Conjugation of Biomolecules with Magnetic Protein Microspheres for

The magnetic HSA/γ-Fe2O3 microspheres were characterized by scanning electron ... (4-6) Frequently the results are inconclusive since it is often dif...
1 downloads 0 Views 2MB Size
Download PDF
Anal. Chem. 2009, 81, 6210–6217

Conjugation of Biomolecules with Magnetic Protein Microspheres for the Assay of Early Biomarkers Associated with Acute Myocardial Infarction Jinyi Wang,*,†,‡ Xueqin Wang,† Li Ren,† Qiang Wang,† Li Li,† Wenming Liu,† Zongfang Wan,† Linyan Yang,‡ Peng Sun,† Lili Ren,‡ Manlin Li,† Heng Wu,† Jinfeng Wang,† and Lei Zhang† College of Veterinary Medicine and College of Science, and Shaanxi Key Laboratory of Molecular Biology for Agriculture, Northwest A&F University, Yangling, Shaanxi 712100, P.R. China This study demonstrates an improved magnetic protein microsphere-aided sandwich fluoroimmunoassay for the analysis of myoglobin and heart-type fatty acid binding protein (H-FABP), early protein markers associated with acute myocardial infarction. In preparation for the assay we constructed superparamagnetic human serum albumin (HSA)/γ-Fe2O3 microspheres, and grafted capture antibodies (monoclonal antimyoglobin 7C3 and antiH-FABP 10E1) onto the protein microspheres using the avidin-biotin system. Then the antibody-carrying microspheres were used in a sequential sandwich fluoroimmunoassay along with detection antibodies (Alexa fluor594-labeled antimyoglobin 4E2 and FITClabeled anti-H-FABP 9F3). The magnetic HSA/γ-Fe2O3 microspheres were characterized by scanning electron microscopy, atomic force microscopy, Fourier transform infrared spectrophotometry, atomic absorption spectrophotometry, and vibrating sample magnetometry. Fluorescence images of the post-immunoassay microspheres recorded using an inverted fluorescence microscope showed that the average fluorescence intensity was correlated with the concentration of cardiac markers, in agreement with the results obtained by an F-4500 FL spectrophotometer; this indicated that the fluoroimmunoassay could be used to semiquantitatively detect both myoglobin and H-FABP. The detection limit was 10 ng/mL for myoglobin and 1 ng/mL for H-FABP. Acute myocardial infarction (AMI), a severe cardiovascular disease, remains one of the leading causes of death in both developing and developed nations.1 According to the latest data issued by the American Heart Association, 7.2 million cases of myocardial infarction were reported in 2003 in the USA, including 865,000 new and recurrent cases and 170,961 deaths, with * To whom correspondence should be addressed. E-mail: [email protected]. Phone: + 86-29-870 825 20. Fax: + 86-29-870 825 20. † College of Veterinary Medicine and College of Science. ‡ Shaanxi Key Laboratory of Molecular Biology for Agriculture. (1) Murray, C. J. L.; Lopez, A. D. Lancet 1997, 349, 1498–1504.

6210

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

approximately 900,000 persons affected annually.2 Patients who delay seeking health care for AMI have fewer and less effective treatment options, which results in higher morbidity and mortality rates.3 The time to diagnosis and treatment is crucial for improving outcomes from AMI. The faster a diagnosis can be made, the faster a patient can either be admitted to the emergency room or sent home with less anxiety. However, patients admitted to emergency rooms with chest pain are usually diagnosed on the basis of their medical history, physical examinations, electrocardiograms, and body surface maps.4-6 Frequently the results are inconclusive since it is often difficult to differentiate between diseases that display similar symptoms in emergency medicine.7,8 Furthermore, approximately one-quarter of all myocardial infarctions are silent, without chest pain or other symptoms.9 All of these factors make AMI a disease with a high rate of misdiagnosis.10 Therefore there is an urgent need for a fast and reliable test to facilitate triage, diagnosis, and adequate treatment strategies. To address these problems and improve the accuracy of detection, cardiac markers, including C-reaction protein,11 CKMB,12 cardiac troponins (troponin I and troponin T),13,14 myelo(2) Thom, T.; Hasse, N.; Rosamond, W.; Howard, V. J. Circulation 2006, 113, e85-e151. (3) Ryan, T. J.; Anderson, J. L.; Antman, E. M.; Braniff, B. A.; Brooks, N. H.; Califf, R. M. J. Am. Coll. Cardiol. 1996, 28, 1328–1419. (4) McClelland, A. J.; Owens, C. G.; Menown, I. B.; Lown, M.; Adgey, A. A. Am. J. Cardiol. 2003, 92, 252–257. (5) Ripa, R. S.; Holmvang, L.; Maynard, C.; Sejersten, M.; Clemmensen, P.; Grande, P.; Lindahl, B.; Lagerqvist, B.; Wallentin, L.; Wagner, G. S. J. Electrocardiol. 2005, 38, 180–186. (6) Drew, B. J.; Schindler, D. M.; Zegre, J. K.; Fleischmann, K. E.; Lux, R. L. J. Electrocardiol. 2007, 40, S15–S20. (7) Parodi, G.; Pace, S. D.; Carrabba, N.; Salvadori, C.; Memisha, G.; Simonetti, I.; Antoniucci, D.; Gensini, G. F. Am. J. Cardiol. 2007, 99, 182–185. (8) Schindler, D. M.; Lux, R. L.; Shusterman, V.; Drew, B. J. J. Electrocardiol. 2007, 40, S145–S149. (9) Kannel, W. B. Cardiol. Clin. 1986, 4, 583–591. (10) Wu, A. H. B.; Apple, F. S.; Gibler, W. B.; Jesse, R. L.; Warshaw, M. M.; Valdes, R. J. Clin. Chem. 1999, 45, 1104–1121. (11) Takahashi, T.; Anzai, T.; Yoshikawa, T.; Maekawa, Y.; Asakura, Y.; Satoh, T.; Mitamura, H.; Ogawa, S. J. Cardiol. 2003, 88, 257–265. (12) Mion, M. M.; Novello, E.; Altinier, S.; Rocco, S.; Zaninotto, M.; Plebani, M. Clin. Biochem. 2007, 40, 1245–1251. (13) Babuin, L.; Jaffe, A. S. J. Can. Med. Assoc. 2005, 173, 1191–1202. (14) Casals, G.; Filella, X.; Bedini, J. L. Clin. Biochem. 2007, 40, 1406–1413. 10.1021/ac9007418 CCC: $40.75  2009 American Chemical Society Published on Web 06/25/2009

