WO1999050673A1

WO1999050673A1 - Solid-phase method for antigen and antibody determinations in bloodgroup serology, and test kit

Info

Publication number: WO1999050673A1
Application number: PCT/NL1999/000181
Authority: WO
Inventors: Eric Marinus Maria Van Der Donk; Pieter Johannes Den Boer; Ronald Victor Wilhelmus Van Eijk
Original assignee: Stichting Sanquin Bloedvoorziening
Priority date: 1998-03-27
Filing date: 1999-03-29
Publication date: 1999-10-07
Also published as: EP1064556B1; ES2255253T3; EP1064556A1; DE69928507T2; AU3059499A; ATE310956T1; NL1008738C2; DE69928507D1

Abstract

A method and test kit for determining in a sample an analyte, which is a bloodgroup antigen present on erythrocytes or an antibody binding to a bloodgroup antigen. The sample is treated with a reagent containing an analyte-binding partner, so that if the sample contains analyte, a complex of bloodgroup antigen present on erythrocytes and antibody bound thereto is formed.

Description

Title: Solid-phase method for antigen and antibody determinations m bloodgroup serology, and test kit

Field of the invention

The invention lies m the field of bloodgroup serology and relates more particularly to a method and test kit for determining bloodgroup antigens, or antibodies directed thereto, in a sample.

Background of the invention

The purpose of bloodgroup assays and antibody examination in clinical work is often to obtain compatible erythrocyte preparations for transfusion. The invention concerns the detection and identification of antibodies and antigens and can be used for bloodgroup assay, antibody screening and identification, and performing cross tests. The test is not based on hemagglutmation but on the solid-phase principle, which does not involve washing steps .

Blood transfusion reactions which are induced by allo-antibodies to erythrocytes are called hemolytic transfusion reactions because they are mostly accompanied by an often very strongly accelerated breakdown of erythrocytes. The point is therefore to prevent hemolytic transfusion reactions by careful bloodgroup examination.

Bloodgroup antibodies are nearly always immunoglobulins of the IgG or IgM type. The antigen- antibody interaction is dependent inter alia on ionic bonding, hydrogen bridges and hydrophobic effects (displacement of water) . The strength of a binding between a binding site of an antibody and an epitope is designated as 'affinity¹. Antibodies which are able to agglutinate erythrocytes under all conditions are called agglutinins or complete antibodies (mostly IgM) . Antibodies that bind to erythrocytes but do not give agglutination (sensitization) are called incomplete antibodies (mostly IgG) .

There are a large variety of serological tests. The most important techniques at present are the tube method, the column test and tests in microtiter plates. A distinction can be made between techniques that are based on hemagglutination and techniques that are based on the solid-phase principle. a. Tests based on agglutination: The tube test is a widely used test, where also prolonged incubations with antibodies are possible. The erythrocytes, after the reaction with antibodies, can be sedimented or be centπfuged to accelerate the agglutination reaction. An important and widely applied te'st is the antiglobulm or Coombs test. The antiglobulm test is based on the principle that erythrocytes loaded with, for instance, antibodies of the IgG type can be agglutinated by antiglobulm serum. In the test, three phases can be distinguished. The first phase is the sensitization phase. During this phase, antibodies bind to the corresponding antigen structures on the erythrocytes (sensitization of erythrocytes) . When the binding is optimal, the second phase, viz. the washing phase, takes place. In this phase, all non-bound antibodies are removed from the incubation mixture. Insufficient removal of non-bound antibodies can lead to inactivation of the antiglobulm serum in that these antibodies bind to the antiglobulm. The third phase is the antiglobulm phase, in which antiglobulm is added to the washed sensitized cells, so that the sensitized cells are coupled to each other (agglutination of the erythrocytes) .

A further simplification of serological tests has been carried through by centrifug g the erythrocytes through a column of (Sephadex) gel or glass beads. The use of transparent, inert, solid particles for distinguishing agglutinates and non-agglutmates was already described by Dalton et al . in 1970 (Becton, Dickinson & Co . US Patent 3,492,396). The system whereby gel material is used for the detection of erythrocyte-antibody reactions has been described by LaPierre et al . (Transfusion 1990; 30: 109-113, European Patents 0 194 212 and 0 305 337). In- this test system, use is made of small columns filled with Sephadex gel . Use can be made of columns not containing any antibodies (e.g. for reverse ABO typing) . As a second possibility, a gel column can also contain antibodies which are directed to certain erythrocyte antigens (e.g. for typing) . For the antiglobulm test, use is made of gel columns which contain antiglobulm serum. After an incubation, the gel columns are centrifuged. In the case of^* a negative reaction, all erythrocytes will end up at the bottom of the tube; if the test is positive, the erythrocyte agglutinates will be captured on top of or in the gel. In the case of weak reactions, erythrocytes will sediment partly.

A major advantage of this test is that in case of the antiglobulm test no washing steps are needed anymore, since during the centrifugation separation of erythrocytes and serum or plasma components takes place. A second advantage is the fact that the results of the test can be properly fixed. The test described is used by DiaMed AG (DiaMed-ID Microtyping System) .

