WO1994013829A1 - Isoelectric focusing differential separation assay - Google Patents

Abstract

A method for detecting and quantitating analytes among closely related substances by reacting the analyte in a test sample with excess labeled binding agent which specifically binds to the analyte to form a complex. The complex and excess labeled binding agent, which have different pIs, are separated by isoelectric focusing, with the location (x1) of the excess labeled binding agent defining an expected location (w1) at which the complex should be found. The focused substances are detected by optically scanning the separation medium (10) before or after release of the applied electric field. A second labeled marker with a known concentration provides a means for normalizing the peak areas which are compared with stored normalized calibration data to determine the amount of target analyte in the test sample.

Description

Description

Isoelectric Focusing Differential Separation Assay

Technical Field

This invention is in the field of detecting analytes using agents which specifically bind to the analyte to form complexes, such as antigen/antibody com- plexes.

Background Art

There is an extensive body of prior art related to analytical techniques based on the formation of com- plexes between specific binding substances, such as anti¬ gens and antibodies, hormones or cell modulators on the one hand and receptors, such as avidin-biotin and the like, on the other hand.

The Journal of Chromatoαraphy 539: 177-185 (1991) and references therein describe the separation of antigen-antibody complexes by capillary zone electropho- resis and isoelectric focusing. In that article, the capillary zone electrophoretic separation of hGH, the an¬ tibody to hGH, and the hGH-antibody complex are shown. Techniques in Protein Chemistry, Academic

Press, Inc., New York, New York (1984) , pp. 456-466, describes the purification of antibodies using high per¬ formance capillary electrophoresis.

U.S. Pat. No. 4,937,200, Kamazawa, describes the elution of antigens from an antibody packed affinity chromatography column wherein one member of the binding pair is bound to a solid support.

U.S. Pat. No. 5,006,473, Bouma, shows the mi¬ gration of an alkaline phosphatase labeled antibody in a liposome embedded electrophoresis media. After electro¬ phoresis, the liposome is lysed and a staining dye or reactant is released. U.S. Pat. Nos. 4,205,058 and 4,301,139 describe a chromato^'graphy column which separates antigen and anti¬ gen-antibody labeled complexes by selectively immobiliz¬ ing the complexes. Radio-labeled antigen competes with unlabeled antigen in a sample for binding with the antibody and the concentration of unlabeled antigen is determined by measuring the percentage of radio-labeled . antigen which migrates on the column. As an example, T4-I¹²⁵ and anti-T4 are reacted and separated on a cross- linked polyvinyl alcohol column. The antibody/T4 complex is retained at the bottom of the column while the T4-I¹²⁵ migrates up the column. The patents represent an example of the direct binding of a labeled hapten T4-I¹²⁵ and antibody and separation of these complexes. in U.S. Pat. No. 4,811,218, M. Hunkapiller et al. teach a DNA sequencing system using a multiple lane electrophoresis apparatus. Fluorescent dyes are attached to molecules moving through the lanes. A moving illumi¬ nation and detection system scans the multiple lanes. Four color data points are recorded for each of several lanes at a particular time at a fixed distance down the gel. Through a complex analytic procedure, the four colors are related to the concentrations of four dye- labeled DNA components. The object is to identify con- centrations of A, C, G or T which are DNA piece endings where A = adenosine, C = cytosine, G = guanine and T = thymine. Peak concentrations of a particular dye label are matched with particular bases in DNA sequences.

In U.S. Pat. No. 4,890,247, Sarrine et al. de- scribe an apparatus which robotically handles a plurality of liquid samples in test tubes, applies the samples to electrophoresis matrices and then carries out electropho- resis. The electrophoretically separated molecules are illuminated with fluorescent light. An analog signal is produced, representing the scanned field of view. A computer stores intensity levels of the analog signal and performs densitometric analysis to read the electro- phoretic data. Densitometry is a conventional prior art technique for reading such data.

In an article entitled "Affinity Electrophore¬ sis" by Vaclav Horejsi, reported in Enzyme Purification and Related Techniques, W. Jakoby ed. , Academic Press, 1984, p. 275, a novel type of electrophoresis is de¬ scribed. One lane of the gel medium is impregnated with immobilized ligands capable of reacting with a migrating macromolecule, while another lane, a control gel, is un- treated. Thus, a comparison can be made, using electro¬ phoresis, between a macromolecule sample retarded by the affinity gel lane and a similar sample in the control gel lane. In a variation of this technique, the gel may in¬ corporate an antibody which interacts with a migrating antigen. The two lanes may be calibrated so that differ¬ ent degrees of retardation, for different concentrations of the migrating macromolecule, are known. Moreover, microscopic beads treated with ligands can be entrapped in the gel and similarly serve as a retardant. Beads have the advantage of tight packing in the gel if they are of the appropriate size. Activation of the gel in¬ volves partial cross-linking so that the gels do not melt on heating. Alternative methods of gel preparation are described, all with the result that a macromolecular re- tardant is immobilized. Electrophoresis proceeds in the usual way.

Various types of pulsed electrophoresis are known for use in separating closely related substances (see Kreger EPA 457,748; Slater EPA 395,319; Agawa EPA 320,937; and Allington EPA 396,053) .

In U.S. Pat. No. 5,137,609, which is assigned to the assignee of the present invention and is incorpo¬ rated herein by reference, Manian et al. teach an elec- trophoresis-based differential separation assay for de- tecting analytes in a test sample. An excess of fluoreε- cently labeled binding agent specific for the analyte is reacted with the test sample to produce a reaction mix¬ ture containing labeled binding agent and labeled binding agent in complex with analyte. These two labeled sub¬ stances are placed on a separation medium and electro- phoretically separated. The differential rate of migra¬ tion of the labeled binding agent and complex are deter- mined by detecting the label as the substances migrate past a fixed detector. The migration time of the labeled binding agent is used to define a time window in which the complex should be found.

