AU2011247142A1 - Ubiquitination assay - Google Patents

Description

UBIQUITINATION ASSAY

FIELD OF THE INVENTION

This invention is directed to assays to measure the activity of ubiquitination enzymes and the use of such assays in screening for agents that modulate this activity.

BACKGROUND OF THE INVENTION

Many diseases are caused by aberrant levels of proteins which cause cellular responses to be either heightened or dampened. Under normal conditions, proteins are targeted for proteasomal degradation by the ubiquitination pathway. Ubiquitination is the post-translational modification of a protein by the covalent attachment of one or more ubiquitin monomers.

The cascade leading to ubiquitination by a ubiquitination ligase (E3) is a complex enzymatic process with numerous proteins involved. There are three different catalytic events: ubiquitin activation; ubiquitin conjugation and ubiquitin ligation.

Ubiquitin is first activated in an ATP-dependent manner by a ubiquitin activating enzyme (El). The C-terminus of a ubiquitin forms a high energy thiolester bond with El. The ubiquitin is then passed to a ubiquitin conjugating enzyme (E2; also called ubiquitin carrier protein), also linked to this second enzyme via a thiolester bond. The ubiquitin is finally linked to its target protein to form a terminal isopeptide bond under the guidance of a ubiquitin ligase (E3). In this process, chains of ubiquitin are formed on the target protein, each covalently ligated to the next through the activity of E3.

El and E2 are structurally related and well characterized enzymes. There are several species of E2 (at least 25 in mammals), some of which act in preferred pairs with specific E3 enzymes to confer specificity for different target proteins.

E3 enzymes contain two separate activities: a ubiquitin ligase activity to conjugate ubiquitin to substrates and form polyubiquitin chains via isopeptide bonds, and a targeting activity to bring the ligase and substrate physically together. Substrate specificity of different E3 enzymes is the major determinant in the selectivity of the ubiquitin-dependent protein degradation process. As the ubiquitination cascade provides important targets for therapeutics, it is crucial to have an effective in vitro ubiquitination assay. In particular, it is important that such an assay can be used for high-throughput screening of potential ubiquitination cascade modulators.

Ubiquitination assays are well known in the art. In US 6,413,725, the level of ubiquitination of a particular substrate is measured by monitoring changes to the molecular weight of the substrate as a marker of ubiquitination activity. The assay is specific for the Cdc53 and Cdc-53 related human cullin E3 ubiquitin ligases and their limited substrates, e.g. Sicl. As the assay relies on changes in molecular weight of the substrate, direct measurement of ubiquitination over the time course of the reaction is difficult and the ubiquitinated substrates are preferably analysed after completion of the reaction, i.e. via SDS PAGE separation. This limitation makes the assay inefficient for high-throughput screening.

EP 1767649 describes a ubiquitination assay which specifically measures the auto-ubiquitination of the enzyme synoviolin. As this assay measures auto- ubiquitination, not only is the ubiquitin ligase limited to synoviolin, but the assay only measures the level of ubiquitination of a single product, the synoviolin itself.

The ubiquitination assay described in EP 1268847 measures E3 ligase activity by monitoring the amount of poly-ubiquitin bound to the ligase itself. The ligase and ubiquitin are labelled with a FRET pair. As the ligase is labelled with one member of the FRET pair, it is directly involved in detection of ubiquitination. Furthermore, the ubiquitination measured is ubiquitination of the ligase itself. Therefore, the assay is limited to use in detection of auto-ubiquitination of a single E3 ligase.

Although assays for ubiquitination already exist in the art, these assays are limited by their specificity to a particular ubiquitin ligase, or their specificity to a particular substrate. Furthermore, if the known assays are used to screen for potential ubiquitination modulators, they are not capable of distinguishing which particular event in the ubiquitination cascade is affected. There is therefore a need in the art for improved ubiquitination assays which overcome the disadvantages associated with the assays known in the art. DISCLOSURE OF THE INVENTION

The inventors have surprisingly found that, by specifically and differentially labelling the various with components of the ubiquitination cascade with different tags, the different stages of the ubiquitination cascade can be investigated directly from a single assay. Additionally it is possible to sequentially determine activity at each stage of the cascade without changing the assay composition.

Accordingly, the invention provides a method of assaying ubiquitination in a sample comprising:

a) combining in a sample under conditions suitable for ubiquitination to take place:

(i) ubiquitin;

and two or more of:

(ii) El;

(iii) E2;

(iv) E3; and

(v) a substrate protein;

b) exposing the sample to a labelled binding partner which is specific for the ubiquitin; and

c) measuring the amount of ubiquitin bound to any one of the components (ii) to (v) in the sample

wherein one or more of components (ii), (iii), (iv), or (v) in the sample comprises an immobilisation tag which facilitates its immobilisation onto a solid surface.

Preferably, all of components (ii), (iii), (iv) and (v) are combined in the sample. Preferably, more than one of components (ii), (iii), (iv) or (v) in the sample comprise an immobilisation tag, and each tag is different. Preferably, two, three or all of components (ii), (iii), (iv) or (v) in the sample are tagged and each immobilisation tag is different. Preferably, the immobilisation tag(s) present on one or more of components (ii), (iii), (iv) or (v) in the sample are selected from the group consisting of a his tag; GST tag; HA tag; c-Myc tag and FLAG tag.

In one embodiment component (ii) in the sample may comprise an immobilisation tag. In this embodiment, none, one, two or three of components (iii), (iv) and (v) may also comprise an immobilisation tag. In another embodiment component (iii) in the sample may comprise an immobilisation tag. In this embodiment, none, one, two or three of components (ii),

(iv) and (v) may also comprise an immobilisation tag.

In another embodiment component (iv) in the sample may comprise an immobilisation tag. In this embodiment, none, one, two or three of components (ii), (iii) and (iv) may also comprise an immobilisation tag.

In another embodiment component (v) in the sample may comprise an immobilisation tag. In this embodiment, none, one, two or three of components (ii), (iii) and (iv) may also comprise an immobilisation tag.

Preferably, one or more of components (ii), (iii), (iv) or (v) in the sample are immobilised on a solid surface. Preferably, only one of components (ii), (iii), (iv) or

(v) in the sample are immobilised on a solid surface.

In one embodiment component (ii) in the sample may be immobilised on a solid surface. In this embodiment, none, one, two or three of components (iii), (iv) and (v) may be immobilised on a solid surface.

In another embodiment component (iii) in the sample may be immobilised on a solid surface. In this embodiment, none, one, two or three of components (ii), (iv) and (v) may be immobilised on a solid surface.

In another embodiment component (iv) in the sample may be immobilised on a solid surface. In this embodiment, none, one, two or three of components (ii), (iii) and (v) may be immobilised on a solid surface.

In another embodiment component (v) in the sample may be immobilised on a solid surface. In this embodiment, none, one, two or three of components (ii), (iii) and (iv) may be immobilised on a solid surface.

Preferably the solid surface is an electrode. Preferably, the electrode comprises a binding member which is specific for one of the immobilisation tags.

Preferably, step (a) further comprises combining a potential modulator of ubiquitination.

Providing an effective in vitro ubiquitination assay has until now been difficult due to

E3 ligases being such a large and varied group of proteins, each member of which may be modulated differently by a variety of different proteins. The inventors have overcome these difficulties and provided an optimised method for assaying ubiquitination as described in detail below.

Step (a)

The methods of the invention are for assaying ubiquitination in a sample. Step (a) involves combining all of the components required for a ubiquitination reaction to occur in a sample. For the avoidance of doubt, this may include any different combination of components (ii). (iii), (iv) and (v) with component (i) under conditions suitable for ubiquitination to take place. For example components (ii) and (iii) may be combined with component (i); components (iii) and (iv) may be combined with component (i); or components (iv) and (v) may be combined with component (i). Preferably all of components (ii), (iii), (iv) and (v) are combined with component (i) in the sample.

Preferably the sample is in the form of a solution. The components of step (a) may be combined in any order. In a preferred embodiment, the ubiquitin is added to the sample last, in order to commence the ubiquitination reaction.

For the avoidance of doubt, a ubiquitination assay according to the invention includes both situations where ubiquitination is occurring and is suspected of occurring. Even if ubiquitination is not in fact occurring, the method is still one "for assaying ubiquitination". For example, the method may be performed on a negative control sample, for example, to confirm the fidelity of the assay. As a further example, if the assay is being used to select for a potential modulator of ubiquitination, the potential modulator may inhibit the process of ubiquitination.

Step (a) takes place in a first reaction container.

Suitable reaction containers for use in the invention can hold a volume of liquid and include, but are not limited to, test tubes, eppendorf tubes, wells in microtitre plates and mixing trays.

Step (a) of the methods of the invention is intended to be incubated for a finite period of time, for example, in the range of about 1 to 300 minutes, e.g. about 1, 2, 5, 10, 20, 40, 60, 120, 180, 240 or 300. An example of a preferred period is an incubation period between 10 and 60 minutes. Preferably the incubation period is 60 minutes. Step (a) may be incubated at any temperature in the range of about 0-99°C, e.g. at around 5, 10, 15, 20, 25, 30, 35, 37, 40, 45, 50, or 60°C, or even higher. Conveniently, step (i) may be incubated at between about 20°C and 40°C, for example, room temperature, 25, 30 or 37°C.

If the ubiquitination reaction is allowed to proceed for a finite period of time, then it may be stopped by, for example, the addition of EDTA, heating, freezing, acidification or alkination (i.e. pH shock), the addition of a known ubiquitination inhibitor, etc.

The sample may be any specimen in which the components of step (a) are combined together to allow a ubiquitination reaction to occur. The sample may contain other components in addition to the components described above. For example, the sample may comprise buffers, salts, metal ions, and other components required to facilitate ubiquitination. If the assay is being used to screen for potential modulators of ubiquitination, the sample will also include a potential modulator.

Ubiquitination assays are well known in the art, and the skilled person will be aware of the conditions which are suitable for allowing a ubiquitination reaction to proceed. For example, a ubiquitination reaction mixture will usually comprise buffers, salts, metal ions, ATP, mono-ubiquitin and the enzymes described herein which are required to facilitate ubiquitination, i.e. one or more of El, E2 and E3.

Step (b)

Step (b) of a method according to the invention involves exposing the sample to a labelled binding partner. This step can take place in the same reaction container or may involve the transfer of the sample to a second reaction container.

If step (b) takes place in the same reaction vessel, then the exposure of the sample to the labelled binding partner may be facilitated by the addition of a slide, rod or beads, coated with the labelled binding partner, to the first reaction vessel or may simply involve the addition of the labelled binding partner to the sample, e.g. as a solution.

Step (b) may occur after the sample has been incubated for a period of time, but may also occur at approximately the same time as the step (a). For example, the labelled binding partner may be in solution with the sample or may be added at a particular time point during the ubiquitination reaction. Therefore, the labelled binding partner may form part of the sample, or may be added to the sample at a later stage. Both of these embodiments are included when we refer to "exposing the sample".

If step (b) occurs after the ubiquitination reaction has been allowed to occur for a finite period of time, the reaction may have been stopped as described above and step (b) may occur minutes, hours, days, weeks or even months after the ubiquitination reaction has been stopped. The skilled reader will also appreciate that the components of step (a) may be mixed together and once the ubiquitination reaction has begun, samples may be taken continually throughout the reaction to produce a series of data over a certain timecourse.

If step (b) of the methods of the invention involves the transfer of the sample to a second reaction container then this transfer can be achieved by any method known in the art including, but not limited to, pipetting, pouring or transfer along microchannels between wells in a microfluidic chip.

Also envisaged by the invention is a first reaction container, separated from the second reaction container by a removable partition. The partition prevents the sample of step (a) gaining access to the labelled binding partner until step (b). In such a case, transfer of the sample to the second reaction container involves the removal of the partition. For example, the mixing container may be an eppendorf tube containing a wax plug separating the reagents used in step (a) and step (b) of the reaction. Transfer of the first mixing product between the first and second reaction container in this example may be achieved by increasing the temperature of the tube sufficiently to melt the wax plug.

Ubiquitination pathway

Ubiquitination of proteins in vivo is normally carried out by three enzymes: El, E2 and a ubiquitin ligase, often referred to in the art as an E3 ligase. The process occurs in three stages. Firstly activation of ubiquitin is undertaken by the ubiquitin activating enzyme El. A ubiquitin-adenylate is produced from ubiquitin and ATP, and is transferred to the active site of El, where AMP is released. The second step involves the transfer of ubiquitin from the active site of El to the active site of E2, a ubiquitin conjugating enzyme, via a trans(thio)esterification reaction. The final step of ubiquitination produces an isopeptide bond between a lysine residue of the target protein and the ubiquitin. This is achieved by an E3 ligase which recognises the target protein, forms interactions with both the ubiquitin bound E2 enzyme and the target protein, and finally forms the isopeptide bond between the lysine and the ubiquitin. Further ubiquitins are then added by the same mechanism to a lysine residue present on the ubiquitin attached to the target protein, and the process is repeated in order to form a chain of poly-Ub.

El

Ubiquitin-activating enzymes, also known as El enzymes, catalyze the first step in the ubiquitination cascade. Ubiquitin-activating enzyme (El) starts the ubiquitination process. The El enzyme along with ATP binds to the ubiquitin protein. The El enzyme then passes the ubiquitin protein to the ubiquitin carrier or conjugation protein (E2).

The skilled person will appreciate that the methods of the invention are suitable for use with any known El enzyme.

Preferably the methods of the invention use ubiquitin-like modifier activating enzyme 1 (UBEl), also known as UBA1, which is an enzyme encoded by the UBA1 gene in humans. UBEl catalyzes the first step in ubiquitin by first adenylating its C-terminal glycine residue with ATP, and thereafter linking this residue to the side chain of a cysteine residue in El, yielding an ubiquitin-El thioester and free AMP.

Preferably the UBEl used in the methods of the present invention is the human protein, which is a 1058 amino acid protein (Pubmed accession no.: NP 003325) having the sequence given in SEQ ID NO:l, or an active fragment or variant thereof.

Preferably the UBEl is present in the sample at a final concentration of 3-7nM, i.e. 3, 4, 5, 6 or 7nM. It is especially preferred that the UBEl is present in the sample at a final concentration of 5nM.

E2

While the nomenclature for E2, also known as ubiquitin-conjugating enzyme is not standardized across species, investigators in the field have addressed this issue and the skilled person can readily identify various E2 proteins, as well as species homologues (See Haas and Siepmann, FASEB J. 11 : 1257-1268 (1997)). These enzymes catalyse the second step in the ubiquitination reaction that targets a protein for degradation via the proteasome. Once activated by El, the activated ubiquitin is then transferred to an E2 cysteine. Once conjugated to ubiquitin, the E2 molecule binds one of several ubiquitin ligases or E3s via a structurally conserved binding region.

