- Email: [email protected]
Detection of HbA1c by boronate affinity immunoassay using bacterial magnetic particles
Biosensors & Bioelectronics 16 (2001) 1089– 1094 www.elsevier.com/locate/bios
Detection of HbA1c by boronate affinity immunoassay using bacterial magnetic particles Tsuyoshi Tanaka, Tadashi Matsunaga * Department of Biotechnology, Tokyo Uni6ersity of Agriculture and Technology, Koganei, Tokyo 184 -8588, Japan Received 7 June 2000; received in revised form 20 March 2001; accepted 22 March 2001
Abstract We have developed a boronate affinity immunoassay system using m-aminophenylboronic acid (mAPB) coupling to bacterial magnetic particles (BMPs). Homobifunctional crosslinker, Bis-(succcimidyl)suberate (BS3), was employed for preparation of mAPB-BMPs conjugates (mAPB-BMPs). Quantities of HbA1c on mAPB-BMPs were evaluated based on luminescence from alkaline phosphatase-conjugated anti-Hb antibody (ALP– antibody) binding to HbA1c on the BMP surface. The binding of HbA1c to mAPB-BMPs occurred gradually and was almost completed within 10 mm. The coupling reaction is enhanced due to static electric interaction between the positive charges on HbA1c and negative charges on BMPs. The amount of HbA1c binding to mAPB-BMPs increased with increasing sodium chloride concentrations in the range of 0 – 100 mM. However, the amount of Hb binding to mAPB-BMPs also increased in high concentration of sodium chloride. The Hb binding to mAPB-BMPs was detached from mAPB-BMPs when Hb–mAPB-BMPs were washed with low salt buffer. This indicates that Hb is nonspecifically adsorbed onto the surface of mAPB-BMPs in high concentration of sodium chloride. These results suggest that selective separation of HbA1c using mAPB-BMPs can be achieved with these conditions. A dose– response curve was obtained between luminescence intensity and HbA1c concentration using a fully automated boronate affinity immunoassay. A linear relationship between luminescence intensity and HbA1c concentration was obtained in the range of 10 – 104 ng/ml. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Bacterial magnetic particles (BMPs); HbA1c; Fully automated system; m-Aminophenylboronic acid (mAPB); Boronic affinity immunoassay
1. Introduction Hemoglobin A1c (HbA1c) has widely been used as a marker for monitoring long-term blood glucose concentration in diabetic patients. Various automated assay systems for measuring HbA1c have been developed. The assays are based on immunoassay (Holownia et al., 1997; Palfrey and Labib, 1994; Weets et al., 1996) using antibody-conjugated beads, boronate affinity chromatography (Cefalu et al., 1994; Hill, 1990) and cationexchange chromatography using chemically modified beads (Chen et al., 1998; Nuttall, 1998). The small size of these beads offers multiple advantages due to faster reaction when applied to small-scale purification pro* Corresponding author. Tel: +81-42-388-7020; fax: + 81-42-3857713. E-mail address: [email protected] (T. Matsunaga).
cesses. Assays using magnetic particles are convenient since the magnetic separation steps are employed. Magnetic bacteria contain magnetic particles from 50 to 100 nm in diameter (Blakemore, 1975; Matsunaga et al., 1991, 1990; Sakaguchi et al., 1993). These bacterial magnetic particles (BMPs) are composed of magnetite (Fe3O4) with a single magnetic domain having a phospholipid membrane covering their surface (Balkwill et al., 1980; Matsunaga and Kamiya, 1987). The amount of antibody coupled to BMPs was about 4-fold higher compared with artificial magnetite particles of the same size because BMPs have superior dispersion characteristics in aqueous solution (Nakamura et al., 1991). Antibody-conjugated BMPs have been applied toward highly sensitive immunoassay for detection of food allergens and specific IgG molecules (Matsunaga et al., 1996; Nakamura and Matsunaga, 1993). The BMP lipid membrane is more negatively charged in low
0956-5663/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 5 6 - 5 6 6 3 ( 0 1 ) 0 0 1 8 7 - 7
1090
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
ionic-strength and low pH aqueous solutions, resulting in efficient particle dispersion (Tanaka and Matsunaga, 2000). On the basis of these properties, BMPs are useful for efficient and selective separation of cationic compounds such as hemoglobin and glycated blood lysates. In this study we have developed boronate affinity immunoassay system using m-aminophenylboronic acid coupled with BMPs and alkaline phosphatase-conjugated anti-hemoglobin antibody. The assay procedures have been adapted to a fully automated immunoassay system that we had already constructed (Tanaka and Matsunaga, 2000).
2. Materials and methods
2.1. Materials Mouse anti-human Hb monoclonal antibody was purchased from Nippon Bio Test Inc. (Tokyo, Japan). Mouse anti-human HbA1c monoclonal antibody and purified Hemoglobin A1c (HbA1c) was purchased from Exocell, Inc. (PA, USA). Hemoglobin (Hb) was purchased from Biogenesis Inc. (NH, USA). Hemolysates were prepared from fresh human whole blood, anticoagulated with EDTA. Plasma was removed by centrifugation (2000×g, 10 min), and an equal volume of water was added to the pellet and incubated for 10 min. The mixture was again centrifuged (2000× g, 10 min) to remove lipids and lipid-soluble material. Bis-(succcimidyl)suberate (B53) was obtained from Pierce Chemical Co. (Rockford, IL, USA). m-Aminophenylboronic acid (APB) was obtained from Sigma Chemical Co. (St. Louis, MO, USA). Other reagents were commercially available analytical reagents or laboratory grade materials. Deionized-distilled water was used in all procedures.
