WO2010028443A1

WO2010028443A1 - Microrna target identification assay

Info

Publication number: WO2010028443A1
Application number: PCT/AU2009/001196
Authority: WO
Inventors: Thomas Preiss; Traude Helene Beilharz
Original assignee: Victor Chang Cardiac Research Institute Ltd
Current assignee: Victor Chang Cardiac Research Institute Ltd
Priority date: 2008-09-11
Filing date: 2009-09-10
Publication date: 2010-03-18
Anticipated expiration: 2011-03-11

Abstract

The invention relates to assays and kits for detecting microRNA (miRNA) targets through determining changes in length of poly(A) tails of messenger (mRNA). In particular, microRNA (miRNA) targets can be detected by treating a cell with a miRNA, or treating a cell to inactivate a miRNA produced by the cell, or treating a cell to prevent a miRNA from being produced by the cell; detecting poly(A) tail lengths of messenger (mRNA) molecules of the treated cell; detecting poly(A) tail lengths of mRNA molecules of a corresponding untreated cell; comparing poly(A) tail lengths on mRNA molecules between the treated cell and untreated cell; and detecting a target of the miRNA from changes in poly(A) tail lengths of mRNA molecules in a treated cell compared with an untreated cell.

Description

MICRORNA TARGET IDENTIFICATION ASSAY

Technical Field

The invention relates to assays to detect targets for microRNAs, microRNA-like molecules or functionally related gene regulators, by measuring changes to the length of the poly(A) tail associated with cellular mRNA.

Background Art

Most eukaryotic messenger RNAs (mRNAs) are synthesized with a terminal adenosine extension called a poly(A) tail which is added to nascent mRNAs at a species-specific length (-200 in humans). This tail can be altered during the lifetime of the mRNA and such alterations in length can result in changes in translation and stability of specific mRNAs. Typically, shortening of the poly(A) tail is associated with translational silence whereas lengthening is associated with translational activation. Thus, tail length control plays a central role in gene expression. For example, during the development of oocytes, maternally deposited mRNAs are stored in a short, oligoadenylated form until oocytes are stimulated to undergo further development, at which point specialized enzymes (cytoplasmic adenylases) are recruited to lengthen the poly(A) tail to activate translation. Such changes in the functional status of maternal mRNA are orchestrated by regulatory complexes that bind to the 3' untranslated region (UTR) of the mRNA. Conversely, cytoplasmic deadenylases can be stimulated to remove the poly(A) tail, which in turn promotes translational silencing. An important emerging example of this is the mechanism of translational repression employed by small, non-coding regulatory RNAs, called microRNAs (miRNA; ~ 22 nucleotides in length). In most known cellular situations, binding of miRNA to the 3'UTR of target mRNAs stimulates mRNA deadenylation that is not necessarily accompanied by complete decay of the mRNA. Furthermore, the widely used method of gene knockdown by RNA interference deploys short interfering RNA (siRNA), or precursors thereof, to cells, which are designed to lead to cleavage of a perfectly matching cellular RNA target. It is known that siRNAs can function like miRNAs and vice versa, and a common problem with RNA interference approaches are the so-called off-target effects elicited by siRNAs. It is commonly accepted that such off-target effects originate to a significant extent through the siRNA acting on other cellular RNA in a miRNA-like fashion. Thus miRNA expression, either endogenous or exogenously added, induces a shortening of poly(A) tails of mRNAs. Conversely, introduction of miRNA-antisense oligonucleotides, by inhibiting the mi RN A-d riven deadenylation, allow long-tailed mRNAs to accumulate. It has been reported that under specific cellular conditions miRNAs can alternatively stimulate translation of their targets. It is currently unknown whether this stimulation is associated with a lengthening of the poly(A) tail but this may well be the case. Any such alterations in mRNA poly(A) tail length can potentially be used to identify targets of miRNAs and miRNA-like molecules, which is a significant challenge in the field.

Changes in poly(A) tail length can profoundly alter the fate of gene expression by changing the rate of protein synthesis by the translation machinery. In previous research, the present inventors have been instrumental in pioneer work using microarrays to sample mRNAs from translating polysomes versus non-translating fractions (Beilharz TH and T Preiss, Brief Funct Genomic Proteomic, 2004. 3: 103-11; Preiss T, J Baron-Benhamou, W Ansorge, and MW Hentze, Nat Struct Biol, 2003. 10: 1039-47) and in fractionation of mRNA based on poly(A) tail length to understand the functional importance of tail length regulation (Beilharz TH and T Preiss, RNA, 2007. 13: 982-97; Lackner DH, TH Beilharz, S Marguerat, J Mata, S Watt, F Schubert, T Preiss, and J Bahler, MoI Cell, 2007. 26: 145-55). Because the length of the poly(A) tail associated with mRNAs in the transcriptome is indicative of their functional expression in the cell, tail length regulation is likely to be a feature of every situation where alteration of gene expression is appropriate, including during the cell cycle, throughout development, in response to environmental insult, and in the disease process. While current genome-wide technologies can measure total mRNA abundance differences very effectively, determination of the functional pool of mRNA is more complicated. Because changes in adenylation state reflect on mRNA utilisation, the present invention can be used as a valuable tool for the differentiation between the translationaly active from inactive pools of mRNA within the cell.