A disadvantage of this test is that often problems occur in demonstrating weak agglutinates, for instance in the case of weak ABO-incompatibilities (Cummins & Downham, Lancet 1994; 343:1649 and Philips et al . , Transfusion Medicine 1997; 7:47-53), possibly in that so-called "shear forces" during the centrifugation cause larger agglutinates to disintegrate into several small agglutinates or, in an extreme case, into individual (sensitized) erythrocytes. In spite of the sieve action of the gel, only weak to negative reactions are observed then, resulting in an increase of tne number of false- negative reactions. b. Tests based on the solid-phase principle: Another approach is the use of the solid-phase principle as an alternative to direct and indirect agglutination reactions for bloodgroup assay, antibody screening, antibody identification and cross tests. Application and advantages of the use of solid-phase techniques in the areas mentioned have been described by Rosenfield (Abstracts, 15th Cong. Int. Soc. Blood Trans., Paris pp. 27-33, 1976; US Patent 4,275,053, 1981) . Here, inter alia erythrocytes were used which had been coupled to the surface of plastic tubes. More recently, systems have been described by Plapp et al . (Am. J. Cl . Path. 82: 719-721, 1984), Bayer et al . (US Patent 4,608,246, 1986), Rachel et al . (Transfusion 25: 24-26, 1985), Plapp et al . (The Lancet 1465-1466, 1986) and Uthemann et al . (US Patent 4,925,786, EP 363 510 and Transfusion 30: 114-116, 1990). Microtiter plates in combination with the solid-phase principle are used by, among others, Biotest AG (Solidscreen II for antibody diagnostics), Immucor Inc. (Immunocapture Capture-R systems for antibody screening and identification) and CLB (Microtype for typing erythrocyte antigens) and have further been described by Llopis et al . (Vox Sangumis

1996; 70:152-156 and Vox Sangumis 1997; 72:26-30), Ohgama et al . (Transfusion Medicine 1996; 6:351-359), Chateau et al. (La Gazette de la Transfusion 1997; 131:38-41, FR 94 05123) and Den Boer et al . (WO 98/02752). The solid-phase tests are not limited to microtiter plates. Pernell (Sanofi Pasteur, EP 0 594 506) describes an affinity gel test m which an lmmunoglobulm-bmdmg substance (such as protein A, protein G or anti -human IgG) is immobilized on the gel. Incubation of erythrocytes and antibodies occurs above the gel. After th s incubation phase, centrifugation of the gel column takes place. If antibodies are bound to the erythrocytes, these sensitized erythrocytes will bind to the lmmunoglobulm-bmdmg gel material and will hence bind to the upper part of the gel column. Cells to which no antibodies are bound will not be bound and end up at the bottom of the column. This principle of binding of sensitized cells to a solid phase has previously been described by Pharmacia (Cell Affinity Chromatography; Uppsala, Sweden; 1984) . Pharmacia describes therein the purification of erythrocytes on the basis of the presence or absence of certain bloodgroup antigens using protein A-sepharose 6MB. Erythrocytes with the A antigen on the surface (bloodgroup A erythrocytes) to which anti-A antibodies are bound, can be separated by binding to said gel material (via protein A) from erythrocytes that do not possess the A antigen. The test described by Sanofi Pasteur is therefore a combination of this principle and the centrifugation step for separating erythrocytes, whether or not sensitized, and non-bound antibodies as used, for instance, in the column system of DiaMed.

A disadvantage of the above-described affinity gel test, where the ligands described are protein A, protein G or an antiglobulm, is that no complement factors are bound. Some antibodies are complement-binding (e.g. anti-Jk antibodies) : the binding of a single antibody molecule to a particular erythrocyte antigen leads to the binding of many complement molecules on the erythrocyte membrane. Consequently, many allo-antibodies, for instance, can be demonstrated with much higher sensitivity with anti -complement than with anti-immunoglobulm antibodies. In fact, many examples have been described of antibodies that could be demonstrated exclusively with anti-complement (Mollison et al . , Blood Transfusion in Clinical Medicine, Blackwell Scientific Publications 1992) . Accordingly, tests that do not display anti -complement activity involve a chance of false-negative reactions.

The disadvantage of the use of protein A as lmmunoglobulin-bmdmg material is that antibodies of the type IgG3 cannot be demonstrated (the fact is that protein A only binds the IgG subclasses IgGl, IgG2 and IgG4) , although these may be clinically relevant. IgM antibodies, which are also clinically relevant, are alleged not to be bound by protein G and only partly by protein A.

Brief summary of the invention

The invention now provides a method for determining in a sample an analyte selected from a bloodgroup antigen present on erythrocytes or an antibody binding to such a bloodgroup antigen, comprising treatment of the sample in an incubation zone of a reaction vessel with a reagent containing an analyte-binding partner, which analyte-bmdmg partner, m the case where the analyte is a bloodgroup antigen present on erythrocytes, is an antibody capable of binding to the bloodgroup antigen, and in the case where the analyte is an antibody binding to a bloodgroup antigen, is the bloodgroup antigen present on erythrocytes, wherein in the incubation zone, if the sample contains the analyte, either a complex is formed of bloodgroup antigen present on erythrocytes and antibody bound thereto, or a complex is formed of bloodgroup antigen present on erythrocytes and antibody bound thereto, as well as a complex of complement factors bound to these erythrocytes if the antibody is complement-binding, further comprising separation of erythrocytes, complexed or not, from non-bound antibodies using a separation medium, located in the reaction vessel under the incubation zone, of a density higher than that of the fluid containing the antibodies but lower than the density of erythrocytes, whereby erythrocytes pass through the separation medium and non-bound antibodies remain m the incubation zone, separation of complexed erythrocytes from non-complexed erythrocytes by binding complexed erythrocytes in an immobilization zone of the reaction vessel to binding substance immobilized this zone, which binding substance comprises IgM-bmdmg substance and/or complement-binding - substance, and optionally also IgG-bmding substance, and discharging non-complexed erythrocytes to a collection-, zone of the reaction vessel, and detection of erythrocytes in the immobilization zone and/or collection zone.

