of the triple helix The sequence of oligonucleotide 5′-TT (CTT) ₆ -3 ′ (SEQ ID NO: 2) was modified and these various oligonucleotides (FRB15, FRB16 and The ability of FRB17) to form a stable triple helix with the internal sequence ID1 (5'-AA GAA GCA TGC AGA GAA GAA-3 '; SEQ ID NO: 1) of plasmid pXL3426 was tested.

As is clear from the results shown in Table 3 above, the oligonucleotide sequence was increased so as to increase the number of T ^* AT and ⁺ C ^* GC canonical triads in the central inner region N of the triple helix structure and simultaneously reduce the number of non-canonical triads. The stability of the triple helix structure can be increased by modification.

Internal sequence of FGF1 gene capable of forming a non-canonical triad influence oligonucleotide _{5'-TT (CTT) 6 -3} '( SEQ ID NO: 2) a stable triple helix on the stability of the triple helix structure ID1 ( The sequence of plasmid pXL3426 containing SEQ ID NO: 1) was modified, and two consecutive identical T ^* GC non-canonical triads and one 5'-side C ^* AT non-canonical triad were introduced into the central region N. In another experiment, five consecutive non-canonical triads were introduced (pXL3676): C ^* AT, T ^* GC, T ^* GC, C ^* AT and T ^* GC. Finally, plasmid pXL3426 was modified to introduce two consecutive T ^* TA type non-canonical triads and one 5'-side C ^* TA non-canonical triad (pX3713).

As is clear from the purification results of various plasmids shown in Table 4 above, a stable triple helix structure is formed when the non-canonical center region forms T ^* CG, T ^* GC, C ^* AT and C ^* TA type triads. It is formed. As is apparent from the binding yields of plasmids pXL3675 and pXL3676 to the affinity gel, a stable triple helix is formed even when two non-canonical T ^* GC triads are introduced in succession. That is, the front and rear triads may be canonical or non-canonical. On the other hand, when two consecutive T ^* TA type triads and one C ^* TA triad are successively introduced, the triple helix structure is completely destabilized.

Construction of plasmids containing cassettes encoding SeAP, hαFP, FIX and GAX genes The genes used in these experiments to demonstrate the activity of the compositions of the invention include, for example, the human gene encoding FIX factor (Kurachi et al., Proc Natl.Acad.Sci.U.S.A.79 (1982) 6461), human gene encoding secreted alkaline phosphatase SeAP (Millan et al., J. Biol. Chem., 261 (1986) 3112), α-fetoprotein The human gene hαFP (Gibbs et al., Biochemistry 26 (1987) 1332), and the human gene (Gorski et al., Mol. Cell. Biol., 13 (1993) 3722) encoding GAX. These genes were PCR amplified using a plasmid or cDNA library (Clontech) and then cloned downstream of the eukaryotic CMV E / P promoter of the pCOR plasmid derived from pXL3296, upstream of the SV40 late poly A signal sequence. The gene encoding secreted alkaline phosphatase (SeAP) was introduced into the pCOR plasmid derived from pXL3296 to produce plasmid pXL3402 (FIG. 4). The gene encoding α-fetoprotein (hαFP) was introduced into the pCOR plasmid derived from pXL3296 to prepare plasmid pXL3678 (FIG. 5). The gene encoding GAX was introduced into the pCOR plasmid derived from pXL3296 to prepare plasmid pXL3207 (FIG. 6). The gene encoding FIX factor was introduced into the pCOR plasmid derived from pXL3296 to prepare plasmid pXL3388 (FIG. 7).

5 for forming a variety of target gene and stable triple helical structure - by measuring the capacity obtained with a plasmid having various genes use _'(CTT) 7 -3' type oligonucleotide, the oligonucleotide 5'-TT (CTT) 6 _-3 'were investigated interactions (SEQ ID NO: 2) with functionalized triplex interactions gel and various sequences. The genes tested were i) human gene encoding factor IX, ii) secreted alkaline phosphatase SeAP gene, iii) α-fetoprotein (αFP) human gene and iv) human GAX gene.

As is clear from the results shown in Table 5 above, when a (CTT) _n- type oligonucleotide is used, the target gene does not contain a 5 ′-(GAA) _n −3 ′ type target sequence complementary to the oligonucleotide. Even in this case, it is possible to form a stable triple helix structure with the target sequence. An isolated T ^* TA triad (GAX gene) without affecting the triple helix structure even if the base contained in the T ^* CG, C ^* CG, T ^* GC and C ^* AT type triads is present in the center of the target sequence. It is the same even if exists.

