Method and computer system for determining one or more parameters characterising the interaction between antigen and antibody

Abstract

The invention is a method and computer system for determining one or more parameters quantitatively characterising the interaction between an antigen and an antibody. The method comprises producing a test region carrying an antigen that produces a first luminescence spectrum, bringing into contact with the test region a sample solution comprising an antibody that forms a complex with the antigen and produces a second luminescence spectrum different from the first luminescence spectrum, recording an image (30) or images of the test region with a digital imaging device, producing a 2D histogram (40) from intensity values of the image (30), performing curve fitting on the points of the 2D histogram (40), and determining the one or more parameters by means of the fitted curve.

Description

[0001] METHOD AND COMPUTER SYSTEM FOR DETERMINING ONE OR MORE PARAMETERS CHARACTERISING THE INTERACTION BETWEEN ANTIGEN AND ANTIBODY

[0002] TECHNICAL FIELD

[0003] The present invention is a method and a computer system aimed at determining one or more parameters quantitatively characterising the interaction between an antigen and an antibody. More particularly, the invention relates to a system and a method wherein molecular interactions are measured utilising high-resolution digital imaging methods.

[0004] BACKGROUND ART

[0005] Serology is the branch of science studying the dissolved components of blood, i.e. , the quality and quantity of the molecules in blood serum and blood plasma. Functional immunomics is the study, systematization and interpretation of the extremely diverse components of the immune system and their interactions and at the biological (systemic) level.

[0006] Antibodies in circulation bind specifically to their targets, in the sense that of the millions of the encountered molecular structures they only recognise those that have been produced by the immune response leading to their production. Binding leads to the modification of the conformation of binding partners, and spontaneous binding always leads to a lower-energy state of the molecule. The greater the difference between the free energy of the initial and final states, the stronger the binding, i.e., the higher the affinity. Because binding energy varies over multiple orders of magnitude, binding is not a yes / no event but exhibits a continuum of states and is fine-tuned. Likewise, antibody specificity - specific binding to the related antigen - is also not a yes / no event but a rather complex phenomenon. Antibodies crossreact, i.e., are able bind to various targets but with different affinity.

[0007] When presented with an immunological stimulus, the organism initiates a response that gradually wears off after the cause triggering the response has disappeared. Depending on the type of the stimulus, the organism may increase the concentration of antibodies (thymus-independent and primary responses), may modify the quality of antibody effector functions (class switch), or may modify the strength of antibody binding (affinity maturation), typically making use of a combination of these processes. These characteristics of the antibody response measurable in the blood vary depending on the time elapsed between stimulus and sampling.

[0008] The number of antibody molecules being bound to the target antigen depends on three biochemical parameters: the number of binding locations (epitopes) on the antigen molecules, concentration, and affinity. In an ideal case, i.e., in the case of a simple binary interaction, an estimation of these parameters would be sufficient for a quantitative characterisation of the antibodies. The serum under examination, however, contains an unknown number of antibodies of various clonotypes of which both the concentration and the affinity are unknown. As a result of that, quantitatively determining the interactions between a large number of different antibodies and antigens presents a serious challenge.

[0009] A simple and straightforward tool for characterising antibody reactivity is titration. The serum to be examined is gradually diluted such that the initial binding signal observable during the measurement gradually decreases and finally disappears. A titration curve is obtained from these measurements, to which curve a mathematical function is fitted; the parameters of the function can be used as measurement results. However, this method is not applicable for determining biological units related to the antibodies of the serum.

[0010] Due to these challenges, in spite of some successful efforts at standardisation, serology is still in need of devising quantitative methods (Prechl, J. Why current quantitative serology is not quantitative and how systems immunology could provide solutions. BIOLOGIA FUTURA 72, 37-44 (2021 )).

[0011] Microarray analysis considers a microsurface or microspot as the physical unit of the measurement. Planar microsurfaces can be produced applying contact or nocontact microarray printing techniques. Contact printing applies pins for transferring the solutions from the source plate to the solid surface, by physically touching the surface with the pins to transfer the solution to the surface. In no-contact printing, droplets of the solution to be printed are formed, and these droplets are fed to identifiable locations on the solid surface. Contact printing produces microsurfaces with morphologies determined by the shape of the pins. Droplets produce circularshaped microsurfaces. In prior art approaches, signals in each microsurface are averaged, and the averages of the parallel microsurfaces are further averaged to obtain a single number for characterising the interaction.

[0012] Suspension microarrays may also be produced utilising beads (microbeads, microspheres, microparticles) with diameters in the micrometre range. Microbeads can be separated based on their size, colour, fluorescence, shape, or other physical properties. The antigen molecules can be attached to the surface of the separable microbeads by chemical (covalent bonds) or physical (adsorption) means. Separability of the microbeads can be ensured by labelling the molecules attached to their surface, in which case they produce a signal proportionate to the quantity of the attached molecules under appropriate measurement conditions.

[0013] As a further, deeper analytic step, the relation between antigen density distribution and bound antibody density distribution is investigated. The binding of serum antibodies to the target antigen depends both on affinity and concentration. Microsurfaces capture very low quantities of antibodies, so the reactions follow the rules of ambient analyte immunoassays. The density of bound antibodies only depends on affinity and the initial concentration of the antibodies which is practically constant. Reducing antigen density results in primarily capturing high-affinity antibodies. On the other hand, as antigen density is gradually increased, antibodies with gradually decreasing affinity are also captured. Using the concepts of physical chemistry, on the microsurfaces the chemical potential of antibodies can be examined in the presence of antigens with variable density (surface concentration).

[0014] The term “ambient analyte immunoassay” refers to immunoassays performed under special conditions. Its essential meaning is that the measurement does not significantly change the quantity of the components under examination in the test solution. Immunoassays make use of the binding specificity and binding strength of antibody-antigen interactions. The reaction conditions change over the course of these assays, because the concentration of bound molecules increases and the concentration of non-bound molecules decreases as a result of binding. In ambient analyte immunoassays, the microsurfaces of a first reactant are used in a reaction volume that contains a second reactant in a quantity that is at least two orders of magnitude greater than the quantity of the first reactant (Roger Ekins, Chapter 2.5 - Ambient Analyte Assay, Editor(s): David Wild, The Immunoassay Handbook (Fourth Edition), Elsevier, 2013, pp. 109-121 ). Consequently, the amount of the second reactant can be considered near-constant during the reaction. Therefore, ambient analyte assays are also called “mass-independent” assays because they are affected by affinity and concentration but not by mass (quantity).