An important goal in the study of miRNA, and their exploitation in disease therapy, is to identify mRNA targets for the hundreds of miRNAs encoded by the human and other genomes. Bioinformatic predictions are problematic because of the limited complementarity required for miRNA-target interaction. Algorithms employ additional constraints (e.g. conservation, multiplicity of sites within the 3' UTR) to raise prediction accuracy above noise level. These constraints are based on working definitions of a miRNA target but many true miRNA targets ultimately may not comply with these constraints. Bioinformatic target predictions therefore comprise of many false positive and negative assignments. The challenge is to develop facile experimental techniques for genome-wide identification of miRNA targets. An elegant solution is via pull-down of affinity-tagged miRNA or RISC components, followed by analysis of enriched mRNAs. But these approaches are technically demanding, which limits their broad application. Microarray- based transcriptome profiling, in the presence and absence of a miRNA, relies on miRNAs causing target destabilisation. While technically simple, this approach canhot distinguish between transcriptome changes due to direct miRNA-target interaction and secondary effects. It also misses miRNA targets that are predominantly regulated at the translational level. Several methods have been developed to measure mRNA poly(A) tail lengths on purified RNA samples. The classical approach to sizing the poly(A) tail of a specific mRNA uses oligonucleotide-mediated digestion by RNase H to liberate a smaller 3' fragment from the mRNA under study, followed by Northern blotting to detect and size this fragment. Any length of this fragment in excess of the known 3' UTR sequence length is then attributable to the presence of a poly(A) tail. Alternatively, a modified reverse-transcriptase PCR method termed the PoIy(A) Test (PAT) may be used. Its most advanced form, the Ligation Mediated Poly(A)-Test (LM-PAT, relies on the serial annealing and ligation of 5' phosphorylated oligo(dT)_12-i₈ linkers along the length of the poly(A)-tail and subsequent ligation of an oligo(dT) anchor primer, followed by reverse transcription. The resulting cDNA library is used to template PCR reactions employing a gene-specific forward primer and an anchor-specific reverse primer to generate a spectrum of amplicons that reflect the distribution of poly(A) tail lengths present on the mRNA. In previous work, the present inventors have used chromatography on poly(U) sepharose to fractionate cellular mRNA based on poly(A) tail length and employed microarray analysis of the resulting chromatography fractions to acquire poly(A) tail length information on mRNA in a massively parallel fashion. This method was termed polyadenylation state array analysis or PASTA (Beilharz TH and T Preiss 2007; Lackner et al 2007; Beilharz TH and T Preiss, Methods 2009. 48:294-300).

The present invention relates to the adaptation of global or gene-specific technologies to measure mRNA poly(A) tail lengths as a method to identify miRNA targets and to monitor miRNA activity. In addition, the present invention also provides new methodology to measure poly(A) tail lengths in a global or gene-specific fashion that can be used in miRNA target identification assays, as well as in the characterisation of other gene regulators that function through the modulation of mRNA tail lengths and their targets. Disclosure of Invention

In general, the present invention involves using measurements of mRNA poly(A) tail length as a means to identify and characterise miRNA targets.

The invention in this regard relates generally to detecting or measuring changes of poly(A) tail lengths on mRNA molecules that result from miRNA action on a cell.

In a first aspect, the present invention provides an assay for detecting miRNA targets comprising:

(a) treating a cell with a miRNA;

(b) detecting poly(A) tail lengths of mRNA molecules of the treated cell; (c) detecting poly(A) tail lengths of mRNA molecules of a corresponding untreated cell;

(d) comparing poly(A) tail lengths on mRNA molecules between the treated cell and untreated cell; and

(e) detecting targets of the miRNA from changes in poly(A) tail lengths of mRNA molecules in a treated cell compared with an untreated cell.

The cell may be treated by exposing the cell directly with the miRNA or indirectly through an introduced construct that gives rise to the production of the miRNA by the cell.

In a second aspect, the present invention provides an assay for detecting miRNA targets comprising:

(a) treating a cell to inactivate a miRNA produced by the cell;

(b) detecting poly(A) tail lengths of mRNA molecules in the treated cell;

(c) detecting poly(A) tail lengths of mRNA molecules in a corresponding untreated cell; (d) comparing poly(A) tail lengths on mRNA molecules between the treated cell and untreated cell; and

The cell can be treated with an agent that inactivates the miRNA. Preferably, the agent is an oligonucleotide antisense to the miRNA. It will be appreciated that the miRNA can be inactivated directly or prevented from being made by the cell. The analyses may be carried out in a cell that expresses minor levels of the miRNA. In this case, a comparison may be made between a cell that was treated to inactivate the miRNA, and a cell that was treated with additional levels of the miRNA. In this manner, using a combination of the two approaches described in the first and second aspect of the present invention.

In one preferred form of the assay for multiple targets, the cell is treated with an agent that inactivates, or blocks the production of a group of miRNAs, or all cellular miRNAs.

An example of such an agent may be an RNA interference agent that depletes cells of a miRNA processing enzyme such as Drosha.

It is expected that in most cellular contexts a miRNA will cause the deadenylation and repression of its mRNA targets. The data from the assays according to the present invention could be screened for mRNAs that posses a longer (or shorter) poly(A) tail when the miRNA under study is inactive or absent, and have a shorter (or longer) poly(A) tail if the miRNA is present or active. As known cellular conditions typically cause miRNA-mediated tail shortening, the preferred form of the present invention is screening for mRNAs that posses a longer poly(A) tail when the miRNA under study is inactive or absent, and have a shorter poly(A) tail if the miRNA is present or active.

The present invention also provides an assay for screening miRNAs or miRNA- like molecules that are effective in causing deadenylation of a given target mRNA.

In a third aspect, the present invention provides an assay. for detecting an effect on the polyadenylation status of target mRNA.

Examples include, but are not limited to, screening a cellular condition, administration of a substance or other treatment for an effect on the polyadenylation status of target mRNA in a cell. The assay may be used to screen a chemical library for candidate drugs that interfere with the action of a specific miRNA or with specific . components of the cell (e.g. deadenylase enzymes or their cofactors) required for miRNA action.

Preferably, the assay according to the third aspect of the present invention comprises: .

(a) exposing multiple aliquots of cells that express a target mRNA in parallel to a range of test agents; (b) detecting poly(A) tail lengths of target mRNA molecules in each treated cell aliquot;

(c) detecting poly(A) tail lengths of target mRNA molecules in a corresponding untreated cell aliquot; (d) comparing poly(A) tail lengths on target mRNA molecules between the treated cell and untreated cell aliquots; and

(e) identifying the test agents that facilitates or inhibits rηiRNA action on the mRNA target from changes in poly(A) tail lengths of target mRNA molecules in a treated cell compared with an untreated cell. The target mRNA may be endogenous or introduced to the cells.

A cell-free system that can mimic miRNA action in a cell is contemplated by the present invention.

In another preferred form, the present invention can be used as a "Genome-Wide PoIy(A) Test' (GW-PAT) assay. In a fourth aspect, the present invention provides an assay to measure poly(A) tail lengths on individual mRNAs or all mRNAs in a mixture. This assay may be used in the context of miRNA target detection according to the first, second or third aspects of the present invention, but may also in the detection and characterization of other gene regulators that affect the length of poly(A) tail on target mRNA. It will be appreciated that poly(A) tail length can be detected directly on mRNA or indirectly on complementary cDNA, or amplified products. As such, the miRNA target identification assays according to the present invention may be carried out by any suitable means such as Ligation Mediated PoIy(A)-TeSt (LM-PAT), high-precision Northern analysis of isolated RNA, rapid amplification of cDNA ends (RACE), or PASTA (detailed in Background Art; Beilharz TH and T Preiss, 2007; Lackner et al, 2007; Beilharz TH and T Preiss 2009).

