Background of the Invention

Magnetic particles have been used as a tool for separation in biological systems for over twenty years (Robinson PJ et al. in Biotech. Bioengng 15:603-606,1973). Their use in immunoassays has been known for over ten years. The early work by Whitehead et al (U.S. Patents 4,554,088, 4,695,393 and others) described a technique for preparation of the particles which utilized silane amine as the coating material and immobilized biological materials such as antibody on the coated particles.

The current technique used for immobilization as recommended by the current supplier of silanized particles involves activating the magnetic particles with glutaraldehyde, using 2 milligrams of glutaraldehyde per milligram of particles, allowing the reaction to proceed for 3 hours. Following this, the glutaraldehyde activated particles are mixed with antibody for 16 to 24 hours. (See Advanced Magnetics Product Manual, 1993-1994.)

Summary of the Invention

An improved process for the preparation of antibody - coated magnetic particles which are used for biological separations has been invented. The new process involves the use of much lower concentration of glutaraldehyde (approximately ten times less) , and surprisingly requires much less time, than was used previously to activate the particles. The benefits of the novel process are: (1) It is a more efficient method of removing the excessive glutaraldehyde from the reaction system, and it eliminates the quenching step after antibody immobilization. (2) The activation process can be completed in a much shorter time than in the past, even though there is a significant reduction in glutaraldehdye concentration, which allows for the antibody immobilization step to be completed within one working day instead of two. (3) The resulting particles give 15 to 74% improvement of the sensitivity, which is observed in some immunochemical assays.

Brief Description of the Drawings

Figure 1 shows Glutaraldehyde Optimization, in terms of RLU bound to reflect the binding capacity of the finished particles as a function of glutaraldehyde concentration, which is expressed in terms of mg glutaraldehyde/mg PMP.

Figure 2 shows Kinetics of Mouse IgG Uptake by PMP, in terms of % mouse IgG uptake as a function of incubation time, in minutes.

Figure 3 shows Standard Curves for HCG generated on the ACS:180 ^® instrument, in terms of RLU bound as a function of HCG concentration, which is expressed in terms of milli-International Units/ml. Figure 4 shows Standard Curves for AFP generated on the ACS:180 ^® instrument, in terms of RLU bound as a function of AFP concentration, which is expressed in terms of ng AFP/ml. Figure 5 shows Standard Curves for PSA generated on the ACS:180 ^® instrument, in terms of RLU bound as a function of PSA.concentration, which is expressed in units of ng PSA/ml. Figure 6 shows Standard Curves for TSH generated on the ACS:180 ^® instrument, in terms of RLU bound as a function of TSH concentration, which is expressed in units of micro-International Units of TSH/ml .

Figure 7 shows Standard Curves of LH generated on the ACS:180 ^® instrument, in terms of RLU bound as a function of LH concentration, which is expressed in units of milli- International Units of LH/ml.

Detailed Description of the Invention

Magnetic particles have a number of uses in biological processes. Among these are the use of magnetic particles for the separation of biological components (i.e., a preparation function) . In addition, they are also used for analytical purposes. One of the analytical applications in which these particles are used is the immunochemistry area, where antibodies are attached to the magnetic particles. These antibodies are then reacted with an analyte of interest, and the application of a

magnetic field will permit the separation of the analyte which has become attached to the magnetic particle. In a variety of this assay, a second antibody to a different epitope of the the analyte may have attached to it a label (e.g., acridinium ester) which can be detected by some analytical means (e.g., detection of luminescense) . Many variations of these assays exist.

The immunochemical assays that are described herein can be conducted either manually or on an automated instrument, for example those instruments manufactured by Ciba Corning Diagnostics Corp. of Medfield, MA (e.g., the ACS:180 ^® analyzer, which is an automated chemiluminescent immunochemical analyzer) . Those assays run on the Ciba Corning instruments are generally based on the use of a chemiluminescent label, namely an acridinium ester-labeled antibody.

The magnetic particles which have been used in the biological process area have generally been those which have been coated with silane amine to enable the attachment of chemical entities to the surface. Whitehead et al (U.S. Patent No. 4,695,393) describes such a process for preparation of the silane amine-coated magnetic particle (also referred to as paramagnetic particles or PMP) . It should be noted that variations of the Whitehead technology are well-known by those with skill in the art. For example, particles made out of glass or polymeric materials, and which have magnetic cores, have also been utilized for the same purpose. See, for example, U.S. Patent No. 3,652,761 (Weetall) , 3,933,997 (Hersh et al) and literature from Bang Laboratories. In addition,

coating techniques aside from those of Whitehead have also been used. For example, in using the iron-core glass particles, it is not necessary to attach a silane amine coating because the siliceous nature of the glass allows direct attachment via chemical reaction with many of the desired activation materials.