[0015] The scale of the signals originating at the microsurfaces can be mathematically transformed in a way that is required for analysis. Logarithmic-linear plots are often used for characterising binding in receptor-ligand, antigen-antibody, or, in general, protein-protein interactions. These plots contain sigmoid curves, where the parameters of the equation are the minimum signal, the maximum signal, position, and the rate of signal increase. As an alternative, linear-linear and logarithmic- logarithmic plots show hyperboloid curves that, with appropriate modifications of the equations, are suitable for curve fitting. Logistic equations are most frequently applied for fitting sigmoid curves, but the rich class of growth curves can also be adapted for the given conditions.

[0016] The characterisation of immune responses against bacterial or viral infections and the immune responses following vaccination is of crucial importance in clinical practice. Antigen microarrays have been successfully applied for the serodiagnosis of infectious diseases (Mezzasoma et al., Clin. Chem. 2002, 48:121 -130). In another implementation, the published methods are useful for evaluating the characteristics of immune response against infectious agents. Obtaining biologically relevant information in antigen-microarray experiments is important for many applications, from the screening of autoimmune sera to the evaluation of vaccination efficacy; thus, the method has a wide array of possible uses.

[0017] Tests for determining the conditions of antigen-antibody complexes, for detecting the formation of the antigen-antibody complex, and for determining the quantity of the detected protein are known in literature. Such assays can be among others the Western blot assay, immunoprecipitation, immunofluorescence, immunocytochemistry, immunohistochemistry, fluorescence activated cell sorting (FACS), fluorescence in situ hybridization (FISH), immunomagnetic assays, ELISA (enzyme linked immunosorbent assay), ELISPOT (enzyme-linked immunosorbent spot) (Coligan, J. E., et al., eds. 1995. Current Protocols in Immunology. Wiley, New York.), agglutination assays, flocculation assays, cell scanning, etc., which are all well-known to the skilled person.

[0018] In the publication „Deep physico-chemical characterization of individual serum antibody responses against SARS-CoV-2 RBD using a dual titration microspot assay” c. bioRxiv 2023.03.14.532012, A. Kovacs et al. describe a fluorescent dualtitration microspot immunoassay for measuring antibody levels and neutralization potential against SARS-CoV-2 that provides the physico-chemical parameters that are both necessary and sufficient to quantitatively characterise the humoral immune response. The authors used the recombinant receptor binding domain of SARS- CoV-2 as antigen on microspot arrays and varied the concentration of both the antigen and serum antibodies from infected persons to obtain a measurement matrix of binding data. The binding curves were fitted with the help of a novel algorithm for determining the thermodynamic variables of binding. A standard state was defined for the system of serum antibodies and the antigen, and it was demonstrated how a normalised generalised logistic function can be associated with the thermodynamic activity, standard concentration, and activity coefficient.

[0019] In their publication „Absolute Quantitation of Serum Antibody Reactivity Using the Richards Growth Model for Antigen Microspot Titration” c., Sensors 2022, 22, 3962, K. Papp et al. propose novel protein microspot assays for quantifying polyclonal antibody reactivity. Like in the method set forth in the article by A. Kovacs et al., they describe a dual-titration assay developed for testing IgG and IgA binding. For evaluating the binding data of both dilution sequences two generalised logistic functions are applied for simultaneously estimating the antibody reactivity of two immunoglobulin classes.

[0020] The above-described technical solutions focus on dual titration, which involves the dilution of the antigen and antibody under examination. In these solutions, the emission by fluorescence-labelled secondary antibodies is detected after the buildup of molecular interactions. Because the fluorescent emission of the immobilised antigen is not measured, the theoretical concentration of the antigen is used for evaluating the results, so the information from inhomogeneities forming on surfaces is lost during the evaluation of the measurement data.

[0021] The relationship between antigens and antibodies is analysed from a thermodynamic point of view in the publication by J. Prechl, Statistical thermodynamics of self-organization of the binding energy super-landscape in the adaptive immune system” c., 2306.04665v4 2024, according to which from a physical perspective the humoral adaptive immune system can be characterised as a self-organising antibody binding-energy landscape. In a binding process, the “super-landscape” created by the fusion of binding energy landscapes can be described by the distribution of interaction energies and deformation parameters of thermodynamic potentials in the system. The deformation parameter characterizes the adaptive network of interactions in the system and the asymmetry of generalized logistic distributions obtained in immunoassays. The author sets forth that in a heterogeneous system, provided that heterogeneity refers to the number of different molecules making up the chemical potential vector, the deformation parameter coefficient can be used for the antigen subset of the configuration space. In the configuration space, thermodynamic states can be distinguished by vector directions, where the directions correspond to the molecular structures (conformations), and thus, to the specificities of binding. In the article, Prechl examines the behaviour of the adaptive humoral immune system from a theoretical point of view and sketches a thermodynamic model for providing a basis of serological immunoassays for a practical purpose.

[0022] This theoretical work does not contain a teaching in relation to the practical possibilities of examining inherently inhomogeneous systems effectively, i.e., obtaining as much information as possible. DISCLOSURE OF THE INVENTION

[0023] In light of the known approaches the need has arisen for providing a method and a computer system for characterising antigen-antibody interaction that can be applied in a simpler and quicker manner for estimating the affinity of serum components using universally accepted biological units and parameters.

[0024] The object of the invention is to provide a such method for assaying immunological reactivity that allows for the quantitative characterisation of the antibodies constituting the humoral immune system and is able to characterise serum antibody composition for all selected antigens using universally accepted units and parameters, in a more thorough manner with respect to prior art techniques.

[0025] A further object of the invention is to provide a novel diagnostic method and computer system that better exploit the possibilities offered by high-resolution images and digital imaging techniques.

[0026] Another object of the invention is to make use of the entirety, or as much as possible, of the wealth of information provided by the inhomogeneity that is inherently present or is produced intentionally in conventional assays. Preferably, the data reduction allowing for performing calculations quickly is also implemented differently compared to the prior art; i.e., data reduction is performed after the histogram processing of the recorded images, in such a way that retains to the greatest possible extent the information applicable for parameter determination.

[0027] The objects set in relation to the invention have been fulfilled by the method according to claim 1 , by the computer system according to claim 13, by the computer program product according to claim 14, and by the computer-readable storage medium according to claim 15. Preferred embodiments of the invention are defined in the dependent claims.

