GB2459384A

GB2459384A - Homogeneous immunoassay comprising z-axis fluorescence imaging

Info

Publication number: GB2459384A
Application number: GB0907066A
Authority: GB
Inventors: Homanaz Ghafari; Quentin Sean Hanley
Original assignee: Nottingham Trent University
Current assignee: Nottingham Trent University
Priority date: 2008-04-24
Filing date: 2009-04-24
Publication date: 2009-10-28
Also published as: GB0807466D0; GB0907066D0

Abstract

An assay such as a capture ELISA or other antibody binding assay comprises providing a support surface to which first reagent 10 (anti-human IgG) is immobilized prior to the overlaying of a second reagent 14 (fluorescently labeled anti-human IgG, for example to detect human IgG molecules 12), the assay being characterised in that a series of measurements of the label intensity (plotted graph curve) along the z-axis (e.g. confocal optical slice images in the z-series), so as to obtain an assay response curve by which to determine the contribution 20 of the portion of the second reagent 14 that is bound directly or indirectly to the first reagent 10.

Description

INTELLECTUAL

. .... PROPERTY OFFICE Application No. GB0907066.5 RTM Date:17 August 2009 The following terms are registered trademarks and should be read as such wherever they occur in this document: Sigma-Aldrich Nalgene Nunc Lab-Tek Leica Intellectual Property Office is an operating name of the Patent Office www.ipo.gov.uk Title -Improvements relating to Assays including a Labelled Reagent This invention relates to assays including a labelled reagent, and in particular chemical and biochemical assays including a labelled reagent.

Chemical and biochemical assays are analytical methods that are widely used for diagnosis, chemical analysis and screening. In particular, chemical and biochemical assays generally involve the detection of the specific binding of two or more reagents. For example, a specific type of biochemical assay is an immunoassay. Immunoassays generally involve the detection of the specific binding of an antibody to its antigen, and typically enable measurement of the concentration of the bound or unbound, antibody or antigen. Research efforts in relation to chemical and biochemical assays are generally directed toward increasing the sensitivity of the assay, decreasing the processing time, and reducing the amount of reagents and samples required.

Detection of one or more of the reagents of the assay can be achieved by a variety of methods. One of the most common is to label the relevant reagent. In particular, an enzyme or a fluorophore is typically attached to the relevant reagent, which may then be detected by light absorbance or fluorescence detection. Where the assay is an immunoassay, these techniques are typically referred to as enzyme-linked immunosorbent assays (ELISA) and fluorescence-linked immunosorbent assays (FLISA) respectively.

Examples of assays involving a labeled reagent include competitive and non-competitive assays. For instance, in a typical competitive immunoassay, an antigen in a sample competes with a labeled antigen to bind with antibodies. The amount of labeled antigen bound to the antibody site is then measured. In this method, the response will be inversely proportional to the concentration of antigen in the sample. In a sandwich immunoassay, which is a typical non-competitive immunoassay, an antigen in the sample is bound to a primary antibody, and then a labeled, secondary antibody is bound to the antigen. The amount of labeled antibody bound to the antigen site is then measured. Unlike the competitive method, the results of the non-competitive method will be directly proportional to the concentration of antigen in the sample.

In addition, assays are either homogeneous or heterogeneous. In particular, a heterogeneous assay includes one or more steps of removing unbound reagent from the assay, typically by means of a washing step. In contrast, homogeneous assays do not require such washing steps, such that unbound reagent remains within the assay during measurement of the label intensities.

Heterogeneous assays are typically more sensitive and/or accurate because background signals are significantly reduced or eliminated. However, homogeneous assays are more convenient and much less time consuming, and hence suitable for processing a large number of samples.

Existing homogeneous assays typically utilise support surfaces provided by a plurality of beads. However, beads are susceptible to motion during measurement unless immobilised. Furthermore, the shape of the bead requires complex procedures to remove background signals and determine the label intensity arising over the entire volume of the bead.

There has now been devised an improved assay which overcomes or substantially mitigates the above-mentioned and/or other disadvantages

associated with the prior art.

According to a first aspect of the invention, there is provided an assay comprising the steps of: (a) providing a support surface; (b) immobilising a first reagent on the support surface; (c) applying an overlayer to the first reagent, the overlayer including a labelled second reagent; (d) taking a series of measurements of the label intensity along an axis that extends through both the support surface and the overlayer, so as to obtain an assay response curve; and (e) determining the contribution to the assay response curve of that portion of the second reagent that is bound either directly or indirectly to the first reagent.

According to a further aspect of the invention, there is provided a kit for performing an assay, which kit comprises an assay including a first reagent immobilised on a support surface, and an overlayer applied to the first reagent, the overlayer including a labelled second reagent, the kit further comprising apparatus that includes means for taking a series of measurements of the label intensity of the assay along an axis that extends through both the support surface and the overlayer, so as to obtain an assay response curve, and means for determining the contribution to the assay response curve of that portion of the second reagent that is bound either directly or indirectly to the first reagent.

By "reagent" is meant either a substance that is to be detected, measured or examined, ie a substance contained in a sample, or a substance that is used to detect, measure or examine the sample. Which reagent is being detected, measured or examined will differ between different configurations of assay.

Nevertheless, the labelled second reagent will typically be a substance that is used to detect, measure or examine a sample, which typically comprises either the first reagent or an additional, third reagent.