peroxidase,15 myoglobin,16 and fatty acid binding protein (FABP),17 have been employed to diagnose AMI. A protein marker to be used in a diagnostic assay should be present in measurable concentrations in the early post-AMI stage and also have a high clinical selectivity. In this case, the use of smaller proteins would lead to earlier detection.18,19 Studies have shown that myoglobin (17.8 kDa) and FABP (15 kDa) are two relatively small proteins that show elevated levels in the bloodstream shortly after an acute myocardial infarction. Furthermore, a combined analysis of these two markers can specifically discern an injury to the myocardium after AMI onset.20,21 Therefore these two markers are suitable for an early rapid screening test for an AMI diagnosis with a high sensitivity and a high negative predictive value.22 The technique commonly used for the analysis of disease biomarkers is an immunoassay based on the reaction between an antibody to the biomarker and the antigen (biomarker) itself.23-25 Conventional enzyme-linked immunosorbent assays and radioimmunoassays have proved to be effective techniques for the analysis of cardiac protein markers.26,27 However, they often require multistep processing of samples, as well as laboratories equipped with proper instruments and well-trained personnel, making these techniques time-consuming and expensive options. Therefore, challenges remain in finding a rapid, simple, and sufficiently sensitive process for detecting cardiac markers. In recent years, a superparamagnetic polymer microsphere-aided immunoassay, with apparent advantages over previous carriers, has been widely used in clinical analyses and biomedical research.28-31 The high available surface-to-volume ratio of the microspheres provides more binding sites for analytes. Moreover, the microspheres can be dispersed homogeneously in a solution to reduce diffusional distances and facilitate the binding of analytes, resulting in shorter assay preparation and detection (15) Baldus, S.; Heeschen, C.; Meinertz, T.; Zeiher, A. M.; Eiserich, J. P.; Mu ¨nzel, T.; Simoons, M. L.; Hamm, C. W. Circulation 2003, 108, 1440–1445. (16) Matveeva, E. G.; Gryczynski, Z.; Lakowicz, J. R. J. Immunol. Methods 2005, 302, 26–35. (17) Chan, C. P.; Sum, K. W.; Cheung, K. Y.; Glatz, J. F. C.; Sanderson, J. E.; Hempel, A.; Lehmann, M.; Renneberg, I.; Renneberg, R. J. Immunol. Methods 2003, 279, 91–100. (18) Meng, X. Z.; Ming, M.; Wang, E. Y. Forensic Sci. Int. 2006, 160, 11–16. (19) Groot, M. J. M.; Wodzig, K. W. H.; Simoons, M. L.; Glatz, J. F. C.; Hermens, W. T. Cardiovasc. Res. 1999, 44, 315–324. (20) Mo ¨ckel, M.; Gerhardt, W.; Heller, G.; Klefisch, F.; Danne, O.; Maske, J.; Mu ¨ ller, C.; Sto ¨rk, T.; Frei, U.; Wu, A. H. Clin. Chim. Acta 2001, 303, 167– 179. (21) Chan, C. P.; Sanderson, J. E.; Glatz, J. F.; Cheng, W. S.; Hempel, A.; Renneberg, R. Z. Kardiol. 2004, 93, 388–397. (22) Brogan, G. X. J.; Friedman, S.; McCuskey, C.; Cooling, D. S.; Berrutti, L.; Thode, H. C. J.; Bock, J. L. Ann. Emerg. Med. 1994, 24, 665–671. (23) Pulli, T.; Ho ¨yhtya¨, M.; So ¨derlund, H.; Takkinen, K. Anal. Chem. 2005, 77, 2637–2642. (24) Bromberg, A.; Mathies, R. A. Anal. Chem. 2003, 75, 1188–1195. (25) Park, J.; Kurosawa, S.; Aizawa, H.; Hamano, H.; Harada, Y.; Asano, S.; Mizushima, Y.; Higaki, M. Biosens. Bioelectron. 2006, 22, 409–414. (26) Stone, M. J.; Willerson, J. T.; Gomez-Sanchez, C. E.; Waterman, M. R. J. Clin. Invest. 1975, 56, 1334–1339. (27) Katus, H. A.; Remppis, A.; Looser, S.; Hallermeier, K.; Scheffold, T.; Kubler, W. J. Mol. Cell Cardiol. 1989, 21, 1349–1353. (28) Dungchai, W.; Siangproh, W.; Lin, J. M.; Chailapakul, O.; Lin, S.; Ying, X. T. Anal. Bioanal. Chem. 2007, 387, 1965–1971. (29) Biagini, R. E.; Smith, J. P.; Sammons, D. L.; MacKenzie, B. A.; Striley, C. A. F.; Robertson, S. K.; Snawder, J. E. Anal. Bioanal. Chem. 2004, 379, 368–374. (30) Matsunaga, T.; Maeda, Y.; Yoshino, T.; Takeyama, H.; Takahashi, M.; Ginya, H.; Aasahina, J.; Tajima, H. Anal. Chim. Acta 2007, 597, 331–339. (31) Zou, M.; Gao, H.; Li, J.; Xu, F.; Wang, L.; Jiang, J. Anal. Biochem. 2008, 374, 318–324.

