Method for diagnosis of high-affinity binders and marker sequences

Abstract

The present invention relates to a novel method for diagnosing high-affinity binders, in particular antibodies or autoantibodies, and the identification, characterization and selection of marker sequences and diagnostic use thereof, in particular in the form of a panel. The invention also relates to a singleplex assay in which the discovered selection of marker sequences is used in the form of a panel and high-affinity binders are detected using a single signal.

Claims

1. A method for the identification and/or selection of marker sequences suitable for use in a singleplex assay for detection and/or diagnosis and/or stratification of a disease in a subject, comprising the following steps: a) mixing n marker sequences placed on a solid substrate with a sample containing high-affinity binders and detecting a single signal resulting from the interactions between the n marker sequences and the high-affinity binders, b) mixing at least n1 marker sequences having statistically sufficient signal intensity identified from a) with said sample containing high-affinity binders and detecting a single signal resulting from the interaction between the at least n1 marker sequences and the high-affinity binders, c) optionally repeating step b) with at least n-k marker sequences, wherein k>1, and d) selecting nk marker sequences having statistically sufficient signal intensity identified from b) or c) for use in a singleplex assay for detection and/or diagnosis and/or stratification of a disease in a subject, wherein the nk marker sequences selected in d) comprise at least 4 different marker sequences.

2. The method of claim 1, wherein the marker sequences are obtained from biological material.

3. The method of claim 2, wherein the biological material is selected from the group consisting of tissue, native sources, cells, bacteria, viruses, phages, prions, plants, animals and humans.

4. The method of claim 1, wherein the marker sequences are mRNA, si-RNA, microRNA, cDNA, peptide, protein, or originate from an expression library.

5. The method of claim 4, wherein the protein is an antigen or an autoantigen.

6. The method of claim 4, wherein the expression library is an mRNA, si-RNA, microRNA, cDNA, peptide or protein expression library.

7. The method of claim 1, wherein the high-affinity binders are antibodies.

8. The method of claim 1, wherein the high-affinity binders are autoantibodies.

9. The method of claim 1, wherein the n-k marker sequences selected in d) are applied to a substrate.

10. The method of claim 9, wherein the substrate is a filter, a membrane, a magnetic or fluorophore labelled bead, a silicon wafer, glass, metal, plastic, a chip, a mass spectrometry target or a matrix.

11. The method of claim 1, wherein the sample containing high-affinity binders is a bodily fluid or tissue extract.

12. The method of claim 11, wherein the high affinity binders are antibodies or autoantibodies and the bodily fluid is blood, whole blood, blood plasma, blood serum, patient serum, urine, cerebrospinal fluid, or synovial fluid.

13. The method of claim 1, wherein the signals are detected with the aid of radioactive or fluorescently labelled antibodies by means of a bioanalytical method or a mass spectrometry method, wherein the bioanalytical method is optionally Western blotting (1D and 2D), immunohistochemistry, antibody arrays, Luminex, ELISA, immunofluorescence, radioimmunoassays, and wherein the mass spectrometry methods are optionally MRM (multi reaction monitoring) or AQUA (absolute quantification).

14. The method of claim 1, wherein the n-k marker sequences selected in d) comprise at least 10 different marker sequences.

15. The method of claim 1, wherein the marker sequences are antigens, parts of antigens, haptens or proteins.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIGS. 1 to 4 show the mean value MFIs of a multiplex assay (a number of marker sequences deliver a number of individual signals) and singleplex assays in comparison, wherein 10 different antigens (proteins 1 to 10) and the serums RA00029 (FIG. 1), RA0037 (FIG. 2), RA00046 (FIG. 3) and PRO244 (FIG. 4) were used. The individual values for proteins 1 to 10 were measured in the multiplex assay and the mean values tar the multiplex assay were calculated from this and compared with the measured value (single signal) for the singleplex assay.

(2) FIG. 5 shows the comparison of the median MFIs of multiplex and singleplex assay, wherein 10 different antigens (proteins 1 to 10) were used. 10 l/ml (median of the counted heads) were used as a coupling control. Set 1 (multiplex assay) with 10 proteins at different bead regions is compared with set 2 (singleplex assay) with 10 proteins at the same bead region.

