WO1991012530A1

WO1991012530A1 - Radiative transfer fluorescence assay

Info

Publication number: WO1991012530A1
Application number: PCT/GB1991/000165
Authority: WO
Inventors: William Arthur Bains; John Andrew Golby
Original assignee: Pa Consulting Services Limited
Priority date: 1990-02-16
Filing date: 1991-02-05
Publication date: 1991-08-22
Also published as: EP0594606A1; JPH05506094A; GB9003593D0; AU7237691A

Abstract

A fluorescence assay for determining the presence in a sample of an analyte or binding partner to which the analyte binds specifically and selectively, comprises contacting the sample with a quantity of binding partner or analyte labelled with a first fluorescent material A and with a quantity of binding partner or analyte labelled with a second fluorescent material B, a unit of substance labelled with A and a unit of substance labelled with B both being capable of binding simultaneously to a single particle of analyte of binding partner or being capable of binding to each other, A and B being selected such that B is capable of being stimulated by radiative transfer fluorescence by light emitted from A when suitably stimulated, and B is not significantly stimulated by light which acts to stimulate A; if necessary adding to the sample a quantity of the analyte or binding partner; and monitoring light emitted from B in response to suitable stimulation of A. A preferred combination is the dye Styryl 9(A) and Nd3+(B) in the form of Nd:glass particles.

Description

Title: RADIATIVE T-RA SFER FLDC_SESCENCE ASSAY

DESCRIPTION

^"Field of the invention

This invention relates to fluorescence assays, using fluorescent materials as labels. The invention is particularly applicable to immunoassays, based on the affinity of antibodies and antigens, but is also applicable to other types of assay using other pairs of reactants that bind specifically and selectively to each other.

Background to the invention

In a typical fluorescence sandwich immunoassay, a sample containing an unknown quantity of an antigen of interest is contacted with the corresponding antibody (monoclonal or polyclonal) immobilised on a solid support, and with a supply of fluorescently labelled antibody to the antigen. Any antigen in the sample will bind to the immobilised antibody and to the labelled antibody, resulting in formation of a fluorescently labelled immobilised "sandwich". After washing to remove any unbound reagents, if necessary, an indication can be obtained of the presence or amount of the antigen by monitoring light emitted by the fluorescent label material in response to

SUBSTITUTE SHEET suitable stimulation.

In conventional assays, one molecule of fluorescent label material is typically associated with each molecule of antibody, so that only a small amount of light is emitted for each molecule of bound antigen, which limits significantly the sensitivity of the technique.

In a typical agglutination assay, a sample containing an unknown quantity of an antigen of interest is introduced into a solution of the corresponding antibody labelled with monodisperse latex spheres in the size range 0.1- 1.0um. Any antigen in the sample will bind to the labelled antibody and multiple binding causes the latex spheres to bind together.

Any binding can be detected by monitoring the optical properties of the solution, specifically the angular distribution of light scattered from an incident beam, this test is homogenous, ie does not require a separate washing stage, but is relatively insensitive because it requires many thousands or millions of binding events.to cause a detectable change in the scattered light.

The present invention aims to provide a homogeneous assay, i.e. one that does not require a washing stage prior to detection. The invention also aims, in preferred embodiments at least, to provide an assay capable of greater sensitivity than conventional agglutination and fluorescence assays.

Summary of the invention

According to the present invention there is provided a

SUBSTITUTESHEET fluorescence assay for determining the presence in a sample of an analyte or binding partner to which the analyte binds specifically and selectively, comprising contacting the sample with a quantity of binding partner or analyte labelled with a first fluorescent material A and with a quantity of binding partner or analyte labelled with a second fluorescent material B, a unit of substance labelled with A and a unit of substance labelled with B both being capable of binding simultaneously to a single particle of analyte or binding partner or being capable of binding to each other, A and B being selected such that B is capable of being stimulated by radiative transfer fluorescence by light emitted from A when suitably stimulated, and B is not significantly stimulated by light which acts to stimulate A; if necessary adding to the sample a quantity of the analyte or binding partner; and monitorin _*g light emitted from B in response to suitable stimulation of A.

