WO2011089393A1

WO2011089393A1 - Detection of target nucleic acids based on oligonucleotide hybridization and chemical ligation

Info

Publication number: WO2011089393A1
Application number: PCT/GB2011/000075
Authority: WO
Inventors: Andrew Thompson
Original assignee: Trillion Genomics Limited
Priority date: 2010-01-23
Filing date: 2011-01-21
Publication date: 2011-07-28
Also published as: GB201001088D0

Abstract

The present invention relates inter alia to methods for detection of target nucleic acids and oligonucleotide probes for use in such methods. In one aspect of the invention, the method comprises the steps of: providing a first oligonucleotide probe with a 5' terminus that comprises an azide group (or alternatively an alkyne group) and a first probe target recognition sequence capable of hybridising to a first region of a target nucleic acid; providing a second oligonucleotide probe with a 3' terminus that comprises an alkyne group (or alternatively an azide group) and a second probe target recognition sequence capable of hybridising to a second region of the target nucleic acid which is adjacent the first region, such that the 5' terminus of the first oligonucleotide when hybridised to the first region is adjacent the 31 terminus of the second oligonucleotide probe when hybridised to the second region; hybridising the first and second oligonucleotide probes to the target nucleic acid under hybridisation conditions; and ligating opposing terminal alkyne and azide groups of the oligonucleotide probes by cycloaddition to form a ligated oligonucleotide probe.

Description

DETECTION OF TARGET NUCLEIC ACIDS BASED ON

OLIGONUCLEOTIDE HYBRIDIZATION AND CHEMICAL LIGATION

The present invention relates inter alia to methods for detection of target nucleic acids and oligonucleotide probes for use in such methods.

Nucleic acids may be detected by contacting them with labelled oligonucleotide probe molecules under controlled conditions and detecting the labels to determine whether specific binding or hybridisation has taken place. Various methods of labelling probes are known in the art, including the use of radioactive atoms, fluorescent dyes, luminescent reagents, electron capture reagents and light absorbing dyes. Most of these labelling systems are limited in the number of probes that can be detected simultaneously. For many applications, however, it is desirable to detect multiple probes simultaneously, which is referred to as multiplexing. Fluorescent labelling schemes permit the labelling of a relatively small number of molecules simultaneously - typically four labels can be used simultaneously and possibly up to eight - although fluorescent dye encoded beads where two or more dyes are mixed in different ratios can achieve high levels of multiplexing (Fulton RJ et al., Clin Chem 43:1749-1756, "Advanced multiplexed analysis with the FlowMetrix™ system." 1997). However the costs of the detection apparatus and the difficulties of analysing the resultant signals limit fluorescence detection schemes. Mass tags can also be readily multiplexed to high levels, enabling analysis of 30 or more probes (Kokoris M et al., Mol Diagn. 5(4):329-40, "High-throughput SNP genotyping with the Masscode system." 2000).

Multiplexed assays require more than just multiple tags to be effective. Many nucleic acid probe- binding assays do not function well when multiplexed because of problems of cross- hybridisation which arises due to the high risk of cross hybridisation of primers to incorrect templates leading to cross-amplification of templates and hence to incorrect results. This is a particular issue for polymerase chain reaction (PCR)-based assays, for which it is very costly and time-consuming to optimise reactions involving multiple primer pairs.

However, various nucleic acid probe binding assay methods that enable high-order multiplexing have been developed in the art. Such assay methods include an oligonucleotide ligation assay (OLA), for example as described in US patent document US4,988,6 7. The OLA disclosed in US4, 988,617 is an assay for determining the sequence of a region of a target nucleic acid such as DNA or RNA which has a known possible mutation in at least one nucleotide position in the sequence. In this type of assay, two oligonucleotide probes that are complementary to immediately adjacent segments of a target molecule which contains the possible mutation(s) near the segment joint are hybridised to the target molecule. A ligase enzyme is then added to the juxtaposed hybridised probes. Assay conditions are selected such that when the target molecule is correctly base paired, the probes will be covalently joined by the ligase, and if not correctly base paired due to a mismatching nucleotide(s) near the segment joint, the probes are not capable of being covalently joined by the ligase. The presence or absence of ligation is detected, and is indication of the present or absence of a specific sequence of the target molecule.

A similar assay to the OLA of US4,988,617 is disclosed in EP0185494. In the EP0185494 method, the formation of a ligation product depends on the capability of two adjacent probes to hybridise under high stringency conditions, rather than on the requirement of correct base- pairing in the joint region for the ligase to function properly as in US4,988,617. Other discloses relating to ligase-assisted detection of nucleic acids include EP0330308, EP0324616, EP0473155, EP0336 731 , US4,883,750 and US5,242,794.

Ligation-mediated assays have a number of advantages over conventional hybridisation based assays. The reaction is more specific than conventional hybridisation as it requires several independent events to take place to give rise to a positive signal. Ligation reactions rely on the spatial juxtaposition of two separate probe sequences on a target nucleic acid sequence, and this is less likely to occur in the absence of the appropriate target molecule even under non- stringent reaction conditions. This means that standardised reaction conditions can be used, which allows for automation of the assay. In addition, due to the substrate requirements of ligases, incorrectly hybridised probes with terminal mismatches at the ligation junction are ligated with very poor efficiency. This means that allelic sequence variants can be distinguished with suitably designed probes. Also, the ligation event creates a unique molecule, not previously present in the assay, which enables a variety of useful signal generation systems to be employed to detect the event. High specificity makes ligation-based assays easier to multiplex, as described for example in US2003/0108913. Novel chemical ligation methods allow the ligase enzyme to be dispensed with in ligation- mediated assays. The so-called "click reaction" or Copper catalyzed Azide Alkyne Cycloaddition (CuAAC) reaction allows rapid, chemo-specific chemical ligation of oligonucleotides in a template directed fashion (Kumar R et al„ J Am Chem Soc. 129(21 ):6859-64, "Template- directed oligonucleotide strand ligation, covalent intramolecular DNA circularisation and catenation using click chemistry." 2007). The use of chemical ligation enables a wider range of template-directed ligation assays using nucleotide analogues that are not recognized by ligases but which can offer enhanced binding specificity, such as Peptide Nucleic Acid ("PNA"; Nielsen PE, Curr Opin Biotechnol. 12(1):16-20, "Peptide nucleic acid: a versatile tool in genetic diagnostics and molecular biology." 2001) or Locked Nucleic Acid ("LNA"; Vester B & Wengel J, Biochemistry. 43(42): 13233-41 , "LNA (locked nucleic acid): high-affinity targeting of complementary RNA and DNA." 2004). In addition, the ligation reaction can proceed in the presence of hybridisation reagents that would inhibit a ligase enzyme. Furthermore, chemical ligation can be helpful when the target molecule is RNA, as RNA templated DNA ligation is very inefficient (Stougaard et al., BMC Biotechnol. 7:69, "In situ detection of non-polyadenylated RNA molecules using Turtle Probes and target primed rolling circle PRINS." 2007).

Further improvements in specificity and multiplexing can be achieved using circularising probes. Circularising probes usually comprise a single linear oligonucleotide probe (a "linear circularising probe" or "LCP"), for example about 70 nucleotides in length or greater, in which two probe sequences that are to be hybridised to a target molecule are located at either end (i.e. at the 5' terminus and the 3' terminus) of the probe molecule. The probe sequences are designed so that when they bind to their target sequences of the target molecule, the termini of the probe are adjacent to each other. The probe termini can then be ligated to form a closed circular loop (a "closed circularised probe" or "CCP"). Since both probe sequences are linked to each other in the LCP, when one probe sequence binds to its target sequence, binding of the other probe sequence to its target sequence takes place with rapid kinetics. This ensures that intramolecular ligation is much more likely than inter-molecular ligation, reducing cross-ligation of probes to very low levels. In addition, cross-ligated probes are linear and it is highly unlikely that two or more probes will cross-ligate to form a circular species. Similarly, mismatched probes, i.e. probes that have bound to a target that does not exactly match the probe sequence, are unable to ligate and therefore will not be circularised. This means that correctly reacted probes can be distinguished from incorrectly reacted probes by the fact that correctly reacted probes are circular. The ability to resolve correctly matched probes allows large numbers of probes to be used simultaneously in a single reaction.

An important feature of using circularising probes is the ability to detect a signal from circularised probes (CCPs) rather than from non-circularised probes (LCP, or other incorrectly hybridised non-circular probes), and various methods for this have been disclosed in the prior art. For example, circularised probes may be detected by the ability of the probes to undergo linear rolling circle amplification (RCA; also referred to as rolling circle replication or "RCR"). The RCA method is used in JP4262799 and JP4304900, which disclose contacting a sample in the presence of a ligase with an LCP. A correctly hybridised probe is circularised by ligation to form a CCP and acts as a template in a RCA polymerisation reaction. A primer, which is at least partially complementary to the circularised probe, together with a strand-displacing nucleic acid polymerase and nucleotide triphosphates are added to the circularised sequences and a single stranded nucleic acid is formed which has a tandemly repeated sequence complementary to the circularised probe and at least partially to the template. The amplification product (also referred to herein as an "amplicon") is then detected either via a labelled nucleotide triphosphate incorporated in the amplification, or by an added labelled nucleic acid probe capable of hybridising to the amplification product (i.e., the amplicon)

Other methods based on RCA of circularised probes have been disclosed in US5, 854,033, US6,344,329, US6.210,884 and US6, 183,960. A notable difference between the disclosure of these US patents and the disclosure of JP 4262799 and JP 4304900 is the use of hyper- branching RCA in the US patents. In hyper-branching RCA, a second primer that is at least partially complementary to the single stranded nucleic acid amplification product of RCA of a CCP is added to the reaction. This results in a further geometric amplification of the single stranded nucleic acid amplification product. A further method for resolving circularised probes such as CCPs from non-circularised probes such as LCPs is disclosed in W095/22623. That method exploits the property of circularised probes that they are not susceptible to degradation by exonucleases, while unreacted, linear probes are susceptible to degradation. In addition, cyclisation of a probe to form a circularised probe "locks" the probe onto its target, i.e. the probes are resistant to being separated from their target. This allows circularised probes to be distinguished from linear probes by subjecting the probes to non-hybridising conditions. This approach to the use of circularising probes is sometimes referred to as padlock probe technology.

The CuAAC reaction has also been applied to circularising probes (see WO2008/120016) combining many of the advantages of probe circularisation with the advantages of chemical ligation discussed above.

However, higher levels of specificity using chemical ligation methods in their application to nucleic acid analysis and enabling support for single nucleotide polymorphism detection is still desirable. One objective of the present invention is accordingly to provide improved nucleic acid template-directed chemical oligonucleotide ligation methods and oligonucleotides for nucleic acid analysis.

According to the present invention, there is provided in a first aspect a method for detecting the presence or absence of a target nucleic acid, comprising the steps of:

(1 ) providing a first oligonucleotide probe with a 5' terminus and a 3' terminus, wherein the 5' terminus of the first oligonucleotide probe comprises an azide group or an alkyne group, and wherein the first oligonucleotide probe comprises at its 5' terminus a first probe target recognition sequence that is complementary to and capable of hybridising to a first region of the target nucleic acid;

(2) providing a second oligonucleotide probe with a 5' terminus and a 3' terminus, wherein the 3' terminus of the second oligonucleotide probe comprises an azide group, an alkyne group or a free 3' hydroxyl group, and wherein the second oligonucleotide probe comprises at its 3' terminus a second probe target recognition sequence that is complementary to and capable of hybridising to a second region of the target nucleic acid which is adjacent the first region of the target nucleic acid such that the 5' terminus of the first oligonucleotide when hybridised to the first region is adjacent the 3' terminus of the second oligonucleotide probe when hybridised to the second region, provided that where the free 3' hydroxyl group is present on the second oligonucleotide probe then the 3' terminus of the first oligonucleotide probe is blocked to prevent extension with a polymerase and there is a gap of one or more nucleotides in length between the 3'-terminus of the second oligonucleotide probe and the 5'-terminus of the first oligonucleotide probe when the first and second oligonucleotide probes are hybridised to the target nucleic acid;

(3) hybridising the first and second oligonucleotide probes to the target nucleic acid under hybridisation conditions;

(4) where the free 3' hydroxyl group is present on the 3' terminus of the second oligonucleotide probe, extending the 3'-terminus of the second probe with a polymerase in the presence of at least one nucleotide triphosphate that is modified with an azide group or an alkyne group, thereby introducing the azide group or the alkyne group into the 3' terminus of the second oligonucleotide probe, such that according to the method, one of the oligonucleotide probes comprises a terminal azide group and the other oligonucleotide probe comprises a terminal alkyne group which are adjacent to each other when the first and second oligonucleotide probes are hybridised to the target nucleic acid;

(5) ligating the terminal alkyne group and the terminal azide group of the oligonucleotide probes by cycloaddition to form a ligated oligonucleotide probe; and (6) detecting the presence or absence of the ligated oligonucleotide probe, thereby detecting the presence or absence of the target nucleic acid.

The first oligonucleotide probe may be hybridised to the target nucleic acid prior to, sequentially, or after hybridisation of the second oligonucleotide probe to the target nucleic acid in step (3).

According to a second aspect of the invention, there is provided a method for detecting the presence or absence of a target nucleic acid, comprising the steps of:

(1) contacting the target nucleic acid under hybridising conditions with a first oligonucleotide probe having a 5' terminus and a 3' terminus, wherein the 5' terminus of the first oligonucleotide probe comprises an azide group or an alkyne group, the 3' terminus of the first oligonucleotide probe comprises an azide group, an alkyne group or a free 3' hydroxyl group, and the wherein the first oligonucleotide probe comprises a first probe sequence complementary and capable of hybridising to a first target sequence in the target nucleic acid;

(2) contacting the target nucleic acid under hybridising conditions with a second oligonucleotide probe having a 5' terminus and a 3' terminus, wherein the second oligonucleotide probe comprises two terminal target recognition sequences that are complementary to and capable of hybridising to two regions of the target nucleic acid located on each side of the first target sequence, wherein the 5' terminus of the second oligonucleotide probe comprises an azide group or an alkyne group, and wherein the 3' terminus of the second oligonucleotide probe comprises an azide group, an alkyne group or a free 3' hydroxyl group;

(3) extending any free 3' hydroxyl groups at the 3' termini of the oligonucleotide probes with a polymerase in the presence of at least one nucleotide triphosphate comprising a sugar residue having a 3' position modified with an azide group or an alkyne group, thereby introducing the azide group or the alkyne group into the 3' terminus of the or each of the oligonucleotide probes, such that according to the method when the first and second oligonucleotide probes are hybridised to the target nucleic acid, a terminal azide group on one oligonucleotide probe is adjacent to a terminal alkyne on the other oligonucleotide probe;

(4) ligating the terminal alkyne and azide groups of the oligonucleotide probes by cycloaddition to form a circularised probe which interlocks with the target nucleic acid through catenation;

(4) optionally, separating the circularised probe from uncircularised probe; and

(5) detecting the presence or absence of the circularised probe, thereby detecting the presence or absence of the target nucleic acid. The first oligonucleotide probe may be hybridised to the target nucleic acid prior to, sequentially, or after hybridisation of the second oligonucleotide probe to the target nucleic acid.

According to the method of either the first or second aspects, the first and/or second oligonucleotide probe may comprise an immobilisation function for immobilising the oligonucleotide probe to the target nucleic acid.

According to the method of either the first or second aspects, the first and/or second oligonucleotide probe may comprise one or more tag molecules.

The immobilisation function and the one or more tag molecules may be located on different oligonucleotide probes. Here, the method may further comprise the step of separating the ligated oligonucleotide probe hybridised to the target nucleic acid from unligated first and second oligonucleotide probes using the immobilisation function on one of the oligonucleotide probes. The method may comprise stringent washing.

The immobilisation function may be a cross-linking agent, for example a photo cross-linking agent such as psoralen. The oligonucleotide probe comprising the cross-linking agent may be cross-linked to the target nucleic acid prior to hybridisation of the other oligonucleotide probe.

The method may comprise a washing step prior to hybridisation of the other oligonucleotide probe.

The immobilisation function may be a biotin function.

According to the first and aspects of the invention, there may be a gap of one or more nucleotides in length between the adjacent termini of the first and second oligonucleotide probes when hybridised to the target nucleic acid, which gap is filled before coupling the terminal alkyne group and the terminal azide group of the oligonucleotide probes.

According to a third aspect of the invention, there is provided a method for detecting the presence or absence of a target nucleic acid, comprising the steps of:

(1) contacting the target nucleic acid under hybridising conditions with an oligonucleotide probe having a 5' terminus and a 3' terminus and further comprising two terminal nucleic acid target recognition sequences that are complementary to and capable of hybridising to two neighbouring regions of the target nucleic acid, in which the 5' terminus of the oligonucleotide probe comprises an alkyne group or an azide group and the 3' terminus of the oligonucleotide probe comprises a free 3' hydroxyl group, and in which there is a gap of one or more nucleotides in length between the hybridisation site of the 5' terminus of the probe and the 3'- temninus of the oligonucleotide probe;

(2) extending the 3'-terminus of the oligonucleotide probe when hybridised to the target nucleic acid with a polymerase in the presence of at least one nucleotide triphosphate comprising a sugar residue having a 3' position modified with an azide group or an alkyne group, thereby introducing the azide group or the alkyne group into the 3' terminus of the oligonucleotide probe, provided that according to the method the oligonucleotide probe when hybridised to the target nucleic acid comprises adjacent terminal alkyne and azide groups;

(3) ligating the terminal alkyne and azide groups of the hybridised probe by cycloaddition to form a circularised probe which interlocks with the target nucleic acid through catenation;

(5) detecting the presence or absence of the circularised probe, thereby detecting the presence or absence of the target nucleic acid. The probe may comprise an immobilisation function.

The probe may comprise a cleavable group which allows the circularised probe to be re- linearised. In the methods of the invention comprising an option step of separating the circularised probe from uncircularised probe, this step may include an exonuclease treatment to degrade uncircularised probe. An exonuclease treatment may also be applied in any method of the invention to degrade unhybridised probe. According to all methods of the invention, the target nucleic acid may be fixed in a tissue section.

In situ hybridisation may be employed to detect the target nucleic acid. The tag may be a mass tag detectable by mass spectrometry, for example MALDI mass spectrometry or electrospray mass spectrometry.

The or each oligonucleotide probe may comprise a nucleic acid analogue. The or each oligonucleotide probe may comprise an artificial mismatch.

According to a fourth aspect of the invention, there is provided a pair of oligonucleotide probes (also referred to herein as "paired oligonucleotide probes" or "POPs") for detecting the presence or absence of a target nucleic acid, in which the pair consists of a first oligonucleotide probe and a second oligonucleotide and in which:

the first oligonucleotide probe has a 5' terminus and a 3' terminus, wherein the 5' terminus of the first oligonucleotide probe comprises an azide group or an alkyne group, and wherein the first oligonucleotide probe comprises at its 5' terminus a first probe target recognition sequence that is complementary to and capable of hybridising to a first region of the target nucleic acid; the second oligonucleotide probe has a 5' terminus and a 3' terminus, wherein the 3' terminus of the second oligonucleotide probe comprises an azide group, an alkyne group or a free 3' hydroxyl group, and wherein the second oligonucleotide probe comprises at its 3' terminus a second probe target recognition sequence that is complementary to and capable of hybridising to a second region of the target nucleic acid which is adjacent the first region of the target nucleic acid such that the 5' terminus of the first oligonucleotide when hybridised to the first region is adjacent the 3' tenninus of the second oligonucleotide probe when hybridised to the second region, provided that where the free 3' hydroxyl group is present on the second oligonucleotide probe then the 3' terminus of the first oligonucleotide probe is blocked to prevent extension with a polymerase and there is a gap of one or more nucleotides in length between the 3'-terminus of the second oligonucleotide probe and the 5'-terminus of the first oligonucleotide probe when the first and second oligonucleotide probes are hybridised to the target nucleic acid;

and wherein the T_m of the first probe target recognition sequence is substantially the same as the T_m of the second probe target recognition sequence.

The free 3' hydroxyl group when present on the 3' terminus of the second oligonucleotide probe may be extensible when the second oligonucleotide probe is hybridised to the target sequence with a polymerase that in the presence of at least one nucleotide triphosphate modified with an azide group or an alkyne group introduces the azide group or the alkyne group into the 3' terminus of the second oligonucleotide probe, such that within the pair of oligonucleotide probes, one of the oligonucleotide probes comprises a terminal azide group and the other oligonucleotide probe comprises a terminal alkyne group which are adjacent to each other when the first and second oligonucleotide probes are hybridised to the target nucleic acid. The or each target recognition sequence may comprise a discontinuity. Any of the discontinuities discussed below are applicable here.

The first and/or second oligonucleotide probe may comprise a tag. The first and/or second oligonucleotide probe may comprise an immobilisation function. The immobilisation function may be a cross-linking agent or biotin.

According to a fifth aspect of the invention, there is provided an oligonucleotide probe (also referred to herein as a "linear circularising probe" or "LCP") for detecting the presence or absence of a target nucleic acid, in which:

the oligonucleotide probe comprises a 5' terminus and a 3' terminus and further comprises two terminal nucleic acid target recognition sequences that are complementary to and capable of hybridising to two neighbouring regions of the target nucleic acid, wherein the 5' terminus of the oligonucleotide probe comprises an azide group or an alkyne group, and the 3' terminus of the oligonucleotide probe comprises an azide group, an alkyne group or a free 3' hydroxyl group; and

the T_m of the target recognition sequences of the oligonucleotide probe are substantially the same as each other. The free 3' hydroxyl group when present on the 3' terminus may be extensible when the oligonucleotide probe is hybridised to the target sequence with a polymerase that in the presence of at least one nucleotide triphosphate modified with an azide group or an alkyne group introduces the azide group or the alkyne group into the 3' terminus of the second oligonucleotide probe, such that the oligonucleotide probe comprises a terminal azide group and a terminal alkyne group which are adjacent to each other when the oligonucleotide probe is hybridised to the target nucleic acid.

