Conjugation of Biomolecules with Luminophore-Doped Silica

Anal. Chem. 2001, 73, 4988-4993

Conjugation of Biomolecules with Luminophore-Doped Silica Nanoparticles for Photostable Biomarkers Swadeshmukul Santra,† Peng Zhang,† Kemin Wang,‡ Rovelyn Tapec,† and Weihong Tan*,†

Department of Chemistry and McKnight Brain Institute, The University of Florida, Gainesville, Florida 32611-7200, and College of Chemistry and Chemical Engineering, Hunan University, Changsha, Hunan, China

A new molecular conjugation method has been developed to label biomolecules with optically stable metalorganic luminophores, such as tris(2,2′-bipyridyl)dichlororuthenium(II) hexahydrate (Rubpy), which are otherwise not possible for direct linking with the biomolecules. Unique biochemical properties of the biomolecule can, thus, be associated with photostable luminophores. This opens a general way to conjugate desired biomolecules using a sensitive signal transduction method. It also promotes the application of excellent luminescent materials, especially those based on photostable metalorganic luminophores, in biochemical analysis and biomolecular interaction studies. The conjugation method is based on uniform luminophore-doped silica (LDS) nanoparticles (63 ( 4 nm). These nanoparticles have been prepared using a water-in-oil (W/O) microemulsion method. The controlled hydrolysis of tetraethyl orthosilicate (TEOS) in W/O microemulsion leads to the formation of monodisperse LDS nanoparticles. The luminophores are doped inside the nanoparticles, and the particle’s silica surfaces can be used to covalently bind with biomolecules. The luminophores are well-protected from the environmental oxygen when they are doped inside the silica network. As an example, we used an antibody for leukemia cell recognition. The antibody was first immobilized onto the luminophore-doped nanoparticle through silica chemistry and then was used for leukemia cell identification by an optical microscopy imaging technique. The leukemia cells were identified easily, clearly, and with high efficiency using these antibody-coated nanoparticles. The advantages of using small, uniform luminophore-doped nanoparticles are discussed. Molecular probes for biomolecular recognition are of great importance in the fields of chemistry, biology, and medical sciences as well as in biotechnology. These probes have been used in studies of biological functions and in ultrasensitive detection of biological species responsible for many diseases. The demand to develop highly sensitive, nonisotopic analysis systems for biological applications, such as DNA sequencing, clinical diagnostics etc, has driven nanomaterials toward biomedical fields * Phone: 352-846-2410. Fax: 352-846-2410. E-mail: [email protected]. † The University of Florida. ‡ Hunan University.

4988 Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

and biotechnology.1-6 On the other front, the development of nontoxic and biocompatible luminescent materials for cell staining and visualization has been critically important in cellular biology and in ultrasensitive immunoassay. Recent advances in the semiconductor nanocrystals (quantum dots) open a promising field toward the development of a new generation of luminescent biomarkers. These luminescent dots have been functionalized to couple biomolecules.1,2,6 It has been suggested that these markers are better labeling agent than commonly used organic dyes. However, these luminescent quantum dots have not been extensively used as optical probes because of their poor solubility in water (unless they are modified), agglutination, blinking properties, and moderate quantum yields. Fluorescent latex particles7-10 (e.g., polystyrene particles) have also been used as labeling materials. However, their large size, tendency to agglomerate, swelling, and dye leaking have prevented their effective application as luminescent biomarkers for ultrasensitive biochemical analysis. Synthesis of dye-doped micrometer-sized silica particles by the Stober method has also been reported.11 The formation mechanism of such dye-doped silica particles clearly demonstrates the feasibility of incorporating both hydrophobic and hydrophilic dyes. However, no further investigations on photostability, size-tuning, and bioconjugation of those dye-doped silica particles were made. Their applications for biomarkers have not been pursued. In an effort to prepare efficient and uniform biomarkers, we have developed novel luminophore-doped silica nanoparticles. Effective biomarkers should combine the excellent biospecificity of designated biomolecules with optically stable luminophores, (1) Bruchez, M. Jr.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science 1998, 281, 2013. (2) Chan, W. C. W.; Nie, S. Science 1998, 281, 2016. (b) Taylor, J. R.; Fang, M. M.; Nie, S. M. Anal. Chem. 2000, 72, 1979. (3) Shenton, W.; Davis, S. A.; Mann, S. Adv. Mater. 1999, 11, 449. (4) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A. Science 1997, 277, 1078. (5) Mucic, R. C.; Storhoff, J. J.; Mirkin, C. A.; Letsinger, R. L. J. Am. Chem. Soc. 1998, 120, 12674. (6) Mitchell, G. P.; Mirkin, C. A.; Letsinger, R. L. J. Am. Chem. Soc. 1999, 121, 8122. (7) Harma, H.; Lehtinen, P.; Takalo, H.; Lovgren, T. Anal. Chim. Acta 1999, 387, 11. (8) Giunchedi, P.; Conte, U.; Chetoni, P.; Saettone, M. F. Eur. J. Pharm. Sci. 1999, 9, 1. (9) Adler, J.; Jayan, A.; Melia, C. D. J. Pharm. Sci. 1999, 88, 371. (10) Bourel, D.; Rolland, A.; Leverge, R.; Genetet, B. J. Immunol. Methods 1988, 106, 161. (11) Shibata, S.; Yano, T.; Yamane, M. Jpn. J. Appl. Phys. 1. 1998, 37, 41. (b) Shibata, S.; Taniguchi, T.; Yano, T.; Yamane, M. J. Sol-Gel Sci. Technol. 1997, 10, 263. 10.1021/ac010406+ CCC: $20.00