In a variation of PASTA, the constituent mRNAs in poly(U) chromatography fractions could further be identified and quantified by some form of 'sequence census' approach such as deep sequencing methodology instead of microarray analysis. More preferably, measurements of poly(A) tail length may be performed using new poly(A) tail length assays, loosely based on an LM-PAT approach. One advantage of these new assays is that there is no requirement for amplification in the assay. Another advantage is that the assays do not require a separation step using chromatography or electrophoresis.

In one particularly suitable genome-wide or massively parallel configuration (termed herein 'Genome-Wide PoIy(A) Test¹, GW-PAT), the assay comprises: (a) providing 5' phosphorylated oligo(dT) primers having a detectable label to mRNA isolated from cellular material;

(b) allowing the oligo(dT) primers to anneal to a poly(A) tail of the mRNA;

(c) ligating the annealed oligo(dT) primers to form labeled poly(dT) oligonucleotide having a length substantially corresponding to the length of the poly(A) tail; and (d) assigning a length to the poly(A) tail according to the amount of detectable label on the poly(dT) section.

The assay may further comprise:

(e) providing an oligo(dT)-anchor primer that is detectably labeled; and

(f) annealing the oligo(dT)-anchor primer to the 3' end of a poly(A) tail, followed by ligation of the oligo(dT)-anchor primer to the preassembled poly(dT) section.

The label can be the same or different to the label on the oligo(dT) linkers. If the labels on the anchor and linker are different, it is possible to differentiate between the two entities.

To facilitate detection of the mRNA-associated labelled poly(dT), a reverse transcription step is carried out to form cDNA molecules complementary to the mRNA. The resulting cDNA pool is referred to as the 'test' cDNA library, and is purified away from reaction components and mRNA template.

In a parallel reaction, a 'reference' cDNA library is generated from a separate aliquot of the RNA isolated from cellular material under study using a VN/oligo(dT)- anchor primer (V: G, T, or C; N: G, A, T, or C) (Beilharz TH and T Preiss 2009; Beilharz TH, Humphreys DT, Clancy JL, Thermann R, Martin DIK, Hentze MW and T Preiss, PLoS ONE, 2009 4(8) e6783). The VN/oligo(dT)-anchor primer is detectably labelled with a label distinct from the label used for the generation of the test cDNA library.

The two cDNA libraries are mixed and hybridised to a microarray suitable for two-colour detection of mRNA expression ("transcriptome profiling").

Following washing, scan and image analysis, there will be a ratio between signals in both channels for each detectable feature of the microarray. This ratio will relate directly to the extent of poly(A) tail present on the original mRNA detected at that feature.

Other possible assay configurations include:

5' phosphorylated Oligo(dT) primer: A particularly suitable length of the labelled oligo(dT) primer would be from 12 to 18 thymidylates, provided as a mixture of equal proportions. However, a population of primer of uniform size may also be suitable. The use of primers of a shorter length, such as hexamer dT, should also be possible and may improve labelling density as well as allowing a more finely graded relationship between signal intensity and length of poly(A) tail. Oligo(dT)-anchor primer & VN/oligo(dT)-anchor primer: A particularly suitable structure for these primers would be (in 5' to 3' direction) a unique anchor sequence segment that can efficiently form a reverse priming site for PCR, followed by a segment of 12 thymidine residues (plus VN in the case of the VN/oligo(dT)-anchor primer). However, a shorter dT segment (down to hexamer length) may also be suitable, as long as conditions can still be achieved to efficiently hybridise the primer to the 3' end of poly(A) tails.

The assay may also be configured such that there is no need for the provision of anchor primers. However, the inclusion of the anchor primers provides several advantages. First, the test and reference cDNA libraries can then be further purified or immobilised using a nucleic acid of sequence complementary to the anchor linked to a solid support. This can serve as a clean-up step, or be used to obtain a global measure of the cellular polyadenylation state (i.e. the fluorescence in the purified sample is an aggregate signal derived from all poly(A) tails in the sample. Secondly, the test and reference cDNA libraries may then also be used for conventional single target LM-PAT, for instance to verify the initial microarray data. Thirdly, the label on the oligo(dT)- anchor primer may contribute to overall signal intensity from the completed poly(dT) assembly. Fourthly, if the label on the oligo(dT)-anchor primer used for generating the test cDNA library is distinct from the label used on the oligo(dT) primer, it is possible to differentiate between the two entities. This could be used as an internal reference on the microarray, thus dispensing with the need to generate and analyse a separate reference cDNA library. For the latter use, it would also suffice to use a differently labelled oligo(dT) primer without anchor to hybridise to 3' terminal stretches of the poly(A) tail.

Ligation and reverse transcription steps: Ligation of the poly(dT) assembly and extension of it by reverse transcription generates a stable molecular copy of the poly(A) tail configuration on each mRNA. While this has many advantages, it will also be possible to dispense with the reverse transcription step or both the reverse transcription and ligation steps, provided that the initial mRNA/labelled oligo(dT) hybrid is assembled in the presence of an immobilised oligonucleotide complementary to the body of the mRNA. An example of this could be to jointly hybridise a mRNA sample and labelled oligo(dT) on the surface of a microarray exhibiting features that are antisense to sequences within the body of mRNAs. A suitable reference signal for mRNA abundance could be obtained from the additional provision in solution, of a differentially labelled oligonucleotide complementary to a separate sequence within the body of the mRNA. A variant of this approach may be that the annealing of the soluble components will be done separately, and the resulting hybrid transferred to the microarray surface under conditions that preserve the hybrid.

Capturing the labelled poly(dT) oligonucleotide on a solid support: The above mentioned microarrays may be of any commonly used type, provided that they are suitable for the detection of mature mRNA. As described above, the assay can use dual colour microarray hybridisation and detection. It will however, be possible to adapt the procedure to a single colour microarray platform (such as Affymetrix), by analysing test and reference cDNA libraries on two separate arrays and obtaining signal ratios from the comparison of both array scans. The solid support may further be a bead, membrane, slide, nucleic acid miroarray, or any other surface capable of presenting a nucleic acid primer that can select individual mRNA or cDNA molecules from a mixture. As described above, the assay can use microarray technology to generate polyadenylation state data for multiple mRNAs in a parallel fashion. It is however, possible that other technology, which operates in a similarly parallel fashion as microarray, may substitute for microarrays. Further, it may be desirable to reconfigure the assay to measure the polyadenylation state of just one or a small number of mRNA at any given time. In these cases, a solid support in the form of a bead, membrane, or other coated surface (e.g. multiwell plate) may be particularly suitable to capture the labelled mRNA/oligo hybrid or cDNA molecules.