2.2. Preparation of BMPs BMPs were isolated from Magnetospirillum magneticum AMB-1. Approximately 1.4 mg wet cells suspended in 20 ml PBS (10 mM phosphate buffered saline, pH 7.4) were disrupted by three passes through a French pressure cell at 1300 kg/cm2 (Ohtake Works Co. Ltd., Tokyo, Japan). BMPs were collected from the disrupted cell fraction using a columnar Neodymium– boron (Nd – B) magnet (diameter: 22.5 mm, height: 12.5 mm) producing an inhomogeneous magnetic field (0.5 T at the surface). BMPs were pelleted at the bottom of the tube using the magnet and the supernatant removed. The collected BMPs were washed with PBS using an ultrasonic bath at least three times and stored at 4 °C in PBS containing 0.01% (w/v) sodium azide until required.
2.3. Immobilization of m-aminophenylboronic acid onto BMPs Immobilization of m-aminophenylboronic acid onto BMPs was performed using the homobifunctional reagent BS3, which contained N-hydroxysuccinimide reacted with primary amines. BMPs (1 mg) were suspended in 1 ml of B53 (1 mM). The suspension was dispersed by sonication and incubated for 0.5 h at room temperature with pulsed sonication (1-min pulses at 5-min intervals). Modified BMPs were separated magnetically from the reaction mixture using a Nd– B magnet and washed three times with 1.0 ml PBS. The modified BMPs were dispersed in 1 ml of mAPB (1 mM) and incubated for 0.5 h at room temperature with pulsed sonication (1-min pulses at 5-min intervals). Excess mAPB was removed from mAPB-BMP conjugates (mAPB-BMPs) by three washes with PBS.
2.4. Specific detection of HbA1c using mAPB-BMPs Fig. 1 shows the schematic diagram of the boronate affinity immunoassay using mAPB-BMPs and alkaline phosphatase-conjugated Hb antibody (ALP–antibody). mAPB-BMPs (500 mg) were suspended with 1 mg/ml HbA1c (100 ml) and incubated for 20 min at room temperature. As a binding buffer, 20 mM N-[2-hydroxyethyl]piperazine-N%-[3-propanesulfonic acid] (EPPS) buffer containing 10 mM MgCl2 with varying sodium chloride concentration was used. After three washes with EPPS buffer containing various sodium chloride concentration (100 ml), HbA1c –mAPB-BMPs were mixed with alkaline phosphatase-conjugated anti-Hb antibody (ALP –antibody). ALP-antibodies were prepared according to the previous report (Tanaka and Matsunaga, 2000). The HbA1c –mAPB-BMP complexes were finally suspended in 100 ml of Lumi-phos 530 and the luminescence intensity was measured by using a luminometer (Lucy-2, Aloka, Tokyo, Japan). The condensation of HbA1c in hemolysate to mAPB-BMPs was also investigated by isoelectric gel electrophoresis. mAPB-BMPs (500 mg/ml) were suspended in 1 ml of hemolysate (1 mg/ml) and incubated for 20 min. BMPs were magnetically separated with a Nd–B magnet and resuspended in 100 ml of EPPS buffer containing 200 mM sorbitol. The supernatant and elutant were subjected to native polyacrylamide gel electrophoresis with PhastGel™ Gradient 5–8.5 using PhastGel native buffer strips. Electrophoresis, focusing and silver staining of the gel were performed using Pharmacia phastsystem™ according to the manufacturer’s instructions.
2.5. Fully automated boronate affinity immunoassay system Boronate affinity immunoassay of HbA1c was performed on a fully automated immunoassay system
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
1091
Fig. 1. Schematic diagram of the boronate affinity immunoassay using mAPB-BMPs and alkaline phosphatase-conjugated anti-Ho antibody (ALP– antibody).
(Tanaka and Matsunaga, 2000). Automated immunoassay system contains an immunoreaction station (96 well microtiter plate), luminescence detection station (8× 3 well), disposable tip rack, an automated eight pipettor that is able to attach and detach a strong magnet to a tip surface. This system permits the simultaneous incubation and separation of eight samples in an assay. Each reagent is applied to eight rows of the microtiter plate as follows: well 1, HbA1c solution; well 2, mAPB-BMPs suspension; well 3, EPPS containing 0.1% BSA and 0.05% Tween 20 (washing buffer); well 4, alkaline phosphatase-conjugated anti-Hb antibody (ALP –antibody); wells 5– 7, EPPS containing 0.1% BSA and 0.05% Tween 20 (washing buffer); well 8, luminescence substrate (Lumi-phos 530). HbA1c solution (50 ml; well 1) was transferred to 50 ml of mAPBBMPs suspension (well 2) using an automated eight pipettor. The mixture was dispersed by the pipettor and incubated for 20 min. The HbA1c – mAPB-BMPs were separated magnetically using a Nd– B magnet on the inner surface of the pipetting tip during the aspirating and dispensing. The pellet was transferred to well 3 and subject to automated resuspension (20 cycles of pipette action) in 100 ml washing buffer. After magnetic separation, the HbA1c –mAPB-BMPs were transferred into well 4 to immunoreact with ALP– antibody. After immunoreaction (20 min), the HbA1c – APB-BMP and ALP –antibody complexes were separated magnetically. The BMPs were washed three times by repeated pipetting in washing buffer (well 5, 6 and 7). After magnetic separation, the complexes were finally suspended in 100 ml of Lumi-phos 530 (well 8). The suspension was transferred to a well in luminescence detection station and luminescence intensity was measured.
