WO2012102616A1

WO2012102616A1 - Heterodirectional polynucleotides

Info

Publication number: WO2012102616A1
Application number: PCT/NL2012/050046
Authority: WO
Inventors: Stefan Matysiak; Klaus Hellmuth
Original assignee: Flexgen B.V.
Priority date: 2011-01-28
Filing date: 2012-01-27
Publication date: 2012-08-02

Abstract

The current invention now allows the circularization of a polynucleotide via complementary motifs within the polynucleotide that can hybridize with each other, e.g. via classic Watson- Crick base pairing. This is achieved by having a heterodirectional design, i.e. a polynucleotide comprising at least one intramolecular switch, a 3'-3' or 5'-5' covalent linkage, such that the polynucleotide comprises at least two polynucleotide segments in opposite direction.

Description

Title: Heterodirectional polynucleotides

Field of the invention: The current invention relates to chemistry. In particular to polynucleotides and to methods of synthesizing polynucleotides.

Background

Polynucleotide synthesis is well known in the art. It is of particular interest to circularize polynucleotides, as this may stabilize the structure of the polynucleotide. It is also of interest to circularize polynucleotides to increase the biological stability of polynucleotides.

Circularization can be achieved by subjecting a polynucleotide to a complex chemical reaction, such that the 5' and 3' end of the polynucleotide are covalently joined.

For example a Cu(l)-catalyzed alkyne-azide cycloaddition (CuAAC) reaction applied to 5'- alkyne-3'-azide-labeled hairpin loop oligonucleotides proceeds in high yield under mild conditions in as little as 5 min. However, so far such chemical cyclisation reactions are limited to very short sequences. (El-Sagheer and Brown, Curr Protoc Nucleic Acid Chem, 2008, Dec;Chapter 4:Unit 4.33, El-Sagheer and Brown, Int. J. Pept. Res.Ther. 2008, Angew Chem Int Ed Engl. 2002 Oct 4;41 (19):3704-7; 3523. Smietana M, Kool ET).

Alternatively, circularization may be achieved by splicing or via ligation. For example, a synthesized polynucleotide may have both their 5' and 3' ends hybridized with a short second polynucleotide such that a double stranded sequence is formed having a nick, which nick may be repaired enzymatically the "nick", thus forming a circular single stranded poly-nucleotide with a hybridized second polynucleotide, upon removal of the second polynucleotide, a single stranded circular polynucleotide is obtained. Several similar alternative methods utilizing enzymatic ligation of a nicked double stranded DNA structure have been used. (Giusto and King, J Bio Chem, 2004:Vol. 279, No. 45, pp. 46483-9, US2005/0176940A1 , Umekage and Kikuchi, J Biotechnol, 2009: 139(4): 265-72, Hartig and Kool, NAR 2004:Vol. 32, No. 19 e152).

These methods have the drawback that they all require complex additional steps in order to obtain a polynucleotide that has a circularized structure.

Summary

The current invention now allows the circularization of a polynucleotide via complementary motifs within the polynucleotide that can hybridize with each other, e.g. via classic Watson- Crick base pairing. This is achieved by having a heterodirectional design, i.e. a

polynucleotide comprising at least one intramolecular switch, a 3'-3' or 5'-5' covalent linkage, such that the polynucleotide comprises at least two polynucleotide segments in opposite direction. The heterodirectional polynucleotide can base pair, thus forming a circular structure similar to circular structures known in the art, but without requiring covalent ligation. Such heterodirectional polynucleotides can be synthesized e.g. via standard solid phase synthesis and do not require further modifications. Moreover, the heterodirectional polynucleotides with the two complementary regions show resistance against enzymatic degradation.

Figure 1. A schematic of a reaction cycle for the chemical synthesis in the 3' to 5' direction. Figure 2. Different monomers for both the 3' to 5' chemical synthesis and the 3' to 5' chemical synthesis. A) and B) show thymidine-5'-NPPOC-3'cyanoethoxy-amidite and thymidine-3'-NPPOC-5'cyanoethoxy-amidite respectively, C) and D) show the corresponding DMT analogues respectively. In E) linker molecule DMT-hexane-Diol phosphoramidite is shown with a DMT group suitable for the monomers shown in C) and D), alternatively at the position of the DMT group an NPPOC may be selected in case synthesis is carried out with the monomers of A) and B).

Figure 3. Secondary structures of polynucleotides. A) A typical hairpin structure of a unidirectional polynucleotide wherein two sections of the polynucleotide basepair. The structure has a 5' end and a 3' end. B) Heterodirectional polynucleotide with one directional switch, a 3'-3' covalent linkage or 5'-5' covalent linkage. The polynucleotide has two 3'-ends or two 5' ends and the secondary structure forms a circular structure very different from the hairpin structure of A).

Figure 4. Cleavable linkers. A cleavable linker can be introduced as a succinyl- phosphoamidite (A) with standard coupling chemistry, or as a succinate (B) via peptide coupling chemistry, at the beginning of the synthesis. Examples of cleavable linkers attached to a solid support are shown in C.