times. Additionally, the superparamagnetism of the microspheres allows them to be separated easily from the solution during assay procedures with little or no residual magnetization once the field is removed. All of these merits indicate a promising future for superparamagnetic microspheres in rapid, high-throughput screening in clinical analysis.32 Nevertheless, shortcomings such as biodegradability and toxicity limit further development of polymer microspheres for applications in biological analysis, especially for online and point of care testing.33,34 Here we describe an improved magnetic protein microsphereaided sandwich immunoassay for the analysis of cardiac markers, myoglobin and heart-type fatty acid binding protein (H-FABP). In contrast to conventional magnetic polymer microsphere-based protocols, the magnetic protein microspheres possess excellent biodegradability, non-toxicity, and biocompatibility,35,36 and can partially mimic conditions in whole blood samples for clinical use.37 Using this approach, serial dilutions of samples containing the cardiac biomarkers myoglobin and H-FABP were analyzed. EXPERIMENTAL SECTION Instruments. A scanning electron microscope (SEM, JSM6701F, Japan), transmission electron microscope (TEM, Hitachi H-8100, Japan), and atomic force microscope (AFM, AJ-IIIa, China) were employed to characterize the morphology and structure of the prepared γ-Fe2O3 nanoparticles and HSA/γ-Fe2O3 microspheres. The crystal structure of γ-Fe2O3 nanoparticles was analyzed with an X-ray diffractometer (Philips D/Max-2500, Holland) using a monochromatic X-ray beam with nickelfiltered Cu-KR radiation. The IR spectrum of HSA/γ-Fe2O3 microspheres was recorded using a Fourier transform infrared spectrophotometer (FT-IR, Nicolet NEXUS 670, U.S.A.). Magnetic measurements of γ-Fe2O3 nanoparticles and HSA/γ-Fe2O3 microspheres were carried out on a vibrating sample magnetometer (LAKESHORE-7304, U.S.A.) by changing H between +1375 and -1375 Oe. The iron concentration in the HSA/γFe2O3 microspheres was determined by an atomic absorption spectrophotometer (Hitachi 180-80, Japan). Fluorescence images of the post-immunoassay microspheres were taken by an inverted fluorescence microscope (Olympus CKX41, Japan) equipped with a high-resolution CCD camera (MicroPublisher 5.0 RTV, U.S.A.). Fluorescence intensity in aqueous suspension was recorded using a spectrofluorophotometer (Hitachi F-4500, Japan). Materials and Reagents. Superparamagnetic γ-Fe2O3 nanoparticles utilized in this study were prepared from magnetite (Fe3O4) according to methods proposed elsewhere.38,39 (For more detailed information on their synthesis and characterization see Supporting Information). Monoclonal capture antibodies (antimyoglobin 7C3 and anti-H-FABP 10E1), detection antibodies (32) Farrell, S.; Ronkainen-Matsuno, N. J.; Halsall, H. B.; Heineman, W. R. Anal. Bioanal. Chem. 2004, 379, 358–367. (33) van der Voort, D.; Pelsers, M. M.; Korf, J.; Hermens, W. T.; Glatz, J. F. J. Immunol. Methods 2004, 295, 1–8. (34) Haik, Y.; Cordovez, M.; Chen, C. J.; Chatterjee, J. Eur. Cell. Mater. 2002, 3, 41–44. (35) Kramer, P. A. J. Pharm. Sci. 1974, 63, 1646–1647. (36) Lee, T. K.; Sokoloski, T. D.; Royer, G. P. Science 1981, 213, 233–235. (37) Gupta, P. K.; Hung, C. T. J. Microencapsulation 1989, 6, 427–462. (38) Qu, S. C.; Yang, H. B.; Ren, D.; Kan, S. H.; Zou, G.; Li, D. M.; Li, M. H. J. Colloid Interface Sci. 1999, 215, 190–192. (39) Sun, Y. K.; Ma, M.; Zhang, Y.; Gu, N. Colloids Surf. A: Physicochem. Eng. Aspects 2004, 245, 15–19.

