Chemiluminescence Enzyme Immunoassay UsingBacterial Magnetic Particles

Tadashi Matsunaga,* Masashi Kawasaki, Xie Yu, Noriyuki Tsujimura, and Noriyuki Nakamura

Department of Biotechnology, Tokyo University of Agriculture and Technology, Koganei, Tokyo 184, Japan

A novel chemiluminescence enzyme immunoassay usingbacterial magnetic particles (BMPs) has been developedfor highly sensitive and rapid detection of immunoglobulinG. Antibody was immobilized onto BMPs using theheterobifunctional reagents sulfosuccinimidyl 6-[3′-(2-pyridyldithio)propionamido]hexanoate (sulfo-LC-SPDP)and sulfosuccinimidyl 4-(N-maleimidomethyl)cyclohex-ane-1-carboxylate (sulfo-SMCC). For the highly sensitiveimmunoassay method using these BMPs, a good relation-ship was obtained between the luminescence intensity andmouse IgG concentration in the range of 1-105 fg/mL.Furthermore, in order to reduce assay time and to simplifyoperations, a rapid chemiluminescence enzyme immu-noassay method has been developed. The rapid methodwas completed in 10 min. A linear relationship wasobtained between the luminescence and mouse IgGconcentration in the range of 10-1000 ng/mL.

Radioimmunoassay still plays a major role in medicine andrelated areas1 due to its sensitivity, but radioactive materials areunwelcome in posttreatment. Therefore, enzyme immunoassay(EIA) has become an important analytical method in clinicaldiagnostics2 because of its sensitivity, specificity, and generalapplicability. Improvement of EIA in terms of reducing assay timeand simplifying operations is one of the major trends in develop-ment of immunoassay technology.3

The use of magnetic particles in immunoassay enables separa-tion of the bound and free analyte by application of a magneticfield. For example, proteins can be attached covalently to solidsupports, such as magnetic particles, preventing the desorptionof antibody during assay conditions. Because the particles aredispersed evenly throughout the reaction mixture, they allow rapidreaction kinetics without the need for continuous mixing orshaking, provide for the precise addition of antibody, and facilitateease of use. The magnetic particles serve as both the solid supportand the means of separation in the system. Amine-terminatedmagnetic particles (∼1 µm diameter) developed by AdvancedMagnetics, Inc. (Cambridge, MA) are available commercially andhave been used for solid phase immunoassay.4-7 The determi-nation of chlorothalonil (2,4,5,6-tetrachloro-1,3-benzenedicarboni-

trile), a broad spectrum fungicide, was carried out using magneticparticles-based enzyme immunoassay with a detection limit of 0.07ng/mL.4

Magnetic bacteria have been isolated from fresh and marinesediments and are known to produce magnetic particles.8-11

Much research has been carried out regarding the mechanismof production of these magnetic particles and their function asnavigational compasses in vivo. The bacterial magnetic particles(BMPs) are small in size (50-100 nm) and disperse very wellbecause they are covered with a stable lipid membrane.12

Enzymes and antibodies have been immobilized on BMPs usingboth bifunctional reagents and glutaraldehyde and have beenfound to have higher activities than those immobilized ontoartificial magnetic particles.13 On the basis of these properties,BMPs have been applied to fluoroimmunoassay,14-16 mRNArecovery,17 and DNA carriers.18 In this study, to develop highlysensitive immunoassay using antibody-immobilized BMPs, weemployed alkaline phosphatase instead of fluorescein isothio-cyanate for a label to chemiluminescence EIA. Furthermore, inorder to reduce assay time and to simplify operations, wedeveloped a chemiluminescence EIA using antibody-immobilizedBMPs for the determination of mouse immunoglobulin G (IgG).