3. Results and discussion
3.1. Binding condition of HbA1c onto mAPB-BMPs Fig. 2 shows the time-course of HbA1c binding onto mAPB-BMP surface. EPPS containing 10 mM sodium chloride was used as a binding and washing buffers. HbA1c immediately bound to the mAPB-BMP surface. The reaction was almost completed within 10 min. A similar result was obtained from the previous report (Nustad et al., 1984), where tosyl activated particles resulted in two step binding of antibodies due to hydrophobic interactions. In this experiment, these results
Fig. 2. Time-course of HbA1c binding to mAPB-BMP surface. mAPB-BMPs (500 mg) were suspended in HbA1c solution (1 mg/ml) and incubated 20 min at room temperature. The HbA1c –mAPBBMPs were washed with EPPS buffer containing 10 mM sodium chloride. The HbA1c – mAPB-BMPs were dispersed in ALP –antibody solution (10 mg/ml) and incubated for 20 min at room temperature. Luminescence intensity was measured 1 min after the luminescence reaction.
1092
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
Fig. 3. Luminescence intensity of HbA1c –mAPB-BMP and ALP– antibody complexes when the binding of HbA1c to mAPB-BMPs was performed with EPPS buffers containing various sodium chloride concentrations. The binding of HbA1c to mAPB-BMPs were performed using 500 mg/ml mAPB-BMPs and 1 mg/ml HbA1c. The immunoassay process using ALP –antibody was performed under the same conditions as described above.
suggest that rapid completion of the coupling reaction occurs as a result of static electric interaction between positive charges on HbA1c and negative charges on BMPs. Fig. 3 shows the luminescence intensity of HbA1c – mAPB-BMP and ALP –antibody complexes when the binding of HbA1c to mAPB-BMPs was performed with EPPS buffers containing various sodium chloride concentrations. Quantities of HbA1c on mAPB-BMPs were evaluated based on luminescence from ALP– antibody binding to HbA1c on the BMP surface. The amount of HbA1c binding to mAPB-BMPs increased with increasing sodium chloride concentrations in EPPS buffer. Excess HbA1c is nonspecifically adsorbed onto the surface of mAPB-BMPs in high concentration of sodium chloride due to static electric interaction. Therefore, detachment conditions of HbA1c from mAPB-BMPs were investigated using EPPS buffers containing various concentrations of sodium chloride as a washing buffer. Fig. 4 shows the luminescence intensity when Hb–mAPB-BMPs or HbA1c – mAPB-BMPs were washed using EPPS buffers with and without sodium chloride. Non-treated BMPs were used as controls. The luminescence intensity of HbA1c – mAPBBMPs and ALP –antibody complexes decreased slightly when washing HbA1c-APB-BMPs without sodium chloride. However, the luminescence intensity of Hb– mAPB-BMP and ALP – antibody complexes extremely decreased when mAPB-BMPs were reacted with Hb and washed using EPPS without sodium chloride. A similar result was obtained in control experiment. This suggests that Hb is nonspecifically adsorbed onto the surface of mAPB-BMPs and detached from the surface by washing with low ion-strength buffer. These results suggest that binding of HbA1c onto mAPB-BMPs was via boronic acid and cis-diol. Selective separation of
Fig. 4. Luminescence intensity of when Hb – mAPB-BMPs or HbA1c – mAPB-BMPs were washed using EPPS buffers with and without sodium chloride. The washing steps were performed in EPPS buffer containing various concentrations of sodium chloride. The immunoassay process using ALP –antibody was performed under the same conditions as described above.
HbA1c using mAPB-BMPs was achieved with these conditions.
3.2. Condensation of HbA1c in hemolysate onto mAPB-BMPs Fig. 5 shows the correlation between amount of mAPB-BMP and the luminescence intensity for determination of HbA1c. The intensity at 0 mg/ml mAPBBMPs indicates luminescence on Lumi-phos 530. The luminescence in the presence of 10 mg/ml HbA1c increased with increasing mAPB-BMPs concentration and reached a plateau above 500 mg/ml. This result suggests that 500 mg of mAPB-BMPs are capable of condensing 10 mg/ml HbA1c most efficiently. Isoelectric gel electrophoresis was performed to confirm the condensation of HbA1c from hemolysates on mAPB-BMPs. Supernatant in mAPB-BMPs suspensions after reaction with 100 ml hemolysate (1 mg/ml)
Fig. 5. Correlation between amount of mAPB-BMP and the luminescence intensity for determination of HbA1c EPPS buffer with and without sodium chloride was used as a binding and a washing buffer, respectively. The experiments were performed under various amount of mAPB-BMPs and under the same immunoassay conditions as described above, except for HbA1c concentration (10 mg/ml).
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
1093
In conclusion, we developed a detection procedure for HbA1c using mAPB-BMPs and ALP –antibody. This method may be applied toward the detection of minor glycated Hb as well as HbA1c. This work also suggests that BMPs have the potential to control attachment and detachment of charged compounds with selectivity and efficiently. Furthermore, determination of HbA1c concentrations with a fully automated boronate affinity immunoassay using mAPB-BMPs was successfully done. Fig. 6. Relationship between HbA1c concentration and luminescence intensity using mAPB-BMPs (A) and antibody-conjugated BMPs (B). EPPS buffer with and without sodium chloride were used as a binding and a washing buffer, respectively. The experiments were performed under various HbA1c concentration and under the same immunoassay conditions as described above.