The terminal alkyne group and azide group of the oligonucleotide probe when hybridised to the target nucleic acid may be ligatable by cycloaddition to form a circularised probe which interlocks with the target nucleic acid through catenation. The oligonucleotide probe may comprise a cleavable group which allows re-linearisation of the circularised probe. The or each target recognition sequence of the oligonucleotide probe may comprise a discontinuity. Any of the discontinuities discussed below are applicable here.

The oligonucleotide probe may comprise a tag. The oligonucleotide probe may comprise an immobilisation function.

The oligonucleotide probe may comprise a cross-linking agent.

The pair of oligonucleotide probes and/or the oligonucleotide probe according to the invention may comprise one or more primer binding sequences.

The or each probe may comprise one or more probe identification sequences.

The or each probe may comprise a microarray address sequence.

According to a further aspect of the invention, there is provided a kit comprising a pair of oligonucleotide probes as defined herein or the oligonucleotide probe as defined herein.

In another aspect, there is provided a microarray comprising discrete locations each having a microarray address sequence complement which is capable of binding to a microarray address sequence of an oligonucleotide probe as defined herein.

There is also provided a kit comprising a microarray as defined herein and a pair of oligonucleotide probes as defined herein or an oligonucleotide probe as defined herein.

The term "target nucleic acid" is also referred to herein as a "target" or a "target sequence" (and their plural forms).

As used herein, T_m is defined as the temperature at which 50% of an equimolar solution of an oligonucleotide or portion thereof and its perfect complement are hybridised in a duplex. By "substantially the same" is meant that the T_m of the first portion is identical with, or up to 10% different from, for example up to 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2% or 1 % different from, the T_m of the second portion. The term "oligonucleotide probe", "probe" and "oligonucleotide" (and their plural forms) are used interchangeably herein and are deemed to be synonymous unless otherwise indicated or unless otherwise clear from context. As used herein, the terms "nucleic acid", "oligonucleotide", "oligonucleotide probe", "probe" or similar wording refer to polymers composed of naturally occurring nucleotides, polymers composed of synthetic or modified nucleotides (i.e. nucleotide analogues), or a combination of natural, synthetic and/or modified nucleotides. Furthermore, these terms encompass polymers including non-nucleotide structures such as linkers. A polynucleotide that is a RNA or DNA may include naturally occurring moieties such as the naturally occurring bases and ribose or deoxyribose rings, or they may be composed of synthetic or modified moieties as described elsewhere herein. The linkage between nucleotides is commonly the 3-5' phosphate linkage, which may be a natural phosphodiester linkage, a phosphothioester linkage, and other synthetic linkages. Examples of modified backbones include phosphorothioates, chiral phosphorothioates, phosphorodithioates, phosphotriesters, aminoalkylphosphotriesters, methyl and other alkyl phosphonates including 3'-alkylene phosphonates, 5'-alkylene phosphonates and chiral phosphonates, phosphinates, phosphoramidates including 3'-amino phosphoramidate and aminoalkylphosphoramidates, thionophosphoramidates, thionoalkylphosphonates, thionoalkylphosphotriesters, selenophosphates and boranophosphates. Additional linkages include phosphotriester, siloxane, carbonate, carboxymethylester, acetamidate, carbamate, thioether, bridged phosphoramidate, bridged methylene phosphonate, bridged phosphorothioate and sulfone internucleotide linkages. Other polymeric linkages include 2'-5' linked analogs of these (see for example US6,503,754 and US6,506,735). The monosaccharide may be modified by being, for example, a pentose or a hexose other than a ribose or a deoxyribose. The monosaccharide may also be modified by substituting hydryoxyl groups with hydro or amino groups, by esterifying additional hydroxyl groups, and so on. Further backbone modifications suitable for use in the invention are described below.

Further aspects, non-limiting embodiments and advantages of the present invention will now be described with reference to the following drawings, in which: Fig. 1(a)-(b) illustrates template-directed oligonucleotide chemical ligation according to the invention by copper catalyzed azide alkyne cycloaddition (CuAAC). In Fig. 1(a), each of the two probes (5' probe designated as "5'P"; 3' probe designated as "3'P") on the top of the target sequence ("TS") is shown, by convention, in the direction of 5' on the left to 3' on the right of the figure. Similarly, the target sequence on the bottom is shown in the direction of 3' on the left to 5' on the right of the figure. This convention is used in all the drawings where relevant;

Fig. 2(a)-(c) illustrates template-directed oligonucleotide CuAAC after gap filling with a polymerase to introduce an azide function. In Fig. 2(a), a 5' probe is designated as "5'P", a 3' probe as "3'P", and the target sequence as "TS". Reaction "A" requires gap filling polymerase and AZT triphosphate, while reaction "B" is CuAAC;

Fig. 3(a)-(f) illustrates template-directed oligonucleotide chemical ligation by CuAAC after SNP selective gap filling with a proofreading polymerase to introduce an azide function. In Figs 3(a) and (d), "P1" designates probe 1 , "P2" designates probe 2, "PS" designates a polymorphic site, "LNA" designates an LNA nucleotide, and "TS" a target sequence. Reactions "A" and "A"' require gap filling polymerase plus ATP, GTP, CTP and AZT triphosphate, while reactions "B" and "B"' are CuAAC;

Fig. 4(a)-(d) illustrates the structures of a variety of commercially available nucleotides or nucleotides whose synthesis has been described in the literature;

Fig. 5 schematically illustrates the hybridisation of two oligonucleotide probes of one embodiment of the invention to their target nucleic acid sequences in a target sequence ("TS"). The 5' probe is shown bearing an optional immobilisation function ("IF") while the 3' probe is shown bearing a tag ("T") function. Each probe also comprises a Target Recognition Sequence ("TRS1" and "TRS2"). "G" denotes a gap;

Fig. 6 schematically illustrates the hybridisation of two oligonucleotide probes of a further embodiment of the invention to their target nucleic acid sequences in a target sequence ("TS"). The 5' probe is shown bearing a primer binding sequence ("PBS") function while the 3' probe is shown bearing a further primer binding sequence ("PBS"') and a probe identification sequence ("PIS"). Each probe also comprises a Target Recognition Sequence ("TRS1" and "TRS2"). "G" denotes a gap; Fig. 7 illustrates a directly labelled linear circularising probe ("LCP") according to another embodiment of the invention. The probe comprises two Target Recognition Sequences ("TRS1" and "TRS2"; marked as the grey regions) at either end of the probe. An intermediate sequence ("IS") is shown in white. A tag ("T") is shown linked to the probe sequence. In some embodiments, more than one tag may be linked to a probe of the invention;

Fig. 8 illustrates hybridisation of an LCP as shown in Fig. 7 to its target nucleic acid sequence ("TS"). It can be seen that the TRS regions 1 and 2 are designed here to hybridise in juxtaposition on the target sequence, leaving a small gap ("G"), which may for example be a missing phosphodiester linkage or a space of one or more nucleotides;

Fig. 9 illustrates an LCP according to another embodiment of the invention. The probe comprises two Target Recognition Sequences ("TRS1" and "TRS2"; marked as the grey regions) at either end of the probe. An intermediate sequence ("IS") is shown in white. A probe identification sequence ("PIS"; marked as a black region) is present within the intermediate sequence. The probe identification sequence is designed to uniquely identify the probe. In some embodiments, more than one probe identification sequence may be present;

Fig. 10 schematically illustrates the use of a directly labelled LCP in a method according to the invention. In Fig. 10a, "P1" denotes circularising probe 1 with a tag 1 ("T1") and "P2" denotes circularising probe 2 with a tag 2 ("T2"). Step A denotes hybridisation of the probes in the presence of a target with a target sequence ("TS"). Step B denotes ligation (and optional gap filling) in the present of the target. Exonuclease is added in step C, while step D involves capture and wash. The process continues in Fig. 10b, where "B" denotes a bead and "CO" a capture oligonucleotide. In step E, tags are cleaved from captured probes, while step F denotes injection of solute into ESI-MS and detection of mass tags;

Fig. 11 schematically illustrates the use of an LCP that comprises a PI sequence in a different method according to the invention. In Fig. 11a, "P1" denotes circularising probe 1 with a probe identification sequence 1 ("PIS1") and "P2" denotes circularising probe 2 with a probe identification sequence 2 ("PIS2"). Step A denotes hybridisation of the probes in the presence of a target with a target sequence ("TS"). Step B denotes ligation (and optional gap filling) in the present of the target. Probes are captured in step C, where "B" denotes a bead and "RCP" a rolling circle primer, while step D involves performing rolling circle amplification (RCA). The process continues in Fig. 11 b, where in step E mass tagged tag complement oligonucleotides, comprising probe detection sequences ("PDS1" and "PDS2") and mass tags ("MT1" and "MT2") are hybridised. In step F, unhybridised tag complement oligonucleotides are washed away, and in step G, tags are cleaved from captured tag complement oligonucleotides. In step H, solute is injected into ESI-MS and mass tags detected; Fig. 12 schematically illustrates the use of an LCP that comprises a probe identification sequence ("PIS") and a primer binding sequence ("PBS") in another method according to the invention. In Fig. 12a, "P1" denotes circularising probe 1 with a probe identification sequence 1 ("PIS1") and "P2" denotes circularising probe 2 with a probe identification sequence 2 ("PIS2"). "PBS" refers to the primer binding sequences on each probe. Step A denotes hybridisation of the probes in the presence of a target with a target sequence ("TS"). Step B denotes ligation (and optional gap filling) in the present of the target. Probes are captured in step C, where "B" denotes a bead and "CO" a capture oligonucleotide. Further steps in the procedure are explained below. In Fig. 12b, "P" refers to a primer, "POL" to polymerase, "MT1" and "MT2" to mass tags 1 and 2, respectively, and "PDS1" and "PDS2" to probe detection sequences land 2, respectively. In Fig. 12c, the final step depicted in injection of solute into ESI-MS and detection of mass tags;

Fig. 13 schematically illustrates the use of an LCP that comprises a cleavable linker according to the invention. In Fig. 13a, "P1" denotes circularising probe 1 with a mass tag ("Tf '), while "P2" denotes circularising probe 2 with a mass tag ("T2"). Both probes have a cleavable group ("C") and a capture sequence ("CS"). Step A denotes hybridisation of the probes in the presence of a target with a target sequence ("TS"). Step B denotes ligation (and optional gap filling) in the present of the target. A cleavage agent is added in step C. In Fig. 13b, "B" denotes a bead. Step D denotes capturing and washing, while in step E mass tags are cleaved from captured probes. In step F, solute is injected into ESI-MS and mass tags detected; and

Fig. 14 schematically illustrates the use of an LCP involving gap oligonucleotides according to another embodiment of the invention. A tag on the LCP is denoted "T", and the target recognition sequences "TR1" and "TR2". A gap oligonucleotide ("GO") has an immobilisation function ("IF"). A target sequence is indicated as "TS".

Copper catalyzed azide alkyne cycloaddition (CuAAC) ligation

Azide alkyne cycloaddition can be used to conjugate two molecules and is advantageous because both of the alkyne and azide functionalities are relatively stable. However, these two functions react very slowly by themselves. Fortunately, it has recently been demonstrated that catalysis of the reaction of these two functional groups, particularly with copper, can lead to orders of magnitude improvements in reaction rate and also regio-specificity in the reaction (Rostovtsev et al., Angewandte Chemie-lntemational Edition 41 , 2596-2599, "A stepwise Huisgen cycloaddition process: Copper(l)-catalyzed regioselective "ligation" of azides and terminal alkynes" 2002; Tornoe et al., Journal of Organic Chemistry 67, 3057-3064, "Peptidotriazoles on solid phase: [1 ,2,3]-triazoles by regiospecific copper(l)-catalyzed 1 ,3-dipolar cycloadditions of terminal alkynes to azides." 2002). Thus, the CuAAC reaction allows rapid coupling of two suitably functionalized biomolecules or tags under very mild conditions using stable functional groups. The reaction has been used to couple peptides to oligonucleotides (Gogoi K et al., Nucleic Acids Res. 35(21 ):e139, "A versatile method for the preparation of conjugates of peptides with DNA/PNA analog by employing chemo-selective click reaction in water." 2007) and oligonucleotides to oligonucleotides (Kumar et al., 2007, supra).

The CuAAC conjugation approach has an advantage that the reactive groups for conjugation are stable and can be introduced easily particularly during automated peptide and oligonucleotide synthesis. The reagents are also fairly stable and can be stored and transported easily. The final conjugation can be controllably initiated by the addition of the necessary catalyst to drive the reaction.

Ligation mediated assays

Template-directed ligation mediated assays as used in the present invention have a number of advantages over conventional hybridisation based assays, as discussed above. Chemical ligation has further advantages, in particular CuAAC ligation, as it occurs under mild conditions but at a meaningful rate only in the presence of a catalyst. This means that probes can be hybridised under controlled conditions prior to ligation, as for enzymatic ligation, but with a wider range of buffers, solvents, temperatures and nucleic acid modifications than enzymatic ligation, thereby extending the range of probes and probe conditions that can be used.

As noted above, improvements in specificity and multiplexing ability in chemical ligation methods can be achieved using LCPs which form CCPs. Large numbers of LCPS can be used simultaneously in a single reaction (Hardenbol et al., Nature Biotechnology 21(6): 673 - 678, 2003) but can be designed for specific sequences, rather than relying on polyadenylation. This makes LCPs useful for analysis of bacterial and viral RNA, which is encompassed by the present invention. Also encompassed by the invention is the ability to analyze numerous species simultaneously allows the analysis of viral RNA and bacterial simultaneously with human mRNA, for example allowing expression changes in both host and infectious agent to be analyzed simultaneously during studies of infection.

As described above, padlock probes have been used to label RNA and DNA in tissue sections since they "lock" onto their targets and the tissue section can then be washed stringently to remove non-specifically bound probes. This is useful when densely-labelled probes are used as non-specific binding could otherwise result in a significant signal.

A number of specific problems are associated with analysis of gene expression. Expression analysis typically involves the analysis of RNA species. RNA can be converted to cDNA by reverse transcription and numerous methods for such conversion are known in the art (Wang J. et al., Biotechniques 34(2):394-400, "RNA amplification strategies for cDNA microarray experiments." 2003; Petalidis L. et al., Nucleic Acids Res. 31(22): e142, "Global amplification of mRNA by template-switching PCR: linearity and application to microarray analysis." 2003; Baugh L.R. et al., Nucleic Acids Res. 29(5):E29, "Quantitative analysis of mRNA amplification by in vitro transcription." 2001). However, it has been shown that target mediated ligation of LCPs can be performed with RNA targets directly, thus avoiding the need for conversion of RNA to cDNA (Nilsson M. et al., Nat Biotechnol. 18(7):791-793, "Enhanced detection and distinction of RNA by enzymatic probe ligation." 2000). Unfortunately, mRNA is not a good substrate for the ligases used to lock padlock probes onto target sequences (Stougaard et al., 2007, supra). In addition, ligases typically do not like DNA analogues, which are useful to enhance binding affinity and specificity. Chemical ligation like the CuAAC reason, as used in the present invention, is sufficiently specific and controllable and addresses these above problems. Paired oligonucleotide probes and linear circularising probes

According to certain aspects of this invention, paired oligonucleotide probes (POPs) consisting of first and second oligonucleotide probes are used for detection of target nucleic acids. Linear circularising probes (LCPs) are used in other aspects of the invention. Each probe in a POP pair comprises a target recognition sequence (TRS), as depicted for example in Figs 5 and 6. In each probe of a POP pair, the TRS may be at either the 5' end or 3' end of the sequence but within the pair, one probe should have the TRS at the 3' end and the other probe should have the TRS at the 5' end. LCPs comprise two TRS's which hybridise to two neighbouring regions of a target sequence except where one or more gap oligonucleotides or other means are used to bridge a gap between hybridised TRS's. In all cases, two TRS's are required for the methods of the invention, as depicted for example in the accompanying Figs 5, 6, 7, 8 and 9 where the two TRS's are designated TRS1 and TRS2.

The size of each of TRS1 and TRS2 may vary and be independent of each other. In some examples, one TRS may be designed to hybridise to an allelic sequence, e.g. a target nucleic acid sequence which may have one of two or more variations at a specific nucleotide. The TRS which hybridises to the allelic sequence is designated TRS1 for the description below, although it will be understood that this is an arbitrary designation and the TRS which hybridises to the allelic sequence may be at the 5' end or the 3' end of a probe. The present invention may be used for example to determine which of two or more single nucleotide polymorphisms (SNPs) is present in a target sequence, by using a set comprising a mixture of two or more POPs or LCPs, each of which has a TRS1 specific for one SNP and a TRS2 which will may be identical for each member of a set of POPs or LCPs.

The length of the TRS1 and position of the allelic nucleotide may be selected to allow a first TRS1 of a first POP or LCP which is completely complementary to its target nucleic acid to hybridise to that target nucleic acid and be ligated to a first TRS2 when hybridised, whilst a second TRS1 of a second POP or CLP in the same set but which differs by only a single residue does not hybridise to the target nucleic acid sufficiently to undergo ligation with a second TRS2 when the target nucleic acid is specific for the first TRS1.

A TRS may be between 15 and 25 nucleotides in length, though shorter lengths, for example of from 4 or more nucleotides, are also encompasses. The precise size and composition of the TRS may be selected by a person of skill in the art taking into account the specific nature of the target.

After a TRS1 and TRS2 have hybridised to the target molecule and any missing nucleotides between the LCP ends have been filled, adjacent probe ends are connected to each other by chemical ligation. In the case of an LCP, based on appropriate selection of the probe sequence as well as the combined length of any gap filling nucleotides or oligonucleotides, the resultant circular CCP molecule formed will be wound around and interlock with the target nucleic acid. The CCP circularised sequence may be about 70 nucleotides or greater, for example about 70 to 100 nucleotides, in length, such as a probe of about 80 or about 90 nucleotides in length. These dimensions apply for example where the CCP comprises nucleotide linkages. An LCP comprising non-nucleotide linkages may have different steric limitations and therefore such an LCP may be shorter.

It is sufficient for the purposes of certain aspects of the present invention that the TRS or TRS's of a probe comprises nucleotides or optionally functionally analogous structures that can undergo ligation but that all or part of non-TRS components of each probe has a non-nucleotide chemical composition referred to herein as an "intermediate structure" or "intermediate sequence" of the probe. The intermediate structure may for example comprise residues selected from peptides or proteins, carbohydrates or other natural or synthetic polymers. Such an intermediate structure of non-nucleotide nature may be used to enhance stability and/or allow easier introduction of labels or tags into the probe. A non-nucleotide intermediate structure may be selected to avoid a secondary structure in the probe and/or avoid mis-hybridisation with a target nucleic acid. If, however, a probe comprises only nucleic acids, the combined lengths of the component sequences of each LCP may be such that the strands will leave a DNA double helix on the same face 10 or a multiple of 10 bases apart (10 bases representing approximately one turn of a DNA double helix). Having a gap of one or more nucleotides between an adjacent TRS1 and TRS2 when hybridised to the target nucleic acid may be advantageous as the gap filling step may improve specificity of the recognition reaction. It is noted that a gap is not essential for certain aspects the invention, as these aspect may be performed where TRS1 and TRS2 are designed to bind in immediate juxtaposition on the target molecule, whereupon their termini can be directly ligated to circularise the LCP to form a CCP.

There are a number of advantages of using an LCP that can form a covalently closed circular molecule (a CCP) upon correct hybridisation to the target nucleic acids, rather than using conventional labelled linear probes. First, each target nucleic acid requires only a single, synthetic probe molecule (LCP). Second, the hybridisation and ligation reaction can provide high specificity of detection, if configured to use a proof-reading polymerase (discussed further below). Third, the circularisation of a correctly hybridised LCP provides a number of ways by which a correctly matched probe can be distinguished from on or more incorrectly matched LCP: a CCP catenates with the target nucleic acid, thereby becoming substantially insensitive to denaturants; the ends of the CCP become unavailable to exonuclease digestion; and CCPs can mediate rolling circle amplification (RCA). Finally, the simultaneous presence of two terminal TRS's on one molecule in the case of an LCP confers kinetic advantages in the hybridisation step. Illustrated in Fig. 7 is an LCP according to one aspect of this invention, in an embodiment where the probe is directly conjugated to a tag. The two termini of the probe comprise the TRS's of the probe. The 3' terminus of the probe preferably comprises a free hydroxyl group, an azide or an alkyne while the 5' terminus of the probe comprises either an azide or an alkyne. Fig. 8 illustrates the same directly labelled probe as shown in Fig. 7, hybridised to a target nucleic acid sequence, such as a DNA strand, via two TRS end segments of the probe, designated TRS1 and TRS2. TRS1 and TRS2 are complementary to two respective almost contiguous sequences of the target nucleic acid: a gap is present between TRS1 and TRS2 when hybridised to the target nucleic acid. In this example, the 3' terminus of the probe has a 3' hydroxyl group. The gap may be a missing phosphodiester linkage or it may comprise a space of one or more nucleotides. If the gap comprises a space of one or more nucleotides, it may be bridged by a second oligonucleotide probe or it may be filled by polymerase activity in the presence of nucleotide triphosphates so that an azide or alkyne is introduced into the 3' terminus of the probe, rendering the probe competent for CuAAC ligation.