EXPERIMENTAL SECTIONMaterials. Sulfosuccinimidyl 6-[3′-(2-pyridyldithio)propiona-

mido]hexanoate (sulfo-LC-SPDP) and sulfosuccinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate (sulfo-SMCC) werepurchased from Pierce (Rockford, IL), and Lumi-phos 530, whichincludes lumigen PPD [4-methoxy-4-(3-phosphonophenyl)spiro-(1,2-dioxetane-3,2′-adamantane] disodium salt (3.3 × 10-4 M) asfluorophore, was obtained from Wako Pure Chemical Industries,Ltd. (Osaka, Japan). Bovine serum albumin (BSA) and mouseIgE were purchased from Seikagaku Co. (Tokyo, Japan). MouseIgG, goat anti-mouse IgG antibody, and alkaline phosphatase-

(1) Ekins, R.; Chu, F.; Micallef, J. J. Biolumin. Chemilumin. 1989, 4, 59-78.(2) Diamandis, E. P.; Christopoulos, T. K. Clin. Chem. 1991, 37, 625-636.(3) Kricka, L. J.; Phil, D.; Path, F. R. C. J. Clin. Immunoassay 1993, 16, 267-

271.(4) Lawruk, T. S.; Gueco, A. M.; Jourdan, S. W.; Scutellaro, A. M.; Fleeker, J.

R.; Herzog, D. P.; Rubio, F. M. J. Agric. Food Chem. 1995, 43, 1413-1419.(5) Luk, J. M. C.; Lindberg, A. A. J. Immunol. Methods 1991, 137, 1-8.(6) Schlaeppi, J.-M. A.; Kessler, A.; Fory, W. J. Agric. Food Chem. 1994, 42,

1914-1919.(7) Yeung, J. M.; Newsome, W. H. Bull. Environ. Contam. Toxicol. 1995, 54,

444-450.

(8) Blakemore, R. P. Science 1975, 190, 377-379.(9) Matsunaga, T.; Tadokoro, F.; Nakamura, N. IEEE Trans. Magn. 1990, 26,

1557-1559.(10) Matsunaga, T.; Sakaguchi, T.; Tadokoro, F. Appl. Microbiol. Biotechnol.

1991, 35, 651-655.(11) Sakaguchi, T.; Burgess, J. G.; Matsunaga, T. Nature 1993, 365, 47-49.(12) Balkwill, D. L.; Maratea, D.; Blakemore, R. P. J. Bacteriol. 1980, 141, 1399-

1408.(13) Matsunaga, T.; Kamiya, S. Appl. Microbiol. Biotechnol. 1987, 26, 328-332.(14) Nakamura, N.; Hashimoto, K.; Matsunaga, T. Anal. Chem. 1991, 63, 268-

272.(15) Nakamura, N.; Burgess, J. G.; Yagiuda, K.; Kudo, S.; Sakaguchi, T.;

Matsunaga, T. Anal. Chem. 1993, 65, 2036-2039.(16) Nakamura, N.; Matsunaga, T. Anal. Chim. Acta 1993, 281, 585-589.(17) Sode, K.; Kudo, S.; Sakaguchi, T.; Nakamura, N.; Matsunaga, T. Biotechnol.

Tech. 1993, 7, 688-694.(18) Takeyama, H.; Yamazawa, A.; Nakamura, C.; Matsunaga, T. Biotechnol. Tech.

1995, 9, 355-360.

Anal. Chem. 1996, 68, 3551-3554

S0003-2700(96)00369-1 CCC: $12.00 © 1996 American Chemical Society Analytical Chemistry, Vol. 68, No. 20, October 15, 1996 3551

conjugated anti-mouse IgG antibody (ALP-Ab) were purchasedfrom Cosmo Bio (Tokyo, Japan). Other reagents were of analyti-cal-reagent or laboratory grade. Deionized, distilled water wasused in all procedures.

Preparation of Bacterial Magnetic Particles (BMPs).BMPs were isolated from the magnetic bacterium Magnetospir-illum sp. AMB-1 by the following method. Wet cells (∼1.4 mg)suspended in 10 mL of water were disrupted by three passesthrough a French pressure cell at 1300 kg/cm2 (Ohtake WorksCo. Ltd., Tokyo, Japan). BMPs were collected from the disruptedcell fraction by using a neodymium-boron (Nd-B) magnet (10mm × 10 mm × 6 mm) that produced an inhomogeneousmagnetic field (0.37 T on the surface of the magnet). BMPs werecollected at the bottom of the tube due to the presence of themagnet, and the supernatant was removed. The collected BMPswere washed with 10 mM phosphate-buffered saline (PBS, pH7.4) using ultrasonic cleaner CA 4481 (Kaijo Denki Co. Ltd.,Tokyo, Japan) at least three times and kept at 4 °C in PBScontaining 0.01% sodium azide before use.