Acknowledgements
and elutant after washing with 10 ml EPPS buffer containing 200 mM sorbitol were used as applied samples. At least 10 mg/ml of proteins was required to detect bands by silver-staining. The protein band, originating from HbA1c (pI value: 7.0– 7.5), was observed only in elutant samples (data not shown). These results suggest that the Hb fractions are condensed by an order of 10 due to binding of HbA1c onto the surface of mAPB-BMPs.
References
3.3. Determination of HbA1c using a fully automated boronate immunoassay On the basis of the above results, a dose– response curve was obtained between luminescence intensity and HbA1c concentration using a fully automated boronate affinity immunoassay. Fig. 6A shows a relationship between HbA1c concentration and luminescence intensity using ALP–antibody and mAPB-BMPs. A logarithmic relationship was obtained in the range of 10–104 ng/ml. This range was similar to that obtained from chemiluminescence enzyme immunoassay (CLEIA) using anti-HbA1 antibody-conjugated BMPs and ALP-conjugated anti-Hb antibody (Fig. 6B). However, the luminescence intensity obtained from CLEIA was lower as compared with the boronate affinity immunoassay because the specific binding of HbA1c via antigen– antibody interactions decreased by washing with low salt buffer. Therefore, the boronate affinity immunoassay is better suited for detection of glycated Hb including HbA1c, compared with CLEIA. The effects of hemolysate from patients with hyperglycemia on assays was disregarded in this study. HbA1c should be expressed as a percentage of the total hemoglobin measured by photometry. Therefore, the specificity of this method using various hemolysates should be further analyzed.
We would like to thank Mr Yoza Brandon for suggestions to this manuscript.
Balkwill, D.L., Maratea, D., Blakemore, R.P., 1980. Ultrastructure of a magnetotactic spirillum. J. Bacteriol. 141, 1399 – 1408. Blakemore, R.P., 1975. Magnetotactic bacteria. Science 190, 377 – 379. Cefalu, W.T., Wang, Z.Q., Bell-Farrow, A., Kiger, F.D., Izlar, C., 1994. Glycohemoglobin measured by automated affinity HPLC correlates with both short-term and long-term antecedent glycemia. Clin. Chem. 40, 1317 – 1321. Chen, D., Crimmins, D.L., Hsu, F.F., Lindberg, F.P., Scott, M.G., 1998. Hemoglobin Raleigh as the cause of a falsely increased hemoglobin A1C in an automated ion-exchange HPLC method. Clin. Chem. 44, 1296 – 1301. Hill, R.P., 1990. Semi-automated colorimetric method for measuring glycohemoglobin, with reduction of nitroblue tetrazolium, evaluated. Clin. Chem. 36, 2131 – 2133. Holownia, P., Bishop, E., Newman, D.J., John, W.G., Price, C.P., 1997. Adaptation of latex-enhanced assay for percent glycohemoglobin to a Dade Dimension analyzer. Clin. Chem 43, 76 – 84. Matsunaga, T., Kamiya, S., 1987. Use of magnetic particles isolated from magnetotactic bacteria for enzyme immobilization. Appl. Microbiol. Biotechnol. 26, 328 – 332. Matsunaga, T., Tadokoro, F., Nakamura, N., 1990. Mass culture of magnetic bacteria and their application to flow type immunoassays. IEEE Trans. Magnet. 26, 1557 – 1559. Matsunaga, T., Sakaguchi, T., Tadokoro, F., 1991. Magnetite formation by a magnetic bacterium capable of growing aerobically. Appl. Microbiol. Biotechnol. 35, 651 – 655. Matsunaga, T., Kawasaki, M., Yu, X., Tsujimura, N., Nakamura, N., 1996. Chemiluminescence enzyme immunoassay using bacterial magnetic particles. Anal. Chem. 68, 3551 – 3554. Nakamura, N., Matsunaga, T., 1993. Highly sensitive detection of allergen using bacterial magnetic particles. Anal. Chim. Acta 281, 585 – 589. Nakamura, N., Hashimoto, K., Matsunaga, T., 1991. Immunoassay method for the determination of immunoglobulin G using bacterial magnetic particles. Anal. Chem. 63, 268 – 272. Nustad, K., Johansen, L., Ugelstad, J., Ellingsen, T., Berge, A., 1984. Hydrophilic monodisperse particles as solid-phase material in immunoassays: comparison of shell and-core particles with compact particles. Eur. Surg. Res. 16, 80 – 87. Nuttall, F.Q., 1998. Comparison of percent total GHb with percent HbA1c in people with and without known diabetes. Diabetes Care 21, 1475 – 1480.
1094
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
Palfrey, S., Labib, M., 1994. A rapid automated method for measuring glycated haemoglobin on the Beckman CX7 analyser. Ann. Clin. Biochem. 31, 293 –295. Sakaguchi, T., Burgess, J.G., Matsunaga, T., 1993. Magnetite formation by a sulphate-reducing bacterium. Nature (London) 365, 47– 49.
Tanaka, T., Matsunaga, T., 2000. Fully automated chemiluminescence immunoassay of insulin using antibody-protein A-bacterial magnetic particle complexes, Anal. Chem., 72, 3518 – 22. Weets, I., Gorus, F.K., Gerlo, E., 1996. Evaluation of an immunoturbidimetric assay for haemoglobin A1c on a Cobas Mira S analyser. Eur. J. Clin. Chem. Clin. Biochem. 34, 449 – 453.