In spite of the increased specificity and sensitivity of this method there are still several problems. The labeled substances are separated on the basis of their charge to mass ratio making it difficult to distinguish between large molecules with small charge differences. In addition the labeled substances are mov- ing through the separation medium during detection which causes the peaks to spread, thereby decreasing resolution of the peaks. Further, use of the fixed detector neces¬ sitates waiting for the slowest migrating peak to move completely past the detector which may take 30 minutes or more.

An object of the invention is to devise a rapid, sensitive and high resolution method for sepa¬ rating large molecules of similar molecular weight with small charge differences. A further object is to in- crease the resolution and sensitivity of differential separation assays.

Summary of the Invention

The above objects have been achieved in a method for detecting and quantitating target analytes which specifically bind to a labeled binding reagent to form a complex. The binding agent is labeled with a de¬ tectable marker so that binding between the labeled binding agent and analyte provides a reaction mixture which contains labeled binding agent and a complex of analyte and labeled binding agent. The amount of labeled binding agent in the reaction mixture is in excess of what will react with the target analyte. The reaction mixture is placed on a suitable medium and the labeled substances are separated by isoelectric focusing. The term "separation medium" as used herein refers to the materials and material containers within which isoelec- trie focusing takes place. These include slab gels, gel filled or free solution capillaries, and any other mate¬ rials or containers suitable for isoelectric focusing. Besides binding with the target analyte, the excess amount of binding agent, which has a different focus point than the analyte/binding agent complex, serves as a reference point or marker for the complex. Differences in focus point arise because of differences in isoelec¬ tric point (pi) . In a preferred embodiment, a second labeled marker of known concentration is included which has a pi different from the labeled binding agent and complex.

More specifically, in one embodiment, the in¬ vention encompasses a method for detecting and quantitat¬ ing a target analyte in a test sample. One first pro- vides a labeled binding agent which specifically binds to a target analyte to form a complex having a characteris¬ tic isoelectric point different from the characteristic isoelectric point of unbound labeled binding agent. One also provides a separation medium for isoelectric focus- ing on which the labeled binding agent and complex are focused to different locations. Then one contacts the test sample with an amount of the labeled binding agent in excess of what will react with the target analyte to form a complex with analyte in the test sample. One then applies the reaction mixture of labeled binding agent and test sample to the separation medium.

This is followed by separating the complex and excess labeled binding agent by isoelectric focusing and scanning the separation medium to detect the presence of the label. One then records the relative spatial distri¬ bution and peak areas of the detected label and compares the recorded relative spatial distributions to the known relative spatial distributions of the complex and unbound labeled binding agents. Identification of a peak corre¬ sponding to the complex indicates the presence of the target analyte in the test sample. The area under the complex peak is then normalized and compared with the normalized peak areas of peaks containing known concen¬ trations of the target analyte to determine the concen¬ tration of the target analyte in the test sample.

For example, the analyte may be an antigen and labeled binding agent a fluorescently labeled antibody or the active fragment of an antibody such as an Fab. This method can be applied to the simultaneous measurement of multiple analytes or to the differential measurement of multiple isoforms of the same analyte.

In another embodiment, one again provides a specific binding agent which binds to the analyte to be detected. Next, one provides an excess of a labeled sub¬ stance which competes with the analyte for binding to said specific binding agent, the complex of the labeled substance and specific binding agent having a character- istic pi different from the characteristic pi of the un¬ bound labeled substance. One can now perform a competi¬ tion reaction of the labeled substance and the analyte in the test sample for the specific binding agent to form a labeled complex, the presence of analyte leading to a reduced amount of the complex. Then, one applies the reaction mixture to a separation medium suitable for iso¬ electric focusing of the complex and the unbound labeled substance. Next, the complex and excess labeled sub¬ stance are separated by isoelectric focusing. One then scans the separation medium to detect the presence of label and records the relative spatial distribution and peak areas of the detected label. The recorded relative spatial distributions are compared to the known relative spatial distributions of unbound labeled substance and complex to identify a complex peak the area under the complex peak is normalized so that one can compare the normalized peak area to the normalized complex peak area of a sample which does not contain the analyte. A finding of a reduced amount of complex indicates presence of the analyte in the test sample.

A preferred assay utilizes a second labeled marker of known concentration which has a pi different from the labeled binding agent and complex and therefore focuses at a different location. This second labeled marker serves as a quality control check on the system. For example, if the separation media or other reagent in the test kit are not operative, the failure of the second labeled marker to focus as expected can be readily detected and any results from that assay are discarded. Also, if the labeled binding agent or peak corresponding to complex failed to appear in the proper relationship to the second labeled marker, this test would be discarded as faulty. The second labeled marker, having a known concentration, can also be used to normalize the peak areas between runs.

Thus the present invention takes advantage of the specificity of reaction of specific binding ole- cules, the ability of capillary isoelectric focusing to separate small amounts of materials with high resolution, the sensitivity of detecting labels such as fluorescent labels, and a quality control and signal normalizing agent to provide rapid, sensitive and high resolution quantitative results. Another advantage of the invention is rapid determination of the results of isoelectric focusing without waiting for mobilization of the target substance or production of detectable materials by the separated biochemicals. The invention also discriminates between isoforms of the target analyte and is particular¬ ly useful in separating closely related molecules with small charge differences. The term "closely related" means not only heavy molecules of close molecular weight, but also molecules whose charge/mass ratio or other char- acteristic is such that both molecules exhibit similar migration rates, e.g., using electrophoresis, so that previous separation efforts have been difficult. Detailed Description of the Invention