Paired oligonucleotide probes for ligation assays

The invention in certain aspects relates to a method for the detection of one or more target nucleic acids by the ligation of a pair of probes complementary to the target, where the probes are ligated to each other upon hybridisation to the target nucleic acid(s) by azide alkyne cycloaddition. Template directed or template mediated ligation to effect circularisation of oligonucleotide probes using the CuAAC reaction has been demonstrated previously (Kumar et al., 2007, supra).

CuAAC ligation is used according to various aspects the invention for detection of DNA or RNA sequences in a biological sample. Methods are described in more detail below for a single nucleic acid target but multiple sets of probes can be used to detect multiple targets. Indeed, one of the advantages is the ability of the invention to be employed in a multiplexed format. The invention can be applied to multiple target nucleic acids even if the examples discussed herein are directed, in order to explain various features of the invention, to a single target nucleic acid. In one step of the first aspect of the invention, a first oligonucleotide probe having a first probe target recognition sequence (TRS) which is complementary (in full or partly) to a first region of a target nucleic acid in a sample is hybridised to the target nucleic acid. Prior to, after, or simultaneously with the first probe, a second oligonucleotide having a second probe TRS which is complementary (in full or partly) to a second region of a target nucleic acid in a sample is hybridised to the target nucleic acid. This is depicted in Fig. 1(a). In this figure, the bottom strand is the target nucleic acid while the two oligonucleotide probes are on top. The second probe may be tagged if desired. The second probe is designed to hybridise to the target nucleic acid adjacent to the first probe oligonucleotide. The second probe may be designed to hybridise adjacent to either the 5" or 3' terminus of the first probe. The second probe may for example comprise either an azide group (also referred to herein as an "azide function") at the 5' terminus or an azide group at the 3' terminus of the probe depending on which terminus of the first probe it is intended to ligate with. If the first probe has an alkyne group (also referred to as an "alkyne function") at its 3' terminus, the second probe may bear the azide function at the 5' terminus and the second probe may be designed to be complementary (in full or part) to a site on the target nucleic acid such that the second probe hybridises to the target on the 3' side of the first probe. If the alkyne function is at the 5' terminus of the first probe then the second probe may comprise an azide function at the 3' terminus and the second probe may be designed to be complementary to a site on the 5' side of the first probe.

Unlike with enzymatic ligation reactions, the first probe and second probe do not need to be immediately adjacent to each other for chemical ligation to take place. The alkyne and azide functions may for example have flexible linkers that will tolerate a small gap (for example a gap of about 1 to 3 nucleotides in length) between the oligonucleotides if that is desirable. However, it is usually preferred that short linkers are used and the two probes are designed to be as close to each other as possible when hybridised to the target nucleic acid to ensure maximum specificity for the reaction.

In the embodiment shown in Fig. 1 , the distinction between the first probe and the second probe is arbitrary, and the probes are referred to as a 5' probe and a 3' probe. It can be seen in Fig. 1 that the 5' probe has an azide function at its 3' terminus, while the 3' probe has an alkyne at its 5' terminus. After hybridisation of the probes, unhybridised probes are washed away with hybridisation buffer. The probes are then ligated to each other by the CuAAC reaction, as shown in Fig. 1(b). In this way, a new oligonucleotide sequence is generated in situ in the presence of the target. First and second probes may comprise additional sequences in addition to the target recognition portions of the probe. For example, they may comprise primer sequences and/or one or more address sequences (see below) and/or probe identification sequences that will uniquely identify each member of a probe pair.

As noted above, the two probes when hybridised to the target nucleic acid may have a gap between them. The 3' terminus of the first oligonucleotide probe may comprise a free 3' hydroxyl group, which can be extended by a polymerase in a gap filling reaction which may introduce an azide or an alkyne function that is present on an appropriately modified nucleotide triphosphate. This is discussed in more detail in the section below on gap-filling reactions, with reference to Figs 2 and 3.

The present invention in some aspects is directed to the use of CuAAC ligation for diagnostic detection of sequences present in biological samples. Here, the first probe may comprise an immobilisation function while the second probe may comprise a tag, as depicted in Fig. 5. Correct ligation of the probes in the presence of their target nucleic acid will link the immobilisation function to the tag allowing the tag to be captured. If the immobilisation function is biotin then the ligated probes can be captured onto an avidinated solid support allowing non- ligated tagged probes to be washed away. Alternatively, the probe may be covalently linked to a solid support via the immobilisation function.

The present invention may in some aspects be directed to the use of CuAAC ligation for in situ hybridisation of oligonucleotide probes to one or more target nucleic acids which may b fixed in a tissue section. Here, a target nucleic acid may be detected in a multi-step process. Starting from a fixed tissue section that has been prepared for in situ hybridisation (as elaborated below), a first probe complementary (in full or partly) to a target nucleic acid is hybridised to the target nucleic acid, as illustrated in Fig. 5. Again, the first probe comprises an immobilisation function, which may be a cross-linking agent such as psoralen or a psoralen derivative. According to the invention, psoralen can be introduced into the 5' terminus of an oligonucleotide probe during standard automated phosphoramidite DNA synthesis using for example the Psoralen C2 Phosphoramidite reagent (2-t4'-(hydroxymethyl)-4,5',8-trimethylpsoralen]-ethyl-1-0-(2- cyanoethyl)-(N,N-diisopropyl)-phosphoramidite) from Glen Research, Inc. Alternatively, psoralen modified deoxyadenosine can be positioned at an appropriate adenosine within the oligonucleotide probe sequence as published (U. Pieles and U. Englisch, Nucleic Acids Res. 17, 285-299, "Psoralen covalently linked to oligodeoxyribonucleotides: synthesis, sequence specific recognition of DNA and photo-cross-linking to pyrimidine residues of DNA." 1989; Pieles U et al., Nucleic Acids Res. 17(22):8967-78, "Preparation of a novel psoralen containing deoxyadenosine building block for the facile solid phase synthesis of psoralen-modified oligonucleotides for a sequence specific crosslink to a given target sequence." 1989). In addition, the first probe may comprise an azide function or an alkyne function at either of its 3' or 5' termini (see examples below for preparation of azide and alkyne terminated oligonucleotides).

After hybridisation of the first probe, the tissue section is washed (for example, with hybridisation or washing buffer) to remove any unhybridised first probe. When psoralen is used as the cross- linking agent, the hybridised first probe is then photo-cross-linked to the target nucleic acid strand by irradiation at a wavelength of 350 nm, thus covalently linking the first probe to the RNA, which in turn is fixed to the tissue section. After cross-linking of the first probe, a second tagged oligonucleotide complementary to the target nucleic acid may be hybridised to the target nucleic acid. The second probe may comprise either an azide function at the 5' terminus or an azide at the 3' terminus of the probe. If the first probe has an alkyne function at the 3' terminus, the second probe may bear the azide function at the 5' terminus and the second probe may be designed to be complementary (in full or partly) to a site on the target nucleic acid that is on the 3' side of the first probe when hybridised. If the alkyne function is at the 5' terminus of the first probe then the second probe may comprise an azide function at the 3' terminus and the second probe may be designed to be complementary (in full or partly) to a site on the target nucleic acid that is on the 5' side of the first probe when hybridised. As described above, a (fillable) gap may be present between the termini to be ligated. After hybridisation of the second probe, unhybridised probe may be washed away with hybridisation buffer. The second probe may then be ligated to the first probe by the CuAAC reaction. In this way, a tagged probe is covalently linked to the target nucleic acid allowing further (very) stringent washes to be used to remove any unhybridised and unligated probes, thus improving specificity of the assay. In addition, since two separate hybridisation events must take place (and the probes can be designed to hybridise at different temperatures), high specificity is ensured.

Where the adjacent termini of two probes have a gap between them, the 3' terminus of the first probe may comprise a free 3' hydroxyl group which may be extended by a polymerase in a gap filling reaction that can be used to introduce an azide or an alkyne function on an appropriately modified nucleotide triphosphate. This is discussed in more detail in a section below on gap- filling reactions.

The location of the various probe functionalisations can be varied. For example, the first probe may carry the alkyne group at its 5' terminus, while the psoralen may be at the 3' terminus or elsewhere (for example, central) in the probe. The second tagged probe may then be designed to bind 5' to the first probe and can bear an azide group at the 3' terminus. Alternatively, if a gap-filling reaction is to be used, then the second probe would bear a free 3' hydroxyl group. Interrupted linear circularising probes and gap oligonucleotides

In the second aspect of this invention, a linear circularising probe (LCP) has two terminal target recognition sequences (TRS's) which are designed to hybridise with a target nucleic acid to leave a gap between the termini of the LCP. The gap may then be filled by hybridisation with another oligonucleotide probe, also referred to herein as a "gap oligonucleotide", that hybridises between the two TRS regions of the LCP, as depicted in Fig. 14. The LCP and gap oligonucleotide may be hybridised to the target nucleic acid simultaneously or in either order.

In this aspect, the LCP may comprise either an alkyne function or an azide function at each terminus. Similarly, the gap oligonucleotide may comprise an azide function at its 3' and 5' termini if the LCP comprises an alkyne function at both of its 3' and 5' termini, or the gap oligonucleotide will comprise an alkyne function at both 3' and 5' termini if the LCP comprises an azide function at both of its 3' and 5' termini.

Gap filling with a polymerase to introduce an alkyne or an azide function, as discussed herein, is also possible with this aspect of the invention, if the 3' terminus of either or both of the LCP or the gap oligonucleotide has a free 3' hydroxyl group.

In Fig. 14, the gap oligonucleotide is shown to comprise an immobilisation function while the LCP comprises a tag. After successful ligation, the circularised probe will comprise both a tag and an immobilisation function, allowing correctly circularised probes to be captured onto a solid support and thereby linking the tag to the support. Any unreacted tagged LCPs can be washed away, thus separating the reacted (circularised) probe from unreacted probes. Design of ligation probes

Probes of or for use in the invention may be designed in various ways, depending on their intended application. For SNP detection, for example, the 5" probe of a POP pair (or the 3' TRS of an LCP) may be designed so that the 3'-end nucleotide of the probe is hybridised to the polymorphic nucleotide of the SNP. Different 5' probes of POP pairs may thus be designed to hybridise with and recognise each possible base change at a polymorphic site, while the 3'- probe may be generic. Alternatively, if a stop-base design is used, the 5' probe may be designed so that the 3'-end nucleotide hybridises adjacent (just before) the polymorphic nucleotide of an SNP and there is a one nucleotide gap between the 5' probe and the 3' probe (or 3' TRS and 5' TRS of an LCP) that must be filled by a polymerase. Thus either a generic 5' or 3' probe may be used.

A polymorphism may be detected by setting up four different reactions with polymerase and a single different nucleotide triphosphate added to each reaction. Extension and ligation will only occur in the reaction where the specific nucleotide triphosphate added is complementary to the polymorphism.

The TRS on either side of the polymorphism may be selected so that the melting temperature of each TRS is substantial the same as each other. In a multiplexed assay, the melting temperature of each TRS in one probe may be substantial the same as the melting temperatures of each of the TRS's of other probes in the multiplexed assay.

The use of asymmetric TRS's in a LCP is also encompassed by the present invention. It has been suggested in the prior art that circularising probes with asymmetric melting temperatures may be more specific, i.e. where one TRS has a higher melting temperature than the other and where the hybridisation reaction is optimised to maximise specificity of the binding of the TRS with the higher melting temperature (Szemes M et al., Nucleic Acids Res. 33(8):e70, "Diagnostic application of padlock probes-multiplex detection of plant pathogens using universal microarrays." 2005).

One class of sequences that is difficult to resolve from each other are gene fusions and splice variants. In a process called splicing (or cis-splicing), non-coding introns or other sequences are removed from pre-mRNA (also known as heterogeneous nuclear RNA [hnRNA] or immature RNA), and the remaining exons which usually encode for polypeptides are joined together to produce mature mRNA. Another splicing process called trans-splicing involves ligation of two or more exons from different genes. Differential splicing by extending or skipping of one or more exons, or retaining of one or more introns, creates alternative mRNA splice variants of the same gene. Many genes which are subject to splicing have a modular structure in which domains or "blocks" of a gene sequence can be recombined in different ways to give alternative mRNA splice variants. The presence of such conserved or shared domains makes it difficult to distinguish different splice variants in hybridisation-dependent gene expression profiling assays.

Additionally, chromosomal translocations or deletions may result in two "parent" genes being fused to form a "chimeric" gene (also referred to as a "gene fusion") and a mutant chimeric gene product may be expressed. The chimeric gene comprises part of the sequence of one parent gene and part of the sequence of another parent gene. Chimeric genes are found in certain cancers such as leukemia, where the presence of particular translocations or deletions is associated with the severity of the disease. One characteristic fusion found in leukemia is a BCR-ABL gene fusion ("B2A2"; see Bohlander, 2000, Cytogenet Cell Genet. 91(1-4):52-6). The best mode of treatment by targeting and killing cancer cells which express chimeric genes may be determined following identification of the presence of these fused sequences. It would be useful to be able to reliably detect a chimeric gene (or its mRNA) in a simple assay.

Consider for example a gene comprising the exons A, B and C, with two splice variants "A-C" and "A-B-C". Detecting the presence of both variants can be achieved with probes for exon A or C. The presence of the second variant could be determined using a probe for exon B, while the presence of the first variant could be identified by the presence of A or C but not B. However, if both variants are present in a sample, the presence of the variant "A-C" could not be ascertained with exon-specific probes for exon A and exon C only, as these exons are shared by the variants.

The problem of detecting a splice variant or a chimeric gene mRNA is compounded when attempting to perform in situ hybridisation, as corresponding chromosomal gene sequence(s) and pre-mRNA may additionally be present. Also, amplification of the target may be difficult or impossible, so any probe used should be highly sensitive. Furthermore, generally speaking it is necessary to have high affinity probes for in situ hybridisation to assist binding in the presence of extensive secondary structure and auxiliary proteins bound to RNA and also to ensure that the probe remains bound to the target after extensive wash steps to remove unbound probe. Achieving both high affinity and high specificity is very challenging. It has been proposed that boundary-spanning oligonucleotide probes complementary to sequences flanking juxtaposed domain sequences can be used to detect splice variants or gene fusion mRNAs (Kane et al., 2000, Nucleic Acids Res. 28: 4552-4557). The experimental results for that approach using 50-mer oligonucleotides showed significant cross-hybridisation of different splice variants to the boundary-spanning probes. This should not be unexpected since splice variants that share one exon spanned by a boundary-spanning probe can hybridise to part of the sequence of the probe.

Ligation of a pair of probes across a splice junction has previously been proposed and demonstrated (Yeakley J et al., Nat Biotechnol. 20(4):353-8, "Profiling alternative splicing on fiber-optic arrays." 2002). In the Yeakley et al. approach, ligation probes were designed so that a TRS of a 5' probe binds 20 bases of the 3' exon in a splice junction and a TRS of a 3' probe binds 20 bases of the 5^* exon of the splice junction. The probes were thus designed to be length symmetrical with the ligation point at the junction boundary. An advantage of the ligation approach, applicable to the present invention, is that novel splice junctions between exons can be discovered as the probes for each exon can ligate to each other independently. If they are appropriately screened, new splice variants can be detected. For example if a 5' probe bears a unique address sequence for binding to a microarray, while a 3' probe bears a different unique address sequence for detection by hybridisation with a set of labelled probe identification sequences, then any combination of 5' and 3' exon ligations can be distinguished.

The invention also provides improved splice junction and gene fusion ligation probes, where the TRS of the first probe and the TRS of the second probe of a probe oligonucleotide pair have substantially the same melting temperature (T_m). By having the same T_m, the TRS's (or TRS "regions") of the probes may have different lengths. The probe design proposed here may enhance the discrimination of any sequence compared to the standard approach of probes with symmetrical lengths. Similarly, the TRS's at each end of a circularising probe can be designed in a similar manner. The T_m of the TRS's for the first and second probes, or the TRS at the termini of a circularising probe, may be calculated theoretically, for example using the Nearest Neighbour method (see Breslauer et al., 1986, Proc. Natl Acad. Sci. USA 83(11 ): 3746-50; SantaLucia et al., 1996, Biochemistry 35(11 ): 3555-62; Xia et al., 1998, Biochemistry 37(42): 14719-35; Kierzek et al., 2006, Nucleic Acids Res. 34(13): 3609-14; McTigue et al., 2004, Biochemistry. 43(18): 5388- 405). As a further enhancement over known probe designs, each TRS of each oligonucleotide probe in a pair according to the invention, or each TRS at the termini of a circularising probe of the invention, comprises at least one discontinuity relative to the target nucleic acid (or target sequences thereof). Such a probe will typically bind to a target nucleic acid with a lower affinity and consequently with a lower T_m than a corresponding oligonucleotide probe that is fully complementary to the same target. Thus, the oligonucleotide probe of the present invention will be slightly less sensitive than a corresponding fully complementary oligonucleotide probe of the same length. However, the difference between the T_m of a duplex formed by the oligonucleotide of the invention and its target nucleic acid compared to the T_m of the duplex formed by the oligonucleotide of the invention and a nucleic acid variant comprising only one domain of conserved or identical sequence will be larger than that for a corresponding fully complementary oligonucleotide which lacks a discontinuity. This feature of the oligonucleotide of the invention allows, for example, enhanced discrimination of the target vs variant nucleic acids, as described in International Patent Application No. PCT/GB2009/002748 filed on 24 November 2009 in the name of Trillion Genomics Limited et al. and entitled "Oligonucleotides".

According to the invention, a discontinuity present in the terminal TRS's of a circularising probe of may be positioned adjacent nucleotide 2 to 20, for example adjacent nucleotide 3 or nucleotide 4 or adjacent nucleotide 8 to 12, of the oligonucleotide, relative to a splice junction, gene fusion boundary or polymorphism site of a target nucleic acid. The discontinuity may be present in the centre of the TRS. For example, in a TRS of 15 bases, the discontinuity may be located at base 8. In a TRS with an even number of nucleotides, the discontinuity may be at a position near the centre of the TRS. For example, in a 20-mer TRS, the discontinuity may be present at base 10 or 11.

The discontinuity may for example be positioned so that the T_m of a portion of the TRS upstream of the discontinuity is substantially the same as a portion of the TRS downstream of the discontinuity.

Each discontinuity may be of a length equivalent to 1 to 5 nucleotides, for example 1 or 2 nucleotides.

Each discontinuity may comprise a natural nucleotide which is non-complementary to a base at a corresponding position in the target nucleic acid. Alternatively, the discontinuity may comprise an artificial mismatch. The TRS regions of the oligonucleotide probes of the present invention may comprise a mixture of one or more natural nucleotide mismatches and one or more artificial mismatches. The artificial mismatch may comprise a universal base analogue or an abasic mismatch.

Each discontinuity may be a non-nucleotide spacer, for example polyethylene glycol, a phosphoramidite spacer such as a C3 phosphoramidite spacer, or an amino acid.

The oligonucleotide may comprise a natural nucleotide or a nucleotide analogue, for example a 2-O-methyl analogue, a bridged nucleic acid monomer (such as a locked nucleic acid [LNA] monomer), a peptide nucleic acid (PNA) monomer, a universal nucleoside, or a combination of any of these. Further suitable nucleotide analogues are described below.

Nucleotide analogues suitable for use in the oligonucleotide probe of the present invention may have enhanced binding affinity compared with their native DNA or RNA counterparts. For example, for the purposes of binding to RNA, the analogue may have an enhanced affinity for RNA such as a 2'-modified RNA analogue (for example, 2-O-Methyl RNA, 2' Fluoro-RNA and 2- O-ethyl RNA). Bridged nucleic acid monomers comprise a linkage from the 2' position in the ribose ring to the 4'-position. Bridged nucleic acid monomers (for example, LNA monomers) are suitable analogues for oligonucleotides intended for binding and recognition of DNA targets. An oligonucleotide probe comprising both one or more 2'-modified RNA analogues and one or more bridged nucleic acid monomers is also encompassed.

An oligonucleotide comprising PNA, and a chimeric oligonucleotide comprising PNA and DNA, are suitable for binding to a DNA target. Due to lower solubility than other analogues, PNA oligonucleotides longer than 20 bases are currently difficult to manufacture so PNA probes may be shorter and the discontinuities may be position closer to the splice junction, gene fusion boundary or polymorphism site of the target nucleic acid (for example, 3 to 4 bases rather than 8 to 12 bases). Chimeric DNA-PNA oligonucleotide probes may be longer.

When native DNA probes are used with symmetrical T_ms and artificial mismatches, a gap of one or nucleotides may be introduced between the TRS regions in an LCP or POP probe. This may be filled by a polymerase as discussed herein. In this situation, the probe termini will not meet exactly at the splice junction, gene fusion boundary or polymorphism site, but rather there will be a gap between the termini. Preferably the gap coincides with the splice junction, gene fusion boundary or polymorphism site, i.e. the probes or probe binds on either side of the splice junction, gene fusion boundary or polymorphism site. This approach may be useful in detecting splice junctions which often comprise single nucleotide polymorphisms. The above approach to probe design can be applied for more generic targets, i.e. the design of T_m symmetric probes or TRS's, with or without artificial mismatches, may be used for any target sequence whether or not it comprises a splice junction, gene fusion boundary or polymorphism site. Gap-filling

As described, above the TRS portions of an LCP, or a pair of oligonucleotide probes or an LCP in combination a gap oligonucleotide, may hybridise to a target sequence so that there is a gap between the TRS termini of the probe(s) when hybridised. This gap may be filled by extending a free 3' hydroxyl group at the 3' terminus of the probe using a polymerase and one or more nucleotide triphosphates, where at least one of the nucleotide triphosphates is modified to carry an azide function or an alkyne function. Alternatively or additionally, a larger gap may be filled by one or more gap oligonucleotides as disclosed herein.