(3) FIG. 6 shows the comparison of the mean value MFIs of multiplex and singleplex assay, wherein 10 different antigens (proteins) were used. 10 l/ml (mean value of the counted beads) were used as a coupling control. Set 1 (multiplex assay) with 10 proteins at different bead regions is compared with set 2 (singleplex assay) with 10 proteins at the same bead region.

(4) FIG. 7 shows the spots on a microarray on which proteins and protein mixtures (marker sequences) are immobilised, said proteins acting as antigens and having been incubated with serum. RA00037.

(5) FIG. 8 shows the spots on a microarray on which proteins and protein mixtures (marker sequences) are immobilised, said proteins and protein mixtures acting as antigens and having been incubated with serum RA00033.

(6) FIG. 9 shows the calculated mean value compared with the measured value of the signal intensities with use of serum RA00037 with different proteins and protein mixtures as marker sequence or panel of marker sequences.

(7) FIG. 10 shows the calculated mean value compared with the measured value of the signal intensities, with use of serum RA00033 with different proteins and protein mixtures as marker sequence or panel of marker sequences.

(8) FIG. 11 shows the result of microarray experiments. The signal intensities of the used serums (RA0026 so RA0046) with the individual used marker sequences P1 to P9 and P0 are specified.

(9) The invention will be explained hereinafter by examples. The invention is not limited to the examples however, and in principle high-affinity binders, types of assays and detection methods can be applied universally for a wide range of marker sequences.

(10) ELISA and PUPA Luminex assays are used in the examples in addition to antigens (proteins) as marker sequences and also serums that contain antigens and autoantigens as high-affinity binders.

(11) Assays based on Luminex technology allow the simultaneous quantitative determination of up to 100 parameters in an individual sample, wherein multiplex assays are known in the prior art. Advantages of the multiplex format are high sensitivity and specificity, exact quantification and suitability for automation.

(12) A preferred embodiment of the invention concerns singleplex assays (single-signal assays). In the singleplex assay (single-signal assay) according to the invention, only one parameter is determined in contrast to the multiplex assay. In the singleplex assay according to the invention, it is not a selected marker sequence that is used, but a panel of marker sequences obtained by combining a plurality of marker sequences to form a single marker (panel, of markers, marker combination). An advantage with the use of a panel of marker sequences in a singleplex assay is the additional signal amplification caused by the combination of the marker sequences in a single panel. In the singleplex assay, the mean value of the intensity of the combined marker sequences is then preferably measured. Marker sequences will also be referred to hereinafter simply as markers.

Example 1: Additional Signal Amplification by Marker Combination in a Single Panel of Marker Proteins

(13) Theoretical signal intensity, preliminary consideration

(14) In a singleplex assay with a panel of marker sequences formed from protein 1 and protein 2 (2-way combination), in which protein 1 has a signal intensity of 10,000 and protein 2 has a signal intensity of 30,000, the calculated mean value of the intensity MFI of the signals from protein 1 and protein 2 is 20,000.

(15) In a singleplex assay with a 5-way combination of protein 1, protein 2, protein 3, protein 4 and protein 5 in a panel, the calculated signal intensity of the panel of marker sequences corresponds to the mean value of the individual signals of the 5 proteins.

(16) Test Execution

(17) The used beads are coated with the proteins acting as antigen. These coated beads are incubated with different serums, such that the antibodies present in the serum can bind to the antigens. This involves comparing the signal intensities of a multiplex measurement with the signal intensity of the singleplex measurement according to the invention.

(18) In the multiplex assay 10 different bead regions are coated with 10 different proteins, wherein 100 beads/bead region, that is to say a total of 1,000 beads, are used. 45 serums (each replicated twice) are then tested in three different settings: as 10 Plex with the proteins 1-10 and as 5 Plex, once with the proteins 1-5 and once with the proteins 6-10. When evaluating the data, the median MFI (median fluorescent intensity) or mean value MFI (mean fluorescent intensity) of all proteins is determined.

(19) In the singleplex assay 10 proteins are tested in the same bead region, wherein 1,000 are located in a bead region. 45 serums (each replicated twice) are tested in 3 settings: as singleplex with the proteins 1-10, as singleplex with the proteins 1-5 and as singleplex with the proteins 6-10. When evaluating the data, the median MFI or mean value MFI of the 1,000 beads (mix) is taken into consideration.

