EP0918990A1

EP0918990A1 - An assay which measures phosphatidylinositol 3,4,5 trisphosphate mass at high sensitivity

Info

Publication number: EP0918990A1
Application number: EP97927287A
Authority: EP
Inventors: Peter Charles University of Dundee DOWNES; Ian Howard University of Dundee BATTY; Jeroen University of Dundee VAN DER KAAY
Original assignee: University of Dundee
Current assignee: University of Dundee
Priority date: 1996-06-21
Filing date: 1997-06-23
Publication date: 1999-06-02
Also published as: JP2000513091A; GB9613107D0; AU3183697A; WO1997049990A1

Abstract

A method of assaying phosphatidylinositol 3,4,5 trisphosphate (Ptd Ins(3,4,5)P3) in a sample, the method comprising the steps of: (a) converting phosphatidylinositol 3,4,5 trisphospate present in the sample to inositol 1,3,4,5 tetrakisphosphate (Ins(1,3,4,5)P4); and (b) measuring the amount of inositol 1,3,4,5 tetrakisphosphate. Conveniently, the amount of inositol 1,3,4,5 tetrakisphosphate is measured by binding to a polypeptide which binds inositol 1,3,4,5 tetrakiphosphate selectively. Preferably, the phosphatidylinositol 3,4,5 trisphosphate is converted to inositol 1,3,4,5 tetrakisphosphate by alkaline hydrolysis.

Description

AN ASSAY WHICH MEASURES PHOSPHATIDYLINOSITOL 3.4.5 TRISPHOSPHATE MASS AT HIGH SENSITIVITY

Phosphatidylinositol 3,4,5 trisphosphate is a second messenger which is the principal product of the receptor-regulated enzyme, PI 3-kinase. The role of PI 3-kinase in cell signalling has been well established over the past years. This enzyme phosphorylates phosphoinositides on the 3 position of the inositol ring. In vitro, phosphatidylinositol 3-kinase uses phosphatidylinositol, phosphatidylinositol4P and phosphatidylinositol (4,5)P₂ as substrate. However Stephens et al (1991) Nature 351, 33-39 have shown that in vivo phosphatidylinositol(4,5)P₂ is phosphorylated to phosphatidylinositol 3,4,5 trisphosphate which is subsequently dephosphorylated to phosphatidylinositol(3,4)P₂. The transient phosphatidylinositol 3,4,5 trisphosphate response in stimulated cells and the fact that there are no known pic (phospholipase C) enzymes that can metabolise this lipid lead to the hypothesis that this lipid is a second messenger itself. At present, however, there are only a few reports on potential effector enzymes for phosphatidylinositol 3,4,5 trisphosphate.

Measurement of phosphatidylinositol 3,4,5 trisphosphate levels has so far been done by labelling cells with [³H] inositol or [³²P]orthophosphate. Cell-labelling with inositol (Ins) is often not very efficient and labelling to isotopic equilibrium takes days. Labelling cells with [³²P]orthophosphate is inconvenient because of the precautions needed for the large amounts of radioactivity required. Both labelling procedures suffer from the drawback of elaborate, expensive and time-consuming high-performance liquid chromatography (HPLC) analysis.

There remains the need for an improved assay for phosphatidylinositol 3,4,5 trisphosphate. A straightforward, highly specific and sensitive method has therefore been developed as described in more detail below which allows the detection of picomolar amounts of phosphatidylinositol 3,4,5 trisphosphate. The assay is based on the cleavage of the polar headgroup of phosphatidylmositol 3,4,5 trisphosphate to produce Ins(l,3,4,5)P₄ (inositol 1,3,4,5 tetrakisphosphate) of which the mass can be measured by binding to a Ins(l,3,4,5)P₄-binding compound, for example using an isotope dilution assay. The preparation of suitable reagents, including a highly specific Ins(l,3,4,5)P₄ binding protein (which can, for example, be obtained from mammalian cerebellum or by recombinant DNA technology) and a high specific activity Ins(l,3,4,5)P₄ radioisotope (prepared from Ins(l,4,5,)P₃ and [γ-³²P]adenosine triphosphate using a recombinant Ins(l,4,5)P₃ 3- kinase from rat brain (kind gift C. Erneux)) or the use of [³H]Ins(l ,3,4,5)P₄ are described, as well as the comparison of phosphatidylinositol 3,4,5 trisphosphate mass measurements with cell labelling studies in different cell-types. It has been established that measurements of phosphatidylinositol 3,4,5 trisphosphate by the methods of the invention, including the isotope dilution assay, match well with measurements obtained by cell-labelling procedures.

An advantage of the described method lies in the fact that it is very sensitive, simple, cheap and rapid (routinely 200 samples can be processed in one day) compared with cell-labelling procedures. Moreover, it allows measurements of phosphatidylinositol 3,4,5 trisphosphate in tissues and cells which are not suitable for cell-labelling. Indeed the first measurements of phosphatidylinositol 3,4,5 trisphosphate were performed in rat-skeletal muscle tissues after insulin stimulation in vivo.

A first aspect of the invention provides a method of assaying phosphatidylinositol 3,4,5 trisphosphate (Ptd Ins(3,4,5)P₃) in a sample the method comprising the steps of:

(a) converting phosphatidylinositol 3,4,5 trisphospate present in the sample to inositol 1,3,4,5 tetrakisphosphate (Ins(l,3,4,5)P₄); and

(b) measuring the amount of inositol 1,3,4,5 tetrakisphosphate.

Typically, the sample is a sample of lipids extracted from the cell in which the level of PtdIns(3,4,5)P₃ is to be measured. Lipids may be extracted using any suitable methods such as those described in the Examples. By "suitable method" we mean that the method is compatible with the assay methods disclosed herein. Suitably, the cell is a cell in culture or it may be a cell from any suitable organism. Typically, the cell is a mammalian cell in culture and the method of the invention is used to measure PtdIns(3,4,5)P₃ in the cell following stimulation of the cell in a suitable way. Thus, the method will find uses as a research tool in studying second messenger pathways and the like.