© 2001 American Chemical Society Published on Web 09/13/2001

Figure 1. Schematic representation of antibody immobilization process onto the luminophore-doped silica nanoparticle’s surface.

such as metalorganic luminophores. However, the metalorganic luminophores often cannot be directly linked with biomolecules such as an antibody or an enzyme. We are developing a molecular conjugation method that can be used to link an antibody with optically stable luminophore-doped nanoparticles. These nanoparticles have a silica surface and can, thus, be conjugated with desired biomolecules through many existing molecular immobilization methods based on silica chemistry.12-14 To demonstrate the application of such a method and the luminophore-doped silica nanoparticles as efficient biomarkers for bioimaging, we have modified a LDS nanoparticle surface with an antibody. The antibody for leukemia cell binding is covalently immobilized onto the nanoparticle’s surface. This immobilization technique can be extended easily to other biomolecules, for example, proteins, enzymes, peptides, and DNA.1,2,12-18 The newly prepared antibody immobilized nanoparticles are used to label human leukemia cells sensitively through antigen-antibody recognition. EXPERIMENTAL SECTION Materials. Tris(2,2′-bipyridyl)dichlororuthenium(II) hexahydrate (Rubpy), tetraethyl orthosilicate (TEOS), cyanogen bromide (CNBr), Triton X-100 (TX-100), n-hexanol, and cyclohexane were purchased from Aldrich Chemical Co. Inc. Ammonium hydroxide (28-30 wt %) and poly(methyl methacrylate) (PMMA) were obtained from the Fisher Scientific Co. Trimethoxysilylpropyldiethylenetriamine (DETA), a silanization reagent, was purchased from Sigma Chemical Co. All other chemicals were of analytical reagent grade. Mouse anti-human CD10 antibody in RPMI 1640 fetal calf serum and antibiotics and mononuclear lymphoid cells (∼2 million/mL) as a suspension in the cell culture medium were obtained from the Biotechnology Facility of the University of Florida. PBS buffer (pH 6.8) was used during the antibody immobilization and cell experiments. Distilled, deionized water (Easy Pure LF, Barnstead Co.) was used for the preparation of all aqueous solutions. Nanoparticle Synthesis and Surface Modification with Antibody. Nanoparticle Synthesis. The W/O microemulsion was (12) Tan, W.; Kopelman, R.; Barker, S.; Miller, M. Anal.Chem, 1999, 71, 606A. (b) Liu, X.; Tan, W. Mikrochim. Acta 1999, 131, 129. (13) Fang, X. H.; Liu, X.; Schuster, S.; Tan, W. J. Am. Chem. Soc. 1999, 121, 2921. (b) Liu, X.; Tan, W. Anal. Chem. 1999, 71, 5054. (14) Ingersoll, C. M.; Bright, F. V. Anal. Chem. 1997, 69, 403A. (15) Andres, R. P.; Bielefeld, J. D.; Henderson, J. I.; Janes, D. B.; Kolagunta, V. R.; Kubiak, C. P.; Mahoney, W. J.; Osifchin, R. G. Science, 1996, 273, 1690. (16) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Science, 1995, 270, 1335. (17) Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J. Nature, 1996, 382, 607. (18) Shenton, W.; Pum, D.; Sleytr, U. B.; Mann, S. Nature, 1998, 389, 585.