The labelling system having a detectable label can be a fluorescent label, enzyme/substrate system or radioactive label. Any suitable nucleic acid synthesis approach may be used to generate the modified oligonucleotides required for the procedure. In a preferred approach, the fluorescent label is selected from available fluorophors that emit at wavelengths captured by standard laboratory slide scanners. Examples are the Alexa-[555] and Alexa-[647] fluorophores.

5' phosphorylated oligo(dT) linkers consisting of 12 to 18 thymidylate residues labelled with Alexa-[555] might be particularly suitable for this poly(A) tail assay. The labelling can be direct, by the incorporation of fluorophore-derivatised thymidylates into the oligonucleotide. The oligo(dT)-anchor primer or VN/oligo(dT)-anchor primer may be labelled in this way with Alexa-[555] and Alexa-[647], respectively. One or several fluorophore groups may be attached to each oligonucleotide to improve signal strength while avoiding signal quenching. Labelling may also be achieved using oligonucleotides linked to a hapten such as biotin, with attachment of the fluorophore to the hapten occurring at a later stage (e.g. during microarray staining and detection). In this case the oligonucleotides may be synthesised with a biotin-derivatised thymidylate residue.

The detection system: Detection of the oligo/poly(dT) assemblies may be done either at the mRNA/labelled oligo(dT) hybrid or cDNA level, depending on the exact assay configuration. The principle of detection (fluorescence, luminescence, radioactivity) will vary depending on the labelling system chosen. The choice of imager/reader equipment will also depend on whether a genome-wide or gene-specific assay format is implemented. In a preferred form, fluorescently labelled test and reference cDNA libraries are hybridised to a conventional microarray and the array is scanned using dedicated equipment such as the Axon GenePix scanner. Following image analysis, the ratio between signals in both channels for each detectable feature of the microarray will provide a measure of the extent of poly(A) tail present on the original mRNA detected at that feature. All mRNAs detectable in the assay may then be ranked by that ratio in order of increasing propensity of having a long tail. The ratio may further be measured under multiple different cellular or experimental conditions (e.g. with or without active miRNA) and any change in the ratio will then be indicative of a change in mRNA polyadenylation state between conditions.

Further configurations of the assay may require autoradiography/phosphor storage imaging equipment to detect radioactive labels, or instruments to detect chemiluminescence. Genome-wide or otherwise massively parallel formats of the assay will require instruments capable of acquiring multiple readings either in parallel or in quick succession (e.g. microarray scanners or multiwell plate readers). Single gene- specific versions of the assay will be able to employ equipment capable of measuring one sample at a time. For instance, using a fluorescent labels on the oligonucleotides it will be possible to obtain a quantitative measure of poly(A) tail length simply by purifying the oligo/poly(dT) assemblies derived from a single mRNA by capture with a gene- specific antisense oligonucleotide on solid support, followed by spectrophotometric quantification of co-captured fluorescence. In a fifth aspect, the present invention provides a kit for use in detecting changes in length of poly(A) tails of mRNA comprising:

5' phosphorylated oligo(dT) primer mix having a detectable label; oligo(dT)-anchor primer having a detectable label capable of binding to the 3' end of a poly(A) tail of mRNA; and a universal anchor sequence for affinity purification, immobilisation, and/or optionally PCR verification; and

VN/oligo(dT)-anchor primer having a detectable label capable of binding to the 3" end of a poly(A) tail of a mRNA; and a universal anchor sequence for affinity purification, immobilisation, and/or optionally PCR verification.

The kit may firther comprise one or more of; reagents required to perform ligation reactions to join multiple primers annealed to a poly(A) segment; reagents required to perform reverse transcription reactions to generate test and reference cDNA libraries; and immobilised primers capable of binding to specific sequences in mRNA. The label may be a hapten.

Preferably, the kit contains primers for a plurality of different mRNAs (e.g. in an arrayed format).

Preferably, the kit further includes instructions to perform a gene-specific or genome-wide assay to detect length of poly(A) tails. Multiple variations of such a kit are possible that provide a collection of reagents tailored to implement the various alternate oligo/poly(dT) assembly detection strategies detailed above and summarised in Figure 7.

In a sixth aspect, the present invention provides use of the kit according to the fifth aspect of the present invention in a genome wide assay to detect mRNA poly(A) tail lengths.

Preferably, the kit is used for detecting targets of miRNAs, or other post- transcriptional gene regulators that affect mRNA poly(A) tail length. Throughout this specification, unless the context requires otherwise, the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element, integer or step, or group of elements, integers or steps, but not the exclusion of any other element, integer or step, or group of elements, integers or steps.

Any discussion of documents, acts, materials, devices, articles or the like which has been included in the present specification is solely for the purpose of providing a context for the present invention. It is not to be taken as an admission that any or all of these matters form part of the prior art base or were common general knowledge in the field relevant to the present invention as it existed in Australia before the priority date of each claim of this specification.

In order that the present invention may be more clearly understood, preferred embodiments are described with reference to the following drawings and examples.