Preferably, the phosphatidylinositol 3,4,5 trisphosphate present in the sample is converted to Ins(l,3,4,5)P₄ by alkaline hydrolysis, most preferably by using a solution of potassium hydroxide at 100°C although any suitable alkali solution may be used. The amount of potassium hydroxide (or other suitable alkali reagent) and the conditions of hydrolysis are chosen by reference to the amount of lipids. Suitable amounts and conditions are described in the Examples and can, in any case, be readily ascertained by the person skilled in the art using the information contained herein.

We have found that alkaline hydrolysis of phosphatidylinositol 3,4,5 trisphosphate reproducibly produces around 11 % of Ins(3,4,5)P₃, 4% of

GroPIns(3,4,5)P₃, 62% of Ins(l,3,4,5)P₄ and 24% of Ins(2, 3,4, 5)P₄ under the conditions described. Thus, the major product of alkaline hydrolysis of phosphatidylinositol 3,4,5 trisphosphate is inositol 1 ,3,4,5 tetrakisphosphate.

Preferably, the pH of the hydrolysate formed following alkaline hydrolysate is adjusted by addition of a suitable acid or buffer so mat the final pH of the solution containing the hydrolysate is compatible with the binding of inositol 1,3,4,5 tetrakisphosphate to the binding compound. Preferably, the pH is adjusted to a pH which does not denature a Ins( 1 ,3 ,4,5)P₄-selective binding protein. Suitable methods of adjusting the pH and identifying a suitable pH are readily achieved by the person skilled in the art without inventive effort.

Preferably, the inositol 1,3,4,5 tetrakisphosphate is measured by binding it to a compound which selectively binds inositol 1 ,3,4,5 tetrakisphosphate; typically the amount is determined by competition with or displacement of a known amount of radiolabelled Ins(l ,3,4,5)P₄ or Ins(l ,3,4,5)P₄ labelled in any other detectable way.

By "selectively binds inositol 1,3,4,5 tetrakisphosphate" we include the meaning that the compound binds inositol 1,3,4,5 tetrakisphosphate in preference to other mositol-contaiiiing compounds produced upon alkaline hydrolysis of a lipid sample from a cell. In particular, it is preferred if the compound which selectively binds inositol 1 ,3,4,5 tetrakisphosphate at least 50-fold more strongly than it binds Ins(l,4,5)P₃ (which is derived from the alkaline hydrolysis of phosphatidylinositol(4,5)P₂). It is preferred if the compound binds Ins(l ,3,4,5)P₄ at least 100-fold more strongly than Ins(l ,4,5)P₃, more preferably at least 500-fold more strongly, still more preferably at least 1000-fold, yet still more preferably at least 2000-fold and in further preference at least 4000-fold (or more) more strongly. Preferably, the compound which selectively binds inositol 1,3,4,5 tetrakisphosphate is a binding protein from mammalian cerebellum, for example rat or sheep cerebellum. A suitable such protein is described in Reference 7, incorporated herein by reference.

The inclusion of eώylenediamine tetraacetate (EDTA) in the preparation buffer of the Ins(l,3,4,5)P₄-selective binding protein may improve the selectivity of the protein for Ins(l,3,4,5)P₄ over Ins(l,4,5)P₃ and so it is preferred although not essential if the Ins(l,3,4,5)P₄-binding protein is prepared in the presence of EDTA. It is also preferred but not essential if the binding of Ins(l,3,4,5)P₄ to the said binding protein is in the presence of EDTA.

A further preferred Ins(l,3,4,5)P₄-binding protein is the InsP₄-GAP protein that selectively binds Ins(l,3,4,5)P₄ and which has been purified to homogeneity from pig platelet membranes. The cDNA encoding the InsP₄-GAP protein has been cloned (see Cullen et al (1995) "Identification of a specific Ins(l,3,4,5)P₄-binding protein as a member of the GAP1 family" Nature 376, 527-530, incorporated herein by reference, and Cullen et al (1995) Biochem. J. 305, 139-143, incorporated herein by reference). Since the cDNA encoding this protein has been cloned it is possible to produce large quantities of substantially pure protein by recombinant DNA techniques which are well known in the art.

It will be appreciated that other Ins(l,3,4,5)P₄-selective binding proteins may be discovered and that they may also be useful in the method of the invention. In particular, different isoforms of the known Ins(l,3,4,5)P₄- selective binding proteins may be found and, indeed, it may be the case that the cerebellum Ins(l,3,4,5)P₄-selective binding proteins may be isoforms of InsP₄-GAP. It will be appreciated that the mammalian cerebellum Ins(l,3,4,5)P₄- selective binding proteins may also be produced by recombinant DNA technology once the cDNAs have been cloned.

Although it is preferred if the Ins(l,3,4,5)P₄-selective binding proteins have the structure as found in nature, it will be readily appreciated that the binding proteins may be modified, for example by point mutation, deletion, insertion or fusion and that they are nevertheless useful in the practice of the invention provided that they retain their Ins(l,3,4,5)P₄- selective binding ability.

In a preferred embodiment the Ins(l,3,4,5)P₄-binding compound is a fusion between InsP₄-GAP and a moiety which aids its purification and/or allows it to be bound to a solid substrate. For example, fusion with glutathione-S-transferase allows the fusion protein to be bound to glutathione and therefore readily purified by glutathione affinity chromatography. Other binding tags are known in the art such as the Myc tag which is recognised selectively by an anti-Myc monoclonal antibody and the His_n tag which bind Ni²⁺.

The amount of Ins(l,3,4,5)P₄ generated or present in a sample may be measured by binding to the binding compound by any suitable method. Preferably, the amount of Ins(l,3,4,5)P₄ is measured using an isotope dilution assay in which there is competition between radiolabelled Ins(l,3,4,5)P₄ added to the system and the unlabelled Ins(l,3,4,5)P₄ derived from the PtdIns(3,4,5)P₃ in the cell. Typically, the radiolabelled Ins(l ,3,4,5)P₄ is labelled with ³²P or ³³P but it may be labelled with any suitable isotope such as ³H or ¹⁴C although ¹⁴C is less preferred.

In a particularly preferred embodiment the amount of Ins(l ,3,4,5)P₄ is measured using a scintillation proximity assay in which there is competition between radiolabelled Ins(l,3,4,5)P₄ added to the system and the unlabelled Ins(l,3,4,5)P₄ derived from the PtdIns(3,4,5)P₃ in the cell.