prepared first by mixing 1.77 mL of TX-100, 7.5 mL of cyclohexane, 1.8 mL of n-hexanol, and 340 µL of water. Aqueous luminophore solution was then added in such a way that the water-surfactant molar ratio was kept constant at 10. The final luminophore concentration in the mixture was 1.2 mM. In the presence of 100 µL of TEOS, a polymerization reaction19 was initiated by adding 60 µL of NH4OH. The reaction was allowed to continue for 24 h. After the reaction was completed, the LDS nanoparticles were isolated by acetone, followed by centrifuging and washing with ethanol and water several times to remove any surfactant molecules. We did ultrasonication while washing LDS nanoparticles to remove any physically adsorbed luminophores from the particle’s surface. About 35 mg of completely dry orange-colored LDS nanoparticles were obtained over acetone from 11.7 mL of microemulsion. For post silica coating, LDS nanoparticles were first redispersed in microemulsion, and then 100 µL of TEOS and 60 µL of NH4OH were added. After 24 h of aging, post coated LDS nanoparticles were collected in pure form. The thickness of the post silica coating was controlled by the amount of TEOS that was added and the aging time of polymerization reaction. We were able to make a coating as thick as 180 nm with this post coating treatment. Chemical Activation of the LDS Nanoparticle Surface. Dried LDS nanoparticles were suspended in 9.0 mL of 2 M sodium carbonate solution (activation buffer) by ultrasonication. A solution of CNBr in acetonitrile (1.0 g of CNBr dissolved in 0.5 mL of acetonitrile) was then added dropwise to the particle suspension (5 mg/mL) under stirring for 5 min at room temperature. Activated particles were washed twice with ice-cold water and twice with PBS buffer (pH 6.8). Covalent Immobilization of the Antibody onto the LDS Nanoparticle Surface. Antibody immobilization onto the silanized surface was then followed by using the cyanogen bromide pretreated LDS nanoparticles, as shown in Figure 1.20 Mouse anti-human CD10 antibody in RPMI 1640 fetal calf serum and antibiotics were used in this study. Its activity was first tested by Goat anti-mouse IgG, a polyclonal secondary antibody conjugated to FITC (fluorescein isothiocyanate). A 40-µL portion of antibody diluted in PBS buffer (pH 6.8) was added to the surface-modified particles, and stirring was continued for 24 h at 4 °C. Antibody immobilized nanoparticles were then treated with 10 mL of 0.03 M glycine solution for 30 min. The final product was washed, resuspended in PBS (pH 6.8) buffer, and stored at 4 °C for future usage. Note that after antibody immobilization onto the nanoparticles’ surfaces, there is no change (19) Stober, W.; Fink, A. J. Colloid Interface Sci. 1968, 26, 62. (20) Preparation protocol. Bangs Laboratories Inc., Fishers, IN.

Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

4989

Figure 2. TEM image of Rubpy-doped silica nanoparticles at 70 000× magnification (inset showing a region of 184 000× magnification).