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

6211

Scheme 1. Schematic Presentation of the Process of Analysis of AMI Markers (Myoglobin and H-FABP) by Superparamagnetic Protein Microsphere-Aided Fluoroimmunoassay

(antimyoglobin 4E2 and anti-H-FABP 9F3), human heart myoglobin, and H-FABP were purchased from Hytest Ltd. (Turku, Finland). BCA Protein Assay Kit, EZ-Link-Sulfo-NHS-LC-Biotinylation Kit, and EZ-Label FITC Protein Labeling Kit (which was used to label anti-H-FABP 9F3) were received from Pierce Biotechnology (Rockford, IL). The Alexa Fluor594 Labeling Kit was from Molecular Probes (Eugene, OR) and was used to label antimyoglobin 4E2. Human serum albumin (HSA) and bovine serum albumin (BSA) were purchased from Sigma-Aldrich (St. Louis, MO). Cardiac troponin I (cTnI) was obtained from Biosyn & Biotech Ltd. (Beijing, China). Human serum was from BioMag Ltd. (Beijing, China). Other reagents and chemicals were purchased from local commercial suppliers and were of analytical reagent grade, unless otherwise stated. Deionized (DI) water (Milli-Q, Millipore, Bedford, MA) was used to prepare aqueous solutions. General Process. The superparamagnetic protein microsphere-aided sandwich fluoroimmunoassay (Scheme 1) starts with the preparation of magnetic HSA/γ-Fe2O3 microspheres by combining superparamagnetic γ-Fe2O3 nanoparticles and HSA. Then the capture antibodies (antimyoglobin 7C3 and anti-HFABP 10E1) are grafted onto the protein microspheres using an avidin-biotin protocol. Upon addition of a sample containing antigen (myoglobin or H-FABP), the antigen occupies the antigen-binding site of the antibody, and it can subsequently be assayed using fluoro-labeled detection antibodies (Alexa fluor594-labeled antimyoglobin 4E2, or FITC-labeled anti-HFABP 9F3). 6212

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

Preparation of Magnetic HSA/γ-Fe2O3 Microspheres. Magnetic HSA/γ-Fe2O3 microspheres used in this study were prepared by an improved heat-stabilization process.40,41 Freshly prepared γ-Fe2O3 nanoparticles (60.0 mg) were first dispersed in 0.8 mL of DI water and sonicated for 5 min at room temperature. Then 200.0 mg of HSA and 24.0 mL of cottonseed oil containing 160 µL of sorbitan sesquioleate were added to the γ-Fe2O3 suspension; the mixture was sonicated again until an emulsion was created. The emulsion was then added dropwise over 15 min to 80 mL of cottonseed oil at 140 °C with stirring at 280 rpm, and the mixture was stirred for an additional 20 min under the same conditions. Then the microsphere suspension was cooled to room temperature and extracted five times with diethyl ether to remove the cottonseed oil and byproducts. After separation with an external magnetic field and washing with phosphate-buffered saline (PBS, pH 7.4), the HSA/γ-Fe2O3 magnetic microspheres were stored in a refrigerator at 4 °C until use. Modification of HSA/γ-Fe2O3 Microspheres with Avidin. The HSA/γ-Fe2O3 microspheres were modified with avidin in three steps. All steps were performed at room temperature, and in a laminar flow hood to maintain sterility of all reagents. Caution was used in handling all human biological material. Preparation of Sulfhydrylated Avidin. Avidin solution (360 µL, 10 mg/mL) was first mixed into 9 µL of 13 mg/mL N-succinimidyl-S-acetylthioacetate (SATA) solution and incubated (40) Chatterjee, J.; Haik, Y.; Chen, C. J. Colloid Polym. Sci. 2001, 279, 1073– 1081. (41) Sahin, S.; Selek, H.; Ponchel, G.; Ercan, M. T.; Sargon, M.; Hincal, A. A.; Kas, H. S. J. Controlled Release 2002, 82, 345–358.