Detection of HbA1c by boronate affinity immunoassay using bacterial magnetic particles Tsuyoshi Tanaka, Tadashi Matsunaga * Department of Biotechnology, Tokyo Uni6ersity of Agriculture and Technology, Koganei, Tokyo 184 -8588, Japan Received 7 June 2000; received in revised form 20 March 2001; accepted 22 March 2001
Abstract We have developed a boronate affinity immunoassay system using m-aminophenylboronic acid (mAPB) coupling to bacterial magnetic particles (BMPs). Homobifunctional crosslinker, Bis-(succcimidyl)suberate (BS3), was employed for preparation of mAPB-BMPs conjugates (mAPB-BMPs). Quantities of HbA1c on mAPB-BMPs were evaluated based on luminescence from alkaline phosphatase-conjugated anti-Hb antibody (ALP– antibody) binding to HbA1c on the BMP surface. The binding of HbA1c to mAPB-BMPs occurred gradually and was almost completed within 10 mm. The coupling reaction is enhanced due to static electric interaction between the positive charges on HbA1c and negative charges on BMPs. The amount of HbA1c binding to mAPB-BMPs increased with increasing sodium chloride concentrations in the range of 0 – 100 mM. However, the amount of Hb binding to mAPB-BMPs also increased in high concentration of sodium chloride. The Hb binding to mAPB-BMPs was detached from mAPB-BMPs when Hb–mAPB-BMPs were washed with low salt buffer. This indicates that Hb is nonspecifically adsorbed onto the surface of mAPB-BMPs in high concentration of sodium chloride. These results suggest that selective separation of HbA1c using mAPB-BMPs can be achieved with these conditions. A dose– response curve was obtained between luminescence intensity and HbA1c concentration using a fully automated boronate affinity immunoassay. A linear relationship between luminescence intensity and HbA1c concentration was obtained in the range of 10 – 104 ng/ml. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Bacterial magnetic particles (BMPs); HbA1c; Fully automated system; m-Aminophenylboronic acid (mAPB); Boronic affinity immunoassay
1. Introduction Hemoglobin A1c (HbA1c) has widely been used as a marker for monitoring long-term blood glucose concentration in diabetic patients. Various automated assay systems for measuring HbA1c have been developed. The assays are based on immunoassay (Holownia et al., 1997; Palfrey and Labib, 1994; Weets et al., 1996) using antibody-conjugated beads, boronate affinity chromatography (Cefalu et al., 1994; Hill, 1990) and cationexchange chromatography using chemically modified beads (Chen et al., 1998; Nuttall, 1998). The small size of these beads offers multiple advantages due to faster reaction when applied to small-scale purification pro* Corresponding author. Tel: +81-42-388-7020; fax: + 81-42-3857713. E-mail address: [email protected] (T. Matsunaga).
cesses. Assays using magnetic particles are convenient since the magnetic separation steps are employed. Magnetic bacteria contain magnetic particles from 50 to 100 nm in diameter (Blakemore, 1975; Matsunaga et al., 1991, 1990; Sakaguchi et al., 1993). These bacterial magnetic particles (BMPs) are composed of magnetite (Fe3O4) with a single magnetic domain having a phospholipid membrane covering their surface (Balkwill et al., 1980; Matsunaga and Kamiya, 1987). The amount of antibody coupled to BMPs was about 4-fold higher compared with artificial magnetite particles of the same size because BMPs have superior dispersion characteristics in aqueous solution (Nakamura et al., 1991). Antibody-conjugated BMPs have been applied toward highly sensitive immunoassay for detection of food allergens and specific IgG molecules (Matsunaga et al., 1996; Nakamura and Matsunaga, 1993). The BMP lipid membrane is more negatively charged in low
0956-5663/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 5 6 - 5 6 6 3 ( 0 1 ) 0 0 1 8 7 - 7
1090
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
ionic-strength and low pH aqueous solutions, resulting in efficient particle dispersion (Tanaka and Matsunaga, 2000). On the basis of these properties, BMPs are useful for efficient and selective separation of cationic compounds such as hemoglobin and glycated blood lysates. In this study we have developed boronate affinity immunoassay system using m-aminophenylboronic acid coupled with BMPs and alkaline phosphatase-conjugated anti-hemoglobin antibody. The assay procedures have been adapted to a fully automated immunoassay system that we had already constructed (Tanaka and Matsunaga, 2000).
2. Materials and methods
2.1. Materials Mouse anti-human Hb monoclonal antibody was purchased from Nippon Bio Test Inc. (Tokyo, Japan). Mouse anti-human HbA1c monoclonal antibody and purified Hemoglobin A1c (HbA1c) was purchased from Exocell, Inc. (PA, USA). Hemoglobin (Hb) was purchased from Biogenesis Inc. (NH, USA). Hemolysates were prepared from fresh human whole blood, anticoagulated with EDTA. Plasma was removed by centrifugation (2000×g, 10 min), and an equal volume of water was added to the pellet and incubated for 10 min. The mixture was again centrifuged (2000× g, 10 min) to remove lipids and lipid-soluble material. Bis-(succcimidyl)suberate (B53) was obtained from Pierce Chemical Co. (Rockford, IL, USA). m-Aminophenylboronic acid (APB) was obtained from Sigma Chemical Co. (St. Louis, MO, USA). Other reagents were commercially available analytical reagents or laboratory grade materials. Deionized-distilled water was used in all procedures.