Figure 5. Heterodirectional polynucleotides. A) A heterodirectional polynucleotide is depicted with two complementary regions at both 5'-ends and having a 3'-3' covalent linkage in between the two complementary regions. B) A heterodirectional polynucleotide is depicted with one complementary region at one 5'-end and having a 3'-3' covalent linkage in between the two complementary regions. The design in A) and B) may also be reversed, i.e. having a 5'-5' covalent linkage and having two 3'-ends.

Figure 6. Heterodirectional polynucleotide with primer regions. (A) A heterodirectional polynucleotide is depicted with two complementary regions at both 5'-ends and having a 3'- 3' covalent linkage in between the two complementary regions. The heterodirectional polynucleotide also comprises primer binding sequences. One of the primer binding sequences as depicted overlaps (in part) with the hybdridised regions, but may also be designed such that it does not overlap. With a PCR reaction, the variable region may be amplified, e.g. for sequencing. The first primer (targeting 1) will during the first round of amplification generate a template for the second primer (2). Hence, after the first round of amplification in a PCR reaction, a double stranded template is obtained that is amplified in the subsequent rounds of the PCR reaction. (B) A heterodirectional polynucleotide is depicted with one complementary region at one 5'-end and having a 3'-3' covalent linkage in between the two complementary regions. The polynucleotide section that is outside of the circularized structure has two primer binding sequences, with a DNA barcode in between. With a PCR reaction, the DNA barcode may be amplified, e.g. for sequencing, such that the barcode can be determined and the heterodirectional polynucleotide identified. This may for instance be useful when the variable region does not allow amplification or sequencing. The first primer (targeting 1) will during the first round of amplification generate a template for the second primer (2). Hence, after the first round of amplification in a PCR reaction, a double stranded template with the DNA barcode is obtained that is amplified in the subsequent rounds of the PCR reaction.

Figure 7. Mobility shift assays. A) depicts a typical hairpin structure; B) depicts a circularized heterodirectional polynucleotide structure. C) Gel mobility shift assay for the linear, hairpin and circularized heterodirectional nucleotides as listed in tables 1 and 2.

Definitions

"Polynucleotides" or "oligonucleotides" are well known in the art. Polynucleotides according to the invention may include any polymer or oligomer of pyrimidine and purine bases, preferably cytosine, thymine, and uracil, and adenine and guanine, respectively (See Albert L. Lehninger, Principles of Biochemistry, at 793-800 (Worth Pub. 1982) which is herein incorporated by reference in its entirety for all purposes. The present invention contemplates any deoxyribonucleotide, ribonucleotide or peptide nucleic acid component, and any chemical variants thereof, such as methylated, hydroxymethylated or glycosylated forms of these bases, and the like. The polynucleotides, to which alternatively may be referred to as oligonucleotides, may be heterogeneous or homogenous in composition, and may be chemically synthesized. The polynucleotides may be DNA or RNA, or a mixture thereof, and may exist permanently or transitionally in single-stranded or double-stranded form, including homoduplex, heteroduplex, and hybrid states. A nucleotide is composed of a nucleobase (nitrogenous base), a five-carbon sugar (either ribose or 2'-deoxyribose), and one phosphate group. Ribonucleotides are nucleotides where the sugar is ribose, and deoxyribonucleotides contain the sugar deoxyribose. Nucleotides can contain either a purine or a pyrimidine base. Nucleotide analogues (or analogous nucleotides) are compounds structurally similar (analogous) to naturally occurring nucleotides. An analogue may have any of the nucleobase, five-carbon sugar and/or phosphate group altered. Analogue nucleobases may confer, among other things, different base pairing and base stacking properties. Examples include universal bases, which can pair with all four canonical bases, and phosphate-sugar backbone analogues such as peptide nucleotides, morpholino and locked nucleotides, as well as glycol nucleotides (GNAs) and threose nucleotide acid.

Additional examples of backbone modified nucleotides are phosphorothioates,

phosphorodithioates or boranophosphates. Each of these is distinguished from naturally occurring DNA or RNA by changes to the backbone of the molecule. Examples of deoxyribose modifications are, but are not limited to, 2'-carbamates (Curr. Prot. in Nucleic Acid Chem. 2003, 4.21.1 - 4.21.26) and 2'-0-alkynes such as 2'-0-propargyl or the corresponding dibenzocycloctine version (Org. Lett. 2010 Dec 3;12(23):5486-9. van Delft P, Meeuwenoord NJ, Hoogendoorn S, Dinkelaar J, Overkleeft HS, van der Marel

GA, Filippov DV.) 2'O-alkynes allow post-synthesis modification via "Click-Chemistry" (Chem Soc Rev. 2010 Apr; 39(4): 1388-405. El-Sagheer AH, Brown T ). The conformation of nucleotides can be either the D-conformation or the L-conformation. Natural DNA or RNA is in the D-conformation. By using nucleotides of the L-conformation, polynucleotides can be made having the same physical characteristics in terms of solubility, hybridization stability and selectivity as the D-conformation. When base paired, in the L-conformation a left-helical double-helix is formed. Nucleotides or polynucleotides in the L-conformation may also be referred to as "mirror image" nucleotides or polynucleotides.