for 60 min in a shaker incubator. After removing the unreacted SATA using Slide-A-Lyzer MINI Dialysis Units (MWCO 3,500, Pierce) against PBS (pH 7.4), 36 µL of freshly prepared deacetylation solution (0.5 M hydroxylamine, 25 mM EDTA in PBS, pH 7.4) was then added to the reaction mixture and incubated for 2 more hours. After purification using Slide-A-Lyzer MINI Dialysis Units against PBS (pH 7.4, containing 5.0 mM EDTA), the sulfhydrylated avidin was obtained. Preparation of Maleimide-Modified HSA/γ-Fe2O3 Microspheres. Sulfosuccinimidyl-4-(N-maleimidomethyl)-cyclohexane1-carboxylate (1.19 mg) was first added to 400 µL of 10 mg/mL HSA/γ-Fe2O3 suspension (in PBS, pH 7.4). Then the mixture was incubated for 60 min with gentle shaking. The maleimidemodified HSA/γ-Fe2O3 microspheres were then dialyzed using Slide-A-Lyzer MINI Dialysis Units against PBS (pH 7.4, containing 5.0 mM EDTA) and separated with an external magnet. Preparation of Avidin-Modified HSA/γ-Fe2O3 Microspheres. The prepared sulfhydrylated avidin and maleimidemodified HSA/γ-Fe2O3 microspheres were first mixed together in a molar ratio of 2.5:1 (HSA: avidin) and incubated for 4 h in a shaker incubator. Then, the mixture was centrifuged for 10 min at 10 °C with a speed of 1500 rpm. The supernatant was reserved for a BCA protein assay (see Supporting Information) to determine the amount of avidin incorporated on the magnetic microspheres. The recovered avidin-modified HSA/γ-Fe2O3 microspheres were resuspended in 400 µL of 1% (w/v) BSA solution and incubated for 4 h at 4 °C to block all remaining non-specific protein binding sites. After dialysis against PBS, the avidin-modified HSA/γ-Fe2O3 microspheres were stored in a refrigerator at 4 °C until use. Biotinylation of Antimyoglobin 7C3 and anti-H-FABP 10E1. Biotinylation of antimyoglobin 7C3 and anti-H-FABP 10E1 were carried out using the EZ-link Sulfo-NHS-LC Biotinylation Kit following its instructions with some modifications. First, 30 µL of antimyoglobin 7C3 (5.6 mg/mL) and anti-H-FABP 10E1 (5.8 mg/ mL) were each diluted to 2 mg/mL using PBS, and then 2.27 or 2.35 µL of Sulfo-NHS-LC-Biotin solution (10 mM), respectively, were added to these solutions. The mixtures were incubated for 1 h at room temperature in a shaker incubator at a speed of 300 rpm. After dialysis against PBS (pH 7.4) for 10 h at 4 °C, the biotinylated antimyoglobin 7C3 and anti-H-FABP 10E1 were obtained and stored in a refrigerator at 4 °C. The amounts of biotin on the biotinylated antimyoglobin 7C3 and anti-H-FABP 10E1 were each determined using the 4-hydroxyazobenzene-2-carboxylic acid (HABA) protocol (see Supporting Information). Grafting Capture Antibodies onto HSA/γ-Fe2O3 Microspheres. First, 160 or 110 µL of the prepared avidin-modified HSA/γ-Fe2O3 microspheres (10 mg/mL) were respectively added to 1 mL of the biotinylated antimyoglobin 7C3 solution (0.1 mg/mL) or 1 mL of the biotinylated anti-H-FABP 10E1 solution (0.1 mg/mL). Then they were incubated for 30 min at 37 °C in a shaker incubator at 150 rpm. After washing with PBS (pH 7.4, containing 1% (w/v) BSA), the antibody-grafted HSA/γ-Fe2O3 microspheres were stored at 4 °C in a refrigerator until use.

Superparamagnetic Protein Microsphere-Aided Sandwich Fluoroimmunoassay. A series of 10 µL volumes of capture antibody-grafted microspheres (10 mg/mL) were loaded into 0.2 mL polystyrene cuvettes (coated with 1% (w/v) BSA prior to use) and magnetically separated to remove the suspension medium. Then a volume of serially diluted myoglobin or H-FABP (10 µL in PBS) was pipetted into each cuvette and incubated at 37 °C for 40 min (for myoglobin) or 30 min (for H-FABP) in a shaker incubator. After magnetic separation and washing with PBS (pH 7.4, containing 0.05% (v/v) Tween-20), the antigen-bound microspheres were respectively resuspended in Alexa Fluor594-labeled antimyoglobin 4E2 (50 µg/mL, 15 µL for each reaction cuvette) or FITC-labeled anti-H-FABP 9F3 solution (40 µg/mL, 10 µL for each reaction cuvette). After incubation with gentle shaking at 37 °C for a period (30 min for myoglobin and 25 min for H-FABP), the post-immunoassay microspheres were magnetically separated, washed, and resuspended in PBS. The fluorescence images of the post-immunoassay microspheres were recorded using an inverted fluorescence microscope equipped with a high-resolution CCD camera. Blank controls were run simultaneously during each experiment. Following the same procedure, blood serum samples containing the two protein markers were also studied and the postimmunoassay microsphere suspensions were all analyzed using a spectrofluorophotometer (Hitachi F-4500, Japan) with a 150 W Xenon discharge excitation source.42 A 1 cm path length quartz cuvette was used to measure the fluorescence intensity. With excitation at 494 nm (H-FABP) or 590 nm (myoglobin), fluorescent emission of the immunocomplexes at 520 nm (H-FABP) or 617 nm (myoglobin) was each detected. All optical measurements were carried out at room temperature under ambient conditions. Meanwhile, control experiments for studying the specificity of this assay were also carried out using antibody-conjugated microspheres with cTnI and H-FABP (or myoglobin). Data Analysis and Statistics. The data acquisition and analysis of the fluoroimmunoassay were carried out with ImagePro Plus 6.0 (Media Cybernetics Inc.) and SPSS 12.0 (SPSS Inc.) software. Data are presented as means ± SD for the measured fluorescence intensities of each experiment. The t-test for unpaired values was used to evaluate the significance of differences between the myoglobin or H-FABP concentrations measured in the different tests. P < 0.01 was considered statistically significant. RESULTS AND DISCUSSION Synthesis and Characterization of HSA/γ-Fe2O3 Magnetic Microspheres. As magnetic carriers for this study, superparamagnetic HSA/γ-Fe2O3 microspheres were synthesized using an improved heat-stabilization process. Figure 1a shows a SEM image of the prepared HSA/γ-Fe2O3 microspheres, which indicates that the magnetic HSA microspheres consisted of uniform, round microspheres with an average diameter of 8 µm, which were well distributed even when dried. No agglomerates were found. However, some small solids were attached to the microsphere surfaces, and they did not look smooth. We considered the small solids on their surfaces to be broken microspheres physically attached to the intact microspheres; they could be removed by washing with DI (42) Ma, Q.; Wang, X.; Li, Y.; Shi, Y.; Su, X. Talanta 2007, 72, 1446–1452.

