EP0661979A4 - Cross-linking oligonucleotides for enzyme-mediated triple strand formation. - Google Patents

Abstract

The invention disclosed is directed to cross-linking between specific sites on adjoining oligonucleotides or oligodeoxynucleotides wherein the nucleoside monomers used to effect the cross-linking are (alkylating substituent substituted) pyrazolol[3,4-d]pyrimidine ribosides or 2'-deoxyribosides. This cross-linking is assisted by the presence in mammalian host of a recombination enzyme such as RecA. The cross-linking of either double or triple stranded nucleic acids is expected to have utility in the inhibiting the expression of the targeted nucleic acids in vivo and as a diagnostic tool as well.

Description

03736

1

CROSS-LINKING OLIGONUCLEOTIDES FOR ENZYME-MEDIATED TRIPLE STRAND FORMATION

This application is a continuation-in-part of application Serial No. 07/353,857, filed on May 18, 1989, which in turn is a continuation-in-part of application Serial No. 07/250,474, filed on September 28, 1988 (now abandoned) .

BACKGROUND OF THE INVENTION

This invention relates to nucleoside cross¬ linking agents and to the use of these compounds in the preparation of oligonucleotides. It also relates to derivatives of pyrazolo[3,4-d]pyri_nidine which are useful as nucleic acid bases for the preparation of oligonucleo¬ tides.

Oligonucleotides are useful as diagnostic probes for the detection of "target" DNA or RNA sequences. In the past, such probes were made up of sequences of nucleic acid containing purine, pyrimidine or 7-deazapurine nucleotide bases (U.S. Patent 4,711,955; Robins et al., J. Can. J. Chem. , .60:554 (1982); Robins et a-I.'.-'J. Org. ^'cHe . ,' 48:1854 (1983)). The method for attaching chemical moieties to these bases has been via an acetoxy-mercuration reaction, which introduces covalently bound mercury atoms into the 5-position of the pyrimidine ring, the C-8 position of the purine ring or the C-7 position of a 7-deazapurine ring (Dale et al., Proc. Natl. Acad. Sci. USA_f 7_0:2238 (1973); Dale et al. , Biochemistry. .14:2447 (1975)), or by the reaction of organomercurial compounds with olefinic compounds in the presence of palladium catalysts (Ruth et al., J. Org. Chem.. 4.:2870 (1978); Bergstrom et al., J. Am. Chem. Soc.. 100:8106 (1978); Bigge et al., J. Am. Chem. Soc.. 02..-2033 (1980)). 03736

The sugar component of oligonucleotide probes has been, until the present, composed of nucleic acid containing ribose or deoxyribose or, in one case, natural _/3-arabinose (patent publication EP 227,459). A novel class of nucleotide base, the 3,4- disubstituted and 3,4,6-trisubstituted pyrazolo[3,4-d]- pyrimidines, has now been found which offers several advantages over the prior art. The de novo chemical synthesis of the pyrazolopyrimidine and the resulting nucleotide allows for the incorporation of a wide range of functional groups in a variety of different positions on the nucleotide base and for the use of different sugar moieties. Also, adenine, guanine and hypoxanthine analogs are obtained from a single nucleoside precursor. Additionally, the synthesis does not require the use of toxic heavy metals or expensive catalysts. Similar pyrazolo[3,4-d]pyrimidines are known (Kobayashi, Chem. Pharm. Bull. , 2.1:941 (1973)); however, the substituents on the group are different from those of the present invention and their only use is as xanthine oxidase inhibitors.

The concept of crosslinkable nucleotide probes for use in therapeutic and diagnostic applications is related--t-o-the-pioneering work of B.R. Baker, "Design of Active-Site-Directed Irreversible Enzyme Inhibitors," Wiley, New York, (1967) , who used what was termed "active-εite-directed enzyme inhibitors" in chemothera- peutic applications.

In recent years, the concept of incorporating a crosslink in an oligonucleotide has been sporadically discussed in efforts to develop superior sequence probes. Knorre and Vlassov, Prog. Nucl. Acid Res. Mol. Biol.. 3_2:291 (1985), have discussed sequence-directed cross-, linking ("complementary addressed modification") using an W-(2-chloroethyl)-_v'-methylaniline group attached to either the 3'- or 5'-terminus of oligonucleotides. Summerton and Bartlett, J. Mol. Biol.. 112:145 (1978) 3/03736

have shown that an 8-atom chain, attached to a cytosine residue at its C-4 position and terminating in the highly reactive bromomethyl ketone group, can crosslink to the N-l of guanosine. 5 Webb and Matteucci, Nucleic Acids Res.. 14:7661 (1986) , have prepared oligonucleotides containing a 5- methyl-.-v'.iv'-ethanocytosine base which is capable of slow crosslinking with a complementary strand. In a concept¬ ually related alkylation via a linker arm within a DNA 0 hybrid, Iverson and Dervan, Proc. Natl. Acad. Sci. USA. .85.-4615 (1988), have shown opposite strand methylation, triggered by BrCN activation of a methylthio ether, predominately on a guanine base located two pairs from the base bearing the linker.

15 Oligonucleotides may be used as chemothera- peutic agents to control the expression of gene sequences unique to an invading organism, such as a virus, a fungus, a parasite or a bacterium. In nature, some RNA- expression in bacteria is controlled by "antisense" RNA, ₀which exerts its effect by forming RNA:RNA hybrids with complementary target RNAs and modulating or inactivating their biological activity. A variety of recent studies using plasmid vectors for the introduction of antisense RNAs__int-o—eukaryotic cells have shown that they

-..effectively inhibit expression of mRNA targets in vivo (reviewed in Green, et al., Ann. Rev. Biochem. 55: 569- 597 (1986)). Additionally, a specific mRNA amongst a large number of mRNAs can be selectively inactivated for protein synthesis by hybridization with a complementary

30DNA restriction fragment, which binds to the mRNA and prevents its translation into protein on riboso es (Paterson, et al., Proc. Natl. Acad. Sci 74: 4370-4374 (1977); Hastie et al., Proc. Natl. Acad. Sci. 75: 1217- 1221 (1978)).

In the first demonstration of the concept of 5 using sequence-specific, antisense oligonucleotides as regulators of gene expression and as chemotherapeutic 3/03736

agents, Za ecnik and Stephenson, Proc. Natl. Acad. Sci. USA. 7J5:280 (1978), showed that a small antisense oligo- deoxynucleotide probe can inhibit replication of Rous Sarcoma Virus in cell culture, and that RSV viral RNA translation is inhibited under these conditions

(Stephenson et al. , Proc. Natl. Acad. Sci. USA 75:285 (1978)). Zamecnik et al. , Proc. Natl. Acad. Sci. USA. 12:4143 (1986), have also shown that oligonucleotides complementary to portions of the HIV genome are capable of inhibiting protein expression and virus replication in cell culture. Inhibition of up to 95% was obtained with oligonucleotide concentrations of about 70 μM. Importantly, they showed with labeled phosphate studies that the oligonucleotides enter cells intact and are reasonably stable to metabolism.

Uncharged methylphosphonate oligodeoxynucleo- tides with a sequence complementary to the initiation codon regions of rabbit globin mRNA inhibited the translation of the mRNA in both cell-free systems and in rabbit reticulocytes (Blake et al., Biochemistry 24:6139 (1985) ) . Another uncharged methylphosphonate oligonucleotide analog, an 8-nucleotide sequence complementary to the acceptor splice junction of a mRNA of—_ϊerpe_j— simpiex virus, Type l, can inhibit virus replication in intact Vero cells. However, fairly high concentrations (>25 mM) of this nonionic probe were required for this inhibition.

Although the impact of crosslinking oligonuc¬ leotides in the che otherapeutic field might be of great significance, their impact in DNA probe-based diagnostics is of equally great importance. The ability to covalent¬ ly crosslink probe-target hybrids has the potential to dramatically improve background and sensitivity limits in diagnostic assays as well as permit novel assay formats. 3736

Specific innovations (discussed previously by Gamper et al. , Nucl. Acids Res.. 14, 9943 (1988)) include:

(a) incorporation of a denaturing wash step to remove background; (b) use of the crosslink as an additional tier of discrimination;

(c) crosslinking occurring at or near the melt¬ ing temperature of the expected hybrid to insure exquisite specificity and to sub- stantially reduce secondary structure in the target, thereby increasing the effici¬ ency of hybrid formation; and

(d) novel solution hybridization formats as exemplified by the Reverse Southern protocol.

The concept of crosslinking, however, suggests potential problems that must be circumvented. For instance, the oligonucleotide containing a crosslinking. arm might covalently bond to the target sequence so readily that mismatching of sequences will occur, pos¬ sibly resulting in host toxicity. On the other hand, the crosslinking reaction must be fast enough to occur before correctly matched sequences can dissociate.

T is issue can be addressed by constructing an oligonucleotide that, upon hybridization, results in a duplex whose T_B is just above the physiological temperature of 37°C. Thus, even a single mismatched base will prevent hybrid formation and therefore crosslinkage. The optimization can be accomplished by judicious choice of oligonucleotide length and base composition, as well as position of the modified base within the probe. The probe must be long enough, however, to insure specific targeting of a unique site.

European Patent Application No. 86309090.8 describes the formation of chemically modified DNA probes such as 5-substituted uridinyl in which the substituent does not crosslink but contains a chemical or physical reporter group. WO8707611 describes a process for label¬ ing DNA fragments such as by chemically modifying the fragment followed by reaction with a fluorescent dye. Yabusaki et al. in U.S. Patent No. 4,599,303 disclose a scheme for covalently crosslinking oligonucleotides such as by formation of furocoumarin monoadducts of thymidine which are made to covalently bond to other nucleotides upon photoexcitation. EP 0259186 describes adducts of macromolecules and biotin which can be used as cross- linking nucleic acid hybridization probes. WO8503075 describes crosslinking disulfonic esters useful as nucleic acid fragmentation agents. DE3310337 describes the covalent crosslinking of single-stranded polynucleo- tides to such macromolecules as proteins with the resulting complex subsequently used as a marker in hybridization experiments in the search for complementary sequences in foreign polynucleotides.

A need exists for probe oligonucleotides, con¬ sisting of sufficient base sequences to identify target sequences with high specificity, that are provided with one or more crosslinking arms which readily form covalent bonds with specific complementary bases. Such oligonuc¬ leotides may be used as highly selective probes in hybr-idi_r_-ti-oπ-'-__-Ξays. The oligonucleotides may also be 5used as antisensing agents of RNAs, e.g., in chemo¬ therapy.

SUMMARY OF THE INVENTION This invention is directed to crosslinking agents which accomplish crosslinking between specific sites on adjoining strands of oligonucleotides. The crosslinking reaction observed is of excellent specificity. The invention is also directed to oligo¬ nucleotides comprising at least one of these crosslinking agents and to the use of the resulting novel oligonucleo¬ tides for diagnostic and therapeutic purposes. /03736

More particularly, the crosslinking agents of this invention are derivatives of nucleotide bases with a crosslinking arm and are of the following formula (!'):

R_ - B - (CH₂)_q- (Y)_r - (CH₂)_- A¹ (!•)

wherein.

R_x is hydrogen, or a sugar moiety or analog thereof optionally substituted at its 3' or its 5' oposition with a phosphorus derivative attached to the sugar moiety by an oxygen and including groups Q_{l f} Q₂ and Q₃, or with a reactive precursor thereof suitable for nucleotide bond formation; d is hydroxy, phosphate or diphosphate; 5 Q₂ is =0 or =S;

Q₃ is CH₂-R', S-R^», O-R', or N-R'R"; each of R¹ and R" is independently hydrogen or C_halky1;

B is a nucleic acid base or analog thereof that θis a component of an oligonucleotide;

Y is a functional linking group; each of m and q is independently 0 to 8, inclusive;

-r— ±s~&-or l; and 5 A' is a leaving group.

The invention also provides novel oligonucleo¬ tides comprising at least one of the above nucleotide base derivatives of formula I' .

Nucleotides of this invention and oligonucleo- tides into which the nucleotides have been incorporated may be used as probes. Since probe hybridization is reversible, albeit slow, it is desirable to ensure that each time a probe hybridizes with the correct target sequence, the probe is irreversibly attached to that sequence. The covalent crosslinking arm of the nucleotide bases of the present invention will permanent¬ ly modify the target strand, or cause depurination. As /03736

8 such, the oligonucleotides of this invention are useful in the identification, isolation, localization and/or detection of complementary nucleic acid sequences of interest in cell-free and cellular systems. Therefore, the invention further provides a method for identifying target nucleic acid sequences, which method comprises utilizing an oligonucleotide probe comprising at least one of a labeled nucleotide base of the present invention. The invention further describes methods or inactivating gene function involving combination of a crosslinkable anti-gene ODN and a recombination enzyme. Coating the ODN with a recombination enzyme facilitates the search for homology with in the target gene and subsequent triple strand formation. Crosslinking of resultant triple strand complexes inactivates gene function. A crosslinkable anti-gene nucleoprotein filament that includes (i) a nucleoside crosslinking agent covalently linked to an oligonucleotide (ODN) complementary to a target DNA sequence within a gene, and (ii) a recombination enzyme non-covalently associated with the ODN is also described.