In a preferred aspect of the present invention there is provided a fluorescence assay for determining the presence in a sample of an analyte or binding partner to which the analyte binds specifically and selectively, comprising contacting the sample with a quantity of binding partner labelled with a first fluorescent material A and with a quantity of binding partner labelled with a second fluorescent material B, a unit of binding partner labelled with A and a unit of binding partner labelled with B both being capable of binding simultaneously to a single particle of analyte, A and B being selected such that B is capable of being stimulated by radiative transfer fluorescence by light emitted from A when suitably stimulated, and B is not significantly stimulated by light which acts to stimulate A; and monitoring light emitted

SUBSTITUTESHEET from B in response to suitable stimulation of A.

The phenomenon of radiative transfer fluorescence involves photons passing from a first fluorescent material (A) to a second fluorescent material (B) and an inverse square law applies so that a significant amount of light will only be emitted from the second fluorescent material when sufficiently .close to the first fluorescent material. When carrying out the assay of the invention, this is only likely to occur when a unit of substance labelled with A and a unit of substance labelled with B are bound in close proximity, e.g. by both being bound to a single particle of analyte, so that the emission of light from B is indicative- of the presence of analyte. By monitoring the intensity of light emitted, an indication can be obtained of the quantity of analyte present in the sample.

Because unbound B will not generally emit the light there is no need to remove unbound reagents e.g. by washing. The method thus provides a homogeneous, single step assay.

The method may be used qualitatively or quantitively, in the latter case by comparing results with those for known standards.

In certain embodiments it may be appropriate for reference purposes to obtain results from two similar systems, one with sample and one without sample.

It is preferred to use equal numbers of units of A labelled binding partner and B labelled binding partner. In this case, in the absence of any dual binding scheme, as discussed below, it is probable that half the binding

SUBSTITUTE SHEET events will result in a unit of bidding partner labelled with A and a unit of binding partner labelled with B being bound to a single molecule of analyte, forming an AB pair, rather than AA or BB pairs. Multiple binding events, if present, will increase sensitivity if each A can excite more than one B.

The method of the invention typically finds application in an immunoassay, e.g. using labelled antigens to detect an antibody of interest in a sample, with two labelled antigens binding to the antigen binding sites of each molecule of antibody of interest. Alternatively an antigen of interest may be detected by using two types of differently labelled antibodies, each antibody type binding to a different determinant of the antigen. Such an arrangement would theoretically result in production of 100% AB pairs on binding. It is also possible to use bispecific antibodies to increase the proportion of AB pairs on binding. It is envisaged the method could also be used in a competition assay, for instance with labelled antigen being used to detect antigen in the sample by competing for binding to added antibody. As a further possibility, the A and B materials may be used to label antibody and antigen respectively. In the absence of any target antigen (or antibody) the A and B units would all bind together. However, any target antigen would compete with the labelled antigen for binding sites and thus reduce the signal level.

The method of the invention can also be used for detection of enzymes or DNA species, and for detection of other materials as will be apparent to those skilled in the art. For example, for a DNA test the A arid B labels may be associated with two different DNA probes which will bind

SUBSTITUTE SHEET o specific sites on the target DNA, such that the A and B labels will be sufficiently close for efficient radiative fluorescence transfer i.e. label B should be within a few particle diameters of label A. The technique can also be used to detect molecules such as cholesterol by coating the labels with enzymes which bind to the target substrate. In principle the reverse is also possible, i.e. detection of enzymes by labels coated with substrate.

In order to increase the amount of light emitted and so improve sensitivity of the method, each unit of binding partner is preferably labelled with more than one molecule of fluorescent material A or B. This is conveniently achieved by encapsulating a large number of molecules of A or B in a particle of suitable carrier material and attaching binding partner to the particle. The particle is conveniently in the form of a microsphere or microparticle having a diameter in the range 1 to 100 microns. In this way each unit of binding partner may comprise a large number of fluorophores, possibly in the range of 10β to 1012, providing a substantial amplification effect.