The term "complementary" is defined herein as nucleotides of a polynucleotide region that bind to another polynucleotide region through hydrogen bonds, i.e. nucleotides that are capable of base pairing. Nucleotides that can form base pairs, that are complementary to one another, are e.g. cytosine and guanine, thymine and adenine, adenine and uracil, guanine and uracil. Nucleotide analogues can also form base pairs, e.g. nucleotide analogues can base pair and may even increase the thermal stability of a base pair

(stronger binding) such as is the case for PNA (peptide nucleic acid) or LNA (locked nucleic acid). Complementarity may be expressed by the number of nucleotides of the range over which the polynucleotide region is base paired, starting from the first nucleotide that base pairs and ending at the last nucleotide that base pairs. Base pairing is not required throughout the complete polynucleotide region, i.e. not all nucleotides of the complementary regions have to base pair. Complementarity may thus also be expressed in the percentage of base pairing. The percentage of complementarity is calculated over the length of the complementary region, starting from the first base-paired nucleotide and ending at the last base-paired nucleotide. For example, a polynucleotide may comprise a region of which at least 8 adjacent nucleotides are at least 75% complementary to 8 adjacent nucleotides of a complementary region of the polynucleotide. As long as of 8 adjacent nucleotides of the region of the polynucleotide at least 6 nucleotides can base-pair with 8 adjacent nucleotides of the complementary region and the first and last of the 8 adjacent nucleotides is base paired, such a polynucleotide region has at least 75% complementary.

Moreover, the degree of complementarity that provides binding between the two polynucleotide regions is dependent upon the conditions under which that binding occurs. It is well known that binding, i.e. hybridization, between polynucleotides depends on factors besides the degree of mismatch between two sequences. Such factors include e.g. the GC content of the region, temperature, ionic strength, the presence of chaotropic reagents such as formamide and types of counter ions present. The effect of these conditions upon binding is known to one skilled in the art. Furthermore, conditions are frequently determined by the circumstances of use. For example, when a heterodirectional oligonucleotide according to the invention is made for use in vivo, no formamide will be present and the ionic strength, types of counter ions, and temperature correspond to physiological conditions. Binding conditions can be manipulated in vitro to optimize the utility of the present oligonucleotides. A thorough treatment of the qualitative and quantitative considerations involved in establishing binding conditions that allow one skilled in the art to design appropriate polynucleotide regions for use under the desired conditions may be provided by Beltz et al., 1983, Methods Enzymol. 100:266-285 and by Sambrook et al. 2000, Molecular Cloning 3^rd edition.

As used herein "binding" means that two polynucleotide regions are hybridized to each other. Binding can be detected by either physical or functional properties of the bound polynucleotide regions. Binding between two polynucleotide regions can be detected by any procedure known to one skilled in the art, including both functional or physical binding assays. Physical methods of detecting the binding of complementary regions of

polynucleotides are well known in the art, and include such methods as DNase I or chemical footprinting, gel shift and affinity cleavage assays, and light absorption detection

procedures. For example, a method that is widely used, because it is simple and reliable, involves observing a change in light absorption of a solution containing a heterodirectional polynucleotide at 220 to 300 nm as the temperature is slowly increased. Detailed description of the invention

The current invention relates to a heterodirectional polynucleotide comprising at least one 3'- 3' covalent linkage or 5'-5' covalent linkage, and comprising a first complementary region and a second complementary region that are complementary to each other wherein the first and second region are separated by the at least one 3'-3' covalent linkage or 5'-5' covalent linkage. A heterodirectional polynucleotide is a polynucleotide in which during the synthesis, at least once, the direction of synthesis has been reversed, as is outlined below in more detail. A heterodirectional polynucleotide comprises at least two polynucleotide sections of two different directions, i.e. 5'-3' and 3'-5', which are linked by a 3'-3' covalent linkage or a 5'-5' linkage. For example, a heterodirectional polynucleotide may be represent by the following formula, 5'-(N)n-3'-3'-(N)n-5 or 3'-(N)n-5'-5'-(N)n-3', wherein N represents any monomer, which may for example be A, C, T and G, representing DNA monomeric units comprising adenosine, cytosine, guanidine or thymidine, but other monomers, e.g. natural or analogous nucleotides are also possible and combinations thereof, and n representing the number of nucleotides having a value of at least 1. The number of nucleotides (n) for each section may for example range from 10 nucleotides to about 200 nucleotides. The number of nucleotides for each section (n) may be at least 10, 15, 20, 30, 40, 50, 60, or 70 nucleotides. The number of nucleotides for each section (n) may be at most 10, 15, 20, 30, 40, 50, 60, 70 nucleotides. A heterodirectional oligonucleotide may comprise more than two of such sections, e.g. 5'-(N)n-3'-3'-(N)n-5'-5'-(N)n-3' comprises three sections. The number of nucleotides for each section may be independently selected. Hence, the number of nucleotides for each section, the number of sections being at least 2, may be the same or different.

A 3'-3' or 5'-5' covalent linkage refers to the intramolecular switch between the at least two polynucleotide sections of two different directions of the heterodirectional polynucleotide. Using solid-phase synthesis and the phosphoramidite chemistry developed by Curuthers ( Beaucage, S.L.; Caruthers M.H. {Tetrahedron Letters 22: 1859-1862, 1981) an intramolecular phosphodiester (PD) can be formed.

By switching the nucleotides used in the synthesis, switching from nucleotides suitable for a 5'-3' direction to nucleotides suitable for a 3'-5' synthesis (or the reverse), a 3'- 3' covalent linkage consisting of a phosphodiester covalently links the two 3'-ends.