Use of _(CTT) 7 -3 column functionalized with 'type oligonucleotide -;' 5 for purifying plasmid (SEQ ID NO: 1 5'-AA GAA GCA TGC AGA GAA GAA-3) ' internal sequence ID1 _{5 '- (CTT) 7 -3} ' type using triple helix interaction affinity chromatography support functionalized with oligonucleotides, the 5 on the basis of examples _{8 '- (R) n -} (n) t - (R ′) The purifiability of plasmids with _m −3 ′ type sequences was tested.

Functionalization of plasmid pXL3179 (containing human FGF1 gene with sequence 5′-AA GAA GCA TGC AGA GAA GAA-3 ′) with oligonucleotide 5′-NH _2- (CH ₂ ) _6- (CTT) ₇ -3 ′ And chromatographed on a Sephacryl S-1000 interaction column. To this end, 50 mM acetic acid was added to 10 ml of an affinity column containing oligonucleotide 5′-NH ₂ — (CH ₂ ) _6- (CTT) ₇ -3 ′ covalently linked to Sephacryl S-100 SF as described in Example 3. 9.40 mg of plasmid pXL3179 in 60 ml of sodium, 2M NaCl buffer (pH 4.5) was injected at a flow rate of 30 cm / h. After washing the column with 5 volumes of the same buffer, the bound plasmid was eluted with 2 column volumes of 100 mM Tris / HCl, 0.5 mM EDTA buffer, measured for UV absorbance (260 nm) and on a GenPak-Fax column (Waters). Quantified by ion exchange chromatography. The E. coli genomic DNA content in the initial preparation and the purified fraction was determined by PCR as described in WO 96/18744. 7.94 mg of plasmid pXL3179 was detected in the elution fraction (elution yield 84%) and the contamination rate with E. coli genomic DNA was reduced from 7.8% to 0.2% by the affinity chromatography described. Similarly, the contamination rate with RNA decreased from 43% in the starting plasmid to 0.2% in the purified plasmid.

Various plasmid pXL3179 preparations oligonucleotide _{5'-NH 2 - (CH 2} ) 6 - (CTT) 7 in the other various experiments chromatographed Sephacryl S-1000 affinity column functionalized with 3 ', genomic DNA The content decreased from 0.2% to 0.007%, 0.7% to 0.01%, 7.1% to 0.2%, or 7.8% to 0.1%.

Use of oligonucleotide of type 5'-CCT TTT CCT CCT T-3 '(SEQ ID NO: 12) to form a stable triple helix structure with the therapeutic gene human VEGFB-167 human VEGFB-167 gene (Figure 8) ( By measuring the volume obtained with plasmid pXL3579 with Olofsson et al., J. Biol. Chem., 271 (1986) 19310), the internal sequence of the therapeutic gene (human VEGFB-167) and the oligonucleotide 5'-CCT TTT The interaction of the triple helix interaction carrier functionalized with CCT CCT T-3 ′ form (SEQ ID NO: 12) was tested. The plasmid pXL3579 shown in FIG. 8 contains the VEGFB-167 gene PCR-amplified from a human heart cDNA library (Clontech), and the eukaryotic promoter CMV E / P (−522 /) between the NsiI and XbaI sites of the multiple cloning site of pXL3296. Cloned downstream of +74) and upstream of the SV40 late poly A signal sequence.

As is apparent from the results shown in Table 6 above, a sequence of 5 ′-(R) _n- (N) _t- (R ′) _m −3 ′ type (for example, region 5 ′ of human VEGFB-167 gene in this case) -Using an oligonucleotide (eg oligonucleotide 5'-CCT TTT CCT CCT T-3 '(SEQ ID NO: 12)) targeting GGC AAC GGA GGA A-3' (SEQ ID NO: 13)) A stable triple helix structure can be formed. Furthermore, the human VEGFB-167 gene contains the homopurine sequence 5′-AAA AAA AAA AAG GA-3 ′ targeted by oligonucleotide 5′-TTT TTT TTC CT-3 ′ (Table 6), but oligonucleotide 5′-CCT. The interaction obtained with TTT CCT CCT T-3 ′ is much superior to the interaction obtained with homopyrimidine oligonucleotides. Similarly, homopurine internal sequence 5′-GGA GGA A-3 ′ is not long enough to form a stable triple helix with oligonucleotide 5′-CCT CCT T-3 ′.

Therefore, in this example, the 5 ′-(R) _n- (N) _t- (R ′) _m −3 ′ type is used even when the test double-stranded DNA further shows a homopurine structure having a considerable length. It clearly shows that the arrangement of can form a stable triple helix structure.