The principles and procedures for gap-filling ligation are known in the art and have been used in the method of "gap LCR" (Wiedmann et al., "PCR Methods and Applications", Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, pages S51-S64, 1994; Abravaya et al., Nucleic Acids Res., 23(4):675-682, 1995; EP0439182). In the "gap LCR" processes described in those publications, the gap-filling methods are applied to allow the ligation of two independent nucleic acid probes but the method are equally applicable to the present invention.

Hybridisation of LCPs with gaps followed by gap-filling prior to ligation is advantageous as it provides higher stringency or specificity due to the multiple independent steps have to take place for correct closure of an LCP to form a CCP or for ligation of POPs. Since these independent steps are less likely to occur by chance, gap-filling offers a means for enhancing discrimination between closely related target sequences.

Polymerases for gap-filling according to the invention may be referred to as "gap-filling DNA polymerases" or "gap-filling polymerase". Suitable gap-filling DNA polymerases are discussed below but in brief when they extend the 3' end of a hybridised probe, they preferably do not displace the adjacent hybridised 5' end the same or different probe. When the gap between the two termini is only a single nucleotide, then only the correct expected nucleotide needs to be added to the polymerase gap-filling step to allow extension of the 3' terminus to fill the gap. Provided that the next base is not the same as the missing (gap) nucleotide, then most DNA polymerases can be used to fill the gap. This missing base is sometimes referred to as a "stop base". The use of "stop bases" in the gap-filling operation of LCR is described in EP0439182, for example. The principles of the design of gaps and the ends of flanking probes to be joined, as described in EP0439182, are in general applicable to the design of the gap spaces between the probe termini of the present invention. In the context of CuAAC-mediated ligation of oligonucleotides, the LCP or one of the probes in a POP pair will have an alkyne group or an azide group at the 5' terminus of the probe. The 3' terminus of the LCP or the second probe in a POP set will have a free hydroxyl group as shown in Fig. 2(a). a matching functional group can then be introduced by the gap-filling polymerase. For example, if an alkyne group is introduced into the 5' end of the LCP (see examples below for appropriate methods), the 3'-Azido-3'-deoxythymidine (AZT) triphosphate (Bioron GmbH, Ludwigshafen, Germany) can be used to introduce the azide group into the 3' terminus of the hybridised LCP if the gap between the 3'-TRS and 5'-TRS comprises an adenine base in the target nucleic acid strand, as shown in Fig. 2. The gap filling polymerase will insert the AZT- triphosphate into the gap rendering the LCP competent for CuAAC ligation (see Fig. 2(b)). At this point, the probes can be chemically ligated as shown in Fig. 2(c).

The corresponding 3'-azide triphosphates for adenine, guanine and cytosine can be readily prepared from their corresponding azide nucleosides using published methods ("Nucleoside Triphosphates and Their Analogs: Chemistry, Biotechnology, and Biological Applications", Editor: Morteza Vaghefi, CRC Press; 1 edition, 2005, ISBN: 1574444980). 3'-Azido-2',3'- dideoxyadenosine, 3'-Azido-2',3'-dideoxyguanosine are commercially available (Berry & Associates, Ml, USA). The corresponding cytidine nucleotide, 3'-Azido-3'-deoxycytidine, can also be prepared by published methods (Tanmaya Pathak, Chem. Rev. 102 (5): 1623-1668, "Azidonucleosides: Synthesis, Reactions, and Biological Properties" 2002; US 5,084,445).

With all four 3'-azido nucleotide triphosphates, SNP selective chemical ligation can be achieved. To identify an unknown nucleotide adjacent to a probe hybridisation site, four reactions may be set up for the same sample. One of the four 3'-azide triphosphates is added to each reaction along with a suitable polymerase in an appropriate buffer (usually supplied with the polymerase). The polymerase should extend the probe only where the azide triphosphate is the complementary base to the target nucleic acid nucleotide adjacent to the hybridised probe 3' terminus.

According to the invention, it is also possible to achieve SNP selective chemical ligation with only one modified nucleotide such as AZT triphosphate. This is discussed with reference to Fig. 3. To detect the four possible nucleotides at a known polymorphic site in a target nucleic acid, four different oligonucleotide probes are required. In the case of a POP pair, four different versions of one of the probe pair are required while for circularising probes four different LCPs are required. Each different probe is designed to hybridise to one of the possible nucleotides present at the polymorphic site, with different possible bases positioned at the 3' terminus of the probe. In Fig. 3 two exemplar pairs of probes (Probe 1 and Probe 2) are shown hybridising to a target nucleic acid bearing a polymorphic base, indicated in the figure. Two of the possible probes that could recognize the different bases that might be present at the polymorphic site are shown in Figs 3(a) and (d). The probes marked Probe 1 in Figs 3(a) and (d) have guanidine and adenine nucleotides at the 3' terminus respectively. Probe 1 from Fig. 3 (a) hybridises correctly to the cytosine in the target strand while Probe 1 in Fig. 3 (d) does not.

Certain types of proofreading gap filling polymerase will only extend from probes hybridised at the polymorphic nucleotide in the target nucleic acid if the probes are correctly hybridised. Such proofreading polymerases, which are also generally applicable to other aspects of the invention, include Vent DNA polymerase, Phusion DNA polymerase (available from New England Biolabs), Pfu DNA Polymerase, and Klenow Fragment polymerase (Fermentas)(Di Giusto DA & King GC, Nucleic Acids Res. 32(3):e32. "Strong positional preference in the interaction of LNA oligonucleotides with DNA polymerase and proofreading exonuclease activities: implications for genotyping assays." 2004). The polymorphic probes will be extended by the proofreading polymerase in the presence of the three normal dNTPs and the one azido-dNTP.

In Fig. 3 it can be seen that the CTP, GTP and ATP and AZT triphosphate are added with the polymerase. Probe 2 which has a 5' alkyne is designed to bind immediately downstream of the first adenine in the target nucleic acid strand. This means that if Probe 1 is correctly hybridised, as shown in Fig. 3(a), the polymerase will extend it by incorporating standard dNTPs until the azide modified nucleotide is incorporated, thus rendering the complex competent for CuAAC ligation. At this point Probe 1 and Probe 2 can be ligated together (or in the case of a circularising probe, both termini of the probe can be ligated generating a circular species). However, if Probe 1 is incorrectly hybridised at the polymorphic site, as shown in Fig. 3(d), then the polymerase will not extend, as shown in Fig. 3(e), so the complex will not be competent for ligation and will not be ligated after addition of copper catalyst, as shown in Fig. 3(f).

The Probe 1 sequences shown in the embodiment illustrated in Fig. 3 have a single LNA nucleotide at the position immediately next to the polymorphic base of the target nucleic acid. In general according to the invention, the use of an LNA nucleotide in an oligonucleotide probe at a position adjacent to the 3' terminal nucleotide renders the probe extremely resistant to degradation by the exonuclease activity of proofreading polymerases (Di Giusto DA & King GC, Nucleic Acids Res. 32(3):e32. "Strong positional preference in the interaction of LNA oligonucleotides with DNA polymerase and proofreading exonuclease activities: implications for genotyping assays." 2004). Exonuclease resistance is important in certain aspects of the invention where proofreading enzymes that degrade the 3' terminus of mismatched probes and then extend them are used, as this will rendering the assay unreliable. Exonuclease resistance may additionally or alternatively be achieved according to the invention by incorporation of phosphorothioate linkages in the probe (Di Giusto & King, 2003, supra). Additionally or alternatively, other modifications may be incorporated into the probe to provide exonuclease resistance, for example 2'-0-methyl RNA nucleotides, 2'-fluoro RNA nucleotides, phosphorodithioate modifications and boranophosphate backbone modifications, for example in the first and second nucleotides from the 3' end of the probe.

Exonuclease resistance may be employed in both the first and second probe of a POP pair. Similarly, if circularising probes are used with an intervening or gap oligonucleotide, the 3' termini of both oligonucleotides may be modified if proofreading polymerases are to be used. According to the invention, exonuclease resistance may be used to distinguish ligated probes from unligated probes. For example, if the 5' probe in a probe pair is designed without exonuclease resistance moieties at its 3'-terminus, while the 3' probe is modified at the 3" terminus with exonuclease resistant nucleotides, then correct ligation of the probe pair will produce a ligation product that is substantially exonuclease resistant at the 3' terminus. A suitable exonuclease, such as exonuclease III from E. coli, will degrade any unligated 5' probes after the ligation reaction.

When pairs of oligonucleotides hybridise with a target nucleic acid to leave a gap that is to be filled by a polymerase, the 3'-probe may be blocked at its 3' terminus. Such blocking may be effected by a phosphate group, for example as shown in Fig. 3 for the 3'-probe. Alternatively, a dideoxynucleotide could be incorporated into the 3' terminus of the probe. The 3' terminus of probes is also a suitable location for linking tags or immobilisation functions, such as biotin, if these are desired, as they will act as blocking groups. The oligonucleotides shown in Fig. 3 are very short and in practice much longer oligonucleotides are likely to be used, typically with a TRS of about 10 to 15 bases. Similarly, oligonucleotide probes are likely to comprise additional sequences that are not intended to hybridise to the target nucleic acid, such as primer sequences, address and probe identification sequences, as elaborated elsewhere herein.

An alkyne modified nucleotide triphosphate may be used to render a complex of two oligonucleotides on a target nucleic acid competent for chemical ligation. In this context, the second probe must comprise an azide at its 5'-terminus. The preparation of alkyne modified nucleotides is well known and a variety of alkyne-modified nucleosides is commercially available. The uridine nucleosides 5-(Propargyloxy)-2'-deoxyuridine and 5-ethynyl-2'- deoxyuridine are commercially available (Berry & Assoiciates, Inc.) and these can be readily modified to form the corresponding nucleotide triphosphates using methods known in the art ("Nucleoside Triphosphates and Their Analogs: Chemistry, Biotechnology, and Biological Applications", Editor: Morteza Vaghefi, CRC Press; 1 edition, 2005, ISBN: 1574444980). Uridine is an analogue of thymidine and will substitute for thymidine in most assays so the alkyne- modified nucleotide triphosphates discussed above can be used to introduce an alkyne at the 3' hydroxyl of a probe according to this invention in a template directed manner with a polymerase as discussed above. WO2008/120016 discloses alkyne-modified thymidine and adenine nucleosides (see for example Fig. 12 of WO2008/120016) that can be converted to the corresponding nucleotide triphosphate.

Where rolling circle amplification (RCA) is to be used according to the invention, gap-filling may be performed with a different DNA polymerase from the polymerase used for RCA, because the gap-filling polymerase may interfere with RCA. The gap-filling polymerase may alternatively or additionally be removed by extraction or inactivated thermally or with a neutralizing antibody prior to performing RCA. Such inactivation is analogous to the use of antibodies for blocking Taq DNA polymerase prior to PCR (Kellogg et al., Biotechniques 16(6): 1134-1137, 1994). Alternatively, if a polymerase that is not thermally stable is used, then it can be deactivated by elevating the sample temperature to 98°C for a few minutes prior to RCA. As illustrated in Figs 5(a) and (b), after hybridisation, gap-filling and ligation of LCPs to form CCPs, the CCPs (and any unreacted and incorrectly reacted LCPs) may be captured onto a solid phase support by a tethered oligonucleotide. The capture step may alternatively be performed with a biotinylated oligonucleotide, which can be subsequently captured onto an avidinated solid support. The gap-filling polymerase may then be removed by washing the solid support and disposing of the liquid phase. Similarly, if the target nucleic acid is captured onto a solid support, ligation of LCPs to form CCPs will leave the CCPs catenated with the target nucleic acid and thus locked onto the solid support. After ligation, both the gap-filling polymerase and unreacted LCPs can be washed away. This situation arises for example during in situ hybridisation where a target has been fixed in a tissue section. If the assays are being performed in situ on a tissue section, then the probes or probe may be captured on the tissue section, so washing away gap-filling polymerase prior to further assay steps is straightforward.

Azide and alkyne modified nucleotides

Fig. 4 illustrates a number of nucleotides that are commercially available or for which published synthesis methods are available. The nucleotides shown in Figs 4(a)-(c) are commercially available (Berry & Associates, Inc for all compounds in Figs 4(a) and (c)), while AZT triphosphate in Fig. 4(b) is available from Bioron, GmbH). Fig. 4(d) shows nucleotides with published syntheses but which are not yet known to be commercially available.

Preparation of 5' alkyne modified nucleotides and oligonucleotides

Natural thymidine can be converted to the corresponding oxetane compound using a published method (Horwitz, J. P.; Chua, J.; Urbanski, J.A.; Noel, M., J. Org. Chem. 28: 942-944, 1963). Similarly, the oxetane derivative can be converted to the 5' acetylene derivative shown in Fig. 4(d)(i) by reaction with Lithium Acetylide using a published procedure (M. Yamaguchi, Y. Nobayashi, Hirao I, Tetrahedron Lett., 24(46): 5121 -5122, "The alkynylation reaction of oxetanes" 1983; M. Yamaguchi, Y. Nobayashi, Hirao I, Tetrahedron, 40(21): 4261 -4266, "A ring opening reaction of oxetanes with lithium acetylides promoted by boron trifluoride etherate", 1984; Hiroyuki Isobe et al., Org. Lett., 10 (17): 3729-3732, "Triazole-Linked Analogue of Deoxyribonucleic Acid (^TLDNA): Design, Synthesis, and Double-Strand Formation with Natural DNA", 2008).

Similarly, the synthesis of the nucleotide shown in Fig. 4(d)(ii) has been published (El-Sagheer AH, Brown T, J Am Chem Soc. 131(11 ): 3958-64., "Synthesis and polymerase chain reaction amplification of DNA strands containing an unnatural triazole linkage." 2009). Both of the nucleotides shown in Figs 4(d)(i) and 4(d)(ii) can be converted to the corresponding 3'-phosphoramidate monomers, which can in turn be coupled to the 5' terminus of an oligonucleotide during automated oligonucleotide synthesis using published methods (El- Sagheer et al. 2009, supra).

Preparation of 3' alkyne modified nucleotides and oligonucleotides

Synthesis of the 3'-alkyne-modified thymidine analogue shown in Fig. 4(d)(iii) is published in Nuzzi et al. (QSAR Comb. Sci. 26 (11-12): 1191-1199, "Model Studies Toward the Synthesis of Thymidine Oligonucleotides with Triazole Internucleosidic Linkages via Iterative Cu(l)-Promoted Azide-Alkyne Ligation Chemistry." 2007). This can be converted to the corresponding nucleotide triphosphate using published methods ("Nucleoside Triphosphates and Their Analogs: Chemistry, Biotechnology, and Biological Applications", Editor: Morteza Vaghefi, CRC Press; 1 edition, 2005, ISBN: 1574444980). Similarly, the compound in Fig. 4(d)(iii) can be converted to the corresponding 5' phosphotriester monomer, which can in turn be coupled to the 3' terminus of an oligonucleotide during automated 5" to 3' oligonucleotide synthesis using the published method for conversion of AZT to a phophotriester (El-Sagheer et al. 2009, supra).

Preparation of 3' azide-modified nucleotides and oligonucleotides

AZT (available from various sources including SigmaAldrich) can be converted to the corresponding 5' phosphotriester monomer, which can in turn be coupled to the 3' terminus of an oligonucleotide during automated 5' to 3' oligonucleotide synthesis using published methods (El-Sagheer et al. 2009, supra). It is expected that the same methods used to modify AZT for incorporation into an oligonucleotide will be applicable to the corresponding cytosine, adenine and guanine nucleotides shown in Figs 4(d)(iv), 4(a)(ii) and 4(a)(iii) respectively. The synthesis of the cytosine-like nucleotide shown in Fig. 4(d)(iv) is published (Tanmaya Pathak, Chem. Rev. 102 (5): 1623-1668, "Azidonucleosides: Synthesis, Reactions, and Biological Properties" 2002; US 5,084,445).

Preparation of 5' Azide-modified nucleotides and oligonucleotides

5'-Azido-5'-deoxythymidine shown in Fig. 4(a)(iv) is commercially available (Berry & Associates, Inc). This nucleotide can be converted to the corresponding 3'-phosphoramidate monomer, which can in turn be coupled to the 5' terminus of an oligonucleotide during automated oligonucleotide synthesis using the published methods for 5'-alkyne nucleotides (El-Sagheer et al. 2009, supra).

Multiple ligation reactions

The quantity of covalently circularised probe (i.e. CCP) may be increased by repeating the CuAAC reaction step and/or any gap filling reaction steps according to the methods of the invention one or more times. This may be achieved by dehybridising (or melting) ligated probes from their target nucleic acid after each ligation step. Hybridisation may then be repeated by restoring hybridisation conditions. Additional (fresh) probe may be added or if a sufficient excess of unligated probe is present this will tend to hybridise in preference to the previously ligated probes.

Repeating the ligation and/or gap filling reaction steps is feasible with POPs and LCPs despite the fact that CCPs become concatenated with their target. When these reaction steps are repeated with LCPs, it is possible that under appropriate conditions the same target sequence will mediate closure of multiple LCPs to form CCPs, as the CCPs can become threaded on the target molecule. This is because the CCPs may move, or wander, to some extent along the target molecule during the dehybridising step, making the target sequence available for a renewed hybridisation by a non-circularised probe.

If non-hybridising conditions are to be used to separate CCPs from unreacted and incorrectly ligated LCPs, and if multiple probe hybridisation and closure cycles are to be used, it may be necessary to use a target nucleic acid molecule that is reasonably large. Also, the target nucleic acid sequence to be detected will suitably at a sufficient distance from the ends of the target molecule that the CCPs remain linked to the target molecule. For practical purposes, the target sequence of the target molecule may be a distance of at least about 200 base pairs from the nearest terminus depending on whether and/or how the target sequence is bound to a solid phase support. If the target sequence is free in solution, a longer distance may be required, especially in the case of long-lasting denaturing washes.

If the target nucleic acid is sufficiently large, the CCP molecule will remain linked to the target molecule even under conditions that would release or degrade any hybridised non-cyclised LCPs. This is one way in which a circularisation reaction produces a selectively detectable species, indicating the presence of the target nucleic acid in a sample. Conditions that will denature or degrade a hybridised but non-cyclised probe include heat, alkali, guanidine hydrochloride, urea and other chemical denaturants or exonuclease activity, the latter degrading the free ends of any unreacted LCPs.

Directly labelled circularising probes

Figs 5, 7, 8, 10, 12, 13 and 14 schematically show aspects of this invention in which directly labelled probes are used. In Fig. 10, a method for resolving correctly closed circularised probes (CCPs) from unreacted probes (LCPs) is shown. In this figure, the method is shown for two probes that recognise different alleles of a single target nucleic acid sequence (also referred to herein as "target sequence" or "target"). Each probe, designated Circularising Probe 1 (also termed "LCP1" herein) and Circularising Probe 2 (also termed "LCP2" herein), is covalently linked to and identified by a unique tag, Tag 1 and Tag 2, respectively. Upon contacting the target sequence with LCP1 and LCP2, only LCP1 is capable of hybridising with the target to form a ligatable complex and so in the presence of ligase only LCP1 is ligated to form CCP1. The unreacted LCP2 and any remaining LCP1 can then be degraded by exonuclease activity while CCP1 is protected by virtue of being circular.

According to the invention, gene 6 exonuclease of phage T7 provides a useful tool for the elimination of excess LCPs and any unreacted gap oligonucleotides. This exonuclease digests DNA starting from the 5'-end of a double-stranded structure. It has been used successfully for the generation of single-stranded DNA after PCR amplification (Holloway et al., Nucleic Acids Res. 21 :3905-3906 (1993); Nikiforov et al., PCR Methods and Applications 3:285-291(1994)). Alternatively, E. coli Exonuclease I digests DNA starting from the 3'-end of single-stranded DNA. If a "capture" or immobilisation sequence is incorporated into an LCP design, the surviving CCP can be captured onto a solid phase support, as illustrated in Fig. 10b. The support can then be washed and exonuclease digested LCP2 and unreacted LCP1 , which cannot hybridise to the solid support, can be separated from the captured CCP1. After washing away LCP2 and its corresponding tag(s), the tag(s) on CCP1 can be cleaved from the CCP molecule. If according to the invention a tag is linked to a probe via a trypsin-cleavable linkage, the tag may be cleaved by trypsin into solution. The solution containing cleaved tags may then be injected into a mass spectrometer for detection of the tags, as illustrated in Fig. 10b.

Although only two tags have been shown in the schematic diagram in Figs 10a and b, many thousands of different LCPs of the present invention can be used together, as has been demonstrated in principle previously (Hardenbol et al., 2003, supra). In a further aspect of the invention, shown schematically in Fig. 13, a tagged LCP comprises a cleavable group. The cleavable group (denoted "C" in Fig. 13a) may be positioned between the tag and the portion of the LCP that will allow it to be captured onto the solid support (i.e. the immobilisation function). In Fig. 13a, an immobilisation function in the form of a capture sequence is present allowing the LCP to be captured by hybridisation to a tethered or biotinylated oligonucleotide. It would additionally or alternatively be possible to directly biotinylate the LCPs to create an immobilisation function. The presence of a cleavable group means that a CCP may be cleaved after its formation from an LCP. The cleavage step may take place before or after the CCPs are captured onto a solid phase support. In Fig. 13b, cleavage is shown taking place before the capture step. The cleavage step ensures that the tagged portion of any unreacted LCPs is not retained on the solid support, as a tagged portion of the LCP is linked to the capture sequence by the cleavable group. The ligation of LCPs to form CCPs means that the tag is linked through the ligated portion of the probe so that after the cleavage step the tag remain linked to the immobilisation function-containing part of the probe (here, a capture sequence or biotinylated portion of the probe). In this way, tags will only be captured for correctly closed CCPs allowing the tags from unreacted LCPs to be washed away, as illustrated in Fig. 13b.