The term "analyte" refers to a large variety of chemical substances for which there is a specific binding partner. It is contemplated that the present assay may be applied to the detection of any analyte for which there is a specific binding partner. The analyte usually is a peptide, protein, carbohydrate, glycoprotein, ster¬ oid, or other organic molecule for which a specific bind- ing partner exists in biological systems or can be syn¬ thesized. The analyte, in functional terms, is usually selected from the group consisting of antigens and anti¬ bodies thereto; haptens and antibodies thereto; and hor¬ mones, vitamins, metabolites and pharmacological agents, and their receptors and binding substances. Examples of analytes are immunologically-active polypeptides and pro¬ teins of molecular weights between 1,000 and 4,000,000, such as antibodies and antigenic polypeptides and pro¬ teins, as well as haptens of molecular weights between 100 and 1,500. Representative of such antigenic polypep¬ tides are angiotensin I and II, C-peptide, oxytocin, vasopressin, neurophysin, gastrin, secretin, and gluca- gon. Representative of antigenic proteins are insulin, chorionic gonadotropin (e.g. HCG) , carcinoembryonic anti- gen (CEA) , myoglobin, hemoglobin, follicle stimulating hormone, human growth hormone, thyroid stimulating hor¬ mone (TSH) , human placental lactogen, thyroxine binding globulin (TBG) , intrinsic factor, transcobalamin, enzymes such as alkaline phosphatase and lactic dehydrogenase, and hepatitis, HTLV-III, influenza, herpes, and other viral associated antigens. Representative of antibody ligands are those antibodies of the IgG, IgE, IgM and IgA classes specific for any of the antigens or haptens, or a class thereof, herein described. The class of hapten ligands is exemplified by thyroxine (T₄) , triiodothyro- nine (T₃) , the estrogens such as estriol, prostaglandins, vitamins such as biotin, vitamin B₁₂, folic acid, vitamin E, vitamin A, and ascorbic acid (vitamin C) , and drugs such as carbamazepine, quinidine, digoxin, digitoxin, theophylline, phenobarbital, primidone, diphenylhydan- toin, morphine, nicotine, and so forth.

DNA, RNA, and their complementary binding sequences and binding proteins can be determined by the method of this invention. Cytokines such as interleu- kins, interferons, G-CSF, GM-CSF, M-CSF, tumor necrosis factors (TNF) , erythropoietin and the like are represen¬ tative of cytokines that may be determined by methods of this invention.

This discussion of analytes is intended to il¬ lustrate the large number and variety of chemical sub¬ stances which have specific binding agents or for which specific binding agents can be made. The term "labeled binding agent" refers to sub¬ stances which specifically bind to the analyte and which have a detectable label. The label may be covalently linked or bound to the binding agent indirectly through another specific binding reaction, for example, a labeled goat antihuman antibody could be used to label a human antibody. Those skilled in this art will recognize a wide variety of techniques to label proteinaceous, as well as non-protein specific binding substances. For instance, fluorescent dye labeling of proteins in general and antibodies or antigens in particular is well known. For the determination of small analytes, such as those having a molecular weight of 100-1500, it is necessary to prepare a reagent which is a conjugate of the analyte to a labeled carrier. The free analyte in the test sample and analyte on the labeled carrier com¬ pete for the specific binding partner prior to separa¬ tion.

Current Protocols in Immunology, edited by John E. Coligan, Ada M. Kruisbeck, David H. Margolies, Ethan M. Shevach, and Warren Strober, John Wiley & Sons, N.Y. , 1991, extensively describes methods for obtaining poly- clonal and monoclonal antibodies and Fab fragments there¬ of, means for fluorescently labeling these antibodies and the reactions of these antibodies with antigens. Also described are electrophoresis and electrophoresis media as well as other separation techniques.

A large number of cytokines and monoclonal antibodies to these cytokines are known, for example, interleukin (l , 13, 2, 3, 4, 5, 6, 7, 8, 9, 10) ; inter- feron ( , β , γ) ; granulocyte/macrophage colony stimulat¬ ing factor (CN-CSF, G-CSF, M-CSF) ; tumor necrosis factor (TNF and β ) ; and transforming growth factor and their monoclonal antibodies are known. Antibodies to creatine kinase (skeletal, cardiac, and brain) are described in U.S. Pat. No. 4,353,982 (Gomez et al.) .

Lipopolysaccaride (LPS) , an endotoxin, is a major outer membrane component of cell walls of Gram- negative bacteria. Monoclonal antibodies to LPS are well known, see U.S. Pat. No. 5,092,235 (Williams et al. and references therein) .

A specific binding substance may be labeled with any of a variety of dyes, such as fluorescein dyes or rhoda ine dyes by conventional chemical techniques. Representative fluorescent dyes for making the labeling binding agent are fluorescein isothiocyanate (emission at 520 nm) , 4-chloro-7-nitrobenzo- -oxa-l-diazole (emission 550 nm) , tetramethylrhodamine isothiocyanate (emission 580 nm) , Texas Red (emission 610 nm) . These dyes are available from Molecular Probes, Inc. (Eugene, Oregon) , and can be synthesized with a variety of functional groups to accommodate the binding of these dyes to various chemical functional groups. For example, fluorescein 5 or 6 succinimidylcarboxylate, fluorescein 5 or 6 iodoacetamide, and fluorescein 5 or 6 maleimide are available. Similar functional groups are available for tetramethyl-rhodamine dyes.

It is well known to indirectly label a spe- cific binding molecule, such as an antibody by specifi¬ cally binding a second labeled antibody to the antibody reagent, such that the detected components are labeled antibody-antibody complex and labeled antibody-antibody/ antigen complex.

The term "separation media" refers to electro¬ phoresis, such as slab gel electrophoresis or capillary electrophoresis. Media, such as polyacrylamide, cellu¬ lose acetate, agar gel, and agarose gel, are suitable for slab gel electrophoresis. The medium in capillary elec¬ trophoresis is generally a free solution and separation medium refers to the capillary container for the free solution. In isoelectric focusing a pH gradient is established in the separation medium by inclusion of a mixture of carrier ampholytes: These separation tech¬ niques are well known to those skilled in this art.

In its simplest form, the invention involves the reaction of a labeled binding agent (LB*) and an ana¬ lyte [A] to form a complex A/LB* and separation of these species on a separation media, measuring the differential migration of A/LB* and LB* on the separation media by detecting the label and then using the difference in migration to identify the analyte by comparison with calibration data or other data which establishes expected migration data for the analyte.