in the optical properties for the silica nanoparticles doped with Rubpy. Instrumentation. Emission spectra were recorded on the Perkin-Elmer Spectrofluorometer, model LS50B. Silica nanoparticle size was measured using the Hitachi transmission electron microscope (TEM). Optical and fluorescence images of the LDS nanoparticles and leukemia cells were taken by a fluorescence microscope (Olympus, Japan) assembled with an intensified CCD (charge coupled device) camera. The microscope is equipped with an argon laser for excitation. This system has been used for effective single-molecule imaging and for single-cell studies.21,22 RESULTS AND DISCUSSION Nanoparticle Characterization. Microemulsion method yielded uniform luminophore-doped silica nanoparticles. These nanoparticles have been characterized with microscopy and spectroscopic methods. The LDS nanoparticles are uniform in size, 63 ( 4 nm in diameter (Figure 2), as characterized by TEM. At higher resolution (inset of Figure 2), luminophore particles are also visible as dark dots embedded inside the silica sphere as a result of the presence of the heavy metal ruthenium atom. The low polarity of the microemulsion water pool allows luminophore particles to aggregate while the silica network is being formed. We have not observed such dark dots in the pure silica particles at the same TEM resolution. Spectrofluorometric measurements were also used to characterize the nanoparticles. The excitation and the emission spectra of the pure Rubpy and LDS nanoparticles were measured in the aqueous solution. Pure Rubpy shows emission at 594 nm when excited at the 458 nm excitation band maxima in aqueous solution. The excitation spectra remain the same for the pure Rubpy and LDS nanoparticles in the aqueous solution, but the emission band maximum of the LDS nanoparticles shifts by 7 nm toward the (21) Fang, X.; Tan, W. Anal. Chem. 1999, 71, 3101. (22) Bui, J.; Zelles, T.; Lou, H. J.; Gallion, V.; Phillips, M. I.; Tan, W. J. Neurosci. Methods, 1999, 89, 9.

4990

Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

longer wavelength when compared with the pure Rubpy luminophore (Figure 3). We also measured the effect of oxygen pressure on the emission intensity of the pure Rubpy and LDS nanoparticles in the solid state. Rubpy is known to be a good oxygen-sensing reagent, because the luminescence of the luminophore can be quenched by oxygen.23 Our results show that there was no evident change in the emission intensity for the LDS nanoparticles when the air pressure was increased from 1 to 8 psi (data shown in supplement). Only when the pressure was further increased could a decrease in emission intensity be observed. This clearly shows that the luminophores are well-protected inside the silica network in normal atmosphere. Photobleaching Experiments. Rubpy is known to be a photosensitive compound. To investigate whether the Rubpy molecules still remain photosensitive when doped inside the silica network, we measured emission intensity with respect to time and excitation intensity. We studied photobleaching both in aqueous suspension (solution phase) and in a thin film of PMMA (solid phase). The purpose of doing this experiment in solution was to observe the stability of these particles when they are exposed to an aqueous environment for biological applications. The results from solution studies showed that there was basically no photobleaching for the nanoparticles over a long period of continuous intensive excitation with a 150 watt xenon lamp. The results were compared with a rhodamine 6G (Rh6G) aqueous solution. A 200µL portion of the sample solution was taken in a confined cuvette, and the spectrofluorometer was used for this experiment. With this setup, no noticeable photobleaching was observed (Figure 4) for LDS nanoparticles in solution for a period of 1 h. However, a slight decrease in fluorescence intensity was observed for the Rubpy, and a significant decrease was observed for the Rh6G molecules. It is worth noting that in a solution-phase photobleaching experiment, the luminophore doped nanoparticles keep on as a result of Brownian motion. Therefore, the particles are not continuously exposed to the excitation source. To test the photobleaching property in a more rigid situation, we immobilized the particles in PMMA films. The luminophores in the PMMA solid matrix are continuously exposed to optical excitation. A thin film sample doped with the nanoparticles was prepared on a microscope cover slip. A 0.1% PMMA solution in toluene was used to prepare the thin film matrix by spin coating. The cover slip with the thin film sample was then mounted over the microscope stage with laser excitation [a focused 500 mW argon laser (488 nm) in the visible range was used with appropriate filters (575nm-long pass filter and 610-nm band-pass filter for emission)]. We observed photobleaching for all of the samples tested, as shown in Figure 5. As expected, photobleaching was much more severe for Rh6G (Figure 5a) when compared with both pure Rubpy (Figure 5b) and LDS nanoparticles (Figure 5c). We believe that the luminophores, even in LDS nanoparticles (nanocomposites), were not immune from photobleaching under intense excitation by a focused laser beam. There are two possibilities for this photobleaching in the solid state. The first one is a minute amount of oxygen penetration into the silica network, and the second one is that some luminophores remain very close to the silica particle surface, because the surface area/volume ratio is quite high for (23) Hartmann, P.; Leiner, M. J. P.; Lippitsch, M. E. Sensors Actuat. B-Chem. 1995, 29, 251.

Figure 3. Luminescence excitation (left panel) and emission spectra (right panel) recorded in the aqueous phase for the pure Rubpy (b, d) and LDS nanoparticles (a, c).