Immobilization of Antibody onto BMPs. For the im-mobilization of antibody onto BMPs, we modified the methodreported by Hashida et al.19 At first, sulfo-SMCC (0.06 mg) wasadded to 1 mL of anti-mouse IgG antibody solution (1 mg/mL)and incubated for 2 h at room temperature. After incubation, thesample was purified using a NAP-10 column (Pharmacia, Uppsala,Sweden), eluting with PBS according to the manufacturer’sinstructions. On the other hand, 1.2 mg of sulfo-LC-SPDP wasadded to 2 mL of BMPs suspension (1 mg/mL). The suspensionwas then dispersed by sonication and incubated for 2 h at roomtemperature. After the incubation, the sulfo-LC-SPDP-modifiedBMPs were separated magnetically from reaction mixture usinga Nd-B magnet and washed three times with 1.0 mL of PBS.The sulfo-LC-SPDP-modified BMPs were dispersed in 2 mL of 20mM dithiothreitol in phosphate buffer containing 100 mM NaCland incubated for 1 h at room temperature. After being washedthree times, the modified BMPs were incubated with the sulfo-SMCC-modified anti-mouse IgG solution for 12 h at 4 °C. Anti-mouse IgG antibody-immobilized BMPs (Ab-BMPs) were washedwith PBS three times to remove excess antibody.

Highly Sensitive Chemiluminescence EIA of Mouse IgGUsing Ab-BMPs and ALP-Ab. Mouse IgG solution (500 µL)was mixed with Ab-BMPs (50 µg) in the test tube and incubatedfor 1 h at room temperature. Antigen-antibody complex wascollected with a Nd-B magnet and washed three times with PBS.BMPs were dispersed by sonication in 200 µL of alkalinephosphatase-conjugated antibody solution (60 ng/mL) and incu-bated for 1 h at room temperature. After incubation, antigen-antibody complex was collected with a Nd-B magnet and washedfive times by PBS to remove excess alkaline phosphatase-conjugated antibody. Lumi-phos 530 (300 µL) was then added toantigen-antibody complex and dispersed by sonication. After 30min of incubation at 37 °C, the luminescence intensity wasmeasured using a BLR-301 luminescence reader (Aloka Co., Ltd.,Tokyo, Japan).

Rapid Chemiluminescence EIA of Mouse IgG. A rapid andsimple chemiluminescence EIA procedure was developed asfollows. Ab-BMPs (100 µg) and 10 µL of ALP-Ab (100 µg/mL)was added to 500 µL of mouse IgG solution. The mixture was

dispersed by sonication and incubated for 5 min at room temper-ature. After incubation, antigen-antibody complex was separatedmagnetically from the reaction mixture using a Nd-B magnetand washed five times with 0.5 mL of PBS. Lumi-phos 530 (500µL) was then added to the antigen-antibody complex anddispersed by sonication. Luminescence intensity was measuredat 37 °C using a BLR-301 luminescence reader, and the integratedvalues for 5 min were evaluated as luminescence.

RESULTS AND DISCUSSIONPreparation of BMPs and Immobilization of Anti-IgG

Antibody on BMPs. When BMPs were prepared by the Frenchpress and ultrasonication method, they were well dispersed inbuffer. The size of 70 wt % of BMPs ranged from 50 to 120 nm(mean size 100 nm), and the other 30 wt % was aggregated in thesize range of 250-700 nm. Modified antibody was immobilizedon BMPs activated with SPDP. The extent of antibody couplingwith BMPs was 54 µg/mg of particles.

Highly Sensitive Chemiluminescence EIA Using Ab-BMPand ALP-Ab. Optimum Assay Conditions. Figure 1 shows therelationship between the luminescence intensity and the amountof magnetic particles. The luminescence intensity increases withincreasing amount of Ab-BMPs in the range from 10 to 30 µg. Itwas suggested that, when more than 30 µg of antibody-conjugatedBMPs was employed, IgG was sufficiently bound to Ab-BMPs.But, when more than 70 µg of antibody-conjugated BMPs wasused, the luminescence intensity decreased. Optimum amountsof Ab-BMPs were 30-70 µg. It was suggested that, when theamount of BMPs was high, BMPs aggregated in the buffer andblocked luminescence themselves.