Alternatively, a linker molecule may also be used to reverse the synthesis direction. Instead of switching during synthesis switching from nucleotides suitable for a 5'-3' direction to nucleotides suitable for a 3'-5' synthesis, a linker molecule is used as an intermediate. This linker molecule has compatible chemistry with the nucleotides used for the polynucleotide synthesis. For example, a linker molecule may have a phosphoramidite group (PA) and a protecting group (PG). The linker molecule may thus be defined as PA-X-PG. Hence, when this linker molecule is used in a heterodirectional polynucleotide, the linker molecule will be part of the "3'-3' or 5'-5' covalent linkage". An example of a linker molecule may be for instance DMT-hexane-Diol phosphoramidite (figure 2E). The 3' end of a polynucleotide is covalently linked to the linker molecule via a phosphodiester bridge, and the 3' end of the other polynucleotide is also covalently linked to the linker molecule via a phosphodiester bridge The resulting polynucleotide may be represented by 5'-N-3'- phosphodiester -X - phosphodiester -3'-N-5'. In the case of the linker DMT-hexane-Diol phosphoramidite, the resulting polynucleotide may be represented by 5'-N-3'- phosphodiester -C₁₂ - phosphodiester -3'-N-5'. Hence, a "3'-3' or 5'-5' covalent linkage is understood to comprise not only a phosphodiester, but can also comprise other types of covalent linkages, as long as two polynucleotide regions are covalently conjugated to each other via their 3' ends, or via their 5' ends, such a intramolecular covalent linkage may be contemplated.

The complementary regions that are part of different polynucleotide sections can bind to each other to form a circularized structure, e.g. such as depicted in figure 3 B. The heterodirectional polynucleotide comprises at least two regions that are complementary to each other such that these regions bind. The two complementary regions being separated by at least one covalent linkage means that the linkage may be directly adjacent to the complementary region but doesn't have to be. As long as in between the two

complementary regions that are part of the same polynucleotide somewhere there is at least one 3'-3' or 5'-5' covalent linkage, it is understood that these two regions are separated by a 3'-3' or 5'-5' covalent linkage. It is preferred that the number of 3'-3' or 5'-5' covalent linkages between the complementary regions is uneven, i.e. 1 , 3, 5, etc, such that the complementary regions are in opposition direction, i.e. one is in the 5'-3' direction and the other in the 3'-5' direction, such that the two complementary regions can form a circular type of structure such as depicted e.g. in figure 1 B. In addition, a polynucleotide section of one direction may also consist entirely of a complementary region.

This structure is very different from conventional "circularized polynucleotides" which are circularized via ligation or chemical linkage and that both require additional steps. The heterodirectional polynucleotide of the invention can be formed automatically, it is a thermodynamically stable structure, that does not require additional steps. The structure is also very different from the structure formed from a conventional unidirectional

polynucleotide having complementary regions. Such a structure forms a "pan-shaped" polynucleotide molecule, having a forked like structure, with a (partially) hybridized stem and a loop (figure 3A). The heterodirectional polynucleotide of the invention can have a more relaxed state as it does not have to bend as much as compared to a unidirectional circularized polynucleotide and can have a more stable secondary structure. In addition, such a circular heterodirectional molecule can have increased enzymatic stability, e.g. being less prone to enzymatic degradation.

The covalent linkage of the heterodirectional polynucleotide may comprise or consist of a phosphodiester bond. During the synthesis of a heterodirectional polynucleotide according to the invention it may be preferred to use nucleotides and/or nucleotide analogues and linker molecules that have compatible chemistry. The 3'-3' or 5'-5' covalent linkage of the heterodirectional polynucleotide may consist of a phosphodiester bond when only nucleotide or nucleotide analogues are used that have compatible chemistry. In case linker molecules are used in the synthesis, a nucleotide of one polynucleotide section is covalently linked via a phosphodiester bond to the linker molecule, which in its turn is covalently linked to the next polynucleotide section in opposite direction also via a phosphodiester bond. Hence, such a 3'-3' or 5'-5' covalent linkage comprises the linker molecule and two phosphodiester bonds that covalently link two 3' or two 5' ends.

The heterodirectional polynucleotide according to the invention may have two 3'-ends or two 5'-ends. Such heterodirectional polynucleotides may have one 3'-3' of 5'-5' covalent linkage. However, such heterodirectional polynucleotides may also have 3, 5, 7 covalent linkages. Hence, it is not excluded to have more than one intramolecular directional switch. The heterodirectional polynucleotide according to the invention may have one 3' end comprising the first complementary region and the other 3' end comprising the second complementary region. Alternatively, the heterodirectional polynucleotide according to the invention may have one 5' end comprising the first complementary region and the other 3' end comprising the second complementary region. This way, the heterodirectional polynucleotide is circularized and has both ends of the polynucleotide base paired and with both ends of the nucleotide being of the same type, i.e. both 3' or both 5' (see i.a. figure 3 B).