Where the fluorescent material is an organic dye solution, large numbers of molecules may be encapsulated in cross- linked gelatin or other protein shells by forming emulsions in water-immiscible solvents, microcapsules of protein and carbohydrate such as gum arabic (as is used in carbonless copying paper), by in-situ polymerisation of a polymer such as polyacrylamide, polystyrene or nylon by precipitation of polymer around droplets of dye or by other methods. See "Microcapsule Processing and Technology", A. Konder and J.W. VQΠ Valkenburg, Marcel Dekker Inc., 1979 for further details.

SUBSTITUTESHEET Where the fluorescent material comprises a solid state fluόrophore such as neodymium ions (Nd 3+) or other laser media, these are conveniently doped into a glass or similar host, which may be produced in the form of microspheres or ground into small particles.

Proteins, including antibodies, may be bound to particles by a wide range of known techniques. One technique is non-covalent adsorption, as described in "Antibodies: a laboratory manual" by E. Harlow and D.Lane, Cold Spring Harbour Lab., 1988, chapter 14. A range of covalent links may also be used e.g. to an adsorbed protein via glutaraldehyde, p,p'difluoro-m.m'-dinitrophenylsulphone (FNPS) or similar bifunctional reagent: see "Handbook of experimental immunology:1, immunochemistry" ed D.M. Weir, L.A. Herzenberg, Blackwell 1986, p 35.26. Proteins may also be bound directly to "activated" plastics following treatment with cyanogen bromide, tocyl chloride etc (see Harlow and Lane pp 528-9) or directly to beads with active surface groups such as tocyl, chloride, etc. (as produced by, for example. Bangs Lab Inc. Car el, Indiana, U.S.A).

DNA probes may be bound to such beads by labelling the DNA with biόtin and binding this to avidin, itself bound to beads as described above (see e.g. T. Hultman, S. Stahl, E. Homes, M. ϋhler, "Direct solid phase sequencing of genomic and plasmid DNA using magnetic beads as solid support", Nucleic Acids Research 1*989 in press), or covalently by cross-linking using UV light (see E.W.K. Landjian, BioTechnology 5_^, 165-167,(1987), "Optimised hybridisation of DNA blotted and fixed to nitrocellulose and nylon membranes"), or by introducing a reactive functional group into the DNA (see e.g. J.N. Kremsky, J.L.

SUBSTITUTESHEET Wooters, J.P. Dougherty, R.E. Meyers, M. Collins, E.L. Brown "Immobilisation of DNA via oligonucleotides containing an aldehyde or carboxylic acid group at the 5 ' terminus", Nucleic Acids Research JJ^ 2891-2909 (1987); B. Connolly, "The syntheis of oligonucleotides containing a primary amino group at the 5' terminus" Nucleic Acids Research _15_, 3131-3139 (1987); S. Pocket, B.Arcangioli, T. Huynh-Dinh, "Solid supported ligation primer", Nucleic Acids Research, 16, 1619). DNA may also be linked to a support (which has been previously derivatized with a linker DNA) using enzymatic or chemical means (Pocket et al ibid. )

When considering glass microparticles, proteins, including antibodies, may be bound to glass essentially as to plastics, as described above. Specifically, proteins may be adsorbed to clean glass surfaces or cross-linked to covalently activated glass. DNA may be adsorbed non- covalently or cross-linked to activated glass, again essentially as described above; see e.g. S.S. Ghosh and F.G. Musso "Covalent attachment of oligonucleotide to solid supports" Nucleic Acids Research 13, 5353-5372, (1987).

Suitable pairs of fluorescent materials A and B can be selected such that A is stimulated by light of a first wavelength (or within a first range of wavelengths) L . and emits light of a second, longer wavelength (or range of wavelengths) L₂ which in turn stimulates B to emit light of a third, longer wavelength (or range of wavelengths) A->2 > It is also necessary for L-_j and L₂ to be distinct from each other, with no overlap, so that undesired stimulation of B on exposure to light of wavelength L__j does not occur.

SUBSTITUTESHEET One preferred pair of materials which satisfies these requirements is as follows:

A is the dye Styryl 9, e.g. as available commercially from Lambda Physik and other manufacturers. Graphs showing the absorption and fluorescence characteristics of Styryl 9 dye are given in "The Performance of Dyes and Dye Mixtures of Rhodamine 6G/Rhodamine 560 and of Styryl 8/Styryl 9 for Single Mode CW Ring Laser Operation", by J.J.L. Mulders, and L.W.G-. Steenhuysen in Optics Communications, volume 54 number 5, pages 295-298, July 1985. Brielfy, L__| is about 650mm and L₂ is 700-800 mm.