This invention also provides novel substituted pyrazolo{-3-r4-d.^ρyrimidines which are useful as a nucleotide base in preparing nucleosides and nucleotides, rather than the natural purine or pyrimidine bases or the deazapurine analogs.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 depicts a modified deoxyuridine residue of an oligodeoxynucleotide crosslinked via an aceta idopropyl sidear to a deoxyguanosine residue located two sites away from the complementary base along the 5' direction.

Figure 2 depicts an autoradiogram of ³²P-labeled HPV target and crosslinked product following cleavage at /03736

the 3' side of the crosslinked guanosine. Lane 1: ³²P-labeled 15-mer size marker. Lane 2: 24 hour reaction at 20°C. Lane 3: 72 hour reaction at 20°C. Lane 4: 24 hour reaction at 30^βC. Lane 5: 72 hour reaction at 30^βC. Reactions were quenched with 2-aminoethanothiol and treated with piperidine solution to effect cleavage.

Figure 3 depicts an autoradiogram of ³²P-labeled HPV target and crosslinked product showing hybrid separation by denaturing polyacrylamide gel electrophor- esis. Lane 1: Control ³²P-labeled CMV target. Lane 2: 24 hour reaction at 20°C. Lane 3: 72 hour reaction at 20°C. Lane 4: 24 hour reaction at 30°C. Lane 5: 72 hour reaction at 30°C. Reaction solutions were treated with 2-aminoethanothiol, which quenches the iodoacetamido group.

DETAILED DESCRIPTION OF THE INVENTION A. Crosslinking oligonucleotides This invention provides novel substituted nucleotide bases with a crosslinking arm which are useful in preparing nucleosides and nucleotides and are useful as crosslinking agents. The substituted bases are of the fpllowing—form-ϊia (I*) :

Ri - B - (CH₂)_q- (Y)_r - (CH_z) _m - A' (I')

wherein,

R_α is hydrogen, or a sugar moiety or analog thereof optionally substituted at its 3' or its 5¹ position with a phosphorus derivative attached to the sugar moiety by an oxygen and including groups Q_x, Q₂ and Q_3/ or with a reactive precursor thereof suitable for nucleotide bond formation; Qi is hydroxy, phosphate or diphosphate;

Q₂ is =0 or -=S;

Q₃ is CH₂-R\ S-R', O-R¹, or N-R'R"; 3/03736

10 each of R' and R" is independently hydrogen or C__.₆al_cyl;

B is a nucleic acid base or analog thereof that is a component of an oligonucleotide; Y is a functional linking group; each of m and q is independently 0 to 8, inclusive; r is 0 or 1; and A' is a leaving group. In the practice of the present invention, the sugar moiety or analog thereof is selected from those useful as a component of a nucleotide. Such a moiety may be selected from, for example, ribose, deoxyribose, pentose, deoxypentose, hexose, deoxyhexose, glucose, arabinose, pentofuranose, xylose, lyxose, and cyclopentyl. The sugar moiety is preferably ribose, deoxyribose, arabinose or 2 '-O-methylribose and embraces either ano er, α or β .

The phosphorus derivative attached to the sugar moiety is conveniently selected from, for example, mono- phosphate, diphosphate, triphosphate, alkyl phosphate, alkanephosphonate, phosphorothioate, phoεphorodithioate, and the like.

—A—re-a-etive precursor suitable for internucleo- tide bond formation is one which is useful during chain extension in the synthesis of an oligonucleotide. Reactive groups particularly useful in the present invention are those containing phosphorus. Phosphorus- containing groups suitable for internucleotide bond formation are preferably alkyl phosphorchloridites, alkyl phosphites or alkylphosphora idites. Alternatively, activated phosphate diesters may be employed for this purpose.

The nucleic acid base or analog thereof (B) may be chosen from the purines, the pyrimidines, the deaza- purines and the pyrazolopyrimidines. It is preferably selected from uracil-5-yl, cytosin-5-yl, adenin-7-yl, /03736

11 adenin-8-yl, guanin-7-yl, guanin-8-yl, 4-aminopyrrolo- [2,3-d]pyrimidin-5-yl, 2-amino-4-oxopyrrolo[2,3-d]pyri- midin-5-yl, 4-aminopyrazolo[3,4-d]pyrimidin-3-yl or 4- amino-6-oxopyrazolo[3,4-d]pyrimidin-3-yl, where the purines are attached to the sugar moiety of the oligonucleotides via the 9-position, the pyrimidines via the l-position, the pyrrolopyri idines via the 7-position and the pyrazolopyri idines via the l-position.

The functional linking group Y may be chosen 10 from nucleophilic groups such as oxy, thio, amino or chemically blocked derivatives thereof, for example trifluoroaceta ido, phthalimido, CONR' , NR'CO, and S0₂NR', where R¹ Such functionalities, including aliphatic or aromatic amines, exhibit 15nucleophilic properties and are capable of serving as a point of attachment of the -(CH^-A¹ group. Amino groups and blocked derivatives thereof are preferred.

The leaving group A' may be chosen from, for example, such groups as chloro, bromo, iodo, S0₂R^MI, or o S*R'"R"", where each of R'" and R"" is independently Ci.galkyl or aryl or R¹" and R"" together form a C_X._B- alkylene bridge. Chloro, bromo and iodo are preferred. The leaving group will be altered by its leaving ability. ep.end__ng_o_ι,;__-_e- nature and reactivity of the particular 5 leaving group, the group to be used is chosen in each case to give the desired specificity of the irreversibly binding probes.

Examination of double-stranded DNA by ball- and-stick models and high resolution computer graphics 0 indicates that the 7-position of the purines and the 5-position of the pyrimidines lie in the major groove of the B-form duplex of double-stranded nucleic acids. These positions can be substituted with side chains of considerable bulk without interfering with the hybrid- 5 ization properties of the bases. These side arms may be introduced either by derivatization of dThd or dCyd, or by straightforward total synthesis of the heterocyclic 3/03736

12 base, followed by glycosylation. These modified nucleo¬ sides may be converted into the appropriate activated nucleotides for incorporation into oligonucleotides with an automated DNA synthesizer. With the pyrazolo[3,4-d]- pyrimidines, which are analogs of adenine, the crosslink¬ ing arm is attached at the 3-position, which is equivalent to the 7-position of purine.

The crosslinking side chain should be of sufficient length to reach across the major groove from a 0purine 7- or 8-position, pyrimidine 5-position, pyrrolo- pyrimidine 5-position or pyrazolopyrimidine 3-position and reacting with the N-l of a purine (preferably guanine) located above (on the oligomer 3 '-side) the base pair containing the modified analog. Thus, the side chain should be of at least three atoms, preferably of at least five atoms and more preferably of at least six atoms in length. A generally preferred length of the side chain is from about 5 to about 9 carbon atoms.

To optimize strand crosslinking, it would be desirable to have the target strand base which is being attacked paired to the first or second base which is on the 3 ' side of the modified base in the.oligonucleotide containing the crosslinking arm. For example, in the case where_the_«target strand base under attack is a guanine, the target sequence for a probe containing a modified uracil should contain the complement GZA (preferably GGA) , where Z is any base, with the probe oligonucleotide containing UZC (preferably UCC) , where U is dϋrd 5-substituted with the crosslinking arm. In oligonucleotides containing crosslinking adenine deriva¬ tives, for example, the adenine-modified AZ^XC triplet would target GZ^XT, where Z¹ is any base.

It has been found that when the modified base containing the crosslinking arm is a uracil and the target sequence is GGA, alkylation of the second guanine on the target's 5' side of the crosslinker-modified base pair is the exclusive action observed (as shown in Figure 1) . The crosslinking reaction seems to be very specific for the "best fit" of electrophile to nucleophile, i.e., two or more guanine residues may need to neighbor the complement of the modified base to discover the site of ⁵ alkylation.

Two classes of modified 2'-deoxynucleosides have demonstrated particular usefulness in the present invention for incorporation into oligonucleotides as sequence-directed crosslinking agents. The first class o is the 5-substituted-2'-deoxyuridines whose general structure is presented below:

The 5- (substituted) -2 ' -deoxyuridines may be prepared""by-~_:_ϊ_n:outes shown in Schemes 1 and 2.

5

3/03736

14

Scheme 1 ;

• cecxyπ -ose (XXII)

For example, the general procedure of Robins et al. (J. Can. J. Chem.. .6(_:554 (1982); J. Org. Chem..

4j_:l854 (1983)) may be adapted, as shown in Scheme 1, to the palladium-mediated coupling of a substituted 1-alkyne (XXI) to 5-iodo-2'-deoxyuridine (XX) to give the acetylene-coupled product (XXII) . The acetylenic dUrd analog XXII is reduced, with Raney nickel for example, to give the saturated compound (XXIII) , which is then used 3/03736

15 for direct conversion to a reagent for use on an automated DNA synthesizer, as described below.

Scheme 2:

i deoxyribose

(XXVII)

When 5-chloromercurio-2'-deoxyuridine (XXIV) is used as a starting compound, it cannot be directly coupled to an olefin group to give the olefinic compound (XXVII) by palladium-catalyzed coupling with functional- 93/03736

16 ized olefins. Instead, as shown in Scheme 2, a substi¬ tuted alkene (XXV) and 5-chloromercurio-2'-deoxyuridine (XXIV) are reacted together with methanol to give the alpha-methoxy adduct (XXVI) , which is converted to the olefinic compound XXVII by trifluoroacetic acid and trifluoroacetic anhydride. Reduction gives the saturated compound (XXIII) , to be converted to the DNA synthesizer- ready reagent as described below.

B. Pyrazolo[3,4-d,]pyr__midines o The second class of modified nucleoside is a group of 2'-deoxy-4-aminopyrazolo[3,4-d]pyrimidine derivatives. The general structure of these derivatives is presented below:

_tt__."___ove compounds are derived from a novel group of derivatives of 3,4-disubstituted and 3,4,6- trisubstituted pyrazolo[3,4-d]pyrimidines. The 3,4-di¬ substituted and 3,4,6-trisubstituted pyrazolo[3,4-d]pyri- midines and their synthesis are disclosed in commonly owned, copending application Serial No. 250,474, the entire disclosure of which is incorporated herein by reference. They have the following formula (I) :

(I)

3736

I⁷ wherein,

Rj is hydrogen, or a sugar moiety or analog thereof optionally substituted at its 3• or its 5' position with a phosphorus derivative attached to the sugar moiety by an oxygen and including groups Q_{l f} Q₂ and Q₃, or with a reactive precursor thereof suitable for nucleotide bond formation; provided that when R₃ is hydrogen, then R-*. cannot be hydrogen;

Q_x is hydroxy, phosphate or diphosphate; Q₂ is =0 or =S;

Q₃ is CH₂-R', S-R^», O-R', or N-R'R"; each of R' and R" is independently hydrogen or C_j.galkyl;

R₃ is hydrogen or the group -W-(X)_n-A; each of W and X is independently a chemical linker arm;.