In another embodiment of the invention where the A/LB* complex is too complicated, the reaction can be determined by measuring the decrease in the size of the LB* peak.

In another embodiment, multiple analytes can be detected by using different labeled binding agents that specifically bind to each analyte and detecting the LB* and A/LB* complex for each label.

In another aspect of this invention, two dif¬ ferent labeled binding agents can be bound to one analyte to form a sandwich complex and the migration of the sand¬ wich complex can be compared with either or both of the labeled binding agents.

Small molecules such as haptens can be detected by binding the hapten to a labeled carrier [C*] such as a polypeptide to form a conjugate in which the hapten or the carrier is labeled (HC*) . The hapten in a test sam¬ ple is allowed to compete with HC* for antihapten anti¬ body and the mixture is separated on the separation media. The species HC* and antibody/HC* complex are de¬ tected after they separate on the media. This reagent is referred to as a "hapten conjugated labeled carrier."

All of these embodiments are preferably prac¬ ticed by the inclusion of a second labeled marker which has a pi different from the labeled binding agent and complex and therefore focuses at a different location. The fluorescent dye can be bound to a protein or other substance which will affect its pi so that it will focus at a desired location in a particular medium. The labeled marker, having a known concentration, provides for normalization of channel to channel variability in antibody reaction and detection of signal amplitude. It provides an early warning that the particular assay is grossly incorrect. The labeled marker also provides a reference point for discrimination of both the labeled binding agent peak and complex peak, thus further as¬ suring the quality of the assay. The second labeled marker also provides a means for quantitation of the analyte by normalizing peak areas. This invention is most advantageously applied in the diagnosis of molecular variants in proteins. For example, creatine-kinase (CK) MB isoforms (MB2 and MB1) have been used in the early diagnosis of cardiac muscle injury following acute myocardial infarction. The determination of the MB2 and MB1 isoforms is also useful in determining the onset of acute cardiac allograft rejection as well as injury following coronary artery bypass grafting. The determination of CKMM isoform is important in monitoring atropic skeletal muscle changes. There is also known to be an increase in mitochondrial creatine kinase in patients with cerebrovascular damage and it is critical that there be a rapid assay to assess such damage so that drug therapy can begin.

The determination of alkaline phosphatase iso¬ forms is important in liver, bone, and kidney disease, as well as liver transplant rejection. Those skilled in the medical arts will recognize a large number of macromole- cules with charge-based isoforms or isovariants having clinical diagnostic significance which can be discrimi¬ nated by using the method of this invention.

Brief Description of the Drawings

Fig. 1 is a plan view of the apparatus of the present invention.

Fig. 2 is a top view of a thin film slab gel arrangement for electrophoresis.

Fig. 3 is a side view of a capillary arrangement for electrophoresis.

Fig. 4 is a plot of detector signals from complex and free fluorescent labeled binding agent. Fig. 5 is a plot of overlapping detector signals of different wavelength from complex and free fluorescent labeled binding agent.

Fig. 6 is a top view of a multiple lane arrangement for electrophoresis. Fig. 7 is a plot of detector signals from free fluorescent labeled binding agent and fluorescent labeled binding agent complexed with two different isoforms of a target analyte.

Fig. 8 shows the isoelectric focusing on a thin film slab gel of Cy5 labeled human serum albumin (HSA) and Cy5 labeled human serum albumin complexed with an antibody.

Fig. 9 shows the isoelectric focusing in a capillary of Cy5 labeled Fab and Cy5 labeled Fab complexed with CKMB2. Best Mode for Carrying Out the Invention

With reference to Fig. 1, a single lane iso¬ electric focusing electrophoresis apparatus 10 having a negative electrode 16 and a positive high voltage elec- trode 15 at an opposite end is shown. The electrophore¬ sis apparatus consists of a conventional single lane 18 having a substrate 17 and a separation medium 19. When the separation medium is a thin film gel the substrate is usually a self-supporting material which may be glass, mylar (trademark) or any well known gel support. The gel itself is usually polyacrylamide or argarose, although other gel materials such as synthetic acrylamide substi¬ tutes may also be used. Uniform polymerization and free¬ dom from bubbles and irregularities are desirable proper- ties. The gel may be hydrated or rehydratable. When the separation medium is a free solution capillary the sub¬ strate is a grooved insulator such as acrylic. The cap¬ illary itself is usually borosilicate glass, although other clear low reflection materials may also be used. Rectangular capillaries are preferred when the label is to be detected by optical scanning. The reaction mixture is applied to the separation medium by placing it on top of the gels or wicking it into the capillaries. High voltage is then applied to the separation medium at elec- trodes 15 and 16 and charged ions migrate to their iso¬ electric point.

The test sample is a fluid, frequently a frac¬ tionated blood sample. Blood may be preprocessed to re¬ move constituents which will interfere with the assay. Removal may be by filtering, absorption, centrifuging or precipitating either the desired or undesired components so that a desired target analyte may be obtained for electrophoresis. The desired target analyte must be one for which there is a specific binding agent. Fluorescent tags such as those commercially available are manufac¬ tured by Molecular Probes Inc. of Oregon which special¬ izes in dyes or dye beads that can be covalently attached to binding agents. Where target analytes are found in larger structures, such as pathogenic agents, then such a dye binding agent conjugate would be appropriate for tracking that pathogenic agent. Monoclonal antibodies can now be manufactured so that the behavior of this binding agent is uniform and predictable for many assays. Monoclonal antibodies are more expensive than polyclonal antibodies, but the antibodies have greater specificity, are directed toward single epitopes, are easy to produce in large quantities and are generally more useful in precise separation of bound and free material.