A cleavable group may be a type IIS restriction endonuclease recognition sequence, in which case the immobilisation function such as a capture sequence may also serve as the cleavage site by providing the restriction sequence. In this situation, a tethered or biotinylated oligonucleotide may be hybridised with the LCPs and CCPs prior to cleavage to form a double stranded substrate for the restriction endonuclease. Alternatively, the cleavable group may be chemically cleavable. Replacement of one of the phosphodiester linkages in the backbone of an LCP with 3'-(N)-phosphoramidate or a 5'-(N)-phosphoramidate, results in a linkage that is more susceptible to acid hydrolysis than the rest of the probe. Alternatively, a uracil residue can be incorporated into the phosphodiester backbone. This residue is a substrate for the enzyme uracil deglycosylase, which depurinates this residue. The depurinated residue is then much more susceptible to hydrolysis than the rest of the probe molecule. As a further alternative, the probe can be constructed with nucleotides that are resistant to degradation by an exonuclease leaving only one or a few native nucleotides with natural phosphodiester linkages that can be cleaved by a DNAse or other nuclease.

Where the LCP of the invention is used for in situ hybridisation, for example, an immobilisation function in the form of a capture sequence or biotin may not be necessary for use with a cleavable group in an LCP because ligation to a target nucleic acid fixed in a tissue section, will leave the resultant CCP concatenated with a fixed target. After correct ligation, the ligated probe will have a higher binding affinity for the target than the unligated probe (or any pair of TRS's). Cleavage of the cleavable group will allow unligated sequences to be washed away more easily than the ligated probes.

Indirect detection of circularising probes

In another aspect, each different probe of the invention may comprises a unique probe identification (PI) sequence by which it can be identified, for example through hybridisation with an appropriate probe detection sequence ("PDS").

The probe detection sequence may be labelled with any appropriate label or tag (see below). Preferred label or tags include fluorophores and mass tags. PI sequences may be incorporated in an intermediate region of an LCP or other probe. Each PI sequence should uniquely identify a corresponding LCP. The PI sequence allows detection by a corresponding PDS such as a tagged PDS. The PI sequences, when amplified during RCA for example, may result in tandemly repeated sequences that are complementary to the sequence of a tagged PDS probe. It may be desirable to have two or more PI sequences on an LCP or other probe as this will, for example, increase a detection signal from correctly hybridised tagged PDS probes. There is no theoretical limit to the number of PI sequences that can be present in an LCP or other probe except the practicality of synthesising and using very large probes comprising large numbers of PI sequences. When there are multiple PI sequences on a single probe, they may have the same sequence or they may have different sequences with each different sequence complementary to a different PDS probe. A probe such as an LCP may suitably containing PI sequences that have the same sequence such that they are all complementary to a single PDS probe. The PI sequences can each be any length that supports specific and stable hybridisation between the PI sequences and PDS probes. For practical purposes, a length of 10 to 35 nucleotides is suitable, for example a length of 15 to 20 nucleotides.

Similarly, the PDS's may have a complementary sequence whose length is similar to the PI sequences. However, the PDS may comprise additional non-hybridising sequence, for example to which one or more tags may be linked. The PDS may be a branched oligonucleotide. For example, the PDS may comprise multiple sequences complementary to its PI sequence, in addition to comprising a tag. Such a PDS may be in the form of a Y-shaped oligonucleotide of a structure described by Suzuki Y. et al. (Nucleic Acids Symp Ser. 2000;(44): 125-126, "Synthesis and properties of a new type DNA dendrimer.") and comprising three copies of the PDS. A second Y-shaped branched oligonucleotide comprising three copies of the PI sequence when added to the tripartite PDS probe will assemble a dendrimer in which very large numbers of copies of the PDS, and consequently its associated mass tag, will be present. If the tripartite PDS sequence is present in excess, then the dendrimer will have free PDS sequences available for hybridisation to the PI sequences present in correctly circularised CCPs, for example. In this way signal amplification can be achieved without amplifying the target nucleic acid or CCPs.

Fig. 11 illustrates an aspect of the invention in which LCPs are identified after circularisation by the ability of the resultant CCPs to be selectively amplified by RCA. In Fig. 11a, a schematic of a method of detecting DNA sequence variants is illustrated in which a pair of LCPs that identify different alleles of a DNA sequence is used. The LCPs in this assay are identifiable by their unique Probe Identification sequences. In Fig. 11a, a preferred embodiment of the invention is illustrated for a pair of probes that detect different variants of a single target molecule. In the first step, the pair of LCPs are contacted with their target nucleic acid sequence. Only one of the LCPs matches the target sequence correctly and hybridises to form a duplex, so that in the next step ligation only occurs at this correctly hybridised duplex converting the LCP into a CCP. This circular sequence is now a substrate for RCA, depicted in Fig. 11b.

In some embodiments of this aspect of the invention, the unreacted LCPs may be degraded by exonuclease at this stage (not shown in Figs 11 ).

In the next step, the CCP is hybridised to a captured primer to form a CCP/primer duplex. A polymerase may then extend the primer generating a tandem repeated sequence complementary to the CCP where the tandemly repeated complement is captured on a solid phase support such as a bead. In alternative embodiments, the primer sequence may be biotinylated rather than linked directly to a bead or other solid phase support. Here, the biotinylated product of the linear extension of the primer can then be captured onto an avidinated solid phase support after the extension reaction. Captured tandem repeat sequences contain the complement of PI sequence(s) present in the LCP sequence. In the final steps of the method as depicted in Fig. 11(b), these complements of the PI sequences are probed with mass tagged PDS's. Since the targets of the PDS probes are captured on a solid phase support, the correctly hybridised PDS probes will be captured onto the support by the hybridisation reaction, allowing unhybridised PDS probes to be washed away. After washing away unhybridised PDS probes, the mass tags on the correctly hybridised PDS probes can be cleaved off for (or during) detection by mass spectrometry.

Microarravs and suspension arrays

A probe of the invention may comprise a PI sequence that is complementary to a PDS on or contactable with a solid support such as a microarray. In embodiments where the PDS is located or locatable on an array, it may be referred to as a microarray address sequence. A microarray address sequence may have a sequence that is complementary to an oligonucleotide at a specific discrete location on a planar array.

Alternatively, other array formats can be used to detect probes of the invention such as suspension arrays (Fulton RJ et al., Clin Chem 43:1749-1756, "Advanced multiplexed analysis with the FlowMetrixT system." 1997; Eriksson R et al., J Microbiol Methods. 78(2):195-202, "Multiplex and quantifiable detection of nucleic acid from pathogenic fungi using padlock probes, generic real time PCR and specific suspension array readout." 2009). A suspension array is related to a microarray in that an array of oligonucleotide address sequences is present on a solid support. The support may be a microsphere that is uniquely identifiable by the presence of two or more fluorescent dyes present at unique ratios in the microspheres. Thus each address sequence is linked to a unique microsphere that can be identified in fluorescence activated flow sorter. In these embodiments, the probes or amplification products of ligated probes may be labelled while also comprising a PDS or microarray address sequence.

Where directly labelled probes according to the invention are used, it is possible to hybridise probes that form as a result of template mediated ligation to oligonucleotides on a microarray or on microspheres. The microarray address sequence of the probe will ensure that each ligated probe or probe amplicon (such as from RCA) hybridises to a discrete location on the microarray.

If mass tags are used to label the probe or probe amplicon, a combination of a distinct microarray address sequence and mass tag can be used to designate a large number of probes that will then be uniquely identifiable by a unique combination of theses features. For example, 1000 discrete microarray address sequences, corresponding to 1000 discrete locations on a microarray, combined with 400 distinguishable mass tags, will allow 400 000 different probes to be uniquely identified in a single assay, thereby providing a high level of multiplexing in a single assay.

Similarly, mass tagged probes according to the invention may be combined with flow sorting using a suspension array. Thus a library of probes may be divided into subsets by using PI sequences that allow capture onto fluorescently encoded microbeads, for example. Further subdivision can then be effected by directly labelling the probes with mass tags or a further probe identification sequence. The fluorescent microbeads may then be used to sort a library of probes into subsets that may be analysed further with mass tags or other means.

In alternative aspects where LCPs are detected through a PI sequence which is distinct from the microarray address sequence, the microarray address sequence may be used to ensure that subsets of CCPs in a library of CCPs hybridise to distinct locations on an array. After hybridisation, the correctly hybridised microarray probe sequence can be extended using an appropriate polymerase to effect RCA of the hybridised CCPs. Thus, the microarray address sequence may also act as a binding site for a rolling circle primer (RCP; see also below) which may be immobilised at a discrete location on a planar array surface. In this way, a spatially resolved "captured library" of CCP sequences can be generated. The captured library may then be probed by hybridisation with PDS sequences that recognize the PI sequence complements generated by the RCA that takes place at each array location.

One aspect the invention provides a microarray having discrete locations (for example, about 100 to 1000 discrete locations), each location comprising a microarray address sequence complement which is capable of binding a microarray address sequence located on a probe according to the invention. Each microarray address sequence complement may be unique. In another aspect, the invention provides a kit comprising such a microarray together with a probe or probe pair of the invention where each probe comprises a microarray address sequence which is capable of hybridising to a microarray address sequence complement in the microarray.

A microarray according to the invention may comprise an array of wells or microtitre plates, for example, such that each well contains a single immobilised (or capture) oligonucleotide that is a member of the array. A sample containing ligated probes or amplicons (such as from RCA of CCPs) of this invention may be added to each well and allowed to hybridise to the immobilised oligonucleotide located in the well. After a predetermined time, unhybridised probes or amplicons may be washed away. Various detection techniques may then be applied. Fluorescently labelled probes or amplicons can be detected directly in a fluorimeter. The probe may contain a second PI sequence as defined above, in which case a labelled probe detection sequence may be applied to the captured probes in a form of sandwich assay. Alternatively, the probe may contain a mass tag, in which case the hybridised probe may be melted off the capture oligonucleotide. Released probes may then be loaded into a capillary electrophoresis mass spectrometer or they can be injected into the ion source of an electrospray mass spectrometer. Alternatively, if the mass tag comprises or is linked to the probe via a photocleavable linker, the mass tag can be cleaved and desorbed directly by laser in a MALDI instrument.

A microarray of the invention may be synthesised combinatorially on a glass "chip" according to the methodology of Southern or that of Affymetrix, Santa Clara, California ( see for example: A.C. Pease et al. Proc. Natl. Acad. Sci. USA. 91 , 5022 - 5026, 1994; U. Maskos and E.M. Southern, Nucleic Acids Research 21 , 2269 - 2270, 1993; E.M. Southern et al, Nucleic Acids Research 22, 1368 - 1373, 1994) or using related ink-jet technologies such that discrete locations on the glass chip are derivatised with one member of the hybridisation array.

Branched oligonucleotides

Oligonucleotides probes are typically linear polymers of nucleotides but for many of the applications of the invention, there are advantages in using branched oligonucleotides. Introduction of branched structures into nucleic acids, produces Y-shaped and comb-shaped branched structures (see for example Reese C.B. & Song Q., Nucleic Acids Res. 27(13):2672 - 2681 , "A new approach to the synthesis of branched and branched cyclic oligoribonucleotides." 1999; Horn T. et al., Nucleic Acids Res. 25(23):4835 - 4841 , "An improved divergent synthesis of comb-type branched oligodeoxyribonucleotides (bDNA) containing multiple secondary sequences." 1997; Braich R.S. & Damha M.J., Bioconjug Chem. 8(3):370 - 377, "Regiospecific solid-phase synthesis of branched oligonucleotides. Effect of vicinal 2', 5'- (or 2',3'-) and 3', 5'- phosphodiester linkages on the formation of hairpin DNA." 1997; Horn T. & Urdea MS., Nucleic Acids Res. 17(17):6959 - 6967, "Forks and combs and DNA: the synthesis of branched oligodeoxyribonucleotides." 1989), while iteration of a branching process will result in a dendrimeric probe. Branched oligonucleotides may be used to enable signal amplification without resorting to nucleic acid amplification, particularly comb-oligonucleotides in which a primary, sequence- specific, linear oligonucleotide probe is linked to a series of secondary oligonucleotides, which carry one or more labels (Horn T. et al., Nucleic Acids Res. 25(23):4842-4849, "Chemical synthesis and characterization of branched oligodeoxyribonucleotides (bDNA) for use as signal amplifiers in nucleic acid quantification assays." 1997).

The prior art describes various methods to produce branched oligonucleotides. In particular, phosphoramidate reagents that introduce two or more branch points into oligonucleotide that allow extension of the oligonucleotide by standard oligonucleotide synthesis (M.S. Shchepinov et al., Nucleic Acids Res 25: 4447-4454, "Oligonucleotide dendrimers: synthesis and use as polylabelled DNA probes." 1997; M.S. Shchepinov et al., Nucleic Acids Res 27: 3035-41 , Oligonucleotide dendrimers: stable nano-structures." 1999) are commercially available (for example the "Symmetric Doubler Phosphoramidite" from Glen Research Corporation, Sterling, Virginia, USA). These linkers however are limited by steric factors on the solid phase synthesis resin. Controlled Pore Glass, the typical substrate, is available in various pore sizes and it is the pore size that imposes the most significant constraint.

Thus, there are advantages to be gained by synthesizing branched oligonucleotides as individual branches by automated solid phase synthesis and then assembling the final branched structure in the solution phase. However, to achieve efficient coupling without side reactions is not trivial. The CuAAC reaction offers an excellent mechanism for achieving this goal.

The probes of the invention may have a comb structure which comprises a primary sequence including the TRS or TRS's and secondary oligonucleotides branched off the primary sequence. These secondary oligonucleotides may for example all comprise an identical sequence. The secondary oligonucleotides may comprise or act as the PI sequence. In the case of LCPs, after circularisation of the primary sequence and removal of unreacted probes, the CCP circularised sequence may be probed with mass tagged probe detection sequences. Since the comb structure allows multiple PI sequences to be incorporated into the probe, signal amplification without amplification of the target sequence or the probe sequence is achievable.

Alternatively, probes may have a comb structure as defined above but where the secondary oligonucleotides are directly labelled or tagged. In this way multiple tags can be incorporated into the probes of the present invention. Where pairs of probes are used according to the invention, one of the pair of probes may comprise a comb oligonucleotide, where the secondary oligonucleotides comprise either address sequences or are directly tagged (for example, with one or more mass tags). The other probe of the probe pair may be untagged may comprise an immobilisation function, such as biotin, a cross-linker and/or a microarray address sequence for solid phase capture onto beads or a microarray.

Captured libraries

Although only two tags have been shown in the schematic diagram in Figs 10, many thousands of different LCPs can be used together as has been demonstrated in principle previously (Hardenbol et al., 2003, supra). If multiple probes are used in a method shown in Fig. 10, the result of the RCA step in which the circularised probes sequences are copied onto beads generates a "captured library" of circularised probes that represents information in the probed sample. Captured libraries have a number of advantages. After appropriate washing steps the library can be archived for future analysis. In addition, the library can be probed multiple times with the same mass tagged PDS probes to give signal amplification. In some embodiments, the captured library may be probed in multiple sequential assays rather than in a single step using multiple distinct libraries of tagged PDS probes. In this way the same tags can be used to detect different PI sequences in the captured library. For the purposes of archiving captured libraries, it may be desirable to synthesise the captured libraries with exonuclease resistant nucleotide analogues that are compatible with polymerases such as boranophosphate nucleotides! or alpha-thio deoxynucleotide triphosphates.

For long term storage, it may be preferable to generate captured libraries with covalently tethered oligonucleotides rather than with biotinylated oligonucleotides that are later captured onto avidinated beads to avoid the risk of sample loss by dissociation of the non-covalent biotin/avidin complex.

Rolling circle amplification (RCA)

As discussed above, LCPs are useful particularly for in situ detection of nucleic acid targets in tissue sections. Not only do they lock on to their targets after ligation and formation of CCPs, permitting stringent washing conditions to be applied, but they may also be extended in situ by RCA (also referred to as rolling circle replication) using the fixed target nucleic acid as a primer (Stougaard et al., 2007, supra). It has been shown that a variety of polymerase enzymes will copy DNA with triazole linkages in the DNA backbone (El-Sagheer et al. 2009, supra), so this chemistry is envisaged to work with RCA. In certain aspects of the invention, RCA is applied to CCPs generated by target mediated ligation of LCPs. To effect RCA, a circular single-stranded CCP DNA molecule is contacted with one or more rolling circle primers (RCPs) that hybridise to primer binding site(s) in the CCPs. Extension of the RCPs by a strand displacing polymerase results in tandem repeats of the complement of the CCP sequence, as illustrated in Fig. 11(a) and (b). It can be seen from this figure that in some embodiments the RCP may be immobilised on a solid phase support or be capable of being immobilised on a solid support after extension and RCA of hybridised CCPs, for example by using a biotinylated RCP. Alternatively, a target nucleic acid such as a messenger RNA molecule fixed in a tissue section may be a suitable target for RCA. Fig. 11(a) and (b) shows a schematic of a method of the invention comprising the following steps:

(a) mixing one or more LCPs with a target nucleic acid under conditions promoting hybridisation, resulting in LCP-target duplexes;

(b) if necessary filling any gaps between the LCP termini;

(c) ligating the termini of the LCP (and optional gap oligonucleotides, if present) to generate a CCP;

(d) contacting an RCP or generating a primer from the target nucleic acid under conditions that promote hybridisation with the CCP, resulting in an RCP-CCP duplex;

(e) contacting the RCP-CCP duplex with a DNA polymerase under conditions promoting extension of the RCP to produce the complement of the CCP sequence, such that continuous extension of the RCP (via RCA) results in formation of tandem repeats of the complement of the CCP sequence.

Although Fig. 11(a) and (b) shows an embodiment in which only two LCPs are present, multiple different LCPs may be present in a single reaction. Those LCPs that are ligated to form CCPs will be able to support RCA and may thus generate captured tandem repeats of their complement on a solid support. The solid support bound complement sequences for a number of different CCPs are referred to herein as a captured library. The TRS's of the LCP may hybridise to the target nucleic acid sequence, with or without a central gap to be filled by one or more gap nucleotides or oligonucleotides.

A target nucleic acid may be fixed in a tissue section, in which case the target can be used to prime the RCA reaction. The CCP then acts as a template for RCA. If the target is RNA and the probe hybridises near the 3' terminus of the target, a suitable polymerase such as Phi29 will degrade any unhybridised bases back to the duplex formed by the probe with the target and will initiate RCA once the duplex of RNA/probe is reached, i.e. when the target nucleic acid is priming the circularised probe. Alternatively, the target nucleic acid may be cleaved by incorporating the binding site for a type IIS restriction endonuclease like Fokl into the probe ( see for example: US 6,558,928; Lagunavicius A et al., RNA. 15(5):765-71 , "Novel application of Phi29 DNA polymerase: RNA detection and analysis in vitro and in situ by target RNA-primed RCA." 2009). If DNA detection in situ is desired, a related technique where a conventional restriction type II endonuclease is used to cleave the target DNA after ligation providing a priming site for in situ RCA (Lohmann JS et al., BMC Mol Biol. 8:103, "Detection of short repeated genomic sequences on metaphase chromosomes using padlock probes and target primed rolling circle DNA synthesis." 2007).

In other examples, target DNA can be cleaved with a type II restriction endonuclease and an exonuclease to render chromosomal DNA accessible before hybridisation with a probe of this invention. This may be followed by use of a polymerase such as Phi 29, whose 3'-to-5' exonuclease activity will degrade unwanted single stranded target DNA until a duplex with the probe is reached, i.e. so that the target can act as a primer enabling RCA (Larsson C et al., Nat Methods 1(3):227-232, "In situ genotyping individual DNA molecules by target-primed rolling- circle amplification of padlock probes." 2004).

Where the invention employs RCA, each LCP may comprise a primer binding sequence (PBS) if an exogenous primer is to be provided. The PBS is complementary to at least a part of the rolling circle primer (RCP). An RCP is an oligonucleotide having sequence complementary to the primer binding sequence of an LCP or CCP. This complementary sequence of the RCP is also referred to herein as the complementary portion of the RCP. Each LCP should have at least one PBS, although if the LCPs are small, i.e. less than 100 nucleotides in length then only a single PBS may be present. This allows RCA to initiate at a single site on CCPs. The PBS and the corresponding RCP can have any desired sequence as long as they are complementary to each other. In general, the sequence of the PBS and the RCP should be chosen so that they are not significantly similar to any other portion of the LCP or any other LCP to be used simultaneously. The PBS can be any length that supports specific and stable hybridisation between the PBS and the RCP. For this purpose, a length of 10 to 35 nucleotides, for example 16 to 20 nucleotides, is suitable for the PBS and the complementary portion of the RCP. The PBS can be positioned anywhere in an LCP, for example between terminal TRS's. The PBS may be positioned adjacent to the 5' TRS, with the TRS and the PBS for example separated by three to ten, such as six, nucleotides. This position prevents the generation of any other spacer sequences, such as detection tags and secondary target sequences, from unligated LCPs during DNA replication. An RCP may comprise an additional sequence located at the 5' end of the RCP that is not complementary to any part of the LCP or CCP. This additional sequence may be referred to as the displacement region of the RCP. The displacement region serves to facilitate strand displacement during RCA. The displacement region may be a short sequence such as 4 to 8 nucleotides in length. The displacement region provides an unhybridised region of already displaced sequence that assists the strand displacing polymerase to start displacing the extended RCP in RCA.