A is an intercalator, a metal ion chelator, an electrophilic crosslinker, a photoactivatable cross¬ linker, or a reporter group; each of R_< and R₆ is independently H, OR, SR,

NHOR, NH₂, or NH(CH₂)_tNH₂;

R is H or Cj.βalkyl n is zero or one; and _.t ,3s.zero to twelve. The synthesis of 3,4-disubstituted and 3,4,6- trisubstituted pyrazolo[3,4-d]pyrimidine nucleosides and their use as reagents for incorporation into nucleic acids either enzymatically or via chemical synthesis offers several advantages over current procedures. The de novo chemical synthesis of the nucleotide allows for the incorporation of a wide range of functional groups (e.g., NH₂, SH, OH, halogen, COOH, CN, CONH₂) and the use of different sugar moieties. Also, adenine, guanine, and hypoxanthine analogs are obtained from a single nucleo- side precursor. And, the synthesis does not require the use of toxic heavy metals or expensive catalysts. /03736

18

In the practice of the present invention, the sugar moiety or its analog is selected from those useful as a component of a nucleotide. Such a moiety may be selected from, for example, pentose, deoxypentose, 5 hexose, deoxyhexose, ribose, deoxyribose, glucose, arabinose, pentofuranose, xylose, lyxose, and cyclo- pentyl. The sugar moiety is preferably ribose, deoxy¬ ribose, arabinose or 2'-O-methylribose and embraces either anomer, α or β . o The phosphorus derivative attached to the sugar moiety is conveniently selected from, for example, mono- phosphate, diphosphate, triphosphate, alkyl phosphate, alkanephosphonate, phosphorothioate, phosphorodithioate, and the like. 5 A reactive precursor suitable for internucleo- tide bond formation is one which is useful during chain extension in the synthesis of an oligonucleotide. Reactive groups particularly useful in the present invention are those containing phosphorus. Phosphorus- ocontaining groups suitable for internucleotide bond formation are preferably alkyl phosphorchloridites, alkyl phosphites or alkylphosphoramidites. Alternatively, activated phosphate diesters may be employed for this purpose.. In the above formula I, a chemical linker arm

(W alone or together with X) serves to make the functional group (A) more able to readily interact with antibodies, detector proteins, or chemical reagents, for example. The linkage holds the functional group away from the base when the base is paired with another within the double-stranded complex. Linker arms may include alkylene groups of 1 to 12 carbon atoms, alkenylene groups of 2 to 12 carbon atoms and 1 or 2 olefinic bonds, alkynylene groups of 2 to 12 carbon atoms and 1 or 2 acetylenic bonds, or such groups substituted at a terminal point with nucleophilic groups such as oxy, thio, amino or chemically blocked derivatives thereof 3/03736

19

(e.g., trifluoroacetamido, phthalimido, CONR', NR'CO, and S0₂NR*, where R¹ = H or C_j.galkyl) . Such functionalities, including aliphatic or aromatic amines, exhibit nucleo¬ philic properties and are capable of serving as a point of attachment of the functional group (A) .

The linker arm moiety (W alone or together with X) is preferably of at least three atoms and more preferably of at least five atoms. The terminal nucleo¬ philic group is preferably amino or chemically blocked derivatives thereof.

Intercalators are planar aromatic bi-, tri- or polycyclic molecules which can insert themselves between two adjacent base pairs in a double-stranded helix of nucleic acid. Intercalators have been used to cause frameshift mutations in DNA and RNA. It has also recently been shown that when an intercalator is covalently bound via a linker arm ("tethered") to the end of a deoxyoligonucleotide, it increases the binding affinity of the oligonucleotide for its target sequence, resulting in strongly enhanced stability of the comple¬ mentary sequence complex. At least some of the tethered intercalators also protect the oligonucleotide against exonucleases, but not against endonucleases. See, Sun et a_-_.-_--.Nucl_ri-c^A.--_ts Res.. JL :6149-6158 (1987); Le Doan et al., Nucleic Acids Res. _r .15:7749-7760 (1987). Examples of tetherable intercalating agents are oxazolopyridocar- bazole, acridine orange, proflavine, acriflavine and derivatives of proflavine and acridine such as 3-azido-6- (3-bromopropylamino)acridine, 3-amino-6-(3-bromopentyl- amino)acridine, and 3-methoxy-6-chlσro-9-(5-hydroxy- pentylamino)acridine.

Oligonucleotides capable of crosslinking to the complementary sequence of target nucleic acids are valu¬ able in chemotherapy because they increase the efficiency of inhibition of mRNA translation or gene expression control by covalent attachment of the oligonucleotide to the target sequence. This can be accomplished by cross- /03736

20 linking agents being covalently attached to the oligonuc¬ leotide, which can then be chemically activated to form crosslinkages which can then induce chain breaks in the target complementary sequence, thus inducing irreversible damage in the sequence. Examples of electrophilic crosslinking moieties include alpha-halocarbonyl compounds, 2-chloroethylamines and epoxides.

When oligonucleotides comprising at least one nucleotide base moiety of the invention are utilized as a probe in nucleic acid assays, a label is attached to detect the presence of hybrid polynucleotides. Such labels act as reporter groups and act as means for detecting duplex formation between the target nucleotides and their complementary oligonucleotide probes.

A reporter group as used herein is a group which has a physical or chemical characteristic which can be measured or detected. Detectability may be provided . by such characteristics as color change, luminescence, fluorescence, or radioactivity; or it may be provided by the ability of the reporter group to serve as a ligand recognition site. The pyrazolopyrimidines of the present inven¬ tion of f.ormul______-where Rj is hydrogen may be prepared by the procedures outlined below and as set forth by Kobayashi in Chem. Pharm. Bull., £1:941-951 (1973), the disclosure of which is incorporated herein by reference. ^R3 ^R3 i _____- ' RCOCI + CH(CN)—=→^» HO-C=C(CN). ^ CH0-C=C(CN) ra (mi AY) ^{3 2}

(YD H (YD) 03736

21

In general, alononitrile (III) is treated with acyl halide (II) in the presence of a base to yield acylmalononitrile (IV) , which is subsequently methylated with dimethyl sulfate or diazo ethane, for example, to give the substituted methoxymethylenemalononitrile (V) . This, compound is then reacted with hydrazine hydrate in boiling alcohol to give the 3-substituted-5-aminopyra- zole-4-carbonitrile (VI) , which is treated with cold concentrated sulfuric acid to give the 3-substituted-5- 0 minopyrazole-4-carboxamide (V I) .

The carboxamide (VII) may alternatively be prepared by treating cyanoacetamide (XII) with acid halide (II) to give the acylcyanoacetamide (XIII) , which is then methylated, and the resulting methoxy compound 5 (XIV) is reacted with hydrazine hydrate.

0 ^R3 O ^R3 0

R COCi + CH - C - NH ^KO - C-C.C-NH.→-CH.O-C-C-C.NH.

3 i 2 2 | ^{2 3} | 0 CN ra (Xil)

(YII) 5

Syntheses of pyrazolo[3 , 4-d]pyrimidines are accomplished from the two pyrazole intermediates , VI, and

VII . Thus, 3 , 4-disubstituted pyrazolo[3 , 4-d]pyrimidines

(VIII and X) are obtained by treating the corresponding VI and VII with boiling formamide. Alternatively, VI may be treated with dialkoxymethyl ester of a carboxylic acid, at room temperature or above room temperature, and then with ammonia to give VIII, and VII may be treated with dialkoxymethyl ester of a carboxylic acid (without subsequent ammonia treatment) , at room temperature or above room temperature, to give compound X. 3 , 4 , 6-Tri- substituted pyrazolo [ 3 , 4-d] pyrimidines (IX and XI) are obtained by fusing the corresponding VI and V I with urea and thiourea (H₂N)₂C=R₆ (where R₆ is 0 or S) . Alternative¬ ly, VI and VII may be treated with an alkyl xanthate salt such as potassium ethyl xanthate and with alkyl halide such as methyl iodide, at a temperature above room temperature, followed by oxidation by a peroxide such as m-chloroperbenzoic acid (MCPBA) and subsequent treatment with ammonia to give IX and XI, respectively, where R₆ is NH₂.

(VIII) (IX) (X) (XI)

The compounds of formula I may be recovered from the reaction mixture in which they are formed by established procedures.

In the compounds of formula I where _: is a sugar moiety, the sugar may be either added to the 1- position of the pyrazole VI or VII prior to further treatment or added to the l-position of the pyrazolo[3,4- d]pyrimidine VIII, IX, X or XI. To add the sugar, the pyrazole or pyrazolopyrimidine is treated with sodium hydride and then with the glycosyl halide of the blocked sugar. Oligonucleotides of the present invention may comprise at least one and up to all of their nucleotides from the substituted pyrazolo[3,4-d]pyrimidines of formula I and/or at least one and up to all of their nucleotides from the substituted nucleotide bases of formula I' .

To prepare oligonucleotides, protective groups are introduced onto the nucleosides of formula I or formula I' and the nucleosides are activated for use in the synthesis of oligonucleotides. The conversion to protected, activated forms follows the procedures as described for 2'-deoxynucleosides in detail in several reviews. See, Sonveaux, Bioorσanic Chemistry. 14:274-325 (1986); Jones, in "Oligonucleotide Synthesis, a Practical Approach", M.J. Gait, Ed., IRL Press, p. 23-34 (1984).

The activated nucleotides are incorporated into oligonucleotides in a manner analogous to that for DNA 0and RNA nucleotides, in that the correct nucleotides will be sequentially linked to form a chain of nucleotides which is complementary to a sequence of nucleotides in target DNA or RNA. The nucleotides may be incorporated either enzymatically or via chemical synthesis. The nucleotides may be converted to their 5'-O-dimethoxy- trityl-3'- (N,Jf.-diisopropyl)phosphoramidite cyanoethyl ester derivatives, and incorporated into synthetic oligonucleotides following the procedures in "Oligonucleotide Synthesis: A Practical Approach", supra . The _^■*.-protecting groups are then removed, along with the other oligonucleotide blocking groups, by post-synthesis aminolysis, by procedures generally known in the a t.

In a preferred embodiment, the activated nucle tides-may_^be used directly on an automated DNA synthesizer according to the procedures and instructions of the particular synthesizer employed. The oligonucleo¬ tides may be prepared on the synthesizer using the standard commercial phosphoramidite or H-phosphonate chemistries. In another preferred embodiment, the amino- pyrazolopyrimidine nucleotide triphosphates may substitute for an adenine using the nick translation procedure, as described by Langer et al. , Proc. Natl. Acad. Sci. USA. 211:6633-6637 (1981), the disclosure of which is incorporated herein by reference.

The leaving group, such as a haloacyl group, may be added to the aminoalkyl tails (-CH₂)_-Y) following incorporation into oligonucleotides and removal of any blocking groups. For example, addition of an α-haloacet- amide may be verified by a changed mobility of the modified compound on HPLC, corresponding to the removal 5 of the positive charge of the amino group, and by sub¬ sequent readdition of a positive charge by reaction with 2-aminoethanethiol to give a derivative with reverse phase HPLC mobility similar to the original aminoalkyl- oligonucleotide. ₁0 In specific embodiments, each of the following electrophilic leaving groups were attached to an amino- propyl group on human papillomavirus (HPV) probes: bromoacetyl, iodoacetyl and the less reactive but cσnfor- mationally more flexible 4-bromobutyryl. Bromoacetyl^' and 15iodoacetyl were found to be of equal reactivity in crosslinking.

An oligonucleotide probe according to the invention includes at least one labeled substituted pyrazolo[3,4-d] yrimidine nucleotide moiety of formula I 20and/or at least one labeled substituted nucleotide base of formula I' .

Probes may be labeled by any one of several methods typically used in the art. A common method of detection—-is- he-use of autoradiography with ³H, ¹²⁵I, ³⁵S, 25¹⁴c, or ³²P labeled probes or the like. Other reporter groups include ligands which bind to antibodies labeled with fluorophores, chemiluminescent agents, and enzymes. Alternatively, probes can be conjugated directly with labels such as fluorophores, chemiluminescent agents, 30enzymes and enzyme substrates. Alternatively, the same components may be indirectly bonded through a ligand- antiligand complex, such as antibodies reactive with a ligand conjugated with label. The choice of label depends on sensitivity required, ease of conjugation with 5the probe, stability requirements, and available instrumentation. The choice of label dictates the manner in which the label is incorporated into the probe. Radio¬ active probes are typically made using commercially available nucleotides containing the desired radioactive 5 isotope. The radioactive nucleotides can be incorporated into probes, for example, by using DNA synthesizers, by nick-translation, by tailing of radioactive bases to the 3' end of probes with terminal transferase, by copying M13 plasmids having specific inserts with the Klenow ° fragment of DNA polymerase in the presence of radioactive dNTP's, or by transcribing RNA from templates using RNA polymerase in the presence of radioactive rNTP's.

Non-radioactive probes can be labeled directly with a signal (e.g., fluorophore, chemiluminescent agent 5 or enzyme) or labeled indirectly by conjugation with a ligand. For example, a ligand molecule is covalently bound to the probe. This ligand then binds to a receptor molecule which is either inherently detectable or covalently bound to a detectable signal, such as an oenzyme or photoreactive compound. Ligands and anti- ligands may be varied widely. Where a ligand has a natural "antiligand", namely ligands such as biotin, thyroxine, and cortisol, it can be used in conjunction with.,j.ts—labeled-,- naturally occurring antiligand. 5Alternatively, any haptenic or antigenic compound can be used in combination with a suitably labeled antibody. A preferred labeling method utilizes biotin-labeled analogs of oligonucleotides, as disclosed in Langer et al., Proc. Natl. Acad. Sci. USA. 78.:6633-6637 (1981), which is Oincorporated herein by reference.