5'T CCT CTC CCT C-3 '(SEQ ID NO: 14) type oligo for separating non-deleted VEGFB-186 cDNA by forming a stable triple helix structure with the target sequence from the modified VEGFB-186 cDNA Use of Nucleotides During the fermentative production phase of the human VEGFB-186 gene, this gene was found to be a field of translocation and gene deletion, particularly at the level of exon 6A. Therefore, by sequential mutagenesis PCR, 510 (A / C), 513 (C / T), 516 (A / T), 519 (C / T) and 522 (C / T) with respect to the translation start point +1. A silent point mutation was introduced into the level. The amino acid sequence of the protein VEGFB-186 contained in amino acids 170 to 174 is not changed. On the other hand, the VEGFB-186 gene (VEGFB-186m) thus modified can form a stable triple helix structure with the oligonucleotide of the sequence 5′T CCT CTC CCT C-3 ′ (SEQ ID NO: 14). It has the cDNA sequence 5'-A GGA GCG GGA G-3 (SEQ ID NO: 15). This stable triple helix interaction is advantageously used to carry out the method of the invention and to isolate modified VEGFB-186 genes that are not translocated and deleted after fermentation production.

  In order to illustrate the purification method of modified VEGFB-186 gene by triple helix interaction, two plasmids pXL3601 and pXL3977 as shown in FIG. 9 were used. The gene VEGFB-186 was first PCR amplified from a human heart cDNA library (Clontech) and then the eukaryotic promoter CMV E / P (-522 /) between the NsiI and XbaI sites of the multiple cloning site of pXL3296 (Example 1.2). The plasmid pXL3601 was created by cloning downstream of +74) and upstream of the polyadenylation signal of the late region of the SV40 virus. This plasmid was modified by sequential mutagenesis PCR to generate plasmid pXL3977 in which the VEGFB-186 gene was modified at the level of exon 6A as described above.

  Triple helix interaction carrier functionalized with oligonucleotide 5′T CCT CTC CCT C-3 ′ (SEQ ID NO: 14) by measuring the volume obtained with plasmid pXL3777 carrying the modified human VEGFB-186 gene (FIG. 9) Were used to test the interaction of the target sequence 5′-A GGA GCG GGA G-3 (SEQ ID NO: 15) in the VEGFB-186m gene. Since the sequence of VEGFB-186m includes the sequence of VEGFB-167, plasmid pXL3579 containing the human VEGFB-167 gene as described in Example 10 is used as a negative control.

As is clear from the results shown in Table 7 above, 5 ′-(R) _n- (N) _t- (R ′) _m −3 ′ type sequence (in this case, region 5 ′ of the human modified VEGFB-186 gene) When an oligonucleotide (for example, oligonucleotide (SEQ ID NO: 14)) is used with -A GGA GCG GGA G-3 (SEQ ID NO: 15) as a target, a stable triple helix structure can be formed with the region of the target gene, Therefore, it can be purified effectively. Oligonucleotide 5′-TTT CCT CTC CCT C-3 ′ (SEQ ID NO: 16) can also be used for purification of the modified human VEGFB-186 gene.

Schematic diagram of plasmid pXL3179. Schematic diagram of plasmid pXL3296. Schematic diagram of plasmid pXL3426. Schematic diagram of plasmid pXL3402. Schematic diagram of plasmid pXL3678. Schematic diagram of plasmid pXL3207. Schematic diagram of plasmid pXL3388. Schematic diagram of plasmid pXL3579. Schematic representation of plasmids pXL3601 and pXL3977.

Sequence listing

Claims (34)