In this embodiment it is preferred if the compound which selectively binds Ins(l,3,4,5)P₄ is a substantially pure protein, for example InsP₄-GAP. In this way the density of Ins(l,3,4,5)P₄-binding sites proximal to the scintillant in the scintillation proximity assay will be high compared to the use of only a partly pure preparation (for example, the rat cerebellum Ins(l,3,4,5)P₄ binding protem as described above). Conveniently the scintillant is immobilised in or on a suitable substrate (for example, agarose beads or the like) and the Ins(l,3,4,5)P₄-selective binding compound may conveniently be immobilised in or on the scintillant- containing substrate using methods well known in the art. The binding site of the binding compound should be sufficiently close to the scintillant so mat upon the selective binding of radiolabelled Ins(l,3,4,5)P₄ the emission from the radioisotope causes detectable levels of scintillation.

Scmtillation proximity assays are described in more detail in US Patent No 4,568,649 to Amersham International, incorporated herein by reference.

Conveniently the radiolabelled Ins(l,3,4,5)P₄ added is labelled with an isotope which is useful in a scintillation proximity assay. Suitably, the isotope is one with a short path length emission such as ³H or ¹²⁵I and, therefore, the background scintillation from unbound radiolabelled Ins(l ,3,4,5)P₄ is low.

The use of substantially pure Ins(l,3,4,5)P₄ binding protein in a scintillation proximity assay is preferred so that the signal/noise ratio is minimised. A second aspect of the invention provides the use of a compound which selectively binds inositol 1,3,4,5 tetrakisphosphate in an assay for phosphatidylinositol 3,4,5 trisphosphate. The preferred compounds which selectively bind inositol 1,3,4,5 tetrakisphosphate are mose preferred in the first aspect of the invention.

It is preferred but not essential if the inositol 1,3,4,5 tetrakisphosphate- selective binding protein is prepared in the presence of EDTA; it is further preferred but not essential if the binding of inositol 1 ,3,4,5 tetrakisphosphate to the binding protein is in the presence of EDTA.

Before the present invention a compound which selectively binds Ins(l,3,4,5)P₄ has not been used in an assay for PtdIns(3,4,5)P₃.

As noted above in relation to the Ins(l ,3,4,5)P₄ selective binding proteins, especially those from mammalian cerebellum, EDTA may improve the selectivity of binding of Ins(l,3,4,5)P₄, for example when EDTA is used in the preparation of the binding protein.

Thus, a third aspect of the invention provides a method of assaying inositol 1,3,4,5 tetrakisphosphate the method comprising binding said inositol 1,3,4,5 tetrakisphosphate to a protein which selectively binds inositol 1,3,4,5 tetrakisphosphate characterised in that the protem is prepared in the presence of ethylenediamine tetraacetate. Preferably, the assay is carried out in the presence of EDTA. More preferably the protein is a mammalian cerebellum inositol 1,3,4,5 tetrakisphosphate binding protein or a substantially pure Ins(l,3,4,5)P₄-selective binding protein such as InsP₄-GAP.

A fourth aspect of the invention provides the use of ethylenediamine tetraacetate in an assay for inositol 1,3,4,5 tetrakisphosphate.

A fifth aspect of the invention provides the use of ethylenediamine tetraacetate in an assay for phosphatidylinositol 3,4,5 trisphosphate wherein said phosphatidylinositol 3,4,5 trisphosphate is converted to inositol 1,3,4,5 tetrakisphosphate and the amount of inositol 1,3,4,5 tetrakisphosphate is measured.

Suitably, in the fourth and fifth aspects of the invention the assay comprises binding inositol 1,3,4,5 tetrakisphosphate to an mositol 1,3,4,5 tetrakisphosphate binding protein from mammalian cerebellum or to a substantially pure InsP₄-GAP protein.

A sixth aspect of the invention provides a kit of parts for assaying phosphatidylinositol 3,4,5 trisphosphate comprising (1) radiolabelled inositol 1,3,4,5 tetrakisphosphate; and (2) a compound which selectively binds 1,3,4,5 tetrakisphosphate.

Preferably, the kit further comprises non-radiolabelled inositol 1,3,4,5 tetrakisphosphate which, conveniently, is present in a suitable form and in suitable quantised amounts to be used as a standard in order to, for example, generate a standard curve. The substrate of which the compound is associated further comprises scintillant attached to or embedded therein or otherwise associated with the substrate.

A seventh aspect of the invention provides a kit of parts for assaying phosphatidylinositol 3,4,5 trisphosphate comprising (1) radiolabelled inositol 1,3,4,5 tetrakisphosphate; (2) phosphatidylinositol 3,4,5 trisphosphate; and (3) a compound which selectively binds inositol 1,3,4,5 tetrakisphosphate. Preferably, the phosphatidylinositol 3,4,5 trisphosphate is present in a suitable form and in suitable quantised amounts to be used as a standard in order, for example, to generate a standard curve.

It is particularly preferred if the compound which selectively binds inositol 1,3,4,5 tetrakisphosphate in the kits of the sixth and seventih aspects is a substantially pure protein, for example InsP₄-GAP.

In a particularly preferred embodiment the compound which selectively binds to inositol 1,3,4,5 tetrakisphosphate in the kits of the sixth and seventh aspects is bound to a substrate for use in a scintillation proximity assay.

The binding of the Ins(l,3,4,5)P₄-selective compound to the substrate comprising the scintillant may be achieved using methods well known in the art and they will depend on the nature of the substrate. Conveniently, the well known streptavidin/biotin or wheat germ agglutinin systems may be used, although the binding compound may be covalently cross-linked to the substrate using methods known in the art.

Conveniently, the radiolabelled Ins(l,3,4,5)P₄ is suitable for use in a scintillation proximity assay. It is particularly preferred if the radiolabelled Ins(l,3,4,5)P₄ is [³H]Ins(l,3,4,5)P₄.