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

6213

Figure 1. SEM (a, 3500 × magnification) and AFM (b) images of HSA/γ-Fe2O3 microspheres. (c) The magnetic hysteresis loop of HSA/γFe2O3 microspheres at 300 K (Ms ) 10 emu/g, and Hc ) 11 Oe). (d) Temperature-dependent magnetism of HSA/γ-Fe2O3 microspheres under 1000 Oe.

water. To further study the surface morphology of the HSA/ γ-Fe2O3 microspheres, we characterized their surfaces using an atomic force microscope after washing three times with DI water. The AFM image (Figure 1b) showed that the surfaces of the prepared HSA/γ-Fe2O3 microspheres were very smooth and uniform. FT-IR analysis (see Supporting Information) showed that the characteristic absorption bands of the prepared HSA/γ-Fe2O3 microspheres were at 2925 cm-1, 2855 cm-1 (methyl, methylene group), 1630 cm-1 (carbonyl group), 1379 cm-1 (carbon-carbon double bond), and 3436 cm-1 (hydroxyl group). All of these absorption bands confirmed the presence of human serum albumin in the microspheres. Also, the characteristic adsorption band of Fe-O was observed at 669 cm-1. Overall, the FT-IR spectrum provided supportive evidence that HSA/γ-Fe2O3 microspheres had been obtained. Analysis of the iron concentration using an atomic absorption spectrometer showed that the γ-Fe2O3 content in the HSA/γ-Fe2O3 microspheres was 37.8%. Figure 1c is the magnetization curve of the prepared HSA/γFe2O3 microspheres, which was similar to that of the γ-Fe2O3 nanoparticles (see Supporting Information) and demonstrated a symmetrical hysteresis loop. This phenomenon is characteristic of superparamagnetic materials; in other words, the HSA/γ-Fe2O3 microspheres become magnetized in the presence of a magnetic field. Once the field is removed, only a minimal residual magnetization remains within the particles. This makes these particles ideal for magnetic immunoassay; if the microspheres retained substantial residual magnetization they would ag6214

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

glomerate, making immunoassay impossible. Additionally, the temperature-dependence of the magnetization of the HSA/γFe2O3 microspheres (Figure 1d) showed a magnetization loss phenomenon, which is also a typical characteristic of superparamagnetic materials.43 The protein microspheres still retained good magnetism at routine immunoassay temperatures, usually below 50 °C. The saturation magnetization (Ms) of the prepared HSA/ γ-Fe2O3 microspheres was 10 emu/g with a very small coercivity (Hc ) 11 Oe). Compared with the γ-Fe2O3 nanoparticles, the HSA/γ-Fe2O3 microspheres exhibited a relatively low saturation magnetization. However, satisfactory magnetic-responsive aggregation and redispersion properties could be clearly seen in a study on the magnetic performance of the protein microspheres in DI water with external magnetic fields (see Supporting Information). When the external magnet was removed, the magnetic microspheres could be fully dispersed by gentle shaking. Therefore, the HSA magnetic microspheres could be easily attached and recycled by an external magnetic field, but remained very suitable for the next magnetic fluoroimmunoassay of AMI protein markers. Grafting Capture Antibodies onto HSA/γ-Fe2O3 Microspheres. Because of its high affinity and stability,44 the biotin-avidin system was employed to graft the capture antibodies (antimyoglobin 7C3 and anti-H-FABP 10E1) onto the HSA/γ-Fe2O3 microspheres in the current study. The whole process was (43) Ding, X. B.; Li, W.; Zheng, Z. H.; Zhang, W. C.; Deng, J. G.; Peng, Y. X.; Chan, A. S. C.; Li, P. J. Appl. Polym. Sci. Symp. 2001, 79, 1847–1851. (44) Diamandis, E. P.; Christopoulos, T. K. Clin. Chem. 1991, 37, 625–636.