The present invention also provides a test kit suitable for use in a method for determining an analyte m a sample, the analyte being a bloodgroup antigen present on erythrocytes or an antibody binding to such a bloodgroup antigen, comprising:

(I) a reagent containing an analyte-bmdmg partner which, in the case where the analyte is a bloodgroup antigen present on erythrocytes, is an antibody capable of binding to the bloodgroup antigen, and in the case where the analyte is an antibody binding to a bloodgroup antigen, is the bloodgroup antigen present on erythrocytes,

(n) a reaction vessel comprising an incubation zone, an immobilization zone and a collection zone, there being immobilized in the immobilization zone a binding substance which comprises IgM-bmdmg substance and/or complement- binding substance, and optionally also IgG-bmdmg substance, and

(m) a separation medium having a density lower than the density of erythrocytes but higher than the density of an antibody-containing fluid.

Brief description of the drawings

Fig. 1: this figure shows a test card with six microcolumns on it. In a microcolumn (9) closed at the bottom, a matrix of inert particles (1) is provided to support a thin layer of affinity gel matrix (2) . On this affinity gel matrix there is a layer of separation medium of high density (3) . The incubation compartment (4) is formed by a widening at the top of the microcolumn. A microcolumn test system (7) can be integrated into a base plate (8) and be placed a centrifuge using supporting and guiding elements (11, 12, 13) .

The reference numerals denote:

1 matrix of inert particles

2 affinity gel matrix 3 separation medium of high density

4 incubation compartment

5 microcolumn wall

6 microcolumn bottom

7 test card 8 base plate

9 microcolumn

10 recess

11 top surface

12 slot 13 support

Fi . 2 : this figure shows the various possible reaction strengths. How many sensitized erythrocytes will bind to the affinity gel matrix depends on various parameters, such as the antigen density on the erythrocytes, the titer of the antibody m question, and the affinity of the antibody for the erythrocyte antigen m question. In general, in reactions weaker than 4+, erythrocytes will also be found at the bottom of the microcolumn.

The strength indications denote: 4+ all cells bound to the affinity matrix, no cells at the bottom 3+ more cells bound to the affinity matrix than at the bottom

2+ as many cells bound to the affinity matrix as at the bottom 1+ fewer cells bound to the affinity matrix than at the bottom 0.5 few cells bound to the affinity matrix, many cells at the bottom +/ - hardly any cells bound to the affinity matrix, nearly all cells at the bottom neg no cells bound to the affinity matrix, all cells at the bottom

10

Detailed description of the invention

The invention relates to a method for demonstrating antibodies and antigens, the application of this method lying mainly in the field of bloodgroup serology: bloodgroup determination, antibody screening and antibody identification, and cross tests.

The invention relates to an improvement of the solid-phase test as described by Pernell (Sanofi Pasteur, EP 0 594 506) . A disadvantage of this affinity gel test of Sanofi Pasteur, where an immunoglobulin-bindmg substance (such as protein A, protein G or anti-human IgG) is immobilized on the gel matrix, is that no complement factors are bound and that IgM-antibodies are not bound or only partly so. In serology, however, in addition to IgG-antibodies, IgM-antibodies are also relevant, as are erythrocytes which are loaded with complement factors (for instance, in demonstrating complement-binding antibodies such as anti-Jk antibodies, also referred to as complement-dependent antibodies) . To be able to also demonstrate these IgM-antibodies and/or complement - dependent antibodies with a high sensitivity, in the present invention, in addition to IgG-bindmg substances also IgM-bmding and complement -binding substances can be immobilized on the gel matrix, which leads to a lower number of false-negative reactions. The invention further encompasses the possibility that, immobilized on the gel matrix, an IgM-bindmg substance alone, or a complement - binding substance alone, or both, but without IgG-bmding substance, are utilized. According to the invention, sensitized erythrocytes can be bound directly to this affinity matrix without intervention of an anti -globulin serum.

Above the affinity matrix there is a fluid medium having a density between the density of erythrocytes and the density of the antibody-containing fluid. The presence of this layer provides that no separate washing step is 11

needed to separate non-bound antibodies from the erythrocytes sensitized or not. Above this layer, incubation of erythrocytes and antibodies takes place. After the incubation phase, a centrifugation phase takes place, such that the erythrocytes are centrifuged through- the high-density medium. The antibody-containing fluid remains above this high-density medium. In this manner, therefore, the erythrocytes are separated from non-bound antibodies, coming, for instance, from the serum or plasma. If complement factors and/or IgG antibodies and/or IgM antibodies are bound to the erythrocytes, these sensitized erythrocytes will bind to the affinity matrix and so will bind to the upper part of the gel column. Cells which have not been sensitized will not be bound to the affinity matrix and can readily pass it.

According to the invention, it has been found to be of essential importance that in addition to the optional IgG-bmdmg substance, complement -binding substance and/or IgM-bmdmg substance be immobilized on the solid phase and not be added loosely to the solid phase, as is the case, for instance, in the agglutination test described by Lapierre. Using a combination of IgG-bmdmg, IgM-bmdmg and complement-binding substances immobilized on the gel matrix, an unexpected positive result was achieved. Firstly, it was found possible to demonstrate, m addition to IgG antibodies, complement -dependent antibodies with a high sensitivity. In addition, it also proved possible to demonstrate agglutinins (bloodgroup antibodies of the IgM type, of importance in performing cross tests with serum from the recipient and erythrocytes from the donor) with a high sensitivity, including weak agglutinins.