In a preferred embodiment the kits of parts of the sixth and seventh aspects further comprise means for converting phosphatidylmositol 3,4,5 trisphosphate in a sample to inositol 1,3,4,5 tetrakisphosphate. Suitably the means comprise an alkaline solution. Preferably, the means comprise potassium hydroxide. In a still further preferred embodiment the kit of parts comprises details of the conversion factor for converting phosphatidylinositol 3,4,5 trisphosphate to inositol 1,3,4,5 tetrakisphosphate. As is discussed in more detail elsewhere, we have deteraiined the identity and relative amounts of the products of the alkaline hydrolysis of phosphatidylmositol 3,4,5 trisphosphate. This information is preferably included in the kit of parts in order for the operator of the assay method to be able to determine the PtdIns(3,4,5)P₃ content of the original sample.

A further aspect of the mvention provides a kit of parts for assaying inositol 1,3,4,5 tetrakisphosphate comprising (1) radiolabelled inositol 1,3,4,5 tetrakisphosphate; (2) non-radiolabelled inositol 1,3,4,5 tetrakisphosphate; and (3) a substantially pure protein which selectively binds inositol 1,3,4,5 tetrakisphosphate.

Preferably, said substantially pure protein is InsP₄-GAP. Preferably, the substantially pure protein which selectively binds Ins(l,3,4,5)P₄ is prepared in the presence of EDTA.

In a preferred embodiment the kits of parts of the invention further comprise a solution for binding said inositol 1,3,4,5 tetrakisphosphate to said binding protein wherein said solution comprises EDTA.

A further aspect of the mvention provides a substrate for a scintillation proximity assay comprising a scintillant and a compound which selectively binds inositol 1,3,4,5 tetrakisphosphate.

The substrate may be any suitable substrate for use in a scintillation proximity assay and the scintillant may be any suitable scintillant. Suitable substrates and scintillants are known in the art. The compound which selectively binds inositol 1,3,4,5 tetrakisphosphate is preferably a substantially pure protem; in a particularly preferred embodiment the protein is InsP₄-GAP.

The invention will now be described in more detail with reference to the following Figure and Examples in which:

Figure 1 shows the displacement curves of Ins(l ,4,5)P₃ and Ins(l ,3,4,5)P₄. The Kj for Ins(l,3,4,5)P₄ and Ins(l ,4,5)P₃ was 2.5 and 12500mM, respectively. Thus, Ins(l ,4,5)P₃ was a 5000-fold less potent ligand for the receptor and was therefore excluded as a significant cross-reactant.

Example 1; Measurement of phosphatidylinositol 3.4,5 trisphosphate using an isotope dilution assay and cerebellum Ins(1.3.4.5)P₄-binding protein

Materials

[γ-³²P] Adenosine triphosphate (3000 Ci/mmol) and [³H]Ins(l ,4,5)P₃ (20-60 Ci/mmol) were from Amersham International pic, Amersham Place, Little

Chalfont, Bucks, HP7 9NA, UK. Ins(l,4,5)P₃ and Ins(l,3,4,5)P₄ were from Cell Signals Inc, 4053 Tates Creek Suite 138, Lexington, Kentucky

40517, USA. Adenosine triphosphate (special quality) and hexokinase were from Boehringer Mannheim HTTP://BIOCHEM.BOEHRINGER- MANNHEIM.COM, Bell Lane, Lewes, East Sussex BN7 IL6, UK.

GroPIns(3,4,5)P₃ was prepared from phosphatidylmositol 3,4,5 trisphosphate (generous gift from Roy Gigg) by deacylation for 20 minutes in memylamine (25%)/methanol/butanol (42.8/45.7/11.5 v/v/v) at 53 °C. Phosphatidylinositol 3,4,5 trisphosphate can be obtained using methods known in the art and is believed to be commercially available from Upstate Biotechnology, Inc (UBI). A parallel reaction with [³²P]phosphatidylinositol 3,4,5 trisphosphate internal standard showed a 95% conversion to GroPIns(3,4,5)P₃.

Methods

Partial purification of recombinant rat brain inositol phosphate (Ins(l,4,5)P₃) 3-kinase

Expression and purification of recombinant rat brain Ins(l ,4,5)P₃ 3-kinase was carried out according to (1) with minor modifications. A single E. coli colony containing the Bluescript plasmid with the cloned DNA insert (C5) was grown in LB medium in the presence of 50μg of ampicillin per ml at 37 °C to an A^ of 1.5. After dilution with preheated medium (30 °C) to an A^ of 0.5, the culture was incubated with ImM IPTG for 2 hours at 30°C. Bacteria were harvested (1200g, 15 min) and resuspended in cold lysis buffer (50mM Tris-HCI pH 8.0, ImM EDTA, 0.4mM PMSF, 0.4mM benzamidine, 5μM leupeptin, 5μM peptstatin and calpain inhibitors I and II at 5μg/ml). After sonication in ice, Triton X- 100 was added to a final concentration of 1 % (v/v) and the lysate was shaken for 1 hour at 4°C. After 30 minute centrifugation at 15,000g the supernatant was applied to a CaM-Sepharose (CaM is calmodulin) column and eluted as described in (1). The active fraction was concentrated up to 0.5mg/ml using an Amicon centriprep 30 filter.

Preparation of [3-³²P]Ins(l,3,4,5)P₄

[3-³²P]Ins(l ,3,4,5)P₄ was prepared from Ins(l ,4,5)P₃ and [γ-³²P]adenosine triphosphate using the partially purified recombinant Ins(l ,4,5)P₃ 3-kinase described above. Overnight incubations were in lOOμl at room temperature and contained lOOμM Ins(l,4,5)P₃, 20mM magnesium chloride, 84mM Hepes (pH 7.5), l .lmM calcium chloride, ImM EGTA, lμM CaM (CaM is calmodulin), lmg/ml BSA and 500μCi [γ- ³²P]adenosine triphosphate (3000 Ci/mmol). Excessive [γ-³²P]adenosine triphosphate was removed by a subsequent hexokinase reaction using D- glucose as ³²P0₄ acceptor after which the resulting [³²P0₄]D-glucose 6- phosphate was removed by dialysis as described in (2). Briefly, after addition of 2μl of 5mM glucose and lμl hexokinase (1.5U) a 30-minute incubation at room temperature was terminated by addition of 0.8ml lOmM EDTA acid and 2-minute boiling. The sample was centrifuged for 1 minute at 13,000g and the supernatant was dialysed 3 times for 45 minutes against 600ml of lOmM HEPES pH 7.0. The retentate was then applied to an HPLC partisphere-SAX column and eluated using a gradient made of water and 1.0M NH₄H₂PO₄ pH 3.8 with the following breakpoints: t=0 0% , 2 min 0%, 5 min 75 % , 10 mm 75 % , 40 min 100% and 48 min 100% . Fractions of 1.75ml were collected and Cerenkov counting revealed the final yield of [3-³²P]Ins(l,3,4,5)P₄.