The assay and associated apparatus according to the invention is advantageous principally because it enables accurate determination of a response curve for that portion of the labelled, second reagent that is bound either directly or indirectly to the immobilised first reagent, without removal of the overlayer containing unbound, labelled reagent. The present invention therefore reduces the number of wash steps that are necessary for accurate measurements, and also enables wash steps to be eliminated entirely to provide homogeneous assays according to the invention. Furthermore, the present invention provides a much less complex assay than prior art homogeneous assays, which typically include support surfaces provided by beads or the like.

A plurality of assays may be provided, and the assays are preferably conducted substantially simultaneously, or sequentially. In particular, the assays are preferably provided in the same support structure, which preferably comprises an array of support surfaces that each carry a separate assay. For example, the plurality of assays may have the form of a microarray. These support surfaces may or may not be co-extensive. As discussed in more detail below, this array of support surfaces may include one or more support surfaces arranged in one or more planes.

The support surface of the assay is preferably planar in form, and hence extends along orthogonal x-and y-axes. Furthermore, the axis along which the series of label intensity measurements are taken, ie the measurement axis, is preferably substantially linear. In this arrangement, the measurement axis is preferably orientated substantially perpendicularly to the support surface, such that the measurement axis preferably extends along a z-axis, which is orthogonal to the x-and y-axes of the plane of the support surface. This arrangement of the assay enables a plurality of label intensity measurements to be taken for each position along the measurement axis. In particular, a plurality of measurements may be taken in a plane that is perpendicular to the measurement axis, ie the x-y plane, such that the assay response curve may constitute data that is the proportional to the sum of those measurements for each position along the measurement axis.

In this arrangement, the support surface is preferably very flat.

The support surface is conveniently provided by the surface of a support of transparent material, such as glass or a suitable plastics material, and may have the form of a slide or a microtitre plate. In particular, the support surface may be conveniently provided at the base of a well. Where a plurality of assays is provided, the slide preferably includes an array of wells, with each assay being prepared within a separate well.

The first reagent is preferably immobilised on the support surface by applying the first reagent within a suitable solvent to the support surface, preferably incubating the assay, and preferably washing the assay to remove any free first reagent.

The first reagent is preferably not labelled at all, but in any case is preferably not labelled in such a way that interferes with measurement of the labelled, second reagent.

The assay may be a competitive assay. By "competitive assay" is meant an assay in which a plurality of reagents of the assay, including a labelled reagent, are capable of binding to a binding site, and the plurality of reagents compete to bind to that binding site. Where the assay is a competitive assay, the labelled, second reagent preferably competes with a third reagent to bind to the first reagent. In this form of assay, the overlayer preferably includes both the labelled, second reagent and the third reagent, and the response curve for that portion of the second reagent that is bound either directly or indirectly to the first reagent will be inversely proportional to the concentration of the third reagent within the overlayer. Unlike conventional competitive assays, there is no need to remove the overlayer before taking the label intensity measurements. Where the competitive assay is an immunoassay, the first and second reagents will typically be antibodies, and the third reagent will typically be an associated antigen.

The assay may be a non-competitive assay. By "non-competitive assay" is meant an assay in which a labelled reagent of the assay does not compete with any other reagent to bind to a binding site. A typical non-competitive assay is a sandwich assay, in which a third reagent is bound to the immobilised first reagent, and the labelled, second reagent is capable of binding to the third reagent. Where the non-competitive assay is an immunoassay, the first reagent will typically be an antibody (often referred to as the "primary antibody"), the second reagent will typically be an antibody (often referred to as the "secondary antibody"), and the third reagent will typically be an associated antigen.

This type of assay may be homogeneous, in which the labelled, second reagent and the third reagent are applied to the first reagent simultaneously as constituents of the overlayer, and the overlayer is not removed before taking the label intensity measurements.

Alternatively, the non-competitive assay may be heterogeneous, in which the third reagent is applied to the first reagent, the assay is preferably incubated, the unbound third reagent is preferably removed, most preferably by washing, following incubation, and then the overlayer including the labelled, second reagent is added to the first reagent and the bound third reagent. Again, unlike conventional non-competitive assays, there is no need to remove the overlayer before taking the label intensity measurements.

Each of the above described assays preferably comprises a layer of the first reagent, which is approximately one molecule in thickness, that is immobilised on the support surface, and a layer of the labelled, second reagent that is bound to the first reagent and is also approximately one molecule in thickness. As discussed above, the labelled, second reagent may be bound directly to the first reagent, or may be bound via a layer of a third reagent, which is also approximately one molecule in thickness. Each of the above described immunoassays preferably therefore results in a layer of approximately 2 or 3 molecules in thickness, which includes labelled, secondary reagent that is bound either directly or indirectly to the first reagent. This active layer will typically, therefore, be approximately SOnm in thickness, but may have a thickness of up to SOOnm, depending upon the size of the relevant molecules.

The overlayer preferably comprises a solvent and the labelled, second reagent, and may also include an antigen, as discussed above. The overlayer is preferably of substantially greater thickness than the active layer discussed above, and is most preferably at least 102, i03 or i04 times greater in thickness.

Once the overlayer has been added to the assay, the assay is preferably incubated, before the label intensity measurements are taken.