[0028] In our experiments, we have recognised that, in addition to the microarrays customarily applied for characterising antibody binding, such microspots or microbeads can also be used which carry the antigen under examination at different densities or with different density distributions. We have recognised that inhomogeneous antigen density, more precisely, inhomogeneous surface density, can be made use of appropriately as a rich source of information not only on intentionally inhomogeneously produced microsurfaces but also on other inherently inhomogeneous microsurfaces. We have recognised that the inhomogeneity of microsurfaces can be characterised by means of two-dimensional histograms, and that the mathematical analysis of the two-dimensional binding signal leads to the ability to determine general physical-chemical units and parameters. We recognised that by determining the two biochemical parameters determining the quantity of antibodies bound to the antigen, i.e., the concentration and the affinity of the antibodies, the third related parameter, i.e., molecular epitope density, also becomes measurable.

[0029] BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Preferred embodiments of the invention will be explained in the following with reference to the accompanying drawings, where

[0031] Fig. 1 is a schematic view of a test region applicable in the method according to the invention that is configured as a microarray and carries an antigen, Fig. 2 is a schematic view of a test region applicable in the method according to the invention that is configured as a microsurface or microspot and carries an antigen,

[0032] Fig. 3 shows a luminescence intensity image according to a first example and a 2D histogram based thereon,

[0033] Fig. 4 shows a luminescence intensity image according to a second example and a 2D histogram based thereon,

[0034] Fig. 5 shows a luminescence intensity image according to a third example and a 2D histogram based thereon,

[0035] Fig. 6 shows a luminescence intensity image according to a fourth example and a 2D histogram based thereon,

[0036] Fig. 7 is a sequence of 2D histograms recorded at 5, 100, and 1000-times dilutions of the sample solution, starting from the example according to Fig. 6,

[0037] Fig. 8 is a schematic depiction of antigen-microbead conjugation efficiency on a theoretical level and in a practical implementation, Fig. 9 is a schematic view of exemplary antigen-antibody binding interactions, Fig. 10 is a 2D histogram according to an embodiment comprising microbeads,

[0038] Fig. 11 depicts a 2D histogram according to another schematic example, showing selected partial regions with data reduction, and also showing the reduced 2D histogram,

[0039] Fig. 12 is an exemplary 2D histogram with a curve fitted thereto,

[0040] Fig. 13 shows microbead 2D histograms corresponding to various dilutions and curves fitted to reduced 2D histograms obtained through data reduction, Fig. 14 shows the theoretical curves of the antibody binding signal on a log- lin and a log-log plot,

[0041] Fig. 15 shows schematically the 2D histograms and the corresponding fitted curves obtained in a further example applying various dilutions,

[0042] Fig. 16 shows luminescence intensity images and 2D histograms obtained for various dilutions of two antibody types in a microarray-type test region, and

[0043] Fig. 17 is a schematic 3D model visualisation achievable with the present invention.

[0044] MODES FOR CARRYING OUT THE INVENTION

[0045] The invention is a method for determining, in a computer system, one or more parameters quantitatively characterising the interaction between an antigen and an antibody. The method according to the invention comprises the application of a test region carrying an antigen that produces in a first luminescence spectrum. The device that comprises the test region for the method can be commercially available, or alternatively it can be produced during the method. In this context, the term “test region” refers to arranging on a surface such (a) molecular element(s) that are capable of binding other molecules.

[0046] Fig. 1 shows a prior art test region 20 that is produced by printing, applying sequential dilution of antigen solutions, using known concentrations. Alternatively, this test region 20 can also be considered as an inhomogeneous region produced intentionally, for widening the scope of obtainable information. Although the test region according to Fig. 1 can be applied for the purposes of the present invention, we have recognised according to the invention that, due to the inhomogeneity inherently present in the microspots of microarrays, or even in the separately produced microspots, the antigen- 10 carrying test region 21 according to Fig. 2 comprising only a single spot or microspot can also be utilised.

[0047] The meaning of term “molecular element” is “antigen”. According to the invention, the term “antigen” includes such substances that, for example upon administration into vertebrates, are capable of eliciting an immune response, i.e., eliciting the production and release of antibodies specifically binding to them. Antigens, as defined herein, include molecules and / or components that are bound by an antibody in a specific fashion to form an antigen-antibody complex. According to the invention, antigens include, but are not limited to, peptides, polypeptides, proteins, nucleic acids, DNS, RNS, and saccharides, and combinations, fractions, or mimetics thereof. The term “autoantigen” means “self antigen”, i.e., a molecular component of the organism that elicits an antibody production response from the immune system.

[0048] The antigen 10 according to the invention has a first luminescence spectrum. In an exemplary embodiment of the invention, the antigen having the first spectrum has an autoluminescence spectrum. In another feasible embodiment, the antigen is modified, or optionally, labelled, such that it has the first luminescence spectrum.

[0049] Luminescence is a physical phenomenon in which a molecule emits light when excited; it can be for exam pie fluorescence, i.e., the emission of light at a wavelength that is different from the wavelength of the illumination applied for excitation. In some preferred embodiments according to the invention luminescence is fluorescence. Synthetic peptides with own fluorescence can be produced by incorporating appropriate amino acid residues (tryptophan, tyrosine) or real fluorophores, or by subsequent chemical bonding of fluorescent molecules. By attaching fluorophores to antigens, the antigens can be labelled in a fluorescent manner. Fluorescencebased detection has the advantage that it is sensitive and can be implemented in a multiplexed fashion, i.e., it is possible to detect a very low number of molecules, while molecules of various types can also be detected at once. According to the invention, in addition to fluorescence-based labelling, luminescent labelling and detection can also be applied.

[0050] The molecular elements are preferably arranged spatially or are separated in an identifiable manner such that at least one or more different such elements, more preferably at least 10 elements, and still more preferably at least 100 elements, or even 1000 elements can be found in a given measurable reaction. In a preferred embodiment, in the case of a spatial surface arrangement, the reaction producing the interaction may occur over a substrate surface with a surface area of a few cm2. In a preferred embodiment, the identifiable microbeads in a suspension can be present in a volume of a few hundredth cm3.

[0051] The test region is arranged on a substrate surface. The meaning of the term “substrate surface” is “solid substrate”, to which the various elements of the invention are physically attached, thereby immobilising the elements of the present invention.