(20) The principle of the measurement and evaluation is based on Luminex technology. The beads are coloured using two fluorescent dyes (red and infrared), which emit in different ranges of the optical spectrum. The combination of these two dyes in ten different concentration stages in each case leads to one hundred shades of red and infrared that are spectrally distinguishable. Each of the resultant fluorescent intensities defines a population of bead regions. The fluorescent coding of the bead regions forms the basis for the identification by the analysis device and the precise assignment. The clear assignment of the individual bead regions is the basis for the multiplex analysis, wherein each bead region represents an individual test (fluorescent coding of the head regions). In the singleplex only one head region is used, such that an individual test is carried out and neither an assignment nor labelling with the red and infrared fluorescent dye is necessary, in contrast to the multiplex assay.

(21) The beads are coated with the proteins 1 to 10 to be tested, said proteins acting as antigen. The beads are then incubated with the serums to be tested. The better the extent to which the antibodies present in a serum bind to the antigen (s) immobilised on the bead (marker proteins), the higher is the measured fluorescent signal. The specific detection of the binding of the antibody to the heads is achieved via a detection molecule (conjugate). This conjugate has a high specific affinity to the bound antibody from the serum and is coupled to a fluorescent dye (for example phycoerythrin), which emits in the wavelength range of green light. This spectral range differs from those of the internal dyes, such that the classification of the beads and the quantification of the antibodies can be executed in parallel in the multiplex assay.

(22) Evaluation of the Results:

(23) The general standard evaluation for the measurement of Luminex beads is the median MFI (median fluorescent intensity). Here, individual high and/or low signals are ignored systematically.

(24) The median splits the total quantity (number) of measured intensities into two halves. The median is therefore the intensity value that lies in the middle of all measured intensities. It specifies the average intensity of the measured intensities, that is to say the typical intensity for a specific protein, here a specific antibody. With the specification of the median, only the average value is taken into consideration, whereas individual high and low intensity signals that lie to the right and left of the median remain unconsidered.

(25) The mean value M (arithmetic mean) is calculated as follows:
M=1/nXi=x1+x2+ . . . +xn/n

(26) For the multiplex measurement of a protein 1 on bead 1 (signal intensity 300), protein 2 on bead 2 (signal intensity 600) and protein 3 on bead 3 (signal intensity 30,000), a calculated median of 600 MFI (median fluorescent intensity) and a calculated mean value of 10,300 MFI (mean fluorescent intensity) are thus given.

(27) For a singleplex measurement of protein 1 (signal intensity 300), protein 2 (signal intensity 600) and protein 3 (signal intensity 30,000) on a head, the measured median is 600 MFI and a measured mean value is 10,300 MFI.

(28) The values of median MFI and mean value MFI in the multiplex and singleplex assay are consequently identical.

(29) In the singleplex measurement according to the invention, some high signals should not be ignored, in contrast to the conventional multiplex measurement. With the singleplex measurement according to the invention, the mean value is therefore preferably determined, and not the median.

(30) In FIG. 1, the MFI mean values of the individual proteins 1 to 10 of the multiplex assay are illustrated. The mean values of the multiplex assay for proteins 1-10, 1-5 and 6-10 have been calculated from these values. These calculated mean values of the multiplex assay were compared with those from the measured values of the singleplex assay for proteins 1-10, 1-5 and 6-10.

(31) The following serums were tested: RA-00037 (positive), RA-00029 (positive), RA-00046 (positive) and PRO-244 (negative). In the case of serum RA-00029 it was found that the calculated mean values of the multiplex assay coincide largely with the measured mean values of the singleplex assay (that is to say are comparable), more specifically for proteins 1-10 and also for proteins 1-5 and 6-10.

(32) Similar effects were found for serums RA-00037 (FIG. 2), RA-00046 (FIG. 3) and PRO-244 (FIG. 4)the mean values of the multiplex assay calculated from the measured intensities are comparable with the measured intensities of the singleplex assay (evaluation on the basis of mean values).

(33) These experiments have shown that the mean value MFI of the multiplex assay (proteins on different bead regions) as 5 plex and as 10 plex is comparable to the mean value of the respective singleplex assay (protein panel on one bead region).

Example 2: Comparison of Multiplex Assay and Singleplex Assay with Respect to the Median Intensities and the Mean Value Intensities

(34) Test set-up and execution as in example 1 with protein 1 on bead 1, protein 2 on bead 2 and protein 3 on bead 3 in the multiplex assay. The median MFI and mean value MFI were determined once for the multiplex assay from the measured intensities of proteins 1 to 3.

