WO1996041176A1 - Immunoassay for determining blood cell activation - Google Patents

Abstract

A method for determining the activation level of platelets and other types of blood cells that undergo activation by ELISA and other immunoassay techniques includes the step of reacting a sample containing the blood cells in a liquid phase with an excess quantity of an activation-specific primary antibody prior to allowing the cells in the sample to bind to a solid surface. This differs from typical ELISA procedures, wherein the cells are bound to the surface of a plastic test tube or microtiter plate prior to addition of the primary antibody. It has been discovered according to the invention that binding the cells to a surface prior to reaction with the primary antibody changes the activation level of the cells, making the assay less accurate. In addition, whole platelet ELISA, like flow cytometric analyses, can sensitively detect activated platelets.

Description

IMMUNOASSAY FOR DETERMINING BLOOD CELL ACTIVATION

TECHNICAL FIELD This invention relates to an immunoassay, particularly an enzyme-linked immunoabsorbent assay (ELISA) for monitoring activation of platelets and other types of blood cells which undergo activation.

BACKGROUND OF THE INVENTION The use of fluorescent-activated flow cytometry with specific activating antibodies has allowed investigators to directly monitor platelet activation and subsequent microaggregate formation. While accurate and specific, flow cytometry is extremely expensive, requiring an expensive machine and maintenance, and a dedicated technician. P-selectin (e.g., CD62) is used as a means of detecting activated platelets in the systemic circulation by flow cytometry. P-selectin (also known as GMP-140 or PADGEM) is an a-granule protein which is expressed on the surface of endothelial cells and on platelets only after activation and release. P-selectin has been implicated in recognition of monocytes, macrophages and neutrophils during the inflammatory response, and in particular, as a first step in the regulation of leukocyte migration through endothelium.

Platelet assay methods using other well-known assay systems, such as ELISA, radioimmunoassay (RIA) and Western blot analysis have been described in Berman et al.. Methods in Enzymolocfy 169:314, 1989. Berman et al. discuss an ELISA assay which involves fixing platelets and placing the fixed platelets directly in a microtiter plate for the purpose of screening for activation specific antibodies.

Current assays for monitoring platelet activation in patients who may be at risk for thrombosis or thrombosis- related conditions involve indirect assessment of this activation by measuring systemic released platelet products (factors present in the circulation) . Such products include platelet factor 4 and β-thromboglobulin, or the thromboxane B2, a stable product of platelet activating factor thromboxane A₂. Unfortunately, circulating levels of these f ctors can undergo metabolically-produced changes, and as such these factors are considered indirect measures of platelet activation. As such, a need remains for a direct platelet activation assay which can be carried out on patient samples without undue complexity or expense.

SUMMARY OF THE INVENTION

A method according to the invention for determining the activation level of blood cells by an immunoassay of the type wherein a primary antibody undergoes specific binding with the blood cells and a labelled secondary antibody undergoes specific binding with the primary antibody-blood cell complex, such as ELISA, includes the step of reacting a sample containing the blood cells in a liquid phase, e.g. an aqueous suspension, with an excess quantity of an activation-specific primary antibody prior to allowing the cells in the sample to bind to a solid surface. This differs from typical ELISA procedures, wherein the cells are bound to the surface of a plastic test tube or microtiter plate prior to addition of the primary antibody. It has been discovered according to the invention that binding the cells to a surface prior to reaction with the primary antibody changes the activation level of the cells, making the assay less accurat .

The activated platelet ELISA assay of the invention is a relatively simple procedure for quantitation of the amount of activated platelets in the bloodstream of patients who may be at risk for thrombosis or thrombosis- related conditions. Accordingly, a method for controlling the activation level of platelets or other blood cells in a patient's bloodstream according to the invention involves the steps of obtaining a sample of platelets or other blood cells which undergo activation, immunoassaying the sample with an antibody that selectively binds to activated platelets or other blood cells while the sample preferably remains in liquid suspension, correlating the results with a standard to determine the degree of activation of the platelets or cells prior to the assay, and administering a drug to the patient in an amount effective to maintain activation at a predetermined level. For platelets, if the assay reveals an elevated level of activation, the physician then administers an anticoagulant or antiplatelet agent or increase the anticoagulant or antiplatelet agent dosage. If the opposite is true, the anticoagulant or antiplatelet agent dosage can be decreased.

According to a preferred aspect of the invention, the assay includes the steps of:

(A) obtaining a sample of platelets from a living subject;

(B) adding an excess quantity of a platelet activation-specific primary antibody to the platelets in an aqueous suspension and allowing the primary antibody to react with the platelets for a time sufficient to allow substantially complete specific binding between the antibody and activated platelets;

(C) forming a complex between the platelets bound to the primary antibody and a secondary antibody bound to an enzyme, in which the secondary antibody reacts by specifically binding to the primary antibody;

(D) adding a substrate to the platelet complexes, which substrate reacts with the enzyme to produce a visible indicator;

(E) measuring the extent to which the visible indicator is present; and

(F) correlating the results with a standard, such as a standard curve, to determine the degree of activation of the platelets prior to the assay. For purposes of the invention method, an "excess" quantity is an amount sufficient to react with all of the available binding sites. In one embodiment, step (C) further comprises adding a labelled secondary antibody to the suspension which reacts with and specifically binds to the primary antibody. The label is one of a pair of specific binding substances, such as biotin, which binds specifically to the other substance, such as streptavidin. The platelets are separated from unbound labelled secondary antibody, and an enzyme labelled with the other of the pair of specific binding substances is added to the platelets so that the enzyme becomes bound to the secondary antibody. The platelets are then separated from unbound enzyme. In the alternative, the secondary antibody may be linked to the enzyme in advance, and the resulting complex reacted directly with the platelet having primary antibody bound thereto.