Alternative preparation of [3- P]Ins(l,3,4,5)P₄

[3-³²P]Ins(l ,3,4,5)P₄ was prepared from Ins(l,4,5)P₃ and [7-³²P] ATP using recombinant Ins(l ,4,5)P₃ 3-kinase partially purified as described above. A reaction mixture of 200 μl, containing 100 μM Ins(l,4,5)P₃, 20 mM MgCl₂, 50 mM Tris-HCl (pH 7.5), 1.018 mM CaCl₂, 1 mM EGTA, 10 μM CaM, 1 mg/ml bovine serum albumin, 1 mCi of [γ-³²P]ATP (3000 Ci/mmol), and 20 μl of enzyme, was incubated for 1 h at 37 °C, and the reaction was terminated by the addition of 0.8 ml of 10 mM EDTA followed by 2 min of boiling. The sample was applied to a HPLC Partisphere-SAX column, eluted with a nonlinear gradient made of water and 1.0 M NH₄H₂P0₄ (pH 3.8). Fractions (2 ml) containing 13- ³²P]Ins(l,3,4,5)P₄ were identified by Cerenkov counting, pooled, and dialysed three times against 1000 ml of water for 45 min (the desalting of inositol polyphosphates by dialysis is described in more detail in Ref 2). This effectively removed the inorganic phosphate which would otherwise interfere in the Ins(l,3,4,5)P₄ binding assay. The recovery of [³²P]Ins(l,3,4,5)P₄ following dialysis was approximately 50% . Starting with 100 μM Ins(l ,4,5)P₃ and 1 mCi of [7-³²P]ATP ( -3000 Ci/mmol), 0.5 mCi of [³²P]Ins(l,3,4,5)P₄ with a specific radioactivity of - 3000 Ci/mmol was routinely produced.

Cell culture and lipid extraction

1321N1 Astrocytoma cells were grown as described previously (3). Petri dishes (60mm X 25mm) with cells grown to confluence were washed twice with Krebs Ringer HEPES (KRH) buffer (or preferably modified Krebs Henseleit buffer) and incubated for 30 minutes in KRH buffer at 37°C. Cells were stimulated with insulin at lOμg/ml for 7.5 minutes or with thrombin peptide at O.lmM for 30 seconds after which the medium was aspirated and cells were quenched with 1ml 10% TCA. After 15 minutes on ice dishes were scraped and washed once with 0.5ml 10% TCA. After 5 minutes of centrifugation at 13,000g the pellet was washed twice with 0.75ml TCA/EDTA (5% ImM). Lipids were extracted as described by Folch et al (5). Briefly, after extraction for 20 minutes on ice in 0.75ml of CHCl₃/MeOH/HCI (40/8/1: v/v/v) phases were split by addition of 0.25ml CHCI₃ and 0.45ml 0.1 M HCI. After a 1 minute centrifugation at 13,000g, the lower phase was transferred to a screw-cap tube and the upper phase was reextracted once with 0.45ml synthetic lower phase. The lower phases were pooled. Samples were dried down, resuspended by vortexing in 50μl 1.0M potassium hydroxide and boiled for 30 minutes. After neutralisation with HAc, the fatty acids were removed by extraction with butan-1-ol/petroleum ether (40- 60°C)/ethylacetate (20/4/1: v/v). Samples were dried down and resuspended in water or dilute acetic acid so that the pH is about 5 and aliquots were directly used in the radioreceptor assay.

Ins(l,3,4,5)P₄ radioreceptor assay

The binding protein for Ins(l,3,4,5)P₄ was obtained from rat cerebellum (and from sheep cerebellum which has the same features). The Ins(l ,3,4,5)P₄ binding protein was prepared following (4) using an altered homogenisation buffer: 20mM sodium bicarbonate pH 8.0, ImM DTT, 2mM EDTA, 0.5 mM PMSF and 0.5mM benzamidine. The inclusion of EDTA in the homogenisation buffer appeared to increase the selectivity of the Ins(l,3,4,5)P₄ receptor for Ins(l,3,4,5)P₄ over Ins(l,4,5)P₃.

Ins(l,3,4,5)P₄ concentration was deteimined as in (4) which is an adaptation of (6). Samples of 40μl were added to 80μl of 50mM NaAc, 50mM KH₂P0₄ pH 5.0, 2mM EDTA and 100000 dpm [³²P]Ins(l ,3,4,5)P₄. After 30 minutes incubation on ice with 40μl cerebellar protein (5mg/ml) rapid filtration using GF/C filters removed the unbound [³²P]Ins(l,3,4,5)P₄. Filters were washed three times wim 3ml of ice cold 25mM NaAc, 25mM KH₂P0₄, pH 5.0, ImM EDTA and 5mM sodium bicarbonate. Filters were extracted with 4ml scintillant and radioactivity was counted after over 2 hours. Results

[³²P]Ins(l,3,4,5)P₄ yield and specific activity

Routinely, a 50% yield of [³²P]Ins(l,3,4,5)P₄ with a specific radioactivity of - 3000 Ci/mmol was obtained starting with a 100 micromolar Ins(l ,4,5)P₃ concentration and 500μCi[7-³²P] adenosine triphosphate (3000 Ci/mmol). The hexokinase reaction and subsequent dialysis of [³²P]glucose 6-phosphate in order to remove the excessive [γ-³²P]adenosine triphosphate resulted in a loss of approximately 2% of the [³²P]Ins(l,3,4,5)P₄. The [³²P]Ins(l,3,4,5)P₄ in the retentate was HPLC purified as described above and yielded 125μCi [³²P]Ins(l,3,4,5)P₄ in 1.0M NH₄H₂P0₄ pH 3.8 after desalting by dialysis (three times 20 minutes against 750 ml lOmM HEPES pH 7.1).