Enzymes of interest as reporter groups will primarily be hydrolases, particularly phosphatases, esterases, ureases and glycosidases, or oxidoreductases, particularly peroxidases. Fluorescent compounds include fluorescein and its derivatives, rhodamine and its derivatives, dansyl, umbelliferone, rare earths, etc. 3/03736

26

Chemiluminescers include luciferin, acridinium esters and 2,3-dihydrophthalazinediones, e.g., luminol.

The specific hybridization conditions are not critical and will vary in accordance with the investi- 5 gator's preferences and needs. Various hybridization solutions may be employed, comprising from about 20% to about 60% volume, preferably about 30%, of a polar organic solvent. A common hybridization solution employs about 30-60% v/v formamide, about 0.5 to 1M sodium 10 chloride, about 0.05 to 0.1M buffers, such as sodium citrate, Tris HC1, PIPES or HEPES, about 0.05% to 0.5% detergent, such as sodium dodecylsulfate, and between 1- 10 mM EDTA, 0.01% to 5% ficoll (about 300-500 kdal) , 0.1% to 5% polyvinylpyrrolidone (about 250-500 kdal) , and ¹⁵0.01% to 10% bovine serum albumin. Also included in the typical hybridization solution will be unlabeled carrier nucleic acids from about 0.1 to 5 mg/ml, e.g., partially fragmented calf thymus or salmon sperm, DNA, and/or partially fragmented yeast RNA and optionally from about 20 o.5% to 2% wt./vol. glycine. Other additives may also be included, such as volume exclusion agents which include a variety of polar water-soluble or swellable agents, such as anionic polyacrylate or polymethylacrylate, and charged-sacch_-_5_Ldic polymers, such as dextran sulfate.

25 The particular hybridization technique is not essential to the invention. Hybridization techniques are generally described in "Nucleic Acid Hybridization, A Practical Approach", Hames and Higgins, Eds., IRL Press, 1985; Gall and Pardue, Proc. Natl. Acad. Sci. , U.S.A..

30.63.:378-383 (1969); and John et al., Nature, ,22J_:582-587

(1969) . As improvements are made in hybridization techniques, they can readily be applied.

The amount of labeled probe which is present in the hybridization solution may vary widely. Generally, 5 substantial excesses of probe over the stoichiometric amount of the target nucleic acid will be employed to 03736

27 enhance the rate of binding of the probe to the target DNA.

Various degrees of stringency of hybridization can be employed. As the conditions for hybridization become more stringent, there must be a greater degree of complementarity between the probe and the target for the formation of a stable duplex. The degree of stringency can be controlled by temperature, ionic strength, the inclusion of polar organic solvents, and the like. For example, temperatures employed will normally be in the range of about 20° to 80°C, usually 25° to 75^βC. For probes of 15-50 nucleotides in 50% formamide, the optimal temperature range can vary from 22-65°C. With routine experimentation, one can define conditions which permit satisfactory hybridization at room temperature. The stringency of hybridization is also conveniently varied by changing the ionic strength and polarity of the reactant solution through manipulation of the concen¬ tration of formamide within the range of about 20% to about 50%.

Treatment with ultrasound by immersion of the reaction vessel into commercially available sonication baths can oftentimes accelerate the hybridization rates. Af-te_f—hybridization at a temperature and time period appropriate for the particular hybridization solution used, the glass, plastic, or filter support to which the probe-target hybrid is attached is introduced into a wash solution typically containing similar reagents (e.g., sodium chloride, buffers, organic solvents and detergent), as provided in the hybridization solution. These reagents may be at similar concentra¬ tions as the hybridization medium, but often they are at lower concentrations when more stringent washing condi¬ tions are desired. The time period for which the support is maintained in the wash solutions may vary from minutes to several hours or more. 93/03736 ^r^

28

Either the hybridization or the wash medium can be stringent. After appropriate stringent washing, the correct hybridization complex may now be detected in accordance with the nature of the label. 5 The probe may be conjugated directly with the label. For example, where the label is radioactive, the support surface with associated hybridization complex substrate is exposed to X-ray film. Where the label is fluorescent, the sample is detected by first irradiating 0 it with light of a particular wavelength. The sample absorbs this light and then emits light of a different wavelength which is picked up by a detector ("Physical Biochemistry", Freifelder, D., W.H. Freeman & Co., 1982, pp. 537-542) . Where the label is an enzyme, the sample 5 is detected by incubation with an appropriate substrate for the enzyme. The signal generated may be a colored precipitate, a colored or fluorescent soluble material, or photons generated by bioluminescence or chemilumin- escence. The preferred label for dipstick assays generates a colored precipitate to indicate a positive reading. For example, alkaline phosphatase will dephos- phorylate indoxyl phosphate which then will participate in a reduction reaction to convert tetrazolium salts to highly, colored—and .insoluble formazans. Detection of a hybridization complex may require the binding of a signal generating complex to a duplex of target and probe polynucleotides or nucleic acids. Typically, such binding occurs through ligand and antiligand interactions as between a ligand-conjugated Probe and an antiligand conjugated with a signal. The binding of the signal generation complex is also readily amenable to accelerations by exposure to ultrasonic energy.

The label may also allow indirect detection of the hybridization complex. For example, where the label is a hapten or antigen, the sample can be detected by using antibodies. In these systems, a signal is gener- ated by attaching fluorescent or enzyme molecules to the antibodies or in some cases, by attachment to a radio¬ active label. (Tijssen, P., "Practice and Theory of Enzyme Immunoassays, Laboratory Techniques in Bio- chemistry and Molecular Biology", Burdon, R.H., van Knippenberg, P.H. , Eds., Elsevier, 1985, pp. 9-20).

The amount of labeled probe present in the hybridization solution may vary widely, depending upon the nature of the label, the amount of the labeled probe that can reasonably bind to the cellular target nucleic acid, and the precise stringency of the hybridization medium and/or wash medium. Generally, substantial probe excesses over the stoichiometric amount of the target will be employed to enhance the rate of binding of the probe to the target nucleic acids.

The invention is also directed to a method for identifying target nucleic acid sequences, which method comprises utilizing an oligonucleotide probe including at least one labeled substituted nucleotide moiety of formula I and/or formula I¹.

In one embodiment, the method comprises the steps of:

(a) denaturing nucleic acids in the sample to be-tested (b) hybridizing to the target nucleic acids an oligonucleotide probe including at least one labeled substituted nucleotide moiety of formula I or formula I' , wherein the probe comprises a sequence complementary to that of the target nucleic acids; (c) washing the sample to remove unbound probe;

(d) incubating the sample with detection agents; and

(e) inspecting the sample. The above method may be conducted following procedures well known in the art. An assay for identifying target nucleic acid sequences utilizing an oligonucleotide probe including at least one labeled substituted nucleotide moiety of formula I and/or formula I' and comprising the above 5 method is contemplated for carrying out the invention. Such an assay may be provided in kit form. For example, a typical kit will include a probe reagent component comprising an oligonucleotide including at least one labeled nucleotide moiety of formula I or formula I' , the o oligonucleotide having a sequence complementary to that of the target nucleic acids; a denaturation reagent for converting double-stranded nucleic acid to single- stranded nucleic acid; and a hybridization reaction mixture. The kit can also include a signal-generating 5 system, such as an enzyme for example, and a substrate for the system.

C. Enzyme-Mediated Triple Strand Formation Antisense ODNs Oligodeoxynucleotides (ODNs) have great potential as sequence specific pharmaceutical agents for the inhibition of gene expression. For instance, chemically synthesized ODNs may inhibit the expression of specific— ene-pxuducts through formation of duplexes upon hybridization with complementary messenger RNAs (mRNAs) . More specifically, these "antisense" ODNs are believed to inhibit the processing or translation of message primarily through an RNase H-mediated cleavage of the target mRNA sequence. Because of this inhibitory effect, ntisense ODNs may be useful as anti-viral, anti- parasitic, and anti-cancer agents. Further, antisense ODNs provide a unique opportunity for rational drug development, since a genetic target offers both exquisite specificity and universality with respect to potential target sequences.

However, "antisense" technology is beset with certain fundamental disadvantages. One major challenge involves development of antisense ODNs with sufficient potency to warrant in vivo testing. Antisense ODN formulations that exhibit nuclease resistance, rapid cellular uptake, and efficient and stable hybridization to the target RNA sequence are desirable. Improvement of one or more of these properties without concomitant deleterious effects on other properties may be difficult. For instance, antisense ODNs with modified backbones, such as oligonucleoside methylphosphonates and 0phosphorothioates, exhibit excellent nuclease resistance relative to unmodified ODNs, but methylphosphonate ODNs form DNA-RNA hybrids that are refractile to RNase H and phosphorothioate ODNs are poorly taken up into cells.

Another approach to improving effectiveness of 5antisense ODNs involves attaching moieties to ODNs in order to potentiate their antisense activity. For instance, moieties that interact directly with the RNA target upon hybridization (such as pendant intercalating groups that stabilize ODN-target hybrids, ODNs with free 0radical-based RNA cleaving activity or ODNs capable of covalently linking to their targets upon hybridization) may be useful in this regard. Direct cleavage or cross- linkage of an mRNA target by an ODN renders the RNA ■inactive-r 5 Cellular uptake of antisense ODNs is another area of concern. While charged ODNs may be actively taken up into mammalian cells through an energy dependent pathway, augmentation of this pathway to achieve high intracellular concentrations of antisense ODN is _θdesirable. "Endocytic approaches", which enhance receptor-mediated cellular targeting, include conjugation of a cholesteryl group that may act as an lipophilic anchor for the antisense ODN; encapsulation of ODNs into Uposomes; and linkage of ODNs to soluble macromolecular complexes. Application of the above-noted approaches, alone or in combination, may provide an increase in antisense ODN potency in the range of 1000-fold and may significantly reduce the cost of ODN synthesis. 5

Anti-gene ODNs

A variation of the "antisense" approach to rational drug design is termed "anti-gene". Whereas antisense ODNs target single stranded mRNA, anti-gene 0ODNs hybridize with and are capable of inhibiting the function of double-stranded DNA. More specifically, anti-gene ODNs form sequence-specific triple-stranded complexes with a double stranded DNA target and thus interfere with the replication or transcription of 5selected target genes. Except for certain RNA viruses and nucleic acid-free viroids, DNA is the repository for all genetic information, including regulatory control sequences and non-expressed genes, such as dormant proviral DNA genomes. In contrast, the target for (antisense ODNs, mRNA, represents a very small subset of the information encoded in DNA. Thus, anti-gene ODNs have broader applicability and are potentially more powerful than antisense ODNs that merely inhibit mRNA processi-ng-and— ranslation. Anti-gene ODNs in the nuclei of living cells can form sequence-specific complexes with chromosomal DNA. The resultant triplexes can inhibit restriction and/or transcription of the target double stranded DNA. Based on the known stabilities of the two target nucleic acid species (i.e., DNA and RNA), anti-gene interference with DNA functioning has longer lasting effects than the corresponding antisense inhibition of mRNA function. Mammalian cell DNA does not turnover; in fact, cells possess sophisticated pathways capable of repairing lesions in DNA that may arise from environmental insults or from spontaneous rearrangements. In contrast, mRNA is transient and may exist only for minutes within a cell. The constant turnover of an mRNA species and the potentially high copy number of such mRNA species suggest that anti-sense ODNs will provide relatively short term effects. While cellular uptake of anti-gene ODNs may 5 need to be augmented to achieve sufficient intracellular concentrations, once within the cell the ODNs naturally concentrate in the nucleus.

Anti-gene therapy is based on the observation that certain DNA homopolymers can form triple-stranded o complexes. In these triple-standed complexes, the third strand resides in the major groove of the Watson Crick base-paired double helix, where it hydrogen bonds to one of the two parental strands. A binding code governs the recognition of base pairs by a third base (see allowed 5 triplets below). In each case, the third strand base is presented first and is followed by the base pair; hydrogen bonding between the first two bases maintains the third base interaction.

0 A-A-T G-G-C

T-A-T C-G-C

Certain limitations of this base pair recognition code are_..appar-ent—from the allowed triplets. First, there is ₅no capability for the recognition of T-A and C-G base pairs; hence, triple strand formation is restricted to runs of homσpurine bases on one strand and homopyrimidine bases on the other strand of the duplex. Second, if cytosine is in the third strand ("C") , it must be 0protonated to be able to hydrogen bond to the guanine of a G-C base pair. The pKa for protonation of cytosine is 4.6, suggesting that at physiological pH the stability of C-G-C triads is likely to be impaired. The substitution of 5-methyl cytosine or the use of polyvalent cations (such as spermine or spermidine) may stabilize the C-G-C triplets at pH 7.0. Third, in all cases triads are maintained by two hydrogen bonds between the third strand 03736

34 base and the purine residue of the base pair. Hence, triple-stranded complexes are generally less stable than the parental double-stranded DNA, which is maintained by a combination of two (A-T) or three (G-C) hydrogen bonds between purine and pyrimidine pairs.