The invention further contemplates a kit for carrying out the assay of the invention. The kit includes necessary reagents, including standards containing fixed platelets activated at different, predetermined levels for generating a standard curve, and the necessary primary and optionally secondary antibody preparations, as well as written directions giving a protocol for carrying out the assay.

According to an additional aspect of the invention, the principle of measuring P-selectin is used as a specific marker for platelet activation to develop a whole platelet ELISA for detecting activated platelets in platelet-rich plasma (PRP) . The accuracy and specificity of the assay is verified via fluorescence-activated flow cytometric analysis. The quantitative assay is based on the use of a standard curve generated from activation of platelets in pooled PRP with phorbol yristate acetate (PMA) or thrombin receptor peptide. The assay is applied to assess platelet activation in patients diagnosed with unstable angina compared to normal controls. BRIEF DESCRIPTION OF THE DRAWINGS The drawings are graphs of experimental results on a semi-logarithmic scale, wherein:

Figures 1 and 2 are plots of optical density (OD) against collagen concentration (CC) in μg/ml for two examples of assays according to the invention; and

Figures 3 and 4 are plots of optical density (OD) against thrombin concentration (TC) in μg/ml for two examples of assays according to the invention. Figures 5A and 5B are agonist-induced dose response curves for washed platelets. Washed platelets were treated with increasing concentrations of either thrombin (5A) or collagen (5B) and subsequently measured by enzyme-linked indirect antibody detection. Figs. 6A and 6B show a flow cytometric analysis of resting and thrombin-activated platelets. The degree of platelet activation as indicated by CD-62 (6A) or CD-63 (6B) antigen expression was carried out to assess relative changes in platelet activation during preparation of homologous whole blood (WB) , PRP, and washed platelets (PLTS) . The symbols for the respective lines are indicated within the legends. The exception is the resting PRP in Fig. 6A, which is superimposed on the whole blood sample. In Figure 6A: WB, 0.75%, PRP, 0.84%, resting PLTS, 59.9%, and thrombin-activated PLTS, 97.2%; Fig. 6B - WB, 0.39%, PRP, 0.46%, resting PLTS, 8.6%, and thrombin-activated PLTS, 90%.

Figures 7A-7C are agonist-induced dose response curves of PRP. The dose-response curves corresponding to collagen (7A) , thrombin receptor peptide (7B) , and PMA (7C) activation of PRP were measured by flow cytometry.

Figures 8A and 8B are standard curves for PMA- activated platelet ELISA. Figs. 7A and 7B demonstrate curves that have correlation coefficients of 0.976 and 0.993, respectively.

Figures 9A-9D show the results of flow cytometric analysis of samples from a representative control subject (9A) and a patient with unstable angina (9B) together with whole blood resting (9C) and activated (9D) controls. Samples were fixed in 1.8% paraformaldehyde for 30 min and after washing, were stained with CD-61- FITC (FLl) and CD-62-PE (FL2; % of platelets expressing P-selectin) and evaluated for the degree of activated platelets. The resting and activated controls were prepared from whole blood obtained from a normal volunteer where the resting control was immediately fixed and a sample of the blood was activated with 20 micromoles APP for 15 min, and then fixed and stained. The results were (9A) resting control (0.1%), (9B) activated control (57%) ; (9C) patient, control subject (7.5%), and (9D) patient, unstable angina (45%). Figure 10 is a scattergram of flow cytometric and ELISA data on the percent P-selectin expression in fixed PRP samples from control subjects and patients with unstable angina.

DETAILED DESCRIPTION The assay of the present invention can eliminate the need for any equipment other than that commonly found in an ordinary clinical/or research laboratory. The following description deals with a platelet assay, but the procedure is also applicable to samples of other types of cells which undergo activation, for example, blood cells other than red blood cells, such as monocytes, lymphocytes, and granulocytes.

The invention is based in part on the discovery that affixing platelets or the associated primary antibody that binds thereto to a plastic surface as part of the ELISA assay causes surface activation of the platelets, rendering the assay inaccurate. This effect is most pronounced for plastics which have a substantial electrical surface charge, such as polyethylene and polystyrene. However, it has been discovered that an

ELISA assay for platelet activation can be carried out if binding of the primary antibody occurs in aqueous suspension. In practice, this means taking care to avoid binding the platelets to a plastic surface, such as the bottom of a microtiter well, until after the platelets have reacted with the primary antibody. Accordingly, according to assay of the invention, the platelet suspension is placed in a plastic microtiter well or test tube but not centrifuged until after the reaction between the primary antibody and the platelets has taken place. In accordance with one embodiment of the assay method, platelets are obtained from whole blood by standard centrifugation procedures, for example, using a standard laboratory centrifuge at room temperature and spinning for 15 minutes at 350 times gravity (xG) . The assay can be performed with either platelet-rich plasma, or washed platelets. The procedure for washing the platelets may be the one described in Amrani et al., Blood. Vol. 72, No. 3, 1988, pp. 919-924, the contents of which are incorporated by reference herein. In preparing platelets taken from normal individuals for use as standards, washed platelets are preferred to provide better uniformity in the results. For the patient (unknown) sample to be measured, the procedure may be simplified by using platelet-rich plasma. After washing, the platelets are then resuspended in a buffer solution at a pH of about 7.3 to 7.4. In general, buffer solutions according to the invention are formulated to maintain the platelets in a stable condition by maintaining salt concentrations and pH at normal plasma levels. A calcium inhibitor, e.g., a chelating agent such as ethylenediaminetetraamine (EDTA) or citrate, is preferably added to the platelet suspension prior to activation. The calcium inhibitor is added in an amount effective to chelate substantially all calcium ions in the aqueous suspension so that calcium does not bind to FC receptors on the platelet surface, leading to unwanted binding of S12 antibody containing FC regions of the antibody at these sites. The amount of calcium inhibitor added is generally in the range of 2 to 10 mM in the suspension.