The skilled person will appreciate that the methods of the invention are suitable for use with any known E2 enzyme.

Preferably the E2 enzyme used in the methods of the present invention is UbcH3 or an active fragment or variant thereof, preferably human UbcH3, which is also known as CDC34. Human UbcH3 is a 236 amino acid protein (Pubmed accession no.: NP_004350.1) having the sequence given in SEQ ID NO:2.

Preferably the UbcH3 is present in the sample at a final concentration of 750-1250nM, i.e. 750, 775, 800, 825, 850, 875, 900, 925, 950, 975, 1000, 1025, 1050, 1075, 1100, 1125, 1150, 1175, 1200, 1225 or 1250nM. It is especially preferred that the UbcH3 is present in the sample at a final concentration of ΙμΜ.

E3

Some E3 ubiquitin ligases are known to have a single subunit responsible for the ligase activity. Such E3 ligases that have been characterized include the HECT (homologous to E6-AP carboxy terminus) domain proteins, represented by the mammalian E6AP-E6 complex which functions as a ubiquitin ligase for the tumor suppressor p53 and which is activated by papillomavirus in cervical cancer (Huang et al, Science 286 : 1321-26 (1999)).

Single subunit ubiquitin ligases having a RING domain include Mdm2, which has also been shown to act as a ubiquitin ligase for p53, as well as Mdm2 itself. Other RLNG domain, single subunit E3 ligases include: TRAF6, involved in IKK activation; Cbl, which targets insulin and EGF ; Sina/Siah, which targets DCC ; Itchy, which is involved in haematopoesis (B, T and mast cells) ; and LAP, involved with inhibitors of apoptosis.

The best characterized E3 ligase is the APC (anaphase promoting complex), which is a multi-subunit complex that is required for both entry into anaphase as well as exit from mitosis (King et al., Science 274 : 1652-59 (1996) for review).

In eukaryotes, a family of complexes with E3 ligase activity play an important role in regulating Gl progression. These complexes, called SCF's, consist of at least three subunits, SKP 1, Cullins (having at least seven family members) and an F-box protein (of which hundreds of species are known) which bind directly to and recruit the substrate to the E3 complex.

The skilled person will appreciate that the methods of the invention are suitable for use with any known E3 ligase, which in turn can comprise any E3 ligase component.

Preferably the methods of the present invention rely on an E3 ligase comprising the following components:

S-phase kinase-associated protein 2 (Skp2)

Skp2 is an enzyme that in humans is encoded by the SKP2 gene. This gene encodes a member of the F-box protein family which is characterized by an approximately 40 amino acids motif, the F-box. The F-box proteins constitute one of the four subunits of a ubiquitin protein ligase complex called SCF, which functions in phosphorylation- dependent ubiquitination. The F-box proteins are divided into 3 classes: Fbws containing WD-40 domains, Fbls containing leucine-rich repeats, and Fbxs containing either different protein-protein interaction modules or no recognizable motifs (Kipreos ET, Pagano M. (2000). The F-box protein family. Genome Biol l(5):Reviews3002).

The protein encoded by the Skp2 gene belongs to the Fbls class; in addition to an F- box, this protein contains 10 tandem leucine-rich repeats. It specifically recognizes phosphorylated cyclin-dependent kinase inhibitor IB (CDKN1B, also referred to as p27 or KIPl) predominantly in S phase and interacts with S-phase kinase-associated protein 1 (SKP1 or pi 9). Alternative splicing of this gene generates 2 transcript variants encoding different isoforms.

It is preferred that a method according to the present invention utilizes the human isoform-1 variant of Skp2 which is a 424 amino acid protein (Pubmed accession no.: AAK31593), or an active fragment or variant thereof. The wild type sequence is given in SEQ ID NO:3.

Preferably the Skp2 protein is present in the sample at a final concentration of between about 20-30nM, i.e. about 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. It is particularly preferred that the Skp2 is present in the sample at a final concentration of 25nM. Skpl

S-phase kinase-associated protein 1 (SKP1) is an enzyme that in humans is encoded by the Skpl gene. SKP1 is a member of the SCF ubiquitin ligase protein complex. It binds to proteins containing an F-box motif, such as cyclin F, Skp2, and other regulatory proteins involved in ubiquitination. Alternative splicing of this gene results in two transcript variants: "isoform-a" and "isoform-b".

It is preferred that a method according to the invention utilizes the human isoform-a variant of Skpl which is a 160 amino acid protein (Pubmed accession no.: NP_008861), or an active fragment or variant thereof. The wild type sequence is given in SEQ ID NO:4.

Preferably the Skpl is present in the sample at a final concentration of 20-30nM, i.e. about 21, 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. It is particularly preferred that the Skpl is present in the sample at a final concentration of 25nM.

Cullin 1 (Cull)

Cull is a human protein and gene from the cullin family. Cull plays an important role in protein degradation and protein ubiquitination and is an essential component of the SCF (SKP1-CUL1 -F-box protein) E3 ubiquitin ligase complex. In the SCF complex, Cull serves as a rigid scaffold that organizes the SKP1 -F-box protein and RBX1 subunits. The protein is usually neddylated, which enhances the ubiquitination activity of SCF.

Preferably the methods of the invention utilize human Cull, which is a 776 amino acid protein (Pubmed accession no.: NP 003583), or a fragment or variant thereof. The wild type protein has the sequence given in SEQ ID NO:5.

Preferably the Cull is present in the sample at a final concentration of 20-30nM, i.e. about 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. It is particularly preferred that the Cull is present in the sample at a final concentration of about 25nM.

RING-box protein l(Rbxl)

Rbxl is a protein that in humans is encoded by the RBX1 gene. Rbxl is an evolutionarily-conserved protein which interacts with cullins. The protein plays a unique role in the ubiquitination reaction by heterodimerizing with Cull to catalyze ubiquitin polymerization. Preferably the methods of the invention utilize human Rbxl, which is a 108 amino acid protein (Pubmed accession no.: CAG30446), or an actice fragment or variant thereof. The wild type protein has the sequence given in SEQ ID NO:6.

Preferably the Rb l is present in the sample at a final concentration of 20-30nM, i.e. about 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. It is particularly preferred that the Rbxl is present in the sample at a final concentration of 25nM.

Cksl

Cksl is a member of the highly conserved Sucl/Cks family of cell cycle regulatory proteins and is also known as CDC28. All proteins of this family have Cdk-binding and anion-binding sites, but only mammalian Cksl binds to Skp2 and promotes the association of Skp2 with p27 phosphorylated on Thr-187.

Preferably the methods of the invention utilize human Cksl, which is a 79 amino acid protein (Pubmed accession no.: NP_001817) or a fragment or variant thereof. The wild type protein has the sequence given in SEQ ID NO:7.

Preferably the Cksl is present in the sample at a final concentration of 20-30nM, i.e. about 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. It is particularly preferred that the Cksl is present in the sample at a final concentration of 25nM.

Cyclin-dependent kinase 2 (CDK2)

CDK2 is a human gene which encodes a protein which is a member of the cyclin-dependent kinase family of Ser/Thr protein kinases. This protein kinase is highly similar to the gene products of S. cerevisiae cdc28, and S. pombe cdc2. Two alternatively spliced variants and multiple transcription initiation sites of this gene have been reported.

Preferably the methods of the invention utilize human CDK2, which is a 298 amino acid protein (Pubmed accession no.: CAA43985) or an active fragment or variant thereof. The wild type protein has the sequence given in SEQ ID NO:8.

Preferably the CDK2 is present in the sample at a final concentration of 20-30nM, i.e. about 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. It is particularly preferred that the CDK2 is present in the sample at a final concentration of 25nM. Cvclin El

Cyclin-El is a protein that in humans is encoded by the CCNE1 gene. Cyclin El belongs to the highly conserved cyclin family, whose members are characterized by a dramatic periodicity in protein abundance through the cell cycle. Cyclins function as regulators of CDK kinases. Different cyclins exhibit distinct expression and degradation patterns which contribute to the temporal coordination of each mitotic event. This cyclin forms a complex with and functions as a regulatory subunit of CDK2, whose activity is required for cell cycle Gl/S transition.

This protein accumulates at the Gl-S phase boundary and is degraded as cells progress through S phase. Overexpression of this gene has been observed in many tumors, which results in chromosome instability, and thus may contribute to tumorigenesis. Two alternatively spliced transcript variants of this gene, which encode distinct isoforms, have been described.

Preferably the methods of the invention utilize human Cyclin El, which is a 395 amino acid protein (Pubmed accession no.: AAM54043), or an active fragment or variant thereof. The wild type protein has the sequence given in SEQ ID NO:9.

Preferably the Cyclin El is present in the sample at a final concentration of 20-3 OnM, i.e. about 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. It is particularly preferred that the cyclin El is present in the sample at a final concentration of 25nM.

Substrate

The methods of the invention provide a ubiquitination assay which is suitable for use with a range of different substrates. The skilled person will appreciate that the ubiquitination cascade is responsible for the ubiquitination of many cellular proteins. Accordingly, all of these proteins are suitable for use as a substrate in the methods of the invention.

It will be appreciated that different F-box proteins which form part of the E3 ligase are primarily responsible for the specific binding of the different substrates. Therefore, the F-box protein which forms part of the E3 ligase used in the methods of the invention may be selected based on the choice of substrate. Examples of such substrates will be known to the skilled person and, in particular, are described in Deshaies & Joazeiro (Annu Rev Biochem. 2009;78:399-434). Preferably the substrate is p27.

p27

p27, also known as cyclin-dependent kinase inhibitor IB is an enzyme that in humans is encoded by the CDKN1B gene. p27 belongs to the Cip/Kip family of cyclin dependent kinase (Cdk) inhibitor proteins. The encoded protein binds to and prevents the activation of cyclin E-CDK2 or cyclin D-CDK4 complexes, and thus controls the cell cycle progression at Gl. It is often referred to as a cell cycle inhibitor protein because its major function is to stop or slow down the cell division cycle.

The p27Kipl gene has a DNA sequence similar to other members of the "Cip/Kip" family which include the p21Cipl/Wafl and p57Kip2 genes. In addition to this structural similarity the "Cip/Kip" proteins share the functional characteristic of being able to bind several different classes of Cyclin and Cdk molecules. For example, p27Kipl binds to cyclin D either alone, or when complexed to its catalytic subunit CDK4. Likewise, p27Kipl is able to bind other Cdk proteins when complexed to cyclin subunits such as Cyclin E/Cdk2 and Cyclin A/Cdk2.

Preferably the methods of the invention utilize human p27, which is a 199 amino acid protein (Pubmed accession no.: BAA25263) or an active fragment or variant thereof. The wild type protein has the sequence given in SEQ ID NO: 10. Preferably the p27 is phosphorylated at position T187.

Preferably the p27 is present in the sample at a final concentration of 20-30nM, i.e. about 20, 21 , 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. It is particularly preferred that the p27 is present in the sample at a final concentration of 25nM.

Preferred combinations of components

Preferably the Skp2 isoform 1, Skpl, Cull and Rbxl are combined with each other prior to combination with the other components of (a). In this embodiment these four components are thought to form a tetramer. Preferably the Skp2 isoform 1, Skpl, Cull and Rbxl tetramer are present in the sample at a final concentration of 20-30nM, i.e. about 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. It is particularly preferred that the Skp2 isoform 1, Skpl, Cull and Rbxl tetramer is present in the sample at a final concentration of 25nM. Preferably the p27, CDK2 and cyclin El are combined with each other prior to combination with the other components of (a). Preferably the p27, CDK2 and cyclin El are co-expressed in the same host cell and purified as a complex, which is then used in the methods of the invention. If the p27, CDK2 and cyclin El are combined together prior to the combination with the other components, or are co-expressed, then preferably the complex should be present in the sample at a final concentration of 20-30nM, i.e. about 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or about 30nM. Preferably the p27, CDK2 and cyclin El complex are present in the sample at a final concentration of 25nM.

Variants

The preferred El, E2, E3 components and substrate for use in the methods of the invention are described above. However, the skilled person will appreciate that these components can be substituted with any known variant of these components, including active fragments, naturally-occurring alleles, modified variants or variants from species other than Homo sapiens. These variants are all suitable for use in the methods of the invention, provided that they retain activity. By "retain activity" is meant that the normal biological activity of the wild type protein, as a component of the ubiquitination machinery, is retained to such a degree to enable the assay to function at a level adequate for the researcher's requirements e.g. at a level 30%, 40%, 50%, 60%, 70%, 80%, 90% or more of wild type activity.

A variant of the various components may comprise substitutions, additions or deletions in its amino acid sequence compared to SEQ ID NOs: 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10. In a preferred embodiment, the variant comprises an amino acid sequence with greater than 75% sequence identity, i.e. 80%, 85%, 90% or 95% or more to the amino acid sequence of one of SEQ ID NOs: 1, 2, 3, 4, 5, 6, 7, 8, 9 or 10.

A variant may also represent a fragment of the wild type full length protein, truncated, for example, at one or both the N and/or C termini. A truncation may be to a length 50% of the full length sequence, or 60%, 70%, 80%, or 90% or more of the full length sequence, provided that the biological activity of the wild type protein is retained to such a degree to enable the assay to function at a level adequate for the researcher's requirements e.g. at a level 30%, 40%, 50%, 60%, 70%, 80%, 90% or more of wild type activity. A variant may also form part of a larger protein complex. For example, a variant may represent one domain of a multi-domain protein, or may further comprise additional features to facilitate purification, detection and stability of the expressed protein. For example, the variant may comprise a polyarginine tag (Arg-tag), polyhistidine tag (His-tag), FLAG-tag, Strep-tag, c-myc-tag, S-tag, calmodulin-binding peptide, cellulose-binding domain, SBP-tag, chitin-binding domain, glutathione S-transferase- tag (GST), maltose-binding protein, transcription termination anti-termination factor (NusA), E. coli thioredoxin (TrxA) or protein disulfide isomerase I (DsbA).

Ubiquitin (lib)

Ubiquitin is a highly conserved regulatory protein that is present in all eukaryotic cells.

The ubiquitin used in the methods of the invention may comprise any known variant of ubiquitin including naturally occurring alleles or modified variants. As ubiquitin is highly conserved, the ubiquitin used in the present invention may be ubiquitin from any species, for example, human ubiquitin which comprises the amino acid sequence given in SEQ ID NO: 11, mouse ubiquitin which comprises the amino acid sequence given in SEQ ID NO: 12, Arabidopsis ubiquitin which comprises the amino acid sequence given in SEQ ID NO: 13, yeast ubiquitin which comprises the amino acid sequence given in SEQ ID NO: 14, Drosophila ubiquitin which comprises the amino acid sequence given in SEQ ID NO: 15, Dictyostelium ubiquitin which comprises the amino acid sequence given in SEQ ID NO: 16 or bovine ubiquitin which comprises the amino acid sequence given in SEQ ID NO: 17. Preferably, the ubiquitin used in the present invention is human ubiquitin, which comprises the amino acid sequence given in SEQ ID NO: 11, or an active variant or fragment thereof.