It should be noted that a terminology has been developed by those working in this area in order to be able to distinguish between the various coating layers applied. The first layer is "coated" on the particle itself, and a coating material used herein was silane amine. The second layer is described as an "activation step", and an activating material used herein was glutaraldehyde. The third layer is "immobilized" on the particle, and one material immobilized herein was an antibody. This terminology is used herein.

Once the silane amine-coated or similar particles are obtained, a procedure for attaching the antibody or other molecule that will permit attachment to the target molecule is undertaken. Various procedures are utilized for this purpose. For example, a procedure recommended by a manufacturer of magnetic particles includes the reaction of the magnetic particles with a 5% glutaraldehyde solution for 3 hours, removal of excess glutaraldehyde, reaction with the protein to be attached for 16 - 24 hours, followed by removal of excess protein. (See literature from Advanced Magnetics Inc.) This technique is typical of the coating procedures currently utilized.

It has been unexpectedly found that the activating process can be significantly improved. By reducing the amount of glutaraldehyde in the activating process, it has also been found that the activation can be achieved in a much shorter time period. In addition, it has been found that the resulting particles yield enhanced sensitivity when used in some immunochemical assay procedures.

The concentration of glutaraldehyde used until the instant invention was in the range of approximately 1.25 mg to 2.0 mg of glutaraldehyde per milligram of magnetic particle.

Unexpectedly, it has been found that, by reducing the glutaraldehyde concentration to approximately 0.2 mg per milligram of magnetic particles, the improvements were realized. The concentration range where improvement was observed was actually from about 0.05 to about 0.5 milligram of glutaraldehyde per milligram of magnetic particle. The optimal amount of glutaraldehdye was determined by the binding capacity of the resulting particles in a "sandwich" assay, such as the human chorionic gonadotropin (HCG) assay. A large amount of analyte (approx. 1 million mlU/ml) beyond the highest standard (1184 mlU/ml) was introduced to test the binding capacity of the particles carrying anti-HCG antibody activity, which relects the effectiveness of the new process. The added glutaraldehyde was noted to be as effective as the reference when the concentration of glutaraldehyde ranged down to about 0.033 mg/mg PMP, at which point performance began to worsen, as reflected by the progressive loss of the binding activity on the finished particles. (See Fig. 1.) The optimum concentration was observed to be about 0.2 mg

glutaraldehyde per mg of magnetic particle.

The amount of reactive amine per unit of weight of the particle was never a factor until it was determined herein to have been overused in the past. Even at the low concentration of glutaraldehyde used herein (0.2mg/mg magnetic particle) , glutaraldehyde was still about 50 to 100 times excess over the amount of reactive amines on the magnetic particle.

It has also been found that that the glycine quenching required in the previous immobilization procedure is no longer needed because the excess glutaraldehdye in the novel method is much less than the previous method and, therefore, much more easily removed during the washing step. In order to achieve similar efficiency in the previous procedure, we would need at least three more washes than in this novel procedure in order to skip the glycine quenching step. Although this may seem like a small difference, in a setting where a machine is designed to conduct each step, this represents a tremendous saving in design time and saves a considerable time in the laboratory.

The reduction in the activation time of glutaraldehyde was found to have a significant effect on the time required to complete the antibody immobilization process, thus allowing the process to be completed within a normal working day. Coupling of glutaraldehyde to magnetic particles was found to be complete in approximately 1 hour. Reasonably good performance was found when the glutaraldehyde coupling took place for between 30 and 90

minutes, with optimum performance being found at 1 hour. In addition, uptake of the antibody molecule was found to be extremely fast (approximately 10 minutes) , while the subsequent bonding of the protein to the glutaraldehyde was found to be completed in approximately 6 hours. Good performance was found when the protein bonding was allowed to proceed for 4 - 8 hours, with optimum performance found when bonding proceeded for 6 hours. After the reaction with glutaraldehyde and protein, excess reagent is removed from the system and the magnetic particles washed. Thus, the overall process for immobilizing antibody on the magnetic particles was found to be reduced from about 19 -

27 hours which had been considered to be the optimum time in the literature (Weston et al . Biochem Biophys Res Commmun 45:1,1971; Rembaum et al. J. Immuno Methods 24:239,1978) to a time of about 7 hours, a reduction of about 60 - 75% in the time required to complete the antibody immobilization process. This significant reduction of antibody coupling time was attributed to the discovery of the immediate uptake of antibody by the PMP that facilitates the formation of covalent bonding between the glutaraldehyde and the antibody molecules as shown in Figure 2.

An added advantage of the novel immobilization process is the improved sensitivity of the resulting particles used in some immunochemical assays, as shown in Table 1.

.