The platelets are then activated at room temperature with one of a number of known agonists such as thrombin (for washed platelets only) in an amount ranging from 0 units/ml to 1.0 units/ml, or collagen (for either platelet-rich plasma or washed platelets) in an amount ranging from 0 μg/ml to 400 μg/ml. A specific primary antibody which is directed at platelet-activatiorr specific antigens (receptors) is also added, and the mixture is allowed to react for a time sufficient to allow the binding reaction to go to substantial completion, generally 30-60 minutes. Suitable primary antibodies include S12 or CD63, and others which are commercially available. CD63, available from AMAC Inc., Westbrook, ME binds to glycoprotein (GP) 53 lysosomal protein. Antibody S12, described in McEver and Martin, J. Biol. Chem. 259, 9799 (1984), is directed against granule membrane protein 140. After the antigen-antibody reaction has proceeded to completion, one or more inhibitors of platelet activation are then optionally added to minimize further activation and antibody binding. Such inhibitors include prostaglandin El (PGE-L) , preferably used in an amount of 5 ng to 100 ng per ml of carrier liquid, preferably ethanol, typically 10 ng/ l ethanol, and hirudin, a specific thrombin inhibitor used at the same concentration as thrombin.

A secondary antibody is added to link to the platelet-attached primary antibody (i.e., the activation- specific antibody) . The secondary antibody is generally either coupled in advance with an enzyme, as in Example 1 below, or else is labelled with a substance such as biotin. In the latter embodiment, a biotin-labeled anti- mouse immunoglobulin, preferably goat or rabbit anti- mouse Ig depending on the antibody used, is added to link to the platelet-attached primary antibody. The reaction is allowed to proceed in the mixture for a sufficient time, typically 30-60 minutes.

The platelet-antibody mixture is then separated by centrifugation from a supernatant material containing any antibody remaining unbound either directly, or through a dense medium such as sucrose or silicone oil. Centrifugation time may vary from about 2 seconds to 10 minutes depending on the medium used. For example, a 6 minute spin at 10,000 xG is preferred for a 5:1 silicone- oil mixture for washed platelets, and 8 minute spins at 10,000 xG are used for 8:1 silicone-oil mixture.

The platelet pellet is then washed 1 to 6 times with buffer, and an enzyme bound to streptavidin, such as streptavidin peroxidase, is added and allowed to react for 15-60 minutes. Avidin is a molecule that binds specifically and with high affinity to biotin. Preferred enzymes include peroxidase and alkaline phosphatase. If the enzyme is pre-linked to the secondary antibody as in Example 1 below, then the use of labels (specific binding substances) such as biotin and streptavidin may be omitted, along with the associated separation steps. The label for the secondary antibody could also be a

12c fluorescent label, radioactive label such as ³I, or other suitable marker substance. However, an enzyme suitable for ELISA has proven practical and sensitive, and is preferred.

After the reaction period, the pellet is again washed, e.g. 2-6 times, and a substrate such as 2,2- azino-di(3-ethylbenzylthiazoline) sulfonic acid (ABTS) is added to the pellet to develop the color. The samples are removed from the tubes and placed in a microtiter plate, and absorbance is read at a standard wavelength.

The foregoing procedure is used to construct a standard curve by activating platelets to predetermined levels based on the amount of activating agent used. The procedure can then be repeated without prior in vitro activation to determine activation levels for unknown samples. In assaying the unknown (patient) sample, it is preferred to use platelet-rich plasma; the washing procedure is accordingly omitted. Whole blood is centrifuged at a relatively low speed, e.g. 350 xG, to obtain platelet-rich plasma. The steps of adding the activator and inhibitor are likewise omitted. Otherwise, the procedure remains substantially the same. The activation level of the unknown sample is then read from the linear portion of the standard curve.

In the foregoing procedure, each successive step is preferably carried out with the platelets remaining in the same tube or microtiter well. Since polypropylene has less surface charge and less tendency to cause the platelets to bind thereto spontaneously, use of a polypropylene microtiter plate, particularly a Nunc pre- siliconized polypropylene microtiter plate, is most preferred.

With a sufficiently large pool of platelet samples from different donors, it is believed that a generalized standard curve could be developed. Accordingly, instead of generating a standard curve for each determination, the standard according to the invention may be a predetermined standard curve that correlates activation level with the observed color change, fluorescence or similar indication. In either embodiment, the invention provides a method for identifying patients who may be at risk for thrombosis or thrombosis-related conditions, such as artificial heart recipients. Anticoagulant or antiplatelet drugs are commonly administered to such patients. In a method for prevention of thrombosis according to the invention, the assay of the invention is carried out on a patient, such as a person on anticoagulant or antiplatelet drug therapy, and the activation level is compared to a normal activation level. The latter is generally around 5-6%, i.e., in a normal person only 5 to 6 percent of platelets will be activated at any given time. In the patient, who may be on anticoagulant drug therapy, the activation level may become as high as 25-50%; 5 to 10 times the normal level. This can be readily determined by, for example, carrying out the assay of the invention on several normal individuals and comparing the values obtained on the standard curve with that of the patient.

When the assay reveals that the activation level is excessively high, for example, over 25%, the dosage of the anticoagulant drug can be increased, a stronger anticoagulant can be used, or the like. Correspondingly, if the assay indicates that activation level is too low, for example, below 5%, then the dosage of the anticoagulant drug can be decreased, or a weaker anticoagulant can be used. Such a method can thereby prevent strokes and similar complications by detecting elevated platelet activation and permitting remedial drug therapy before clotting actually occurs. Percent activation can be correlated through clinical trials to the optical density determined by the assay, the relationship being roughly linear. The immunoassay used in the method for controlling platelet or other blood cell activation according to the invention need not be the primary and secondary antibody- based assay described above. Any suitable assay system could be used, for example, standard radioimmunoassay or immunofluorescence procedures, so long as the assay is conducted in the fluid phase so that the result of the assay accurately reflects the activation level of the cells in the patient's bloodstream. For example, the primary antibody could be directly labelled with an enzyme, fluorescent label, radioactive isotope or similar label, and the secondary antibody could be omitted.