Cytosine/thy idine-, guanine/adenine- and guanine/thymidine-containing ODNs can sequence- specifically bind to homopurine runs in double-stranded DNA. These recognition motifs are based on Hoogstein or 0reverse Hoogstein base pairing. In the C/T recognition motif, the ODN is parallel to the homopurine strand of the duplex; in the G/A recognition motif, the ODN is anti-parallel to the homopurine strand; in the G/T recognition motif, the ODN may bind parallel or anti- 5parallel to the homopurine strand of the duplex, depending on the G content of the third strand. These recognition motifs may be sequence dependent. The sequence specificity of anti-gene ODNs using the C/T recognition motif permits hybridization of such ODNs to ohomopurine runs in plasmid DNA and in yeast chromosomes. Since ODN binding is restricted to homopurine runs, it would be advantageous to identify additional heterocycles that can recognize the remaining two base pairs, i.e., C- G_. and T-A-.- While guanosine may be used in the third 5Strand to recognize T-A base pairs, this interaction involves only one hydrogen bond and is relatively unstable.

Analogous to anti-sense ODNs, anti-gene ODNs may be modified with a variety of pendant groups designed to augment their activity. Hence, intercalating groups, cleaving agents, and crosslinking moieties may be appended to the termini of anti-gene ODNs. Upon triplex formation, these groups interact with the adjacent duplex and intercalate, cleave or crosslink, respectively. Further, in the C/T recognition motif, substitution of 5-methyl cytosine for cytosine in the third strand ODN significantly stabilizes triplexes formed with "G"-rich /03736

35 homopurine runs. In addition, ODNs with modified backbones, such as oligonucleoside methyl-phosphonates and phosphorothioates, may form triple-stranded complexes.

Enzyme-Mediated Triple Strand Formation

A important disadvantage of triple strand formation as discussed above is the relatively slow 0 inetics of triple strand formation. The claimed invention overcomes this disadvantage through enzyme catalysis of triple strand formation, with recombination enzymes particularly preferred for this purpose. More significantly, enzyme-catalyzed triple strand formation 5provides the immense advantage of universal sequence recognition (in contrast to the A-T and G-C recognition limitation associated with non-enzyme-mediated triple strand formation) .

Briefly, the genetic material of all living organisms is occasionally subject to homologous rearrangement. Recombination between like (or homologous) DNA molecules, which plays an important role in promoting genetic diversity within a species, is of enzymes in both procaryotic and eucaryotic cells. Recombination also plays an important role in DNA repair and in the immune response.

Homologous recombination in E. coll serves as an illustrative example of processes that occur in both procaryotic and eucaryotic cells. Studies of purified recombinational enzymes from E. coli in a defined cell- free system permit division of homologous recombination into three steps. For purposes of this illustration, the "invading" single strand is circular and the target double strand is linear. In the first step, "presynapsis", single stranded circular DNA is coated with the multifunctional protein recA in the presence of ATP. The resultant 36 nucleoprotein filament possesses a right handed helical twist composed of 18 bases per turn and 1 monomer of recA protein per 3.6 nucleotides.

In the second step, "synapsis", the single- stranded circular nucleoprotein filament conducts a two- dimensional search along a linear double-stranded DNA template for homologous sequences. The search concludes with homologous alignment of the two molecules in an initial complex that has no net helical interwinding (the complex is referred to as a "paranemic joint") . Paranemic joints are highly unstable and upon deproteinization readily dissociate. Very rapidly, the DNA double helix is incorporated into the filament, and in so doing, the two DNA molecules become plectonemically coiled (i.e., helically interwound) . In this triple strand form, the newly incorporated third strand is homologous (in sequence and polarity) to one of the parental duplex strands, and complementary to the other parental duplex strand. In the resultant synaptic complex, the three strands of DNA are believed to exist as a true hydrogen-bonded triple strand having a close association with recA. In contrast to paramenic joints, the plectonemically coiled complex is stable upon (removal of recA) .

The plectonemically coiled synaptic complex contains 18 triads per right handed helical turn. While recA is a critical part of the complex, sequence specificity resides entirely in the hydrogen bonding between bases. In a plectonemically coiled synaptic complex, two alternative conformations exist. In one, the negative linear strand of parental DNA is Watson Crick base-paired to the positive linear strand; in the other, the negative linear strand of parental DNA is Watson Crick base-paired to the positive circular single strand. In both conformations, the bases of the third strand are presumed to be hydrogen-bonded in a Hoogstein or reverse Hoogstein fashion to the purines of the other 37 two strands (according to the binding code described above) . The single stranded circular DNA within the triple stranded nucleoprotein filament subsequently displaces the homologous linear strand ("strand 5exchange") . This displacement features a recA-catalyzed, energy-dependent release of the linear positive strand, coupled with hybridization of the circular single strand to the complementary linear strand. During both synapsis and strand exchange, recA remains associated with the locomplex. Finally, recA dissociates from the newly formed double strand and recombination is complete.

Upon deproteinization of synaptic complexes, the protein-free triple stranded DNA complexes are essentially resistant to single strand-specific nucleases 15and exhibit very high T_'s. This is likely attributable to hydrogen bonding of the so-called "third strand" to both parental strands, as well as to the highly underwound state of the complex. The highly underwound, deproteinized synaptic complexes therefore are analogous ₀to highly underwound recA-containing complexes. Thus, recA may catalyze the formation of a DNA triple strand complex not otherwise attainable. In this complex, the two positive strands have identical sequence and positive strand is capable of forming a double stranded hybrid with the negative parental strand.

Recent evidence indicates that recA alters the conformation of the triplex, such that strand switching is allowed. That is, the bases of the third strand can hydrogen bond to the appropriate purine base regardless of which duplex strand contains the purine base.

Several factors may affect the stability of synaptic complexes, including: (1) the length of the invading single strand; (2) the extent of shared homology between the invading single strand and the target duplex (with length of homology and absence of mismatches of particular importance) ; (3) the position of the shared 3/03736

38 homology region with respect to the termini of the target duplex; and (4) the superhelicity, if any, of the duplex.

The following factors may be important when selecting and designing an optimal system for stable 5 triple strand formation with an anti-gene:

Length: An oligomeric 50-mer should have sufficient length to permit the formation of a stable presynaptic complex with recA.

Shared Homology: A minimal shared homology of 1013 bases appears to be necessary for formation of recA-stabilized synaptic filaments. Photo-fixation of a recA-stabilized triple strand by an HMT furan side monoadducted ODN may add further contraints to a minimal or preferred level of shared homology. With photo- 15fixation, a conformational change within the triple strand complex allows the invading third strand to interchange position with the homologous duplex strand. This results in the third strand bases being Watson Crick base-paired to the complementary strand. Thus, in a 0 preferred embodiment, the ODN and target share 50 bases of homology.

Type of Joint: Depending upon the positional relationship of the region of shared homology to the ends (if._a,ny)-o-f—the-targeted double-stranded DNA, the resultant synaptic complex is classified as a proximal joint, a medial joint, or a distal joint. Proximal joints, wherein the recA-stabilized triple strand complex is located at the left hand of a linear duplex, are unstable due to recA-catalyzed strand exchange. During 0Strand exhange, the duplex strand that is homologous to the invading third strand in the triple strand complex is displaced, resulting in a new duplex containing the invading strand and a free single strand. Since strand exchange requires a free 5' end and proceeds with a 5' to 53' polarity, this process readily occurs in proximal joints. 03736

39

By contrast, distal joints and medial joints lack the appropriate ends and do not readily undergo recA-catalyzed strand exchange. As a result, these joints are highly stable structures. Accordingly, within the present invention, formation of recA-stabilized triple strand complexes so as to form meidal or distal joints is preferred.

Superhelicity: Superhelical targets may facilitate the formation of recA-stabilized triple strands. However, high efficiency triple strand formation (even using a DNA target and ODN with a very small region of shared homology) may be obtained using linear targets. Selection of a synthetic, oligomeric double stranded target DNA versus a supercoiled target in the model system will vary with the characteristics of the target duplex DNA.

The present invention combines the recA- catalyzed formation of stable triple strand complexes with synthetic anti-gene ODNs. Enzyme-catalyzed triple strand formation exhibits rapid kinetics and universal DNA sequence recognition. A recA-coated anti-gene ODN serves as a "guide" that seeks homology in a target double s-trand-cBNA sequence; upon recognition and binding of this nucleoprotein filament, recA catalyzes the formation of sequence-specific triple strand complexes. Within the present invention, it is preferred that an anti-gene ODN is least 30-40 nucleotides in length and has a base sequence and polarity identical to either of the two duplex strands in the target DNA. The frequency of triplex formation and the stability of the triple strand complexes following deproteinization may be directly related to the length of the anti-gene ODN. Anti-gene ODNs (of the appropriate polarity) may be used in combination with endogenous recombinatory pathways to form sequence-specific triple strand complexes with chromosomal DNA. Alternatively, anti-gene ODNs of a /03736

40 variety of lengths may be complexed with a recombinational enzyme prior to combination with target double strand DNA.

Using the recombinational machinery of cells to catalyze the formation of hydrogen bonded triple-stranded complexes overcomes several disadvantages associated with current anti-gene approaches. First, the enzyme-mediated recognition motif recognizes all four base pairs, thereby allowing targeting of any double stranded DNA sequence. 0second, the recA-coated, single stranded anti-sense ODN (nucleoprotein filament) searches for target double strand DNA homology much more efficiently than does a small naked anti-gene ODN, thus decreasing the concentration of anti-gene ODN required for efficient 5triple strand complex formation. Third, due to the hydrogen bonding patterns and the novel helical twist involved in enzyme-mediated recognition, the resultant triple strand complex is stable at physiological pH. Fourth, since the cellular recombinational pathway is 0being harnessed, the DNA in higher order chromatin structures will be accessible for targeting. And fifth, the resultant triple strand complexes display increased stability with increased anti-gene ODN length.

-The---ability to conduct an efficient homology 5search is a significant advantage. Preliminary data indicate that ODNs are very inefficient at scanning double stranded DNA for complementary homopurine sequences. For instance, non-enzyme-catalyzed triple strand formation jLn vitro has been analyzed after 60 min at 37°C (F.M. Orson et al. , Nucl. Acids Res. 19:3435-41, 1991) . In contrast, a classical hybridization between two complementary single strands would occur within seconds, rather than hours. Since the human genome contains over 3 X 10⁹ base pairs, the homology search time may be inordinately long, especially if anti-gene ODNs are used at relatively low concentration. The use of presynaptic nucleoprotein filaments, such as those formed 03736

41 between single stranded DNA and recA, that display a weak nonspecific affinity for double stranded DNA effectively reduces the homology search from a three dimensional to a two dimensional process. Furthermore, upon homologous registry with the double strand, the nucleoprotein filament will more likely produce a triple strand complex than the corresponding interaction of double strand and a naked single strand. Because of these factors, triple strand formation between a recA-coated, single stranded ODN and an homologous double strand occurs at a reaction rate that exceeds by 1 or 2 orders of magnitude the calculated rate of spontaneous renaturation of complementary single strands under standard hybridization conditions.

The present invention involves combination of (1) an ODN that is homologous to a portion of one target DNA duplex strand and complementary to the analogous portion of the other target DNA duplex strand; and (2) ^enzyme-catalyzed triple strand formation to achieve inactivation of a target DNA sequence. In a preferred embodiment, the ODN has a crosslinking moiety covalently attached thereto. Upon hybridization to the target DNA sequ.erice_T—cross-linkage of the third strand ODN (the anti- gene ODN) to both parental duplex strands inactivates the target DNA sequence. Depending on the ODN, the crosslinking moiety, the target DNA sequence and the environment and characteristics of the target DNA sequence, inactivation of the parental duplex strands may be permanent. Thus, a single administration of one or more anti-gene ODNs may abolish the expression of integrated retroviral genomes, of episomal herpesvirus genomes, or of mutant oncogenes.

Following triple strand formation and covalent crosslinkage, the modified target DNA no longer supports replication or transcription. Unlike all other lesions in DNA, however, this modification is not repairable. 03736

42

Normally, crosslinked DNA is repaired by a combination of excision repair and homologous recombination. With crosslinked triple strand complexes, however, there will be no undamaged copies of the targeted gene to 5 participate in recombination. By analogy with procaryotic models, the eucaryotic cell may attempt to use a misrepair (or SOS) pathway wherein the crosslink will be removed, but at the expense of mutagenesis. In such case, gene function would be irreversibly silenced o by the resultant mutations.