2.2. Preparation of BMPs BMPs were isolated from Magnetospirillum magneticum AMB-1. Approximately 1.4 mg wet cells suspended in 20 ml PBS (10 mM phosphate buffered saline, pH 7.4) were disrupted by three passes through a French pressure cell at 1300 kg/cm2 (Ohtake Works Co. Ltd., Tokyo, Japan). BMPs were collected from the disrupted cell fraction using a columnar Neodymium– boron (Nd – B) magnet (diameter: 22.5 mm, height: 12.5 mm) producing an inhomogeneous magnetic field (0.5 T at the surface). BMPs were pelleted at the bottom of the tube using the magnet and the supernatant removed. The collected BMPs were washed with PBS using an ultrasonic bath at least three times and stored at 4 °C in PBS containing 0.01% (w/v) sodium azide until required.
2.3. Immobilization of m-aminophenylboronic acid onto BMPs Immobilization of m-aminophenylboronic acid onto BMPs was performed using the homobifunctional reagent BS3, which contained N-hydroxysuccinimide reacted with primary amines. BMPs (1 mg) were suspended in 1 ml of B53 (1 mM). The suspension was dispersed by sonication and incubated for 0.5 h at room temperature with pulsed sonication (1-min pulses at 5-min intervals). Modified BMPs were separated magnetically from the reaction mixture using a Nd– B magnet and washed three times with 1.0 ml PBS. The modified BMPs were dispersed in 1 ml of mAPB (1 mM) and incubated for 0.5 h at room temperature with pulsed sonication (1-min pulses at 5-min intervals). Excess mAPB was removed from mAPB-BMP conjugates (mAPB-BMPs) by three washes with PBS.
2.4. Specific detection of HbA1c using mAPB-BMPs Fig. 1 shows the schematic diagram of the boronate affinity immunoassay using mAPB-BMPs and alkaline phosphatase-conjugated Hb antibody (ALP–antibody). mAPB-BMPs (500 mg) were suspended with 1 mg/ml HbA1c (100 ml) and incubated for 20 min at room temperature. As a binding buffer, 20 mM N-[2-hydroxyethyl]piperazine-N%-[3-propanesulfonic acid] (EPPS) buffer containing 10 mM MgCl2 with varying sodium chloride concentration was used. After three washes with EPPS buffer containing various sodium chloride concentration (100 ml), HbA1c –mAPB-BMPs were mixed with alkaline phosphatase-conjugated anti-Hb antibody (ALP –antibody). ALP-antibodies were prepared according to the previous report (Tanaka and Matsunaga, 2000). The HbA1c –mAPB-BMP complexes were finally suspended in 100 ml of Lumi-phos 530 and the luminescence intensity was measured by using a luminometer (Lucy-2, Aloka, Tokyo, Japan). The condensation of HbA1c in hemolysate to mAPB-BMPs was also investigated by isoelectric gel electrophoresis. mAPB-BMPs (500 mg/ml) were suspended in 1 ml of hemolysate (1 mg/ml) and incubated for 20 min. BMPs were magnetically separated with a Nd–B magnet and resuspended in 100 ml of EPPS buffer containing 200 mM sorbitol. The supernatant and elutant were subjected to native polyacrylamide gel electrophoresis with PhastGel™ Gradient 5–8.5 using PhastGel native buffer strips. Electrophoresis, focusing and silver staining of the gel were performed using Pharmacia phastsystem™ according to the manufacturer’s instructions.
2.5. Fully automated boronate affinity immunoassay system Boronate affinity immunoassay of HbA1c was performed on a fully automated immunoassay system
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
1091
Fig. 1. Schematic diagram of the boronate affinity immunoassay using mAPB-BMPs and alkaline phosphatase-conjugated anti-Ho antibody (ALP– antibody).
(Tanaka and Matsunaga, 2000). Automated immunoassay system contains an immunoreaction station (96 well microtiter plate), luminescence detection station (8× 3 well), disposable tip rack, an automated eight pipettor that is able to attach and detach a strong magnet to a tip surface. This system permits the simultaneous incubation and separation of eight samples in an assay. Each reagent is applied to eight rows of the microtiter plate as follows: well 1, HbA1c solution; well 2, mAPB-BMPs suspension; well 3, EPPS containing 0.1% BSA and 0.05% Tween 20 (washing buffer); well 4, alkaline phosphatase-conjugated anti-Hb antibody (ALP –antibody); wells 5– 7, EPPS containing 0.1% BSA and 0.05% Tween 20 (washing buffer); well 8, luminescence substrate (Lumi-phos 530). HbA1c solution (50 ml; well 1) was transferred to 50 ml of mAPBBMPs suspension (well 2) using an automated eight pipettor. The mixture was dispersed by the pipettor and incubated for 20 min. The HbA1c – mAPB-BMPs were separated magnetically using a Nd– B magnet on the inner surface of the pipetting tip during the aspirating and dispensing. The pellet was transferred to well 3 and subject to automated resuspension (20 cycles of pipette action) in 100 ml washing buffer. After magnetic separation, the HbA1c –mAPB-BMPs were transferred into well 4 to immunoreact with ALP– antibody. After immunoreaction (20 min), the HbA1c – APB-BMP and ALP –antibody complexes were separated magnetically. The BMPs were washed three times by repeated pipetting in washing buffer (well 5, 6 and 7). After magnetic separation, the complexes were finally suspended in 100 ml of Lumi-phos 530 (well 8). The suspension was transferred to a well in luminescence detection station and luminescence intensity was measured.