A variant of ubiquitin may comprise substitutions, additions or deletions in its amino acid sequence compared to SEQ ID NOs: 11, 12, 13, 14, 15, 16 or 17. In a preferred embodiment, ubiquitin comprises an amino acid sequence with greater than 75% sequence identity, i.e. 80%, 85%, 90% or 95% or more to the amino acid sequence of one of SEQ ID NOs: 11, 12, 13, 14, 15, 16 or 17. The ubiquitin of the invention may comprise a tag. Where the ubiquitin comprises a tag it is referred to as "tagged ubiquitin" and the tag is referred to as "the ubiquitin tag". Preferably the ubiquitin is present in the sample at a final concentration of between 1 -3μΜ, i.e. about 1, 1.25, 1.5, 1.75, 2, 2.25, 2.5, 2.75 or about 3μΜ. It is particularly preferred that the ubiquitin is present in the sample at a final concentration of 2μΜ.

Preferably the ubiquitin comprises a tag.

In one embodiment the ubiquitin or ubiquitin tag is recognized by the labelled binding partner. The skilled person will appreciate that, if present, the choice of tag is dependent on the choice of labelled binding partner. For example, if the labelled binding partner is an antibody then the tag may be a protein which is recognizable by the labelled binding partner antibody or vice versa. If the labelled binding partner is not an antibody, then the ubiquitin tag may be a protein with a high affinity for it. In a preferred embodiment, the ubiquitin tag is biotin or a derivative thereof and the labelled binding partner is streptavidin or a derivative thereof.

In another embodiment, the ubiquitin tag is capable of interacting with the labelled binding partner to produce a detectable signal. Preferably in this embodiment the ubiquitin tag and the labelled binding partner form a FRET (fluorescence resonance energy transfer) pair. In this case the ubiquitin tag will be either a donor or acceptor dye, whilst the labelled binding partner will be labelled with the other. Only when the two are brought into close proximity, through the respective binding of the ubiquitin and labelled binding partner, will photon emission be observable.

Labels used to form a FRET pair can be any molecule that may be detected via its inherent fluorescent properties. Examples of fluorescent labels include, but are not limited to fluorescein isothiocyanate, fluorescein/rhodamine, coumarin/fluorescein, Alexa fluors, IAEDANS, BODIPY FL, europium cryptate, Cy dyes, XL665 and Oregon green. In a preferred embodiment the two labels which constitute the FRET pair are europium cryptate and XL665. In addition to FRET partners. FRET/Quench methods may be used in the invention. These methods rely on a donor dye and another molecule which accepts energy but does not emit, effectively quenching the donor dye, for example iron quench.

Poly-ubiquitin (poly-Ub)

It will be appreciated that poly-Ub comprises at least two ubiquitin monomers covalently attached to one another. Poly-Ub of the present invention is a polymer formed from ubiquitin as described above. Poly-Ub comprises two or more labelled ubiquitins covalently attached to one another e.g. 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 or 15 or more. Poly-Ub may be free poly-Ub chains, poly-Ub bound to a substrate protein, or poly-Ub bound to the ubiquitin ligase.

Adenosine-5 '-triphosphate (ATP)

ATP is a multifunctional nucleotide used in cells as a coenzyme. ATP transports chemical energy within cells for metabolism. It is produced by photophosphorylation and cellular respiration and used by enzymes and structural proteins in many cellular processes, including biosynthetic reactions, motility, and cell division. One molecule of ATP contains three phosphate groups, and it is produced by ATP synthase from inorganic phosphate and adenosine diphosphate (ADP) or adenosine monophosphate (AMP). Guanidine-5'-triphosphate (GTP), Cytidine-5 '-triphosphate (CTP), Uridine-5'-triphosphate (UTP) or Thymidine5'-triphosphate (TTP) may also be used in the methods of the invention.

Preferably, the methods of the invention use ATP.

Preferably the ATP is present in the sample at a final concentration of 75-125μΜ, i.e. about 75, 80, 85, 90, 95, 100, 105, 110, 115, 120 or about 125μΜ. Preferably, the ATP is present in the sample at a final concentration of ΙΟΟμΜ.

Labelled binding partner

The labelled binding partner used in the methods of the invention interacts with the ubiquitin either directly or via the ubiquitin tag. The labelled binding partner is specific for the ubiquitin or the tagged ubiquitin. As such, the labelled binding partner may bind to the ubiquitin or to the ubiquitin tag and both possibilities are used interchangeably herein.

Preferably, the labelled binding partner has a higher affinity for ubiquitin than for any of the other components used in the methods of the invention, in the sense that it binds preferentially to ubiquitin.

Preferably the labelled binding partner has a binding affinity (Kd) of less than 10^"6M, i.e. 10^"7M, 10^"8M, 10^"9M, 10^"10M, 10^"nM, 10^"l2M or even 10^"13M or less towards ubiquitin. Preferably, the labelled binding partner has a binding affinity (Kd) of less than 10^"9M, i.e., 10^"10M, 10^"nM, 10^"I2M or 10^"13M towards ubiquitin. The labelled binding partner may comprise one or more molecules which have an affinity for ubiquitin.

The labelled binding partner may comprise an antibody. Antibodies may be of any isotype (e.g. IgA, IgG, IgM i.e. an α, γ or μ heavy chain). Antibodies may have a κ or a λ light chain. Within the IgG isotype, antibodies may be IgGl , IgG2, IgG3 or IgG4 subclass. The term "antibody" includes any suitable natural or artificial immunoglobulin or derivative thereof. In general, the antibody will comprise a Fv region which possesses specific antigen binding activity. This includes, but is not limited to: whole immunoglobulins, antigen binding immunoglobulin fragments (e.g. Fv, Fab, F(ab')2 etc.), single chain antibodies (e.g. scFv), oligobodies, chimeric antibodies, humanized antibodies, veneered antibodies, etc.

Antibodies used as the labelled binding partner of the invention may be polyclonal or monoclonal.

The labelled binding partner may in an alternative embodiment, comprise an aptamer. By 'aptamer' we mean a nucleic acid sequence which, owing to said nucleic acid sequence and the environment in which this sequence is located, forms a three- dimensional structure. The three-dimensional structure confers, in part, binding properties to aptamers which have specificity to a target molecule or molecules. In the present invention, the aptamer shows binding specificity towards ubiquitin molecule.

Aptamers are typically oligonucleotides which may be single stranded oligodeoxynucleotides, oligoribonucleotides, or modified oligodeoxyribonucleotides or modified oligoribonucleotides. A screening method to identify aptamers is described in US 5270163. The term 'modified' in the context of aptamers refers to nucleotides with a covalently modified base and/or sugar or sugar derivative.

In another alternative embodiment the labelled binding partner may comprise a ubiquitin binding domain (UBD). UBDs are modular protein domains that are able to bind non-covalently to ubiquitin. The natural function of UBDs is to transmit information conferred by protein ubiquitination in order to control various cellular events. UBDs usually comprise 20-150 amino acids and can be found as part of ubiquitination enzymes, de-ubiquitination enzymes or ubiquitin receptors. Around 10 different motifs have so far been characterised as ubiquitin binding domains (Hicke et al., 2005), the structures of which are highly varied. All UBDs however contact the same face of ubiquitin, which comprises the amino acid residue isoleucine 44. The different motifs of UBDs have different affinities for mono-Ub and poly-Ub depending on their function in vivo, and some of the UBDs also distinguish between poly-Ub with linkage at lysine 63 and that with linkage at lysine 48.

UBDs are present in all proteins that are capable of binding ubiquitin and therefore all proteins that are involved in the ubiquitination reaction or the recognition of ubiquitinated proteins contain a UBD within them.

Within this embodiment the UBD may be a particular class of UBD which is able to bind to poly-Ub preferentially over mono-Ub. These are known as ubiquitin associated domains (UBAs). UBAs are able to distinguish between poly-Ub and mono-Ub because they have a higher binding affinity for poly-Ub compared to mono- Ub. Preferably, the UBAs form three-helix bundles, with two of these three helices packing against ubiquitin on binding.

The preferential binding affinity for poly-Ub over mono-Ub may be defined by the UBAs' respective dissociation constants (¾) for the two species. Binding affinity of UBAs may typically be between 10 and 1000 times greater for poly-Ub compared to mono-Ub, i.e. 10, 20, 30, 40, 50, 100, 200, 300, 400, 500, 600, 700, 800, 900, 1000 times or more greater for poly-Ub than for mono-Ub. Preferably, binding affinity of a UBA used in the present invention is about 50 times greater for poly-Ub than for mono-Ub. UBAs will typically have a ¾ of about 0.03-9μΜ, i.e. about 0.03, 0.04, 0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1, 2, 3, 4, 5, 6, 7, 8 or 9μΜ for poly-Ub, and a ¾ of about 10-500μΜ, i.e. about 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 150, 200, 250, 300, 350, 400, 450 or 500μΜ for mono-Ub.

Preferred UBAs include, but are not limited to, Rad23-UBA1 (Accession NO.AAB28441, amino acids 146-186), Rad23-UBA2 (Accession No.AAB28441, amino acids 355-395), Dsk2-UBA (Accession No. NP_014003, amino acids 327- 371), Ubpl4-UBA2 (Accession No. CAA85001, amino acids 649-689), Ubcl-UBA (Accession no. CAA39812) and RpnlO-UIM (Accession no. EEU06553, amino acids 223-242). In one preferred embodiment the UBA is derived from the Rad23 protein. In a further preferred embodiment, the UBA comprises protein domain UBA2 of the Rad23 protein and comprises or consists of the amino acid sequence of SEQ ID NO: 18. The UBA used in the methods of the invention may comprise a variant of SEQ ID NO: 18. A variant of the UBA may comprise substitutions, additions or deletions in its amino acid sequence compared to SEQ ID NO: 18. The UBA may comprise an amino acid sequence with a sequence identity greater than 75%, i.e. 80%, 85%, 90% or 95% or more, to the amino acid sequence of SEQ ID NO: 18. Variants of the UBA are included for use in the method of the invention provided that they maintain their ability to be able to bind to ubiquitin and preferably to differentiate between poly-Ub and mono-Ub, according to the definitions provided above.

UBAs may also form part of a larger protein complex. For example, they may represent one domain of a multi-domain protein, or may further comprise additional features to facilitate purification, detection and stability of the expressed protein, for example, a polyarginine tag (Arg-tag), polyhistidine tag (His-tag), FLAG-tag, Strep- tag, c-myc-tag, S-tag, calmodulin-binding peptide, cellulose-binding domain, SBP- tag„ chitin-binding domain, glutathione S-transferase-tag (GST), maltose-binding protein, transcription termination anti-termination factor (NusA), E. coli thioredoxin (TrxA) and protein disulfide isomerase I (DsbA).

UBAs may be provided in the form of multimeric complexes comprising two or more UBAs or in the form of individual entities. The multimeric complex may be dimeric, trimeric, tetrameric, pentameric or hexameric. The UBAs present within the multimeric complex may be the same or different. The multimeric complex may be Tandem Ubiquitin Binding Entities (TUBES; tetrameric) or Tandem Dimeric UBAs (dimeric). A tandem dimeric UBA may comprise one or more of Dsk2 and Rad23, and in one embodiment may comprise or consist of Dsk2 and Rad23.

The methods of the invention may also use one or more fragments of UBAs provided that the fragment maintains the functional ability to be able to bind to ubiquitin and preferably to distinguish between poly-Ub and mono-Ub according to the definitions provided above.

In one embodiment one or more UBAs may be used to measure the level of polyubiquitin bound to any one of the components of the assay in step (c) as described in the PCT application entitled "Ubiquitination Assay" which claims priority from GB 1006604.1 and has the same Applicant as the present application.

In one embodiment, where the labelled binding partner recognises the ubiquitin or the ubiquitin tag the label present on the labelled binding partner may be measured directly or indirectly. Preferred labels include, but are not limited to, fluorescent labels, labelled enzymes and radioisotopes.

In a more preferred embodiment, the labelled binding partner is capable of emitting radiation when electrochemically stimulated. Preferably the radiation is in the visible spectrum. Preferably the label is selected from the group consisting of 3- aminophthalydrazide; a ruthenium chelate and an acridan ester. Preferably the label is tris(2,2'-bipyridyl) ruthenium (II).

As discussed above in relation to the ubiquitin tag, the labelled binding partner is capable of interacting with the ubiquitin tag to produce a detectable signal. Preferably in this embodiment the ubiquitin tag and the labelled binding partner form a FRET (fluorescence resonance energy transfer) pair. In this case the labelled binding partner tag will be with either a donor or acceptor dye, whilst the ubiquitin tag will be labelled with the other. Only when the two are brought into close proximity, through the respective binding of the ubiquitin and labelled binding partner, will photon emission be observable.

Step (c)

In one embodiment, step (c) may be performed two or more times in order to measure the amount of labelled ubiquitin bound to two or more of components (ii) to (v) in the sample. For example, step (c) may be performed twice in order to measure the amount of labelled ubiquitin bound to two of components (ii) to (v) in the sample, three times in order to measure the amount of labelled ubiquitin bound to three of components (ii) to (v) in the sample or four times in order to measure the amount of labelled ubiquitin bound to all of components (ii) to (v) in the sample. Where step (c) is performed less than four times, the amount of labelled ubiquitin bound to any combination of components may be measured. In all embodiments the amount of labelled ubiquitin bound to the components may be measured simultaneously or sequentially in any order. Preferably the component(s) which is to be measured comprise an immobilisation tag. If the amount of labelled ubiquitin bound to each relevant component is to be measured simultaneously, each immobilisation tag should be different to allow each of the relevant component to be separated and the level of ubiquitination measured.

In one embodiment, the composition of the assay is not altered between measurements of the ubiquitination of individual components.

The repetition of step (c) allows a single assay to be used to directly measure different stages of the ubiquitination cascade or to sequentially determine activity at each stage of the cascade without changing the assay composition.

Assay formats

The methods of the invention are suitable for use in a number of different assay formats. For example, the method may be performed in a format wherein all components are present in a homogeneous phase, or alternatively the method may be performed in a format where one or more of the components is immobilised.