Significant increase in P-selectin expression is found in subjects diagnosed with unstable angina. The platelet ELISA assay of the invention provides a sensitive, accurate, and economical means of detecting circulating activated platelets. Fixed platelet samples can be prepared and analyzed fresh or after storage. Either procedure is useful or detecting systemic platelet activation by whole platelet ELISA as well as by flow cytometry.

Experimental assays according to the invention suggest that simple physical manipulation of platelets, i.e., centrifugation of whole blood to PRP or PRP to washed platelet, as well as agonist-induced activation, also leads to selective platelet activation as shown in Figs. 6A, 6B and 7A-7C. Previously, George et al., Transfusion. 28(2), 123 (1988), reported that use of different platelet rotators resulted in differences in platelet activation in platelet concentrates. Use of a 2-rpm tumbler rotator resulted in no loss of platelets and no activation of a.IIbB₃ but a 16% increase in P- selectin expression. However, platelets kept in suspension with a 6-rpm elliptical rotator caused a loss of one-third of the platelets in the form of clumps, a 50% decrease in GPIb in remaining single platelets, and an increase in P-selectin expression. The forgoing data suggest that physical manipulation can lead to selective release of platelet vesicles.

Anticoagulants and inhibitors can also effect the expression of platelet receptors. The possibility that hemolysis and release of APP from red blood cells could induce P-selectin expression during physical manipulation can be minimized by the use of apyrase in either the whole blood samples prior to centrifugation or in PRP during washing procedures. The lack of aIIbB₃ activation as seen by failure to bind fibrinogen, Amrani, et. al., Blood, 72, 919 (1988), also suggests that P-selectin expression is independent of fibrinogen receptor activation. Washed platelets, prior to agonist-induced activation, demonstrate an essentially unactivated fibrinogen receptor, and together with the low levels of GP53 expression in the presence of 50-60% expression of P-selectin (Fig. 6) , suggest that P-selectin can be selectively expressed on the surface of platelets and that platelet vesicle release is a more highly regulated process than previously believed. The following examples illustrate the invention:

EXAMPLE 1 The following solutions were prepared. In each case, sufficient water was added to the listed ingredients to bring the volume of the solution to the level indicated:

ACD Anticoagulant:

Citric acid (MW=210.14) 0.80 gm

Sodium citrate (MW=357.16) 2.20 gm Dextrose (MW=180.16) 2.45 gm

Distilled H₂0 to 100 ml

Wash Buffer:

108 mM NaCl (MW=58.44) 3.16 gm

38 mM KC1 (MW=74.56) 1.42 gm 1.7 mM NaHC0₃ (MW=84.01) 0.071 gm

21.2 mM Na₃ citrate (MW=294.10) 3.12 gm

27.8 mM sucrose (MW=342.30) 4.76 gm

2.35 mM MgCl₂ - 6H₂0 (MW=203.31) 0.112 gm

Distilled H₂0 to 500 ml

Buffer pH was adjusted to 6.5 by titration of the solution with 1 N NaOH. PGE-L and apyrase were added as needed to inhibit activation. For 10 ml wash buffer, 21.2 μl PGE- solution, 3.5 μg/ml and 25 μl, 1.25 units/ml apyrase were used; for 20 ml wash buffer, the amounts were 42.3 μl PGE-^ and 50 μl apyrase.

Platelet Resuspension Buffer:

137 mM NaCl (MW=58.44) 2.00 gm

2 mM KC1 (MW=74.56) 0.05 gm

0.3 mM NaH₂P0₄»H₂0 (MW=137.99) 0.02 gm 2 mM CaCl₂ • 2H₂0 (MW-147.02) 0.07 gm

1 mM MgCl₂ • 6H₂0 (MW-203.31) 0.05 gm

5.5 mM dextrose (MW=180.16) 0.25 gm

5 mM HEPES (MW=238.30) 0.30 gm 12 mM NaHC0₃ (MW=84.01) 0.25 g

0.35% BSA (add just before use) 0.875 gm

Distilled H₂0 to 250 ml

HEPES is an amine commercially available from Sigma. Buffer pH was adjusted to 7.3-7.4 by titration of the solution with 1 N NaOH. PGE_]^ is added as needed as needed to inhibit activation, here in an amount that provided a final concentration of 5 ng/ml.

Pre-siliconized polypropylene Eppendorf tubes were filled with 0.5 ml of silicone oil. The silicone oil was a 5:1 mixture of Hiphenyl DC-550 to methyl DC-200 silicone oil. Fresh blood was collected into 15 ml Vacutainer tubes containing 2.14 ml ACD anticoagulant in a ratio of 9 parts by volume blood to 1 part ACD. Washed platelets were then prepared from the blood samples. Blood was drawn into the prepared Vacutainer tubes and centrifuged at 800 rpm (350 xG) for 15 minutes in an IEC-Centra-7 tabletop centrifuge. The platelet- rich plasma was carefully removed from the red cell layer and placed into a clean 13 ml polypropylene tube. 5.75 ml of platelet-rich plasma was obtained, to which 8.22 μl of PGEi inhibitor solution was added to yield a final concentration 5 ng prostaglandin E_j_ per ml of combined solution. PGE-L solution used in this example contained 3.545 μg of prostaglandin E-^ per ml of 100% ethanol.