The label intensity measurements are preferably taken using an optically sectioning microscope, which is a microscope capable of taking measurements at different depths within an assay, ie positions along a measurement axis, without any need to physically section the assay. Most preferably, the optically sectioning microscope is a confocal microscope, a two-photon microscope, a three-photon microscope, a stimulated emission depletion microscope, or any other suitable optically sectioning means.

The second reagent may include any suitable label. In particular, the second reagent may be fluorescently-labelled. Nevertheless, there are many different labelling technologies that are suitable for use with the invention, including organic fluorophores, lanthanide chelates, quantum dots, fluorescent nano particles and lanthanide doped Si02.

The apparatus for performing the assay preferably includes means for moving the assay and/or the microscope objective, such that measurements are taken at a series of positions along the measurement axis. This movement is preferably, therefore, linear movement, and is preferably orientated substantially perpendicular to the support surface. Furthermore, the apparatus is preferably adapted to take a plurality of measurements for each position along the measurement axis. In particular, a plurality of measurements may be taken in a plane that is perpendicular to the measurement axis, ie the x-y plane. In particular, an image of label intensities in the x-y plane may be obtained, for example by scanning over the x-y plane.

The assay response curve comprises data for the label intensity at different positions along the measurement axes. In particular, the assay response curve data may be derived from individual measurements for each position along the measurement axis and/or the summation of a plurality of measurements taken in a plane for each position along the measurement axis that is perpendicular to the measurement axis, for example a label intensity image.

The numerical aperture (NA) of the objective of the microscope is preferably optimised in order to substantially minimise the Full Width Half Maximum (FWHM), and hence improve the resolution of the measurements. In addition, the pinhole size of the microscope is preferably approximately 40-70%, and more preferably 50-60%, of the Airy disk diameter in the focal plane for the selected NA.

The step of determining the contribution to the assay response curve of that portion of the second reagent that is bound either directly or indirectly to the first reagent may be achieved using a variety of methods. In presently preferred embodiments, this determination is achieved utilising information regarding the contribution to the assay response curve of the unbound, labelled second reagent within the overlayer. This information may be derived from theoretical information, experimental information, or a combination of theoretical and experimental information. In any case, this determination preferably utilises information regarding the expected label intensity as a function of distance along the measurement axis, and hence the expected assay response curve, including separate coefficients for the contributions of bound and unbound second reagent.

In particular, this expected label intensity information is preferably fit to the measured assay response curve in order to determine the contribution to the assay response curve of that portion of the second reagent that is bound either directly or indirectly to the first reagent.

In particular, the response curve for that portion of the second reagent that is bound either directly or indirectly to the first reagent may be determined by convolving ideal objects for (i) an active layer including the first reagent and the bound, labelled second reagent, and (ii) the overlayer, with a calculated point spread function (PSF), and making linear combinations of the two ideal objects until a fit to the assay response curve is obtained, for example using least squares minimization. For instance, the PSF as described by Vandervoort HTM, Brakenhoff GJ. (3-D Image-Formation In High-Aperture Fluorescence Confocal Microscopy -A Numerical-Analysis. Journal Of Microscopy-Oxford 1 990;1 58:43- 54) may be applied to the ideal objects. By "ideal objects" is meant calculations of the expected label intensity of the active layer and the overlayer for particular concentrations of label within those layers.

Alternatively, an empirical approach may be used in which the removal of that component of the assay response curve that arises from unbound second reagent within the overlayer may be achieved using experimental data for the expected contribution of the unbound, labelled second reagent to the assay response curve. In particular, a series of calibration measurements may be taken along the measurement axis for (i) an active layer including the first reagent and the bound, labelled second reagent, in the absence of the overlayer, and (ii) the overlayer in the absence of the active layer. The assay response curve may then separated by making linear combinations of the two series of calibration measurements.

Nevertheless, in presently preferred embodiments, a semi-empirical approach is used, in which the removal of that component of the assay response curve that arises from unbound second reagent within the overlayer may be achieved using a distribution function fitted to experimental data for the expected contribution of the unbound, labelled second reagent to the assay response curve. In particular, this semi-empirical approach may comprise the following steps: (a) taking a series of calibration measurements along the measurement axis for (i) an active layer including the first reagent and the bound, labelled second reagent, in the absence of the overlayer, and (ii) the overlayer in the absence of the active layer; (b) selecting a distribution function that substantially fits each series of calibration measurements; and (c) making linear combinations of the two distribution functions until a fit to the assay response curve is obtained.

In particular, the Cauchy-Lorentz and cumulative Cauchy-Lorentz functions have been found to fit the assay response curve well.

This determination of the response curve for that portion of the second reagent that is bound either directly or indirectly to the first reagent is preferably carried out by a computer.

As discussed above, a plurality of assays may be provided, and the assays are preferably conducted substantially simultaneously, or sequentially. In particular, the assays are preferably provided in the same support structure, which preferably comprises an array of support surfaces that each carry a separate assay. These support surfaces are preferably arranged in the same x-y plane, and may or may not be co-extensive with one another. In addition, however, the present invention also enables an assay to be performed that comprises a plurality of support surfaces that are separated along the measurement axis, eg the z-axis, such that the assays may be multiplexed along the measurement axis.

In particular, the measurement axis may extend through a plurality of separate assays. For instance, each assay may comprise a support surface, an active layer and an overlayer, and a plurality of assays may be stacked with the support of one assay being supported by the overlayer of the immediately adjacent assay.