The use of recombination enzymes in combination- with anti-gene ODNs significantly enhances the efficiency with which the single strand ODN "finds" its complementary target DNA sequence. Accordingly, the 5 efficiency of triple strand formation is greatly increased when the anti-gene ODN is combined with a recombination enzyme (for instance, in a nucleoprotein complex) .

Within the present invention, suitable target DNA sequences include structural genes and both up-stream and down-stream regulatory control sequences. These regulatory sequences may be involved in either transcription or replication. The anti-gene ODN will be determined-and-designed according to the target DNA sequence chosen for alteration of function, and will have a sequence complementary to one of the two strands of the chosen target DNA.

Crosslinkers suitable for use within the invention include photochemical agents, such as psoralens, and chemical crosslinking agents. Preferred chemical crosslinking agents include those described in Section A., above; electrophilic moieties attached to the 3' and/or 5' termini of the ODN; and masked electrophilic moieties attached to the ODN. Photo-crosslinking agents may be useful for topical or extracorporeal applications, for targets accessible to light exposure, as well as for in vitro use. Chemical crosslinkers may be used without /03736

43 limitation and are particularly preferred for use within the claimed invention.

Preferred recombination enzymes include procaryotic and eucaryotic recombination enzymes, such as 5 recA, human recombinase and Drosophila recombinase, with human recombinase particularly preferred.

In a particularly preferred embodiment, an anti-gene ODN is administered to a cell or a host, and upon entry to a target cell nucleus, the anti-gene ODN o combines with recombination enzymes present within the nucleus. In an alternative embodiment, the anti-gene ODN and recombination enzyme are combined ex vivo and then administered to a cell or a host as a nucleoprotein filament. In this embodiment, it may be advantageous to 5 administer the nucleoprotein filament in a liposome.

The following examples are provided to illustrate the present invention without limiting same. o"*RT" means room temperature.

General

Thin layer chromatography was performed on silica g-e-l-60-r*F--254 plates (Analtech) using the following 5Sθlvent mixtures: A- 90% methylene chloride:10% methanol; B- 50% ethyl acetate:50% hexanes; C- 70% ethyl acetate: 10% methanol:10% water:10% acetone; D- 50% ether:50% hexanes. Flash chromatography was performed using 60 F 254 silica (Merck) . Oligonucleotides were synthesized on 0an Applied Biosystems Model 380B Synthesizer. Oligo¬ nucleotides were isotopically labeled using T4 Polynucleotide kinase (BRL) and τ-³²P-ATP (New England Nuclear) .

5 03736

44

EXAMPLE 1: 6-(Tritylamino)caproic Acid.

6-Aminocaproic acid (26 g, 0.2 mole) was dissolved in dichloromethane (200 mL) by the addition of triethylamine (100 mL) . Trityl chloride (120 g, 0.45 mole) was added and the solution stirred for 36 hr. The resulting solution was extracted with IN HC1 and the organic layer evaporated to dryness. The residue was suspended in 2-propanol/lN NaOH (300 mL/100 mL) and refluxed for 3 hr. The solution was evaporated to a thick syrup and added to dichloromethane (500 mL) . Water was added and acidified. The phases were separated, and the organic layer dried over sodium sulfate and evapor- ated to dryness. The residue was suspended in hot 2- propanol, cooled, and filtered to give 43.5 g (58%) of 6- (tritylamino)caproic acid, useful as an intermediate compound.

EXAMPLE 2:

5-(Tritylamino)pentylhydroxymethylenemalononitrile.

To a dichloromethane solution of 6-(tritylamino)- caproic -a-cid—(*_^€-.-0 g, 53 mmole) and triethylamine (20 mL) in an ice bath was added dropwise over 30 min isobutyl- chloroformate (8.3 mL, 64 mmole). After the mixture was stirred for 2 hr in an ice bath, freshly distilled malononitrile (4.2 g, 64 mmole) was added all at once. The solution was stirred for 2 hr in an ice bath and for 2 hr at RT. The dichloromethane solution was washed with ice cold 2N HC1 (300 mL) and the biphasic mixture was filtered to remove product that precipitated (13.2 g) . The phases were separated and the organic layer dried and evaporated to a thick syrup. The syrup was covered with dichloromethane and on standing deposited fine crystals of product. The crystals were filtered and dried to give 6.3 g for a total yield of 19.5 g (87%) of the product, which is useful as an intermediate.

/03736

46

EXAMPLE 3 :

5-(Tritylamino)pentylmethoxymethylenemalononitrile.

A suspension of the malononitrile of Example 2 (13 g, 31 mmole) in ether/dichloromethane (900 mL/100 mL) , cooled in an ice bath, was treated with a freshly prepared ethereal solution of diazomethane (from 50 mmole of Diazald® (Aldrich Chemical Company) ) . The solution was stirred for 6 hr and then neutralized with acetic acid (10 mL) . The solution was evaporated to dryness and the residue chromatographed on silica gel using dichloro¬ methane/acetone (4/1) as the eluent. Fractions containing product were pooled and evaporated to a syrup. The syrup was triturated with dichloromethane to induce crystallization. The crystals were filtered and dried to give 8.3 g (61%) of chromatographically pure product, useful as an intermediate compound.

EXAMPLE 4: 5-Amino-3-[ (5-tritylamino) entyl]pyrazole- -carbonitrile.

To a methanol solution (100 mL) of the product of Example 3 (7.0 g, 16 mmole) in an ice bath was added hydrazine- ono-hydrate (7.8 mL, 160 mmole) dropwise over is min. After stirring for 30 min in an ice bath, the solution was evaporated to dryness. The residue was suspended in cold methanol and filtered to give 7.1 g (100%) of 5-amino-3-[ (5-tritylamino)pentyl]pyrazole-4- carbonitrile, useful as an intermediate, after drying. An analytical sample was prepared by recrystallization from water. 3736

47

EXAMPLE 5 :

5-Amino-ι-(2-deoxy-3,5-di-o-toluoyl-.-D-erythropentofur- anosyl)-3-[(5- ritylamino)pentyl]pyrazole- -carbonitrile.

5 An ice cold solution of the carbonitrile from

Example 4 (3.5 g, 8 mmole) was treated with sodium hydride and stirred for 30 min at 0-4°C. l-Chloro-1,2- dideoxy-3,5-di-O-toluoylribofuranose was added and the solution stirred for 1 hr at 0-4°C. The solution was o poured into a saturated solution of sodium bicarbonate and extracted with dichloromethane. The organic layer was dried over sodium sulfate and evaporated to dryness. The residue was flash chromatographed on silica gel using toluene/ethyl acetate (5/1) as eluent. Two major 5products were isolated and identified as the N-l and N-2 isomers in 57% (3.6 g) and 20% (1.2 g) N-l and N-2 yields, respectively. Approximately 1 g of a mixture of N-l and N-2 isomers was also collected. Overall yield of glycosylated material was 5.8 g (92%). The N-l iso er, 05-amino-l-(2-deoxy-3,5-di-0-toluoyl-/_-D-erythropentofur- anosyl)-3-[(5-tritylamino)pentyl]pyrazole-4-carbonitrile, was used without further purification in Example 6.

EXAMPLE -£■_. l-(2-Deoxy-/?-D-erytI_ropentofuranosyl)-3-[5-(tritylamino)- pentyl]pyrazolo[3,4-d]pyrimidin-4-amine.

To a toluene (100 mL) solution of the pyrazole- 4-carbonitrile of Example 5 (3.5 g, 4.4 mmole) was added diethoxymethyl acetate (1.1 mL, 6.7 mmole). The solution was kept at 80-90°C for 5 hr and then evaporated to a syrup. The syrup was dissolved in dichloromethane (10 mL) and added to ice cold methanolic ammonia (100 mL) in a glass pressure bottle. After two days at RT the contents of the bottle were evaporated to dryness. The residue was dissolved in methanol and adjusted to pH 8 with freshly prepared sodium methoxide to complete the 3/03736

48 deprotection. After stirring overnight the solution was treated with Dowex®-50 H+ resin, filtered, and evaporated to dryness. The residue was chromatographed on silica gel using acetone/hexane (3/2) as eluent to give 2.0 g 5 (77%) of analytically pure product.

EXAMPLE 7:

1-(2-Deoxy-_/3-D-erythropentofuranosyl)-3-[5-(tritylamino)- pentyl] yrazolo[3 ,4-d.]pyri___idin-4-__mine 5'-monophosphate.

10

To an ice cold solution of the pyrazolo- pyrimidin-4-amine of Example 6 (250 mg, 0.43 mmole) in trimethyl phosphate (5 mL) was added phosphoryl chloride (50 μ ) and the solution was kept at 0-4°C. The reaction

15was monitored by reversed phase HPLC using a linear gradient from 0 to 100% acetonitrile in water over 25 min. After stirring for 5 hr, an additional aliquot of phosphoryl chloride (25 μL) was added and the solution was stirred another 30 min. The solution was poured into

200.IM ammonium bicarbonate and kept in the cold overnight. The solution was then extracted with ether and the aqueous layer evaporated to dryness. The residue was dissolved in water (5 mL) and purified by reversed phase HPLC using a__,-22_nm X 50cm C18 column. The column was

25equilibrated in water and eluted with a gradient of 0 to 100% acetonitrile over 20 min. Fractions containing the desired material were pooled and lyophilized to give 160 mg (56%) of chro atographically pure nucleotide.

0 EXAMPLE 8: l-(2-Deoxy-.-D-erythropentofuranosyl)-3-{5-[ (6-biotin- __mido)hexanamido]pentyl}pyrazolo[3,4-d]pyrimidin-4-amine 5*-monophosphate.

An ethanol solution (10 mL) of the nucleotide of Example 7, palladium hydroxide on carbon (50 mg) , and cyclohexadiene (1 mL) was refluxed for 3 days, filtered. 3/03736

49 and evaporated to dryness. The residue was washed with dichloromethane, dissolved in DMF (1.5 mL) containing triethylamine (100 mL) , and treated with N-hydroxy- succinimidyl biotinylaminocaproate (50 mg) . After 5 stirring overnight an additional amount of N-hydroxy- succinimidyl 6-biotinamidocaproate (50 mg) was added and the solution was stirred for 18 hr. The reaction mixture was evaporated to dryness and chromatographed following the procedure in Example 7. Fractions were pooled and 0 lyophilized to give 80 mg of chromatographically pure biotinamido-substituted nucleotide.

EXAMPLE 9:

1-(2-Deoxy-.-D-erythropentofuranosyl)-3-[5-(6-biotin- 5 amido)hexanamidopentyl]pyrazolo[3,4-d]pyri_αidin-4-amine 5•-triphosphate.

The monophosphate of Example 8 (80 mg, ca. 0.1_ mmole) was dissolved in DMF with the addition of 0triethylamine (14 μL) . Carbonyldiimidazole (81 mg, 0.5 mmole) was added and the. solution stirred at RT for 18 hr. The solution was treated with methanol (40 μL) , and after stirring for 30 min tributylammonium pyrophosphate (0-«_.5_S in_-0.__5_.mL-DMF) was added. After stirring for 24 5hr another aliquot of tributylammonium pyrophosphate was added and the solution was stirred overnight. The reaction mixture was evaporated to dryness and chromato¬ graphed following the procedure in Example 8. Two products were collected and were each separately treated ø ith cone, ammonium hydroxide (1 mL) for 18 hr at 55°C. UV and HPLC analysis indicated that both products were identical after ammonia treatment and were pooled and lyophilized to give 35.2 mg of nucleoside triphosphate. 03736

50

EXAMPLE 10: NICK-TRANSLATION REACTION

The triphosphate of Example 9 was incorporated into pHPV-16 using the nick tanslation protocol of Langer et al. (supra) . The probe prepared with the triphosphate of Example 9 was compared with probe prepared using commercially available bio-ll-dUTP (Sigma Chemical Co) . No significant differences could be observed in both a filter hybridization and in in situ smears.

More specifically, the procedure involved the following materials and steps:

Materials: DNase (ICN Biomedicals) - 4μg/mL

DNA polymerase 1 (U.S. Biochemicals) - 8

U/mL pHPV - 16 - 2.16 mg/mL which is a plasmid containing the genomic sequence of human papillomavirus type 16. 10X-DP - IM Tris,pH7.5(20mL) ; 0.5M DTT(80 mL) ; IM MgCl₂(2.8 L) ; H₂0(17mL) Nucleotides - Mix A - 2mM each dGTP, dCTP, TTP (Pharmacia)

Mix U - 2mM each dGTP, dCTP, dATP Bio-ll-dUTP - 1.0 mg/mL (BRL) Bio-12-dAPPTP - 1.0 mg/mL

Steps:

To an ice cold mixture of 10X-DP (4 mL) , pHPV-16 (2 mL) , nucleotide mix A (6 mL) , Bio-12-dAPPTP (2 ___L) , and H₂0 (20 mL) was added DNase (1 mL) and DNA polymerase 1 (2.4 mL) . The reaction mixture was incubated at 16°C for 1 hr. The procedure was repeated /03736

51 using Bio-ll-dUTP and nucleotide mix U in place of Bio- 12-dAPPTP (comprising the triphosphate of Example 9) and nucleotide mix A.