Brief Description of the Drawings

Figure 1 shows prior art LM-PAT assay for measurement of poly(A) tail dynamics in eukaryotic cells. A) Schematic of the LM-PAT assay oligo(dT)₁₂-i8 linkers are serially ligated by T4 DNA ligase at 42°C followed by ligation of an anchor and reverse transcription. As a separate reaction, a 3' UTR anchored TVN-PAT primer is used to generate cDNA with a fixed, short poly(A) tail (SEQ ID NOs: 10 - 19). B) Subsequent reverse transcription and PCR generates a spectrum of amplicons that reflect the distribution of poly(A) tail lengths present on the mRNA. C) PAT amplicons are analysed by laser scanning of ethidium bromide stained 2% high-resolution agarose gels. Samples are derived from cells that were transfected with a Renilla Luciferase reporter construct that responds to let-7 miRNA (3xb, exhibiting three 3' UTR let-7 target sites), or a control construct that has mutated target sites (mut) (Beilharz et al, 2009). The 3xb mRNA is rapidly deadenylated in response to let-7 binding, generating an oligoadenylated mRNA, whereas the mut mRNA retains a long poly(A) tail. ACTB and GAPDH are endogenous mRNAs not known to be targeted by let-7. Figure 2 shows deadenylation of miRNA targets is reversed by inhibition of miRNA with antisense LNA oligonucleotides. A) Cells were transfected as in Figure 1 , except that anti-let-7 LNA or a control LNA targeting the irrelevant miRNA miR-499 was further added to the cells. The 3xb mRNA is rapidly deadenylated in response to let-7 binding, generating an oligoadenylated mRNA. This is the situation seen in cells co- transfected with a control LNA targeting the irrelevant miRNA miR-499. Co-transfection of anti-let-7 LNA abrogates this deadenylation and allows longer-tailed forms of 3xb mRNA to accumulate, analogous to those accumulated for the mut mRNA that cannot be bound by let-7. The endogenous let-7 target HMGA2 mRNA is similarly protected from deadenylation by anti-let-7 LNA but not by anti-miR-499 LNA. GAPDH is not targeted by these miRNAs. Arrowheads indicate short, oligoadenylated mRNAs. B) Similar to A) except that cells were transfected only with the LNAs and LM-PAT assays performed for HMGA2, IGF2BP1 (both known let-7 targets), as well as GAPDH as a control. Figure 3 shows separation of mRNA on poly(U)₁₂o beads. A) Total HeIa cell RNA

(input) was bound to beads before thermal elution of mRNA (schematic). (SEQ ID NOs: 20 - 27) B) poly(U) chromatography elution profiles of selected mRNAs and relevant controls (TVN-PAT: control for short poly(A) tails; Total RNA: mRNA in unfractionated cellular total RNA). GAPDH and ACTB are a control mRNAs with long poly(A) tails. HMGA2 is a let-7 target known to be presenting in untreated HeLa cells with short tails. Shown are LM-PAT products.

Figure 4 illustrates the logic of miRNA target identification based on poly(A) tail length changes. A) miRNA target candidates will be selected from genome-wide (e.g. microarray) data by requiring the target to be short-tailed in the presence of a miRNA and gaining in tail length in its absence (or vice versa, see text for details). Shown is the example of specifically interfering with a single miRNA (e.g. miR-X). B) While neither required nor necessarily always appropriate, it may be advantageous to further parse the experimentally identified candidates by requiring them to exhibit bioinformatically predicted target sites for the miRNA(s) under study. Shown is the example of comparing cells with a full complement of miRNAs or without them (as achievable by knock-down of the processing enzyme Drosha).

Figure 5 shows an outline of a screen for compounds that affect miRNA function. Cells expressing a miRNA target and a control mRNA, are exposed to a library of compounds (e.g. a chemical library) and then poly(A) tail lengths of the mRNAs are measured. Compound Y is effective in selectively blocking miRNA action.

Figure 6 shows the novel GW-PAT assay principle using two-colour microarray detection. Serially ligated fluorescent Alexa-[555]-oligo(dT) linkers generate a length measurement that is based on the intensity of signal. To normalise out any intensity variation introduced by mRNA abundance a reference sample is generated with the VN/oligo(dT)-anchor primer derivatised with Alexa-[647]. Competitive hybridisation on microarray slides affords a colour intensity -based differential between oligo-adenylated and long-tailed mRNAs. Treatment -dependent changes in adenylation state are measured on separate microarray-slides. (SEQ ID NOs: 20 - 32).

Figure 7 shows a schematic outlining multiple variations on the novel poly(A) tail length assay principle.

Figure 8 shows specialised requirements for the novel poly(A) tail length assay. A) poly(A) tail length measurement is achieved by serial ligation of labelled oligo(dT)i_2-i₈ linkers. In this case the magenta star represents Alexa-[555]-dT but this could equally be biotin or another capture hapten. B) The oligo(dT)-anchor primer is likewise labelled with a 5'Alexa-[555] (or alternative detection hapten). C) The VN/oligo (dT)-anchor primer is labelled with a 5'Alexa-[647] (or alternative detection hapten). (SEQ ID NOs: 33 - 35).

Mode(s) for Carrying Out the Invention The term microRNA (miRNA) as used herein includes any short nucleic acid that acts in a miRNA-like manner. The nucleic acid may either be produced by cells themselves or generated through the introduction of a construct or agent into cells such as in the RNA interference procedure (e.g. short interfering RNA or siRNA). The miRNA-like molecule may carry out a physiological function, or be involved disease processes (as either a therapeutic or the target of a therapeutic intervention). The miRNA-like molecule may otherwise be of interest in the context of explaining, selecting or modifying cellular or organism properties. The term 'mRNA¹ as used herein includes any polyadenylated RNA species (protein-coding or noncoding) that can be targeted by a miRNA or that could be regulated by changes to its poly(A) tail. The inventors were the first to describe a role of the poly(A) tail in miRNA action

(Humphreys DT, Westman B, Martin DIK and T Preiss. Proc Natl Acad Sci (USA) 2005; 102:16961-6), and it has now been determined that miRNAs commonly, perhaps universally, cause target mRNA deadenylation (Beilharz et al, 2009).

In one form, the assay combines the ease of use of a microarray approach based on total RNA as input with selection based on mRNA deadenylation, a characteristic of miRNA action. In a first step, two cellular conditions are established that differ in cellular miRNA content in a predictable and verifiable way (Figure 4). For instance, the first cellular condition could simply be a cell line known to express a miRNA of interest. The second condition would be that same cell line after the introduction of an oligonucleotide antisense to that miRNA, which leads to its inactivation. Alternatively, cells that normally do not express a certain miRNA, or low levels of it (condition 1 ), could be transfected with a construct leading to the expression of the miRNA (condition 2). A third and more comprehensive approach would be to compare a cell line before and after the removal of a miRNA processing enzyme such as Drosha (Figure 4), by RNAi or other suitable intervention.

Total RNA is extracted from each of the two cellular conditions and subjected to genome-wide tests for poly(A) tail length. From this genome-wide survey, mRNAs are then selected that change from a shorter poly(A) tail length in the presence of the miRNA to a longer poly(A) tail length in its absence (Figure 4).

The genome-wide measurement of poly(A) tail lengths could be accomplished using poly(U) sepharose chromatography (see Figure 3) followed by microarray analyses of tail length-fractionated mRNAs (using the published PASTA approach). Alternatively, the constituent mRNAs in the chromatography fractions could be identified and quantified by a deep sequencing methodology. Finally, it could be performed using a novel poly(A) tail length assay described above in the present specification.