Most preferably, however, a separator layer is disposed between the support of one assay and the overlayer of the immediately adjacent assay, for example comprising microscope oil or the like.

The assays may therefore be provided in a support structure including an array of support surfaces arranged in one x-y plane, and one or more further arrays of support surfaces, each arranged in a different x-y plane, wherein each support surface carries a separate assay. Clearly, this arrangement would significantly increase the speed with which a plurality of assays may be processed.

Nevertheless, where an array of assays are arranged in a plane, the array of assays may be arranged on a common support surface. In this arrangement, the overlayer of each assay may be the same, in which case these overlayers may form a unitary body. Alternatively, the overlayers of at least some of the assays may have different constituents from one another, in which case these overlayers are preferably separated from one another.

Where a plurality of assays are provided, at least some of the assays preferably have different constituents from one another. For instance, the first reagent may differ between at least some of the assays, the labelled, second reagent may differ between at least some of the assays, and/or any other constituent of the assays, for example a third reagent, may differ between at least some of the assays. For instance, different first reagents may be arranged in a particular order in the array, for example first reagents that are members of similar or identical groups may be located in proximity to one another. The array of assays may, for instance, be arranged in a grid pattern. By way of example, each row of the grid pattern may represent a group of different first reagents that share the same or similar characteristics, for example with specificity for the same or similar reagent.

The assay according to the invention may be a chemical assay, involving the detection and/or measurement of a chemical compound. However, the assay is preferably a biological assay involving the detection and/or measurement of a biological molecule. Hence, the first reagent and/or the second reagent are preferably biospecific reagents. By the term "biospecific reagent" is meant a reagent that is capable of binding to another biological molecule, for example via biological interactions. Hence, the first reagent and/or the second reagent are preferably biological molecules. By the term "biological molecule" is meant any organic molecule with biological activity and/or specificity.

By way of example, the biological molecule may be a part of a cell, organelle, virus, a phage, micro-organism,, etc. The biological molecule may be a macromolecule from a living organism. For example, the biological molecule may comprise a nucleic acid molecule, which may be single-stranded or double-stranded. For example, the biological molecule may be DNA or cDNA.

Alternatively, the biological molecule may be RNA (eg mRNA, rRNA, tRNA, or The biological molecule may comprise a plurality of amino acids, such as a peptide, a polypeptide or a protein. The biological molecule may be a receptor, an enzyme or an antibody. For example, the first reagent and/or the second reagent may an antibody, such that the assay is an immunoassay.

Examples of the invention will now be described in greater detail, by way of illustration only, with reference to the accompanying figures, in which Figure 1 shows an example of the z-axis response from a heterogeneous sandwich immunoassay according to the present invention, and includes schematic diagrams of that immunoassay; Figure 2 shows the z-axis response from a heterogeneous sandwich immunoassay, with its overlayer removed, for a variety of objectives with different magnification and NA; Figure 3 shows the z-axis response from a heterogeneous sandwich immunoassay, with its overlayer removed, for a variety of pinhole sizes; Figure 4 shows the variation of the full-width-half-maximum (FWHM) of the z-axis response of a confocal microscope as a function of pinhole size for a 40x 0.75 NA objective; Figure 5 shows the results from a heterogeneous sandwich immunoassay according to the present invention, including the separated signals for the active layer and the overlayer; Figure 6 shows the response for the primary antibody of a heterogeneous sandwich immunoassay according to the present invention; Figure 7 shows the concentration response for the secondary antibody of a heterogeneous sandwich immunoassay according to the present invention; Figure 8 shows the concentration response for the antigen of a heterogeneous sandwich immunoassay according to the present invention; Figure 9 shows the z-axis response of a heterogeneous sandwich immunoassay according to the present invention for a range of low concentrations of antigen; Figure 10 shows the z-axis response of a heterogeneous sandwich immunoassay, with the overlayer removed, for a range of concentrations of antigen; Figure 11 shows the z-axis response of a homogeneous sandwich immunoassay according to the present invention for a range of concentrations of antigen; Figure 12 shows the z-axis response of a z-axis multiplexed assay according to the present invention having two stacked assays; and Figure 13 shows the z-axis response of a z-axis multiplexed assay according to the present invention having three stacked assays.

These examples demonstrate heterogeneous and homogeneous sandwich immunoassays according to the present invention, using a human IgG model system. The z-axis responses of these immunoassays show a label intensity variation originating from an active layer on a glass surface in the presence of an overlayer containing unbound, fluorescently-labelled reagent. In addition, the parameters affecting the response of the immunoassay according to the invention were also investigated. In particular, the z-axis responses of the active layers generated by an immunoassay according to the invention have been studied in relation to changes of numerical aperture and pinhole size.

These results are general to a wide range of assays according to the present invention, including assays for the detection of genetic material and infectious diseases. While these examples utilise traditional organic fluorophores, other labelling technologies, such as lanthanide chelates, quantum dots, fluorescent nano particles, and lanthanide doped Si02 may be used. In addition, although a confocal detection method has been employed for these examples, other optical sectioning methods, such as two-and three-photon excitation, and stimulated emission depletion (STED), may also be used.