Homogeneous phase assay

As described above, step (a) involves combining all of the components required for a ubiquitination reaction to occur in a sample. In a preferred embodiment, the samples are in a homogeneous phase, that is, one where at least two reactants or even all reaction components are in the fluid (solution) phase. This assay format lends itself to application in high throughput screening of ubiquitination reactions.

This embodiment of the invention preferably exploits a labelled binding partner and ubiquitin tag which are capable of interacting to produce a detectable signal, e.g. a FRET pair. As discussed above, only upon binding of these two components will a detectable signal be emitted.

Immobilised assay

In another embodiment of the invention, the assay may be performed in a format wherein one of the components of the sample is immobilised on a solid surface. Preferably the component to be immobilised comprises an immobilisation tag which facilitates its immobilisation onto the surface. Preferably the solid surface comprises a binding member which is specific for the immobilisation tag. Preferably, it is the substrate which comprises an immobilisation tag.

This embodiment of the invention preferably comprises an additional wash step, (V), which occurs between steps (b) and (c) and removes unbound components from the solid surface.

The immobilisation of a component to the solid surface allows for the direct detection of the labelled binding partner. For example, immobilising the substrate will also immobilise any ubiquitin bound to the substrate and in turn any labelled binding partner bound to the ubiquitin. The amount of labelled binding partner present is proportional to the amount of ubiquitin bound to the substrate. If the label present on the labelled binding partner is, for example, a dye, this will be visible.

In a preferred embodiment, the solid surface is an electrode. The labelled binding partner comprising a label which is capable of emitting radiation, preferably in the visible spectrum, upon stimulation is particularly suitable for use with this embodiment of the invention. If ubiquitination of the substrate has occurred, its immobilisation onto the electrode will bring bound ubiquitin, and in turn labelled binding partner into close proximity with the electrode. Passing a current through the electrode will cause stimulation of the labelled binding partner, in turn causing a signal to be emitted (See Figure 51). The skilled person will appreciate that this embodiment of the invention is based on the principals of electrochemiluminescence.

Electrochemiluminescence

Electrochemiluminescence, or electrogenerated chemiluminescence (ECL) is a kind of luminescence produced during electrochemical reactions in solutions. In electrogenerated chemiluminescence, electrochemically generated intermediates undergo a highly exergonic reaction to produce an electronically excited state that then emits light. ECL excitation is caused by energetic electron transfer (redox) reactions of electrogenerated species. Such luminescence excitation is a form of chemiluminescence where one/all reactants are produced electrochemically on the electrodes.

ECL is usually observed during application of potential (several volts) to electrodes of electrochemical cell that contains solution of luminescent species (polycyclic aromatic hydrocarbons, metal complexes) in aprotic organic solvent (ECL composition). ECL combines analytical advantages of chemiluminescent analysis (absence of background optical signal) with ease of reaction control by applying electrode potential. Enhanced selectivity of ECL analysis is reached by variation of electrode potential thus controlling species that are oxidized/reduced at the electrode and take part in the ECL reaction.

ECL generally uses ruthenium complexes, especially [Ru (Bpy)3]²⁺ (which releases a photon at ~620 nm) regenerating with TPA (Tripropylamine) in a liquid phase or at a liquid-solid interface. It can be used as monolayer immobilized on an electrode surface (made for example of nafion, or special thin films made by the Langmuir- Blogett technique or the self-assembly technique) or as a co-reactant or more commonly as a tag and used in HPLC, Ru tagged antibody based immunoassays, Ru Tagged DNA probes for PCR etc, NADH or H₂0₂ generation-based biosensors, oxalate and organic amine detection and many other applications and can be detected from picomolar sensitivity to a dynamic range of more than six orders of magnitude. Photon detection is done with PMT or silicon photodiode or gold coated fibre-optic sensors. These modifications are included within the scope of the invention.

Solid surface

Where the methods of the invention require immobilisation onto a solid surface, the surface may be any surface or support upon which it is possible to immobilise one of the components of step (a) or the labelled binding partner, including one or more of a solid support (e.g., glass such as a glass slide or a coated plate, silica, plastic or derivatized plastic, paramagnetic or non-magnetic metal), a semi-solid support (e.g., a polymeric material, a gel, agarose, or other matrix), and/or a porous support (e.g., a filter, a nylon or nitrocellulose membrane or other membrane). In some embodiments, synthetic polymers can provide a solid surface, including, e.g., polystyrene, polypropylene, polyglycidylmethacrylate, aminated or carboxylated polystyrenes, polyacrylamides, polyamides, polyvinylchlorides, and the like. In preferred embodiments, the solid surface comprises a microtitre immunoassay plate or other surface suitable for use in an ELISA.

The surface of the substrate or support may be planar, curved, spherical, rod-like, pointed, wafer or wafer-like, or any suitable two-dimensional or three-dimensional shape on which the second binding partner may be immobilised, including, e.g., films, beads or microbeads, tubes or microtubes, wells or microtitre plate wells, microfibers, capillaries, a tissue culture dish, magnetic particles, pegs, pins, pin heads, strips, chips prepared by photolithography, etc. In some embodiments, the surface is UV- analyzable, e.g., UV -transparent.

Immobilisation tag/Binding member

The immobilisation tag is intended to be bound by the binding member. As such, the interaction between the two is a specific binding mechanism.

Immobilisation may be achieved in any number of ways, known in the art, described herein, and/or as can be developed. For example, immobilisation may involve any technique resulting in direct and/or indirect association of the immobilisation tag with the binding member, including any means that at least temporarily prevents or hinders release of the immobilisation tag (and the attached component) into a surrounding solution or other medium. The means can be by covalent bonding, non-covalent bonding, ionic bonding, electrostatic interactions, Hydrogen bonding, van der Waals forces, hydrophobic bonding, or a combination thereof. For example, immobilisation can be mediated by chemical reaction where the immobilisation tag contains an active chemical group that forms a covalent bond with the binding member. For example, an aldehyde-modified support surface can react with amino groups in protein receptors; or amino-based support surface can react with oxidization-activated carbohydrate moieties in glycoprotein receptors; a support surface containing hydroxyl groups can react with bifunctional chemical reagents, such as N,N dissuccinimidyl carbonate (DSC), or N-hydroxysuccinimidyl chloroformate, to activate the hydroxyl groups and react with amino-containing receptors. In some embodiments, support surface of the substrate may comprise animated or carboxylated polystyrenes; polyacrlyamides; polyamines; polyvinylchlorides, and the like. In still some embodiments, immobilization may utilize one or more binding-pairs to bind or otherwise attach the immobilisation tag to the binding member, including, but not limited to, an antigen- antibody binding pair, hapten/anti -hapten systems, a avidin-biotin binding pair; a streptavidin-biotin binding pair, a folic acid/folate binding pair; photoactivated coupling molecules, and/or double stranded oligonucleotides that selectively bind to proteins, e.g., transcriptional factors.

The skilled person will appreciate that a number of the immobilisation options described above are also suitable for immobilising any one of the components of step

(a) without the need for an immobilisation tag. The binding member may be immobilised on the surface of a reaction container or to beads which are added to the first reaction container (e.g. superparamagnetic polystyrene beads - Dynabeads M-450). Once binding of the immobilisation tag to the binding member is achieved, unbound reactants can be washed from the immobilised component. Detection of the component is then possible via the labelled binding partner.

Preferably, the binding member has a higher affinity for the immobilisation tag than for any of the other components used in the methods of the invention, in the sense that it binds preferentially to the immobilisation tag. The skilled person will appreciate that the same is true vice versa, i.e. the immobilisation tag preferentially binds to the binding member.

Preferably the binding member has a binding affinity (Kd) of less than 10^"6M, i.e. 10^{" 7}M, 10^"8M, 10^" M, 10^'10M, 10^"nM, 10^"12Μ or even 1(T¹³M or less towards the immobilisation tag. Preferably, the binding member has a binding affinity (Kd) of less than 10^"9M, i.e., 10^"10M, 10^"nM, 10^"12M or 10^"13M towards the immobilisation tag.

Preferably the immobilisation tag is selected from the group consisting of a polyarginine tag (Arg-tag), polyhistidine tag (His-tag), FLAG-tag, Strep-tag, c-myc- tag, S-tag, calmodulin-binding peptide, cellulose-binding domain, SBP-tag„ chitin- binding domain, glutathione S-transferase-tag (GST), maltose-binding protein, transcription termination anti-termination factor (NusA), E. coli thioredoxin (TrxA) and protein disulfide isomerase I (DsbA).

Order of contact with solid surface

Where the embodiments of the invention require immobilising any one of the components of step (a) onto a solid surface, the immobilisation may occur, before the components of step (a) are combined; after the components have been combined; after the components have been combined and the sample has been incubated for a finite period of time; or after the sample has been exposed to the labelled binding partner.

For example, step (a) may be carried out first and the ubiquitination reaction allowed to proceed for a finite period of time. The sample may then be exposed to the solid surface or to the labelled binding partner. If it is exposed to the solid surface first, it will then be exposed to the labelled binding partner. If it is exposed to the labelled binding partner first, it will then be exposed to the solid surface. Exposure to the solid surface in either case allows for the immobilisation of any one of the components of step (a).

Alternatively any one of the components of step (a) may be immobilised to the solid surface first, then the remainder of the components of step (a) combined in a sample which is in contact with the solid surface. The ubiquitination reaction is allowed to proceed and then exposed to the labelled binding partner.

Preferably, the methods of the invention involve completing step (a), allowing the reaction to proceed for a finite period of time, completing step (b), and then exposing the sample to a solid surface.

Screening for modulators of ubiquitination

One of the components of the sample may be a potential modulator of ubiquitination. By measuring levels of ubiquitination both in the presence and absence of a potential modulator, it is possible to assess the effect that any potential modulator has on ubiquitination. The modulator may affect one or more of the components involved in ubiquitination. For example, the modulator may affect El, E2 or E3 directly or may affect the binding of these components to other components of the ubiquitination pathway. The possibility of using different substrates allows for modulators of ubiquitin ligase activity to be screened for their differing effects on different substrates.

Modulator

A modulator may be a compound which can increase or decrease the level of ubiquitination. By potential modulator it is meant that the compound either acts or is suspected to act as a modulator.

As E3 ligases comprise multiple protein components, all of which form interactions with neighbouring components as well as with other members of the ubiquitination pathway, the ubiquitin ligase provides many targets for potential modulators of activity. Potential areas of the ubiquitination pathway which may be targeted by modulators of ubiquitination may include but are not limited to E3 ligase interaction with the substrate, interaction between ubiquitin and the substrate, interactions between two protein subunits of the ubiquitin ligase, interactions between El enzyme and E2 enzyme, interaction between E2 enzyme and ubiquitin ligase, or a combination of these.

The ability of the methods of the present invention to provide an accurate and reliable ubiquitination assay allows for a large number of potential modulators to be screened effectively.

Controls

To detect ubiquitination using the methods of the invention a reference point is typically needed i.e. a control. Analysis of a control sample gives a standard level of ubiquitination and against which the sample can be compared.

A negative control sample may contain all components of the sample with a non-functional protein component that is essential for the reaction to take place. The negative control may also contain all but one of the components required for ubiquitination, e.g. ATP may be omitted from the reaction. Positive controls may also be used as a reference point to which a sample can be compared. For example, a positive control sample may contain components which give a known amount of ubiquitination over a set period of time.

Potential modulators may up-regulate or down-regulate the level of ubiquitination. To detect such up- or down-regulation, a reference point is typically needed i.e. a control. Analysis of a control sample gives a standard level of ubiquitination and therefore a standard level of ubiquitination against which the modulated sample can be compared.

A positive control sample may contain all components of the sample but lack the potential modulator. This can then provide a basal level of ubiquitination in the absence of the potential modulator. Negative controls may also be used as a reference point to which a sample containing a modulator can be compared. A negative control sample may contain a modulator which has a known effect on the level of ubiquitination.

High throughput screening

Optionally the methods of assaying provided by the invention may be used in the high throughput screening of potential modulators of ubiquitination. High throughput screening is a method of screening which allows high numbers of similar tests to be performed in parallel quickly and efficiently. High throughput screening may be achieved using a multi-well plate, which are well known in the art and include but are not limited to plates which comprise a grid of, for example, 96, 384, 1536, or 3456 small, open wells. These wells can each contain different samples to be tested, and the test can be conducted on all of the samples contemporaneously.

Kits

The invention also includes kits which are suitable for carrying out the methods of the invention. Preferably the kits will comprise the components required for step (a) of the methods of the invention as described above in addition to a labelled binding partner. The variations and preferred embodiments described in relation to these components are also intended to be reflected in preferred embodiments of the kits. That is, if a particular component is preferred for the methods of the invention, it is also preferred for the kits.

General

The term "comprising" encompasses "including" as well as "consisting" e.g. a composition "comprising" X may consist exclusively of X or may include something additional e.g. X + Y.

The term "about" in relation to a numerical value x means, for example, x+10%.

The word "substantially" does not exclude "completely" e.g. a composition which is "substantially free" from Y may be completely free from Y. Where necessary, the word "substantially" may be omitted from the definition of the invention.

Various aspects and embodiments of the present invention will now be described in more detail by way of example.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 : List of complexes, proteins, manufacturing names, expression systems and accession numbers used throughout this study.

Figure 2: Ubiquitin biotinylation. Success of biotinylation and optimal conditions determined by ELISA (Streptavidin plate, anti-Ub primary antibody (R+D Systems, MAB701, 0.1^g/ml), anti species-HRP secondary (Sigma, A5278, 1 μg/ml)

Figure 3: Ubel and Ubel UbcH3 - Gel based. Incubated El Ub/ATP and El E2/Ub/ATP (lOOhr incubation) tested using SDS PAGE gels. A band shift indicative of the presence of Ubiquitinated Ubel has been observed in both El and E1/E2 assays. No band associated with Ubiquitinated UbcH3 is visible. This band shift is ATP dependent. Antibody used: Mono- and poly-Ubiquitinated conjugates, mouse mAb horseradish peroxidise conjugate (FK2H) - Enzo Life Sciences, Cat no. PW0150, Batch no. X08036, supplied at lmg ml, used at 1/1000 dilution.

Figure 4: Pre-incubated Ubel/UbcH3 - HTRF. Pre-incubated El/Ub/ATP and El/E2 Ub/ATP (20, 40 and 100 hour incubation) tested using HTRF platform. Anti- 6His-K was employed for El and E2 detection; Anti-FLAG-K for E2; Biotinylated Ub; and Streptavidin XLent! for biotin. FRET signal indicative of the presence of Ubiquitinated Ubel and UbcH3 has been observed. This signal is enzyme and ATP dependent for the E1/E2 assays, as well as time dependent for the El assay.