The target shortlist from the tail length comparison will be substantially enriched for verifyable miRNA targets but it could be further parsed against lists of miRNA candidates derived from other, more error prone methods such as bioinformatic predictions mRNA targets (publicly available on several web sites; Figure 4).

In another form, the assay can measure tail length changes of a single target mRNA of interest and be configured to screen a variety of cellular conditions or experimental interventions for an effect on the polyadenylation status of the target. The purpose of this could be, for instance, to screen a chemical library for candidate drugs that interfere with the action of a specific miRNA or with specific components of the cell (e.g. deadenylase enzymes or their cofactors) required for miRNA action. Alternatively, the screen could be used to determine which of a variety of miRNA-like molecules has the most pronounced effect on this target. miRNAs are not the only gene regulators that acts through changes in the adenylation state of their target mRNAs. Other examples include, but are not limited to, the CPE-binding protein during oocyte maturation and brain activity or the PUF RNA- binding proteins in yeast and other eukaryotes (Figure 5). The methods described above for use with miRNA are further applicable to the use with any gene regulator that effects changes in the adenylation state in a similar manner. MATERIALS & METHODS Cell culture

HeLa cells are maintained in DMEM with 5% FCS, supplemented with glutamine and penicillin/streptomycin.

Cell transfection

Cells were seeded in a 24-well plate and transfections were performed in triplicate at -60-70% confluency by adding a preincubated (30min, 25°C) transfection solution (200 μl) and 300 μl of Opti-MEM I (Invitrogen) to each well. The 200 μl^'of transfection solution (in Opti-MEM I) contains 1 μl Lipofectamine 2000 (Invitrogen), 600 ng of Firefly luciferase (F-luc) pGL3 plasmid (Promega), and 30 ng of Renilla luciferase- encoding (R-luc) encoding plasmid pRL-3xbulge ('3xb', a target of let-7 family miRNAs) or pRL-3xbulgemut ('mut', seed region mutant control) (Pillai RS, SN Bhattacharyya, CG Artus, T Zoller, N Cougot, E Basyuk, E Bertrand, and W Filipowicz, Science, 2005. 309: 1573-6; Clancy JL, M Nousch, DT Humphreys, BJ Westman, TH Beilharz, and T Preiss, Methods Enzymol, 2007. 431 : 83-111; Beilharz TH, Humphreys DT, Clancy JL, Thermann R, Martin DIK, Hentze MW and T Preiss, PLoS ONE, 2009 4(8) e6783). Medium is replaced ~8 hours after transfection and cells are harvested at times specified. For anti-miRNA LNA transfections, 200 μl of transfection solution contains 5 pmoles of antisense LNA-modified oligonucleotide. LNA modified anti-miR oligonucleotides are obtained from Exiqon (MA, USA) with the following sequences: hsa-let-7a δ'-AACTATACAACCTACTACCTCA-S' (SEQ ID NO: 1 ), hsa-miR-499 5'-TTAAACATCACTGCAAGTCTTAA-3' (SEQ ID NO: 2). When using synthetic miRNA, the transfection solution contains 1 pmole of duplex miRNA. The Synthetic duplex let-7a sense sequences used is: UGAGGUAGUAGGUUGUAUAG UU (SEQ ID NO: 3).

Combinations of the above listed additions to the transfection mix can also be used (see Figure 1 and Figure 2).

RNA isolation HeLa cell total RNA was purified using Trizol (Invitrogen) according to the manufacturer's instructions. Ligation Mediated PoIy(A)-TeSt (LM-PAT)

To measure poly(A) tail lengths of individual mRNAs the Ligation-Mediated PoIy(A) Test (LM-PAT) assay can be used as described (Salles FJ and S Stickland. PCR Methods Appl 1995 4:317-321 ; Salles et al. Methods in Enzymology 1999 17:38- 45; Beilharz TH and T Preiss, 2007; Clancy et al 2007; Beilharz TH and T Preiss 2009). Briefly, RNA is first incubated with 5' phosphorylated oligo(dT)₁₂-i8 primers in the presence of T4-DNA ligase at 42°C. This creates a poly(dT) copy of each mRNA's poly(A) tail within the sample. Addition of an excess of anchor-(dT)₁₂ primer and further incubation at 12°C favours annealing of the anchor primer to unpaired poly(A) ends and ligation to the poly(dT) stretch. This assembly is used to prime synthesis of first-strand cDNA by reverse transcription. Aliquots of this cDNA are then used as templates in PCR reactions with primers specific to the mRNA 3'-UTR of choice and the anchor region. To generate a size marker for the shortest possible LM-PAT product, cDNA is additionally synthesised with an anchor-(dT)12VN primer to clamp the anchor to the 3'UTR-poly(A) junction. PCR-products generated from this 'clamped' cDNA library are labelled 'TVN-PAT' in Figure 1C. LM-PAT PCR products are resolved by agarose gel electrophoresis and detected by fluorescence scans using an FLA-5100 imager and MULTIGAUGE software (Fujifilm) or similar device. The oligo(dT)-anchor primer used in the RT step and common reverse PCR primer is GCG AGC TCC GCG GCC GCG TTT TTT TTT TTT (SEQ ID NO 4). The VN/oligo(dT)-anchor primer used in the RT to generate a size marker for the shortest possible LM-PAT product is GCG AGC TCC GCG GCC GCG TTT TTT TTT TTT VN (SEQ ID NO: 5). Gene-specific primers are as follows; R-luc-PAT, GGC CGC TTC CCT TTA GTG AG (SEQ ID NO: 6); H.s GAPDH- PAT, GGA CCA CCA GCC CCA GCA AG (SEQ ID NO: 7); H.s HMGA2-PAT, CGC TTG CTT GTT GAA AAT ATT TCT CTA G (SEQ ID NO: 8); H.s ACTB-PAT, GAG CCT TCG TGC CCC CCC TTC C (SEQ ID NO: 9).