In some embodiments of the invention where RCA is used, gene 6 exonuclease of phage T7 may be added after the ligation reaction together with the DNA polymerase to be used to effect RCA. To protect the RCA product from degradation, the RCP may comprise one or more phosphorothioate linkages at the 5' end, to render the RCP and its extension products resistant to the exonuclease (Nikiforov et al., 1994, supra). The exonuclease will degrade excess LCP molecules as they can become associated with the RCA DNA product and interfere with hybridisation of PDS probes. The use of exonuclease digestion is a suitable step for eliminating unreacted LCPs and gap oligonucleotides, for example.

Features of RCA as discussed in the prior art introduction are applicable to the present invention. Polymerase chain reaction (PCR) amplification of liqated probes

As noted above, it has been shown that some polymerase enzymes will copy DNA with triazole linkages in the DNA backbone (El-Sagheer et al., 2009, supra), meaning that this chemistry may be used with RCA and PCR. Accordingly, in another aspect of the invention, correct closure of LCPs to form CCPs or ligation of a POP pair is detected by PCR. Here, the LCP or the individual probes in a POP pair comprise a pair of PCR primer binding sequences (PPBS). In an LCP, the PPBS sites may be oriented so that a first primer which hybridises with one of the PPBS sites copies across the ligation junction that is formed where the LCP is converted to the CCP by target mediated ligation. A second primer which hybridises with the other PPBS site will not become accessible to the primer unless the correct ligation event has taken place. Similarly, with a POP pair, the PPBS may be located at the termini of the probes that are distal from the ligation site, so that the first primer must bind to its PPBS site in the first probe and then copy across the ligation junction to copy the PPBS site in the second probe and make this site accessible to the second primer.

Fig. 12a-c illustrates a PCR-based method for the detection of probe circularisation (i.e. a CCP) using mass tags. This figure illustrates the method a pair of probes but in practice multiple different probes could be used simultaneously. As shown in Fig. 12a, in the first stage of the method a pair of LCPs are hybridised with the target nucleic acid. Ligation leads to closure (i.e. circularisation) of only one correctly hybridised probe. The probes are captured onto a solid phase support by an oligonucleotide that also comprises a restriction site for a type II restriction endonuclease. Cleavage of the captured probes by the endonuclease results in the formation of a linear structure in which parts of the LCP sequence have been rearranged. A similar process, using uracil degiycosylase to cleave the circularised probes, is described by Hardenbol et al. (2003, supra) and is referred to as "molecular inversion". This results in the PPBS sites on the probe being in the correct orientation to enable exponential amplification of the CCPs only in the rearranged probes that have been correctly ligated by target mediated ligation. In Fig. 12b, the primer sequences are added along with mass tagged PI complement sequences. PCR is then effected with a thermostable polymerase with 5' to 3' exonuclease activity, which will release mass tags from correctly hybridised PI complement sequences during the PCR reaction as shown in Fig. 12c. After the PCR reaction, released mass tags can be analysed by mass spectrometry.

With ligation of two oligonucleotide probes, there is no need for the second cleavage step described above. The capture step would also be optional. After ligation, polymerase, primers and nucleotides may be added and PCR may be initiated without any isolation of ligated probes as long as the ligation buffer is compatible with PCR. If the ligation buffer is not compatible, then a purification step may be conducted. Ligated products can be separated from unligated products by size if desired or by capture using an immobilisation function such as a capture sequence in the probe or through biotinylation of one of the probes. DNA polymerases

When using a DNA polymerase to fill gaps according to the invention, strand displacement by the DNA polymerase is generally not suitable. Similarly, 5'-to-3' exonuclease activity is generally not suitable. However, 3'-to-5' exonuclease activity, also referred to as proof-reading activity, may be desirable in some instances. As noted above, such DNA polymerases are referred to herein as "gap-filling DNA polymerases" or "gap-filling polymerases". Thermostability of gap- filling polymerases is another feature that may be useful but is not essential to all embodiments of the invention. In the context of this invention, the gap-filling polymerase is able to incorporate azide and alkyne modified nucleotides. A wide variety of polymerases have been engineered to incorporate modified nucleotides so it is expected that many commercially available polymerases will be able to incorporate the azide and alkyne modified nucleotides. It is known for example that 3' azido-nucleotides are incorporated by a variety of polymerases including viral reverse transcriptases (Reardon JE, J Biol Chem. 265(33):20302-7, "Human immunodeficiency virus reverse transcriptase. Substrate and inhibitor kinetics with thymidine 5'-triphosphate and 3'-azido-3'-deoxythymidine 5'-triphosphate." 1990) and human polymerases (Copeland WC et al. J Biol Chem. 267(30):21459-64, "Human DNA polymerases alpha and beta are able to incorporate anti-HIV deoxynucleotides into DNA." 1992). Suitable gap-filling DNA polymerases for use in the invention include T7 DNA polymerase (Studier et al., Methods Enzymol. 185:60-89 (1990)), DEEP VENT™ DNA polymerase (New England Biolabs, Beverly, Mass.) and T4 DNA polymerase (Kunkel et al., Methods Enzymol. 154:367-382 (1987)). A particularly suitable type of gap-filling DNA polymerase is the Thermus flavus DNA polymerase (MBR, Milwaukee, Wis.). Another suitable gap-filling DNA polymerase is the Stoffel fragment of Taq DNA polymerase (Lawyer et al., PCR Methods Appl. 2(4):275-287 (1993), King et al., J. Biol. Chem. 269(18):13061-13064 (1994)). As mentioned above, gap-filling polymerases may also be proof-reading polymerases (Cline J et al., Nucleic Acids Res. 24(18): 3546-51 , "PCR fidelity of pfu DNA polymerase and other thermostable DNA polymerases." 1996) of which the Pyrococcus furiosus and Vent polymerases (Stratagene and New England Biolabs respectively) are suitable. Similarly, the E. coli Klenow fragment has proof-reading activity (DNA Polymerase I, Large (Klenow) Fragment, New England Biolabs), which is also known to incorporate AZT as a chain terminator so is likely to accept other related azido-nucleotides (Elwell LP et al., Antimicrob Agents Chemother. 31(2):274-80, "Antibacterial activity and mechanism of action of 3'-azido-3'-deoxythymidine (BW A509U)." 1987). Reverse transcriptases may be used as gap-filling polymerases according to the invention if RNA target nucleic acids are to be detected.

Another type of DNA polymerase may be used for RCA. Suitable polymerases must perform rolling circle replication of primed single-stranded circles. Such polymerases are referred to herein as rolling circle DNA polymerases. For RCA, it is preferred that a DNA polymerase be capable of displacing the strand complementary to the template strand, termed strand displacement, and lack a 5' to 3' exonuclease activity. Strand displacement results in synthesis of multiple tandem copies of the ligated CCP. Any 5' to 3' exonuclease activity may result in the destruction of the synthesized strand.

DNA polymerases for use in the invention methods may be highly processive. The suitability of a DNA polymerase may be readily determined by assessing its ability to carry out RCA. Suitable rolling circle DNA polymerases include bacteriophage Phi29 DNA polymerase (US5, 198,543 and US5.001 ,050), phage M2 DNA polymerase (Matsumoto et al., Gene 84:247 (1989)), phage PhiPRDI DNA polymerase (Jung et al., Proc. Natl. Acad. Sci. USA 84:8287 (1987)), VENT™ DNA polymerase (Kong et al., J. Biol. Chem. 268:1965-1975 (1993)), Klenow fragment of DNA polymerase I (Jacobsen et al., Eur. J. Biochem. 45:623-627 (1974)), T5 DNA polymerase (Chatterjee et al., Gene 97:13-19 (1991 )), PRD1 DNA polymerase (Zhu and Ito, Biochim. Biophys. Acta. 1219:267-276 (1994)), and T4 DNA polymerase holoenzyme (Kaboord and Benkovic, Curr. Biol. 5:149-157 (1995)).

A suitable polymerase is the exonuclease(-) BST thermostable DNA polymerase available from New England Biolabs (Mass, USA). Bacillus stearothermophilus (BST) is a thermophilic bacterium whose polymerase is highly processive and can be used at elevated temperature (65°C). A Klenow-like fragment without exonuclease activity is available (Phang S.M. et al., Gene. 163(1 ):65-68, "Cloning and complete sequence of the DNA polymerase-encoding gene (Bstpoll) and characterisation of the Klenow-like fragment from Bacillus stearothermophilus." 1995; Aliotta J.M. et al., Genet Anal. 12(5-6): 185-195, "Thermostable Bst DNA polymerase I lacks a 3'->5' proofreading exonuclease activity." 1996) and it has been shown that this polymerase is highly effective for RCA (Zhang D.Y. et al., Gene. 274(1-2):209-216, "Detection of rare DNA targets by isothermal ramification amplification." 2001).

The RCA polymerase for use in the present invention may in particular be the Phi29 DNA polymerase (which has proof-reading activity) and exo(-) BST DNA polymerase. Strand displacement may be facilitated through the use of a strand displacement factor such as a helicase. It is considered that any DNA polymerase that can perform RCA in the presence of a strand displacement factor is suitable for use in the present methods, even if the DNA polymerase does not perform RCA in the absence of such a factor. Strand displacement factors useful in RCA include BMRF1 polymerase accessory subunit (Tsurumi et al., J. Virology 67( 2):7648-7653 (1993)), adenovirus DNA-binding protein (Zijderveld and van der liet, J. Virology 68(2): 1158-1164 (1994)), herpes simplex viral protein ICP8 (Boehmer and Lehman, J. Virology 67(2):711-715 (1993); Skaliter and Lehman, Proc. Natl. Acad. Sci. USA 91(22):10665- 10669 (1994)), single-stranded DNA binding proteins (SSB; Rigler and Romano, J. Biol. Chem. 270:8910-8919 (1995)) and calf thymus helicase (Siegel et al., J. Biol. Chem. 267:13629-13635 (1992)).

The ability of a polymerase to carry out RCA may be determined by using the polymerase in a rolling circle replication assay such as those described in Fire and Xu (Proc. Natl. Acad. Sci. USA 92:4641-4645 (1995)).

In aspects of the invention where 5" exonuclease activity of a DNA polymerase is used to degrade mass tagged PDS probes during PCR amplification of CCPs, a polymerase with the relevant 5' exonuclease activity is used. Taq polymerase is widely used for this purpose (Livak K.J., Genet Anal., 14(5-6): 143-9, "Allelic discrimination using fluorogenic probes and the 5' nuclease assay." (1999)) although a variety of other polymerases has been assessed for this purpose and would be applicable with these embodiments of the invention (Kreuzer K.A. et al., Mol Cell Probes., 14(2): 57-60 (2000)).

RNA polymerases

Where linear RCA is conducted in methods of the invention, an RNA polymerase may be used to effect the replication reaction. An RNA polymerase which can carry out transcription in vitro and for which promoter sequences have been identified may be used. Here, the promoter sequences are used as the PBS and a DNA primer is also used. The primer is extended by a non-displacing polymerase, i.e. with the same characteristics as a gap-filling polymerase to produce a double stranded circular product with a nick. The nick may be ligated if desired. The RNA polymerase is then added to the promoter site and will initiate transcription if ribonucleotide triphosphates are present. Stable RNA polymerases without complex requirements are suitable. Particularly suitable are T7 RNA polymerase (Davanloo et al., Proc. Natl. Acad. Sci. USA 81 :2035-2039 (1984)) and SP6 RNA polymerase (Butler and Chamberlin, J. Biol. Chem. 257:5772-5778 (1982)) which are highly specific for particular promoter sequences (Schenborn and Meirendorf, Nucleic Acids Research 13:6223-6236 (1985)). Other RNA polymerases with this characteristic are also suitable. Because promoter sequences are generally recognized by specific RNA polymerases, the LCP or other probes of the invention may comprise a promoter sequence recognised by a RNA polymerase that is used. Numerous promoter sequences are known and any suitable RNA polymerase having an identified promoter sequence can be used. Promoter sequences for RNA polymerases can be identified using established techniques. Preparation of tissue samples for in situ hybridisation

Prior to in situ hybridisation, tissue samples or tissue sections may be "fixed" to preserve their living structures. A fixative helps to prevent tissue auto-digestion, inhibits bacterial or fungal growth, and renders the tissue resistant to damage during subsequent processing, embedding, and sectioning stages. Fixatives which may be used for in situ hybridisation include formaldehyde, paraformaldehyde, and glutaraldehyde. These compounds fix tissue by introducing cross-links between different tissue components, maintaining their structure. Tissue may be immersed in a solution of the fixative, and allowed to "soak" so the fixative can penetrate the structure. To prepare tissues for mounting on suitable solid surface, such as glass for optical imaging (Tholouli E et al„ Biochem Biophys Res Commun. 348(2):628-36, "Imaging of multiple mRNA targets using quantum dot based in situ hybridisation and spectral deconvolution in clinical biopsies." 2006) or stainless steel for MALDI imaging (Lemaire R et al., J Proteome Res. 6(6):2057-67, "Tag-mass: specific molecular imaging of transcriptome and proteome by mass spectrometry based on photocleavable tag" 2007; Seeley EH & Caprioli RM, Proc Natl Acad Sci U S A. 105(47): 18126-31 , "Molecular imaging of proteins in tissues by mass spectrometry." 2008), the tissues may be embedded in a solid medium so that thin sections may be cut from them. Paraffin may be used as an embedding medium. The thin sections may then be mounted on a mounting surface, resulting in a mounted tissue section. However, for the purposes of imaging mass spectrometry, embedding the tissue in paraffin may be avoided as this may interfere with mass spectrometric detection or the paraffin could be removed prior to analysis. Removal of paraffin is typically required for in situ hybridisation in any case.

The mounted tissue section may then be subjected to other treatments prior to hybridisation with a probe of the invention. Treatment of the mounted tissue section may include one or more of the following steps:

- addition of 1% H2O2 in methanol, or levamisole, which renders any enzyme that is naturally present in the tissue sample inactive so that it cannot react or interfere with labelling reactions or cause autodigestion of the tissue;

- addition of RNase, which degrades endogenous RNA from tissues in DNA probe/DNA target hybridisations. A DNA probe may bind to any RNA that is present in the sample, causing a false positive; RNase digests the RNA so that it cannot hybridise to the DNA probe. RNase should not be added if the targets for in situ hybridisation are RNA. Similarly, DNase can be added if only RNA targets are to be detected;

- addition of hydrochloric acid, which is considered to extract proteins from the sample and partially hydrolyze the target nucleic acid, improving the hybridisation results;

- addition of a detergent such as for example Triton X-100 and/or sodium dodecyl sulphate (SDS). Detergents remove lipid membrane components to increase the target nucleic acid accessibility; and/or

- addition of a protease (such as Proteinase K or pepsin), which degrades protein and unmasks the target nucleic acids. Protease should not be added if simultaneous immunocytochemical detection, another feature of the invention, is intended.

One goal of hybridisation according to the invention is to ensure maximum reaction of the probe with the target nucleic acid, while minimising the degree of potential nonspecific interaction with other nucleic acids and other cellular components. This goal can usually be met by adjusting the components of the hybridisation solution, and by hybridising at an optimal temperature for the appropriate length of time. The probes may be hybridised in a hybridisation buffer (discussed below).

According to the invention, a pre-hybridisation step, in which the tissue section is incubated for a period of time with the hybridisation solution (minus the probe), may be used. This step may help to minimise background staining. If pre-hybridisation is performed, the probes in a hybridisation buffer may be added to the tissue after the pre-hybridisation step.

During the hybridisation step, the tissue section which includes the target nucleic acid, together with the hybridisation solution and labelled probes, are incubated for a period of time to allow hybridisation to take place. Most hybridisation reactions are complete in four hours, but it may be convenient to allow the slides to incubate overnight. During hybridisation, the labelled probes may bind non-specifically to nucleic acids that are only partially homologous to the probe, forming incomplete hybrids. Such incomplete hybrids are less stable than perfectly matched hybrids, and may be dissociated by washing with a washing solution. Stringency of the washing may be adjusted by varying the concentrations of components in the washing solution, and/or the temperature of the washing (see also below).

In the various aspects of this invention, additional steps are used during in situ hybridisation, for example if additional ligation and/or in situ RCA (as described herein) are used. Hybridisation buffers, washing solutions and hybridisation conditions

Probes of the present invention may be dissolved in a hybridisation buffer which may comprise one or more of the following components:

(1 ) Salt. The salt may for example be sodium chloride which provides monovalent cations (Na+) to the solution. Other salts can be used including volatile organic salts such as ammonium acetate and others discussed below. Salt helps to regulate the degree of natural electrostatic repulsion between the probe and the target nucleic acid;

(2) Anionic or neutral macromolecules. Such macromolecules, for example Bovine Serum Albumin (BSA), may be used to reduce non-specific probe binding (i.e. reduce the incidence of probe binding to nucleic acids other than the target nucleic acid);

(3) EDTA (Ethylene Diamine Tetra-acetic Acid). EDTA chelates magnesium ions which are known to activate nuclease activity that could degrade target nucleic acids;

(4) pH Buffering components. These may be used to maintain the pH between 7 and 8. Examples include TRIS HCI, phosphate or citrate. A typically hybridisation (and washing) buffer may comprise Sodium Chloride Sodium Citrate (SSC). 1 x SSC is defined as 150mM NaCI + 15mM sodium citrate; and/or

(5) Detergents/Chaotropes/Denaturants, which include sodium dodecyl sulphate (SDS), Formamide, Guandine HCI, amongst others. Formamide is a helix-destabilizing agent that allows hybridisation to proceed at lower temperatures, enhancing tissue preservation. Methods of the invention encompass template directed chemical ligation of oligonucleotides. As such, the hybridisation of the oligonucleotide probes to their target nucleic acid templates can take place under conditions that would be profoundly inhibitory of enzymatic ligation. Chemical ligation will tolerate the presence of chaotropes and denaturants such as Guanidinium chloride, Formamide and Dimethyl Sulfoxide. The use of such reagents in hybridisation buffers is advantageous in many circumstances as discussed in the literature (Van Ness J & Chen L, Nucleic Acids Res. 19(19):5143-51 , "The use of oligodeoxynucleotide probes in chaotrope- based hybridisation solutions." 1991; Bains W, Genet Anal Tech Appl. 11(3):49-62, "Selection of oligonucleotide probes and experimental conditions for multiplex hybridisation experiments." 1994).

Stringency relates to the specificity of a probe binding to its target nucleic acid. It is a factor that may be optimised for best results prior to chemical ligation. Stringency may be adjusted by varying the concentration of the salt in the hybridisation solution and the temperature of hybridisation. When operating at conditions that provide low stringency (low temperature, high salt concentration), the probe is more likely to bind, but sometimes non-specifically (i.e. the probe may bind to sequences other than the target nucleic acid). As conditions are changed to increase stringency (higher temperature, lower salt concentration), binding becomes more specific, until conditions are too stringent and binding does not occur at all. For each case, it is recommended to begin with low stringency conditions and increase stringency to optimise the results.

In general, three environmental factors may influence stringency. Increasing hybridisation temperature increases stringency while reducing temperature decreases stringency. Conversely, increasing salt concentration decreases stringency while reducing salt concentration increases stringency. Other factors that influence stringency, such as increasing formamide content, may decrease the melting point of DNA, thus lowering the temperature at which a probe/target duplex forms. Adding 1 % formamide may lower the melting temperature by about 0.5 - 0.7°C. In hybridisation analyses, a two stage wash protocol may be used. A low stringency wash is first performed to remove non-specifically bound probe, followed by a high stringency wash to remove undesired hybrids of low homology. All wash solutions may be pre-warmed to the desired temperature prior to being added to the sample, otherwise low homology hybrids may not be disrupted during short washes.

Examples of different stringency wash protocols for use in the invention are set out below. Low Stringency Wash

Probe Target Duration Wash Buffer Temperature

All types All types 5 minutes 2XSSC, Room temp.

0.1% SDS

High Stringency Washes

Probe Target Duration Buffer Temperature

DNA DNA 15 minutes 0.5 x SSC, 65°C

0.1% SDS

RNA DNA 15 minutes 0.1X SSC, 68°C

0.1 % SDS

RNA RNA 15 minutes 0.1X SSC, 68°C

0.1% SDS

DNA DNA 15 minutes 1XSSC, Same Temp as

Oligonucleotide 0.1% SDS Hybridisation

Where the tags of the invention are mass tags, the following factors are relevant. Mass spectrometric detection is susceptible to inhibition by inorganic salts, detergents and other substances, so care should be taken to ensure that salts and detergents are not present in the samples in inhibitory levels when the samples are analysed by mass spectrometry. When samples are to be analysed by mass spectrometry it is preferable to use buffers that are compatible with mass spectrometry. For electrospray and MALDI mass spectrometry the final step of sample preparation may leave the sample in buffer such as ammonium citrate, ammonium acetate, trimethylammonium acetate, tetramethylammonium acetate, trimethylammonium citrate, triethylammonium acetate or tetraethylammonium acetate. Other buffers and conditions suitable for mass spectrometric analysis according to the invention are disclosed in US 6,361 ,940. These components may also be used in the hybridisation buffers to replace both sodium chloride and pH buffering components. Mass spectrometry is not particularly tolerant of chaotropes such as guanidinium chloride. Small amounts of DMSO and formamide are tolerated but in general it is preferable that these are washed away from samples prior to analysis by mass spectrometry. Provided mass spectrometry inhibiting materials are removed prior to analysis, these materials may be used in the method steps prior to mass spectrometry analysis. Where hybridisation takes place on a solid support such as a bead or a fixed tissue section, washing can be applied relatively easily.