Recovery of phosphatidylinositol 3,4,5 trisphosphate and the products formed after alkaline hydrolysis

The recovery of phosphatidylinositol 3,4,5 trisphosphate was checked by spiking the lipids during CHCI₃/MeOH/HCl extraction with

[³²P]phosphatidylinositol 3,4,5 trisphosphate made from phosphatidylinositol(4,5) P₂ and [γ-³²P]adenosine triphosphate using a phosphatidylinositol 3-kinase from U937 cells. After a single backextraction of the upperphase with synthetic lowerphase 98% of phosphatidylinositol 3,4,5 trisphosphate was recovered. The results from

HPLC analysis of the products from alkaline hydrolysis of

[³²P]phosphatidylinositol 3,4,5 trisphosphate are shown in table 1. The amount of potassium hydroxide required for efficient hydrolysis was found to be 50μl of IM potassium hydroxide for 30 minutes at 100°C which yielded 62% of Ins(l ,3,4,5)P₄. Although samples contained no more than 0.6mg of lipid and should therefore not have more than 2μmoles of esterbonds (assuming an average MW of 1000 and 4 esterbonds per lipid molecule), lower concentrations of potassium hydroxide resulted in lower yields of Ins(l,3,4,5)P₄ (0.2M and 0.5M potassium hydroxide yielded 29% and 48% respectively). Hydrolysis in 50μl IM potassium hodroxide subsequently neutralised with HAc resulted in 125mM KAc (pH 5.0) final concentration in the Ins(l,3,4,5)P₄ radioreceptor assay. Since the Ins(l,3,4,5)P₄ binding protein tolerates 125mM KAc without loss of sensitivity or selectivity (data not shown), hydrolysis was routinely done with IM potassium hydroxide.

TABLE 1

Products after alkaline hydrolysis of phosphatidylmositol 3,4,5 trisphosphate in l.OM potassium hydroxide at 100°C for 30 minutes.

Selectivity and sensitivity of the Ins(l,3,4,5)P₄ receptor from rat cerebellum

Alkaline cleavage of the polar headgroup of phosphatidylinositol 3,4,5 trisphosphate yields Ins(l,3,4,5)P₄ which can be detected in an isotope dilution assay using the rat cerebellar Ins(l,3,4,5)P₄ receptor and thus provides an indirect measure of the actual phosphatidylinositol 3,4,5 trisphosphate concentration. Sample values were corrected for phosphatidylinositol 3,4,5 trisphosphate recovery (98%) and for the final yield of Ins(l,3,4,5)P₄ (62%) after alkaline hydrolysis (see above).

Another issue to be dealt with was the affinity of the receptor for Ins(l,4,5)P₃ since this is the polar headgroup derived from alkaline hydrolysis of phosphatidylinositol(4,5)P₂, which is present at approximate 1000-fold excess in unstimulated cells. Figure 1 shows the displacement curves of Ins(l,4,5)P₃ and Ins(l,3,4,5)P₄. The K,, for Ins(l,3,4,5)P₄ and Ins(l,4,5)P₃ was 2.5 and 12500mM respectively. So Ins(l,4,5)P₃ was a 5000-fold less potent ligand for the receptor and was therefore excluded as a significant cross-reactant.

The above describes the basis of a radioligand displacement assay capable of measuring phosphatidylinositol 3 ,4 ,5 trisphosphate levels reliably in the subpicomolar range. The key developments which enabled the design of this novel procedure are summarised below.

1. A previously published radioligand displacement assay for Ins(l ,3,4,5)P₄ has been exploited to measure phosphatidylinositol

3,4,5 trisphosphate which is extracted from tissues and cells as an intact lipid and then subjected to alkaline hydrolysis to give

Ins(l,3,4,5)P₄ in high yield. A similar procedure had been published by which a related molecule, phosphatidylinositolP₂ (which is much more abundant than phosphatidylinositolP₃) could be measured following alkaline hydrolysis to InsP₃ and use of a different radioligand binding assay for InsP₃. However, it has been established, unexpectedly, that the present method can be used to give Ins(l ,3,4,5)P₄ in high yield and the yield has been quantified to allow precise calculation of the amount of phosphatidylinositol 3,4,5 trisphosphate in the original extract.

2. In order to use mis method to measure phosphatidylinositol 3,4,5 trisphosphate extracted from cells, the assay needs specifically measure Ins(l,3,4,5)P₄ in the face of a 1000-fold excess of

Ins(l ,4,5)P₃. It was found that this could be achieved most reliably if the Ins(l,3,4,5)P₄ binding protein is prepared in buffers containing ethylenediaminetetra-acetic acid (EDTA) which is a significant difference from the original publication.

3. The use of ethylenediaminetetraacetic acid (EDTA) in buffers used for the preparation of binding protein is desirable but not essential. The combination of two approaches, alkaline hydrolysis of parent lipid to yield the corresponding inositol phosphate and exploitation of a radioligand binding assay for the inositol phosphate is not obvious.

Example 2: Measurement of phosphatidylinositol 3.4.5 trisphosphate using recombinant IP4-GAP

The application of GAP1^1P4BP for Ins(l,3,4,5)P₄ mass determination

A protem that specifically binds Ins(l,3,4,5)P₄ has been purified to homogeneity from pig platelets membranes. Peptide sequencing allowed the design of DNA primers and using the polymerase chain reaction method, the cDNA of the Ins(l ,3,4,5)P₄ binding protein was cloned from a cDNA library. Based on homology the protein was identified as a GTPase activation protein (GAP), specifically a member of the GAP1 family and therefore named GAP1^1P4BP (and here referred to as IP4-GAP). The IP4-GAP is cloned in the pGEX 4T-2 vector as a GST-fusion protein. IP4-GAP is produced following IPTG induction of transformed E. coli BL21 (together with co-expressed chaperones GroES and GroEL introduced by cotransfection with a different plasmid). The full length clone can be used, in our case we use a truncated form of the IP4-GAP, GST ΔC2-GAP1^1P4BP. The glutathione-S-transferase tag allows simple and fast purification from the E. coli ly sates. Cells were pelleted and resuspended in lysis buffer (53 mM Na2HP04, 3 mM NaH2P04, 45 mM NaCl, 1 mM EDTA, 1 mM EGTA and 0.1 % 6-mercaptoethanol) after sonication, Triton X-100 was added to 2% by volume. The IP4-GAP is trapped overnight at 4°C on glutathione-agarose beads which are washed two times in lysis buffer and two times in wash buffer (75 mM Tris (pH 8.0) and 0.3 M NaCl).