Materials The reagents used in these examples were (i) human IgG, goat anti-human IgG (y-chain specific) (Lot no. 095k6026), (ii) goat anti-human IgG (y-chain specific) FITC conjugate (Lot no. 086H8822), and (iii) bovine serum albumin minimum 98% (BSA) (Lot no. 095K1 113). The buffer solutions used in these examples were: (i) phosphate buffered saline pH7.4 (PBS) (Lot no. 069K8204); and (ii) NaN3 (Lot no.49H2527). Each of the reagents and buffer solutions were purchased from Sigma-Aldrich Company Ltd (Dorset, UK).

The solid supports used for these examples were chambered cover glass slides (Lab-Tek chambered #1.0 Borosilicate, Lot no.583345 1104, Nalgene Nunc international, USA).

A confocal scanner head (TCS NT; Leica Microsystems, Heidelberg GmbH, Germany) installed on an inverted microscope (DM IRBE, Leica Heidelberg GmbH, Germany) was used for the optical sectioning of the immunoassay. The 488 nm line of an argon ion laser was used for excitation and the emission fluorescent intensity was detected using a 530/30 nm bandpass filter. The objectives used were: lOx NA 0.3, 1 6x NA 0.5, 40x NA 0.6, 40x NA 0.75, and 63x NA 0.7.

Preparation of Heterogeneous Sandwich Immunoassay The heterogeneous sandwich immunoassays were conducted by immobilizing anti-human IgG as the primary antibody on the chamber slide overnight at 4° C. The remaining sites for protein binding were blocked with a solution of 1% BSA (1% BSA in PBS with 0.02% NaN3) for 2 hours. After washing with PBS, human IgG was added as antigen and incubated for 2 hours. The chambers were emptied and washed with buffer. Anti-human IgG-F ITO was added to the chambers as the labelled, secondary antibody and incubated for 1 hour. All steps from blocking to final incubation were performed at room temperature under continuous shaking.

Preparation of Homogeneous Sandwich Immunoassay The homogeneous sandwich immunoassays were conducted by immobilizing anti-human IgG as the primary antibody on the chamber slide overnight at 4° C. The remaining sites for protein binding were blocked with a solution of 1 % BSA (1% BSA in PBS with 0.02% NaN3) for 2 hour. Human IgG (antigen) and goat anti-human IgG-FITC (labelled, secondary antibody) were mixed and incubated for 1 hour, added to the coated chamber slides, and incubated for 2 hours at room temperature. All steps from blocking to final incubation were performed at room temperature under continuous shaking.

Method of Analysis The heterogeneous and homogeneous sandwich immunoassays both comprise an active layer which is approximately SOnm thick due to the size of three stacked IgG molecules, and an overlayer of considerably greater thickness. This active layer (20) is shown schematically in Figure 1, with a schematic representation of anti-human IgG molecules (10) bound to the surface of the chamber slide, human IgG molecules (12) bound to the anti-human IgG molecules, and labelled anti-human IgG-FITC molecules (14) bound to the human IgG molecules. As shown in Figure 1, the z-axis is orientated perpendicularly to the active layer (20) and the surface of the slide. Figure 1 also schematically shows the unbound labelled secondary antibodies (14) in the overlayer of the assay.

In the confocal microscope, the planar assay format can be thought of as the sum an active layer and an overlayer. For the calculation of ideal objects, the active layer is taken to be 50 nm thick while the overlayer is taken to consist of a uniform fluorescent object beginning 50 nm above the glass surface and extending away from the active layer. The active layer and overlayer differ only in the concentration of fluorophores, designated Ci and 02, respectively. More rigorously, the two fluorescent objects (0) in the microscope can be given as a function of position along the z-axis: 01(z) = C1 for 0 <z <50 nm, and 01(z) = 0 everywhere else and 02(z) = C2 for 50 nm �= z <°°, and 02(z) = 0 elsewhere. The overall object function is the sum of the two: 0(z)= 01(4+02(z) (1) For the purpose of the assay, the key to making accurate measurements of surface bound fluorescence is the ability to selectively determine the concentrations Cl and 02.

In the microscope, the object function is not observed directly. Instead, the measured object (M) depends on the object function convolved with h, the confocal point spread function (PSF): M(z)=h�0(z) (2) Since the diffraction limited resolution along the z-axis in a confocal microscope observed at high numerical aperture is of the order of 500 nm, the surface bound layer is thin relative to the resolution of the microscope. This has the consequence that although the object as modelled consists of exclusive zones containing either Ci or C2, each label intensity measurement contains signals arising from both the overlayer and the active layer at every position along the z-axis. For large z, the contribution of the active layer can be made arbitrarily small. A similar condition does not hold for the overlayer; the overlayer will always contribute significantly to the active layer due to blurring by the confocal PSF. M can be broken down into the parts corresponding to the two composing objects: M(z)=h�01(z)+h�02(z) (3) In order to obtain information selectively from the active layer, the ideal objects (Equations 1 and 2) may be convolved with a computed point spread function.

The concentrations may then be adjusted until a fit to the assay response curve is obtained using least squares minimization. For this study, the PSF as described by Vandervoort HTM, Brakenhoff GJ. (3-D Image-Formation In High-Aperture Fluorescence Confocal Microscopy -A Numerical-Analysis. Journal Of Microscopy-Oxford 1990;158:43-54) was applied to idealized objects as in equation 1.

Alternatively, an empirical approach may be used in which the z-axis response of the microscope is measured for an active layer in the absence of an overlayer, and for the overlayer in the absence of the active layer. The assay response curve may then synthesized by making linear combinations of the two sets of calibration measurements.