The complementary regions of the heterodirectional polynucleotide may have at least least 4 nucleotides. Having complementary regions of 4 nucleotides, that have at least 100% complementarity allows the complementary regions to bind, and hence form e.g. a circularized heterodirectional polynucleotide. The complementary regions may also be larger than 4 nucleotides. For example, the complementary regions may also be 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14, 15, 16, 17, 18, 19, 20, 21 , 22, 23, 24, 25 nucleotides wherein the percentage of complementarity is at least 80%. The number of complementary nucleotides and the percentage of complementarity may be selected depending on the conditions of interest. For example, when an heterodirectional polynucleotide is to be used in cell culture, under in vivo conditions, the complementary regions may have at least 6 nucleotides. For example, under the conditions as used in the example section, the number of base pairs in the

complementary regions may be at least 4 nucleotides. The larger the complementary region and the higher the percentage of complementarity, the greater the thermodynamic stability and thus the binding of the complementary regions will be. As long as the complementary regions result in the binding of the complementary regions, such complementary regions are suitable for the invention.

The total length of the heterodirectional polynucleotide may be at most 250 nucleotides. For example, a heterodirectional polynucleotide according to the invention may have complementary regions of 4-25 nucleotides, said complementary regions being separated at least by one 3'-3' or 5'-5' covalent linkage, and the number of nucleotides in between the two complementary regions ranging from 5-240 nucleotides. The number of nucleotides in between the two complementary regions is preferably at least the number of nucleotides of the complementary region. For example, when each of the complementary regions has 10 nucleotides, the number of nucleotides in between the two complementary regions is at least 10 nucleotides.

The heterodirectional polynucleotide according to the invention may comprise natural and/or analogous nucleotides.

The heterodirectional polynucleotides may be an aptamer. For example, circularized heterodirectional polynucleotide structures similar to as depicted in figure 3A and 5A may be suitable to be used as aptamers. Aptamers are polynucleotide structures that can bind to a specific target molecule. Aptamers are usually created by selecting them from a large random sequence pool, but natural aptamers also exist in riboswitches. Aptamers may also be used as a scaffold for attaching compounds or ligands and positioning these for interaction with a target. Suitable target molecules are for example proteins. Aptamers may be designed or selected for interaction with a protein, such that a function of the protein, e.g. enzyme activity or signal transduction of a receptor, is blocked and/or enhanced.

It is also envisaged that heterodirectional polynucleotides according to the invention may comprise a label that allows to detect and/or identify a specific heterodirectional polynucleotide out of a large pool of heterodirectional polynucleotides. The label may also merely be the sequence of a variable region within the circular structure (figure 6A) For example, a barcode sequence may be incorporated in the heterodirectional polynucleotide sequence that enables identification (see figure 6B). The barcode may be flanked by one or two primer binding sites to allow amplification and/or sequencing. The identifier sequence may be incorporated within a circular structure of the heterodirectional polynucleotide, but may also be outside the circular structure. Barcodes may have a size of ranging from several nucleotides up to 25 nucleotides. Primer binding sites may for example range from 10-30 nucleotides.

In addition, the label may also comprise a fluorescent dye, a fluorescence quencher, an energy-transfer pair, a quantum dot, or a chemiluminescent precursor. The label may also comprise a fluorescein, a rhodamine, or a cyanine. In any case, the use of a label in a heterodirectional polynucleotide may aid in the detection of the heterodirectional

polynucleotide for screening purposes, e.g. in the screening of libraries. The libraries may be present in the form of an array, or may be present in solution.

Also, a method is provided for synthesizing a heterodirectional polynucleotide according to the invention comprising the steps of:

- providing first nucleotides with a protecting group at the 5' position and a phosphoramidite group at the 3' position; - providing second nucleotides with the phosphoramidite group at the 5' position and a protecting group at the 3' position;

- synthesizing a first section with the first nucleotides and continue the synthesis of a second section with the second nucleotides;

- providing a heterodirectional polynucleotide comprising the first section and the second section.

Alternatively, the method for synthesizing a heterodirectional polynucleotide according to the invention comprises the steps of:

- providing first nucleotides with the phosphoramidite group at the 5' position and a protecting group at the 3' position;

- providing second nucleotides with a protecting group at the 5' position and a

phosphoramidite group at the 3' position;

Polynucleotides can be enzymatically synthesized, in vitro or in vivo. Enzymes synthesize DNA and RNA in the 5' to 3 'direction. Polynucleotides may also be chemically synthesized. Standard chemical synthesis is carried out in the 3' to 5' direction. Chemical synthesis may be carried out using monomers with an 5' protecting group and a 3' phosphoramidite group. The synthesis can be carried out on a solid support, e.g. an array or a bead, with the 3'-end bound to the solid support. The 5' protecting group is removed in a deprotection step, generating a primary 5' hydroxyl group. The protecting group can be a photolabile, or an acid labile group. Examples of a photolabile group are DMBOC

(Dimethoxybenzoinyloxycarbonyl) as described in M.C. Pirrung et al., J.Org.Chem. 63 (1998), 241) or NPPOC (Nitrophenylpropyloxycarbonyl) as described in US5763599 and Giegrich et al. , "New photolabile protecting groups in nucleoside and nucleotide chemistry- synthesis, cleavage mechanism and applications", Nucleosides and Nucleotides, 17(9), 1987-1996, which are incorporated herein by reference. Also a class of silyl ethers may be used to protect the 5'-hydroxyl (5 -SIL) in combination with an acid-labile orthoester protecting group on the 2'-hydroxyl (2'-ACE) 2 (Scaringe, et al. J. Am. Chem. Soc, 120, 11820-1 1821 (1998).