Figure 5: 'Live' Ubel assay (60 minutes incubation) - HTRF. 'Live' assays (60 minute incubation) for Ubel at higher concentrations (>250 nM) show a good signal which is dependent on the Ub-biotin concentration. Anti-6His-K was employed for El detection; biotinylated Ub; and Streptavidin XLent! for Biotin. At lower El concentrations, the excess of Ub-biotin inhibits the detection of a signal.

Figure 6: Pre-incubated Ubel/UbcH3-ECL. Pre-incubated El/Ub/ATP and El/E2/Ub/ATP (100 hour incubation) tested using ECL platform. Used Anti-His capture for El (Millipore, 16-255, 1/500 - 1/4000 dil); anti-FLAG capture for E2 (Millipore, MAB3118, 1/500 - 1/4000); biotinylated Ub; and streptavidin sulfo tag (MSD, R32AD-1, ^g/ml). Signal of the presence of ubiquitinated Ubel and HbcH3 has been observed.

Figure 7: Live Ubel assay (60 minutes inc.) - ECL. Initial 'live' assays (60 minute incubation) for Ubel show good signal (> 1000000) but very high background level (>250000) (Max signal/background 4-5 fold). Subsequent Ubiquitin only controls indicates that at elevated concentrations of ubiquitin, there is non-specific ubiquitin binding and consequently high background.

Figure 8: Live Ubel assay (60 minutes inc.) - ECL. 'Live' assays (60 minute incubation) for Ubel at lower concentrations of Ub (2μΜ) still show good signal (> 1000000) and lower background (<10000) (Max signal/background -100 fold). Observed 'lag' in signal increase with increasing Ubel concentration. Signal is low at low concentrations of Ubel (<25mM). Approximately linear increase in signal after 25nM.

Figure 9: ATP and Ubiquitin titration (Ubel Assay) - ECL. 'Live' Ubel assay El/Ub/ATP, varying ATP and Ubiquitin concentrations. Uses Anti-His capture for El (Millipore, 16-255, 1/1000 dil); lOOnM Ubel; biotinylated Ub; streptavidin Sulfo tag (MSD, R32AD-1, ^g/ml). Saturating ubiquitin and ATP conditions - 2μΜ ΙΛ/500μΜ ATP.

Figure 10: Ubel timecourse- ECL. El assay appears to reach completion after a very short time period. Impossible to monitor in current format. Could reduce the amounts of ATP Ub or decrease reaction temperature to see reaction progress.

Figure 11 : Ubel inhibitor study PYR41- ECL. Inhibitor study. Titration of PYR41 (Calbiochem 662105) from 300μΜ to ΙΟηΜ. Used anti-His capture for El (Millipore, 16-255, 1/1000 dil); 50nM El; 2μΜ Ub; 500μΜ ATP; 30- minute pre-incubation with cpd.; 30 minute reaction time. IC₅₀ of 1.3μΜ - Similar to literature IC₅₀ value: BiogenNova quote IC50 value as ~5uM.

Figure 12 (a): Pre-incubated UbH3-ECL. Used pre- incubated His-El/FLAG, c-Myc, HA-E2 Ub/ATP (24 hour incubation) and was tested using ECL platform. Used Anti- FLAG, c-Myc and HA capture for E2 (Millipore, MAB3118 (FLAG), 05274 (c-Myc), 05-904 (HA), all 1/500 - 1/4000); biotinylated Ub (ΙμΜ); ATP 500 μΜ; E1:E2 pre inc. 1 μΜ:1 μΜ; streptavidin sulfo tag (MSD, R32AD-1, 1 μξ/ιοϊ). Signal indicative of the presence of Ubiquitinated UbcH3 has been observed.

Figure 12 (b): 'Live UbcH3 - ECL. Used Anti-HA capture for E2 (Millipore, 05- 904(HA), 1/500); 12.5 to 200nM El; 6.25 to 200mM E2; 2 μΜ Ub; 500 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, 1 μ^ιηΐ). Initial 'live' assays (60 minute incubation) for UbcH3 show good signal (> 1500000) at high El concentrations. Max signal appears to be dependent on the amount of El in the assay. HA and c-Myc UbcH3's perform better than the FLAG version.

Figure 13: 'Live' UbcH3 Assay - ECL. Used Anti-HA or Anti-c-Myc capture for E2 (Millipore, 05-904(HA), 1/500) (Millipore, 05-274 (c-Myc), 1/500); 3.125 to 25nM El; 6.25 to 200mM E2; 2 μΜ Ub; 500 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, 1 μg/ml). Some indication that non-specific binding of Ubel leads to high signal. 'Live' UbcH3 assay repeated using low concentrations of Ubel. Figure 14 (a) Pre-incubated SCF^skp2 - ECL. Used pre-incubated El/E2/E3/p27 Ub/ATP (24 hour incubation) tested using ECL platform. Used anti- FLAG capture for p27 (Millipore, MAB3118 (FLAG), 1/500 to 1/4000); E1:E2:E3 based on Xu publication (40nM El: 5 μΜ E2 : 40nM E3; biotinylated Ub (10 μΜ), 500 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, 1 μg/ml). Signal indicative of the presence of Ubiquitinated p27has been observed.

Figure 14 (b) 'Live' SCF^skp2- ECL (b). Used 'live' El/E2/E3/p27/Ub/ATP (60 minute incubation) tested using ECL platform. Used c-Myc and HA capture for p27 (05274 (c-Myc), 05-904 (HA), 1/500); 5nM El ; 4 to lOOOnM E2 (HA or c-Myc); 50nM E3; 3.125 to ΙΟΟηΜ p27; 2μΜ Ub; 500 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, 1 g/ml). Signal indicative of the presence of ubiquitinated p27 has been observed.

Figure 15: Ubel and Ubel+UbcH3 assay timecourse - HTRF. El assay: 1 μΜ NiNTA Ubel + 1 μΜ Ub + 2mM ATP. El + E2 assay: 1 μΜ NiNTA Ubel + ΙμΜ Flag UbcH3 + 1 μΜ Ub + 2mM ATP. Both the El and El + E2 assays appear to reach completion after a very short time period, which is in line with the ECL data. Used Anti -6HisK for El + E2 detection; Anti-FLAG-K for E2; biotinylated Ub; streptavidin XLent! for biotin. Impossible to monitor in current format. Samples were diluted to lOnM El in xl reaction buffer (RB).

Figure 16: E2 and ubiquitin titration (Ubel + UbcH3 assay) - HTRF. 'Live' Ubel + UbcH3 assay - lOOnM El/ varying [E2] / 200nM Ub/ 2mM ATP. 60 minute room temperature incubation. 'Live' Ubel + UbcH3 assay - lOOnM El/ 1 μΜ E2/ varying [Ub]/ 2mM ATP, 60 min room temperature incubation. Samples diluted to lOnM El in xl reaction buffer. Used Anti-6His-K for El + E2 detection; anti FLAG - K for E2 detection; biotinylated Ub; streptavidin XLent! for biotin. Optimal El, E2 and Ubiquitin conditions are lOOnM Ubel / 1 μΜ UbcH3 / 400nM Ub.

Figure 17: Ubel and UbcH3 construct comparison (E1+E2 assay) - HTRF. 'Live Ubel and UbcH3 assay: El/E2/Ub/2mM ATP, varying El and E2 constructs at: lOOnM El + 1 μΜ E2 + 400nM Ub. Used anti-6His-K for El + E2 detection; anti FLAG-K, anti HA-K or anti cMyc-K for E2 detection; biotinylated Ub; streptavidin XLent! for biotin. NiNTA-Ubel and HA-UbcH3 in a ratio of lOOnM El + 1 μΜ E2 + 400nM Ub assayed at 2mM ATP; and diluted 10-fold gives the best FRET-signal. Figure 18: Ubel Inhibitor study PYR41 in El and E1+E2 assay -HTRF. Inhibitor study: 30 minutes pre-incubation with PYR-41. Assays were immediately diluted followed by addition of 2mM ATP. Titration of PYR-41 (Calbiochem 662105) from 300 μΜ to 0.01 μΜ, IC₅₀ = 43 μΜ - El (1 μΜ El, 2 μΜ Ub (Anti-6His-K).

IC₅₀ = 16 μΜ - E1/E2 (lOOnM El, 1 μΜ E2, 400nM Ub (Anti -6-HIS-K)) IC₅₀ = 20 μΜ - E1/E2 (as above Anti-FLAG-K). BiogenNova quote IC₅₀ value as ~5 μΜ.

Figure 19: Western blot — Anti phospho p27 - E3 assay. Incubated El/E2/E3/p27Ub/ATP, Anti-phospho-p27 antibody (Millipore 06-996 1/1000). Anti species HRP (Cell Signalling #7074 1/1000). Definite shift in the position of the phospho-p27 band consistent with the poly-ubiquitination of p27.

Figure 20: Western blot - Anti Phospho p27. Incubated FLAG-p27/CDK2/cyclinE with CD 2/cyclinE (0.1 :0.1 mg/ml) and Mg-ATP (2 μΜ), anti phospho-p27 antibody (Millipore 06-996 1/1000). Anti species HRP (Cell signalling #7074 1/1000). Incubation of p27/CDK2/CyclinE with CDK2/cyclinE and ATP leads to a significant increase in the amount of phospho p27 as detected by Western blot. Max. signal is seen even after 20 minutes.

Figure 21 : 'Live' sCF^Skp2/cksl Assay - ECL. 'Live' El/E2/E3/p27/Ub/ATP (60 minute incubation) tested using ECL platform. No FLAG tagged E2/p27 due to problems encountered in E2 assay. Used c-Myc and HA capture for p27 (Millipore 05-274 (c- Myc), 05-904(HA), 1.500); 5nM El; 4 to lOOOnM E2 (HA or c-Myc); 50nM E3/Cksl; 3.125 to lOOnM p27 (HA of c-Myc); 2 μΜ Ub; 500 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1 1 μg/ml). Signal indicative of the presence of Ubiquitinated p27 has been observed. c-Myc tagged p27 gives significantly higher signal. Optimum signal at 50nM p27, 1 μΜ E2.

Figure 22: 'Live' sCF^{skp2 Cksl} Assay - ECL El vs. E2 titration. Used 60 minute incubation; 25nM c-Myc p27; anti c-Myc capture for p27 (Millipore 05-274, 1/500); 0.78 to 50nM El; 15 to lOOOnM E2 (HA); 25 nM E3/Cksl ; 2 μΜ Ub; 500 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, 1 μ§/ηύ). El- hyperbolic increase in signal with [El]. Optimum signal after 25 nM. However, optimum signal/background at 6.25 nM El . E2- linear increase in signal with [E2]. Final assay concentration set at 1 μΜ.

Figure 23 (a): 'Live' sCF^Skp2/Cksl Assay - ECL ubiquitin titration. Used 60 minute incubation; 25nM c-Myc p27; anti c-Myc capture for p27 (Millipore 05-274, 1/500); 5nM El ; 1 μΜ E2 (HA); 25nM E3/Cksl; 5 to 0.08 uM Ub; 1000 to 15 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, 1 μg/nll). Hyperbolic increase in signal with [Ub]. Optimum signal after 2 μΜ ubiquitin. Largely dependent of ATP concentration. K_m values between 0.46 and 0.58 μΜ. Final assay concentration is set at 2 μΜ.

Figure 23 (b): 'Live' sCF^Skp2/Cksl Assay - ECL ATP titration. Used 60 minute incubation; 25nM c-Myc p27; anti c-Myc capture for p27 (Millipore 05-274, 1/500); 5nM El; 1 μΜ E2 (HA); 25nM E3/Cksl ; 2 μΜ Ub; 1000 to 7 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, 1 μ§Λ^η1). Initial experiments indicated a high signal which was independent of ATP concentration (lowest [ATP] of 21 μΜ) - indicative of tight binding. Repeat experiment showed similar profile with only a moderate decrease at 7 μΜ ATP. K_m estimated at 4 μΜ. Final assay concentration set at 100 μΜ.

Figure 24: 'Live' sCF^{skp2 Cksl} Assay - ECL E3 tetramer/Cksl titration. Used 60 minute incubation; 25 nM c-Myc p27; anti c-Myc capture for p27 (Millipore 05-274, 1/500); 5nM El; 1 μΜ E2 (HA); 3 to 100 nM E3/Cksl ; 2 μΜ Ub; 100 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, 1 μ£ ηι1). Hyperbolic increase in signal with [E3/Cksl]. Optimum signal at 50nM E3/Cksl . Re-plotting the data between 0 and 25 nM E3/Cksl indicates a linear relationship. Final assay concentration set at 25nM.

Figure 25: 'Live' sCF^{skp2 CksI} Assay - ECL Cksl titration - FIXED E3 tetramer. Used 60 minute incubation; 25 nM c-Myc p27; anti c-Myc capture for p27 (Millipore 05- 274, 1/500); 5nM El ; 1 μΜ E2 (HA); 25nM E3 tetramer; 3 to 100 nM Cksl ; 2 μΜ Ub; 100 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1 , 1 μ^ιηΐ). Linear increase in signal with [Cksl] up to 50nM followed by plateau. Optimum signal at 50nM Cksl. Final assay concentration set at 25nM.

Figure 26: 'Live' sCF^Skp2/Cksl Assay - ECL Skpl/Skp2 vs Cull/Rbxl titration. 'Live' El/E2/E3/p27/Ub/ATP (60 minute incubation) tested using ECL platform. Co- expressions of Sk l/Skp2 and Cull/Rbxl instead of tetramer. Used c-Myc and HA capture for p27 (Millipore 05-274 (c-Myc), 05-904 (HA), 1/500); 5nM El ; ΙμΜ E2 (HA or c-Myc), 25mM Cksl ; 25nM p27; 2μΜ Ub; ΙΟΟμΜ ATP; 50 to 1.5nM Skpl/Skp2; 50 to 3.1 nM Cull/Rbxl ; streptavidin sulfo tag (MSD, R32AD-1, ^g/ml). Optimum signal at 50 nM Cull/Rbxl and 25nM Skpl/Skp2.

Figure 27: Comparison of the Skpl/Skp2 and Cull/Rbxl co-expressions with the Skpl/Skp2/Cull/Rbxl tetramer. Indicates very little difference in max signal strength. Figure 28: 'Live' sCF^Skp2/cksl Assay - ECL time course. Used 10 to 100 minute incubation. 25nM c-Myc p27; anti c-Myc capture for p27 (Millipore 05-274, 1/500); 5nM El; ΙμΜ E2 (HA); 25nM E3 tetramer; 25nM Cksl ; 2μΜ Ub; ΙΟΟμΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, \μ§/πύ). Linear increase in signal with time up to 100 minutes. Experiment repeated to check reproducibility. Comparable signal strength/profile seen for both experiments. High y-intercept may be due to 'stop' method. Final reaction time set at 60 minutes.