Accordingly, the invention provides a composite solid phase and a method for determining m a sample an analyte selected from a bloodgroup antigen present on erythrocytes or 'an IgG antibody and/or IgM antibody and/or complement- dependent antibody binding to such a bloodgroup antigen. 12

The invention further provides a test kit comprising:

(1) a reagent containing an analyte-bmding partner which, in the case where the analyte is a bloodgroup antigen present on erythrocytes, is an antibody capable of binding to the bloodgroup antigen, and in the case ■ where the analyte is an IgG antibody and/or IgM antibody and/or complement-dependent antibody binding to a bloodgroup antigen, is the bloodgroup antigen present on erythrocytes,

(ii) a reaction vessel comprising an incubation zone, an immobilization zone and a collection zone, there being immobilized in the immobilization zone a binding substance which comprises IgM-bmding substance and/or complement-binding substance, and optionally also IgG-bmding substance, and

(iii) a separation medium having a density lower than the density of erythrocytes but higher than the density of an antibody-containing fluid. The invention described here has as an advantage over existing techniques that the sensitivity of the test, owing to the combination used of the IgG-bmding, IgM-bmdmg and complement-binding substances immobilized on the gel matrix, is unexpectedly high. Both agglutinins, including weak agglutinins, and complement - dependent antibodies can be demonstrated with a highly increased sensitivity over the conventional microcolumn tests. Accordingly, this leads to a lower number of false-negative reactions. The invention described can be applied to the current solid-phase tests within bloodgroup serology, including screen tests for antibodies, identification of antibodies, cross-tests, antigen determination, bloodgroup determination and antibody titration. The invention can be used, for instance, for determining the following bloodgroup antigens and antibodies to those 13

antigens: ABO, Kell (K) , cellano (k) , Duffy (Fy^a and

Fy^b) , Kidd (Jk^a and Jk^b) , Lutheran (Lu^a and Lu^b) , the

Rhesus system (D, E, e, C, c) , MNS , Lewis (Le^a and Le^b) , et cetera. By way of example, the invention can be used for the - following assays:

- Tests in which use is made of an antiserum of a known specificity and erythrocytes of unknown antigen composition (bloodgroup assay, antigen assay) . - Tests in which use is made of sera having therein an unknown antibody or unknown antibodies and erythrocytes having a known antigen composition (antibody screening, antibody identification, allo-antibody assay (also referred to as reverse grouping) . Here, often use is made of (a panel of) erythrocytes of known antigen composition. In addition, the test can also be used for the detection of auto-antibodies . In this case, antibodies are already bound to the corresponding antigens on the erythrocytes, so the erythrocytes are already sensitized and can bind directly to the composite solid phase.

- Tests utilizing serum from the recipient and erythrocytes from the donor to demonstrate any antibodies the serum from the recipient (cross test) .

- Tests performed m accordance with the present invention are all based on the same principle: erythrocytes to which antibodies are bound (sensitized erythrocytes) can bind to the composite solid phase, in contrast with non-sensitized erythrocytes.

The sample will typically consist of blood or a material derived therefrom, such as blood plasma, blood serum or a blood fraction, for instance erythrocytes isolated from blood. However, since the invention can also be used for other purposes, such as selecting hybridomas which produce monoclonal antibodies that bind to a bloodgroup antigen, the sample can also consist of 14

materials of a different kind, for instance of a supernatant of a hybridoma.

It is essential that during the incubation of antibodies with erythrocytes and the centrifugation step there be no contact between this incubation mixture and the composite solid phase, for which reason a medium of high density has been applied to the composite solid phase. Additionally of essential importance are: the choice of immunoglobulin-bindmg and complement -binding substances, the choice of the material to increase the density of the medium, and the choice of gel matrix material. These points will be briefly explained here, a. Immunoglobulin-bindmg substances. The choice from immunoglobulm-binding substances that can be immobilized on the solid phase is wide. Useful are, for instance, polyclonal or monoclonal antibodies directed to immunoglobulm G (IgG) , immunoglobulm A (IgA) and/or immunoglobulm M (IgM), or the Fab or F(ab) '₂ fragments thereof. In order to increase the sensitivity, or to detect a widest possible range of antibodies, it is also possible to immobilize a combination of several immunoglobulin-bindmg substances on the solid phase.

It is also possible to use lmmunoglobulm-bmdmg substances of a different origin, for instance from bacteria. lmmunoglobulm-bmdmg bacterial proteins are used, for instance, in immunology, biochemistry, and biotechnology. A number of lmmunoglobulm-bmdmg proteins have been isolated and characterized. The most important are protein A (Forsgren et al . J. Immunol. 1966; 97: 822) and protein G (Bjorck et al . J. Immunol 1984; 133: 969; Reis et al . J. Immunol. 132: 3091) .

For the tests involving IgG antibodies (antibody identification, antibody screening and cross tests) , use can be made of a solid phase with an IgG-bmding 15

substance of a high affinity, so that optimum binding of sensitized cells to a solid phase is ensured. As immunoglobulin-bindmg substance, in addition to anti -human immunoglobul , again protein G can be used. A solid phase on which both IgG-bmding substances and IgM-bmding substances are immobilized is preferred, because in that case a wider range of antibody specificities can be demonstrated. Any IgA-alloantibodies are, normally speaking, always accompanied by IgG antibodies of the same specificity. b. Complement-binding substances.

When complement is activated after the binding of complement -binding antibodies to an erythrocyte antigen, the activation products of the complement factors C4 and especially C3 , C4b and C3b, respectively, are bound on the erythrocyte membrane. Accordingly, useful as complement-binding substances are, for instance, polyclonal or monoclonal antibodies directed against antigenic determinants on relevant complement factors, such as, for instance, C3c, C3d and C3g on the complement factor C3b, or combinations hereof. c . Substances for increasing the density of the medium can be many kinds of substances, including, e.g., albumin, dextran, Ficoll, lodixanol , sodium diatrizoate, Percoll, etc. The density of the medium should be higher than that of the antibody-containing fluid, but lower than that of erythrocytes, so the density must lie between 1.01 and 1.09 g/ml , preferably between 1.02 and 1.06 g/ml. d. Gel matrix material.