[0052] The term “solid substrate” is taken to refer non-exclusively to a non-liquid material that can be a membrane, plate, a gel, glass, plastic, ceramic, metal, or mixtures thereof. The test region is preferably made on a surface made of a polymeric or elastomeric material, or other suitable membrane materials. Such materials are for example polystyrene, nitrocellulose, Nylon, or modified variants thereof. The present invention considers the use of any such substances that can be used for Northern, Southern, or Western blotting according to the prior art.

[0053] “Immobilisation” means binding the molecular elements according to the invention to the substrate surface by means of covalent bonds and / or non-covalent attractive forces, such as hydrogen bonding interactions, hydrophobic attractive forces, and ionic forces.

[0054] The substrate surfaces according to the invention are preferably capable of binding proteins at high quantities and at high spatial density, with minimal denaturation of the proteins, while at the same time - in the case of the use as set forth above - are not capable of binding such proteins that are not the proteins of interest. In the method according to the invention, during the binding to the solid substrate, the variance of binding can preferably be intentionally increased. By facilitating the diffusion of the substances (antigens) to be bound a continuous distribution is produced on the surfaces by the diffusion gradient itself. In the case of a spatial surface arrangement, the diffusion of the antigen molecules can be facilitated by increasing moisture content and by providing appropriate physical and chemical properties to the surface. For labelling microbeads, a concentration gradient can be preferably produced during conjugation in the liquid containing the beads. By increasing the variance of binding, a continuum can be provided, the heterogeneity of concentrations collectively allowing for a wide-scale analysis.

[0055] In the method according to the invention, a sample solution comprising an antibody 11 that forms a complex with the antigen and produces a second luminescence spectrum different from the first luminescence spectrum is brought into contact with the test region 20, 21.

[0056] According to the invention, the second luminescence spectrum being different from the first luminescence spectrum is taken to mean that the luminescence intensities corresponding to these spectra are separately detectable. The difference may manifest itself as the two spectra being located in wavelength ranges overlapping to the smallest possible extent, or even in non-overlapping wavelength ranges, i.e. , the wavelength ranges being separate from each other. Instead of or in addition to that, the difference may also be manifested in that the spectra have peaks at different wavelengths, but other differentiating features can also be utilized. These differences, taken either separately or combined together, make it possible to separately detect antigen and antibody surface densities.

[0057] The “samples” according to the invention may include samples of various types and / or origin, such as blood samples and other fluid samples of biological origin. The term “sample” is taken to cover clinical samples and also includes cell supernatants, cell lysates, serum, plasma, biological fluids, and pure or enriched samples originating from any of these. The sample may come from microorganisms, such as bacteria, yeast, viruses, prions, mould, fungi, from plants, and from animals, including mammals, for example humans. These samples can be produced applying methods known from the prior art, such as lysis, fractioning, purification, including affinity purification, FACS (Fluorescence Activated Cell Sorting), laser microdissection, or preparative centrifugation.

[0058] In the sense it is used in the present invention, the term “antibody” refers to polyclonal or monoclonal antibodies or immunoglobulin molecules, which belong to the group of glycoprotein-type molecules. Furthermore, the term “antibody” also refers to intact immunoglobulin molecules, chimeric immunoglobulin molecules, or Fab or F(ab')2 or scFv or single-domain antibody fragments. Such antibodies and antibody fragments can be produced applying well-known techniques, including those set forth by Harlow and Lane (Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. (1989)) and Kohler et al. (Nature 256: 495-97 (1975)), or the techniques disclosed in the documents US 5,545,806, US 5,569,825, and US 5,625,126.

[0059] The antibodies included in the present invention may be of any isotype (i.e., in humans, according to their heavy chains, belonging to the classes IgG, IgA, IgD, IgE and IgM, and to the subclasses lgG1 , lgG2, lgG3, lgG4, lgA1 , lgA2, and according to their light chains to the kappa or lambda classes). In the present invention, antibodies include any glycoform, i.e., any molecular structure modified with carbohydrate side chains. The antibodies produced in vertebrates, their chains, their isotypes and glycoforms are known to the skilled person, and are also objects of the present invention.

[0060] According to the present invention, the term “binding” refers to antigen / antibody binding, and other non-random association between antigens and antibodies. In this description, the term “specifically bound” refers to such antibodies or other ligands that do not substantially cross-interact with antigens other than the given antigen or antigens. For example, it may refer to the “specific bonding” of an antibody to a group of antigens sharing a common epitope. Epitopes are surface portions in a molecule that produce the non-covalent interactions with the corresponding surface portion of the antibody molecule.

[0061] In an exemplary embodiment of the invention, the antigen 10 has an autoluminescence spectrum being the first luminescence spectrum. In an alternative embodiment, the antigen 10 is modified or labelled such that it produces the first luminescence spectrum. In an exemplary embodiment of the invention the antibody producing the second luminescence spectrum has an autoluminescence spectrum. In another preferred embodiment, the antibody is modified and / or labelled such that it produces the second luminescence spectrum.

[0062] The secondary antibodies applied for detecting bound serum antibodies can be fluorescence-labelled by chemical conjugation of fluorescent dyes. Other affinity reagents adapted for fluorescence detection of serum antibodies are aptamers, affimers, various scleroproteins, recombinant antibody receptors, and other appropriately designed and labelled molecules.

[0063] The term “secondary antibody” used according to the invention refers to antibodies produced in a given species against antibodies of another species. Utilising secondary antibodies, the antibodies originating from the latter species can be detected / labelled in immunoassays. Murine monoclonal anti-human antibodies and goat polyclonal anti-human antibodies are examples for secondary antibodies recognising human antibodies.

[0064] According to the invention, “polyclonality” refers to the serum antibodies having multiple origins, in the sense that multiple genetically different B-cells (clones) contribute to producing antibodies against the antigen in the organism. Polyclonality contributes to the complexity of serum antibody-binding reactions, and also to reference antibody products (monoclonal, polyclonal and their mixtures) being impossible to be applied as a resource for using proper biochemical units in serological assays (i.e., presently, arbitrary units are used instead).

[0065] In an exemplary embodiment of the invention, the test region 20, 21 is formed on a flat, fixed surface. In other embodiments, the test region 20, 21 may be disposed on the surface of microbeads 22 (see Fig. 8). Of course, test regions 20, 21 implemented in other ways are also conceivable.