Table 1

SENSITIVITY IN IMMUNOCHEMICAL ASSAYS USING FORMER VS, NOVEL TECHNIQUES FOR MANUFACTURING COATED MAGNETIC PARTICLES

Assay Current Novel

HCG 2.59 3.22

PRL 2.72 3.15

TSH 2.01 3.45

(Data is ratio of signal measured, in RLU (relative light units, with reference to a light emitting standard) , of the lowest concentration standard to the zero standard of each assay. Solid phase is the only variable in this study. )

It is apparent that numerous variations in the process set forth above may be made by those with ordinary skill without departing from the spirit and scope thereof. The examples shown below illustrate various aspects of the invention but are not intended to limit the usefulness thereof.

Example 1 - Preparation of Antibody Magnetic Particles used in the ACS™ HCG (Human choriqnic gonadotropin) assay

10 mg of PMP was dispensed into a 12x85 mm polystyrene test tube. The PMP was washed twice, using 0.5 ml of 10 mM sodium acetate buffer, pH 5.5, each time, and the PMP were

magnetically separated before the supernatant was removed. The PMP were resuspended in 80 μl of 10 mM sodium acetate buffer, pH 5.5. A 1:12 dilution of a 25% glutaraldehdye solution from a commerical source with 10 mM sodium acetate buffer, pH 5.5, was made, and 80 ul of the diluted glutaraldehyde was transferred to the PMP in the test tube. The test tube was covered with two layers of parafilm and placed on a rotary shaker and mixed at 200 rpm for one hour at room temperature. At the end of one hour, PMP were magnetically separated and supernatant was removed. The PMP was washed five times with 0.5 ml 10 mM sodium acetate buffer, pH 5.5, each time. An antibody solution of purified monoclonal mouse anti-human chorionic gonadotrophin was prepared at 3.8 mg per ml concentration and 0.5 ml of the solution was added to the glutaraldehyde-activated PMP particles. PMP were mixed continuously for six hours at room temperature the same way as before. At the end of sixth hour, PMP were magnetically separated and supernatant was removed. The PMP were then serially washed in 0.5 ml of each of the following wash solutions: 10 mM sodium acetate buffer, pH 5.5, 1 M NaCl, and 10 mM sodium phosphate buffer, pH 7.4. After the last wash with sodium phaosphate buffer, PMP were resuspended in 2 ml of a "heat stress buffer" (usually 1.7% of BSA in 50 mM sodium phosphate buffer, pH 6.5) and transferred to a 15-ml polystyrene centrifugal tube with a screw-on cap. The PMP was heat stressed at 50 C for 14-16 hours. The PMP was magnetically separated and the supernatant was removed. The PMP was washed twice with 2ml each time of the "TSH buffer " ( 0.1% BSA, 0.001% BGG in 10 mM Phosphate Buffered Saline pH 7.4 ) . The final PMP were resuspended and stored in 1 ml of solid

phase buffer, which contains BSA, BGG, normal mouse serum, normal sheep serum and antibiotics . Both test lot and manufacturer (reference) lot solid phases were diluted to the same working concentration in the solid phase buffer. Both solid phase reagents were run for binding performance on an ACS:180 ^® analyzer. The result of standard curves of ACS™ HCG produced by the test and reference solid phases is shown in Fig.3. There is no significant difference between the two standard curves. To further qualify the equivalency between these two solid phases, a "hook" sample with HCG value at one million mlU/ml was included in the run and again no significant difference (less than 5% difference in total counts) was found between the two. Therefore, we concluded that solid phase made by the new protocol gave equivalent performance as the reference solid phase did. This procedure can be easily scaled up proportionally. We did an half gram batch and found it was no difference from a 10 mg batch in terms of performance.

Example 2 - Preparation of Particles for ACS™ AFP (Alpha feto-protein) assay

Solid phase was prepared by the same protocol as in ACS™ HCG except the antibody offered was monoclonal mouse anti- human alpha feto-protein, and the antibody loading was 2 mg per 10 mg of PMP .The binding result between test and reference is shown in Figure 4.

Example 3 - Preparation of Particles for ACS™ PSA (prostate specific antigen) assay

Solid phase was prepared by the same protocol except the antibody offered was monoclonal mouse anti-human prostate- specific antigen and the antibody loading was 1.2 mg per 10 mg of PMP. The binding result of the test lot and the reference lot is shown in Figure 5.

Example 4 - Preparation of Particles for ACS™ TSH (thyroid stimulating hormone) assay

Solid phase was prepared by the same protocol except the antibody offered was affinity purified sheep anti-human thyroid stimulating hormone and the antibody loading was 1.6 mg per 10 mg of PMP. The result is shown in Figure 6.

Example 5 - Preparation of Particles for ACS™ LH (Luteinizing hormone) assay

Solid phase was prepared by the same, protocol except the antibody offered was monoclonal mouse anti-human luteinizing hormone and the antibody loading was 2.0 mg per 10 mg of PMP. The result is shown in Figure 7.

In all cases solid phase antibody made by the new protocol is either equal to or better than the solid phase made by the corresponding reference protocol in terms of binding characteristics. The new protocol is far better than the reference protocol in terms of simplicity and time saved.