(35) These were compared with the measured intensities of the median MFI and the mean value MFI of the singleplex assay. In the singleplex assay, proteins 1, 2 and 3 were immobilised on a single bead (bead 4).

(36) FIG. 5 shows the comparison of the median of multiplex and singleplex assay of individual proteins 1 to 10. In set 1 the proteins 1-10, 1-5 and 6-10 are coupled to different bead regions, and in set 2 proteins 1 to 10, 1-5 and 6-10 are coupled separately to the same bead region. The values for the median MFI of the multiplex and singleplex assays are practically identical for proteins 1 to 10, 1-5 and 6-10. 10 g/ml (median of the counted beads) were used as a coupling control.

(37) FIG. 6 shows the comparison of the mean value of the multiplex and singleplex assay of individual proteins 1 to 10. In set 1 proteins 1-10, 1-5 and 6-10 are coupled to different bead regions, and in set 2 proteins 1 to 10, 1-5 and 6-10 are coupled separately to the same bead region. The values both for the mean value MFI of multiplex and singleplex assay are practically identical. 10 g/ml (mean value of the counted beads) were used as a coupling control.

Example 3: SUPA Microarray Evaluation

(38) The signal intensities of multiplex and singleplex assay were compared. The way in which the signal intensity changes with a reduction of marker sequences by the combination of a plurality of marker sequences to form a single marker panel was examined.

(39) The following consideration forms the starting point: the sum of signal intensities 15,000 and 5,000 gives a signal intensity of 20,000. The mean value of the signal intensity from 15,000 and 5,000 is a signal intensity of just 10,000.

(40) In microarray experiments the antigen/antibody interactions of 10 proteins and 20 serums were examined on protein microarrays (see FIG. 11). The examined antigens (proteins) are denoted by P1 to P0. The serums RA00026-RA00033, RA00035-RA00046 were examined for activity of autoantibodies contained in these serums.

(41) On the whole, only low signal intensities were found for protein/autoantibody interactions in the individual serums, wherein the individual serums differ considerably in part in terms of their activity with respect to individual proteins and also signal intensity. Two very high signal intensities of 15,643 (with protein 9) and 16,034 (with protein 0) were found with the serum RA00037. On the whole, serum RA000333 demonstrated average activitya signal intensity of 1,367 with protein 1, 1,573 with protein 3, 5,010 with protein 8 and 2, 631 with protein 9. Signal intensities with an MFI (median fluorescent intensity) of more than 10,000 are referred to as high signal intensities. Signal intensities with an MFI (median fluorescent intensity) from 1,000 to 10,000 are referred to as average signal intensity, and signal intensities with an MFI of less than 1,000 are referred to as low signal intensities.

(42) The serums RA00037 and RA00033 were characterised further with respect to their reactivity to different protein combinations.

(43) Table 2 gives an overview of the used protein combinations and the used naming system.

(44) TABLE-US-00001 TABLE 2 Tested Naming number system Explanation 1 Protein 10 P6 Protein 6 combination of 45 P67 Proteins 6 and 7 2 proteins combination of 20 P678 Proteins 6, 7 and 8 3 proteins combination of 8 P6789 Proteins 6, 7, 8 and 9 4 proteins combination of 2 P67890 Proteins 6, 7, 8, 9 and 0 5 proteins combination of 1 P1234567890 Proteins 10 proteins 1, 2, 3, 4, 5, 6, 7, 8, 9 and 0

(45) Serum RA00037 and Serum RA00033 were tested with the different proteins in the combinations specified in Table 2. 16 replicates per protein or protein mixture were tested. The threshold value at which the background differs from the signal was then determined statistically, and all signals below the threshold value (detection call) were set to 0. The detection call algorithm helps to decide whether or not a spot on a microarray is lit. To this end, the distribution of the local background over the microarray is calculated.

(46) In each case 5% at the lower end (lowest values) and 5% of the upper end (highest values) are removed from this distribution. The rest of the data is then logarithmised to the base of 10.

(47) The threshold value is given from she mean value of the logarithmised data plus twice she standard deviation of the logarithmised data.

(48) All spots of which the intensities lie below this value are set to zero.

(49) With the serum RA00037, the strongest signal was measured with a signal strength of 16034 with P0 (FIG. 7). The background, with a signal strength of 1134 and the dynamic range, thus lies at 1.15in other words between log 10(1134) and log 10(16034). On the whole, 35 different interactions were detected in the case of RA00037 and the used protein combination.