The labeled binding agent is applied in excess so that the reaction with the -analyte will be driven to completion, or nearly to completion in a reasonable or convenient amount of time. The amount of excess labeled binding agent should not be more than 20 times the maxi¬ mum expected amount of analyte, although the number may range between 2 and 50, approximately. The complex of labeled binding agent and analyte should have a different isoelectric point from that of the free labeled binding agent.

In isoelectric focusing a pH gradient is estab¬ lished in the separation medium by inclusion of carrier ampholytes. Carrier ampholyte mixtures which produce linear pH gradients in a variety of ranges are com er- cially available, such as PHARMALYTE (trademark) carrier ampholytes (copolymers of glycine, glycylglycine, amines and epichlorhydrin available from Pharmacia LKB Biotech¬ nology, Piscataway, New Jersey) or ISOGEL (trademark) ampholytes (available from FMC Bioproducts, Rockland, Maine) . In preferred embodiments the ampholytes are added to the reaction mixture before it is applied to the separation medium. Each solute migrates under the influ¬ ence of the electric field to a position in the separa¬ tion medium where the pH is equal to the solute's iso- electric point. Once focusing is complete, and while the electric field remains intact, the separation medium is optically scanned to detect the labeled substance. A strongly emitting light source, such as light emitting diode or laser 23 is used to generate a beam 25. The LED 23 has an output power of about 50 milliwatts and a wavelength band which will excite fluorescence in the fluorescent labeling material. Such excitation radiation is known as actinic radiation. The beam is intercepted by a focusing lens 27 which directs the beam through a slit aperture and barrier 29. Light emerging from the slit is divergent and is intercepted by the collimating lens 31. The beam is then directed onto a reflecting surface 33 which is part of a dichroic mirror 35. Di- chroic mirror 35 is chosen to selectively reflect light at the wavelengths emitted by light source 23 while transmitting light at the wavelengths emitted by the fluorescent labeled binding agents. The reflected beam is directed toward a focusing lens 37. Light passing through the focusing lens carries an image of the slit 29 which is directed onto separation medium 19. The image of slit 29 can be scanned along the longitudinal axis of separation medium 19 by moving separation medium 19 rela¬ tive to lens 37.

Fluorescent light emitted by labeled binding agent and complex, and some reflected light from the separation medium, travels in a retrobeam to focusing lens 37. Note that the focusing lens is used by light traveling in each direction. From there, the retrobeam is directed to reflecting surface 33 which is a part of dichroic mirror 35. Light reflected from the separation medium is reflected toward light source 23 while fluores- cent light is passed through. The fluorescent light is then directed by a mirror 41 through a filter 43 which rejects any light other than the desired wavelength from the fluorescent label. Light transmitted through the filter is directed toward a focusing lens 45. From there the beam is directed to a light detector, such as photo- multiplier tube 47 with a slit located at the image plane of the separation medium. The focused locations of the labeled substances are measured relative to one end of the separation medi¬ um. Each target substance and the corresponding labeled binding agent are subject to the same procedure in the calibration run. In calibration runs mean locations are determined. Then, the standard deviation is determined for the location of the free binding agent and the loca¬ tion of the complex. , In the present invention, it is necessary to know the expected locations for complex and free binding agent for specific target substances because those locations will be used to search for target analyte in a sample where the target analyte may or may not be present. The relative locations of the complex and free binding agent in calibration runs are used to establish a spatial window for the expected location of the complex. The location of the free binding agent is used to search for the complex in test samples. If the search reveals that a peak is present within a standard deviation or two of the expected location then that peak is identified as the complex.

The output of the photomultiplier tube is main¬ tained in a buffer memory 49 and a ratio may be formed between the signals representing labeled complex and free labeled binding agent. A data reader 50 is connected to the buffer memory 49 for receiving recorded signals which represent the fluorescent peaks. The data reader is a computer which correlates the various peaks. Each peak is recorded in order to search for complex and free labeled binding agent in the recorded data. Normally the location of the free labeled binding agent is established from prior calibration locations. Once the position of the free labeled binding agent peak is known, a search is conducted for the corresponding complex which should be located a certain distance away, within a spatial window defined by statistical limits. A peak within this window is identified as the complex, i.e. the target analyte. Next the peak areas of the identified peaks are examined and normalized in computer 50. The method whereby free labeled binding agent is correlated with complex is ex¬ plained further below. The computer also stores calibra¬ tions of known concentrations of target analyte so that the normalized peak areas may be compared in order to obtain an estimate of the unknown concentration.

In Fig. 2, the top view of a slab gel 11 shows that the image 29' of slit 29 scans between positive high voltage electrode 15 and negative voltage electrode 16. In a preferred embodiment electrodes 15 and 16 are dis- posable conductive polymer films in direct contact with thin film slab gel 11. Elimination of electrode reser¬ voirs is possible because optical scanning of the separa¬ tion medium obviates mobilization of the focused materi¬ als prior to detection. In one preferred embodiment the reaction mixture is placed on top of hydrated gel 11. Small volumes are placed approximately in the middle while larger volumes may uniformly cover the surface. In a second preferred embodiment of the invention the reac¬ tion mixture is used to rehydrate gel 11 giving a uniform initial distribution throughout the gel. In operation, the high voltage applied to electrode 15 causes migration of bound and free labeled binding agents, which are posi¬ tive or negative charged molecules to a point where they encounter a pH at which their net charges become 0, their isoelectric points, and migration halts. The free labeled binding agent will reach a position corresponding to its pi which is different from the pi of the bound labeled binding agent.