B is the solid state fluorophore Nd 3+ in the form of

Nd:glass, e.g. having a composition of 66 wt % Si0₂, 5wt %

Nd₂0₃, 16 wt % Na₂0, 5 wt % BaO, 2 wt % 1₂0₃ and 1 wt %

Sb_jO,. Such material has an absorption free window around

650nm (the excitation wavelength L.. for Styryl 9), has excitation peaks L₂ at 740nm and 810nm (i.e. 20nm wide absorption peaks centred on those wavelengths), a fluorescence wavelength L₃ of 1060nm and a quantum efficiency of about 0.4. In addition it has a long fluorescence lifetime, between 0.1 and 1 ms, allowing time gating against dye A fluorescence, as will be described below. This enables particle sizes to be optimised for minimum background due to direct excitation, as will also be described below.

The transfer efficiency of fluorescence transfer is very low, and it is accordingly important to optimise conditions to enhance the overall gain as far as possible. In particular, it is important to reduce background signals as far as possible.

SUBSTITUTE SHEET Background signals can arise from three main sources:

1. detection of L^ photons,

2. detection of L₂ photons,

3. direct excitation of B by L_. photons.

Time gating may be used to eliminate the possibility of background signals arising from the first two of these sources. This involves using a fluorescent material B with a longer fluorescence decay time than material A and detecting L., photons after the expiry of the fluorescence of A (ie after a few fluorescence lifetimes of A). The preferred pair of materials mentioned above satisfy this requirement, as Styryl 9 has a fluorescent decay of about 10 - 15 ns, while that of Nd: glass is between 0.1 and 1 ms.

Having eliminated sources 1 and 2, conditions such as relative sphere size, number density etc., can then be adjusted to minimise background signals from source 3, without having to worry about the conflicting effects of such adjustment on sources 1 and 2. Calculations based on a simple theoretical model suggest good results can be obtained with the radius of particles labelled with A (r,) being 10 times the radius of particles labelled with B

(r ). Suitable particle sizes may be

The increase in fluorescence on binding and hence sensitivity can be increased by adding a material C which absorbs the A fluorescence but does not fluoresce itself. This has the effect of reducing the background signal due to unbound B in solution fluorescing without affecting the

SUBSTITUTESHEET signal in the bound state

Material C preferably has low absorbance of light of the wavelength used to excite A, although this is not essential. For the preferred materials A and B discussed above, one suitable material C is the dye known as S101756 which is commercially available from ICI Colours and Fine Chemical. This dye is soluble in non-polar solvents and has a high molar extinction coefficient (about 120,000) peaking at 770nm. A dye with similar characteristics, but which is water soluble, known as S116510, is also available, as are broadband infrared absorbing dyes S109186, and S109564 (see ICI Technical Data Sheets). There are various alternative dyes, including several laser dyes such as I 1 0, IR132 which could be used provided the fluorescence of these could be quenched while maintaining their absorption properties.

Problems can arise due to non-specific binding, i.e. binding between A and B labelled substances in the absence of the target analyte, due to extraneous biological material in the sample. It is not possible to reduce this effect other than by careful selection of the labelled analytes or binding partners. However, the magnitude of the effect should be monitored, in order not to overestimate the amount of analyte present. One way of monitoring non-specific binding is to label an antigen (different from that labelled with A and B) with a third material. This antigen will be chosen to have the same degree of non-specific binding to A as B, and the label will be chosen either to absorb L~ and fluoresce at a different wavelength to B, e.g. L., or alternatively to absorb L₂ and not fluoresce. In the first case, a second detector and filter could monitor the amount of light at

SUBSTITUTESHEET Ϊ14, and hence determine the amount of non-specific binding. In the second case, the detector monitors the amount of light at L₂, which will decrease if non-specific binding occurs. It will be seen that in the second case, the label is required to have the same properties as material C which is used to increase the fluorescent gain, and it is thus possible to use microspheres containing dye C coated with an appropriate non-specific antigen.*