[0066] In the course of bringing the test region 20, 21 into contact with the sample solution, the formation of complexes between the antigens and the antigen-recognising antibody molecules can be brought about under various circumstances. Preferably, physiological pH, salt concentration and temperature are applied. By controlling the conditions of this incubation step, the bindable or non-bindable components can be intentionally controlled. Dilution of the sample facilitates the binding of high-affinity substances. Conditions favourable for the activation of the complement system result in the accumulation of complement components that have a serious effect on the cell recognition of the complexes formed in such a manner.

[0067] The analysis of the antibodies bound to the test region carrying the antigen can be quantitative, semi-quantitative, or qualitative. According to the invention, “detection” refers to detecting the presence, absence, and / or quantity of the complexes to be detected. In the context of this description, “absence of binding” and “absence of detection” also includes insignificant detection levels.

[0068] After bringing the sample solution into contact with the test region 20, 21 , the test region 20, 21 is preferably excited applying monochromatic light; in an exemplary preferred embodiment the excitation is carried out applying a laser or LED light source. Preferably, there is no overlap, or there is a minimal overlap between the excited and emitted wavelength ranges of the antigen and the antibody.

[0069] After applying the excitation, an image or images of the test region 20, 21 are taken with a digital imaging device, the image or images including first luminescence intensity values determined by the first luminescence spectrum and second luminescence intensity values determined by the second luminescence spectrum. In an exemplary embodiment, the first luminescence intensity value and the second luminescence intensity value are recorded in the same image such that excitations being the first luminescence spectrum and the second luminescence spectrum are applied simultaneously. In another embodiment, the first luminescence intensity value and the second luminescence intensity value are recorded in two separate images such that excitations being the first luminescence spectrum and the second luminescence spectrum are applied separately.

[0070] According to the invention, the term “image analysis” refers to obtaining high- resolution images of planar or spatial microsurfaces by scanning the surfaces applying a fluorescence laser scanner or a CCD camera, whereby digital images having a resolution on the order of micrometres / pixel can be obtained. In these images, the measurement results preferably being 1 -100 micrometres of a solid surface or microbeads with a size of 0.1 -10 micrometres. The analysis of the signal intensities of measured events, their relative and absolute coordinates, and the signal patterns of various fluorescent channels is called image analysis or data analysis.

[0071] In the present invention, the intensity values of individual pixels are handled separately, allowing for an analysis of signal intensity distribution. The inhomogeneity of the antigen on the microsurfaces thus provides information that would otherwise be lost if conventional averaging approaches were applied.

[0072] On the left of Figs. 3, 4, 5, 6, luminescence intensity images 30, 31 , 32, 33 determined by the first and second luminescence spectra are shown that are taken of the test regions 20, 21 disposed on flat surfaces applying a digital imaging device.

[0073] On the right of Fig. 3, a 2D histogram 40 is included, showing points having x and y coordinates that are determined by a respective pair consisting of a first luminescence intensity value and of a second luminescence intensity value that belong to each other on the basis of the test region. With the help of 2D histograms, inhomogeneities can be visualised more prominently. On the right of Figs. 4, 5, 6 the 2D histograms 41 , 42, 43 corresponding to the luminescence intensity images 31 , 32, 33 are shown. In an exemplary embodiment of the invention, the 2D histogram has log-scale x and y axes. The signals of the two fluorescent channels (x axis: antigen, y axis: antibody) are shown on a log scale.

[0074] In the 2D histogram shown on the right of Fig. 5, a rectangular selection area identifies pixels that have antigen and antibody fluorescence values in the given range. In the luminescence intensity image 32 shown on the left there can be seen that pixels corresponding to this range can be found in almost every region in the nominal dilution sequence of the antigen. Due to the surface diffusion of the antigen, antigens with non-nominal density can also be seen in each region. Using the fluorescence of the antigen instead of its nominal concentration, a more accurate image of the quantitative relationships between antigen and antibody can be obtained. 2D histograms visualise two characteristics of a digital image assignable to the pixels of the image, quantifying the frequency of occurrence of the pixels. It preferably shows the number of pixels having a given value according to the coordinates of signals measured in two fluorescent channels. The number of pixels represents the third dimension in the image; it is usually shown in a colour-coded manner, so these images are also referred to as 2D heat maps.

[0075] In a particularly preferred embodiment, during recording the image 30, 31 , 32, 33 or images, the relative position of the test region 20, 21 and the digital imaging device remains the same. This allows for obtaining per-pixel information in a simple manner, such that each pair belongs to a respective pixel, and each pair consists of a first luminescence intensity value and a second luminescence intensity value measured in the pixel corresponding to the pair.

[0076] Fig. 7 illustrates the effect of serum dilution on the binding patterns; 2D histograms corresponding to 5-fold, 100-fold, and 1000-fold dilution are shown, respectively, on the left, in the centre, and on the right of the figure. The dilution of the serum being examined results in a decrease of antibody concentration; accordingly, the signals of the given fluorescent channel also decrease. By diluting various serums to an identical extent, a comparison can be made between their binding patterns; by performing measurements detecting different antibody classes after a reaction, the different bindings thereof can also be shown.

[0077] In an exemplary embodiment, the test region formed on surfaces of microbeads is examined utilising a flow cytometer. Fig. 8 demonstrates the conjugation of a luminescence-labelled antigen to microbeads. At the top of Fig. 8 the theoretical conjugation efficiency based on the nominal concentration is shown, while at the bottom of the figure the practical results are depicted, illustrating intermediateefficiency conjugation. In the embodiment according to the invention, instead of increasing the digital resolution of smaller surface regions (microspots) in a large contiguous region (i.e., adding more pixels), surfaces with micrometre-scale physical size, preferably microbeads are applied. The density of molecules affixed to the surface of these microbeads is modulated by the efficiency of chemical conjugation. As with planar microarrays, heterogeneities and deviations from the nominal concentration occur also in this case due to the randomness of molecular phenomena.

[0078] In the embodiment with microbeads that is schematically shown in Fig 8, the test region 20, 21 is formed on the surface of microbeads 22, with each pair consisting of a first luminescence intensity value and a second luminescence intensity value corresponding to an image representation of a given microbead 22. The parameters of the image recorded applying a digital imaging device are preferably set such that the image representation of the microbead 22 corresponds to a single pixel. Thereby, analogously to the process described above, image processing is simplified to processing pixels. In certain cases, the image of a particular microbead 22 may cover more than one pixel, in which case the combined detected intensity value of the microbead 22 can be calculated by simple summation.