Nucleic acid was isolated by ethanol precipitation and hybridized to pHPV-16 slotted onto nitrocellulose. The hybridized biotinylated probe was visualized by a streptavidin-alkaline phosphatase conjugate with BCIP/NBT substrate. Probe prepared using either biotinylated nucleotide gave identical signals. The probes were also tested in an in situ format on cervical smears and showed no qualitative differences in signal and background.

EXAMPLE 11: 5-Amino-3-[(5-trityla_αino)pentyl]pyrazole-4-carboxamide.

Following the procedure of Example 2, except that cyanoacetamide is used instead of malononitrile, 5- (tritylamino)pentylhydroxymethylenecyanoacetamide is 0prepared from 6-(tritylamino)caproic acid. This is then treated with diazomethane to give the methoxy derivative, following the procedures of Example 3, which is then reacted with hydrazine monohydrate, as in Example 4, to (5-tritylamino)pentyl]pyrazole-4- carboxamide.

EXAMPLE 12:

4-Hydroxy-6-methylthio-3-[(5-tritylamino)pentyl]pyrazolo-

[3, -d] yrimidine.

The carboxamide from Example 11 is reacted with potassium ethyl xanthate and ethanol at an elevated temperature to give the potassium salt of 4-hydroxypyra- zolo[3,4-d]pyrimidine-6-thiol. This salt is then reacted ith iodomethane to give 4-hydroxy-6-methylthio-3-[(5- tritylamino)pentyl]pyrazolo[3,4-d]pyrimidine. 3/03736

52

EXAMPLE 13 :

1-(2-Deoxy-3-D-erythropentofuranosyl)-4-hydroxy-3-[5-

(tritylamino) entyl] yrazolo[3,4-d] yrimidin-6-amine.

5 Following the procedure of Example 5, the pyrazolopyrimidine of Example 12 is treated with sodium hydride and reacted with l-chloro-l,2-dideoxy-3,5-di-0- toluoylribofuranose. The resulting compound is reacted with MCPBA and with ethanolic ammonia, and the toluoyl oprotecting groups are removed to give the product.

EXAMPLE 14: l-(2-Deoxy-/--D-eryt_.ropentofuranosyl)-4-hydroxy-3-[5-(6- biotinamido) exanamidopentyl]pyrazolo[3,4-d] yrimidin-6- 5aX52ne 5'-monophosphate.

Following the procedure of Example 1 , the pyrazolopyrimidine of Example 13 is reacted with phosphoryl chloride to give the corresponding 5'-mono- 0phosphate.

Following the procedure of Example 8, the above 5'-monophosphate is reacted with palladium/carbon and cyclohexadiene, and the residue is reacted with N- hydrpxysucciniaidy1 biotinylaminocaproate to give l-(2- 5deoxy-/?-D-erythropentofuranosyl)-4-hydroxy-3-[5-(6- biotinamido)hexanamidopentyl]pyrazolo[3,4-d]pyrimidin-6- amine 5'-monophosphate.

EXAMPLE 15: ₀1-(2-Deoxy-?-D-erythropentofuranosyl)-4-hydroxy-3-[5-(6- biotinamido)hexanamidopentyl]pyrazolo[3 ,4-cl]pyrimi in-6- amine 5'-triphosphate.

Following the procedure of Example 9, the 5¹- ^nonophosphate of Example 14 is treated with carbonyldi- imidazole and then reacted with tributylammonium pyro¬ phosphate to give the corresponding 5'-triphosphate. 03736

53

EXAMPLE 16 :

1-(2-Deoxy-?-D-erythropentofuranosyl)-3-[5-(tritylamino)- pent l]pyrazolo[3, -d]pyrimidine-4-benzoylamine.

l-(2-Deoxy-/?-D-erythropentofuranosyl)-3-[5-

(tritylamino)pentyl]pyrazolo[3,4-d]pyrimidine-4-amine from Example 6 is reacted with benzoyl chloride and pyridine to give l-(2-deoxy-3,5-di-o-benzoyl-/?-D-erythro¬ pentofuranosyl)-3-[5-(tritylamino)pentyl]pyrazolo[3,4-d]- pyrimidine-4-dibenzoylamine. This is treated with aqueous sodium hydroxide to partially deprotect the compound, giving l-(2-deoxy-/?-D-erythropentofuranosyl)-3- [5-(tritylamino)pentyl]pyrazolo[3,4-d]pyrimidine-4- benzoylamine.

EXAMPLE 17: .

1-(2-Deoxy-?-D-erythropentofuranosyl)-3-[5-(trifluoro- acetamido)pentyl]pyrazolo[3,4-d]pyrimidine-4-benzoyl- amine.

Following the procedure of Example 8, the benzoylamine of Example 16 is treated with palladium hydroxide on carbon and then with trifluoroacetic anhydrid____to_-g___*/e l-(2-deoxy-?-D-erythropentofuranosyl)- 3-[5-(trifluoroacetamido)pentyl]pyrazolo[3,4-d]pyrimi- dine-4-benzoylamine.

EXAMPLE 18: l-(2-Deoxy-5-θ-dimethoxytrityl-/9-D-erythropentofurano- syl)-3-[5-(trifluoroacetamido) entyl]pyrazolo[3,4-d]- pyrimidine-4-benzoylamine 3'-θ-(N,N-diisoproρyl)phos¬ phoramidite cyanoethyl ester.

The compound of Example 17 is reacted with dimethoxytrityl chloride and pyridine to give the corresponding 5'-dimethoxytrityl compound. This compound is then reacted with cyanoethyl chloro-N,N-diisopropyl- 3/03736

54 phosphoramidite (according to the method of Sinha et al., Nucleic Acids Res.. 12.- 539 (1984)) to give the 3*-0- activated nucleoside.

5 EXAMPLE 19:

5-(4-Phthalimidobut-l-yn-l-yl)-2*-deoxyuridine.

5-Iodo-2*-deoxyuridine (354 mg, 1 mmol) was dissolved in 10 mL of dimethylformamide. Cuprous iodide 10 (76 mg, 0.4 mmol), tetrakis(triphenylphosphine)palla- dium(O) (230 mg, 0.2 mmol), and triethylamine (200 mg, 2.0 mmol) were added. 4-Phthalimidobut-l-yne (300 mg, 1.5 mmol) was added all at once and the reaction kept at 60°C for three hours. The clear yellow reaction was then 5evaporated and methylene chloride was added. Scratching of the flask induced crystallization of nearly all of the product which was filtered and recrystallized from 95% ethanol to give 335 mg (78%) of title compound as fine,, feathery needles. 0

EXAMPLE 20:

5-(4-Phthalimidobut-l-yl)-2'-deoxyuridine.

-L____α___Gra_n of deoxyridine from Example 19 was 5dissolved in 95% EtOH and about 3 g of neutral Raney nickel was added. After 48 hours, the catalyst was removed by cautious filtration and the filtrate was evaporated to a solid which was recrystallized from methanol-water to give 960 mg (97%) of the title o compound.

EXAMPLE 21:

5-(3-Iodoacetamidopropyl)-2*-deoxyuridine.

5-(3-Trifluoroacetamidoprop-l-yl)-2'-deoxy¬ uridine (0.3 mmol) is treated with ammonia and then with N-hydroxysuccinimidyl α-iodoacetate (0.5 mmol). The /03736

55 reaction mixture is evaporated to dryness and purified by chromatography to give 5-(3-iodoacetamidopropyl)-2'- deoxyuridine.

EXAMPLE 22:

5-(4-(4-Bromobutyramido)butyl)-2*-deoxyuridine.

Following the procedure of Example 21, 5-(4- phthalimidobut-1-yl)-2'-deoxyuridine, from Example 20, is 0treated with ammonia and then with N-hydroxysuccinimidyl 4-bromobutyrate to give 5-(4-(4-bromobutyramido)butyl)- 2'-deoxyuridine.

Preparation of Synthetic Oligonucleotides 5

EXAMPLE 23:

Phosphoramidite Preparation and DNA Synthesis.

Nucleosides were 5'-dimethoxytritylated, following known procedures, to give around 85% yield, and the 3 *-phosphoramidite was made using diisopropylamino /?- cyanoethylchlorophosphite (as described in "Oligonucleotide Synthesis: A Practical Approach", supra) with di sopr.op._u.ethylamine in methylene chloride. The phosphoramidite was made into a 0.2N solution in acetonitrile and placed on the automated DNA synthesizer. Incorporation of these new and modified phosphoramidites gave incorporation similar to ordinary phosphoramidites (97-99% as judged by assay of the trityl color released ₀by UV.)

Oligonucleotides were removed from the DNA synthesizer in tritylated form and deblocked using 30% ammonia at 55°C for 6 hours. Ten μL of 0.5M sodium bicarbonate was added to prevent acidification during --concentration. The oligonucleotide was evaporated to dryness under vacuum and redissolved in 1.0 mL water. The oligonucleotides were purified by HPLC using 15-55% /03736

56 acetonitrile in 0.1N triethylammonium acetate over 20 minutes. Unsubstituted oligonucleotides came off at 10 minutes; amino derivatives took 11-12 minutes. The desired oligonucleotide was collected and evaporated to dryness, then it was redissolved in 80% aqueous acetic acid for 90 minutes to remove the trityl group. Desalting was accomplished with a G25 Sephadex column and appropriate fractions were taken. The fractions were concentrated, brought to a specific volume, dilution Oreading taken to ascertain overall yield and an analytical HPLC done to assure purity. Oligonucleotides were frozen at -20°C until use.

Following the above procedures, the nucleoside 5-(3-trifluoroacetamidoprop-1-yl)-2'-deoxyuridine was converted to the 5'-O-dimethoxytrityl-3 '-(N,N- diisopropyl)phosphoramidite cyanoethyl ester derivative. This was added to a DNA synthesizer and the following 14- mer oligonucleotide sequence was prepared: 3 '-CT TCC U*TG TAG GTC-5' 0where U¹ is 5-(3-aminoprop-l-yl)-2 '-deoxyuridine (oligo A).

In the same manner, 5-(4-phthalimidobut-l-yl)- 2 '-deoxyuridine was converted to the 5'-O-dimethoxy- trityl-a!___(___-N=diisopropyl)phosphoramidite cyanoethyl ester derivative and added to a DNA synthesizer to prepare the above 14-mer oligonucleotide sequence where U¹ is 5-(4-aminobut-l-yl)-2 '-deoxyuridine (oligo C) .

A corresponding 14-mer oligonucleotide was also prepared where U¹ is the unmodified deoxyuridine.

EXAMPLE 24:

Derivatization of Oligonucleotides.

In general, to add the crosslinking arm to an aminoalkyloligonucleotide, a solution of 10 μg of the aminoalkyloligonucleotide and a 100X molar excess of n- hydroxysuccinimide haloacylate such as α-haloacetate or 03736

57

4-halobutyrate in 10 μL of 0.1 M borate buffer, pH 8.5, was incubated at ambient temperature for 30 min. in the dark. The entire reaction was passed over a NAP-10 column equilibrated with and eluted with distilled water. Appropriate fractions based on UV absorbance were combined and the concentration was determined spectro- photometrically.

Introduction of the haloacyl moiety was examined by HPLC. A Zorbax® oligonucleotide column (Dupont) eluted with a 20 minute gradient of 60% to 80% B composed of: A (20% acetonitrile:80% 0.02 N NaH₂PO and B (1.2 N NaCl in 20% acetonitrile:80% 0.02 N NaH₂PO . The presence of a reactive α-haloacyl moiety was indi¬ cated by return of the retention time of the α-haloacyl- 5amidoalkyl oligonucleotide to the corresponding amino- alkyl oligonucleotide after exposure to IN cysteamine. Introduction of cysteamine created equivalent charge patterns between the aminoalkyl oligonucleotide and the . α-haloacylamido oligonucleotide. 0 Following the above procedure, the 14-mer oligonucleotide:

3--CT TCC UTG TAG GTC-5' where U¹ is 5-(3-aminoprop-1-yl)-2'-deoxyuridine (oligo A, Example-2^-_τ— as-, reacted with n-hydroxysuccinimide α-iodoacetate to give the above 14-mer oligonucleotide where U¹ is 5-(3-iodoacetamidoprop-1-yl)-2'-deoxyuridine (oligo B) .