With reference to Fig. 3, a side view of a pre- ferred embodiment using a capillary 12 is shown. Sub¬ strate 17 is an insulator platform with elevated ends. Metal electrodes 15 and 16 are located in longitudinal grooves in the top surfaces of opposite ends of platform 17. Capillary 12 is placed horizontally on top of plat- form 17 with the ends of capillary 12 located in the grooves containing electrodes 15 and 16. A gap 13 separates capillary 12 from electrode 15 and a gap 14 separates capillary 12 from electrode 16. In one^' preferred embodiment capillary 12 is loaded with a reaction mixture containing ampholytes before being placed on platform 17. The end of capillary 12 is dipped in the reaction mixture and capillary action wicks the solution into the capillary. Loaded capillary 12 is then placed on platform 17 and a drop of basic catholyte is added to gap 14 bridging capillary 12 with electrode 16 and a drop of acidic anolyte is added to gap 13 bridging capillary 12 with electrode 15. In a second preferred embodiment capillary 12 is loaded with reaction mixture after being placed on platform 17. A drop of reaction mixture is placed in gap 13 or 14 and capillary action wicks the solution into the capillary and the other gap making contact with both electrodes. With reference to Fig. 4, a plot of the detec¬ tor signal is shown where the horizontal axis is distance and the vertical axis is amplitude of the detected sig¬ nal. After focusing, the separation medium is scanned to detect the presence of label. At position XI a relative- ly large peak 51 is observed, representing free fluores¬ cent labeled binding agent of a first color. Another signal 54 discussed below is detected at a second loca¬ tion. At a third location, X3 , a weaker signal 53 of the same color is observed. Peak 53 exists in the mid region of a window, Wl, between X2 and X . The existence of window Wl is established by the strong free fluorescent labeled binding agent signal 51. Peak 53 is within win¬ dow Wl and is recognized as a fluorescent labeled complex signal. Peak 54 is not within window Wl and is treated as a false positive or artifact, after being checked to determine whether the signal is not mistaken for the free fluorescent labeled binding agent signal 51. A search of all signals is made to determine the most logical posi¬ tions for free fluorescent labeled binding agent and com- plex. If no signal is found in spatial window Wl, the absence of target analyte is inferred. Each window W acts as a spatial filter, allowing discrimination of spu¬ rious fluorescent signals and noise. Note that all signals are recorded and signal discrimination occurs after recording by analyzing recorded data. Even though separation medium characteristics may vary from lane to lane or run to run, the present invention has immunity to most variations because the complex and free fluorescent labeled binding agent are subject to the same conditions. The ratio of the two signals represented by the areas under peaks 51 and 53 represents an estimate of the ratio of complex to free fluorescent labeled binding agent after normalizing data relative to calibrations, assuming good binding efficiency. At a further location, another peak 55 is observed. This represents another free fluorescent labeled binding agent. This defines another spatial window W2 at a subsequent location and a lesser peak 57 is measured in the window. This is taken to represent a fluorescent labeled complex. Again, the ratio of bound to free label is computed and once again the target analyte associated with the second label may be estimated in concentration. It is possible for the peaks to overlap each other as shown in Fig. 5. Here, the first free fluores¬ cent labeled binding agent peak 61, having a relatively large amplitude, overlaps a second peak 65 of similar amplitude in a test where two different fluorescent la- beled binding agents were used. Second peak 65 is the second free fluorescent labeled binding agent signal. However, because different colors are used, as separated by filter 43 in Fig. 1, the two peaks may be separately observed. Peak 61 establishes the spatial window W3 where a peak 63, representing a fluorescently labeled complex of a color which is the same as that associated with free labeled binding agent peak 61, occurs totally within second peak 65. Nevertheless, because of filter 43, peak 63 may be spatially and optically differentiated from peak 65. The ratio of bound to unbound signal amplitudes appears to be about 2:1. The corresponding molecular amounts of complex and free labeled binding agent are estimated to be in the same ratio. For peak 65, a spatial window W4 is established, but no fluores¬ cent signal is found within the window so the absence of the corresponding target analyte is inferred.

With reference to Fig. 6, a multiple lane electrophoresis cartridge is shown. The cartridge 71 is provided with two lanes 73 and 75. Each of the lanes has a respective slit image 87 and 89 which scan the lanes as indicated. The two lanes are constructed similarly with the separation mediums 74 and 76 being thin film slab gels or free solution capillaries. Lane 73 is used to run a calibrated amount of target analyte and a known amount of free fluorescently labeled binding agent. In lane 75 an unknown amount of target analyte is run with free fluorescently tagged binding agent. After normaliz- ing the peak areas per unit of labeled binding agent, the two lanes may then be compared to determine the amount of unknown analyte in lane 75. In the preferred embodiment normalization is carried out by running a constant known amount of a second labeled marker in each lane during every run. For greater accuracy, multiple runs may be made in lane 73 of various amounts of target analytes so that many normalized peak areas may be stored in the memory. The normalized peak area from a run containing an unknown amount of target analyte may then be looked up and compared with the normalized peak areas of known con¬ centrations, with the best match indicating the amount of target analyte.

A single labeled binding agent may be used to determine the ratio of different isoforms of a target analyte as shown in Fig. 7. Here, the free fluorescent labeled binding agent peak 91 is used to establish two windows, W5 and W6, for two isoforms of the target ana¬ lyte. A peak 93 is located within window W5 and is rec¬ ognized as labeled binding agent bound to one isoform of the target analyte. A second peak 95 is located within window W6 and is recognized as labeled binding agent bound to a second isoform of the target analyte. The ratio of the peak area for peaks 93 and 95 may be used to estimate the relative amounts of the two isoforms of the target analyte in the sample. Both isoforms of the target analyte must contain the epitope recognized by the specific binding agent, but differ in isoelectric point. One of the advantages of the present invention is that analysis of the peaks representing complex and free labeled binding agent can be computed without re¬ leasing the applied voltage. Another advantage is that the present system uses only a single lane of an electro- phoresis apparatus so that run to run non-uniformities are nulled. It is possible to use a second lane in an electrophoresis device as a reference or calibration, but such calibrations may be done beforehand and the results stored in a memory. It is also possible to use a second or third or fourth lane for additional analytes of inter¬ est creating panels of relevant related analytes. In the prior art, analysis of target analytes usually requires post-focusing mobilization or in situ analysis by a plu¬ rality of stains, colored or fluorescent substrates, etc. Using the present invention the analysis may be run in real time as soon as focusing is complete and before the applied voltage is released. This leads to an increase in both resolution and sensitivity. The applied voltage may be maintained during analysis so that the peaks will remain focused. Focusing increases the concentration of the complex peak leading to greater sensitivity. Spatial separation of the peaks may be adjusted without increas¬ ing peak width by varying the pH gradient. Amplitude thresholds may be used as further discrimination against noise and artificial signals.