Removal of a photolabile group can occur for example via UV-irradiation. Removal of an acid labile group can occur for example via a solution of an acid. The now available 5' hydroxyl group of the polynucleotide chain attached to the solid support reacts with an activated 3' phosphoramidite group of a monomer. The reaction results in the formation of a phosphate linkage between the polynucleotide chain on the solid support and the monomer, thereby adding a nucleotide to the polynucleotide chain. A schematic of the chemical synthesis cycle in the 3' to 5' direction is shown in figure 1.

In polynucleotide solid phase synthesis typically phosphodiester bonds between nucleosides are formed. This requires attack of an activated phosphor species on one unit by an hydroxyl group of the second unit. All other, even only weakly nucleophilic groups, have to be protected with permanent protecting groups that will be removed at the end of oligonucleotide synthesis. Standard protecting groups are for example benzoyl for the 6- amino group at adenine, benzoyl at the 4-amino group at cytosine and iso-butyryl for the 2- amino group of guanine. At the end of polynucleotide synthesis, these protecting groups are removed, for example with an aqueous mixture with NH₃ and ethanol at about 50°C overnight. One can replace the benzoyl group by phenoxyacetyl moieties (PAC), which gives nucleobases, which are a easier to deprotect. Polynucleotides prepared using

phenoxyacetyl/iso-butyryl groups may be deprotected using a mixture of NH₃ and ethanol at room temperature overnight. It is understood that these particular protecting groups are selected such that these not removed during synthesis and that these are thus different from the protecting groups used for protecting the 5' or 3' position of the ribose of the nucleotide.

Chemical synthesis of polynucleotides in the reverse direction, from 5' to 3', is also possible. This way polynucleotides attached to a solid support, e.g. an array or a bead, via the 5' end having the 3' end available are possible. This is for example suitable for the preparation of primers attached on an array. The synthesized polynucleotides on an array can remain on the array and be hybridized with a suitable template, and can function e.g. as a primer for DNA polymerase thereby synthesizing a new DNA strand using the template from the polynucleotide on the array. The chemistry of the monomers used for chemical synthesis is reversed as compared to the monomers used for the 3' to 5' synthesis, i.e. monomers are used with an 3' protecting group and a 5' phosphoramidite group. Such 5' to 3' polynucleotide synthesis is described in Wagner T. , Pfleiderer W. "Nucleotides, Part LXV , Synthesis of 2' Deoxyribonucleoside 5' Phosphoramidites: New Building Blocks for the Inverse (5'-3')-Oligonucleotide Approach" Helvetica Chimica Acta Volume 83, Issue 8, pages 2023-2035, August 9, 2000 ; and Christopher D. Claeboe, Rong Gao and Sidney M. Hecht* "3'-Modified oligonucleotides by reverse DNA synthesis" Nucleic Acids Research, 2003, Vol. 31 , No. 19 5685-5691 ; and Beier M, Hoheisel JD. Curr Protoc Nucleic Acid Chem. 2004 Sep;Chapter 12: Unit 12.3. and US 6756492B1 Beier M., Hoheisel J. D., 2004, which are incorporated herein by reference.

An example of two different monomers for both the 3' to 5' chemical synthesis and the 3' to 5' chemical synthesis is shown in figure 2, showing a thymidine-5'-NPPOC-

3'cyanoethoxy-amidite (figure 2A) and a thymidine-3'-NPPOC-5'cyanoethoxy-amidite (figure 2B) and the corresponding DMT analogues (2C and 2D) respectively. The method of the invention relates to the synthesis of a heterodirectional polynucleotide wherein during the synthesis, at least once, the direction of synthesis has been reversed. For example, if the synthesis is started in a 5'- to 3'-direction, the direction during synthesis is changed to the 3'- to 5'-direction. This means, that a phosphodiester linkage is formed between the 3' end of the nucleotide chain attached to the solid support, and the 3' end of the next monomer which is added to the growing chain. Alternatively, when the synthesis is started in a 3' to 5' direction, the direction during synthesis is changed to the 5' to 3' direction. This means, that a phosphodiester linkage is formed between the 5' end of the nucleotide chain attached to the solid support, and the 5' end of the next monomer which is added to the growing chain. Such reversal of synthesis is obtained by switching between the different types of monomers, i.e. switching between monomers suitable for 5' to 3' synthesis and monomers suitable for 3' to 5' synthesis. This way, a heterodirectional oligonucleotide is synthesized which does not comprise a unidirectional oligonucleotide chain with connected nucleotides in only one direction (5'-3'), but comprises sections with different directions.

A heterodirectional oligonucleotide comprises at least two nucleotide sections of the two different directions. As said, a heterodirectional monomer may be represent by the following formula, 5'-(N)n-3'-3'-(N)n-5 or 3'-(N)n-5'-5'-(N)n-3', wherein N represents a monomer, which may for example be A, C, T and G, representing DNA monomeric units comprising adenosine, cytosine, guanidine or thymidine, but other monomers are also possible, and n having a value of at least 1. A heterodirectional oligonucleotide may comprise more than two of such sections, e.g. 5'-(N)n-3'-3'-(N)n-5'-5'-(N)n-3' comprises three sections.