Figure 29: 'Live' sCF^{Sk 2/Cksl} Assay - ECL Z'. Used 60 minute incubation; 25nM cMyc p27; anti c-Myc capture for p27 (Millipore 05-274, 1/500); 5nM El ; ΙμΜ E2 (HA); 25nM E3 tetramer; 25 nM Cksl; 2μΜ Ub; 100 μΜ ATP; Streptavidin sulfo tag (MSD, R32AD-1, μ§/πύ). 1 sector of positive controls with one sector of negative (no Ubiquitin) controls. Total of 3 independent experiments. Signal strength, signal/background ratio, positive control %CV and Z' all within acceptable limits. Some variation in signal strength.

Figure 30 (a): 'Live' sCF^Skp2/Cksl Assay - ECL Inhibitor study - PYR-41 (El inhibitor). Used 60 minute incubation; 25nM c-Myc p27 (phosphorylated in the absence of inhibitor); anti c-Myc capture for p27 (Millipore 05-274, 1/500); 5nM El ; ΙμΜ E2 (HA); 25nM E3 tetramer; 25nM Cksl ; 2μΜ Ub; ΙΟΟμΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, lug/ml; titration of PYR41 (Calbiochem 662105) from 300μΜ to lOnM. IC₅₀ of 55μΜ - significantly less potent than quoted IC₅₀ against the El assay only. BiogenNova quote IC₅o value as ~5μΜ vs. El .

Figure 30 (b): E1/E2 Assay - ECL Inhibitor study - PYR-41. Used 60 minute incubation; 2μΜ Ub; ΙΟΟμΜ ATP. For El assay (Anti-His capture) 50nM El . For E2 assay (Anti HA-capture) 5nM El ; 50nM E2. Used streptavidin sulfo tag (MSD R32AD-1, ^g/ml); titration of PYR41 (Calbiochem 662105) from 300μΜ to lOnM. IC₅o values of 7.8 and 3.2μΜ for the El and E2 assays respectively. Significantly more potent than IC₅₀ against the E3 assay.

Figure 31 : 'Live' sCF^skp2/Cksl Assay - ECL R140 Kinase inhibitor panel (+PYR-41)

Figure 32: 'Live' sCF^{Skp2 Cksi} Assay - ECL R140 Kinase inhibitor panel. Used 60 minute incubation; 25nM c-Myc p27; anti-c-Myc capture for p27 (Millipore 05-274, 1/500); 5nM El ; Ι μΜ E2 (HA); 25 nM E3 tetramer; 25nM Cksl ; 2μΜ Ub; 100 μΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, ^g/ml). Several compounds show up as inhibitors- Rottlerin, EGCG, K252c and AG538.

Figure 33: 'Live' sCF^Skp2/Cksl Assay - ECL R140 Kinase inhibitor panel. Comparison of the sCF^skp2/Cksl inhibition profile against the CDK2/CyclinE inhibition profile highlights very few similarities.

Figure 34: 'Live' sCF^{skp2 Cksl} Assay - ECL follow up IC₅₀s. Used 60 minute incubation; 25nM c-Myc p27; anti c-Myc capture for p27 (Millipore 05-274, 1/500); 5nM El; ΙμΜ E2 (HA); 25 nM E3 tetramer; 25nM Cksl; 2μΜ Ub; ΙΟΟμΜ ATP; streptavidin sulfo tag (MSD, R32AD-1, ^g/ml). Titration of PYR41 (Calbiochem 662105) from 300 μΜ to lOnM (control). Titration of Rottlerin, EGCG, AG538 and K252c from 300μΜ to 10 nM. PYR-41 IC₅₀ of 84 μΜ - Comparable to previous IC₅₀ data.

Figure 35: El, E2 and E3 assay non-specific binding. El assay - anti His capture for specific binding and BSA block only for NSB. Clearly some NSB, however the relative portion of NSB drops as the concentration of El is reduced. NSB signal in the E3 assay is unlikely to be particularly high due to the low concentration of El present (5nM).

Figure 36: El, E2 and E3 assay non-specific binding. E2 assay - Anti HA capture for specific binding and BSA block only for NSB. No indication of non-specific binding in the E2 assay.

Figure 37: El, E2 and E3 assay non-specific binding. E3 assay - anti c-Myc capture for specific binding and BSA block only for NSB. Clearly some NSB, however the relative portion of NSB drops as the concentration of p27 is reduced. Although there is some non-specific binding, it is unlikely to be NSB from the desired captured species.

Figure 38: El, E2 and E3 assay non-specific binding. Multiple approaches were employed in order to minimise the E3 assay NSB signal including: Buffer pH; BSA; Salt; detergents; buffers; p27 E3 concentrations; blocking solutions. None were successful.

Figure 39: Epitope tag, p27 and E2 optimisation in an E3 assay (HTRF). 'Live' E3 assay - lOOnM El/ ΙμΜ E2 (various epitope tags) / lOOnM phospho-p27 (various epitope tags)/ lOOnM E3 / lOOnM Cksl / 400nM Ub / 2mM ATP. 20 hour incubation at room temperature. Sampled diluted to varying [phospho-p27] in lx reaction buffer. Used anti-FLAG-K, HA-K, and c-Myc-K for p27 detection; biotinylated Ub; streptavidin XLent! for biotin. Highest p27 signal derived from FLAG-E2 and c-Myc- E2 with HA-p27.

Figure 40: E3 assay - His-tagged proteins, E2 and p27 detection (HTRF). E3 assay - lOOnM El / ΙμΜ E2/ lOOnM phospho-p27/ lOOnM E3 / lOOnM Cksl / 400nM Ub / 2mM ATP. 60 minutes incubation at room temperature. Samples diluted to 40nM Ub- B in lx reaction buffer. Used anti-His-K for His-tagged protein detection; anti-HA-K for E2-detection; anti-c-Myc-K for p27 detection; biotinylated Ub; streptavidin XLent! for biotin. In the presence of E3 tetramer, a decrease in His-tagged protein and E2 signal strength was observed that did not correspond with an increase in p27 signal strength.

Figure 41: E3 assay - phospho-p27 vs. E3 tetramer titration (HTRF). E3 assay - lOOnM El / ΙμΜ E2 / lOOnM phospho-p27 / lOOnM E3 / lOOnM Cksl / 400nM Ub / 2mM ATP. 60 minute incubation at room temperature. Samples diluted to 40nM Ub- B in lx reaction buffer. Used anti-c-Myc-K and HA-K for p27 detection; biotinylated Ub; streptavidin XLent! for biotin. A bell shaped relationship was exhibited for p27 signal strength with varying E3 tetramer concentration irrespective of the E2 and p27 constructs used. The p27 signal obtained is not sufficient for a reliable assay.

Figure 42: E3 assay - Effect of [Ub-B] on p27 signal (HTRF). E3 assay - 5nM El / ΙμΜ E2 / 50nM phospho-p27 / 25nM E3 / 25nM Cksl / 400nM or 2 μΜ Ub / 2mM ATP. 60 minute incubation at room temperature. Samples diluted to 40nM Ub-B in lx reaction buffer. Used Anti-HA for p27 detection; biotinylated Ub; streptavidin XLent! for biotin. [Ub-B] had no effect on p27 signal strength in the presence or absence of E3 tetramer.

Figure 43: HTRF and ECL platform comparison - E3 titration in an E3 assay. E3 assay - 5nM El / ΙμΜ E2 / 50nM phospho-p27 / 25nM E3 / 25nM Cksl / 2μΜ Ub / 500μΜ ATP. 60 minutes incubation at room temperature. Samples diluted to multiple [Ub-B] in lx reaction buffer for HTRF detection. Used anti-c-Myc-K for p27 detection; c-Myc capture for p27 (Millipore 05-274 (c-Myc), 1/500); biotinylated Ub; streptavidin XLent! for biotin. A hyperbolic increase in p27 signal generated on the ECL platform compared to no significant p27 signal on the HTRF platform indicated problems with detection capabilities of the HTRF platform.

Figure 44: E3 assay: Effect of [NaCl] and [SA Xlent] on p27 signal (HTRF). E3 assay - 5nM El / ΙμΜ E2 / 50nM phospho-p27 / 25nM E3 / 25nM Cksl / 2μΜ Ub / 500μΜ ATP. Incubated for 60 minutes at room temperature. Samples diluted to 40nM Ub-B in lx reaction buffer containing varying [NaCl]. Used Anti-HA-K for p27 detection; biotinylated Ub; streptavidin XLent! for biotin. [NaCl] had a negative impact on p27 signal strength. Increasing [SA XLent] enhanced signal strength, but above ^g/ml the signal to background ratio was not improved.

Figure 45: E3 assay - effect of El + E2 pre-charge at varying [Ub-B] (HTRF). E1 E2 pre-charge assay - 5nM El / ΙμΜ E2 / multiple [Ub-B] / 500μΜ ATP. Followed by E3 assay - [phospho-p27] titration / 25nM E3 / 25nM Cksl / Ub (final combined concentration = 2μΜ) / 500μΜ ATP. 60 minute incubation at room temperature. Samples diluted to 40nM Ub-B in lx reaction buffer. Used anti-HA- for p27 detection; biotinylated Ub; streptavidin XLent! for biotin. Pre-charging El + E2 in the presence of varying concentrations of Ub-B had no effect on p27 signal strength or signal to background ratios.

Figure 46: E3 assay - phospho-p27 vs. Cksl titration, p27 poly - ubiquitin interchain FRET (HTRF). E3 assay - 5nM El / ΙμΜ E2 / [phospho-p27] titration / 25nM E3 / [Cksl] titration / 2μΜ Ub-B / 500μΜ ATP. 60 minute incubation at room temperature. Samples diluted to 40nM Ub-B in lx reaction buffer. Used anti-HA-K for p27 detection; biotinylated Ub; streptavidin XLent! and streptavidin-K for biotin interchain FRET. Both forms of detection exhibited a bell shaped relationship between the concentration of phospho-p27 and signal for each Cksl concentration tested.

Figure 47: E3 assay - [donor and acceptor fluorophore] optimisation for p27 poly- ubiquitin interchain FRET (HTRF). E3 assay - 5nM El / ΙμΜ E2 / ΙΟΟηΜ phospho- p27 / 25nM E3 / 200nM Cksl / 1 or 2μΜ Ub-B / 500μΜ ATP. 60 minute incubation at room temperature. Samples diluted to 40nM Ub-B in lx reaction buffer. Anti-HA- K and anti-HA XL665 for p27 detection; biotinylated Ub; streptavidin XLent! and streptavidin-K for biotin interchain FRET. ΙμΜ Ub-B over 2μΜ Ub-B improves signal strength and using a no ubiquitin control the signal to background ratio improves to 28. However, in the absence of E3 tetramer the ratio falls to 1.2 indicating the SA-K and SA XLent are potentially binding to other ubiquitinated components.

Figure 48: Effect of DTT on interchain and p27 signal detection (HTRF). E3 assay - 5nM El / luM E2 / lOOnM phospho-p27 / 25 n E3 / 200nM Cksl / 1 or 2 μΜ Ub-B / 500μΜ ATP. 60 minute incubation at room temperature. Samples diluted to 40nM Ub-B in lx reaction buffer containing varying [DTT]. Used anti-HA-K and for p27 detection; biotinylated Ub; streptavidin XLent! and streptavidin-K for biotin interchain FRET. Changing the redox environment through the addition of DTT to the diluent had no effect on interchain or p27 signal detection and signal to background ratios.

Figure 49: E3 assay - p27 signal detection using an anti-phospho-p27 antibody and IgG-Europium cryptate (HTRF). E3 assay - 5nM El / ΙμΜ E2 / lOOnM phospho-p27 / 25 nM E3 / 200nM Cksl / ΙμΜ Ub-B / 500μΜ ATP. 60 minute incubation at room temperature. Samples diluted to 40nM Ub-B in lx reaction buffer. Anti-phospho-p27 (Thrl87) antibody (Millipore 06-996) and IgG-K for p27 detection; biotinylated Ub; streptavidin XLent! and for biotin. Using a bridging phospho-specific antibody had no effect on p27 signal, when coupled to a species specific IgG-Europium cryptate and APC.

Figure 50: HTRF and ECL platform comparison - gcF^{Skp2 cksl} component knock-out in an E3 assay. E3 assay - 5nM El / ΙμΜ E2 / lOOnM phospho-p27 / 25nM E3 / 50nM Cksl / Ι Μ Ub-B / 500μΜ ATP. 60 minute incubation at room temperature. Samples diluted to 40nM Ub-B in lx reaction buffer for HTRF detection. Anti- phospho-p27 (Thrl87) antibody (Millipore 06-996) and IgG-K for p27 detection; HA capture for p27 (Millipore 04-904(HA), 1/500); biotinylated Ub; streptavidin XLent! and for biotin. HTRF— No difference were detected in the knock-out samples, apart from the no ubiquitin control suggesting non-specific binding of the fluorophores. ECL - ubiquitination of p27 is occurring under the stated conditions.

Figure 51 : Schematic showing a current being passed through an electrode, which in turn causes stimulation of the labelled binding partner, in turn causing a signal to be emitted. MODES FOR CARRYING OUT THE INVENTION

Coding sequences used throughout this study were cloned and validated against appropriate GenBank entries (figure 1).

Molecular weights of purified proteins were confirmed by SDS PAGE followed by Western blot analysis.

Assay development pilot: components upstream of E3

Initial experiments in the pilot phase focussed on establishing that the components upstream of the E3 are active, and then optimising the levels of each component required so that we limit the variables for the E3 assay.

The majority of assays require the use of biotinylated ubiquitin in order to recruit streptavidin linked reporter molecules. Biotinylation reactions were carried out and assessed by ELISA to determine optimal biotin:ubiquitin ratios (figure 2).

Results showed that using a labelling ratio of either 2:1 or 5:1 biotimubiquitin provided the best signal by ELISA. Higher ratios were detrimental, and lower ratios look to be inefficient.

Positive controls of ubiquitinated El and/or E2

Prolonged incubation of Ubel (El), UbcH3 (E2) and ubiquitin at ratios of 1:1 :20 micromolar were set up to generate a positive control of ubiquitinated El and/or E2 which could then be used in establishing detection conditions for the HTRF and ECL platforms. Initial SDS polyacrylamide gel-based examination of ubiquitinated Eland E1/E2 showed that there was a ubiquitination event occurring, with an apparent shift in the molecular weight of El (figure 3). A decrease in the intensity of the ubiquitin band could also be seen in some samples. The assays were set up with stoichiometric amounts of El and E2 and a 20-fold excess of biotinylated ubiquitin. The band shifts were shown to be ATP dependent.