Nevertheless, best results were obtained using a semi-empirical approach, in which the measurements from the active layer and the overlayer in isolation were fitted to distribution functions that were found to mimic much of the behaviour of planar objects modified by PSFs. In particular, it was found that the Cauchy-Lorentz function and the cumulative Cauchy-Lorentz function are comparatively simple functions that fit the shape of real axial response functions affected by aberrations reasonably well.

In particular, the semi-empirical fit used the following form: / \ 1 (i ____ i Mz)=C1 +C21 -arctanl ° 1+-i+B (4) ((zz)2 y) 2) ltY 1+1 where, z0 is the centre of the active layer, y is the width of the response, and B is a constant which corrects for photomultiplier tube background in the confocal microscope.

The sum of the intensity of each image along the z-axis was computed and plotted against z-axis position. These data were fit to equations (2) and (4) using least squares minimization.

Experiments to Investigate Parameters Affecting the Con focal Z-axis Response Firstly, the parameters affecting the z-axis response of the assays were investigated, and in particular numerical aperture (NA) and pinhole size were investigated.

In relation to the numerical aperture (NA), the intensity of a fluorescent active layer resulting from the human lgG model system after removal of the overlayer was measured in the confocal microscope with five different objectives having NAs that varied from 0.3 to 0.7. The z-axis response was measured with a 74 pm pinhole every 0.4 pm along the z-axis for a total of 50 images (Fig.2). While remaining constant in this set of experiments, the pinhole was below one Airy disk diameter and has little or no effect on the resolution for this set of magnifications and NAs. The full width at half maximum (FWHM) was measured for each objective, and the results are shown in the following table (Table 1).

NA Magnification FWHM (pm) 0.3 10 10.5 0.5 16 7.0

(Table 1)

0.6 40 3.4 0.7 63 2.8 0.75 40 1.8 As expected, the FWHM decreased with increasing NA with all the objectives showing an z-axis response to the active layer even at low NA. However, the improved resolution at high NA is advantageous as it will result in less blurring between the active layer and the overlayer, and thus improve both sensitivity and limits of detection. Based on these results, the 40x NA 0.75 objective was selected for subsequent use.

In relation to the pinhole size, the effect on the intensity and resolution of the z-axis response was studied using the 40x NA 0.75 objective using 50 steps of 0.4 pm along the z-axis (Fig.3). The intensity of the signal increased nearly linearly with pinhole size and the FWHM increased rapidly above 150 pm (Fig. 4). The Airy disk diameter in the focal plane for the 40x NA 0.75 objective is approximately 170 pm and the optimal pinhole size is considered to about 50- 60% of this value which is in good agreement with the measured data. In addition, since the axial resolution does not improve significantly below this optimum, there is no benefit to further reduction in pinhole size. The only effect would be to reduce the signal strength from the active layer. Based on these experiments the pinhole size was set to 74 pm in subsequent experiments.

Heterogeneous Sandwich Assay Human IgG as antigen (10 pg/mI) was added to coated chamber slides with primary antibody (2Opg/ml), and secondary antibody (4Opg/ml) was applied to detect it, as discussed in detail above. The chamber was scanned in the confocal microscope using a 40x NA 0.75 objective and a 74 pm pinhole without removing the overlayer solution.

The z-axis response exhibited three distinct features (Figs. 1 and 5): i) a region of gradually decreasing signal trending to zero which corresponds to the chamber slide glass below the assay (0 to 5 pm); ii) a prominent peak of fluorescence arising from the active layer of concentrated fluorescence on the surface of the plate (5 pm); iii) a region of decreasing fluorescence trending to a constant value resulting from the fluorescence of the overlayer (5 to pm).

The FWHM of the peak was 1.75 pm and the measured z-axis response indicated that the peak height contains signals from both the active layer and the fluorescent overlayer. The presence of overlayer fluorescence indicates that an excess of secondary antibody is present and hence the active layer signal must be carefully separated from the overlayer signal, ie the background signal.

The method used for separation of the response into active layer and overlayer signals was based on distribution functions, as discussed in more detail above.

Separate scans were generated corresponding to the active layer and overlayer alone. After testing a few functions, the Cauchy-Lorentz and cumulative Cauchy-Lorentz functions were found to fit the assay response curves well. This allowed the decomposition of the two objects when presented together (Fig. 5). Excellent reproducibility (�0.04 pm) was seen in localizing the position of the active layer across the bottom of the chamber slide indicating good slide flatness and microscope stability.

Heterogeneous Sandwich Assay -Optimisation The concentration of coating antibody (goat anti-human IgG) was tested from 2.5 to 320 pg/mI at a constant concentration of antigen (10 pg/mI) and secondary antibody (100 pg/mI) (Fig. 6). The observed behaviour appeared to follow a Langmuir adsorption process, and a preferred primary antibody concentration of pg/mI concentration was chosen. The influence of the secondary antibody concentration was investigated between 10 and 100 pg/mI (Fig. 7) at a constant coating of primary antibody (160 pg/mI) and antigen (10 pg/mI). From these results, a preferred secondary antibody concentration of 50 pg/mI was chosen.