Polyadenylation state array (PASTA) analysis

The procedure is based on published protocols (Beilharz TH and T Preiss, 2007; Beilharz TH and T Preiss 2009; and references therein), with the following modifications. HeLa cell total RNA is purified on on poly(U)120 beads. The poly(U)i₂o matrix is generated by _{in vιtro} transcription. of a poly(A)-stretch flanked by sequences for the inclusion of biotinylated cytidylate. The poly(U)i₂o capture RNA is immobilized on streptavidin coated beads (NEB) in 0.1 M NaCI and 10 mM Tris-HCI (pH 7.4) at room temperature and washed extensively to remove low affinity RNA molecules. Then follow washes in elution buffer (EB; 0.1 M NaCI₁ 0.01 M EDTA, 0.5 M Tris-HCI at pH 7.4, 0.2% SDS, 25% formamide) and high salt binding buffer (HSBB; 0.7 M NaCI, 0.01 M EDTA, 0.5 M Tris-HCI at pH 7.4, 0.2% lauryl sarcosine, 12% formamide), one time each for 5 min at 55°C. For each run, 300 mL of HSBB, 50 μg of total HeLa RNA is denatured for 5 min at 70⁰C in 1.5 mL tubes. For binding, tubes are incubated in an Eppendorf

Thermomixer (1100 rpm in a cold room), first for 10 min at 55°C, followed by cooling to 12°C for 60 min (including 25 min ramp-down time). The matrix is washed four times with HSBB at 12°C for 5 min. Each thermal elution step is performed in the Thermomixer, by resuspending the matrix in 300 mL EB at the specified temperature for 5 min. The supernatant is collected from the tubes after 30 seconds in a magnetic field. The eluted mRNA is precipitated in 2 vol ethanol and 1/10 vol of 5 M NaCI, and coprecipitant (glycogen or Pellet Paint, Novagen). mRNA is finally resuspended in H₂O and then subjected to probe generation and microarray analysis as described in (Beilharz TH and T Preiss, 2007) or following the instructions of the microarray manufacturer.

Genome-wide poly(A) test (GW-PAT)

In a preferred configuration of the assay, the first steps of the method follow the described protocol for the Ligation Mediated PoIy(A)-TeSt (LM-PAT) up to the cDNA library stage, with the important modification that the primers employed are fluorescently labelled. The oligo(dT)₁₂-i₈ mix is labelled with Alexa-[555] and the oligo(dT)-anchor primer and VN/oligo(dT)-anchor primer are labelled with Alexa-[555] and Alexa-[647], respectively. The test (Alexa-[555]) and reference (Alexa-[647j) cDNA libraries thus generated are then analysed directly by two colour microarray following the instructions of the microarray manufacturer.

In the traditional LM-Pat assay, the resulting cDNA library serve as a template for PCR amplification of products that reflect the lengths present on a single mRNA in the original sample. These PCR products need to be resolved by gel electrophoresis to generate an interpretable result. The PASTA approach generates genome-wide data but requires a technically demanding poly(U) sepharose chromatography step. The GW-PAT according to the present invention has several advantages. First, it allows the massively parallel detection of tail lengths on multiple distinct mRNAs in one experiment (i.e. genome-wide). Secondly, there is no requirement for amplification in the assay. Thirdly, it does not require a separation step using chromatography or electrophoresis. Fourthly, as detailed above, the assay can be configured in multiple ways to suit different microarray platforms or other capture methodology, and to allow its use in single gene studies or high-throughput screens.

Figure 7 shows several preferred forms of the GW-PAT assay whereby a series of molecular manipulations of starting RNA results in cDNAs with differential associated fluorescent levels depending on the length of the poly(A) tail. However it should be appreciated that any of the steps leading up to this final cDNA can be exploited to modify the assay to suit different purposes. For example direct detection of the RNA assembly is also possible without further cDNA synthesis. Furthermore, in any given experiment samples are prepared from RNA isolated from cells that have been differentially treated to manipulate miRNA levels.

Adaptation of the LM-PAT assay to a genome-wide analysis tool (i.e. the GW- PAT assay) uses detection of the cDNA with a value added to the number of linkers ligated to the poly(A) tail. In the prior art LM-PAT assay, the length measure is direct, shorter or longer PCR amplicons are detected by electrophoresis. However, for most purposes, the actual length is unimportant and an indirect tail length "score" is sufficient. There is no amplification step in the present invention and increased signal is measured as a fluorescent score compared to a control. There are several readily applicable options available for both, labelling of the cDNA, and detection by microarray. An approach to a simple two-colour platform is described in the schematic below, and further possibilities for one-colour arrays are included as variations at the end of this report. To add "value" to the length score of poly(A) tail, the 5' phosphorylated oligo(dT) is internally derivatised with the cy3-equivalent dT-Alexa-[555] (Figure 8). The anchor primer is synthesised with 5' Alexa-[555] fluorophore (Figure 8). Thus the longer the poly(A) tail the stronger the fluorescent signal and associated tail length score. As reference, to normalise out differences in mRNA abundance, a separate sample is reverse-transcribed with the TVN-reference primer derivatised with cy5 equivalent Alexa-[647] (Figure 8). Thus the Alexa-[555] signal over the Alexa-[647] signal will distinguish between short and long-tailed mRNAs, and the spectrum of intensities are used to calculate a tail length score. Fluorescent detection in this regimen is thus of single mRNA molecules with minimally one Alexa-dye attached to the 5' end of the cDNA (see variations below for "brighter" versions of the assay). A tail length score is determined by competitive hybridisation of the Alexa-[647]-reference versus the Alexa-[555]-PAT. Since most slide scanners are set to capture equal fluorescence in each channel short-tailed mRNAs will seem less abundant in the test sample than the reference (even though both should contain a single fluorophore per mRNA molecule) and thus appear red and score low, long-tailed mRNAs will appear green and score high (Figure 6).

The present inventors have instigated a modification to control for the minimal length measured by this approach in a separate reference sample that is reverse transcribed from an 3.'UTR anchored PAT-TVN primer that anneals to the mRNA at the juxtaposition between the 3'UTR and poly(A) tail.

Let-7 binding to the 3'UTR of the 3xb mRNA stimulates deadenylation and accumulation of a short-oligoadenylated form after 6 hours of transient transfection. Two endogenous housekeeping genes are not deadenylated in this way. Co- transfection of antisense let-7 oligonuclotides abolishes this deadenylation and also that of an endogenous mRNA, HMGA2 (Figure 2).