Suitable as matrix material are, for instance, agarose (e.g. Sepharose 4 Fast Flow, Sepharose High Performance (Pharmacia)) or acrylic resin. Also of importance is the size of the particles: in a matrix with larger particles, and hence more interspace, the binding of sensitized erythrocytes will proceed slowly, so that 16

insufficient cells can be bounα. A finer matrix will lead to a more intensive contact of the sensitized cells with the immunoglobulin-binding ana complement -binding substances, so that optimum binding of the cells is possible. Too fine a matrix, however, will lead to clogging. The diameter of the particles will therefore have to lie between 10 and 125 μm, preferably between 20 and 50 μm.

The lmmunoglobulm-bmdmg and complement-binding substances can be directly immobilized on the solid phase through covalent binding, optionally via a spacer arm. Such a spacer arm can have a chain length of, for instance, 5-10 carbon atoms and has as an advantage that the binding substances (ligands) become better accessible to sensitized erythrocytes in that steπc hindrance is reduced. In the use of NHS-activated Sepharose (Pharmacia) as solid phase, for instance, a spacer arm is already present on this matrix material; when now a ligand with primary ammo groups is presented, there arises, with cleavage of N-hydroxysuccmimide from the activated ester compound, a solid phase to which ligands are bound via a spacer of 6 carbon atoms (i.e. a bridge of a chain length of 6 C atoms) . It is also possible, however, first to immobilize, for instance, protein G on the gel matrix and then to couple anti -human complement and/or anti -human IgM to protein G itself, either directly or through so-called ligand bridges, for instance bridge-forming antibodies, which are first coupled to the protein G and are directed to human antibodies.

For carrying out the test, a test system is preferred which consists of a card with microcolumns manufactured from a transparent inert plastic, allowing single and multiple tests to be performed (see also Fig. 1) In these microcolumns, a matrix of inert particles is provided (e.g. glass beaαs (Supelco) or 17

Streamline (quartz base matrix, Pharmacia) ) . This inert matrix serves to support a thin layer of gel matrix on which lmmunoglobulm-bmdmg and complement -binding substances are immobilized. Provided on this affinity gel matrix is a layer of separation medium of a density higher than that of the antibody-containing fluid but lower than that of erythrocytes. This layer is of importance to the invention, since the presence of this layer eliminates the necessity of a separate washing step to separate non-bound antibodies from the (sensitized) erythrocytes. The incubation of erythrocytes and antibodies can take place in a compartment formed by a widening at the top of the microcolumn. After the incubation phase m which antibodies can bind to the cells, the erythrocytes, whether sensitized or not, are separated from non-bound antibodies by centrifugation through the separation medium mentioned, and transported to the affinity gel matrix, whereafter the sensitized cells will adhere to the affinity gel matrix, while the non-sensitized cells will be transported through the affinity gel matrix to the bottom of the microcolumn. The required centrifugation protocol, as regards centrifugal force (g-value) and duration, depends inter alia on the amount of erythrocytes to be used, the density of the separation medium mentioned, and the particle size of the gel matrix. It is possible, for instance, to opt for a so-called linear centrifugation protocol, whereby the g-value rises in a particular time period (for instance 1 mm) to a given maximum value (e.g. 2000xg) , which is maintained for a certain period of time (e.g. 9 mm) . As the g-value rises, the erythrocytes will be transported through the separation medium, allowing the sensitized erythrocytes to bind to the affinity matrix. Thereupon, during the period at maximum g-value, all non-sensitized erythrocytes will be transported through the affinity matrix to the bottom of the microcolumn. For carrying out the test, a so-called discontinuous centrifugation protocol is preferred, whereby during the 1st phase, at a particular g-value (e.g. 10-60 sec at 400-1000xg, preferably 20-40 sec at 400-600xg) , erythrocytes will be transported through the separation medium, whereafter, in - the 2^nd phase, at a lower g-value (e.g. 10-60 sec at 50-400xg, preferably 20-40 sec at 100-300xg) , the sensitized erythrocytes can bind to the affinity matrix, whereafter during the last phase, at maximum g-value (e.g. 1-13 mm at 1000-2500xg, preferably 7-10 mm at 1800-2000xg) all non-sensitized erythrocytes are transported through the affinity matrix to the bottom of the microcolumn. It is of essential importance that all non-sensitized cells end up at the bottom, so as to prevent false-positive reactions, so centrifugation m the last phase should be relative long and at high speed.

Detection of the bound cells on the affinity gel matrix and non-bound cells at the bottom of the microcolumn can simply be done both visually and automatically. The various possible reaction patterns are shown m Fig. 2. How many sensitized erythrocytes will bind to the affinity gel matrix depends on various parameters, such as the antigen density on the erythrocytes, the titer of the antibody m question, and the affinity of the antibody for the erythrocyte antigen in question. In general, m reactions weaker than 4+, cells will also be found at the bottom of the microcolumn.

It has proved to be of essential importance for the invention that the immunoglobulin-binding substances and complement -binding substances be immobilized on the solid phase and not be added loosely to the gel matrix, as is the case, for instance, the column agglutination test described by Lapierre (EP 0 305 337) . Preferably used as immunoglobulm-binding substances are protein G and antibodies directed against human IgM (anti -human IgM) . 19

As complemen -binding substances, preferably antibodies are used which are directed against the human complement factor C3d (anti-human C3d) . Using the combination of protein G, anti -human IgM and anti -human C3d immobilized on the gel matrix, an unexpected positive result was achieved :

Firstly, it proved possible to properly demonstrate complement -dependent antibodies. When, on the other hand, protein G alone was immobilized on the gel matrix, these complement -dependent antibodies were only weakly demonstrable. When immunoglobulin-binding substances and complement-binding substances were not immobilized but were added loosely to the gel matrix (in this case anti -human IgG and anti -human C3d, whereby m fact an indirect antiglobulm test was obtained, as described by Lapierre) , the complement-dependent antibodies were hardly demonstrated, if at all.