In the method of the invention, instead of switching from nucleotides suitable for synthesizing in the 5' to 3' direction to nucleotides suitable for synthesizing in the 3' to 5' direction, alternatively a linker molecule is used. This linker molecule has compatible chemistry with the nucleotides used for the polynucleotide synthesis. For example, a linker molecule may have a phosphoramidite group (PA) and a protecting group (PG) protecting a hydroxyl group. Hence, when this linker molecule is used in a heterodirectional

polynucleotide, the linker molecule will be part of the polynucleotide and will be in between two polynucleotide sections. An example of such a linker molecule is depicted in figure 2C.

When the synthesis is carried out on a solid support, for example by using the Flexarrayer provided by Flexgen BV, Leiden, the Netherlands, the heterodirectional polynucleotides according to the invention may remain on the solid support, for example when these are used subsequently in an array. The heterodirectional polynucleotides may also be released from the solid support. A standard linker for that purpose that has been used is succinic acid. Such cleavable linker can be introduced as a succinyl-phoshoramidite with standard phosphoramidite coupling chemistry (fig.4A) or via peptide coupling chemistry as a succinate (fig.4B) at the beginning of the synthesis. Examples for cleavable linker molecules attached to a solid support e.g. a bead are shown in fig. 4C. Cleavage normally takes place during the final deprotection step with a nucleophile like ammonia or ethylenediamine.

Examples

Example V. Synthesis of oligonucleotide sequences An 8 bottle ABI 394 was used for solid phase synthesis using standard amidite synthesis protocols. Standard and reversed DMT-amidites are commercially available from

Chemgenes Inc. (Wilmington, MA, USA) or Link Technologies (Bellshill, Scotland).

Deprotection took place in cone, aeqeous ammonia at 60°C for 8 h. Purification was done via reverse phase column and TEAAc/Acetonitril gradient with DMT-on and cleavage after purification. The purified and desalted oligonucleotides were analyzed via mass

spectroscopy on a Bruker Microflex or a Shimadzu Axima instrument. Polynucleotides used in the examples are listed below. A schematic of the proposed structure of polynucleotide 1 is shown in figure 7A and of polynucleotide 8 in figure 7B.

Table 1 . Polynucleotides: The sequence number (SEQ No.), the sequence name (SEQ Name) and the sequence of the nucleotides is listed. Nucleotides in capital letters indicate standard D-DNA monomers (5' to 3'), nucleotides in small letters indicate D-DNA monomers in the reverse direction (3' to 5'). Underlined nucleotides indicate changes of the sequence when compared with reference oligonucleotide oFG423, SEQ NO: 1.

Example 2: Gel mobility shift assay

100 μΙ of 5 μΜ oligonucleotides diluted in PBS buffer (137 mM NaCI, 2.7 mM KCI and 10 mM sodium phosphate buffer solution pH7.4, Sigma 79382) was heated for 3 min at 95 °C and slowly cooled to room temperature. An 1 μΙ aliquot was analyzed by capillary

electrophoresis on Agilent DNA-1000 Chip or DNA High Sensitivity Chip (native gels) according to the manufacturer's instructions. Example 3: Enyzmatic stability assay

33 pg/ml oligonucleotide (1 OD260/ml unit single-stranded DNA) were incubated in 10 μΙ of PBSM (137 mM NaCI, 2.7 mM KCI and 10 mM sodium phosphate buffer solution adjusted with NaOH to pH 8.5, 15 mM MgCI2) with 0.1 mU (10 mU/ml final) snake venom

phosphodiesterase (SVP, an enzyme with 3' exonuclease activity, Sigma P3243) diluted in PBSM. Control reactions (t=0) were performed by adding PBSM buffer only. At the indicated time points, the sample was immediately cooled on ice, and the reaction was stopped by the addition of 1 μΙ 0.5 M EDTA (50 mM final). An aliquot of this sample was diluted 1 :25 in TE buffer (10 mM Tris pH 8.0, 1 mM EDTA), and an 1 μΙ aliquot was analyzed by capillary electrophoresis on Agilent Small RNA Chip (denaturing gel) according to the manufacturer's instructions.

Example 4: Serum stability assay

66 pg/ml oligonucleotide (2 OD260/ml unit single-stranded DNA) in 5 μΙ of PBS buffer (137 mM NaCI, 2.7 mM KCI and 10 mM sodium phosphate buffer solution pH 7.4) were incubated with 5 μΙ fetal bovine serum (FBS, Lonza Verviers DE14-801 F). Control reactions (t=0) were performed by adding PBS buffer only. At the indicated time points, the sample was stored in a freezer at -20°C until all the samples were collected. 5 μΙ of formamide was added to each sample, the mixture was vortexed, heated at 80°C for 5min and cooled on ice. An aliquot of this mixture was diluted 1 :5 in TE buffer (10 mM Tris pH 8.0, 1 mM EDTA), and an 1 μΙ aliquot was analyzed by capillary electrophoresis on Agilent Small RNA Chip (denaturing gel) according to the manufacturer's instructions.

The results of examples 1-3 were summarized in table 2.