Synthesis of oligonucleotides

Oligonucleotide probes of the invention may be produced by chemical synthesis using standard oligonucleotide synthesis methods known in the art. Methods may be purely synthetic, for example, the cyanoethyl phosphoramidite method (Beaucage & Caruthers, 1981 , Tetrahedron Lett. 22: 1859-1862; McBride & Caruthers, Tetrahedron Lett. 24: 245-248). Synthetic methods useful for making oligonucleotides are also described by Ikuta et al. (1984, Ann. Rev. Biochem. 53: 323-356) (phosphotriester and phosphite-triester methods), and Narang et al. (1980, Methods Enzymol. 65: 610-620) (phosphotriester method). PNA molecules may be made using known methods such as those described by Nielsen et al. (1994, Bioconjug. Chem. 5: 3-7).

The H-phosphonate method for oligonucleotide synthesis may be used. This method was first reported by Hale et al. (1957, J. Chem. Soc. 3291) and revisited later by Sekine and Hata (1975, Tetrahedron Lett. 16: 1711), Sekine et at. (1979, Tetrahedron Lett. 20: 1145), Garegg et al. (1985, Chemica Scripta 25: 280), and Garegg et al. (1986, Chemica Scripta 26: 59). The H- phosphonate method involves condensing the 5' hydroxyl group of the nascent oligonucleotide with a nucleoside having a 3' phosphonate moiety. Once the entire chain is constructed, the phosphite diester linkages are oxidized with t-butyl hydroperoxide or iodine to yield the corresponding phosphotriester. See, for example, Froehler (1993, "Oligodeoxynucleotide Synthesis", Methods Mol. Biol. Vol. 20, Protocols for Oligonucleotides and Analogs, p. 63-80, S. Agrawal, Ed., Humana Press), and Uhlmann & Peyman (1990, Chem. Rev. 90: 543). The H- phosphonate method allows the production of modified backbones such as a phosphorothioate backbone (Stawinski & Stramberg, 2005, Methods Mol Biol. 288: 81-100), which is suitable for the present invention.

Since the oligonucleotides of the invention may comprise different portions or regions, smaller subsequences of the oligonucleotides may be synthesised and assembled by ligation (Borodina et al., 2003, Anal Biochem. 318(2): 309-313). Synthesis of oligonucleotides comprising artificial mismatch discontinuities Artificial mismatches in the probes of the invention may comprise natural nucleotides that are known to be non-complementary to the base at the appropriate position in the target sequence. Alternatively, universal nucleoside analogues may used such as 1-(2'-Deoxy-.beta.-D- ribofuranosyl)-3-nitropyrrole, which maximizes stacking interactions while minimizing hydrogen- bonding interactions without sterically disrupting a DNA duplex (Nichols et al., 1994, Nature 369: 492; and Bergstrom et al., 1995, J.A.C.S. 117: 1201 ). Similarly, the analogues 1-(2'-Deoxy-beta- D-ribofuranosyl)-5-nitroindole and 1-(2'-deoxy-beta-D-ribofuranosyl)-4-nitroimidazole may be used. However, nitropyrrole and nitroindole analogues are reported to be most favorable as universal nucleotides as they show the least discrimination for base pairing with natural nucleotides (Bergstrom et al., 1997, Nucleic Acids Res. 25(10): 1935-1942).

Abasic spacers may also be used such as the "D-spacer" in which deoxyribose residues without a base are introduced into the sequence of the probe oligonucleotide (see for example Takeshita et al., 1987, J Biol Chem. 262(21 ): 10171-10179).

The abasic spacer and the universal nucleotides are available as deoxyribose nucleotides but other sugar modifications, in particular 2'-modified sugars, are also envisaged.

Synthesis of oligonucleotides comprising non-nucleic acid discontinuities

A 3-carbon spacer may be introduced into oligonucleotides as a discontinuity by employing a "C3 phosphoramidite spacer". A C3 spacer will separate a pair of phosphodiester groups by approximately the same distance as ribose in terms of the backbone configuration. Longer discontinuities may be introduced, for example as 9-atom, 12-atom and 18-atom spacers which are commercially available as phosphoramidites (for example from Glen Research Corporation, Sterling, Virginia, USA).

Several possibilities exist to introduce discontinuities in PNA oligonucleotides. Most FMOC amino acids may be used. Preferred spacers include glycine, alanine, beta-alanine, serine and lysine. Two alpha amino acids together will have approximately the same internucleotide distance as a single PNA nucleotide. Longer spacings can be introduced using more alpha amino acids or by introducing longer amino acids such as amino hexanoic acid.

Nucleic acid analogues

The oligonucleotide probes of the invention may incorporate "artificial mismatches", so the overall binding affinity of the probes is reduced compared to unmodified probes. It may therefore be useful to use nucleic acid analogues with enhanced binding affinity compared to natural phosphodiester deoxyribosenucleic acids. It is known that RNA analogues with certain modifications at the 2' position of the ribose ring show enhanced binding affinity for RNA targets compared to corresponding DNA RNA hybrids (see Cummins et al., 1995, Nucleic Acids Res. 23(11 ): 2019-24). These RNA analogues also show reduced binding affinity for DNA compared to DNA/DNA hybrids (Tsourkas et al., 2003, Nucleic Acids Res. 31(6): 5168-74). The ability to bind preferentially to RNA over DNA with enhanced melting temperature makes 2'-modified analogues particularly useful for in situ hybridisation applications for detection of alternatively spliced RNA in a background of genomic DNA.

2'-0-methyl analogues in particular are readily available as phosphoramidite monomers for automated synthesis and are suitable for use with this invention. Additionally or alternatively, 2'- fluoro-modified analogues may be used. Other nucleic acid analogues for use in oligonucleotide probes of the invention are "bridged" analogues such as locked nucleic acids ("LNA"; Thomsen et al., 2005, RNA. 11(11):1745-8) and 2'-4'-BNA(NC) (Rahman et al., 2008, J Am Chem Soc. 130(14): 4886-96). Bridged nucleic acid analogues show enhanced binding affinity for RNA compared with their natural nucleic acid counterparts, and are thus suitable for in situ hybridisation applications. Bridged analogues also show enhanced binding affinity for DNA compared with their natural nucleic acid counterparts, and are therefore useful for detection of chromosomal targets such as chromosomal translocations and for the detection of labelled cDNAs.

It is not normally desirable to synthesise oligonucleotide probes that are comprised entirely of bridged analogues. Hence LNA monomers are typically introduced every third base into DNA oligonucleotides (Valoczi et al., 2004, Nucleic Acids Res. 32(22): e175; Obernosterer et al., 2007, Nat Protoc. 2(6): 1508-14). LNA monomers may be introduced into 2'-0-methyl oligonucleotide sequences to enhance binding affinity of the resultant oligonucleotide (Kierzek et al., 2005, Nucleic Acids Res. 33(16): 5082-93). When LNA monomers are introduced into 2'-0- methyl oligonucleotides, the LNA monomers may be positioned every second base but in one embodiment not at the 5' end of an oligonucleotide.

PNA is another analogue for use in oligonucleotide probes of the invention (Nielsen et al., 1994, Bioconjug Chem. 5(1 ): 3-7). PNA has enhanced binding affinity for both DNA and RNA targets compared to DNA oligonucleotides. PNA is less soluble than other DNA analogues and it is currently difficult to produce usable PNA oligonucleotides with a length greater than 20 bases. Hence PNA oligonucleotides may be shorter than oligonucleotides made with sugar/phosphate backbones. The invention encompasses oligonucleotides comprising lengths of PNA and DNA (see Uhlmann, 1998, Biol Chem. 379(8-9): 1045-52). These "mixed" oligonucleotides may be longer than PNA-only probes. Linkages between PNA and DNA can be useful points for introducing discontinuities in such mixed oligonucleotides.

A further useful analogue for purposes of this invention is Triazole-DNA (Hiroyuki Isobe et al., Org. Lett. 10 (17): 3729-3732, "Triazole-Linked Analogue of Deoxyribonucleic Acid (^TLDNA): Design, Synthesis, and Double-Strand Formation with Natural DNA", 2008; Nuzzi et al., QSAR Comb. Sci. 26 (11-12): 1191-1199, "Model Studies Toward the Synthesis of Thymidine Oligonucleotides with Triazole Internucleosidic Linkages via Iterative Cu(l)-Promoted Azide- Alkyne Ligation Chemistry." 2007) Oligonucleotide characterisation

For several applications of the oligonucleotide probes, it may be useful to determine the T_m. Higher T_m values correspond to more stable duplexes. The stability of DNA duplexes can be calculated using known methods for prediction of melting temperatures (Breslauer et al., 1986, PNASUSA 83(11): 3746-3750; Lesnick & Freier, 1995, Biochemistry 34: 10807-10815, 1995; McGraw et al., 1990, Biotechniques 8: 674-678, 1990; and Rychlik et al., 1990, Nucleic Acids Res. 18: 6409-6412).

Oligonucleotide labelling

For several applications of the oligonucleotide probes of this invention, directly labelled probes are provided. Labels may include fluorophores, mass tags, enzymatic labels, affinity ligands such as avidin. As discussed in Example 10 below, azide- or alkyne-labelled peptide mass tags can be readily linked to alkyne- or azide-labelled oligonucleotides respectively. Similarly, a variety of azide modified fluorescent dyes are commercially available such as the Click-iT™ Alexa Fluor dyes from Invitrogen Corporation and can be coupled to alkyne modified oligonucleotide probes of the invention, as discussed below. Maleimide activated enzymes such as horseradish peroxidase are also commercially available (Pierce) and can be readily coupled to thiol-derivatised oligonucleotide probes of the invention using published methods (Niemeyer CM et al., Nucleic Acids Res. 22(25):5530-9, "Oligonucleotide-directed self-assembly of proteins: semisynthetic DNA-streptavidin hybrid molecules as connectors for the generation of macroscopic arrays and the construction of supramolecular bioconjugates." 1994). Alkyne and thiol groups are essentially unreactive to each other and can be used in the same oligonucleotide probe, so an alkyne modified probe for ligation may be modified with a thiol to enable labelling with maleimide-activated labels. Detection of mass tags

As described herein, probes or probe amplicons of the invention may be labelled with mass tags.

For detection of mass tagged probes on microarrays, after hybridisation of directly labeled probes or probe amplicons or after RCA and hybridisation of PDS sequences to the microarray, the microamay may be washed to remove salts that may interfere with mass spectrometry. The array may then be coated with a MALDI matrix material such as 3-hydroxypicolinic acid or alpha-cyano-cinnamic acid. Having prepared the microarray in this way, it may then be loaded into a MALDI based mass spectrometer and the mass tags may be cleaved and desorbed from discrete locations on the array by application of laser light to the desired location on the array.

Similarly, if the probes of this invention are used for in situ hybridisation with mass tags, the tissue section may be washed to remove salts that may interfere with mass spectrometry. The tissue section may then be coated with a MALDI matrix material such as 3-hydroxypicolinic acid or alpha-cyano-cinnamic acid.

Where tags are desorbed from a surface by laser according to the invention, appropriate methods for cleaving the tags from their associated probes on the array should be used. In one approach, the tags may be linked to their associated probes or probe amplicons or probe detection sequences through a photocleavable linker. This means that cleavage of the tags can take place at discrete locations on the array by exposure to light of the appropriate frequency. This light may be applied to the whole array prior to analysis by exposing the array to an intense light source. Alternatively, in a MALDI mass spectrometer, the laser used for desorption may be used to cleave the tags.

Alternatively, an acid cleavable linker may be used. Since most MALDI matrix materials are acidic, addition of the matrix will effect cleavage of the mass tags.

In another embodiment, the entire probe label complex can be desorbed, and cleavage of the tags can take place by collision using Post Source Decay in a Time-Of-Flight mass spectrometer or in the mass analyzer of an ion trap instrument or in a collision cell in alternative geometries that are used with MALDI, such as the Q-TOF geometry.

Oligonucleotide Synthesis Examples

In the section below, standard DNA phosphoramidites, solid supports and additional reagents including the C7-aminoalkyl cpg were purchased from Glen Research, Inc, Link Technologies or Applied Biosystems Ltd. oligonucleotides were synthesized on an Applied Biosystems 394 automated DNA/RNA synthesiser using a standard 0.2 or 1 .0 micromole phosphoramidite cycle of acid- catalysed detritylation, coupling, capping and iodine oxidation. Stepwise coupling efficiencies and overall yields can be determined by the automated trityl cation conductivity monitoring facility. All [betaj-cyanoethyl phosphoramidite monomers were dissolved in anhydrous acetonitrile to a concentration of 0.1 M immediately prior to use. The coupling time for normal (A,G,C,T) monomers is typically 25 seconds. Cleavage of the oligonucleotides from the solid support and deprotection was achieved by exposure to concentrated aqueous ammonia solution for 60 min at room temperature followed by heating in a sealed tube for 5h at 55 degrees Celsius.

Copper Catalyzed Azide Alkyne Cycloaddition Reactions

The reaction between an azide and an alkyne to form a disubstituted, 1 ,2,3-triazole is an example of a 1 ,3-dipolar cycloaddition. This reaction has been known for a very long time but the rate of the reaction without catalysis is low and the regioselectivity of the reaction (1 ,4- versus 1 ,5-cycloaddition) was not controllable until recently.

Copper catalysis of the azide alkyne reaction was first introduced by Meldal and colleagues (Tornoe, C. W., Christensen, C. & Meldal, M. (2002). Peptidotriazoles on solid phase: [1 ,2,3]- triazoles by regiospecific copper(l)-catalyzed 1 ,3-dipolar cycloadditions of terminal alkynes to azides. Journal of Organic Chemistry 67, 3057-3064.). This reaction regiospecifically unites azides and terminal alkynes to afford only 1 ,4-disubstituted 1 ,2,3-triazoles at a much higher rate than the uncatalyzed reaction. This reaction functions with organic solvents and appears to require that the alkyne be immobilised on a solid support to allow the reaction to proceed as desired.

Similarly, the so-called Sharpless protocol developed by K. B. Sharpless and others (see V. V. Rostovtsev et al., Angewandte Chemie-lnternational Edition, 41.(14), 2596 (2002) and WO 03/101972) also employs copper (I) catalysis. However, the Sharpless protocol provides a more general reaction procedure that operates in aqueous solution.

The cycloaddition reaction itself can work with a variety of catalysts as reported in WO 03/101972. In particular, the use of ruthenium (II) complexes has been reported to catalyse the cycloaddition of both terminal and internal alkynes, resulting in the formation of 1 ,5- disubstituted 1 ,2,3-triazoles (as opposed to the 1 ,4 regioisomers obtained through Cu (I) catalysis) and 1 ,4,5-trisubstituted-1 ,2,3-triazoles (L. Zhang et al., J. Am. Chem. Soc, 127, 15998 (2005)); also reported in the recent review by Wu & Fokin of the reactivity and applications of catalytic azide-alkyne cycloadditions (Aldrichimica Acta, 40 (1 ), 7 (2007)).

In the context of the present invention, the precise control over regioselectivity of the cycloaddition reaction is not typically particularly critical, i.e. to generate highly labelled conjugated the structures, rather than to provide specific 1 ,2,3- triazole compounds. Rather, the reaction can proceed with both internal and external alkynes, without catalyst and with catalysis other than by Cu(l). The copper-catalysed reaction has received arguably the greatest attention in the chemical literature where it is referred to as the copper-catalysed azide-alkyne cycloaddition (CuAAC) reaction.

Whilst the CuAAC reaction has been used to immobilise DNA on electrode surfaces and chips (Collman et al. (supra); Seo et al. (supra)), to cyclise peptides on resins (Punna et al., Angewandte Chemie-lnternational Edition in English, 44, 2215 (2005)) and to link the termini of oligonucleotide strands as described by Kanan et al. (supra; see also WO2004/016767 and WO2007/011722 ). The cycloaddition reaction can be performed in aqueous solution and a variety of organic solvents. For the purpose of the present invention, it is convenient and advantageous to conduct the cycloaddition reaction in aqueous solution.

All CuAAC ligation reactions reported in the examples hereinafter were carried out at 0.2M aqueous NaCI. However, it will be understood that other solutions may be used. The reaction is catalyzed by a number of possible agents. Whilst the reaction can be effected with catalytic quantities of catalysts, in other words of the order less than 1 molar equivalent versus the substrate(s), for example 0.001 - 10 mol%, and in one embodiment between about 0.1 to 5%, for example 1 to 3 mol% catalyst, greater quantities of catalyst may be used, if convenient, depending on the scale of any given reaction. For example, in some of the examples, which follow, quantities of copper of the order of 100 equivalent vis-a-vis the reactants are used. Nevertheless, dependent upon scale, this can still involve the use of miniscule quantities of copper catalyst. Whilst copper (I) is the catalyst most described in the scientific (and patent) literature, the reaction can also be catalysed by the presence of metallic copper or by a metal ion selected from Au, Ag, Hg, Cd, Zr, Ru, Fe, Co, Pt, Pd, Ni, Rh and W. Further details of appropriate catalysts will be known to the skilled person and are provided in, for example, WO 03/101972. Where the catalyst is copper (I) a variety of sources of copper (I) can be used. Copper (I) is thermodynamically unstable. However; addition of copper (II) salts can be used in combination with a reducing agent, for example ascorbate, so as to generate conveniently the desired copper (I) species in situ. Alternatively, desired quantities of catalytic amounts of copper (I) can be introduced through comproportionation of copper(ll) and copper (0). Further details are provided in WO03/101972 and Wu & Fokin (supra).

Of particular utility in the cycloaddition of azides with internal alkynes, ruthenium (II) complexes may be used to catalyse the cycloaddition reaction, as reported in Zhang et al. (supra).

The present inventors have found that the most convenient catalyst to use for the cycloaddition reaction was copper (I) prepared in situ from copper (II) sulfate and sodium ascorbate. Such reaction, particularly when carried out in 0.2 M aqueous NaCI, was found to be most conveniently conducted in the presence of a water-soluble tris-triazolylamine copper (l)-binding ligand, as has been reported previously (T. R. Chan et al., Org. Lett. (2004) 6 (17), 2853). In particular, it was found that the use of HPTA (tris-(hydroxypropyltriazoylylmethyl)amine) is advantageous insofar as it assists in preventing degradation of oligonucleotides by the copper catalyst. When more than a 5- fold excess of triazolylamine ligand (in particular HPTA) relative to the copper (I) catalyst was used, for example ten or twenty equivalents or even more, virtually no decomposition of the reactant oligonucleotide substrates for the cycloaddition reaction was observed.

Example 1: Synthesis of an azide modified peptide

Synthesis of an internally azide-modified peptide is achieved by incorporation of FMOC- azidohomoalanine or Fmoc-Lysine(epsilon-N3)-OH (both available from AnaSpec). The incorporation of the FMOC-azidohomoalanine or the corresponding lysine amino acid can be performed using standard FMOC amino acid coupling conditions on certain acid-labile resins, such as Rink Amide or Sieber Amide resins. In preferred embodiments a Sieber Amide resin is used, which releases the C-terminal amino acid as an amide in the final deprotection and cleavage step. The azide group is sensitive to extended exposure to TFA, which should be avoided. Standard FMOC-deprotection conditions and coupling conditions do not affect the azide residue. The peptide is then purified by HPLC. See Fig. 2(a) for an example of azide- modified tag peptide according to this invention. This is similar to tags previously described for development of sets of isobaric mass tags (Thompson A et al., Nucleic Acids Res. 35(4):e28, "Electrospray ionisation-cleavable tandem nucleic acid mass tag-peptide nucleic acid conjugates: synthesis and applications to quantitative genomic analysis using electrospray ionisation-MS/MS." 2007).

Synthesis of an N-terminal azide-modified peptide is achieved by reaction of the alpha-amino group of the final amino acid in the peptide with 5-Azido-pentanoic acid (available from Bachem). The incorporation of the 5-Azido-pentanoic acid entity can be achieved using standard coupling conditions on most resins.

Synthesis of a C-terminal azide-modified peptide is achieved by synthesis of the peptide on an FMOC-hydrazinobenzoyl resin (such as 4-Fmoc-hydrazinobenzoyl AM NovaGel available from Merck Biosciences GmbH). The C-terminal amino acid is introduced by removing the FMOC on the FMOC-hydrazinobenzoyl linker using standard deprotection conditions, followed by coupling using standard coupling conditions. The desired peptide is then synthesised as usual. The hydrazinobenzoyl linker is stable to both acid and base. This means that the peptide can be cleaved as a protected fragment or after deprotection of the side chain amino acids. Cleavage is then effected by oxidation of the linker with N-bromosuccinimide (NBS) which converts the hydrazide to the corresponding diazene. Typically, the resin is swollen and washed with dichloromethane (DCM) and then 2-3 equiv. of NBS in DCM is added to the resin in the presence of 10 - 12 equivalents of anhydrous pyridine. The reaction is usually complete in about 10 min. The resin is washed again with DCM. Finally, the peptide is cleaved from the resin by aminolysis with a suitable azido-amine compound.