Scheme 2. Schematic Presentation of the Synthesis of Capture Antibody-Grafted γ-Fe2O3/HSA Microspheres Using the Avidin-Biotin System As a Linkera

a (a) Primary amine groups in avidin were reacted with SATA, and (b) sequentially treated with NH2OH to synthesize the sulfhydrylated avidin. (c) Maleimide-modified HSA/γ-Fe2O3 microspheres were reacted with the sulfhydrylated avidin to form an avidin and HSA/γFe2O3 microsphere conjugate. (d) HSA/γ-Fe2O3 microspheres were modified with a bifunctional cross-linker sulfo-SMCC to generate maleimide-modified HSA/γ-Fe2O3 microspheres. (e) The biotinylated antibodies were bound to the avidin-modified HSA/γ-Fe2O3 microspheres using the reaction between biotin and avidin.

composed of three steps (Scheme 2). First, the S-acetylthioacetic group was introduced into avidin by the nucleophilic reaction between N-succinimidyl-S-acetylthioacetate (SATA) and the terminal amino groups of avidin (a in Scheme 2). After deacylation using hydroxylamine (b in Scheme 2), sulfhydryl groups were left covalently connected to avidin. Then the sulfhydrylated avidin (c in Scheme 2) was reacted with the maleimide-activated HSA/ γ-Fe2O3 microspheres obtained by the reaction between sulfosuccinimidyl-4-(N-maleimidomethyl)-cyclohexane-1-carboxylate (sulfo-SMCC) and HSA/γ-Fe2O3 microspheres (d in Scheme 2). After removing excess SATA and sulfo-SMCC, avidinconjugated HSA/γ-Fe2O3 microspheres were obtained. The amount of avidin grafted onto the HSA/γ-Fe2O3 microspheres was determined using the BCA Protein Assay Kit (it determines how many biotinylated capture antibodies can be used to react with the avidin-modified HSA/γ-Fe2O3 microspheres). The analysis results showed that the amount of bound avidin on the microsphere was 67 mg/g. On the basis of the analytical results and the knowledge that one avidin can bind 4 biotin groups,44 the avidin-modified microspheres were reacted with biotinylated capture antibody (biotinylated antimyoglobin 7C3 or biotinylated anti-H-FABP 10E1) to conjugate the capture antibodies onto the magnetic protein microspheres (e in Scheme 2). Reaction Time Optimization for Biomarker and Its Antibody. In this sandwich fluoroimmunoassay, quantification of a biomarker is based on the intensity of the fluorescent reporter signal. However, the fluorescence intensity may depend on many factors. Although we had obtained high-quality magnetic protein

microspheres, covered with BSA to block all remaining nonspecific protein binding sites, carefully analyzed the amount of avidin on the protein microspheres and biotin conjugated on the capture antibodies, and utilized amounts of fluoro-labeled detection antibodies that were in excess for the highest concentrations of biomarkers analyzed in this study, another factor, the time for reaction between the biomarker and its antibody, could also greatly affect the fluorescence intensity. The reaction time determines if the reaction between biomarker and its antibody (capture antibody or fluoro-labeled detection antibody) is complete. To improve the fluoroimmunoassay sensitivity and accuracy, we also optimized the reaction time in this study (see Supporting Information). The optimal capture time for the reactions between protein markers (myoglobin and H-FABP) and their capture antibody-grafted protein microspheres was 40 min for myoglobin and 30 min for H-FABP; the optimal incubation time for the reactions between fluoro-labeled detection antibodies and protein marker-immobilized microspheres was 30 min for myoglobin and 25 min for H-FABP because the fluorescence intensities seemed to reach stable maxima at those times. Moreover, the sensitivity at these reaction times was much higher than immediately after adding the sample. Superparamagnetic Protein Microsphere-Aided Sandwich Fluoroimmunoassay. Using this improved HSA/γ-Fe2O3 microsphere-aided fluoroimmunoassay, a series of aqueous phase and human serum samples with different concentrations of the AMI early protein markers myoglobin and H-FABP were analyzed. The relationship between the amount of biomarker detected (as revealed by the fluoresAnalytical Chemistry, Vol. 81, No. 15, August 1, 2009

6215

Figure 2. HSA/γ-Fe2O3 microsphere-aided fluoroimmunoassay of cardiac markers myoglobin and H-FABP. Top: Fluorescence images of the post-immunoassay protein microspheres at different cardiac marker concentrations (Left: myoglobin; Right: H-FABP). Bottom (a, b, c, d): The relationship between the average fluorescence intensity and the concentration of cardiac markers obtained by the analyzed fluorescence images using software Image-Pro Plus 6.0 (a, b) and recorded on a spectrophotometer (c, d). Error bars represent ( SD.