To discriminate between two or more fluores- cently labeled binding agents in the same gel lane, dif¬ ferent fluorescent wavelengths may be used, so long as filter 43 in Fig. 1 can adequately resolve the different wavelengths. Multiple tests can be run simultaneously, each test associated with a particular wavelength. Exa ple 1 Detection of Antibody to Human Serum Albumen (HSA)

In the following example an antibody against HSA is the target substance which is detected by tagging with fluorescent HSA. HSA, fraction V, was obtained from Sigma Chemical Company (St. Louis, Missouri). Monoclonal anti-HSA was obtained from Biospacific Inc. (Emeryville, California) . Cy5-labeled HSA was synthesized by the cou- pling of Cy5 fluorescent dye (Biological Detection Sys¬ tems, Pittsburgh, Pennsylvania) to HSA and removal of free dye using gel filtration ^• (Molecular Probes, Eugene, Oregon) . This fluorescent substance is the labeled bind¬ ing agent. Differential separation assay (DSA) isoelectric focusing was done as follows: Cy5-labeled HSA (binding agent) was incubated with monoclonal anti-HSA (target) at a final concentration of 500 ng/ml Cy5-HSA and 250 μg/ml anti-HSA in 1 mM phosphate, 15 mM NaCl, pH 7.2. A control sample consisted of Cy5-HSA alone at 500 ng/ml without added antibody. Reactions were performed in 1.5 ml Eppendorf tubes in a total reaction volume of 20 μl. After incubating the samples at room temperature (20°C) for 30 minutes, 1 μl aliquots were loaded onto pHAST IEF gels (Pharmacia, Piscataway, New Jersey) con¬ taining a pH gradient of 4 to 6.5 and electrophoresis was performed.

The positions of the fluorescent proteins were determined using a post-electrophoretic scan of the gel with a He-Ne laser optic system by moving the gel with a stepper motor. The reflected fluorescence was collected using a R928 PMT and the data was collected using data acquisition software on an IBM (trademark) personal computer. Separation on this type of gel is based on the isoelectric points of the proteins. With reference to Fig. 8, the result is shown as a plot of fluorescence versus distance on the gel. The pH gradient on the gel is also plotted. The Cy5-HSA control peak 97 is focused at 34 mm (pH 4.7) . The immune complex consisting of Cy5- HSA/anti-HSA, on the other hand, has a peak 98 which is focused at 23 mm (pH 5.3) . As shown on the figure, excellent separation (0.6 pH units) is obtained between the immune complex and the excess binding agent (Cy5- HSA) . The peak 99 is residual uncomplexed labeled Cy5-HSA. This example demonstrates that the relevant spatial window for this pair of labeled binding agent (Cy5-HSA) and analyte (anti-HSA) is 11 mm. Peak 97 at 34 mm defines the reference position for the data acquisi¬ tion window in which the immune complex Cy5-HSA/anti-HSA peak should be found.

Example 2

Detection of Proteins Present in Human Blood

Creatine kinase is an enzyme present in various mammalian tissue. It occurs in three different forms known as isoenzymes: CK-MM (skeletal) , CK-MB (cardiac) and CK-BB (brain) . After release from tissue and on cir¬ culation in blood the MM and MB forms themselves break¬ down to smaller fragments known as isoforms or subforms. In the event of myocardial infarction, the MB isoenzyme, present in cardiac muscle, is released in the plasma. Hence, it serves as a specific diagnostic molecular marker for cardiac ischemia or necrosis. The early and rapid detection of this isoenzyme and its isoforms are very crucial for the diagnosis of myocardial infarction for initiating thrombotic therapy.

Separation of Cy5 labeled CK-MB antibody from its immune complex was performed using a capillary isoelectric focusing system. CK-MB2 (human heart) and monoclonal anti CK-MB were obtained from Biospacific (Emeryville, California) . Fab fragments were prepared by digesting the monoclonal anti CK-MB with the enzyme papain. Cy5 labeled Fab was synthesized by the coupling of Cy5 fluorescent dye (Biological Detection Systems, Pittsburgh, Pennsylvania) to Fab and purified by conventional gel permeation and ion exchange methods. This fluorescent substance is the labeled binding agent. Differential separation assay (DSA) was done as follows: Cy5 labeled Fab (binding agent) was incubated with CK-MB2 (target) at a final concentration of 50 μg/ l Cy5-Fab and 1 mg/ml CK-MB2 in 1 mM phosphate 15 mM NaCl pH 7.2. A control sample consisted of Cy5-Fab alone at 50 μg/ml without added CK-MB2. Reactions were performed in 1.5 ml Eppendorf tubes in a total reaction volume of 10 μl. After incubating the samples at room temperature (20°C) for 30 minutes, HSA was added as a carrier at a final concentration of 2 mg/ml. The reaction mixture was then diluted 30-fold with a 2% solution of ampholytes with a 3 to 10 pH range (Biorad Inc., Hercules, Califor¬ nia) in deionized water. Capillary action was used to fill a 50 x 0.3 x 0.03 mm borosilicate glass rectangular capillary (R&S Medical, Mountain Lakes, New Jersey) , coated to suppress electroendosmosis (Capillary Electro- phoresis. Academic Press, Inc., San Diego, California

(1992), pp. 191-214), by dipping its end in the diluted reaction mixture. The capillary was then placed horizon¬ tally on an acrylic platform and platinum electrodes were bonded to the acrylic at the ends of the capillary. A drop of 0.02 M sodium hydroxide was applied to one end of the capillary to bridge it with the cathodic electrode and a drop of 0.02 M phosphoric acid was applied to the other end of the capillary to bridge it with the anode. Electrophoresis was performed at 2 kV constant voltage for 10 minutes with a CZE 1000R high voltage supply

(Spellman, Plainville, New Jersey) . The current was al¬ lowed to drop from 30 to 2 μamps as the focusing took place.