The illuminating source is conveniently a flash lamp or a light emitting diode (LED). These have properties as follows:

Flashlamp LED

Mean power 0.1 - 1 W 10mW

(in 20nm band)

Wavelength Whitelight 660 +_ 15nm

Pulse rate up to 100Hz up to 1MHz

Directivity 4 pi steradians about 1 steradian

Although an LED is less powerful, its higher directivity allows more efficient coupling to the sample, and it can be modulated at kHz frequencies to allow phase sensitive detection. It is also cheaper, and may not require a bandpass filter. For these reasons an LED is the currently favoured source.

The presently preferred light detector is a silicon photodiode, despite the reduced responsivity at 1060nm wavelength. A standard device (BPX 65) has an NEP of

SUBSTITUTE SHEET 1 A - - about 3 x 10 W//HZ. A large area photodiode (1cm ) has an NEP of about 1pW//HΪ^".

It is thought that the assay of the invention is capable of high sensitivity, possibly being able to detect only a few thousand particles of target material under suitably optimised conditions, which can be carried out using relatively cheap, compact apparatus and which has a wide range of applicability, to detection of antibodies, enzymes, DNA etc.

A preferred embodiment of the invention will now be described, by way of example, with reference to the accompanying drawings, in which:

Figure 1 illustrates schematically apparatus for performing the method of the invention; and

Figure 2 illustrates the timing cycle of the apparatus of Figure 1.

Detailed description of the drawings

The apparatus illustrated in Figure 1 comprises a cuvette 10 with capacity 500 nun for receiving reagents, which is located in an ellipsoidal reflector 12.^* An LED 14 and associated lens 16 are located on the axis of the reflector, outside the reflector, for supplying illuminating light onto the cuvette via a suitable aperture 18 in the reflector. LED 14 has the following characteristics.

Mean power 1OmW

Wavelength 660- +_ 15 nm

SUBSTITUTE SHEET Pulse rate up to 1MHz

Directivity about 1 steradian

The apparatus further comprises a silicon photodiode light detector 20, positioned on the reflector axis. Because of relatively poor off axis imaging performance, it is appropriate to use a detector 20 with a relatively large area of about 1cm , having an NEP (noise equivalent power) of about 1pW//HZ^". Alternatively the detector could be positioned directly by the cuvette. As a further possibility dichroic filters may be used so that the detector selectively reflects and transmits the pump radiation and fluorescent radiation from the cuvette, with some cost penalty.

In use a suitable quantity of a sample containing an unknown quantity of a substance of interest, e.g. a particular antibody, is located in the cuvette. In order to determine the presence or quantity of the antibody in the sample, qunatities of antigen corresponding to the antibody and separately labelled with two different fluorescent materials A and B, are added to the cuvette. The fluorescent materials A and B are both in the form of encapsulated microspheres, as described above, with the antigen bound to the surfaces thereof, as also described above.

In the presently preferred arrangement, material A is the dye Styryl 9, as described above, which is encapsulated by one of the techniques described above, e.g. by polymerisation of a polymer such as polyacrylamide around droplets of the dye, to form microsphere with a radius of

100 urn. The number density of the spheres, N, is in the range 10mm —3 to 1000mm—3. Material B is Nd: glass, as

SUBSTITUTESHEET described above, formed or ground into microspheres with a radius of 10um. With both types of sphere, the antigen is bound to the surface by one of the techniques described above, e.g. by non-covalent adsorption.

Approximately equal numbers of A labelled spheres and B labelled spheres are used.

A quantity of the fluorescence-absorbing dye S101756 (material C) is also added to the cuvette, either in the form of a ^'solution or encapsulated as microspheres with a radius of lum. -_f

The cuvette is illuminated by the LED 14 and emitted light monitored by detector 20 in accordance with the timing cycle illustrated in Figure 2. As illustrated, LED 14 is operated for short, discrete intervals 22, during which time detector 20 is switched off. The illumination of the contents of cuvette 10 causes A to fluoresce and consequential fluorescence of B by radiative transfer fluorescence for adjacent pairs of A and B labelled microspheres both bound to the same molecule of antibody of interest. The presence of fluorescence from B, the timing of which is illustrated in Figure 2, thus indicates the presence of the antibody of interest.