The platelets were then pelleted by centrifugation at 700 xG for 5 minutes. The platelet-poor plasma supernatant may be saved as a source of fibrinogen for a platelet aggregation assay, or may be discarded. The platelet pellet usually contains red cells. The platelets were carefully resuspended in wash buffer containing inhibitors, as described above, at one half the original volume of the platelet-rich plasma. Platelets were removed only from around the red cells at the center, and the red cells were discarded.

The platelets were centrifuged at 1500 rpm (500 xG) for 5 minutes to pellet the platelets. The supernatant was poured off, and the platelets were resuspended in the wash buffer and inhibitors. The platelets were then centrifuged again at 500 xG for 5 minutes. The supernatant from the second wash was discarded, and the platelets were resuspended in the wash buffer containing inhibitors. A 0.5 ml sample of the washed platelet suspension was removed to a 4 ml polypropylene tube for counting. Platelet count is platelets/ml times total ml. The platelets were then centrifuged again at 500 xG for 5 minutes. The supernatant from the third wash was discarded, and the platelets were resuspended in platelet resuspension buffer containing no inhibitors.

During the final wash, the platelet count was adjusted to 175,000 platelets per microliter to provide a final concentration of 100,000 platelets/microliter.

Then, EDTA was added to the final suspension of platelets to provide a final concentration of 1 mM (micromole) . A 200 microliter sample of the washed platelets was pipetted into a number of reaction tubes. Collagen from BioData was reconstituted with distilled water to yield 1.9 mg collagen/ml. 63.2 microliters of the reconstituted collagen suspension were pipetted into each reaction tube to provide final collagen concentrations of 0, 25, 75, 100, 150, 200, 250, 300, 350, and 400 micrograms/ml. 36.8 microliters of S12 antibody solution were then gently pipetted into each tube in amounts to provide a final concentration of 2.0 micrograms of the antibody per ml. Prior to addition, the S12 antibody was centrifuged at the top speed of the microcentrifuge (10,000 xG) about for 15 minutes to remove microaggregates. The reaction tubes were allowed to incubate at room temperature for 30 minutes.

25 microliters of PGE-^ solution and 25 microliters of goat anti-mouse IgG heavy and light chains coupled to horseradish peroxidase (GAM IgG(H & L)-HRP, obtained commercially from BioRad) were pipetted into each reaction tube. The GAM IgG(H & L)-HRP contained 0.9 mg/ml IgG and 0.6 mg/ml peroxidase. The PGE-L solution contained PGE_! in an amount calculated to provide a final concentration of 10 ng/ml in the reaction tubes, and the GAM IgG(H & L)-HRP solution was used at a final dilution of 1:300. The GAM IgG(H & L)-HRP was centrifuged before use for 15 minutes to remove microaggregates in the same manner as for the S12 antibody. The reaction tubes were then incubated for 60 minutes as room temperature.

After incubation, 100 microliters from each tube were layered onto the previously prepared silicone oil aliquots in triplicate using wide-bore pipette tips. All tubes were then centrifuged at the top speed of the microcentrifuge (10,000 xG) for 6 minutes at room temperature. The tubes were then decanted. The peroxidase substrate, activated ABTS, was then added to each tube in an amount of 100 microliters per tube, and the tubes were allowed to incubate at 37°C in the dark for 45 minutes. ABTS development was then stopped by addition of 2 mM sodium azide solution (50 microliters per tube) . Optical densities were then read for each tube using a Microelisa reader for absorbance at 405 nm. The results are plotted on a semi-log scale in Figures 1 and 2. Both represent results for platelets from two normal individuals pooled together. Similar results were obtained using thrombin in concentrations varying from 0 to 1.0 units/ml, as shown in Figures 3 and 4. Assay conditions for Figs. 3 and 4 were 100,000 platelets/ml, 1 mM EDTA, 2 micrograms/ml S12, GAM IgG(H & L)-HRP diluted to 1:300, ABTS as substrate. Figure 3 represents a single normal donor, whereas Figure 4 represents platelets from four normal donors pooled together.

Figure 4 shows much less uncertainty and a more linear plot than Figure 3. The four graphs together show that greater convergence occurs as the number of individuals contributing to the platelet pool increases. Accordingly, the standards for use in the present invention are preferably prepared using platelets pooled from several (4 or more) individuals. It should also be possible according to the invention to generate a standard curve from a large number of individuals that can be used as a general standard, potentially eliminating the need to generate a new standard curve each time the assay is performed.

EXAMPLE 2 The following illustrates preparation of fixed platelets for use in a kit according to the invention. Platelets were prepared in accordance with Example 1, except as follows. After activation, the platelets were treated with 1% para-formaldehyde in phosphate-buffered saline (100 millimolar sodium phosphate, 150 millimolar NaCl, pH 7.4) for 5 minutes at room temperature. The platelets were spun down by centrifugation, washed 3 times, then counted. After the third spin, the platelets were resuspended in platelet resuspension buffer containing 5% dimethyl sulfoxide (DMSO) at a concentration of 10⁹ platelets per ml. The platelets were then stored for 1 day at -70°C. Samples were then thawed out using a water bath at 37°C.

The procedure of Example 1 was then followed as described above using the thawed platelets. The results were substantially the same as shown in Figs. 1-4, except that the background level, reflected as the baseline of the graph from the X-axis, was somewhat higher. This example indicates that platelets for use as standards in a kit according to the invention can be prepared in advance and stored frozen until the time of use.