Example 2: Synthesis of an alkyne modified peptide

Synthesis of an internally alkyne-modified peptide is achieved by incorporation of FMOC- propargyl glycine (available from SigmaAldrich). The incorporation of the propargyl glycine can be performed using standard FMOC amino acid chemistry on most resins. In preferred embodiments a PAL resin is used, which releases the C-terminal amino acid as an amide in the final deprotection and cleavage step. The alkyne group is stable to most conditions used in a peptide synthesizer with the exception of reductive conditions, which should be avoided. Standard FMOC deprotection conditions and TFA, typically used for the final deprotection and cleavage steps, do not affect the propargyl glycine residue. The synthesis was performed on a custom-made synthesizer constructed from a Gilson 215 liquid handling robot configured for peptide synthesis. The peptide can then be purified by HPLC. See Fig. 2(b) for an example of alkyne-modified tag peptide according to the invention.

Synthesis of an N-terminal alkyne-modified peptide is achieved by reaction of the alpha-amino group of the final amino acid in the peptide with 5-hexynoic acid (available from SigmaAldrich). The incorporation of the 5-hexynoic acid can be performed using standard coupling conditions on most resins.

Synthesis of a C-terminal alkyne-modified peptide is achieved by synthesis of the peptide on a number of different resins. In one approach, the peptide can be synthesized on an FMOC- hydrazinobenzoyl resin (such as 4-Fmoc-hydrazinobenzoyl AM NovaGel available from Merck Biosciences GmbH). The C-terminal amino acid is introduced by removing the FMOC on the FMOC-hydrazinobenzoyl linker using standard deprotection conditions, followed by coupling using standard coupling conditions. The desired peptide is then synthesised as usual. The hydrazinobenzoyl linker is stable to both acid and base. This means that the peptide can be cleaved as a protected fragment or after deprotection of the side chain amino acids. Cleavage is then effected by oxidation of the linker with N-bromosuccinimide (NBS) which converts the hydrazide to the corresponding diazene. Typically, the resin is swollen and washed with dichloromethane (DCM) and then 2-3 equiv. of NBS in DCM is added to the resin in the presence of 10 - 12 equivalents of anhydrous pyridine. The reaction is usually complete in about 10 min. The resin is washed again with DCM. Finally, the peptide is cleaved from the resin by aminolysis with propargylamine in a suitable solvent such as DMF. Alternatively, the resin can be oxidized using copper acetate in Dimethylfonnamide in the presence of propargylamine (C. R. Millington, et al. Tetrahedron Lett., 39: 7201 , 1998; Peters C. & Waldmann H. J. Org. Chem., 68: 6053, 2003). Alternatively, the peptide can be synthesized on a 4-Hydroxymethylbenzoic acid (HMBA) resin or a 4-(4-Formyl-3-methoxyphenoxy)butyryl (FMPB) resin (Alsina J, Albericio F, Peptide Science 71(4):454 - 477, "Solid-phase synthesis of C-terminal modified peptides" 2003) with aminolysis in the final cleavage step using propargylamine. Example 3: Preparation of an oligonucleotide 5'-labelled with an alkyne function Oligonucleotides synthesized using standard phosphoramidite chemistry are modified at the 5' terminus by coupling of a 5' hexynyl phosphoramidite synthon (6-Hexyn-1-yl-(2-cyanoethyl)- (N,N-diisopropyl)-phosphoramidite, commercially available from Glen Research, Inc) to the 5' hydroxyl group of the oligonucleotide.

Example 4: Preparation of an oligonucleotide 3'-labelled with an alkyne function

Oligonucleotides synthesized using reverse phosphoramidite chemistry (Claeboe et al., Nucleic Acids Research, 31 , 5685-5691 , "3'-Modified oligonucleotides by reverse DNA synthesis" 2003). The necessary 5'->3' phosphoramidate monomers are commercially available (Glen Research, Inc). These oligonucleotides are modified at the 3' terminus by coupling of a 5' hexynyl phosphoramidite synthon (commercially available from Glen Research, Inc) to the 3' hydroxyl group of the oligonucleotide while on the resin. Example 5: Preparation of an oligonucleotide 3'-labelled with an amine function

A standard oligonucleotide is synthesised with a 3' amino group using 3'-Amino-Modifier C7 CPG 500 (2-Dimethoxytrityloxymethyl-6-fluorenylmethoxycarbonylamino-hexane-1-succinoyl- long chain alkylamino-CPG; Glen Research, Inc) following the manufacturer's instructions. The F OC protection group on the amino function is removed during standard ammonia deprotection and cleavage of the oligonucleotide from the solid support.

Example 6: Preparation of Hexynoic acid N-hydroxysuccinimide ester

Dicyclohexylcarbodiimide is added to a suspension of 4-Hexynoic acid and N- hydroxysuccinimide in Dichloromethane (DCM) at room temperature with the reaction left to stir. After about 4 hours, aqueous KCI is added and the organic layer separated, washed with water, dried over sodium sulfate, filtered and the solvent removed in vacuo. The active ester can be purified by column chromatography (99:1 , DCM: methanol).

Example 7: Alternative preparation of an oligonucleotide 3'-labelled with an alkyne function The 3'-amino modified oligonucleotide from example 5 is coupled to the hexynoic acid N- hydroxysuccinimide ester (example 6) to give the 3'-alkyne modified structure, x equivalents of oligonucleotides is incubated with y equivalents of active ester for 4 hours at room temperature in 0.5 M Na2C03/NaHC03 buffer (pH 8.75). The crude oligonucleotide is then purified by Reverse Phase High Performance Liquid Chromatography (RP-HPLC) with final desalting by NAP-10 gel filtration according to the manufacturers instructions (GE Healthcare). Example 8: Preparation of an alkynyl uridine phosphoramidite

One equivalent of 5'-0-(Dimethoxytrityl)-5-(propargyloxy)-2'-deoxyuridine (Berry & Associates, Inc (Dexter, Ml, USA) is dissolved in DCM under argon. Two equivalents of DIPEA is added followed by 1 equivalent of 2-Cyanoethyl Ν,Ν-diisopropyl-chlorophosphoramidite dropwise. The reaction is stirred for one hour under argon and then transferred under argon into a separating funnel containing DCM. The mixture is washed with saturated aqueous KCI. The organic layer is separated, dried over sodium sulphate and the solvent removed in vacuo. The product is then purified by column chromatography.

Example 9: Preparation of an oligonucleotide internally labelled with multiple alkyne functions The alkynyl uridine phosphoramidite described in Example 8 above may be incorporated into an oligonucleotide sequence at any selected point in exactly the same manner as the standard 4 phosphoramidite deoxynucleotides. Multiple incorporations are also possible. The optimum spacing of the alkynyl nucleotides from each other within the sequence will depend on the size of the corresponding azide that will be coupled to the alkyne. Typically 4 to 5 bases is sufficient for a small azide-derivatized peptide (5 to 10 amino acids). This can be optimised empirically by trying different spacings and coupling the desired azide compound as discussed below. The yield of the desired multiply labelled conjugate will indicate whether greater spacing is required. For an example of an internally alkyne-modifed oligonucleotide see Fig. 2.

A variety of other alkyne containing nucleotides have been described for this purpose (WO2008/120016) and are also applicable in the context of this invention. Internally alkyne labelled oligonucleotides may be modified at the 3' and/or the 5' terminus. Standard modifications include introduction of 3' and or 5' amino groups. A 3' amino group can be introduced using 3'-Amino-Modifier C7 CPG 500 (Glen Research, Inc) as discussed in Example 5. Alternatively, the amino group can be introduced at the 5' terminus as discussed in Example 12.

Example 10: Preparation of an oligonucleotide labelled internally with multiple peptide tags

An internally alkyne labelled oligonucleotide of Example 9 is conjugated to an azide modified peptide of Example 1 by the CuAAC reaction (see Fig. 3). Typically equimolar quantities of azide and alkyne are required for the Sharpless protocol. Thus is if there are 10 alkyne sites in the internally labelled oligonucleotide then at least 10 equivalents of peptide tag will be required.

Thus 1 equivalent of oligonucleotide will require at least 10 equivalents of peptide tag in aqueous solution. Copper (I) catalyst is prepared in situ from copper (II) sulfate and sodium ascorbate. This reaction is conducted in the presence of a water-soluble tris-triazolylamine copper (l)-binding ligand, preferably HPTA (tris-(hydroxypropyltriazoylylmethyl)amine). X equivalents of copper catalyst that is typically present in a 10-fold excess relative to the copper catalyst.

The reaction is stopped by desalting with NAP-G25 (X2) gel filtration (GE HealthCare, Ltd) and the conjugate is then purified by reverse phase HPLC.

If an amino group has been introduced into the oligonucleotide this should be unaffected by the CuAAC reaction but if necessary, the MMT group could be left in place until after the CuAAC reaction is complete (see Example 12 below). This can then deprotected and converted to an azide after purification by reaction with lmidazole-1-sulfonyl azide (E. D. Goddard-Borger and R.

V. Stick. "An Efficient, Inexpensive, and Shelf-Stable Diazotransfer Reagent: lmidazole-1- sulfonyl Azide Hydrochloride". Organic Letters 9 (19): 3797-3800, 2007). Alternatively, the free amino group can be reacted with 4 azido-butyrate N-hydroxysuccinimide ester as published (El-

Sagheer et al., Chembiochem. 9(1):50-2, "A very stable cyclic DNA miniduplex with just two base pairs." 2008; WO2008/120016).

Example 11: Preparation of an oligonucleotide conjugated to multiple oligonucleotide peptide tag conjugates

The peptide conjugated oligonucleotide of Example 10 with an azide at either the 5' or 3' terminus is conjugated to an alkyne derivatized oligonucleotide of Example 9 using the CuAAC reaction as discussed above in Example 10. Fig. 4 illustrates this conjugation reaction. Again the alkyne derivatized oligonucleotide may include an amino group if desired. The resulting branched conjugated would probably not be readily purified by HPLC but could be purified by gel filtration.

It should be clear to one of skill in the art that if the alkyne derivatized oligonucleotide (to which the peptide-tagged oligonucleotide is conjugated) also has an amino group, this can be converted to an azide after the coupling so that this process of conjugation could be iterated to generate dendrimers of varying degrees of branching as determined by the number of alkynes introduced into each oligonucleotide, their spacing and the size of the oligonucleotides used.

Example 12: Preparation of an oligonucleotide 5'-labelled with an amine function

An amino group is introduced at the 5' terminus of an oligonucleotide during standard phosphoramidite synthesis using either the 5'-Amino-Modifier C6 (6-(4- Monomethoxytritylamino)hexyl-(2-cyanoethyl)-(N,N-diisopropyl)-phosphoramidite product from Glen Research, Inc or the 5'-Amino-Modifier C6-TFA (6-(Trifluoroacetylamino)-hexyl-(2- cyanoethyl)-(N.N-diisopropyl)-phosphoramidite) product from Glen Research following the manufacturer's instructions. The former reagent is cleaved from the column with MMT group in place supporting Reverse Phase HPLC purification and the MMT is removed post purification acid. The latter reagent is deprotected in the final ammonia cleavage step. If desirable, reagents with longer linkers are also commercially available. Example 13: Preparation of an oligonucleotide either 3- or 5'-labelled with an azide function As discussed in Examples 5 and 13, respectively, amino groups are introduced at the 3'- and 5' terminus of an oligonucleotide. These are converted to an azide after purification by reaction with lmidazole-1-sulfonyl azide (E. D. Goddard-Borger and R. V. Stick. "An Efficient, Inexpensive, and Shelf-Stable Diazotransfer Reagent: lmidazole-1-sulfonyl Azide Hydrochloride". Organic Letters 9 (19): 3797-3800, 2007). Alternatively, the free amino group may be reacted with 4 azido-butyrate N-hydroxysuccinimide ester as published (El-Sagheer et al., 2008, supra; WO2008/120016).

Although the present invention has been described with reference to preferred or exemplary embodiments, those skilled in the art will recognize that various modifications and variations to the same can be accomplished without departing from the spirit and scope of the present invention and that such modifications are clearly contemplated herein. No limitation with respect to the specific embodiments disclosed herein and set forth in the appended claims is intended nor should any be inferred.

All documents cited herein are incorporated by reference in their entirety.

Claims

1. A method for detecting the presence or absence of a target nucleic acid, comprising the steps of:

(1) providing a first oligonucleotide probe with a 5' terminus and a 3' terminus, wherein the

5' terminus of the first oligonucleotide probe comprises an azide group or an alkyne group, and wherein the first oligonucleotide probe comprises at its 5' terminus a first probe target recognition sequence that is complementary to and capable of hybridising to a first region of the target nucleic acid;

(5) ligating the terminal alkyne group and the terminal azide group of the oligonucleotide probes by cycloaddition to form a ligated oligonucleotide probe; and

(6) detecting the presence or absence of the ligated oligonucleotide probe, thereby detecting the presence or absence of the target nucleic acid.

2. The method according to claim 1 , in which the first oligonucleotide probe is hybridised to the target nucleic acid prior to, sequentially, or after hybridisation of the second oligonucleotide probe to the target nucleic acid in step (3).

3. A method for detecting the presence or absence of a target nucleic acid, comprising the steps of:

(1 ) contacting the target nucleic acid under hybridising conditions with a first oligonucleotide probe having a 5' terminus and a 3' terminus, wherein the 5' terminus of the first oligonucleotide probe comprises an azide group or an alkyne group, the 3' terminus of the first oligonucleotide probe comprises an azide group, an alkyne group or a free 3' hydroxyl group, and the wherein the first oligonucleotide probe comprises a first probe sequence complementary and capable of hybridising to a first target sequence in the target nucleic acid;

(5) detecting the presence or absence of the circularised probe, thereby detecting the presence or absence of the target nucleic acid.

4. The method according to claim 3, in which the first oligonucleotide probe is hybridised to the target nucleic acid prior to, sequentially, or after hybridisation of the second oligonucleotide probe to the target nucleic acid.

5. The method according to any preceding claim, in which the first and/or second oligonucleotide probe comprises an immobilisation function for immobilising the oligonucleotide probe to the target nucleic acid.

6. The method according to any preceding claim, in which the first and/or second oligonucleotide probe comprises one or more tag molecules.

7. The method according to claim 6 when dependent on claim 5, in which the immobilisation function and the one or more tag molecules are located on different oligonucleotide probes.

8. The method according to claim 7, further comprising the step of separating the ligated oligonucleotide probe hybridised to the target nucleic acid from unligated first and second oligonucleotide probes using the immobilisation function on one of the oligonucleotide probes.

9. The method according to claim 8, comprising stringent washing.

10. The method according to any of claims 5 to 9, in which the immobilisation function is a cross-linking agent, for example a photo cross-linking agent such as psoralen.

11. The method according to claim 10, in .which the oligonucleotide probe comprising the cross-linking agent is cross-linked to the target nucleic acid prior to hybridisation of the other oligonucleotide probe.

12. The method according to either of claim 10 or claim 11 , comprising a washing step prior to hybridisation of the other oligonucleotide probe.

13. The method according to any of claims 5 to 12, in which the immobilisation function is a biotin function.

14. The method according to any preceding claim, in which there is a gap of one or more nucleotides in length between the adjacent termini of the first and second oligonucleotide probes when hybridised to the target nucleic acid, which gap is filled before coupling the terminal alkyne group and the terminal azide group of the oligonucleotide probes.

15. A method for detecting the presence or absence of a target nucleic acid, comprising the steps of:

(1 ) contacting the target nucleic acid under hybridising conditions with an oligonucleotide probe having a 5' terminus and a 3' terminus and further comprising two terminal nucleic acid target recognition sequences that are complementary to and capable of hybridising to two neighbouring regions of the target nucleic acid, in which the 5' terminus of the oligonucleotide probe comprises an alkyne group or an azide group and the 3' terminus of the oligonucleotide probe comprises a free 3' hydroxyl group, and in which there is a gap of one or more nucleotides in length between the hybridisation site of the 5' terminus of the probe and the 3'- terminus of the oligonucleotide probe;

16. The method according to claim 15, in which the probe comprises an immobilisation function.

17. The method according to either of claim 15 or claim 16, in which the probe comprises a cleavable group which allows the circularised probe to be re-linearised.

18. The method according to any of claims 3 to 17, in which the step of separating the circularised probe from uncircularised probe includes an exonuclease treatment to degrade uncircularised probe.

19. The method according to any preceding claim, in which the target nucleic acid is fixed in a tissue section.

20. The method according to any preceding claim, in which the tag is a mass tag detectable by mass spectrometry, for example MALDI mass spectrometry or electrospray mass spectrometry.

21. The method according to any preceding claim, in which the or each oligonucleotide probe comprises a nucleic acid analogue.

22. The method according to any preceding claim, in which the or each oligonucleotide probe comprises an artificial mismatch.

23. A pair of oligonucleotide probes for detecting the presence or absence of a target nucleic acid, in which the pair consists of a first oligonucleotide probe and a second oligonucleotide and in which:

the first oligonucleotide probe has a 5' terminus and a 3' terminus, wherein the 5' terminus of the first oligonucleotide probe comprises an azide group or an alkyne group, and wherein the first oligonucleotide probe comprises at its 5' terminus a first probe target recognition sequence that is complementary to and capable of hybridising to a first region of the target nucleic acid; the second oligonucleotide probe has a 5' terminus and a 3' terminus, wherein the 3' terminus of the second oligonucleotide probe comprises an azide group, an alkyne group or a free 3' hydroxyl group, and wherein the second oligonucleotide probe comprises at its 3' terminus a second probe target recognition sequence that is complementary to and capable of hybridising to a second region of the target nucleic acid which is adjacent the first region of the target nucleic acid such that the 5' terminus of the first oligonucleotide when hybridised to the first region is adjacent the 3' terminus of the second oligonucleotide probe when hybridised to the second region, provided that where the free 3' hydroxyl group is present on the second oligonucleotide probe then the 3' terminus of the first oligonucleotide probe is blocked to prevent extension with a polymerase and there is a gap of one or more nucleotides in length between the 3'-terminus of the second oligonucleotide probe and the 5'-terminus of the first oligonucleotide probe when the first and second oligonucleotide probes are hybridised to the target nucleic acid;

24. The pair of oligonucleotide probes according to claim 23, in which the free 3' hydroxyl group when present on the 3' terminus of the second oligonucleotide probe is extensible when the second oligonucleotide probe is hybridised to the target sequence with a polymerase that in the presence of at least one nucleotide triphosphate modified with an azide group or an alkyne group introduces the azide group or the alkyne group into the 3' terminus of the second oligonucleotide probe, such that within the pair of oligonucleotide probes, one of the oligonucleotide probes comprises a terminal azide group and the other oligonucleotide probe comprises a terminal alkyne group which are adjacent to each other when the first and second oligonucleotide probes are hybridised to the target nucleic acid.

25. The pair of oligonucleotide probes according to either of claim 23 or claim 24, in which the or each target recognition sequence comprise a discontinuity.

26. The pair of oligonucleotide probes according to any of claims 23 to 25, in which the first and/or second oligonucleotide probe comprises a tag.

27. The pair of oligonucleotide probes according to any of claims 23 to 26, in which the first and/or second oligonucleotide probe comprises an immobilisation function.

28. The pair of oligonucleotide probes according to claim 27, in which the immobilisation function is a cross-linking agent or biotin.

29. An oligonucleotide probe for detecting the presence or absence of a target nucleic acid, in which:

the T_m of the target recognition sequences of the oligonucleotide probe are substantially the same as each other.

30. The oligonucleotide probe according to claim 29, in which the free 3' hydroxyl group when present on the 3' terminus is extensible when the oligonucleotide probe is hybridised to the target sequence with a polymerase that in the presence of at least one nucleotide triphosphate modified with an azide group or an alkyne group introduces the azide group or the alkyne group into the 3' terminus of the second oligonucleotide probe, such that the oligonucleotide probe comprises a terminal azide group and a terminal alkyne group which are adjacent to each other when the oligonucleotide probe is hybridised to the target nucleic acid.

31. The oligonucleotide probe according to either of claim 298 or claim 30, in which the terminal alkyne group and azide group of the oligonucleotide probe when hybridised to the target nucleic acid is ligatable by cycloaddition to form a circularised probe which interlocks with the target nucleic acid through catenation.

32. The oligonucleotide probe according to any of claims 29 to 31 , comprising a cleavable group which allows re-linearisation of the circularised probe.

33. The oligonucleotide probe according to any of claims 29 to 32, in which the or each target recognition sequence comprises a discontinuity.

34. The oligonucleotide probe according to any of claims 29 to 33, comprising a tag.

35. The oligonucleotide probe according to any of claims 29 to 34, comprising an immobilisation function.

36. The oligonucleotide probe according to any of claims 29 to 35, comprising a cross-linking agent.

37. The pair of oligonucleotide probes according to any of claims 23 to 28, or the oligonucleotide probe according to any of claims 29 to 36, in which the or each oligonucleotide probe comprises one or more primer binding sequences.

38. The pair of oligonucleotide probes according to any of claims 23 to 28 or 37, or the oligonucleotide probe according to any of claims 29 to 37, in which the or each probe comprises one or more probe identification sequences.

39. The pair of oligonucleotide probes according to any of claims 23 to 28, 37 or 38, or the oligonucleotide probe according to any of claims 28 to 38, in which the or each probe comprises a microarray address sequence.

40. A kit comprising a pair of oligonucleotide probes according to any of claims 23 to 28, or 37 to 39 or the oligonucleotide probe according to any of claims 29 to 39.

41. A microarray comprising discrete locations each having a microarray address sequence complement which is capable of binding to a microarray address sequence of an oligonucleotide probe as defined in claim 39.

42. A kit comprising a microarray as defined in claim 41 and a pair of oligonucleotide probes according to claim 39 or an oligonucleotide probe according to claim 39.

43. A method for detecting the presence or absence of a target nucleic acid, a pair of oligonucleotide probes, an oligonucleotide, a kit, or a microarray substantially as described herein with reference to the accompanying drawings.