A method for the purification of double-stranded DNA molecules,
Contacting said double-stranded DNA molecule with a third DNA strand,
The double-stranded DNA molecule is circular DNA, and
General formula:
5 '-(R) _n- (N) _t- (R') _m- 3 '
(In the formula, R and R ′ represent a nucleotide sequence composed only of purine bases, n and m are integers of 8 or less, the sum of n + m is 6 or more, and N is a combination of purine bases and pyrimidine bases. And at least one target sequence (t is an integer of 7 or less), and the DNA sequence can interact with the third DNA strand to form a triple helix. Said method.   The method according to claim 1, wherein the interaction between the N region and the third DNA strand forms up to 6 non-canonical triads. The method of claim 2, wherein the non-canonical triad thus formed is selected from T ^* CG, T ^* GC, C ^* AT and C ^* TA triad.   4. The method according to any one of claims 1 to 3, characterized in that the R and R 'regions contain a total of at least two guanines (G).   5. A method according to any one of claims 1 to 4, wherein the R and R 'regions comprise at least two adenines. 6. A method according to any one of claims 1 to 5, characterized in that the R and R 'regions together with a third DNA strand form a canonical triad selected from T ^* AT and + C ^* GC.   The method according to any one of claims 1 to 6, wherein the third DNA strand is of a homopyrimidine type.   8. The method of claim 7, wherein the third DNA strand comprises a poly CTT sequence.   The method according to any one of claims 1 to 8, wherein the third DNA strand is an oligonucleotide.   The target sequence present on the double-stranded DNA molecule comprises the sequence of SEQ ID NO: 1, and the oligonucleotide is selected from the oligonucleotides of the sequences of SEQ ID NO: 2-5. The method according to any one of the above.   The method according to any one of claims 1 to 10, wherein the target sequence is a target sequence that is naturally present on a double-stranded DNA molecule or artificially introduced into a double-stranded DNA molecule. .   12. A method according to any one of claims 1 to 11, characterized in that the naturally occurring target sequence on the DNA molecule is present in the coding sequence of a therapeutic or experimental gene.   The method according to claim 12, wherein the target DNA sequence comprises all or part of the sequence of SEQ ID NO: 1 contained in the human FGF1 gene.   The method according to claim 12, wherein the target DNA sequence comprises the sequence of SEQ ID NO: 6 contained in a human gene encoding factor IX.   13. The method according to claim 12, wherein the target DNA sequence comprises the sequence of SEQ ID NO: 7 or 8 contained in the human gene for secreted alkaline phosphatase.   The method according to claim 12, wherein the target DNA sequence comprises the sequence of SEQ ID NO: 9 contained in the human gene for α-fetoprotein (αFP).   The method according to claim 12, wherein the target DNA sequence comprises the sequence of SEQ ID NO: 10 contained in the human GAX gene.   The method according to claim 12, wherein the target DNA sequence comprises the sequence of SEQ ID NO: 13 contained in the human VEGFB167 gene.   The method according to any one of claims 1 to 18, wherein the oligonucleotide is stably covalently or non-covalently bound to the carrier.   20. A method according to any one of claims 1 to 19, characterized in that the oligonucleotide is grafted to a natural or synthetic polymer.   20. A method according to claim 19, characterized in that the carrier is selected from functionalized latex beads, plastic surfaces and chromatographic carriers.   The method according to claim 21, wherein the chromatographic support is a gel permeation support.   The method according to any one of claims 1 to 22, wherein the solution containing the double-stranded DNA molecule is a cell lysate.   The method according to any one of claims 1 to 23, wherein the cell lysate is a transparent lysate.   25. A method according to any one of claims 1 to 24, wherein the double-stranded DNA molecule is pre-purified.   26. A method according to any one of claims 19 to 25, characterized in that the oligonucleotide has the sequence GAGGCTTCTTTCTCTCTTCTTCTCTT (SEQ ID NO: 11).   27. A method according to any one of claims 19 to 26, wherein the oligonucleotide is attached to the support via a disulfide, thioether, ester, amide or amine linkage. The oligonucleotide is bound to the carrier via an arm composed of a carbon chain (CH ₂ ) _n (where n is an integer of 1 to 18), and the arm is bound to the oligonucleotide via phosphate. 28. A method according to any one of claims 19 to 27, wherein the method is bound and bound to the support via an amine bond.   29. The oligonucleotide is at least one chemically modified so that it is resistant to or protected from nucleases or has increased its affinity for a specific sequence. The method as described in any one of.   The method according to any one of claims 1 to 29, wherein the oligonucleotide comprises a sequence obtained by methylating at least one cytosine at the 5'-position.   31. The method according to any one of claims 1 to 30, comprising at least one step of bringing a solution containing the double-stranded DNA as a mixture with other components into contact with a carrier covalently bound to the oligonucleotide. Method.   32. The method according to any one of claims 1 to 31, wherein a solution containing the double-stranded DNA as a mixture with other components is passed over a chromatographic carrier covalently bound to the oligonucleotide.   The method according to any one of claims 1 to 32, wherein the third DNA strand or the oligonucleotide is labeled. A method for detecting double-stranded DNA, comprising:
Contacting a solution suspected of containing said double-stranded DNA molecule with a labeled third DNA strand capable of forming a triple helix by hybridization with a target sequence of double-stranded DNA,
The target sequence has the general formula:
5 '-(R) _n- (N) _t- (R') _m- 3 '
(In the formula, R and R ′ represent a nucleotide sequence composed only of purine bases, n and m are integers of 8 or less, the sum of n + m is 6 or more, and N is a combination of purine bases and pyrimidine bases. The method according to claim 2, wherein t is an integer of 7 or less, and the double-stranded DNA is circular DNA.