The preferred GW-PAT assay developed by the present inventors can be used in the cell cycle, during development, and in the study of the circadian clock. Of particular and topical interest is its application in the study of miRNA target mRNA identification. For the determination of miRNA targets (including those generated by siRNA binding off target) the effect on tail length can be measured either by expression of a miRNA in a cell that does not normally express it or by introduction of anti-sense oligonucleotides that inhibit the miRNA function and allowing a pool of long-tailed forms to accumulate (Figure 2). Because many mRNAs are regulated in a combinatorial fashion by a suite of miRNA, a cell may need to be treated with the physiologically relevant suite of miRNA antisense molecules to find all the mRNAs targeted by miRNA driven changes during the disease process for example.

There is a formal possibility that a single fluorophore per mRNA is insufficient to accurately detect low abundance molecules. An alternative approach is to derivatise all oligonucleotides with biotin. The "Test" and "Reference" samples would then be hybridised to separate slides and detected by a signal amplifying product such as, by way of example, the genisphere technology whereby many fluorophores are attached to DNA dendrimers (large 3-dimentional DNA matrices derivatised with various haptens or antibodies) are targeted to the cDNA via anti-biotin antibodies. This approach might work well in on Affymetrix arrays. Thus separately biotinylated samples are hybridised to the array and subsequently detected by the highly fluorescent Anti-Biotin / Oyster®- 550 (900). It will be appreciated by persons skilled in the art that numerous variations and/or modifications may be made to the invention as shown in the specific embodiments without departing from the spirit or scope of the invention as broadly described. The present embodiments are, therefore, to be considered in all respects as illustrative and not restrictive.

Claims

Claims:

1. An assay for detecting a microRNA (miRNA) target comprising: treating a cell with a miRNA, or treating a cell to inactivate a miRNA produced by the cell, or treating a cell to prevent a miRNA from being produced by the cell; detecting poly(A) tail lengths of messenger (mRNA) molecules of the treated cell; detecting poly(A) tail lengths of mRNA molecules of a corresponding untreated cell; comparing poly(A) tail lengths on mRNA molecules between the treated cell and untreated cell; and detecting a target of the miRNA from changes in poly(A) tail lengths of mRNA molecules in a treated cell compared with an untreated cell.

2. The assay according to claim 1 wherein the cell is treated by exposing the cell directly with the miRNA or indirectly through an introduced construct causing production of the miRNA by the cell.

3. The assay according to claim 1 wherein the cell is treated with an agent that inactivates the miRNA.

4. The assay according to claim 3 wherein the agent is an oligonucleotide antisense to the miRNA.

5. The assay according to any one of claims 1 to 4 wherein multiple targets are detected in a cell treated with an agent that inactivates, or blocks the production of a group of miRNAs, or all cellular miRNAs.

6. An assay for detecting an effect on polyadenylation status of target mRNA comprising: exposing multiple aliquots of cells that express a target mRNA to a range of test agents in parallel; detecting poly(A) tail lengths of target mRNA molecules in each treated cell aliquot; detecting poly(A) tail lengths of target mRNA molecules in a corresponding untreated cell aliquot; comparing poly(A) tail lengths on target mRNA molecules between the treated cell and untreated cell aliquots; and identifying the test agents that facilitate or inhibit miRNA action on the mRNA target from changes in poly(A) tail lengths of target mRNA molecules in a treated cell compared with an untreated cell.

7. The assay according to claim 6 wherein the target mRNA is endogenous or introduced to the cells.

8. An assay for genome-wide polyadenylation status of mRNA comprising: providing 5' phosphorylated oligo(dT) primers having a detectable label to mRNA from cellular material; allowing the oligo(dT) primers to anneal to a poly(A) tail of the mRNA; ligating the annealed oligo(dT) primers to form labeled poly(dT) oligonucleotide having a length substantially corresponding to the length of the poly(A) tail; and assigning a length to the poly(A) tail according to the amount of detectable label on the poly(dT) section.

9. The assay according to claim 8 further comprising: providing an oligo(dT)-anchor primer having a detectable label; annealing the oligo(dT)-anchor primer to the 3' end of a poly(A) tail; and ligating the oligo(dT)-anchor primer to the annealed poly(dT).

10. The assay according to claim 9 wherein the detectable label on the oligo(dT)- anchor primer is the same as the label on the phosphorylated oligo(dT) primer or the labels are different.

11. The assay according to any one of claims 8 to 10 further comprising: forming a test cDNA library by carrying out a reverse transcription step to form cDNA molecules complementary to the mRNA to facilitate detection of the mRNA-associated labelled poly(dT).

12. The assay according to claim 11 further comprising: forming a reference cDNA library from a separate aliquot of the mRNA isolated from cellular material using a VN/oligo(dT)-anchor primer, wherein V is G, T, or C; and N is G₁ A, T, or C) and the VN/oligo(dT)-anchor primer is detectably labelled with a label distinct from the label used for the generation of the test cDNA library.

13. The assay according to claim 12 wherein the signals of label of the test cDNA library and the label of the reference cDNA library are compared, wherein the ratio of signals from the two labels relates directly to the extent of poly(A) tail present on the original mRNA.

14. The assay according to claim 12 wherein the test cDNA library and the reference cDNA library are mixed and hybridised to a solid support and signals of the labels are measured, wherein the ratio of signals from the two labels relates directly to the extent of poly(A) tail present on the original mRNA.

15. The assay according to claim 14 wherein the solid support is a bead, membrane, slide, nucleic acid microarray, or any other surface capable of presenting a nucleic acid primer that can select individual mRNA or cDNA molecules from a mixture.

16. The assay according to claim 15 wherein the a solid support is a microarray suitable for two-colour detection of mRNA expression.

17. The assay according to any one of claims 8 to 16 wherein the detectable label is a fluorescent label, enzyme/substrate system, or radioactive label.

18. The assay according to claim 17 wherein the fluorescent label is a fluorophor selected from Alexa-[555] and Alexa-[647].

19. A kit for use in detecting changes in length of poly(A) tails of mRNA comprising: 5' phosphorylated oligo(dT) primer mix having a detectable label; oligo(dT)-anchor primer having a detectable label capable of binding to the 3' end of a poly(A) tail of mRNA; and a universal anchor sequence for affinity purification, immobilisation, and/or PCR verification; and

VN/oligo(dT)-anchor primer having a detectable label capable of binding to the 3' end of a poly(A) tail of a mRNA; and a universal anchor sequence for affinity purification, immobilisation, and/or PCR verification.

20. The kit according to claim 19 further comprising one or more of: reagents required to perform ligation reactions to join multiple primers annealed to a poly(A) segment; reagents required to perform reverse transcription reactions to generate test and reference cDNA libraries; and immobilised primers capable of binding to specific sequences in mRNA.