Optimization of buffer pH for assay was also carried out.Maximum luminescence intensity occurred at pH 7.5. Buffer pHaffected BMP dispersion and antibody reactivity. In this case,these results suggested that BMPs dispersion and antibodyreactivity were optimized in neutral pH.

(19) Hashida, S.; Imagawa, M.; Inoue, S.; Ruan, K.-H.; Ishikawa, E. J. Appl.Biochem. 1984, 6, 56-63.

Figure 1. Relationship between luminescence intensity and theamount of magnetic particles. Mouse IgG solution (500 µL: O, 0 pg/mL; 2, 1 pg/mL) was mixed with Ab-BMPs and incubated for 1 h atroom temperature. Antigen-antibody complex was collected with aNd-B magnet and washed three times with PBS. BMPs werecollected and dispersed by sonication in 200 µL of alkaline phos-phatase-conjugated antibody solution (60 ng/mL) and incubated for1 h at room temperature. Lumi-phos 530 (300 µL) was then addedto antigen-antibody complex and incubated for 30 min at 37 °C, andluminescence intensity was measured.

3552 Analytical Chemistry, Vol. 68, No. 20, October 15, 1996

Determination of Mouse IgG Concentration. As noted above,the conditions for sensitive chemiluminescence EIA were opti-mized, and then the measurement of IgG concentration wascarried out. A good relationship was obtained between theluminescence intensity and IgG concentration in the range of1-105 fg/mL (Figure 2). The maximum detectable concentrationof IgG was 105 fg/mL. This is because the immunological reactionof anti-IgG antibody was saturated. The minimum detectableconcentration of IgG was 1 fg/mL (6.7 zmol; 4000 molecules ascalculated from Avogadro’s number). For this reason, antibodywas specifically immobilized onto the BMPs, which were alsodispersed. We have reported the development of fluoroimmu-noassay using Ab-BMP.14 Mouse IgG concentration could bedetected in the range of 0.5-100 ng/mL. In this paper, theminimum detectable concentration of mouse IgG by sensitivechemiluminescence EIA was 5 × 108 times higher than that offluoroimmunoassay using Ab-BMP.14

Rapid Chemiluminescence EIA of Mouse IgG. Immuno-reaction of ALP-Ab, Mouse IgG, and Ab-BMPs. To reduce theantigen-antibody reaction time, the time course of antigen-antibody reaction of ALP-Ab, mouse IgG, and Ab-BMPs wasexamined. Then, 500 µL portions of mouse IgG standard samples(1000 ng/mL) were mixed with ALP-Ab and Ab-BMPs in the testtubes, and the tubes were incubated at room temperature. Afterwashing five times, the luminescence was measured. Figure 3shows the time course of the antigen-antibody reaction of ALP-Ab, mouse IgG, and Ab-BMPs. The luminescence indicates theamount of antigen-antibody complex formed. The increase inluminescence shows that the immunoreaction continues for morethan 20 min. The luminescence in the presence of 1000 ng/mLmouse IgG increases linearly when the immunoreaction time wasless than 5 min. Although extending the immunoreaction timemakes this assay sensitive, the immunoreaction time was fixedat 5 min in the following experiments.

Optimum Assay Conditions. The amount of Ab-BMP was foundto be a significant factor for optimal conditions. The amount ofAb-BMP influenced the antibody amount on the surface of theBMPs and blocked the luminescence during measurement.Therefore, the correlation between the amount of Ab-BMP andthe luminescence on determination of mouse IgG concentrationwas examined, and the results were shown in Figure 4. Whenthe Ab-BMP was 25-100 µg, the luminescence increased with

increasing amount of Ab-BMP in the presence of 1000 ng/mLmouse IgG. This was caused by increasing the amount ofantigen-antibody complex formed. On the other hand, theluminescence decreased with increasing amounts of Ab-BMP inthe range 100-250 µg. This was caused by BMPs blockingluminescence during measurement. The luminescence in theabsence of mouse IgG varied little with increasing amounts ofAb-BMP. This indicates that the most effective amount of Ab-BMP for this assay is 100 µg.

ALP-Ab concentration also affects the luminescence based onALP-Ab bound to IgG-Ab-BMP conjugates. The luminescenceincreased with increasing ALP-Ab concentration until 100 and 300µg/mL when 1000 and 10 ng/mL IgG was employed, respectively.Therefore, 100 µg/mL ALP-Ab was used for further experiments.