[0079] “Multiplexity” refers to the dimensionality, i.e., the diversity of simultaneously performed measurements. Deeper analyses are based on simultaneously acquiring large amounts of different types of data. Multiplex measurements can involve simultaneously measuring multiple antibody classes (IgA, IgG, IgM, IgE) or subclasses (lgG1 , lgG2, lgG3, lgG-4) or simultaneously testing for multiple antigens (e.g. various virus proteins) or for different epitopes of the same antigen. For example, multiplex measurements can be technically implemented by exploiting the channels of fluorescence wavelengths, applying reagents and microbeads with variable fluorescence characteristics, for example in a Luminex system.

[0080] In Fig. 9 a schematic example for the conjugation of a fluorescence-labelled antigen is illustrated for various densities in the case of microbeads 22. In an exemplary embodiment of the invention, secondary antibodies 12 are applied for labelling the antibodies 11 that produce complexes with the antigens 10.

[0081] The 2D histograms shown in Fig. 10 and on the left of Fig. 13 present results related to test regions formed on the surface of microbeads 22. In the 2D histogram shown in Fig.10 fluorescence intensity values corresponding to fluorescent peptides conjugated to microbeads with a diameter of 1 micrometre in a serial dilution, with arrows indicating the fluorescence signals corresponding to nominal conjugation peptide concentrations. Due to the intermediate efficiency, in actual practice the conjugates corresponding to the nominal concentration form a continuum, so measurement data can be obtained also from the intermediate regions without interpolation. Signals of the fluorescent antigens vary over an intensity range of 3 magnitudes.

[0082] Using a wide range of antigen densities makes it possible to quantitatively analyse the dependency of antibody binding on antigen density. By fitting appropriate mathematical functions to the binding curves, the parameters of the binding curves can be identified. The parameter values can be used for characterising serum antibody affinity and serum antibody concentration. It should be noted that, because antigen density is the “known”, i.e. , the controlled variable, the parameters related to antigen density yield values in the universal biochemical units of concentration, and in units derived therefrom. Both the nominal / estimated and the accurate / measured antigen density values can be used for the calculations.

[0083] The method according to the invention preferably comprises performing data reduction on the 2D histogram prior to curve fitting, and carrying out the curve fitting on the 2D histogram reduced by means of the data reduction. In a preferred embodiment of the invention, partial regions are defined in the 2D histogram during data reduction, the points in each partial region along the x and y axes are averaged, and a point corresponding to the average obtained in this manner is substituted for the points in the given partial region. According to an exemplary preferred embodiment of the invention illustrated on the left of Fig. 11 , the 2D histogram 46 is partitioned into partial regions with an identical rectangular shape that may optionally overlap. The result of the corresponding data reduction, i.e., the reduced 2D histogram 47, is shown on the right of Fig. 11 . Of course, any other partial region partitioning with alternative shapes or sizes is also conceivable; these parameters may expediently correspond to the given application or to the given 2D histogram. Fig. 12 illustrates curve fitting performed on a reduced 2D histogram plotted on a log-log scale that results in a fitted curve 50. Conjugation and binding variances produce a continuum in the data.

[0084] Fig. 13 illustrates a microbead example wherein curves are plotted on the 2D histogram 48 reduced applying data reduction, namely for the different dilutions of the antigen 10 shown to the right of the curves. In the measurement, microbeads conjugated with fluorescent peptides were mixed to the serum reacting with the given antigen, followed by determining the fluorescence signals of the bound antigens.

[0085] Fig. 14 schematically illustrates the theoretical shape of the binding signal of the antibody 11 as a function of antigen density according to a generic logistic function. On the left and right of Fig. 14, the function is plotted on a log-linear scale and a loglog scale, respectively. In the figure, Xi stands for inflection point, max stands for maximum signal, d stands for asymmetry parameter, and s stands for initial slope.

[0086] According to the invention, “curve fitting” means that the measurement results, i.e. , the binding signals, can be plotted as a function of antigen density. These experimental data follow theoretical mathematical functions that can be fitted to the data. Curve fitting is computationally intensive, and is performed applying computers and appropriate computer software. The sigmoidal binding curves are usually fitted to logistic functions and to generalised logistic functions. Depending on the mathematical transformations of the binding data, different functions can be fitted if necessary.

[0087] Fig. 15 shows fitted curves 51 , 52, 53, 54, 55, 56, 57 obtained with sample solutions of different dilutions; the one or more parameters quantitatively characterising the interaction between the antigen 10 and antibody 11 are determined with the help of these curves. Preferably, all fitted curves 51 , 52, 53, 54, 55, 56, 57 are used for determining the parameters.

[0088] We have recognised that the curvature of the 2D histogram curve shows a relationship with the signal-decreasing effect of the dilution of the serum: the larger the curvature, the slower the signal intensity decreases with serum dilution at the inflection point. Using a generalised logistic function in the model of antigen to antibody binding, the curvature of the curve observable in the log-log plot, more precisely, the slope of the leftmost asymptote, is determined by the asymmetry parameter nu. In case the logarithmical, generalised logistic function is normalised such that the function’s value in the inflexion point is 1 , and that the effect of serum dilution is modelled with a similar function; the intensity of fluorescence signals (Fl) is given by the following formula: where Cn is the fluorescence signal at the inflection point, x is the molar concentration of the antigen, y is the reciprocal of the serum dilution factor (relative concentration), nu is the asymmetry parameter of antigen titration, and nu’ is the asymmetry parameter of serum dilution.

[0089] With the dilution of a serum being examined, fluorescence decreases; we have recognised that the extent of this decrease is determined by the curvature of the antigen titration curve.

[0090] If the effect of serum dilution can be expressed from curvature change in the function applied for modelling fluorescence signals (fitting function), the serum dilution step can also be omitted. “Dilution” refers to the steps of a titration process, and titration characterises the interactions of the molecules under examination. Conventional serum titration processes only characterise relative relationships, i.e., midpoint titration searches for a twofold decrease of signal intensity on the saturation curve given by a sigmoid function. A titration process described by a generalised logistic function searches for the inflection point in the curve shape, which, unlike in the case of the logistic function, need not be at the midpoint.

[0091] Therefore, during antigen titration, the relative quantities of the antigens are signified not only by signal intensity but also by signal curvature in the 2D histogram. If the effects of serum dilution are determined by the value of the asymmetry parameter, then serum dilution behaviour and antibody quantities in the serum can be calculated even from a single serum dilution data point.