Figure 3. Photostability of post-immunoassay microspheres and free dyes in aqueous phase. Measurements were performed in suspensions of the post-immunoassay microspheres or solutions of the free dyes. The solution was continuously illuminated by a 150 W xenon lamp at the optimal excitation (590 nm for Alexa Fluor594 and 494 for FITC) for 20 min using a spectrofluorophotometer, and the fluorescence intensity was acquired continuously over the 20 min period.

cence intensity) and the actual concentration of the biomarker is shown in Figure 2 and the Supporting Information. For the two biomarkers myoglobin and H-FABP the minimum detectable concentrations were 10 ng/mL and 1 ng/mL, respectively. To compare with the expected concentrations of the two biomarkers in a healthy patient (17 ± 6 ng/mL for myoglobin; 3.0 ± 1.3 ng/mL for H-FABP)45,46 and the variability of upper reference limits across the population after the onset of AMI (49-105 ng/mL for myoglobin; 5.3-12 ng/mL for H-FABP),47 the detectable ranges reported in the current study (45) Panteghini, M.; Pagani, F. Clin. Chem. 1997, 43, 2435–2435. (46) Okamoto, F.; Sohmiya, K.; Ohkaru, Y.; Kawamura, K.; Asayama, K.; Kimura, H.; Nishimura, S.; Ishii, H.; Sunahara, N.; Tanaka, T. Clin. Chem. Lab. Med. 2000, 38, 231–238. (47) Pelsers, M. M. A. L.; Hermens, W. T.; Glatz, J. F. C. Clin. Chim. Acta 2005, 352, 15–35.

6216

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

include them (for more detailed information see Supporting Information). Therefore, the measurement range of this assay provided a reliable clinic metric to some extent. In addition, statistical analysis showed that the differences in fluorescence intensity between different biomarker concentrations were statistically significant (P < 0.01). However, there were no statistically significant differences (P > 0.01) between two repeated experiments at the same myoglobin or H-FABP concentration. All of these results indicate that this assay can rapidly measure changes in these two biomarkers after the onset of AMI, and it also possesses high sensitivity compared with other protocols reported previously (see Supporting Information). Photobleaching Experiment. The purpose of doing this experiment was to observe the stability of these post-immunoassay microspheres when they are exposed to an aqueous environment

for biological applications. Photobleaching is known to depend on solvent interactions and is thought to occur as a bimolecular reaction between the dye and, for example, dissolved oxygen.48,49 To investigate the photostability of the post-immunoassay microspheres, the intensity of the fluorescence was monitored versus time on a spectrofluorometer (Hitachi F-4500, Japan).50,51 Measurements were performed in suspensions of the post-immunoassay microspheres or solutions of the free dye. Figure 3 shows the photobleaching behavior of the post-immunoassay microspheres and free dye. The results showed that there was basically no photobleaching for the post-immunoassay microspheres over a long period of continuous intensive excitation with a 150 W xenon lamp. Specificity of This Assay. An immunoassay is based on the reaction between an antibody to the biomarker and the antigen (biomarker) itself. However, the specificity of antibodies for their antigens is very important for the sensitivity and accuracy of an assay. In this study, the antibodies used were all commercialized monoclonal antibodies, which are intrinsically specific to their antigens. Control experiments for the specificity study of this assay using cTnI and myoglobin (or H-FABP) also proved that the interactions between the antigens and their corresponding monoclonal antibodies were specific (see Supporting Information). CONCLUSION In summary, we have demonstrated an improved protein magnetic microsphere-aided sandwich fluoroimmunoassay for the (48) Song, L.; Hennink, E. J.; Young, I. T.; Tanke, H. J. Biophys. J. 1995, 68, 2588–2600. (49) Hartmann, P.; Leiner, M. J. P.; Lippitsch, M. E. Sensors Actuat. B 1995, 29, 251–257. (50) Lian, W.; Litherland, S. A.; Badrane, H.; Tan, W.; Wu, D.; Baker, H. V.; Gulig, P. A.; Lim, D. V.; Jin, S. Anal. Biochem. 2004, 334, 135–144. (51) Santra, S.; Zhang, P.; Wang, K.; Tapec, R.; Tan, W. Anal. Chem. 2001, 73, 4988–4993.

analysis of the early AMI protein markers myoglobin and H-FABP. The lowest detectable concentrations for myoglobin and H-FABP were 10 ng/mL and 1 ng/mL, respectively. The prepared HSA/ γ-Fe2O3 magnetic microspheres possessed uniform morphology and sufficient superparamagnetic properties to allow great stability and fidelity in this assay. The magnetic protein microsphere-aided protocol could be used as a sensitive and efficient method for detecting the AMI biomarkers myoglobin and H-FABP. The rapid detection that it provides would have far more diagnostic value than would absolute measurements during the first hours after the onset of AMI symptoms. Also, the protocol devised here can be used as a model for establishing general methods for other protein marker assays in the fields of clinical diagnostics and molecular biology. ACKNOWLEDGMENT The authors are very grateful to Dr. Wenhao Yu at the Department of Pathology, Harvard Medical School, and Dr. Jing Zhou at the Department of Genetics and Genomics, School of Medicine, Boston University, for their helpful discussion. The authors would also like to acknowledge funding from the National Natural Science Foundation of China (No. 207 750 59), the Ministry of Education of the People’s Republic of China (NCET08-0464), and the Northwest A&F University. SUPPORTING INFORMATION AVAILABLE Additional information as noted in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

Received for review April 7, 2009. Accepted June 15, 2009. AC9007418

Analytical Chemistry, Vol. 81, No. 15, August 1, 2009

6217