Oligo A and oligo B, as well as the above 14- mer where U¹ is the unmodified deoxyuridine were resolved in the Zorbax column, all of identical sequence, with the following retention times: unmodified 14-mer, 9.31 min; aminopropyl 14-mer (oligo A), 7.36 min; and iodoacet- amidopropyl 14-mer (oligo B) , 10.09 min.

In the same manner, the aminopropyl 14-mer (oligo A) was reacted with N-hydroxysuccinimide 4-bromo- butyrate to give the 14-mer where U¹ is 5-(3-(4-bromo- butyramido)prop-l-yl)-2'-deoxyuridine. /03736

58

Also, the a inobutyl 14-mer (oligo C, Example 23) was reacted with either N-hydroxysuccinimide α-iodo- acetate or N-hydroxysuccinimide 4-bromobutyrate to give the 14-mer where U¹ is 5-(4-iodoacetamidobut-l-yl)-2•- c deoxyuridine or 5-(4-(4-bromobutyramido)but-l-yl)-2'- deoxyuridine, respectively.

Assays EXAMPLE 25: _Q Assay of Crosslinking Reaction.

The reaction of crosslinking a DNA probe to a target nucleic acid sequence contained 1 μg of haloacyl- camidoalkyl probe and 10 ng of ³²P-labeled cordycepin- tailed target in 200 μL of 0.1 M Tris, pH 8.0, and 0.9 M NaCl incubated at 20° or 30°C. Aliquots were removed at 24- or 72-hour intervals and diluted in 20 μL of 10 mM cysteamine to quench the haloacylamido group. These 0solutions were stored at RT, and 1 μL was used for analysis by denaturing polyacrylamide gel electrophoresis (PAGE) .

Following the above procedure, two model oligonucleotide sequences were utilized to evaluate the crosslin__age_-p-.tential of the modified probe to its complement. The sequences, derived from human papilloma- virus (HPV) or human cytomegalovirus (CMV) , are shown below:

HPV System:

5 10 15 20 25 30

I I I I I I

I I I I I I

Target : 5 * -AGA CAG CAC AGA ATT CGA AGG AAC ATC CAG-3 '

Probe : 3 ' -CT TCC UTG TAG GTC-5 « 03736

59 CMV System:

5 10 15 20

I I I I

Target: 5^»-ACC GTC CTT GAC ACG ATG GAC TCC-3' Probe: 3'-GAA CTG TGC UAC CTC-5'

U = 5-[3-(α-iodoacetamido)- or 3-(4-bromo- butyramido)-propyl]-2*-deoxyuridine, or

U = 5-[3-(_t-iodoacetamido)- or 4-(4- bromobutyramido)-butyl]-2'-deoxyuridine.

The target for HPV is a 30-mer, and for CMV it is a 24-mer. The crosslinking probes were a 14-mer for HPV and two 15-mers for CMV. Each probe contained a single modified deoxyuridine designated as U in the sequences above.

Results of the reaction of HPV target with a limiting amount of crosslinking probe containing a

5-(3-iodoacetamidopropyl) sidearm are shown in Figure 2. Analysis of the cleavage pattern on a denaturing PAGE gel showed the loss of the crosslinked hybrid with the con¬ comitant appearance of a discrete low molecular weight (band. The intensity of this band was dependent upon the extent of crosslinkage in the initial reaction. The localization of signal into two discrete bands on the gel strongly argues that no non-seguence-directed alkylation of either target or probe strands had occurred (including intramolecular probe alkylation) . Comparison to an authentic 15-mer run in an adjacent lane suggested that the major cleaved fragment is a 9-mer. Upon close examination of the original auto- radiogram, a slower moving band of very weak intensity was visible. This pattern would be consistent with major alkylation at G-21 and minor alkylation at G-20. An examination of a Dreiding model of the crosslinkable HPV hybrid shows that the 5-(3-iodoacetamidopropyl) sidearm can contact the G-21 residue of the target strand with only minor distortion of the helix. 3/03736

60

If alkylation occurs predominately at a guano- sine on the target strand located two units on the 5' side of the modified-deoxyuridine base pair, the CMV sequence should not react. This result was in fact observed. The absence of reaction with CMV further supports the specificity of crosslinking scheme of the invention.

EXAMPLE 26:

Time and Temperature Dependence.

Time and temperature dependence studies were carried out with the HPV system of Example 25 where U is 5-(3-iodoacetamidoprop-1-yl)-2'-deoxyuridine. The target was ³²P-labeled by cordycepin tailing with terminal transferase (Maniatis et al., "Molecular Cloning - A Laboratory Manual", Cold Spring Harbor Laboratory, 1982, p. 239) and incubated with excess probe in a pH 8.0 Tris buffer at either 20° or 30°C. Aliquots were removed after 0, 24, or 72 hours incubation, quenched with an equivalent volume of 10 mM mercaptoethylamine (which reacts with the iodoacetamide) , and stored at RT for subsequent analysis by denaturing or non-denaturing PAGE.

._ —Gross_inkage of the hybrid, which was monitored by denaturing PAGE, was evident for the 24 and 72 hour time points at both temperatures (see Figure 3) . The amount of crosslinked hybrid increased with both tempera¬ ture and time. Approximately 20% of the hybrid was crosslinked after 72 hours incubation at 30°C. Separate experiments at a range of temperatures indicated that the half-life for crosslinking at 37°C is approximately 2 days, and that the reaction is complete after 24 hours at 58°C. This time-dependent reaction implies that the iodoacetamido moiety does not hydrolyze or react with the buffer. The increased reaction rate at higher temperature indicates that the hybrid is 3736

61 maintained, and subsequently the rate of alkylation shows the expected increase with temperature.

EXAMPLE 27:

Site Specificity of Alkylation.

To elucidate the site specificity of alkyla-• tion, the crosslinked HPV hybrid of Example 25 (where u is 5-(3-iodoacetamidoprop-l-yl)-2'-deoxyuridine) was subjected to a 10% piperidine solution at 90°C for 60 minutes. As shown by Maxam et al. (Proc. Natl. Acad. Sci. USA. 2__.^:-560 (1977), this treatment quantitatively cleaves the target strand 3'- to the site of alkylation. The resulting data indicated that the alkylation of the second guanine above the crosslinker-modified base pair (i.e., the guanine above the target base) was the exclusive action observed, indicating that the cross¬ linking reaction in the HPV model system is remarkably * specific.

93/03736

62

EXAMPLE 28 :

Chemically-Modified Crosslinkable ODNs.

Psoralen-modified, photo-crosslinkable ODNs 5 were prepared using a photochemical procedure (REF??) . Briefly, ODN was hybridized to a complementary DNA sequence and irradiated at 360 nm in the presence of 4 '-hydroxymethy-4,5',8-trimethylpsoralen (HMT) , thereby forming a sequence-specific, interstrand crosslink at a _1Q unique 5'-TpA-3' sequence. Following partial photoreversal of the crosslink by brief exposure to 260 nm light, the HMT furan-side onoadducted ODN was isolated by denaturing PAGE.

Alternatively, the attachment of a psoralen ₁.. derivative to an ODN may be accomplished by chemical conjugation. These synthetic schemes may be used to obtain ODNs that are coupled to a 4,5' ,8-trimethyl- psoralen (TMP) via a linker arm that spans the 5' terminus of the ODN and the 4 • position of the psoralen. 0 In one approach, a modified TMP is reacted with a suitably 5'-activated, deblocked ODN (B.L. Lee et al. , Biochem. 27:3197-3202, 1988) . In a second approach, a TMP phosphoramidite linker compound is added in the last cycle- of-solid—phase synthesis (U. Pieles and U. 5 Englisch, Nucl. Acids Res. 17:285-99, 1989). Similar adducts may be prepared using a thiolated derivative of 4 '-aminomethyl-4,5' ,8-trimethylpsoralen (AMT) (J. Teare and P. Wollenzien, Nucl. Acids Res. 17:3359-72, 1989). In a further alternative, a TMP-modified deoxyadenosine phosphoramidite may be inserted at any position in an ODN , , during solid phase synthesis (U, Pieles et al., Nucl.

Acids Res. 17:8967-78, 1989). In this synthon, TMP is modified at its 4 ' position with linker which is attached to the C8 position of deoxyadenosine. {Analogous w/ chemical XLers of app'n) λ 93/03736

63

EXAMPLE 29 :

Use of Chemically Modified ODNs to Target Double Stranded

DNA.

A prototype, psoralenated ODN that can efficiently crosslink to both Watson and Crick strands when complexed to homologous DNA in a recA-stabilized triple strand is employed. The ODN 5'-modified with a psoralenated tail is synthesized so as to optimize photo- o crosslinkage within a 5•-TpA-3' sequence flanking the recA-stabilized triplex. Briefly, a psoralenated ODN in the presence of a recombination enzyme is used to introduce a sequence-specific crosslink the complementary site in a double stranded DNA target. The ODN, which 5 Hoogstein base pairs to the DNA target, is modified with a 5-hydroxypsoralen moiety or a 4,5' ,8-trimethylpsoralen moiety at its 5' terminus. By linking the psoralen through the 5' position, the furocourmain is able to readily intercalate into a double stranded 5'-TpA-3' 0 sequence immediately adjacent to the triple strand. Both the furan and pyrone rings are properly positioned to photoreact with thymidines. The length of the psoralenated tail may be altered to obtain optimal -• inter.calation-"-cτf-the psoralen into the duplex-triple 5 strand junction.

Claims

/03736

64

WHAT IS CLAIMED IS:

1. A method for altering gene function, comprising: attaching a crosslinking agent to an oligonucleotide (ODN) that is homologous to one strand and complementary to the other strand of target double stranded DNA, wherein the target double stranded DNA corresponds to a structural gene, a regulatory sequence that affects gene function, or a portion thereof, thereby forming a crosslinking ODN; contacting the crosslinking ODN and a recombination enzyme, thereby forming a nucleoprotein filament; combining the nucleoprotein filament and the target double stranded DNA for a time period sufficient for formation of a triple strand complex; and crosslinking the triple strand complex, thereby altering gene function.

2. The method of Claim 1 wherein the -is a nucleoside crosslinking agent.

3. The method of Claim 2 wherein the nucleoside crosslinking agent is represented by the following formula:

R. - B - (CH₂)_q- (Y)_r - (CH₂)_m- A^» (I^«)

wherein,

R_ is hydrogen, or a sugar moiety or analog thereof optionally substituted at its 3' or its 5' position with a phosphorus derivative attached to the sugar moiety by an oxygen and including groups Q_, Q₂ and 3736

65

Q₃, or with a reactive precursor thereof suitable for nucleotide bond formation;

Q_ is hydroxy, phosphate or diphosphate; Q₂ is =0 or =S; Q₃ is CH₂-R^», S-R', O-R', or N-R'R"; each of R' and R" is independently hydrogen or C_.₆alkyl;

B is a nucleic acid base or analog thereof that is a component of an oligonucleotide; Y is a functional linking group; each of m and q is independently 0 to 8, inclusive; r is 0 or 1; and A' is a leaving group.

4. A chemical crosslinking anti-gene nucleoprotein filament comprising: a chemical crosslinking agent covalently linked to an oligonucleotide (ODN) that is homologous to one strand and complementary to the other strand of target double stranded DNA that corresponds to a structural gene, a regulatory sequence that affects gene function, or a portion thereof; and -a—recombination enzyme non-covalently associated with the ODN.

5. The chemical crosslinking anti-gene nucleoprotein filament of Claim 4 wherein the chemical crosslinking agent is a nucleoside crosslinking agent.

6. The chemical crosslinking anti-gene nucleoprotein filament of Claim 5 wherein the nucleoside crosslinking agent is represented by the following formula:

R_. - B - (CH₂)_q- (Y)_r - (CH₂)₀- A' (I') - 3736

66 wherein,

R_x is hydrogen, or a sugar moiety or analog thereof optionally substituted at its 3 ' or its 5' position with a phosphorus derivative attached to the sugar moiety by an oxygen and including groups Q_, Q₂ and Q₃, or with a reactive precursor thereof suitable for nucleotide bond formation;

Q_ is hydroxy, phosphate or diphosphate;

Q₂ is =0 or =S;

Q₃ is CH₂-R^!, S-R¹, 0-R', or N-R'R"; each of R' and R" is independently hydrogen or C_.₆alkyl;

B is a nucleic acid base or analog thereof that is a component of an oligonucleotide;

Y is a functional linking group; each of m and q is independently 0 to 8, inclusive; r is 0 or 1; and

A' is a leaving group.