The gels were also probed with an antibody against ubiquitinated species. The antibody does not recognise free ubiquitin. The Western blots for this experiment showed a clear band representing ubiquitinated El. It was noted at this stage that the intensity of the band for El was markedly increased when E2 was present. Western blot analysis (Anti-mono and polyubiquitinated protein-HRP conjugate) revealed a positive response for Ubel only in the presence of ATP. This indicates ubiquitination of Ubel. No positive response for UbcH3 was observed in the absence of Ubel . When Ubel and UbcH3 were added together, we detected a positive response Ubel, which was time dependent.

Developing the detection methodology - HTRF

Samples were pre-incubated with equimolar amounts of the El and E2, and bio- ubiquitin and a dilution step was adopted to reduce the amount of El in the detect step to lOnM. This resulted in encouraging ratios, indicating that the format was detecting the presence of ubiquitinated El and E2

Figure 4 shows the bell-shaped curve typical of a titration in this format of assay.

The initial experiments were all carried out using prolonged incubation times to try to ensure that the reaction had progressed as far as it was likely to go in order to try to maximise the population of ubiquitinated El or E2 (figure 4).

Subsequent assays brought the incubation times to a more realistic period. The 60 minute live assay (i.e. not pre-incubating) using a 2-way checkerboard for El and bio- ubiquitin showed that the HTRF ratio will form in this time-frame. Samples were again diluted to 10 nM El detection concentrations (figure 5).

The amount of bio-ubiquitin excess over the El concentration is critical, else the signal gets lost in the background.

All HTRF specific reagents were from Cisbio, including:

Streptavidin-Xlent (611SAXLB), Anti-6His-K (61HISKLA), Anti-FLAG-K (61FG2KLA), and were used at concentrations specified by the manufacturer.

Developing the methodology - ECL

Unlike the HTRF format, the ECL assay is a wash format assay. The initial experiments using the pre-incubated material of 1 :1 :20 El:E2:Bio-Ub at 2 mM ATP showed a signal response dependent on the amount of El or E2 present. The experiment enabled the selection of the type of plate (high bind vs low bind) and the selection of capture antibody concentrations for subsequent experiments (figure 6). It also indicated that the E2 was becoming ubiquitinated (figure 7). While the assay format is wash based, it has been shown that the concentration of bio- ubiquitin that can be used is not limitless, and will contribute to a significant background. Reducing the amount of bio-ubiquitin in the assay restores the S/B accordingly. An El titration gives a good response, and there are hints of sigmoidal behaviour (figure 8).

The next steps for this format are the bio-ubiquitin and ATP titrations to establish the concentrations that will be considered non-rate limiting in downstream ubiquitination assays. The E2 also needs to be titrated into the system. There are 3 E2 molecules available, and the best performing one will be selected for the final stage of the pilot study.

Continued optimisation using the ECL assay

Continuing with the El protein optimisation, ATP and ubiquitin titrations were performed to identify the conditions which are saturating, or non-rate limiting, for the El section of the cascade (figure 9).

The results suggest that after 2 micromolar Ubiquitin and at 500 micromolar ATP, there is no additional signal generation. These concentrations have therefore been taken as the saturating conditions for the El protein.

Following the identification of the ATP and ubiquitin concentrations to use, a time- course of signal generation was followed, to determine a linear part of the reaction velocity curve to use for subsequent inhibitor experiments (figure 10).

What was found was that the reaction appears to happen very quickly, in fact within the dead time of setting up the assay. The fast nature will be an advantage in downstream events with the E2 and E3 assays.

For the IC₅₀ determination of Pyr-41, the data shown is that from a 30 minute preincubation and a 30 minute assay. The assay was performed in a separate plate and then transferred. For information purposes, another set of assays were performed within the capture/detect plate and stopped immediately by washing off the assay solution, or washed after 60 min. It was interesting to note that even with the wash at time 0 an IC₅₀ for the compound that was in line with the data shown could still be determined.

Reported values for the inhibitor are in the low micromolar range (figure 1 1). The detection of ubiquitinated E2 was explored, initially using material that had been pre-incubated for 24 hrs. Equimolar El, E2 and ubiquitin were used for this preliminary experiment, and the concentration of capture antibody determined. The results below are representative, and all 3 E2 constructs were tested.

This experiment was followed up with an El vs E2 checkerboard using 2 micromolar ubiquitin and 500 micromolar ATP, concentrations determined to be optimal for the El ubiquitination previously. The assay was this time set up in a live format rather than use material from a prolonged incubation (figure 12).

Lower concentrations of El were tried in an attempt to bring down any background (figure 13). The lower range of El seemed to retain a decent E2 specific signal and also reduced non-specifics.

The assay for the E3 ubiquitination of p27 was trialled, initially using conditions described in the Xu paper (Xu et al, 2007, JBC, 282 15462-470), and then using conditions identified in-house. The former also included a capture antibody titration for the FLAG-tagged p27 (figure 14).

Both experiments demonstrate a signal for the ubiquitinated p27. The latter also indicates that the higher concentrations of p27 may be detrimental.

Continued optimisation using the HTRF format

The progress with the El assay was continued by optimising the conditions for generation of signal.

A time course for the assay was examined initially using equimolar concentrations of the El , E2 and ubiquitin.

Similar to the ECL format, the time-course in the HTRF assay indicates a very fast reaction (figure 15).

E2 was titrated against fixed concentrations of Ub and El. The detection of ubiquitination was configured in two ways, using Anti-6His-K and Anti-FLAG-K. The former should detect the combination of ubiquitinated El and E2 as both carry a 6His tag, while the latter will detect the E2 ubiquitination only. From the titration of El vs E2, concentrations of El and E2 were chosen for the ubiquitin titration, again detected with two configurations to arrive at conditions shown below for El, E2, and Ub (100 nM El, 1000 nM E2, 400 nM Ub) (figure 16).

The samples require dilution to a lower concentration than used in the assay, and by using the dilution methodology, workable ratios have been obtained. The diluent itself has also proven to be important, as diluting in the Stop solution, which contains EDTA (120 mM), does reduce the signal window compared to stopping the assay and then diluting in reaction buffer.

The various constructs of E2 and the two different El purifications were tested under the conditions derived above to see if some of the variables could be reduced.

The bar chart on the next slide shows the different ratios observed for the E1/E2 and the specific E2 signals. A common dilution for detecting for both the El and E2 was identified (1/10 which gives 10 nM El in the detection), and it is this shown in figure 17.

Using conditions identified previously, an IC₅₀ determination for Pyr-41 using the El and E1/E2 assays was performed. The results show a marked difference in inhibitor potency compared to literature values and that determined in the ECL format. This may be a reflection of the higher enzyme concentrations used in the HTRF assay step (figure 18).

Continued optimisation using Western blots

As well as the ECL and HTRF assay formats, material has been examined by using SDS-PAGE and Western blotting to verify the presence (or absence) of ubiquitinated or phosphorylated material (figure 19).

The blots above show results from probing the membranes with an anti-phospho-p27 antibody. There is a clear indication that the p27 band has shifted, which could be interpreted as ubiquitination. The different loadings are 10, 20, 30 & 40 microlitres of sample from right to left. It should be noted that there was not sufficient sample for the 40 microlitre loading (estimated at 20 μί), so the signal is reduced from what would be expected.

On the right hand panel, a smear also appears in the No El lane. While this could mean that the El is not obligate, we have not seen this effect in either of the other assay formats, and so could also be a small contamination during the setup of this assay. Distinct banding can be seen in the lane with all components at 2 micromolar ubiquitin.

Figure 20 also shows anti-phospho-p27, this time from experiments designed to identify conditions to ensure a good level of phosphorylation which is required for the p27 to be ubiquitinated. It would appear that incubating the p27 with additional CDK2/Cyclin E and ATP for 20 minutes would suffice to ensure good phosphorylation.

ECL Format - E3 Ubiquitin Ligase Assay

Continuing with the E3 assay optimisation, p27 and E2 titrations were carried out using multiple tagging options on the E2 and p27 proteins.

The results indicate that the optimum signal was achieved when pairing the HA-E2 with the c-Myc-p27. This combination was used for all further E3 assay optimisation experiments.

Titrating the p27 against the E2 enzymes indicates an optimal signal strength at 1 μΜ E2 and 50 nM p27 (figure 21).

Following the identification of the optimum 'tag' pairing, an El vs. E2 checkerboard titration was performed in order to identify the optimum conditions for these assay components (figure 22).

El - Hyperbolic increase in signal with increasing El concentration with a signal plateau after 25 nM. However, due to the relatively high non-specific signal which originates from the El enzyme (LB1805pl2-15) the signal to background (No E2 background) ratio was taken into consideration when identifying a final assay concentration for the El enzyme in the E3 assay. The maximum signal to background is seen between 3.125 and 12.5 nM El. Final assay concentration of El set at 5 nM.

E2 - Approximately linear increase in signal with increasing E2 concentration. Unlike the El enzyme, non-specific binding is not a problem (LB1805p35-37). Although it appears that the signal will continue to increase with increasing E2 concentration, in the interests of minimising the total amount of protein and the assay costs, the E2 concentration was set at 1 μΜ. Continuing with the E3 assay component optimization, ubiquitin and ATP titrations were performed to identify saturating, or at least non-rate limiting, conditions (figure 23 (a) and (b)).

For each concentration of ATP tested, a hyperbolic increase in signal with increasing ubiquitin concentration was observed, with a signal plateau after approximately 2 μΜ. Fitting the data to a hyperbolic function generated apparent KM values for ubiquitin binding at multiple ATP concentrations. All of the apparent K values were between 0.46 and 0.58 μΜ ubiquitin.

The initial titration of ATP at multiple concentrations of ubiquitin demonstrated a high signal which was independent of ATP concentration (LB1805p80-83). A repeat of this experiment using lower concentrations of ATP(LB1805p87-89) again showed a signal largely independent of ATP concentration with only a moderate decrease at 7 μΜ ATP. The apparent KM value was estimated at 4 μΜ.

Final assay concentrations were set at 2 μΜ ubiquitin and 100 μΜ ATP.

Next, the E3 components Skpl/Skp2/Cull/Rbxl and Cksl were titrated together (figure 24). The concentration of the E3 tetramer (Skpl/Skp2/Cull/Rbxl) was set according to the concentration of the component which was least prevalent in the co- expression (Skp2). The Cksl concentration was set as equimolar to the concentration of the component which was least prevalent in the co-expression (Skp2).

A hyperbolic increase in signal with increasing E3/Cksl concentration was observed, with a signal plateau after 50 nM. Re-plotting the data from 0 to 25 nM E3/Cksl indicates a linear relationship between the signal and the E3/Cksl concentration. Although the maximum signal is observed at 50 nM E3/Cksl, a final assay concentration of 25 nM was chosen to ensure that the assay was within the linear region.

Finally, the Cksl component was titrated against a fixed concentration of the E3 tetramer (25 nM). A hyperbolic increase in signal with increasing Cksl concentration was observed, with a signal plateau after 50 nM. Although the maximum signal is observed at 50 nM Cksl, a final assay concentration of 25 nM was chosen to ensure the assay is within the linear range (figure 25).

Table 1 shows the final 'live' sCF^{skp2 Cksl} ECL assay concentrations. Table 1: Final 'live ' SCF^s*^p2/Cks/ ECL assay concentrations

Trialling different E3 co-expressions

The preceding E3 experiments were all carried out using an E3 comprised of a Skpl/Skp2/Cull/Rbxl tetramer alongside Cksl. Several other co-expressions were also available. Both a Skpl/Skp2 and a Cull/Rb l co-expression were available with sufficient purity/yield to be used in an assay. A titration of these 2 components against each other was carried out in comparison to the tetramer co-expression. The concentration of the Skpl/Skp2 or Cull/Rbxl was set according to the concentration of the component which was least prevalent. The optimum conditions were seen at 25 nM Skpl/Skp2, 50 nM Cull Rbxl in the presence of 25 nM Cksl. The comparison of the Skpl/Skp2 with Cull/Rbxl co-expressions against the Skpl/Skp2/Cull/Rbxl tetramer showed a largely similar max signal strength (figure 26). As all the previous experiments were carried out using the tetramer it was decided that all remaining assay development should be completed using the tetramer. Depending on the expression levels, yields and purities of the different co-expressions, this approach offers an alternative approach.

Comparison of the Skpl/Skp2 and Cull/Rbxl co-expressions with the Skpl/Skp2/Cull/Rbxl tetramer indicates very little difference in max signal strength. The tetramer was used for future experiments for ease of use and to corroborate with previous data (figure 27).

Time-course assays

Following the identification of optimal conditions for all of the assay components, a time-course (0-100 minutes) was carried out to determine a linear part of the reaction velocity curve to set for future inhibitor work (figure 28). The experiment was performed on 2 different days to check the reproducibility of the data. In both experiments, a comparable and an approximately linear relationship is observed between the reaction time and the signal strength. The reaction time for future experiments was set at 60 minutes. It is evident that the data does not pass through the origin as would be expected but intercepts the y-axis. This may be a result of the 'stop' method. The assay stop contains EDTA which sequesters metal ions and hence prevents the Mg-ATP complex necessary for El function. However, the relatively large amount of E2 present means that even after the El has been halted, there may still be some ubiquitinated E2 present capable of p27 biotinylation leading to a 'lag' period after the stop has been added and an artificially high signal at each time point.

In order to assess the reproducibility of the final assay conditions, a Z' assay was carried out (figure 29) - 64 wells (1 sector) of positive controls and 64 wells of negative (no ubiquitin) controls. The experiment was performed 3 different times to assess the reproducibility of the data. In each case the signal strength, signal/background ratio, positive control %CV and Z' values were within acceptable limits (Z'> 0.6 (Zhang, Ji-Hu, Journal of Biomolecular Screening, Vol. 4, No. 2, 67- 73 (1999)), %CV < 15%).

Inhibitor studies

Each of the assays in the p27 ubiquitination cascade (El, E2 and E3) were tested against the El inhibitor PYR-41. No pre-charging of either the El or E2 was carried out prior to the introduction of the compound although in the E3 assay, p27 phosphorylation by CDK2/Cyclin E was carried out in the absence of inhibitor.