EXAMPLE 3 The following reagents were used needed: bovine serum albumin (RIA grade, fraction V) (BSA) ; prostaglandin El (PGE i) ; potato apyrase (grade III) ; ethylene diamine tetraacetic acid (EDTA) ; adenosine-5'- diphosphate (grade X) from equine muscle (ADP) ; acetate (PMA)3 hirudin, dimethylsulfoxide (DMSO), and leupeptin were purchased from Sigma Chemical Co., St. Louis, MO. Tris(hydroxymethyl)aminomethane (Tris) was purchased from Aldrich Chemical Co. Inc., Milwaukee, WI. Calf skin collagen (soluble form) was purchased from BioData Corporation, Hatboro, PA. N-hydroxy ethylpiperazine N'-2- ethane sulfonic acid (HEPES) was purchased from Boehringer Mannheim Biochemicals, Indianapolis, IN. Trasylol was purchased from Mobay Chemical Corp. , New York, NY. Silicone oil, a 5:1 ratio of DC 550 Hi-Phenyl silicone oil (125 cs) : DC 200 methyl silicone oil (1.0 cs) , was obtained from William F. Nye, New Bedford, MA. Na¹²⁵I was purchased from Dupont-New England Nuclear, Boston, MA. Paraformaldehyde was obtained from Polysciences, Inc. , Warrington, PA.

Purified S-12, an antibody specific for P-selectin (6) was provided by Dr. Rodger McEver, Oklahoma Medical Research Foundation, Oklahoma City, OK. Other antibodies against receptor glycoproteins, p3, P-selectin, and GP53 corresponding to CD61-FITC, CD62-PE, and CD63-FITC, respectively, were purchased from Becton-Dickinson (Torrance, CA) . Thrombin receptor peptide was purchased from Bachem. Inc. , Torrance, CA. Goat anti-mouse (GAM) IgG alkaline phosphatase and para-nitrophyenlyphosphate (PNPP) were purchased from Zymed, Inc. Purified S-12 was radioactively labeled with ¹²⁵I using the iodine monochloride technique as previously described for fibrinogen. ¹²⁵I-labeled S-12 had a specific radioactivity of approximately 7.5 μCi/μg protein.

Venous blood was drawn into acid-citrate dextrose anticoagulant from normal donors who had given informed consent. Platelet-rich plasma (PRP) and washed platelets were subsequently prepared as previously described. Radiolabeled ligands were tested for binding to thrombin- or collagen-activated washed platelets as previously described for fibrinogen. Washed platelets at a final concentration of 2 x 108/ml were preincubated at room temperature with ¹²⁵IS12 for 1 hr. In certain experiments, non-radioactively labeled ligands were also added to reaction mixtures to test for non-specific binding. Bound radioactive counts were determined using a Packard Multi-Prias 4 gamma counter (Packard Instrument Co. , Downers Grove, IL) . Background values were obtained in the absence of activation and in the presence of hirudin (1 unit/ml) and PGE 1(5 ng/ml) .

Platelet rich plasma and washed platelets were obtained as indicated above for platelet binding experiments. Platelet aliquots (1.75 x 108/ml, final concentration) were activated with increasing concentrations of thrombin followed by the addition of hirudin (1 unit/ml, final concentration) . Most studies with PRP were carried out by activation with increasing concentrations of PMA; two studies were carried with the thrombin receptor peptide. PMA-activated platelets (1.8 x 108/ml, final concentration) were fixed with paraformaldehyde, washed in 50 mM sodium phosphate buffer, pH 7.4, containing saline (PBS) and resuspended at a final concentration of l X lθ⁹/ml in platelet storage buffer [platelet resuspension buffer (see below) containing 5% dimethyl-sulfoxide, 5 mM EDTA, 400 μM leupeptin] . Fixed platelets can be used immediately or stored at -70°C for up to 10 months.

A standard curve was prepared from PMA- or thrombin receptor peptide-activated, paraformaldehyde-fixed (1%) platelets, which had been diluted with platelet resuspension buffer (137 mM NaCl, 2mM KC1, 0.3 mM Ma phosphate, 2 mM CaC12, 1 mM MgCl2, 5.5 mM glucose, 5 mM Hepes, 12 mM NaHC03, 0.35% bovine serum albumin, pH 7.4) to a stock concentration of 1.75 X 108/ml. To minimize Fc receptor interaction with the primary antibody, EDTA was added to chelate calcium, which is necessary for Fc receptor-IgG interaction. In the assay, the primary antibody, S-12 or CD-62, was added to a further diluted suspension of fixed platelets (4 x 10⁷/ml) , and incubated at room temperature for 30 min. The platelets were centrifuged at 10,000 X g for 3 min, the supernatant solution decanted, and platelets then washed once with TBS-Tween 20. A secondary antibody-GAM IgG (H+L) Fab'2- alkaline phosphatase was added for an additional incubation period of 60 min, followed by centrifugation at 10,000 X g for 3 min. The substrate, PNPP (1 mg/ml), was then added and color developed over 60 min at 37°C. The reaction was stopped by addition of 1 N NaOH, the solution transferred to a microELISA plate, and the absorbance read at 405 nm.

All samples were analyzed using a FAC Star flow cytometer with Consort 30 software (B-D, Becton-Dickenson Immunocytometry Systems, Mansfield, MA) . Calibration procedures, which were carried out daily, used 'Fluorsbrite beads' and 'Calibrate' beads (Polysciences, Inc., Warrington, PA). Controls included normal unstained' platelets and singly stained, with each fluorochrome-labeled antibody, and respective isotype antibodies. Log amplification was selected for forward angle light scatter (FSC) , fluorescence 1 (FLl) , and fluorescence 2 (FL2) during analysis of 10,000 particles per sample. EDTA-anticoagulated whole blood was obtained by venipuncture or from an arterial line from patients with unstable angina or from individuals who demonstrated no apparent acute disease following catherization as well as from normal individuals who had given written informed consent. The first 2 ml of the blood draw was discarded, and blood then collected by syringe into an anticoagulant solution containing EDTA (2 mM) , Trasylol (100 units/ml) , hirudin (1 unit/ml) and PGE1 (5 ng/ml) , final concentrations, respectively. For flow cytometry, resting control platelets were prepared by direct fixation of whole blood with paraformaldehyde (final concentration, 1.8%) for 30 minutes at 4°C. Platelets for the positive control were activated with either ADP or soluble collagen at final concentrations of 20 kM and 190 micrograms/ml, respectively. Whole blood samples from patients or control subjects were incubated at room temperature for 15 minutes and then fixed with paraformaldehyde as above. Fixed samples were then washed two times with PBS containing 2% fetal bovine serum, incubated 20 minutes with CD-61-FITC (B₃ integrin) at approximately 3 microgram/ml and/or CD-62-PE at 0.4 micrograms/ml in the dark at room temperature. The samples were again washed and resuspended in PBS for flow cytometric analysis. The CD-61-FITC was used solely to gate on the platelet population and CD-62-PE to detect only those platelets that were activated.