3. Results and discussion
3.1. Binding condition of HbA1c onto mAPB-BMPs Fig. 2 shows the time-course of HbA1c binding onto mAPB-BMP surface. EPPS containing 10 mM sodium chloride was used as a binding and washing buffers. HbA1c immediately bound to the mAPB-BMP surface. The reaction was almost completed within 10 min. A similar result was obtained from the previous report (Nustad et al., 1984), where tosyl activated particles resulted in two step binding of antibodies due to hydrophobic interactions. In this experiment, these results
Fig. 2. Time-course of HbA1c binding to mAPB-BMP surface. mAPB-BMPs (500 mg) were suspended in HbA1c solution (1 mg/ml) and incubated 20 min at room temperature. The HbA1c –mAPBBMPs were washed with EPPS buffer containing 10 mM sodium chloride. The HbA1c – mAPB-BMPs were dispersed in ALP –antibody solution (10 mg/ml) and incubated for 20 min at room temperature. Luminescence intensity was measured 1 min after the luminescence reaction.
1092
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
Fig. 3. Luminescence intensity of HbA1c –mAPB-BMP and ALP– antibody complexes when the binding of HbA1c to mAPB-BMPs was performed with EPPS buffers containing various sodium chloride concentrations. The binding of HbA1c to mAPB-BMPs were performed using 500 mg/ml mAPB-BMPs and 1 mg/ml HbA1c. The immunoassay process using ALP –antibody was performed under the same conditions as described above.
suggest that rapid completion of the coupling reaction occurs as a result of static electric interaction between positive charges on HbA1c and negative charges on BMPs. Fig. 3 shows the luminescence intensity of HbA1c – mAPB-BMP and ALP –antibody complexes when the binding of HbA1c to mAPB-BMPs was performed with EPPS buffers containing various sodium chloride concentrations. Quantities of HbA1c on mAPB-BMPs were evaluated based on luminescence from ALP– antibody binding to HbA1c on the BMP surface. The amount of HbA1c binding to mAPB-BMPs increased with increasing sodium chloride concentrations in EPPS buffer. Excess HbA1c is nonspecifically adsorbed onto the surface of mAPB-BMPs in high concentration of sodium chloride due to static electric interaction. Therefore, detachment conditions of HbA1c from mAPB-BMPs were investigated using EPPS buffers containing various concentrations of sodium chloride as a washing buffer. Fig. 4 shows the luminescence intensity when Hb–mAPB-BMPs or HbA1c – mAPB-BMPs were washed using EPPS buffers with and without sodium chloride. Non-treated BMPs were used as controls. The luminescence intensity of HbA1c – mAPBBMPs and ALP –antibody complexes decreased slightly when washing HbA1c-APB-BMPs without sodium chloride. However, the luminescence intensity of Hb– mAPB-BMP and ALP – antibody complexes extremely decreased when mAPB-BMPs were reacted with Hb and washed using EPPS without sodium chloride. A similar result was obtained in control experiment. This suggests that Hb is nonspecifically adsorbed onto the surface of mAPB-BMPs and detached from the surface by washing with low ion-strength buffer. These results suggest that binding of HbA1c onto mAPB-BMPs was via boronic acid and cis-diol. Selective separation of
Fig. 4. Luminescence intensity of when Hb – mAPB-BMPs or HbA1c – mAPB-BMPs were washed using EPPS buffers with and without sodium chloride. The washing steps were performed in EPPS buffer containing various concentrations of sodium chloride. The immunoassay process using ALP –antibody was performed under the same conditions as described above.
HbA1c using mAPB-BMPs was achieved with these conditions.
3.2. Condensation of HbA1c in hemolysate onto mAPB-BMPs Fig. 5 shows the correlation between amount of mAPB-BMP and the luminescence intensity for determination of HbA1c. The intensity at 0 mg/ml mAPBBMPs indicates luminescence on Lumi-phos 530. The luminescence in the presence of 10 mg/ml HbA1c increased with increasing mAPB-BMPs concentration and reached a plateau above 500 mg/ml. This result suggests that 500 mg of mAPB-BMPs are capable of condensing 10 mg/ml HbA1c most efficiently. Isoelectric gel electrophoresis was performed to confirm the condensation of HbA1c from hemolysates on mAPB-BMPs. Supernatant in mAPB-BMPs suspensions after reaction with 100 ml hemolysate (1 mg/ml)
Fig. 5. Correlation between amount of mAPB-BMP and the luminescence intensity for determination of HbA1c EPPS buffer with and without sodium chloride was used as a binding and a washing buffer, respectively. The experiments were performed under various amount of mAPB-BMPs and under the same immunoassay conditions as described above, except for HbA1c concentration (10 mg/ml).
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
1093
In conclusion, we developed a detection procedure for HbA1c using mAPB-BMPs and ALP –antibody. This method may be applied toward the detection of minor glycated Hb as well as HbA1c. This work also suggests that BMPs have the potential to control attachment and detachment of charged compounds with selectivity and efficiently. Furthermore, determination of HbA1c concentrations with a fully automated boronate affinity immunoassay using mAPB-BMPs was successfully done. Fig. 6. Relationship between HbA1c concentration and luminescence intensity using mAPB-BMPs (A) and antibody-conjugated BMPs (B). EPPS buffer with and without sodium chloride were used as a binding and a washing buffer, respectively. The experiments were performed under various HbA1c concentration and under the same immunoassay conditions as described above.