IC50 values of 7.8 and 3.2 μΜ were generated for the El and E2 assays respectively - Comparable to the reported value of ~ 5 μΜ (Calbiochem). However, a significantly less potent IC₅₀ of 55 μΜ was seen for the E3 assay (figure 30 (a) and (b)). In order to test the feasibility of screening the E3 assay against a range of inhibitors, the assay was tested against a panel of 41 inhibitors (40 known kinase inhibitors and PYR-41) (figure 31). As expected, the PYR-41 showed a high degree of inhibition (>80%), and several other inhibitors also indicated moderate amounts of inhibition (Rottlerin, EGCG and K252c) (figure 32 and 33). Using the same assay conditions, the E3 assay was tested against these inhibitors (figure 33). Both the Rottlerin and the EGCG inhibitors give IC₅₀ values in the μΜ range. The K252c inhibitor would appear to be a false positive result. One potential concern was that the inhibition was the result of inhibition of the CDK2/Cyclin E (and hence phosphorylation of the p27), despite the fact that phosphorylation was carried out in the absence of inhibitor. Comparison of the inhibition profiles for both the E3 assay and Cdk2/Cyclin E shows that this is not the case (figure 34).

Non-specific binding assays

Non-specific binding has previously been identified as a potential problem for both the biotinylated-ubiquitin on its own and for the El enzyme in the E2 assay. The NSB signal for the ubiquitin was minimized by limiting the amount of ubiquitin in the assay to 2μΜ. Similarly, limiting the concentration of El enzyme in both the E2 and E3 assays to 5 n allows decent E2 specific signal whilst reducing the non-specific signal from the El (figure 35).

Similar experiments have shown that non-specific binding signal is not a problem in the E2 assay. This indicates that under the E2 assay conditions neither the E2 (at any concentration), El (at 5 nM) or the ubiquitin (at 2 μΜ) contributes to a non-specific signal (figure 36).

Similar experiments have also been carried out for the E3 assay (figure 37). We again see a non-specific signal. However, our previous experiments have indicated that this signal is not the result of the El (at 5 nM), the ubiquitin (at 2 μΜ) or the E2 at any concentration. We must infer therefore that the NSB signal is the result of NSB from either the E3 or from the p27 complex. Further experiments have indicated that it is likely to be NSB of the p27 substrate itself. As a result, the non-specific signal is not an issue when testing the E3 assay.

Multiple approaches have been employed to minimize the proportion of non-specific signal (figure 38). HTRF Format - E3 Ubiquitin Li ase Assay

Three different epitope tags were initially employed to probe the sCF^skp2/Cksl assay. The following tagged combinations of E2 and p27 were tested in an E3 assay (Table 2).

Table 2: Taezed combinations of E2 and p27 tested in E3 assay

The results show the variation in signal with p27 concentration for each of the combinations tested, read after 20 hr on a BMG PHERAstar. An increase in signal was detected with increasing concentration of p27 (figure 39). Phospho-p27 ubiquitination and hence signal strength should be maximised following a 20 hr incubation as previously identified in an anti-phospho-p27 Western blot (LB1805p74- 77). The highest signal achieved was from the FLAG-E2 / HA-p27.

Following the identification of best 'tag' pairings, p27, E2 and His-tagged protein signal strength was determined in an E3 assay in the presence or absence of E3 tetramer and phospho-p27 (figure 40).

In the presence of E3 tetramer, a decrease in His-tagged protein and E2 signal strength was observed that did not correspond with an increase in p27 signal strength. If phospho-p27 was excluded from the assay, but E3 tetramer was still present, a decrease in E2 and His-tagged protein signal strength was still observed.

The decrease in E2 signal strength suggests the transfer of Ub-B to another component in the scF^{skp2 Cksl} ubiquitin ligase complex or to the substrate - phospho- p27. Detection may not be possible through steric hindrance by other complex components, tag combinations on the E2 and p27 constructs, or the distance between the fluorophores may be just too great to form a complex.

Continuing with the E3 assay optimization, FLAG-E2 and HA-p27 was tested alongside the HA-E2 and c-Myc-p27 via a phospho-p27 vs. E3 tetramer checkerboard to establish a p27 signal.

As you titrate in more E3 tetramer at increasing concentrations of phospho-p27, a bell shaped relationship was exhibited for p27 signal strength with varying E3 tetramer concentration (figure 41). No difference in trend was observed between the combinations. However as higher p27 signal was being obtained with the FLAG-E2 and HA-p27 combination, this combination was used for subsequent HTRF experiments. Note that the signal for p27 detection is still not sufficient to enable a reliable assay.

Further optimization of the E3 assay was investigated through an ubiquitin titration in the presence or absence of E3 tetramer. No improvement in p27 signal strength was observed as you titrate in more ubiquitin (figure 42).

HTRF/ECL comparison

Next, the extent of p27 signal detection in the HTRF and ECL platforms were compared in an E3 assay via an E3 titration (figure 43).

A hyperbolic increase in signal generated on the ECL platform with increasing E3 tetramer concentration. A linear relationship between the signal and the E3 tetramer concentration was observed up to 25 nM E3 tetramer, with a signal plateau after 50 nM. No significant p27 signal was observed on the HTRF platform, giving a clear indication of problems with the detection capabilities of the HTRF platform.

To improve the detection capabilities, the effect of sodium chloride on scF^{Skp2 Cksl} complex disruption and p27 signal was investigated in an E3 assay, with or without E3 tetramer (figure 44). Signal interference from branched or long p27 poly-ubiquitin chains was also determined by optimising acceptor fluorophore concentration (SA XLent). Improving detection capabilities

The introduction of NaCl had a negative impact on signal strength. Increasing the concentration of acceptor fluorophore enhances signal strength, but above 1.0 μg/ml SA XLent the signal to background ratio is not improved. Therefore 1.25 με ml acceptor fluorophore concentration was used in subsequent experiments.

Pre-charging

As increased acceptor fluorophore enhanced signal strength and signal to background ratio, it indicated that potentially long or branched poly-ubiquitin chains were fonriing on p27. Proximity distance between the donor and acceptor fluorophores influences signal strength. Therefore, if a pre-charge of E1+E2 with biotinylated ubiquitin was initially carried out, followed by an E3 assay in the presence of naked ubiquitin, it was hypothesised that only the ubiquitin molecules at the start of the poly-ubiquitin chain would be biotinylated. Hence SA XLent (acceptor fluorophore) would bind to these hereby reducing the distance to its donor fluorophore.

Therefore, a pre-charge of E1+E2 with varying concentrations of Ub-B, followed by an E3 assay in the presence of naked ubiquitin (so the final ubiquitin concentration is constant) were investigated (figure 45).

Pre-charging E1+E2 in the presence of varying concentrations of Ub-B had no effect on p27 signal strength or signal to background ratios.

As an E1+E2 pre-charge with Ub-B had no effect, polyubiquitin interchain FRET formation was re-investigated using streptavidin labelled with either Europium cryptate or APC in equimolar concentrations, as well as p27 signal, in an E3 assay via a Cksl vs. phospho-p27 checkerboard (figure 46).

The variation in interchain and p27 signal for each Cksl concentration tested is shown below. Both exhibited a bell shaped relationship between the concentration of phospho-p27 and signal. The ratios obtained in the interchain detection were low due to the high europium signal. However, on analysing the APC component, signals of 8000 - 10000 RFU were observed but the trend to the europium/APC ratio didn't alter. Therefore, optimisation of the SA-K and SA XLent concentrations in the detection mix was investigated to give the best signal ratio possible for interchain detection. Optimising [SA-K1 and [SA XLent] for interchain detection

Optimisation of the SA-K and SA XLent concentrations for interchain detection was investigated in a E3 assay in the presence or absence of E3 tetramer and Ub-B; at two Ub-B concentrations. In this assay a different form of FRET architecture was investigated were the donor fluorophore was attached to biotinylated ubiquitin and the acceptor fluorophore was on the epitope tag (figure 47).

1 μΜ Ub-B over 2 μΜ Ub-B improves signal strength and using a No Ubiquitin control the signal to background ratio improves to 28 (+ E3 / No Ub-B). However, when the E3 tetramer is excluded from the assay the signal to background ratio falls to 1.2 (+ E3 / No E3) indicating that SA-K and SA XLent are potentially binding to other ubiquitinated components (El / E2 / Skp2 in E3 tetramer/ free Ub-B) in the sandwich complex, and due to their close proximity are able to form a FRET complex thus impacting falsely on the overall signal and ratio.

Investigating non-specific signals

Since SA-K and SA XLent fluorophores could be binding to other ubiquitinated components, breakage of ubiquitin bonds specific to El, E2 and E3, over ubiquitin inter-bondage in the polyubiquitin chain may improve the non-specific signal seen in a no E3 control.

DTT was investigated as a reducing agent to change the redox environment of the diluent, thereby attempting to specifically break thiol ester bonds between the C- terminal carboxyl group and an active site cysteine on the El and E2 but not the isopeptide bond between the carboxyl group of Gly⁷⁶ on one ubiquitin molecule and the ε-amino group of Lys⁴⁸ on another ubiquitin molecule. An E3 assay was carried out in the presence or absence of E3 tetramer (figure 48).

Changing the redox environment through the addition of DTT to the diluent had no effect on interchain and p27 signal detection and signal to background ratios.

Effect of anti-phospho-p27 dilution and [IgG-K] on Ubiquitinated p27 signal detection

Since the substrate for the sCF^{Skp2 Cksl} assay is p27 phosphorylated on Thr¹⁸⁷, a bridging approach was used to bring the donor and acceptor fluorophores closer together through the use of a rabbit-derived anti-phospho-p27 (Thr 187) antibody (Millipore 06-996). Formation of the FRET complex will be completed through a donor fluorophore of IgG-Cryptate and an acceptor fluorophore of SA XLent. An E3 assay was carried out in the presence or absence of E3 tetramer to investigate this approach (figure 49).

Using a bridging phospho-specific antibody has no effect on p27 signal, when coupled to a species specific IgG-Europium cryptate and APC. Therefore, in the presence of a large sandwich complex it appears that other complex components are somehow hindering association of the antibody and fluorophores to the ubiquitinated phospho- p27, or that the distance between the fluorophores still remains too great.

Knock-out experiments

To confirm that detection issues surround the HTRF platform, a knock-out experiment was carried out where each component in the complex was systematically removed and samples were detected in parallel on the HTRF and MSD ECL platforms (figure 50). As a positive control, optimised samples used to calculate Z' values for the MSD platform assay were also analysed.

No differences were detected in the knock-out samples, apart from the No Ubiquitin controls indicating that in the presence of these complexes non-specific binding of the Streptavidin-Europium cryptate and APC is occurring which is masking any p27 poly- ubiquitinated interchain signal. When these samples are detected in parallel on the MSD ECL platform it is very clear that ubiquitination of p27 is occurring under the stated conditions but due to proximity fluorophore distances being too great or the solution format, it is not possible to detect this ubiquitination using the HTRF platform.

The best signal to noise background obtained using an epitope tagged Europium cryptate and an acceptor APC fluorophore was 2 to 1 irrespective of sample origin.

Claims (31)

Claims

1. A method of assaying ubiquitination in a sample comprising:

a) combining in a sample under conditions suitable for ubiquitination to take place:

(i) ubiquitin;

two or more of:

(ii) El;

(iii) E2;

(iv) E3; and

(v) a substrate protein;

b) exposing the sample with a labelled binding partner which is specific for the ubiquitin;

c) measuring the amount of labelled ubiquitin bound to any one of components (ii) to (v) in the sample

wherein one or more of components (ii), (iii), (iv), or (v) in the sample comprises an immobilisation tag which facilitates its immobilisation onto a solid surface.

2. The method according to claim 1, wherein all of components (ii), (iii), (iv) and (v) are combined in the sample.

3. The method according to claim 1 or claim 2, wherein more than one of components (ii), (iii), (iv) or (v) in the sample comprise an immobilisation tag, and each tag is different.

4. The method according to claim 3, wherein two, three or all of components (ii), (iii), (iv) or (v) in the sample are tagged and each tag is different.

5. The method of claim 1 or claim 2 wherein component (ii) in the sample comprises an immobilisation tag.

6. The method of any one of claims 1 to 5 wherein component (iii) in the sample comprises an immobilisation tag.

7. The method of any one of claims 1 to 6 wherein component (iv) in the sample comprises an immobilisation tag.

8. The method of any one of claims 1 to 7 wherein component (v) in the sample comprises an immobilisation tag.

9. The method according to any preceding claim, wherein the ubiquitin comprises a tag.

10. The method according to claim 9, wherein the ubiquitin tag is recognised by the labelled binding partner.

11. The method according to claim 10, wherein the ubiquitin tag is biotin.

12. The method according to claim 9, wherein the ubiquitin tag is capable of interacting with the labelled binding partner to produce a detectable signal.

13. The method according to claim 12, wherein the ubiquitin tag and labelled binding partner form a FRET pair.

14. The method according to any one of claims 1 to 11, wherein the labelled binding partner is capable of emitting light when electrochemically stimulated.

15. The method according to claim 14, wherein the label is selected from the group consisting of 3-aminophthalydrazide; a ruthenium chelate and an acridan ester.

16. The method according to claim 15, wherein the label is tris(2,2'-bipyridyl) ruthenium (II).

17. The method according to any one of the preceding claims, wherein one or more of components (ii), (iii), (iv) or (v) are immobilised on a solid surface.

18. The method according to claim 17, wherein component (ii) is immobilised on a solid surface.

19. The method according to claim 17 or 18, wherein component (iii) is immobilised on a solid surface.

20. The method according to any one of claims 17 to 19, wherein component (iv) is immobilised on a solid surface.

21. The method according to any one of claims 17 to 20, wherein component (v) is immobilised on a solid surface.

22. The method according to claim 17, wherein only one of components (ii), (iii),

(iv) or (v) are immobilised on a solid surface.

23. The method according to any one of claims 17 to 22, wherein the solid surface is an electrode.

24. The method according to any one of claims 17 to 23, wherein the solid surface comprises a binding member which is specific for one of the immobilisation tags.

25. The method according to claims 17 to 24 which comprises an additional wash step, (b^'), which occurs between steps (b) and (c) and removes unbound components from the solid surface.

26. The method according to any one of the preceding claims, wherein the immobilisation tag(s) present on one or more of components (ii), (iii), (iv) or

(v) are selected from the group consisting of a his tag; GST tag; HA tag; c- Myc tag and FLAG tag.

27. The method according to any one of the preceding claims, wherein step (a) further comprises combining a potential modulator of ubiquitination.

28. The method of claim 1 wherein step (c) is performed two or more times in order to measure the amount of labelled ubiquitin bound to two or more of components (ii) to (v) in the sample.

29. The method of claim 28 wherein step (c) wherein is performed twice in order to measure the amount of labelled ubiquitin bound to two of components (ii) to (v) in the sample.

30. The method of claim 28 wherein step (c) wherein is performed three times in order to measure the amount of labelled ubiquitin bound to three of components (ii) to (v) in the sample.

31. The method of claim 28 wherein step (c) wherein is performed four times in order to measure the amount of labelled ubiquitin bound to all of components (ii) to (v) in the sample.