Secondly, it also proved possible to properly demonstrate agglutinins: - bloodgroup antibodies of the IgM type, which still gave sensitization of the erythrocytes but no visible agglutination anymore m the tube test, were still properly demonstrable in the present test, m contrast to the situation where protein G alone was immobilized on the gel matrix. When anti -human IgM and anti -human C3d were added loosely to the gel matrix, the diluted IgM bloodgroup antibodies were hardly, if at all, demonstrable. When gel matrix alone, without ligands, was used, the diluted IgM bloodgroup antibodies were not demonstrable. This unexpected result could be explained by the fact that the present test is an affinity test, which is not dependent on the sieve action of the gel matrix for holding agglutinates, but wherein individual sensitized erythrocytes can still be bound to the gel matrix. 20

- when weak ABO-mcompatibilities were tested, e.g. A₂B erythrocytes/B-plasma incompatibility, the reactions were properly demonstrable m the present test, m contrast to the situation where protein G alone was immobilized on the gel matrix. When anti-human IgM and anti -human C3d were added loosely to the gel matrix, the reactions were hardly demonstrable, if at all. When gel matrix alone, without ligands, was used, the reactions were not demonstrable. It is a well known given that "neutral" and antiglobulm microcolumn tests have problems m demonstrating weak ABO-mcompatibilities (Cummins & Downham, Lancet 1994; 343:1649 and Philips et al . , Transfusion Medicine 1997; 7:47-53), possibly m that so^*-called "shear forces" during centrifugation cause la"rger agglutinates to disintegrate into several small agglutinates or, in an extreme case, into individual (sensitized) erythrocytes. In spite of the screening action of the gel, only weak to negative reactions are observed then, resulting in an increase of the number of false-negative reactions. As a possible solution to this problem, in the article of Philips et al . the Diamed test is performed according to a modified, discontinuous centrifugation protocol . by lowering the centrifugal force and tripling the centrifugation time, the weak agglutinates now prove to be better demonstrable, however, this modification is qualified as unsuited for routine use. Now, therefore, the present invention has the advantage that smaller agglutinates and even individual sensitized erythrocytes can still be bound to the affinity matrix, so that this kind of weak reactions can still be demonstrated well .

For that matter, it proved possible to demonstrate the weak agglutinins described equally well with a combination of solely protein G and anti-human IgM, immobilized on the gel matrix, and also with anti-human IgM alone, immobilized on the gel matrix. 21

Likewise, it proved possible to demonstrate the complemen -dependent antibodies equally well with a combination of jus protein G and anti-human C3d, immobilized on the gel matrix, and also with anti -human C3d alone, immobilized on the gel matrix.

When these last four variant embodiments are used in addition to the described variant with the immobilized combination of protein G, anti-human IgM and anti-human C3d, it can be discriminated whether a given sample contains IgG-antibodies and/or IgM-antibodies and/or complement -dependent antibodies.

Examples

The invention will now be further explained in and by the following examples. It is of importance to see, however, that these examples are given for illustrative purposes and illuminate the enablement of the invention but do not in any way constitute the definitive conditions of the different assay methods nor limit the scope of the invention in any way.

Example 1

Glass beads (Supelco) were washed twice m a phosphate-buffered saline solution (PBS) and included in a 20 mM phosphate buffer, further containing 70 mM NaCl, 20 mM EDTA, 30 mM glucose, 5% dextran, 1.5% BSA, 0.02% gelatin, pH 8.8 (gel buffer).

10 ml Sepharose HP having immobilized thereon protein G (binding capacity for human IgG: 20-30 mg/ml gel, Pharmacia, Uppsala, Sweden)) was washed according to the instructions of the supplier and included in a buffer containing 20 mM Tris and 150 mM NaCl. To this was added monoclonal mouse IgG, directed to human IgM, to a final concentration of 100 μg/ml gel and monoclonal mouse IgG, directed to human C3d, to a final concentration of

75 μg/ml gel . After an incubation period of 12 hours at 22

4°C the affinity gel was washed twice m gel buffer. After each washing step, the supernatant was checked for the presence of loose anti -human IgM and anti -human C3d: these components were not demonstrable, so that all added anti -human IgM and anti -human C3d had to be coupled to the protein G. This affinity gel was then included in 25 ml gel buffer. Per microcolumn there was added: 15 μl settled glass beads in gel buffer. To this was added: 15 μl affinity gel suspension in gel buffer (20%) . This affinity gel sedimented within 15 min on top of the glass beads. The supernatant gel buffer functions as high-density separation medium.

Erythrocytes of bloodgroup 0⁺ (i.e. bloodgroup 0, rhesus D positive) were sensitized with anti-D and then washed 3x in a modified LISS (low ionic strength saline) solution. Of these erythrocytes, a 0.5% suspension was made in LISS . Then 1 drop of this suspension was introduced into the incubation compartment of the microcolumn. To this was added 1 drop of normal serum from a donor of bloodgroup 0⁺. After an incubation period of 15 min at 37°C the microcolumn was centrifuged according to the following protocol: 30 sec at 500xg, followed by 30 sec at 200xg, followed by 9 mm at 1950xg. Then the result was assessed: all erythrocytes had been bound to the upper side of the affinity gel matrix (reaction strength 4+) . As control, the procedure described was repeated with non-sensitized erythrocytes of bloodgroup 0⁺. These cells were not bound to the affinity gel matrix.

Example 2

Use was made of the test system as described in Example 1.