[0092] In case nu’ can be expressed using nu in the equations describing the collective titration of the serum and the antigen, i.e. nu’=f(nu), then serum titration is only required for determining yi. However, the value yi that identifies the inflection point of serum titration as expressed in relative concentration also exhibits a relationship with nu, because the fluorescence changes accompanying serum antibody titration are determined collectively by yi and nu’. The smaller the value of yi, the higher the antibody titers, because higher dilutions can be applied until the binding signal disappears. In contrast to that, the higher the value of nu’, the smaller the slope of the serum dilution curve, i.e., the higher serum dilutions can be applied until the serum binding signal disappears. That is, yi and nu’ change in opposite directions, which can be expressed in the following formula:

[0093] Using this parametrisation, the following immunological interpretation of the parameters can be provided:

[0094] Inxi is the logarithm of the molar antigen concentration at the inflection point of the antigen titration curve, which corresponds to the equilibrium dissociation constant (KD); nu’ is the asymmetry parameter of serum titration, which corresponds to relative thermodynamic concentration ([Ab] / KD);

[0095] Cn, is the luminescence signal measurable at the intersection of the inflection points determined on both titration axes, which includes specific fluorescence (fluorescence / antibody molecule), the equilibrium concentration of bound antibodies ([Ab]bound) and molecular epitope density.

[0096] Therefore, with identical Inxi and nu’ values and at a given specific fluorescence, different Cn values must be caused by different epitope intensities. We have therefore recognised that by deconvolving the antibody concentration and affinity determining the fractional saturation of the antigen, the molecular epitope density of the antigen can be characterised, i.e., a parameter that conventional serological methods have yet been unable to measure directly. Importantly, this parameter has a relevant immunological implication, as it makes it possible to separate the active but less specific (lower-affinity antibodies targeting a large number of epitopes) and the resting memory (higher-affinity antibodies targeting a few dominant epitopes) phases of the immune response.

[0097] Measurements performed on microsurfaces constitute a special measurement technique from the aspect of the quantitative relationships of interacting molecules: the component in solution is present in a practically infinite amount, because it is present in the solution at such an excess quantity that the binding of a fraction of it to the surface will not result in a significant change of the solution concentration. Therefore, the titration of the component adsorbed to the surface and the titration of the component in solution yield information on completely different thermodynamic constituents of the interactions. It is possible to differentiate enthalpic and entropic components of the binding of the molecules: the former includes the produced non- covalent bonds, while the latter covers the entropy change occurring during the binding process. In our measurement, the produced non-covalent bonds depend on the structure of the antibodies present in the serum and reflect their qualities. Entropy changes depend on serum composition and antibody diversity. By varying the antigen density on the microsurface, we examine how the “binding force” of serum antibodies changes the quantity of bound antibodies - independently of concentration -, which is indicated by the intensity of the signals. By diluting the serum, in turn, the composition, i.e., the entropic effect, can be varied during the measurement. The variations of enthalpy and entropy collectively determine the variation of free energy which in turn determines the outcome of equilibrium titration, i.e., the number of antibody molecules remaining bound to the surface.

[0098] AG = AW - TAS

[0099] Thus, in our measurement we examine the contribution of these two terms of free energy, reflected in our model by the asymmetry parameters

[0100] V = e&H~&Gv' =e&s~&H which, normalised for the change of free energy, is

[0101] , -1~ v =v

[0102] V

[0103] Using this relation, by analysing the high-resolution image of a single antigen spot (Fig. 3), not only the apparent affinity (~logxi=logKD’) can be determined but also the chemical potential of the antibodies, which depends on their quantity (relative concentration). By fitting the formula we obtain the value of nu, from which, according to the above, the values of both nu’ and yi, that is, a relative serum antibody concentration is the thermodynamic titer that can be calculated.

[0104] Fig. 16 illustrates an example wherein the test region comprises partial test regions that are arranged in a microarray and carry antigens with different average surface concentrations. The luminescence intensity images are shown on the left of the figure; the luminescence intensity image 34 showing the IgA heavy-chain antibodies of the serum 184, and the luminescence intensity image 35 showing the IgA lightchain antibodies of serum 184. Both images show different dilutions of the sample solution, i.e., from top to bottom, 5-fold, 25-fold, and 125-fold serum dilution. The 2D histograms 44, 45 corresponding to the luminescence intensity images are shown on the right of the figure. The luminescence intensity image 34 shows stronger signals (IgA), and correspondingly the shape of the signal curve in the 2D histogram 44 has a more prominent curvature. The curvature can also be observed at higher dilutions.

[0105] Preferably, more than one test regions 20, 21 can be applied that are brought into contact with respective sample solutions of different dilutions, or we can apply one test region 20, 21 that is successively brought into contact with sample solutions of different dilutions. In such a way, the amount of information that can be obtained with the 2D histogram step can be increased in a simple manner. It is particularly preferable if the points obtained from all test regions 20, 21 are plotted in a common 2D histogram, performing a respective curve fitting for points corresponding to each dilution, and utilising all curves for determining the parameters.

[0106] Fig. 17 illustrates a visualisation of the above-mentioned model applying three axes. Enthalpic effects and entropic effects, respectively, are traced by means of the titration of the density of immobilised antigen and the titration of serum concentration, which collectively determine the intensity variation of fluorescence signals proportionate to free energy. In both figures, logxi=-12 and logC=9. The values of nu=0.6 (left) and nu=0.3 (right) are, however, different.

[0107] In sum, by combining the known antigen density difference values and the measured bound antibody density difference values, which is quantified by optical detection and image analysis of antigen microsurface density patterns, and by applying appropriate mathematical algorithms and curve fitting, a quantitative biochemical description of serum antibody binding can be provided. It is possible to determine the quality (apparent affinity) and the relative concentration (titer) of the antigen-recognising antibodies in a given dilution from that single dilution point, and from these two data points the chemical potential of the serum antibodies against the tested antigen can be calculated. So, in case the inflection point obtained through antigen dilution (xi) is considered as the apparent (i.e. , determined through measurement) equilibrium constant, the relative titer value can also be converted to an absolute scale, which allows for making quick and effective comparisons of the results of single-chamber measurements, i.e., measurements performed without titrating the serum under examination.