The IP4-GAP can subsequently be eluted with four times 0.4 ml elution buffer (75 mM Tris, 20 mM glutathione, 0.1% Triton x-100 and 1.0 M NaCl (readjusted to) pH 8.0). IP4-GAP eluted or coupled to the agarose beads can be stored at -20°C after adding 50% glycerol by volume (but can not be frozen).

Ins(l,3,4,5)P₄ binding assay using the IP4-GAP

The Ins(l,3,4,5)P₄ binding assay is performed as described in Example 1 but now using IP4-GAP coupled to agarose beads as the binding protein. Using the same assay conditions as in Example 1 the binding protein shows similar selectivity for Ins(l,3,4,5)P₄ over Ins(l,4,5)P₃ and also the same sensitivity. The K,, for Ins(l,3,4,5)P₄ is about 3 to 5 nM and the Kj for Ins(l,4,5)P₃ is over 15000 nM. (The original assay conditions by Cullen et al (1995) Biochem. J. 305, 139-143, result in a K^ for the full length protein of approx 10 nM and for the ΔC2-GAP1^{1P4BP1 of approx 30 nM).

The IP4-GAP coupled to agarose beads can be used in the filtration assay to separate bound from unbound ligand. Filtration using the eluted IP4- GAP cannot be applied to separate bound from unbound radioligand. However, eluted IP4-GAP can be precipitated with PEG and γ-globulins and subsequently spun in a microfuge after which the supernatant (with the unbound radioligand) is aspirated.

The above mentioned variations can be performed with [³²P]Ins(l ,3 ,4,5)P₄ (3000 Ci/mmol). The IP4-GAP allows for an important improvement with respect to the ligand, in that commercially available [³H]Ins(l,3,4,5)P₄ (21 Ci/mmol) can now be used. Although the specific radioactivity is about 150-fold lower, the non-specific binding is sufficiently low to obtain a decent signal. 3 μg of IP4-GAP and 40,000 dpm [³H]Ins(l,3,4,5)P₄ per assay are used and give maximum binding of approx 7000 dpm and non specific binding of approx 400 dpm.

The Ins(l,3,4,5)P₄ radioligand displacement assay is thus performed using recombinant IP4-GAP expressed in E. coli and commercially available [³H]Ins(l ,3,4,5)P₄ allowing measurement of PtdIns(3,4,5)P₃ as described above. Moreover, the assay may be readily adapted to a novel assay format which is developed by Amersham and is called scintillation proximity assay (SPA). Beads with trapped scintillant (fluoro- microspheres) are coupled a protein of interest (in this case the IP4-GAP). When this protein interacts with a [³H] or [¹²⁵I] radiolabelled ligand (in our case [³H]Ins(l ,3,4,5)P₄). These ligands are in close proximity with the scintillant of which the subsequent excitation can be measured in a liquid scintillation counter. The use of the InsP₄-GAP recombinant protein allows [³H]Ins(l,3,4,5)P₄ to be used as isotope. This isotope does not require shielding, has a long half-life and is already commercially available. Moreover the method can be modified into a scintillation proximity assay saving time and costs. The principle is that fluoromicrospheres (small beads) with incorporated scintillant can be derivatized (eg. with Ins(l,3,4,5)P₄ receptor) and then bind specific molecules ([³H]Ins)(l,3,4,5)P₄). No filters, vials, scintillant or vacuum manifold are needed in this embodiment. This embodiment is less wasteful, less cosdy and less time consirming. Use of IP4-GAP, which is easily produced and purified from E. coli cultures, allows measurements with commercially available [³H]Ins(l,3,4,5)P₄. The long half-life of the ligand and the fact that no shielding is required are clear improvements. The assay can also be converted into a scintillation proximity assay which has the advantage of low cost (no filters, vials, scintillant, scintillant waste and vacuum-mamfold) and speed (direct transfer to me counter of the vials or even microtitre plates). The production of large quantity of homogeneous binding protein is fast and simple. The separation from bound and unbound can also be conveniently done by centrifugation rather than filtration (this holds true for both [³²P] and [³H]Ins(l,3,4,5)P₄).

REFERENCES

1) Takazawa, K. et al (1990) "Cloning and expression in Escherichia coli of a rat brain cDNA encoding a Ca²⁺/calmodulin sensitive inositol 1 ,4,5-trisphosphate 3-kinase" Biochem. J. Ill, 107-112.

2) Van der Kaay, J. & Van Haastert, P.J.M. (1995) "Desalting inositolpolyphosphates by dialysis" Analytical Biochemistry 225, 183-185.

3) Batty, I.H. et al (1995) "The mechanism of muscarinic receptor- stimulated phosphatidylinositol resynthesis in 1321 Nl astrocytoma cells and its inhibition by Li⁺" J. Neurochem. 65, 2279-2289.

4) Challis, R.A.J. & Nahorski, S.R. (1991) "Depolarisation and agonist-stimulated changes in inositol 1 ,4,5-trisphosphate and inositol 1 ,3,4,5 tetrakisphosphate mass accumulation in rat cerebral cortex" J. Neurochem. 57, 1042-1051.

5) Folch, J. et al (1957) "A simple method for the isolation and purification of total lipids from animal tissues" J. Biol. Chem. 226, 497-509.

6) Donie et al (1990) "High affinity inositol 1 ,3,4,5-tetrakisphosphate receptor from cerebellum: solubilisation, partial purification and characterisation" FEBS Lett. 268, 194-198.

7) Donie & Resier (1989) FEBS Lett. 254, 155-158.

All of these references are incorporated herein by reference.

Claims

1. A method of assaying phosphatidylinositol 3 ,4,5 trisphosphate (Ptd Ins(3,4,5)P₃) in a sample the method comprising the steps of: (a) converting phosphatidylinositol 3,4,5 trisphospate present in the sample to mositol 1,3,4,5 tetrakisphosphate (Ins(l,3,4,5)P₄); and

(b) measuring the amount of inositol 1 ,3,4,5 tetrakisphosphate.