Of Duffy-a (Fy^a) positive erythrocytes (phenotype Fy^dtb*) a 0.5% suspension LISS was made. Then 1 drop of this suspension was introduced into the incubation 23

compartment of the microcolumn. To this was added 1 drop of antι-Fy^a (polyclonal IgG antiserum, CLB, Amsterdam, Netherlands) . After an incubation period of 15 mm at 37°C the microcolumn was centrifuged according to the protocol of Example 1. After this, the result was assessed: all erythrocytes were bound to the top side of the affinity gel matrix (reaction strength 4+) . As control experiment, corresponding non-sensitized erythrocytes were tested. These cells were not bound to the affinity gel matrix.

This experiment was repeated for the following antigens: Fy° (with antι-Fy° serum with a Fy^a+b* cell), Jk^a (with antι-Jk^a serum and Jk^atb+ cell) , Jk^b (with antι-Jk^b serum and Jk^a*b+ cell) , Kell (with anti-K serum and K⁺k⁺ cell) , cellano (with anti-k serum and r k^**" cell) , Lu^a

(with antι-Lu^a serum and Lu^atb+ cell) , Lu^b (with antι-Lu^b serum and Lu^{a b+} cell), S (with anti-S serum and S⁺s⁺ cell) and s (with anti-s serum and S⁺s⁺ cell) . The results were in agreement with what has been described for Fy^a. In all cases, complete binding of sensitized erythrocytes to the affinity gel matrix occurred, while non-sensitized erythrocytes in the control experiments were not bound to the affinity matrix.

In the examples below, use was made of the following test systems: system 1 : the test system as described in Example 1 having as immobilized ligands on the gel matrix: protein G, anti-human IgM and anti-human C3d. system 2 as system 1, but with protein G alone as immobilized ligand on the gel matrix. system 3 as system 1, but with anti -human IgM alone as immobilized ligand on the gel matrix. To that end, an optimum amount of anti -human IgM was coupled to NHS-activated Sepharose HP according to the instructions of the supplier (Pharmacia) After the coupling, any 24

residual activated groups were inactivated and the affinity gel was washed twice in gel buffer. After each washing step, the supernatant was cnecked for the presence of loose anti-human IgM: this was not demonstrable, so that all added anti -human IgM had to be coupled to the gel matrix. system 4: as system 1, but without immobilized ligands on the gel matrix. To that end, NHS-activated Sepharose HP was inactivated according to the instructions of the supplier (Pharmacia) . system 5: as system 4, except that subsequently per microcolumn optimum amounts of lmmunoglobulm-bmdmg and complement -binding substances, anti -human IgG and anti- human C3d, respectively, were added loosely, such that this was homogeneously distributed over glass beads, gel matrix and supernatant gel buffer. This in fact yielded an anti -globulin test system. system 6: as system 5, except that immediately prior to the deployment of the test the supernatant gel buffer having therein loose anti-human IgG and anti-human C3d was replaced with normal gel buffer, so that the loosely added anti -human IgG and anti -human C3d were disposed only in the section with gel matrix and glass beads . system 7: as system 4, except that subsequently per microcolumn optimum amounts of immunoglobulin-bind g and complement -binding substances, anti -human IgM and anti-human C3d, respectively, were added loosely, such that this was homogeneously distributed over glass beads, gel matrix and supernatant gel buffer. This m fact yielded an anti-globulm test system. system 8: as system 7, except that immediately prior to the deployment of the test the supernatant gel buffer having therein loose anti -human IgM and anti -human C3d was replaced with normal gel buffer, so that the loosely added anti -human IgM and anti-numan C3d were 25

disposed only in the section with gel matrix and glass beads .

Example 3 A number of the above-mentioned test systems were examined for their ability to demonstrate complement - dependent antibodies. For this purpose, two patient sera were tested which were known to contain complement - dependent anti-Jk^a antibodies (IgG) . Of Jk^a+b cells, a 0.5% suspension in LISS was prepared. Then 1 drop of this suspension was introduced into the incubation compartment of the microcolumn. To this was added 1 drop of patient serum. As control, this experiment was carried out with AB-serum instead of patient serum. After an incubation period of 15 min at 37°C, the microcolumn was centrifuged according to the protocol of Example 1. After this, the result was assessed. The reaction strengths are summarized in Table 1:

Table 1 system 1 system 2 system 5 system 6 patient serum 1 3 + 1+ 0.5 0.5 patient serum 2 1 + + / -

AB-serum — — —

Example 4

A number of the above-mentioned test systems were examined for their ability to demonstrate weak agglutinins. To that end, a human monoclonal anti-D antibody of the type IgM (anti-D pelikloon, CLB) was diluted in AB-serum, such that sensitization of 0^* test erythrocytes still occurred, but no visible agglutination occurred anymore m a normal tube agglutination test a physiological salt solution. Of 0^* test erythrocytes (CLB) a 0.5% suspension m LISS was prepared. Then 1 drop 26

of this suspension was introduced into the incubation compartment of the microcolumn. To this was added 1 drop of diluted anti-D. As control, this experiment was carried out with AB-serum. After an incubation period of 15 mm at 37°C, the microcolumn was centrifuged according to the protocol from Example 1. Then the result was assessed. The reaction strengths are listed in Table 2:

Table 2 syst 1 syst 2 syst 3 syst 4 syst 7 syst 8 anti-D, diluted 2+ 0.5 2+

AB-serum

Example 5

A number of the above-mentioned test systems were examined for their ability to demonstrate agglutinins. To that end, an ABO-incompatibility was tested, viz. an A₂B erythrocyte sample with a bloodgroup B plasma which in a normal agglutination test in a physiological salt solution still gave a 1+ reaction. Of the A₂B erythrocytes, a 0.5% suspension in LISS was prepared. Then 1 drop of this suspension was introduced into the incubation compartment of the microcolumn. To this was added 1 drop of the B plasma. As control, this experiment was carried out with A₂B plasma of the A₂B erythrocytes donor (autocontrol) . After an incubation period of 15 mm at 37°C, the microcolumn was centrifuged according to the protocol from Example 1. After this, the result was assessed. The reaction strengths are listed in Table 3:

Table 3 syst 1 syst 2 syst 3 syst 4 syst 7 syst 8

B plasma 2 + 0.5 2 + _ _ + /-

A_^B plasma — — —

It appears from Tables 2 and 3 that with system 2 27

(protein G alone as immobilized ligand on the gel matrix) a reaction is obtained which, though less strong, is still positive. Since with a neutral gel without ligands (system 4) a negative reaction is obtained, this is apparently the result of binding of IgM-erythrocyte complexes to protein G.

Claims

28Cl aims

1. A method for determining m a sample an analyte, selected from a bloodgroup antigen present on erythrocytes or an antibody binding to such a bloodgroup antigen, comprising treating the sample m an incubation zone of a reaction vessel with a reagent containing an analyte-bmding partner, which analyte-bmdmg partner, m the case where the analyte is a bloodgroup antigen present on erythrocytes, is an antibody capable of binding to the bloodgroup antigen, and in the case where the analyte is an antibody binding to a bloodgroup antigen, is the bloodgroup antigen present on erythrocytes, wherein, in the incubation zone, if the sample contains the analyte, either a complex of bloodgroup antigen present on erythrocytes and antibody bound thereto is formed, or a complex of bloodgroup antigen present on erythrocytes and antibody bound thereto is formed, as well as a complex of complement factors bound to these erythrocytes if the antibody is complement-binding, further comprising separating erythrocytes, complexed or not, from non-bound antibodies using a separation medium, located m the reaction vessel under the incubation zone, of a density higher than that of the fluid containing the antibodies but lower than the density of erythrocytes, whereby erythrocytes pass through the separation medium and non-bound antibodies remain in the incubation zone, separating complexed erythrocytes from non-complexed erythrocytes by binding complexed erythrocytes in an immobilization zone of the reaction vessel to binding substance immobilized in this zone, which binding substance comprises IgM-bmding substance and/or complement -binding substance, and optionally also IgG-bmdmg substance, and discharging non-complexed erythrocytes to a collection zone of ' the reaction vessel, and 2 9

detecting erythrocytes m the immobilization zone and/or collection zone.

2. A method according to claim 1, cnaracterized m that the sample consists of blood, blood plasma, blood serum, a blood fraction or a supernatant of a hybridoma.

3. A method according to claim 1 or 2 , characterized in that as IgG-bmdmg substance, an IgG-bmdmg bacterial protein, such as protein A or protein G, is used.

4. A method according to claim 1 or 2 , characterized in that as IgG-bmdmg substance, a recombmant protein having IgG-bmdmg properties, such as protein A/G, is used.

5. A method according to claim 1 or 2 , characterized in that as IgG-bmdmg substance, monoclonal or polyclonal antibody against IgG is used. '

6. A method according to one or more of claims 1-5, characterized in that as IgM-bindmg substance, monoclonal or polyclonal antibody against IgM is used.

7. A method according to one or more of claims 1-5, characterized in that as IgM-bmdmg substance, an IgM-bmding bacterial protein or a recombmant protein with IgM-bmdmg properties is used.

8. A method according to one or more of claims 1-7, characterized m that as complement-binding substance, polyclonal or monoclonal antibody against an activation product of the complement factor C4 or the complement factor C3 is used.

9. A method according to claim 8, characterized m that an antibody is used which is directed against an antigenic determinant on C3b or C4b, e.g. against C3c, C3d or C3g on C3b.

10. A method according to one or more of claims 1-9, characterized m that the immobilized substance comprises both an IgG-bmdmg substance and an IgM-bmdmg substance and a complement -binding substance.

11. A method according to one or more of claims 1-10, characterized that one or more of the binding substances 30

are immobilized in the immobilization zone either by direct covalent binding, optionally via a spacer arm, or via a ligand bridge.

12. A method according to claim 11, characterized in that the spacer arm contains a chain of 5-10, preferably -

6 carbon atoms .

13. A method according to one or more of claims 1-12, characterized in that as immobilization zone, NHS-activated Sepharose is used.

14. A method according to one or more of claims 1-13, characterized m that the analyte is a bloodgroup antigen present on erythrocytes.

15. A method according to one or more of claims 1-13, characterized in that the analyte is an antibody binding to a bloodgroup antigen.

16. A method according to one or more of claims 1-13, characterized m that a cross-test is carried out in that erythrocytes of unknown bloodgroup antigen composition and antibodies of unknown bloodgroup specificity are used.

17. A test kit suitable for use in a method for determining an analyte in a sample, the analyte being a bloodgroup antigen present on erythrocytes or an antibody binding to such a bloodgroup antigen, comprising:

(I) a reagent containing an analyte-bmdmg partner which, the case where the analyte is a bloodgroup antigen present on erythrocytes, is an antibody capable of binding to the bloodgroup antigen, and in the case where the analyte is an antibody binding to a bloodgroup antigen, is the bloodgroup antigen present on erythrocytes, (n) a reaction vessel comprising an incubation zone, an immobilization zone and a collection zone, there being immobilized in the immobilization zone a binding substance which comprises IgM-bmdmg substance and/or complement - binding substance, and optionally also IgG-bmdmg substance, and 3 1

(iii) a separation medium having a density lower than the density of erythrocytes but higher than the density of an antibody-containing fluid.