Heterogeneous Sandwich Assay -Sensitivity Concentrations of antigen from 0.156 to 1 Opg/ml were then tested to study the sensitivity of the heterogeneous immunoassay, using a constant primary antibody concentration of 160 pg/mI and a constant secondary antibody concentration of pg/mI. All the antigen concentrations were detected by confocal readout, as discussed in more detail above. The concentration response was plotted with the intensity obtained from the semi-empirical separation method versus the antigen concentration (Fig. 8). The graph shows a polynomial response with a nearly straight line at low levels followed by departure from linearity at higher concentrations.

The sensitivity experiment was continued with a lower range of antigen concentrations from 0.58 and 150 ng/ml. The z-axis response showed that the lowest detectable concentration by this method was 37 ng/ml (Fig. 9). The intensity of the fluorescent solution above the active layer overwhelmed the intensity of the active layer and it was not possible to detect the active layer's signal with the semi-empirical separation method. After removal of the fluorescent overlayer, the detection limit was less than 2.3 ng/ml. Below this concentration non-specific interactions limited (Fig. 10).

Example 2 -Homogeneous Con focal Sandwich Assay Antigen concentrations from 0.015 to 1 pg/mI were studied with a constant primary antibody (160 pg/mI) and secondary antibody (50 pg/mI), prepared in accordance with the method discussed above. The z-axis response is shown in Fig. 11. The lowest detected concentration with semi-empirical separation method was 250 ng/ml. The z-axis response after removing the fluorescent overlayer confirmed the presence of the active layer. These results demonstrate a homogeneous planer assay according to the invention. However, they confirm a tendency for homogeneous fluorescence assays to be less sensitive than comparable heterogeneous assays.

Example 3-Z-Axis Multiplexed Assay A 30 pm glass sheet was cut into 9x9 mm pieces. Each sheet was used to perform a heterogeneous sandwich immunoassay on one side of each glass piece. A first antibody was coated onto the surface by incubating overnight at 4 C. After washing, unbound sites were and incubated for 2 hours. Antigen was added to the surface followed by incubation and washing, the second labelled antibody was added and incubated 2 hours at room temperature. The immunoassays were readout with confocal microscopy. The glass sheets were attached to the bottom of a chambered cover glass slide. The z-axis responses of the immunoassays on two and three layers are shown in Figures 12 and 13, respectively.