Determination of Mouse IgG Concentration. Mouse IgG con-centration was measured by rapid and simple chemiluminescenceEIA using Ab-BMP and ALP-Ab. Figure 5 shows the relationship

Figure 2. Correlation between luminescence intensity and mouseIgG concentration using ALP-Ab and Ab-BMP. The experiments wereperformed under various mouse IgG concentration and under thesame conditions, except for the amount of BMPs (50 µg).

Figure 3. Time course of luminescence based on antigen-antibodyreaction. Ab-BMPs (100 µg) and ALP-Ab (10 µL, 100 µg/mL) wereadded to 500 µL of mouse IgG solution: O, 0 ng/mL; b, 1000 ng/mL. The mixture was dispersed by sonication and incubated for 5min at room temperature. After incubation, antigen-antibody complexwas separated magnetically using a Nd-B magnet. Lumi-phos 530(500 µL) was then added to the antigen-antibody complex. Lumi-nescence intensity was measured at 37 °C using a luminescencereader. The integrated values in each time were evaluated asluminescence.

Figure 4. Correlation between amount of Ab-BMP and lumines-cence on determination of mouse IgG concentration: O, 0 ng/mL; b,1000 ng/mL. The experiments were performed under the sameconditions, except for the amount of BMPs. The luminescence wasexpressed as the integrated values of luminescence intensity for 5min.

Analytical Chemistry, Vol. 68, No. 20, October 15, 1996 3553

between the luminescence and mouse IgG concentration. Theluminescence increased with increasing mouse IgG concentration.A linear relationship was obtained between the luminescence andmouse IgG concentration in the range 10-1000 ng/mL. Theminimum detectable concentration of mouse IgG was 10 ng/mL.Luminescence was reproducible, with coefficients of variation of7.7% and 8.2%, when samples containing 1000 and 10 ng/mL IgGwere measured eight times, respectively. BMP-synthesized Mag-netospirillum sp. AMB-1 was covered with 98% lipids and 2% othercompounds, including proteins. Lipids consisted of 58% phos-pholipids and 42% other lipids. This lipid membrane makes BMPsnegatively charged. There was slight aggregation in each particleas a result of its own magnetic properties. BMPs were superiorin dispersion to artificial magnetite particles of the same size inaqueous solution.14

Table 1 shows the specificity of the determination of mouseIgG in this assay. The concentration of each protein was adjusted

to 1 µg/mL. Except for mouse IgG, the luminescence scarcelyincreased. The luminescence increased only in the presence ofmouse IgG. Luminescence was observed attributable to non-specific binding between BMP and alkaline phosphatase-conju-gated anti-mouse IgG in the experiments in the case of BSA andIgE. These results suggested that the increase of the lumines-cence was caused by an antigen-antibody-specific reaction.

This method demonstrates the advantages of BMPs for therapid and simple chemiluminescence EIA. The assay describedallows results to be obtained in about 10 min, whereas thefluoroimmunoassay using Ab-BMP14 allows results to be obtainedin about 15 min. Studies have shown the EIA using Ab-BMPs tobe accurate compared with the EIA using antibody-immobilizedartificial magnetite particles, and that the specific determinationof mouse IgG was possible in the presence of other proteins.Furthermore, this method may be generally applicable to thedetermination of any suitable antigen.

Received for review April 16, 1996. Accepted August 1,1996.X

AC9603690

X Abstract published in Advance ACS Abstracts, September 1, 1996.

Figure 5. Correlation between luminescence intensity and mouseIgG concentration using Ab-BMP and ALP-Ab. A immunoreaction wascarried out for 5 min. Concentration of Ab-BMPs was adjusted to 100µg/mL.

Table 1. Specificity of the Determination of Mouse IgG

protein relative luminescencea

none 0.11 ( 0.02mouse IgG 1.00 ( 0.04BSA 0.16 ( 0.03mouse IgE 0.18 ( 0.03IgG + BSA + IgE 1.06 ( 0.04

a Luminescence obtained from mouse IgG was regarded as 1.00.Concentrations of each protein and Ab-BMP were adjusted to 1000ng/mL and 100 µg/mL, respectively.

3554 Analytical Chemistry, Vol. 68, No. 20, October 15, 1996