Acknowledgements
and elutant after washing with 10 ml EPPS buffer containing 200 mM sorbitol were used as applied samples. At least 10 mg/ml of proteins was required to detect bands by silver-staining. The protein band, originating from HbA1c (pI value: 7.0– 7.5), was observed only in elutant samples (data not shown). These results suggest that the Hb fractions are condensed by an order of 10 due to binding of HbA1c onto the surface of mAPB-BMPs.
References
3.3. Determination of HbA1c using a fully automated boronate immunoassay On the basis of the above results, a dose– response curve was obtained between luminescence intensity and HbA1c concentration using a fully automated boronate affinity immunoassay. Fig. 6A shows a relationship between HbA1c concentration and luminescence intensity using ALP–antibody and mAPB-BMPs. A logarithmic relationship was obtained in the range of 10–104 ng/ml. This range was similar to that obtained from chemiluminescence enzyme immunoassay (CLEIA) using anti-HbA1 antibody-conjugated BMPs and ALP-conjugated anti-Hb antibody (Fig. 6B). However, the luminescence intensity obtained from CLEIA was lower as compared with the boronate affinity immunoassay because the specific binding of HbA1c via antigen– antibody interactions decreased by washing with low salt buffer. Therefore, the boronate affinity immunoassay is better suited for detection of glycated Hb including HbA1c, compared with CLEIA. The effects of hemolysate from patients with hyperglycemia on assays was disregarded in this study. HbA1c should be expressed as a percentage of the total hemoglobin measured by photometry. Therefore, the specificity of this method using various hemolysates should be further analyzed.
We would like to thank Mr Yoza Brandon for suggestions to this manuscript.
Balkwill, D.L., Maratea, D., Blakemore, R.P., 1980. Ultrastructure of a magnetotactic spirillum. J. Bacteriol. 141, 1399 – 1408. Blakemore, R.P., 1975. Magnetotactic bacteria. Science 190, 377 – 379. Cefalu, W.T., Wang, Z.Q., Bell-Farrow, A., Kiger, F.D., Izlar, C., 1994. Glycohemoglobin measured by automated affinity HPLC correlates with both short-term and long-term antecedent glycemia. Clin. Chem. 40, 1317 – 1321. Chen, D., Crimmins, D.L., Hsu, F.F., Lindberg, F.P., Scott, M.G., 1998. Hemoglobin Raleigh as the cause of a falsely increased hemoglobin A1C in an automated ion-exchange HPLC method. Clin. Chem. 44, 1296 – 1301. Hill, R.P., 1990. Semi-automated colorimetric method for measuring glycohemoglobin, with reduction of nitroblue tetrazolium, evaluated. Clin. Chem. 36, 2131 – 2133. Holownia, P., Bishop, E., Newman, D.J., John, W.G., Price, C.P., 1997. Adaptation of latex-enhanced assay for percent glycohemoglobin to a Dade Dimension analyzer. Clin. Chem 43, 76 – 84. Matsunaga, T., Kamiya, S., 1987. Use of magnetic particles isolated from magnetotactic bacteria for enzyme immobilization. Appl. Microbiol. Biotechnol. 26, 328 – 332. Matsunaga, T., Tadokoro, F., Nakamura, N., 1990. Mass culture of magnetic bacteria and their application to flow type immunoassays. IEEE Trans. Magnet. 26, 1557 – 1559. Matsunaga, T., Sakaguchi, T., Tadokoro, F., 1991. Magnetite formation by a magnetic bacterium capable of growing aerobically. Appl. Microbiol. Biotechnol. 35, 651 – 655. Matsunaga, T., Kawasaki, M., Yu, X., Tsujimura, N., Nakamura, N., 1996. Chemiluminescence enzyme immunoassay using bacterial magnetic particles. Anal. Chem. 68, 3551 – 3554. Nakamura, N., Matsunaga, T., 1993. Highly sensitive detection of allergen using bacterial magnetic particles. Anal. Chim. Acta 281, 585 – 589. Nakamura, N., Hashimoto, K., Matsunaga, T., 1991. Immunoassay method for the determination of immunoglobulin G using bacterial magnetic particles. Anal. Chem. 63, 268 – 272. Nustad, K., Johansen, L., Ugelstad, J., Ellingsen, T., Berge, A., 1984. Hydrophilic monodisperse particles as solid-phase material in immunoassays: comparison of shell and-core particles with compact particles. Eur. Surg. Res. 16, 80 – 87. Nuttall, F.Q., 1998. Comparison of percent total GHb with percent HbA1c in people with and without known diabetes. Diabetes Care 21, 1475 – 1480.
1094
T. Tanaka, T. Matsunaga / Biosensors & Bioelectronics 16 (2001) 1089–1094
Palfrey, S., Labib, M., 1994. A rapid automated method for measuring glycated haemoglobin on the Beckman CX7 analyser. Ann. Clin. Biochem. 31, 293 –295. Sakaguchi, T., Burgess, J.G., Matsunaga, T., 1993. Magnetite formation by a sulphate-reducing bacterium. Nature (London) 365, 47– 49.
Tanaka, T., Matsunaga, T., 2000. Fully automated chemiluminescence immunoassay of insulin using antibody-protein A-bacterial magnetic particle complexes, Anal. Chem., 72, 3518 – 22. Weets, I., Gorus, F.K., Gerlo, E., 1996. Evaluation of an immunoturbidimetric assay for haemoglobin A1c on a Cobas Mira S analyser. Eur. J. Clin. Chem. Clin. Biochem. 34, 449 – 453.