Claims

Claims 1. An assay comprising the steps of: (a) providing a support surface; (b) immobilising a first reagent on the support surface; (C) applying an overlayer to the first reagent, the overlayer including a labelled second reagent; (d) taking a series of measurements of the label intensity along an axis that extends through both the support surface and the overlayer, so as to obtain an assay response curve; and (e) determining the contribution to the assay response curve of that portion of the second reagent that is bound either directly or indirectly to the first reagent.
2. An assay according to Claim 1, wherein the assay is a biological assay.
3. An assay according to Claim 2, wherein the first and second reagents are biospecific reagents.
4. An assay as claimed in any preceding claim, wherein the support surface of the assay is planar in form, and hence extends along orthogonal x-and y-axes.
5. An assay as claimed in Claim 4, wherein the axis along which the series of label intensity measurements are taken, ie the measurement axis, is substantially linear.
6. An assay as claimed in Claim 5, wherein the measurement axis is orientated substantially perpendicularly to the support surface, such that the measurement axis extends along a z-axis, which is orthogonal to the x-and y-axes of the plane of the support surface.
7. An assay as claimed in Claim 6, wherein a plurality of label intensity measurements are taken for each position along the measurement axis.
8. An assay as claimed in Claim 7, wherein a plurality of label intensity measurements are taken in a plane that is perpendicular to the measurement axis, ie the x-y plane, such that the assay response curve constitutes data that is proportional to the sum of those plurality of measurements for each position along the measurement axis.
9. An assay as claimed in any preceding claim, wherein the first reagent is immobilised on the support surface by applying the first reagent and a suitable solvent to the support surface, incubating the assay, and washing the assay to remove any free first reagent.
10. An assay as claimed in any preceding claim, wherein the assay is a competitive assay.
11. An assay as claimed in Claim 10, wherein the labelled, second reagent competes with a third reagent to bind to the first reagent.
12. An assay as claimed in Claim 11, wherein the overlayer includes the labelled, second reagent and the third reagent, and the response curve for that portion of the second reagent that is bound either directly or indirectly to the first reagent is inversely proportional to the concentration of third reagent within the overlayer.
13. An assay as claimed in any one of Claims 1 to 9, wherein the assay is a non-competitive assay.
14. An assay as claimed in Claim 13, wherein the assay is a sandwich assay, in which a third reagent is bound to the immobilised first reagent, and the labelled, second reagent is capable of binding to the third reagent.
15. An assay as claimed in Claim 14, wherein the assay is homogeneous, in which the labelled, secondary antibody and the third reagent are applied to the first reagent simultaneously as constituents of the overlayer, and the overlayer is not removed before taking the label intensity measurements.
16. An assay as claimed in Claim 15, wherein the assay is heterogeneous, in which the third reagent is applied to the first reagent, and then the overlayer including the labelled, secondary reagent is added to the first reagent and the bound third reagent.
17. An assay as claimed in Claim 16, wherein any unbound third reagent is removed before the overlayer including the labelled, secondary reagent is added to the first reagent and the bound third reagent.
18. An assay as claimed in any preceding claim, wherein the assay comprises an overlayer and an active layer, the active layer including a layer of the first reagent and a layer of the labelled, second reagent that is bound either directly or indirectly to the first reagent.
19. An assay as claimed in Claim 18, wherein the overlayer is of substantially greater thickness than the active layer.
20. An assay as claimed in any preceding claim, wherein the second reagent is fluorescently-labelled.
21. An assay as claimed in any preceding claim, wherein the label intensity measurements are taken using an optically sectioning microscope.
22. An assay as claimed in Claim 21, wherein the optically sectioning microscope is a confocal microscope.
23. An assay as claimed in any preceding claim, wherein the assay response curve comprises data for the label intensity at different positions along the measurement axis, and the assay response curve data is derived from individual measurements for each position along the measurement axis and/or the summation of a plurality of measurements taken in a plane for each position along the measurement axis that is perpendicular to the measurement axis.
24. An assay as claimed in any preceding claim, wherein the step of determining the contribution to the assay response curve of that portion of the second reagent that is bound either directly or indirectly to the first reagent is achieved utilising information regarding the contribution to the assay response curve of the unbound, labelled second reagent.
25. An assay as claimed in Claim 24, wherein the determination of the contribution to the assay response curve of that portion of the second reagent that is bound either directly or indirectly to the first reagent is achieved utilising information regarding the expected label intensity as a function of distance along the measurement axis, including separate coefficients for the contributions of bound and unbound second reagent.
26. An assay as claimed in Claim 25, wherein the expected label intensity information is fit to the assay response curve in order to determine the contribution to the assay response curve of that portion of the second reagent that is bound either directly or indirectly to the first reagent.
27. An assay as claimed in any one of Claims 24 to 26, wherein the response curve for that portion of the second reagent that is bound either directly or indirectly to the first reagent is determined by convolving ideal objects for (i) an active layer including the first reagent and the bound, labelled second reagent, and (ii) the overlayer, with a calculated point spread function, and making linear combinations of the two ideal objects until a fit to the assay response curve is obtained.
28. An assay as claimed in any one of Claims 24 to 26, wherein the response curve for that portion of the second reagent that is bound either directly or indirectly to the first reagent is determined by taking a series of calibration measurements along the measurement axis for (i) an active layer including the first reagent and the bound, labelled second reagent, in the absence of the overlayer, and (ii) the overlayer in the absence of the active layer; and making linear combinations of the two series of calibration measurements until a fit to the assay response curve is obtained.
29. An assay as claimed in any one of Claims 24 to 26, wherein removal of that component of the assay response curve that arises from unbound second reagent within the overlayer is achieved using a distribution function that has been fitted to experimental data for the expected contribution of the unbound, labelled second reagent to the assay response curve.
30. An assay as claimed in Claim 29, wherein the response curve for that portion of the second reagent that is bound either directly or indirectly to the first reagent is determined by: (a) taking a series of calibration measurements along the measurement axis for (i) an active layer including the first reagent and the bound, labelled second reagent, in the absence of the overlayer, and (ii) the overlayer in the absence of the active layer; (b) selecting a distribution function that substantially fits each series of calibration measurements; and (c) making linear combinations of the two distribution functions until a fit to the assay response curve is obtained.
31. An assay as claimed in Claim 30, wherein the distribution functions include a Cauchy-Lorentz function.
32. A plurality of assays as claimed in any preceding claim, wherein assays are conducted substantially simultaneously, or sequentially.
33. A plurality of assays as claimed in Claim 32, wherein the assays are provided in the same support structure, which comprises an array of support surfaces that each carry a separate assay.
34. A plurality of assays as claimed in Claim 33, wherein a plurality of support surfaces are arranged in the same x-y plane.
35. A plurality of assays as claimed in Claim 33 or Claim 34, wherein a plurality of support surfaces are separated along the measurement axis, such that the measurement axis extends through a plurality of separate assays.
36. A plurality of assays as claimed in Claim 35, wherein an array of support surfaces are arranged in one x-y plane, and one or more further arrays of support surfaces are each arranged in a different x-y plane, wherein each support surface carries a separate assay.
37. A kit for performing an assay, which kit comprises an assay including a first reagent immobilised on a support surface, and an overlayer applied to the first reagent, the overlayer including a labelled second reagent, the apparatus further comprising apparatus including means for taking a series of measurements of the label intensity of the sample along an axis that extends through both the support surface and the overlayer, so as to obtain an assay response curve, and means for determining the contribution to the assay response curve of that portion of the second reagent that is bound either directly or indirectly to the first reagent.
38. A kit as claimed in Claim 37, wherein the second reagent is fluorescently-labelled.
39. A kit as claimed in Claim 37 or Claim 38, wherein the means for taking a series of measurements of the label intensity is an optically sectioning microscope.
40. A kit as claimed in Claim 39, wherein the optically sectioning microscope is a confocal microscope.
41. A kit as claimed in Claim 39 or Claim 40, wherein the apparatus includes means for moving the sample and/or the microscope objective, such that measurements are taken at a series of positions along the measurement axis.
42. A kit as claimed in Claim 41, wherein the movement of the sample and/or the microscope objective is linear movement.
43. A kit as claimed in Claim 42, wherein the linear movement of the sample and/or the microscope objective is orientated substantially perpendicular to the support surface.
44. A kit as claimed in Claim 43, wherein the apparatus is adapted to take a plurality of measurements for each position along the measurement axis, such that the plurality of measurements are taken in a plane that is perpendicular to the measurement axis, ie the x-y plane.