For ELISA studies of patient and control subjects, whole blood and PRP was isolated and prepared as above. The PRP was fixed in paraformaldehyde as stated above and used directly in the described ELISA. For positive control samples, whole blood was collected in the absence of inhibitors, PRP prepared and treated with either ADP or soluble collagen at final concentrations of 20 micromoles and 190 micrograms/ml, respectively. The samples were fixed, washed, and included in the ELISA. The data were evaluated by Mann-Whitney non-parametric, two-tailed analysis of unpaired means using the program "Instat" (Graphpad, Torrance, CA) .

Collagen, which does not induce coagulation directly, could cause significant (>90%) P-selectin expression in washed platelets. The result is shown in Table 1, a comparison of P-selectin expression by washed platelets after thrombin or collagen addition.

Tatble 1

Agent Dose P-selectin (pσ/platelet * thrombin (u/ml) 0.01 53

0.02 121

0.1 119 collagen (μg/ml) 50 59

100 88

200 123

*P-selectin was detected using ¹²⁵I anti P-selectin (S- 12), n=3

In direct binding studies using ¹²⁵I-S-12 with washed platelets, collagen, compared to thrombin, produced maximal P-selectin expression, suggesting that a dose-dependent activation of washed human platelets with either thrombin or collagen could be developed as a standard for quantifying P-selectin expression. Using washed human platelets, we first established an assay based on thrombin (Fig. 5A) or collagen activation (Fig. 5B) , each of which produced similar standard curves. The assay depended on a microcentrifuge approach in which resuspended, fixed platelets, after activation with different concentrations of agonist, were incubated with anti-P-selectin for 30 min followed by a washing step, and subsequent incubation of resuspended platelets with enzyme-labeled GAM-FAb'₂. After five washing steps, the substrate is added to produce color. Interassay results at the respective thrombin and collagen concentrations showed a high degree of correlation with r, 0.98 +/- 0.02 (n = 8) . Flow cytometric analysis of thrombin- and collagen-activated washed platelets (Figs. 6A-6B) revealed, however, that P-selectin expression (Fig. 6A) as compared with the expression of activation markers, GP53 (Fig. 6B) and activated αIIbB₃ (data not shown) was significantly affected by the washing procedure itself. Buffer washing procedures, previously used to study fibrinogen binding, showed that fibrinogen was not bound to any significant extent under these conditions in the absence of an agonist such ADP. Nevertheless, platelets demonstrated 50 to 60% maximal P-selectin expression (Fig. 6A) when compared to autologous samples of whole blood and PRP, which demonstrated less than 0.5-0.75% and 0.6-0.8% P-selectin expression, respectively. In contrast, thrombin treatment of washed platelets resulted in 95 to 98% P-selectin expression (Fig. 6A) . These results suggested that the ELISA "zero" point samples were in fact expressing P-selectin at a high level. Compared to P-selectin expression, GP53 expression was increased only 8% in washed platelets and was maximally increased only after treatment with thrombin (Fig. 6B) . It is found that collagen, thrombin receptor peptide and PMA could nearly maximally activate P- selectin in PRP in which the baseline P-selectin level was <l-2% (Figs. 7A-7C) . Collagen induced expression of P-selectin to a maximum of 65-70% (Fig. 7A) , whereas thrombin receptor peptide and PMA at 25 micromole and 0.1 micromole, respectively, induced expression to >=95% (Figs. 7B and 7C) . In addition, PMA produced a more reproducible expression of P-selectin in PRP than either ADP (data not shown) or collagen (Fig. 7A) . Therefore, an ELISA based on a dose-dependent PMA-induced activation of platelets in PRP was established (Figs. 8A, 8B) .

PRP (4 X 10-platelets/ml) was activated with increasing doses of PMA and subsequently fixed and washed. Two representative standard curves are presented as Figs. 8A, 8B. A linear standard range of activation was obtained between 0 and 100 nM PMA. The interassay variability was similar to that obtained with thrombin and collagen activated washed platelets with r, 0.98 +/- 0.03 (n = 12) .

To assess the ability of the platelet ELISA to detect platelet activation in the systemic circulation, 24 patients diagnosed with unstable angina and 12 age- matched control subjects were examined. PRP samples were assessed by both flow cytometric and ELISA analysis within 3 hr of collection and after storage at -70°C in platelet storage buffer for 10 months. Standard curves used to generate the activation values were from platelets that had been prepared either within 3 hrs of usage or stored with the patient samples.

Comparison of flow and ELISA data for either control subjects or angina patient samples is summarized in Table 2: Table 2

Comparison of Flow Cytometric vs ELISA Platelet Activation Analysis of Control Subjects and Unstable Angina Patients

Control Subjects (n=12) Unstable Angina Patients(n=24)

5 (Mean ± SD)

Assayed Assayed Assayed Assayed

Within 3 Hr After 10 Months Within 3 Hr After 10 month

Flow 0.45±0.4 8.0±4 6.1±8 (p=0.02)* 32±20(p=0.001)*

10 ELISA

0.55±0.6 9.6±6 5.5±9 (p=0.05)* 35±28(p=0.001)*

* p value comparing control subject to unstable angina patient samples assayed within 3 h of collection or after 10 months in storage.