In order to take advantage of the time gating effect discussed above, detector 20 is operated in discrete intervals 24 while LED 14 is switched off, with a small time interval 26 being left between switching off LED 14 and switching on detector 20 to allow the fluorescence of A to decay fully (in a few microseconds). A small time interval 28 is left between switching off detector 20 and switching on LED 14 at the start of the next cycle to

SUBSTITUTESHEET avoid any risk of unwanted detection of illuminating light.

It may be desirable to reference the assay against a "blank" cuvette, i.e. one containing the labelled reagents but not the test sample. The apparatus may thus include a second cuvette and associated components as shown in Figure 1, for receiving the labelled reagents but no test sample. Light emitted from both cuvettes is measured periodically, and the amount of target analyte determined by comparing the results.

Some form of mixing/agitation will also be necessary in order for binding to occur in convenient times.cales e.g. a few minutes.

It is thought it may be possible to detect the presence of as few a thousands of target particles of interest.

SUBSTITUTE SHEET

Claims

1. A"fluorescence assay for determining the presence in a sample of an analyte or binding partner to which the analyte binds specifically and selectively, comprising contacting the sample with a quantity of binding partner or analyte labelled with a first fluorescent material A and with a quantity of binding partner or analyte labelled with a second fluorescent material B, a unit of substance labelled with A and a unit of substance labelled with B both being capable of binding simultaneously to a single particle of analyte or binding partner or being capable of binding to each other, A and B being selected such that B is capable of being stimulated by radiative transfer fluorescence by light emitted from A when suitably stimulated, and B is not significantly stimulated by light which acts to stimulate A; if necessary adding to the sample a quantity of the analyte or binding partner; and monitoring light emitted from B in response to suitable stimulation of A.

2. A fluorescence assay for determining the presence in a sample of an analyte or binding partner to which the analyte binds specifically and selectively, comprising contacting the sample with a quantity of binding partner labelled with a first fluorescent material A and with a quantity of binding partner labelled with a second fluorescent material B, a unit of binding partner labelled with A and a unit of binding partner labelled with B both being capable of binding simultaneously to a single particle of analyte, A and B being selected such that B is capable of being stimulated by radiative transfer fluorescence by light emitted from A when suitably

SUBSTITUTESHEET stimulated, and B is not significantly stimulated by light which acts to stimulate A; and monitoring light emitted from B in response to suitable stimulation of A.

3. An assay according to claim 1 or 2, wherein each unit of binding partner is labelled with more than one molecule of fluorescent material A or B.

4. An assay according to claim 3, wherein a large number of molecules of A or B are encapsulated in a particle of suitable carrier material and analyte or binding partner ^• is attached to the particle.

5. An assay according to claim 4, wherein the particle is in the form of a microsphere or microparticle having a diameter in the range 1 to 100 microns.

6. An assay according to any one of the preceding claims, wherein fluorescent material A is the dye styryl 9, and fluorescent material B is the solid state fluorophore Nd 3+ in the form of Nd:glass.

7. An assay according to any one of the preceding claims, wherein fluorescent material B has a longer fluorescence decay time than material A, and wherein light emitted from material B is detected after the expiry of the fluorescence of A.

8. An assay according to any one of the preceding claims, wherein the radius of particles labelled with A (r ) is 10 times the radius of particles labelled with B (r„).

9. An assay according to any one of the preceding claims, further comprising adding to the sample a material C

SUBSTITUTE SHEET which absorbs light emitted from A but does not fluoresce^* itself.

10. An assay according to claim 9, wherein fluorescent material A is the dye Styryl 9, fluorescent material B is the solid state fluorophore Nd 3+ in the form of Ndrglass, and material C is selected from the dyes known as S101756,

S116510, S109186, S109564, IR140, and IR132.

11. An assay according to any one of the preceding claims, wherein non-specific binding is monitored by labelling a further antigen with a third material, the further antigen having the same degree of non-specific binding to A as B, and the label absorbing light emitted from A and either fluorescing at a different wavelength to B, or not fluorescing.

SUBSTITUTE SHEET