SEQ Mod. Nt shift t1/2 (SVP) t1/2 (FBS) Proposed

No. [min] [min] Sec. Struct.

1 - 29 - <5 10-30 stem loop

2 Rev4 29 - 5-20 30-90 open

3 Fwd4 29 - <5 <10 open

4 3T3_Rev4 32 - 5-20 30-90 open

5 3T3_Fwd4 32 - <5 <10 open

6 5T3_Rev4 32 - 5-20 30-90 open 7 5T3_Fwd4 32 - <5 <10 open

8 Rev7 35 ++ >180 30-90 hd7 circle

9 Fwd7 35 - <5 10-30 open

10 Rev7scr 35 - 20-60 30-90 open

1 1 Rev10 41 + >180 300-900 hd10 circle

12 Fwd10 41 - <5 30-90 open

Table 2. Gel-shift and enzymatic stability results: Mod. refers to the modification of the sequences listed in table 1 , the coding in the Mod. column is as follows (Rev) refers to reverse complement nucleotides added after a 3'-3' switch of the synthesis orientation were compared with their unidirectional counterparts, which have the same nucleotide composition, the same order of synthesis, but in forward (Fwd) orientation, i. e. lacking a 3'- 3' switch of the synthesis orientation. The number like for example in (Rev7) indicates the number of nucleotides in reverse orientation. (Fwd 7) indicates that in reference to (Rev7) the same nucleotides in normal direction were incorporated. (n'Tm) means that compared to the reference sequence oFG423, SEQ NO: 1. additional m Thymidine nucleotides were towards the n' end were incorporated. This is also indicated with underlined nucleotides in Table 1. (scr) annotates that the reversed sequence is not complementary anymore to the 5' end of the non-reversed portion as it was scrambled. by design. The size of the polynucleotides is listed in the Nt column, which indicates the number of nucleotides. The result from the mobility shift assay, as shown in figure 7C are summarized in the (shift) column. A mobility listed as "-" denotes no gel shift, "+" denotes a gel mobility shift towards higher mobility lower than 30 %."++" denotes a gel mobility shift towards higher mobility higher than 30 %. The half life of the polynucleotides is listed in columns t1/2(SVP) and (FBS). The SVP column relates to the half life time (in minutes) of the oligonucleotide when treated with snake venome phosphodiesterase in the enzymatic stability assay according to example 2. Time points used were 0, 5, 20, 60 and 180 minutes. The FBS column describes the half life time (in minutes) of the oligonucleotide when treated with fetal bovine serum in the serum stability assay according to example 3. Time points used were 0, 10, 30, 90, 300 and 900 minutes. In the last column the proposed secondary (sec) structures are listed which are named as follows: hdn circle = heterodirectional circular structure induced by a terminal sequence consisting of n reverse nucleotides.

Claims

1. A heterodirectional polynucleotide comprising at least one 3'-3' covalent linkage or 5'-5' covalent linkage, and comprising a first complementary region and a second complementary region that are complementary to each other, wherein the first complementary region and second complementary region are separated by the at least one 3'-3' covalent linkage or 5'-5' covalent linkage.

2. A heterodirectional polynucleotide according to claim 1 , wherein the covalent linkage comprises a phosphodiester bond.

3. A heterodirectional polynucleotide according to claims 1-2, wherein the heterodirectional polynucleotide has two 3'-ends or two 5'-ends.

4. A heterodirectional polynucleotide according to claim 3, wherein one 3' end comprises the first complementary region and the other 3' end comprises the second complementary region.

5. A heterodirectional polynucleotide according to claim 3, wherein one 5' end comprises the first complementary region and the other 5' end comprises the second complementary region.

6. A heterodirectional polynucleotide according to claims 1-5, wherein the first complementary region and the second complementary region are at least 5 nucleotides.

7. A heterodirectional polynucleotide according to claim 6, wherein the first complementary region and the second complementary region are at most 25 nucleotides.

8. A heterodirectional polynucleotide according to claims 1-7, having a length of at most 250 nucleotides.

9. A heterodirectional polynucleotide according to claims 1-8 wherein the heterodirectional polynucleotide comprises natural and/or analogous nucleotides.

10. Method for synthesizing a heterodirectional polynucleotide according to claims 1-0 comprising the steps of:

- providing first nucleotides with a protecting group at the 5' position and a phosphoramidite group at the 3' position;

- providing second nucleotides with the phosphoramidite group at the 5' position and a protecting group at the 3' position;

11. Method for synthesizing a heterodirectional polynucleotide according to claims 1-9 comprising the steps of:

- providing second nucleotides with a protecting group at the 5' position and a phosphoramidite group at the 3' position;

12. Method according to claims 10-1 1 , wherein a linker molecule is provided having the phosphoramidite group and the protecting group, and, wherein after the synthesis of the first section, first the linker molecule is covalently linked to the first section, after which the synthesis is continued with the second section.

13. Method according to claim 10-12, wherein the protecting group is a photolabile or an acid labile group.

14. Method according to claim 10-13, wherein the synthesis is carried out on a solid support.

15. Method according to claims 14-14, wherein the solid support is an array or a bead.

15. Method according to claims 14-15, wherein the synthesized heterodirectional polynucleotides provided on a solid support are released from the solid support.