(50) With serum RA00033, the strongest signal was measured with a signal strength of 5010 with P8 (FIG. 3). The background, with a signal strength of 1522 and the dynamic range, thus lies at 0.52in other words between log 10(1522) and log 10(5010). On she whole, 38 different interactions were detected in she case of RA00033 and the used protein combination.

(51) A detailed analysis of the produced signal intensities of the proteins and protein combinations by interaction with autoantibodies in serums RA00037 and serum RA00033 is shown in Table 3.

(52) TABLE-US-00002 TABLE 3 Serum RA00037 Serum RA00033 signal intensity Naming signal intensity Naming system MFI system MFI P6 0 P6 0 P7 0 P7 0 P8 0 P8 5010 P9 15643 P9 2631 P0 16034 P0 0 P60 5236 P68 2218 P70 4059 P78 2121 P80 5373 P89 3196 P90 9856 P80 1653 P670 2210 P678 0 P680 2490 P689 1809 P690 4344 P680 900 P780 2014 P789 2331 P790 5292 P780 0 P890 4783 P890 926 P6790 3118 P6789 976 P6890 2455 P6890 0 P7890 2893 P7890 978 P67890 2240 P67890 0 P1234567890 0 P1234567890 0

(53) FIG. 9, for serum RA00037, shows the calculated and measured MFIs for the individual proteins and protein combinations in comparison. FIG. 10 shows the same for RA00033.

(54) It has been found that, with the combination of many proteins, the signal strength decreases and the measured signal intensities lie below the calculated signal intensities. For the serums RA00037 and RA00033 tested here, signals can no longer be detected with the combination of 10 proteins in one bead region. With a combination of 5 proteins, signals can be detected, however the detected signal intensity lies below the calculated signal intensities. With the used serums and marker panels, panels of 2 to 3 proteins are optimal. In particular, proteins that demonstrate strong interaction with autoantibodies are suitable for use in a marker panel.

Example 4: Singleplex Assay as SUPA-ELISA (ELISA Multimarker)

(55) The advantages of the ELISA are the low costs, simple application (ELISA applications are known and standardised) and the comparatively quick execution.

(56) ELISAs (enzyme-linked-immunoabsorbent assays) have long been known (for example Immunoassays of endogenous plasma insulin in man, Rosalyn S. Yalow and Solomon A. Berson, From the Radioisotope Service, Veterans Administration Hospital, New York, N.Y., Submitted for publication Mar. 7, 1960; accepted Mar. 22, 1960; J. Clin. Invest. 39: 1157-75, doi:10.1172JCI104130. PMC 441860. PMID 13346364; Enzyme-linked immunosorbent assay (ELISA) Quantitative assay of immunoglobulin G in Immunochemistry, Pergamon Press 1971, Vol. 8 pp. 871-874).

(57) Here, the indirect ELISA is used, described for example in (http:en.wikibooks.org/wiki/Structural_Biochemistry/Protein/Enzyme-Linked_Immuniabsorbent_Assay_%28ELISA%29). Here, an indentation in an ELISA plate (well) is coated with a protein, which acts as an antigen. A specific antibody can bind thereto. The binding is detected by a second (enzyme-linked) antibody, which converts a detectable substrate.

(58) In a singleplex ELISA according to the invention, 10 proteins as antigen are placed in a well of an ELISA plate and the signal of the panel of antigens is determined.

(59) The proteins used as antigens are immobilised overnight (3 times per sample). The proteins are then washed, blocked for two hours with Candor buffer, then washed and incubated for one hour with the serum to be tested. A washing step then follows as well as subsequent incubation for one hour with human AK HRP conjugate. The proteins are then washed again and incubated with substrate for 15 min. The reaction is stopped after 15 min and the result is read out at 450 nm. (Readout Tecan Safire).

(60) Three different serums RA00029, RA00037 and RA00045 are tested.

(61) The serums deliver the same results as in the Luminex assay (Examples 2 and 3). It is necessary to optimise the nest conditions for the ELISA for the marker proteins to be tested and for the serums to be analysed, for example in respect of the used buffer and the coating of the ELISA plates.

(62) The singleplex assay can be used universally, wherein the conditions have to be optimised for the respective assay type and the used detection system in order to obtain a significant single signal of the panel of marker sequences.