The positions of the fluorescent proteins were determined at this point by scanning the capillary with a He-Ne laser optic system by moving the capillary with a stepper motor with the field on. The reflected fluores¬ cence was collected using a R928 PMT and the data was collected using data acquisition software on an IBM (trademark) personal computer.

Separation on this capillary system is based on the isoelectric points of the proteins. The result is shown as a plot of fluorescence versus distance on the capillary in Fig. 9. The Cy5-Fab control peak 101 is focused at 18 millimeters. The immune complex consisting of Cy5-Fab/CK-MB2, on the other hand, has a peak 103 which is focused at 23 millimeters. The peak 105 is residual uncomplexed labeled Cy5-Fab.

Claims

Claims

1. A method for detecting and quantitating one or more analytes in a test sample comprising: a. providing distinguishable labeled binding agents which specifically bind to each analyte to form complexes having characteristic isoelectric points different from the characteristic isoelectric points of the corresponding unbound labeled binding agents; b. providing a separation medium for isoelec¬ tric focusing on which the labeled binding agents and corresponding complexes are focused to different loca¬ tions; c. contacting the test sample with an amount of the labeled binding agents in excess of what will react with each analyte to form complexes with analytes in the test sample; d. applying the reaction mixture of test sample and labeled binding agents to the separation medium; e. separating the complexes and excess labeled binding agents by isoelectric focusing; f. scanning the separation medium to detect the presence of the labels; g. recording the relative spatial distribu¬ tions and peak areas of each of the detected labels; h. comparing said recorded relative spatial distributions to the known relative distribution of com¬ plex peaks and unbound labeled binding agents wherein identification of said complex peaks indicates the presence of each of said analytes; i. normalizing the areas under each of said complex peaks per unit of labeled binding agent; and j . comparing said normalized peak areas to the normalized peak areas of samples containing known amounts of each of said analytes to determine the concen¬ tration of each of said analytes in the test sample.

2. The method according to claim 1 wherein the distinguishable labels are optically detectable.

3. The method according to claim 2 wherein the distin- guishable labels are different fluorescent dyes.

4. The method according to claim 1 wherein the means for normalizing the peak areas is a second labeled marker of constant concentration having a characteristic iso- electric point different form the characteristic isoelec¬ tric point of the labeled binding agents and labeled binding agents in complexes.

5. The method according to claim 4 wherein the second labeled marker is a fluorescently labeled molecule.

6. A method for detecting an analyte in a test sample comprising: a. providing a specific binding agent which binds to the analyte to be detected; b. providing an excess of a labeled substance which completes with said analyte for binding to said specific binding agent, the complex of said labeled sub¬ stance and specific binding agent having a characteristic isoelectric point different from the characteristic iso¬ electric point of the unbound labeled substance; c. performing a competition reaction of said labeled substance and said analyte in the test sample for said specific binding agent to form a labeled complex, the presence of analyte leading to a reduced amount of said complex; d. applying the reaction mixture to a separa¬ tion medium suitable for isoelectric focusing of the complex and the unbound labeled substance; e. separating the complex and excess labeled substance by isoelectric focusing; f. scanning the separation medium to detect the presence of label; g. recording the relative spatial distribu¬ tion and peak areas of the detected label; h. comparing the recorded relative spatial distributions to the known relative spatial distributions of unbound labeled substance and complex to identify the complex peak; i. normalizing the area under the complex peak per unit of labeled binding agent; and j . comparing said normalized peak area to the normalized complex peak area of a sample which does not contain the analyte wherein finding of a reduced amount of complex indicates presence of the analyte in the test sample.

7. The method according to claim 6 wherein the labeled substance is a hapten conjugated labeled carrier, the analyte is a hapten, the specific binding agent is an antibody, or active antibody fragment, specific for the hapten, and the label is optically detachable.

8. The method according to claim 7 wherein the antibody is a monoclonal antibody, or active antibody fragment, and the label is a fluorescent dye.

9. The method according to claim 7 wherein the analyte is in a mammalian body fluid.

10. A method for determining the relative concentrations of multiple isoforms of a target analyte in a test sample comprising: a. providing a labeled binding agent which specifically binds to each isoform of the target analyte to form complexes; b. providing a separation medium for isoelec- trie focusing on which the labeled binding agent and com¬ plex of each isoform are focused to different locations; c. contacting the test sample with an amount of labeled binding agent in excess of what will react with each isoform to form complexes with each isoform in the test sample; d. applying the reaction mixture of labeled binding agent and test sample to the separation medium; e. separating the complexes and excess labeled binding agent by isoelectric focusing; f. scanning the separation medium to detect the presence of the label; g. recording the relative spatial distribu¬ tion and peak areas of the detected label; h. comparing said recorded relative spatial distribution to the known relative spatial distribution of complexes and unbound labeled binding agent wherein identification of a complex peak indicates the presence of an isoform of the target analyte in the test sample; and i. comparing the peak areas of each isoform of the target analyte to determine the relative concentra¬ tion of the isoforms in the test sample.

11. The method according to claim 10 wherein the labeled binding agent is a labeled antibody, or labeled active antibody fragment, and the label is optically detectable.

12. The method according to claim 11 wherein the anti¬ body is a monoclonal antibody, or active fragment of a monoclonal antibody, and the label is a fluorescent dye.

13. The method according to claim 10 further including a second labeled marker of constant concentration having a characteristic isoelectric point different from the char¬ acteristic isoelectric point of the labeled binding agents and labeled binding agents in complexes.

14. The method according to claim 13 wherein the second labeled marker is a fluorescently labeled molecule.