A representative flow cytometric pattern demonstrates the comparison of control subject platelets (Fig. 9C) versus platelets from an angina patient (Fig. 9D) . Resting control (Fig. 9A) and ADP-activated control (Fig. 9B) platelet samples are also shown. Similar samples were prepared during each flow cytometric run in order to control for the blood drawing process and to provide a positive control for maximal expression of platelet P-selectin. Control data from flow cytometric and ELISA analysis showed a mean of approximately 0.5% for control samples evaluated within the first 3 hrs (Table 2) . In contrast, control values after 10 months storage demonstrated 8.0% and 9.6% P-selectin expression by flow and ELISA analyses, respectively. Data obtained on patient samples evaluated by flow cytometric and ELISA analysis were quite comparable (32% +/- 20% vs 35% +/- 28%, respectively) and showed no statistical differences between the two methods indicating that both assay provide the same information. Although the absolute P- selectin values from fresh and stored samples were clearly different, either result demonstrates statistically significant increases in P-selectin expression in individuals with unstable angina compared to the control subjects (Table 2) . The foregoing example illustrates that whole platelet ELISA, like flow cytometric analyses, can sensitively detect activated platelets. With the ability of stable, pre-activated, fixed platelet standards to be prepared and stored frozen for at least 10 months, this assay method is amenable to development as a kit. Lastly, unlike flow cytometry, this method does not require any sophisticated instrumentation or a dedicated technician, and can be adapted to ELISA instrumentation currently used in most pathology laboratories or other clinical laboratory settings.

It will be understood that the foregoing description is of preferred exemplary embodiments of the invention, and that the invention is not limited to the specific forms shown. Modifications may be made in the design of the assay without departing from the scope of the invention as expressed in the appended claims.

Claims

CLAIMS :

1. A method for determining the activation level of blood cells that undergo activation by an immunoassay of the type wherein a primary antibody undergoes specific binding with the blood cells and a labelled secondary antibody undergoes specific binding with the primary antibody-blood cell complex, characterized by reacting a sample containing the blood cells in a liquid phase with an excess quantity of an activation-specific primary antibody prior to allowing the cells in the sample to bind to a solid surface.

2. The method of claim 1, wherein the blood cells are platelets.

3. The method of claim 1, wherein the step of reacting the blood cells with the primary antibody is conducted in aqueous suspension in a vessel having a relatively low binding affinity for the blood cells.

4. The method of claim 3, wherein the vessel has an inner surface made of polypropylene onto which the aqueous suspension is placed.

5. The method of claim 1, wherein the immunoassay is an enzyme-linked immunoabsorbent assay.

6. An assay method for determining platelet activation level, comprising:

(A) obtaining a sample of platelets from a living subject; (B) adding an excess quantity of a platelet activation-specific primary antibody to the platelets in an aqueous suspension and allowing the primary antibody to react with the platelets for a time sufficient to allow substantially complete specific binding between the antibody and activated platelets; (C) forming a complex between the platelets bound to the primary antibody and a secondary antibody bound to an enzyme, which secondary antibody reacts with and specifically binds to the primary antibody; (D) adding a substrate to the platelet complexes, which substrate reacts with the enzyme to produce a visible indicator;

(E) measuring the extent to which the visible indicator is present; and (F) correlating the results with a standard, to determine the degree of activation of the platelets prior to the assay.

7. The method of claim 6, further comprising a step, prior to step (B) , of adding a calcium inhibitor to the aqueous suspension in an amount effective to chelate calcium ions in the aqueous suspension.

8. The method of claim 6, wherein the separating steps each further comprise centrifuging the platelet suspension.

9. The method of claim 6, wherein step (C) further comprises: adding an excess quantity of a labelled secondary antibody to the suspension which reacts with and specifically binds to the primary antibody, wherein the label is one of a pair of specific binding substances; separating the platelets from unbound secondary antibody; adding an excess quantity of an enzyme to the platelets, which enzyme is labelled with the other of the pair of specific binding substances, so that the enzyme specifically binds to the labelled secondary antibody; and separating the platelets from unbound enzyme.

10. The method of claim 9, wherein the pair of specific binding substances are biotin and streptavidin.

11. The method of claim 6, wherein step (C) further comprises: adding an excess quantity of a secondary antibody to the suspension which reacts with and specifically binds to the primary antibody, wherein the secondary antibody has the enzyme bound thereto; and separating the platelets from unbound secondary antibody-enzyme.

12. The method of claim 6, wherein the enzyme is a peroxidase or alkaline phosphatase.

13. The method of claim 6, wherein the standard comprises a standard curve.

14. The method of claim 13, wherein step (F) further comprises generating the standard curve by repeating steps (B) through (D) using standards comprising platelets activated at different, predetermined levels.

15. A test kit for use in determining the activation level of a platelet sample by ELISA, comprising: standards containing fixed platelets activated at different, predetermined levels for generating a standard curve; a platelet activation-specific primary antibody that can react with specific activation binding sites of platelets; a secondary antibody that binds specifically to the primary antibody; an enzyme bound to the secondary antibody or bindable thereto by means of labels on each of the secondary antibody and enzyme; directions for the performance of a protocol for the determination of platelet activation level in the platelet sample by ELISA and correlation to a standard curve generated using the standards.

16. The test kit of claim 15, wherein the platelets of the standards are in a frozen form which can be effectively thawed prior to use.

17. A method for controlling the activation level of platelets or other blood cells in a patient's bloodstream, comprising: obtaining a sample of platelets or other blood cells which undergo activation; immunoassaying the sample with an antibody that selectively binds to activated platelets or other blood cells; correlating the results with a standard to determine the degree of activation of the platelets or cells prior to the assay; and administering a drug to the patient in an amount effective to maintain activation at a predetermined level.