[0108] Based on the recognitions described in detail above, novel diagnostic methods have been developed in conjunction with devices that are able to quickly carry out the methods (e.g. diagnostic methods). The novel devices according to the invention are kits, preferably diagnostic kits that are capable of detecting the disease-related changes of the immune profile. The kit comprises an antigen density array with variable surface antigen density; reagents for carrying out the method; and calculation methods for expressing the serum component reactivity under examination using biochemical units derived from and related to antigen density units and express them as antigen density units. In such a diagnostic kit the antigens included in the antigen microarray can be parasites, microbes, viruses, prions or components thereof. Likewise, in a serological test kit the antigens can be proteins, synthetic peptides, glycoproteins, lipoproteins, lipids, glycolipids, nucleic acids, carbohydrates, and small molecules that are selected from the groups of allergens or autoantigens. In the measurements, the applied samples can be clinical samples, plasma, biological fluids, cells in culture, cell supernatants, cell lysates, or pure or enriched samples originating from any of the foregoing. The detection agent can be fluorescence-labelled or luminescence-labelled antibodies, aptamers, or other affinity reagents that recognise the serum antibodies.

[0109] The diagnostic kit according to the invention is adapted to be applied in the method according to the invention, and comprises a test region 20, 21 that is disposed on a substrate surface and carries an antigen 10 producing a first luminescence spectrum.

[0110] The computer system according to the invention comprises units programmed for carrying out the steps of the method described above. The units can be hardware units, software units, or a combination of hardware and software units.

[0111] The computer program product according to the invention comprises instructions which, when executed by a computer, cause the computer to carry out the steps of the method described above.

[0112] The computer-readable storage medium according to the invention comprises instructions which, when executed by a computer, cause the computer to carry out the steps of the method described above.

[0113] Industrial applicability of the invention follows from the features of the technical solution according to the description above. As it is apparent from the description above, the invention fulfils its objectives in an extremely preferable manner compared to the prior art. The invention is, of course, not limited to the preferred embodiments described in detail above, but further variants, modifications and developments are possible within the scope of protection determined by the claims.

Claims

CLAIMS1. A method for determining, in a computer system, one or more parameters quantitatively characterising an interaction between an antigen (10) and an antibody (11 ), characterised by comprising the steps of:- applying a test region (20, 21 ) that carries an antigen (10) producing a first luminescence spectrum,- bringing into contact with the test region (20, 21 ) a sample solution comprising an antibody (11 ) that forms a complex with the antigen (10) and produces a second luminescence spectrum different from the first luminescence spectrum,- after the step of bringing into contact, recording an image (30, 31 , 32, 33, 34, 35) or images of the test region (20, 21 ) with a digital imaging device, the image (30, 31 , 32, 33, 34, 35) or images comprising first luminescence intensity values determined by the first luminescence spectrum and second luminescence intensity values determined by the second luminescence spectrum,- in a 2D histogram (40, 41 , 42, 43, 44, 45, 46), creating points each having x and y coordinates that are determined by a respective pair consisting of a first luminescence intensity value and of a second luminescence intensity value that belong to each other on the basis of the test region (20, 21 ),- performing curve fitting on the points of the 2D histogram (40, 41 , 42, 43, 44, 45, 46), and- determining the one or more parameters by means of the fitted curve (50, 51 , 52, 53, 54, 55, 56, 57).

2. The method according to claim 1 , characterised in that the antigen (10) has an autoluminescence spectrum being the first luminescence spectrum, and / or the antigen (10) is modified or labelled such that it produces the first luminescence spectrum.

3. The method according to claim 1 or 2, characterised in that the antibody (11 ) has an autoluminescence spectrum being the second luminescence spectrum, and / orthe antibody is modified or labelled such that it produces the second luminescence spectrum.

4. The method according to any of claims 1 to 3, characterised in that during recording the image (30, 31 , 32, 33, 34, 35) or images, the relative position of the test region (20, 21 ) and of the digital imaging device remains the same, each pair belongs to a respective pixel, and each pair consists of a first luminescence intensity value and of a second luminescence intensity value measured in the pixel belonging to the pair.

5. The method according to any of claims 1 to 4, characterised in that the test region (20, 21 ) is formed on surfaces of microbeads (22), and each pair consisting of a first luminescence intensity value and of a second luminescence intensity value corresponds to an image representation of a given microbead (22).

6. The method according to any of claims 1 to 5, characterised in that- the first luminescence intensity values and the second luminescence intensity values are recorded in the same image such that excitations corresponding to the first luminescence spectrum and to the second luminescence spectrum are applied simultaneously, or- the first luminescence intensity values and the second luminescence intensity values are recorded in two separate images such that excitations corresponding to the first luminescence spectrum and to the second luminescence spectrum are applied separately.

7. The method according to any of claims 1 to 6, characterised by applying a 2D histogram having logarithmically scaled x and y axes.

8. The method according to any of claims 1 to 7, characterised by performing data reduction on the 2D histogram (40, 41 , 42, 43, 44, 45, 46) prior to curve fitting, and carrying out the curve fitting on the 2D histogram (47, 48, 49) reduced by means of the data reduction.

9. The method according to claim 8, characterised in that during data reduction partial regions are defined in the 2D histogram, the points in each partial region areaveraged along the x and y axes, and the points in the given partial region are substituted by a point corresponding to the average obtained.

10. The method according to any of claims 1 to 9, characterised in that the test region (20, 21 ) comprises partial test regions that are arranged in a microarray and carry antigens (10) with different average surface concentrations.11 . The method according to any of claims 1 to 10, characterised by applying more than one test regions (20, 21 ) that are brought into contact with respective sample solutions of different dilutions, or applying one test region (20, 21 ) that is successively brought into contact with sample solutions of different dilutions.

12. The method according to claim 11 , characterised by plotting the points obtained from all test regions in a common 2D histogram, performing a respective curve fitting for points corresponding to each dilution, and utilising all curves for determining the parameters.

13. A diagnostic kit for use in the method according to claim 1 , the diagnostic kit comprising a test region (20, 21 ) that is disposed on a substrate surface and carries an antigen (10) producing a first luminescence spectrum.

14. A computer system characterised by comprising units programmed for carrying out the steps of the method according to claim 1 .

15. A computer program product characterised by comprising instructions which, when the program is executed by a computer, cause the computer to carry out the steps of the method according to claim 1 .

16. A computer-readable storage medium, characterised by comprising instructions which, when the program is executed by a computer, cause the computer to carry out the steps of the method according to claim 1 .

Images