2. A method according to Claim 1 wherein in step (a) phosphatidylinositol 3,4,5 trisphosphate is converted to inositol 1,3,4,5 tetrakisphosphate by alkaline hydrolysis.

3. A method according to Claim 2 wherein the alkaline hydrolysis is carried out in the presence of potassium hydroxide at 100°C.

4. A method according to any one of the preceding claims wherein the amount of inositol 1,3,4,5 tetrakisphosphate is measured by binding said inositol 1,3,4,5 tetrakisphosphate to a compound which selectively binds inositol 1,3,4,5 tetrakisphosphate.

5. A method according to Claim 4 wherein the compound which selectively binds inositol 1,3,4,5 tetrakisphosphate is a binding protein from mammalian cerebellum.

6. A method according to Claim 5 wherein the binding protein is rat cerebellum, or sheep cerebellum, inositol 1,3,4,5 tetrakisphosphate binding protein.

7. A method according to Claims 5 or 6 wherein the amount of inositol 1,3,4,5 tetrakisphosphate is measured by an isotope dilution assay using radiolabelled inositol 1 ,3,4,5 tetrakisphosphate.

8. A method accordmg to Claim 7 wherein the radiolabelled inositol 1 ,3,4,5 tetrakisphosphate is [³²P] labelled.

9. A method according to Claim 4 wherein the compound which selectively binds inositol 1,3,4,5 tetrakisphosphate is a substantially pure inositol 1,3,4,5 tetrakisphosphate-selective binding protem.

10. A method according to Claim 9 wherem the substantially pure inositol 1 ,3,4,5 tetrakisphosphate-selective binding protein is InsP₄-GAP.

11. A method according to Claim 9 or 10 wherein the amount of inositol 1 ,3,4,5 tetrakisphosphate is measured by an isotope dilution assay using radiolabelled inositol 1,3,4,5 tetrakisphosphate.

12. A method according to Claim 10 or 11 wherein the inositol 1 ,3,4,5 tetrakisphosphate is [³²P]-labelled.

13. A method according to Claim 9 or 10 wherein the binding of inositol 1 ,3,4,5 tetrakisphosphate to InsP₄-GAP is measured by a scintillation proximity assay.

14. A method according to Claim 13 wherein the inositol 1 ,3,4,5 tetrakisphosphate is [³H]-labelled.

15. Use of a compound which selectively binds inositol 1,3,4,5 tetrakisphosphate in an assay for phosphatidylmositol 3,4,5 trisphosphate.

16. Use according to Claim 15 wherein the compound is an inositol 1,3,4,5 tetrakisphosphate binding protein from mammalian cerebellum.

17. Use according to Claim 16 wherein said binding protein is rat cerebellum, or sheep cerebellum, inositol 1,3,4,5 tetrakisphosphate binding protem.

18. Use according to Claim 15 wherein the compound is a substantially pure inositol 1,3,4,5 tetrakisphosphate-selective binding protein.

19. Use according to Claim 18 wherein the compound is substantially pure InsP₄-GAP protein.

20. A kit of parts for assaying phosphatidylmositol 3,4,5 trisphosphate comprising (1) radiolabelled inositol 1,3,4,5 tetrakisphosphate; and (2) a compound which selectively binds 1,3,4,5 tetrakisphosphate.

21. A kit of parts for assaying phosphatidylinositol 3,4,5 trisphosphate comprising (1) radiolabelled inositol 1,3,4,5 tetrakisphosphate; (2) phosphatidylinositol 3,4,5 trisphosphate; and (3) a compound which selectively binds inositol 1,3,4,5 tetrakisphosphate.

22. A kit of parts according to Claim 20 or 21 wherein the radiolabelled inositol 1,3,4,5 tetrakisphosphate is labelled with [³H].

23. A kit of parts according to any one of Claims 20 to 22 wherein the compound which selectively binds inositol 1,3,4,5 tetrakisphosphate is a substantially pure protein.

24. A kit of parts according to Claim 23 wherein the substantially pure protein is InsP₄-GAP.

25. A kit of parts according to any one of Claims 20 to 24 wherein the compound which selectively binds inositol 1,3,4,5 tetrakisphosphate is bound to a substrate for use in a scintillation proximity assay.

26. A kit of parts according to any one of Claims 20 to 25 further comprising means for converting phosphatidylinositol 3,4,5 trisphosphate in a sample to inositol 1,3,4,5 tetrakisphosphate.

27. A kit according to Claim 26 wherein said means comprises an alkaline solution.

28. A kit according to Claim 27 wherein said alkaline solution comprises potassium hydroxide.

29. A kit of parts according to any one of Claims 20 to 28 further comprising details of the conversion factor on converting phosphatidylinositol 3,4,5 trisphosphate to inositol 1,3,4,5 tetrakisphosphate by alkaline hydrolysis.

30. A kit of parts for assaying inositol 1,3,4,5 tetrakisphosphate comprising (1) radiolabelled inositol 1,3,4,5 tetrakisphosphate; (2) non- radiolabelled inositol 1,3,4,5 tetrakisphosphate; and (3) a substantially pure protem which selectively binds inositol 1,3,4,5 tetrakisphosphate.

31. A kit of parts according to Claim 30 wherein the said substantially pure protein is InsP₄-GAP.

32. A kit of parts according to Claims 30 or 31 wherein the compound which selectively binds inositol 1,3,4,5 tetrakisphosphate is bound to a substrate for use in a scintillation proximity assay.

33. A substrate for a scmtillation proximity assay comprising a scintillant and a compound which selectively binds inositol 1,3,4,5 tetrakisphosphate.

34. A substrate accordmg to Claim 33 wherem the said compound is a substantially pure protem which selectively binds inositol 1,3,4,5 tetrakisphosphate .

35. A substrate according to Claim 34 wherein said protein is InsP₄- GAP.

36. Any novel method for assaying phosphatidylinositol 3,4,5 trisphosphate or inositol 1,3,4,5 tetrakisphosphate as herein disclosed.

37. Any novel kit of parts for assaying phosphatidylinositol 3,4,5 trisphosphate or inositol 1,3,4,5 tetrakisphosphate as herein disclosed.