Figure 4. Photobleaching experiment: (a) LDS nanoparticles, (b) the pure Rubpy and (c)rhodamine 6G (Rh6G) in solution phase with a 150 W xenon lamp excitation source. Experiments were done on a spectrofluorometer.

those nanoparticles. In both cases, additional silica coating on the nanoparticle surface will be a reasonable solution to minimize photobleaching of the luminophores. We, thus, employed a post silica coating onto the LDS nanoparticles (see nanoparticle synthesis section), expecting that further silica coating will completely prevent oxygen molecule penetration to the top surface and, hence, the photobleaching. These particles were synthesized and then were used in the photobleaching experiment. As shown in Figure 5d, our results clearly demonstrated that post silicacoated LDS nanoparticles were highly photostable. There was no photobleaching observed over a long period of intensive laser excitation. This observation, therefore, suggests that further silica coating completely isolates luminophore particles from the outside environment, thereby preventing oxygen molecule penetration. The thickness of the post coating could be as thin as a few nanometers. We have also tested the photostability of the LDS nanoparticles after their conjugation with the antibody for leukemia cell recognition. Similar results to what has been described above were obtained.

Figure 5. Photobleaching experiment: (a) pure Rh6G, (b) pure Rubpy, (c) LDS nanoparticles, (d)post silica-coated LDS nanoparticles, all dispersed in a thin film of PMMA, with excitation from a focused laser beam of a 0.503 W argon laser. Luminescence signal was collected by the ICCD camera coupled with the fluorescence microscope.

Availability of the Pure Silica Surface of LDS Nanoparticles. The availability of the pure silica surface of the LDS nanoparticles has been confirmed by the positive fluorescamine assay.24 The outside silica surface of the LDS nanoparticle has the same properties as that of a silica glass and, thus, can be easily modified by employing existing silica surface chemistry to attach desired functional groups. To demonstrate the general applicability of this approach, we first silanized the nanoparticle surface with DETA,25 a silanization agent that attaches the primary amine group to the silica surface. We then used fluorescamine, a nonfluorescent molecule that becomes highly fluorescent upon reacting with the primary aliphatic amine group, to test the LDS surface. Similar (24) Chung, L. A. Anal. Biochem. 1997, 248, 195. (25) Cordek, J.; Wang, X. W.; Tan, W. Anal. Chem. 1999, 71, 1529. (b) Qhobosheane, M.; Santra, S.; Zhang, P.; Tan, W. Analyst, 2001, 126, 12741278.

Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

4991

Figure 6. Optical and fluorescence images of leukemia cells (5-7 µm in size): (a) optical and (b) luminescence images of leukemia cells incubated with antibody-immobilized luminophore-doped nanoparticles, (c) optical and (d) luminescence images of leukemia cells incubated with bare luminophore-doped nanoparticles as a control.

fluorescence signal enhancement was obtained for the pure silica particles and for the LDS nanoparticles when they were treated with the same concentration of fluorescamine in dimethylformamide solution. The treated surface can be used for biomolecule conjugation through surface immobilization. This method can be a general method for the conjugation of optically stable luminophores with desired biomolecules for important applications, such as biochemical analysis. Application of the LDS Nanoparticles Conjugated with Antibody. The LDS nanoparticles conjugated with an antibody for the recognition of leukemia cells have been prepared and applied for cellular imaging. To demonstrate the biological application of the LDS nanoparticles as effective biomarkers, mouse anti-human CD10 antibody was used. The antibody was covalently immobilized onto the LDS nanoparticle surface using the CNBr pretreated LDS nanoparticles, as shown in Figure 1.20 We then used antibody-conjugated nanoparticles for labeling of human leukemia cells. The mononuclear lymphoid cells (about 2 million/mL) were obtained as a suspension in the cell culture medium. Most of the cells were round-shaped, as shown in Figure 6a. A cell suspension was incubated with CD10-immobilized nanoparticles for 2 h. After incubation, unbound nanoparticles were washed away with PBS buffer (pH 6.8). The cell suspension was then imaged using both optical microscopy and fluorescence microscopy. As shown in Figure 6b, all of the cells in the field of view of the microscope were labeled, as indicated by the bright emission of the luminophore-doped particles. The optical image (Figure 6a) correlated well with the fluorescence image (Figure 6b). The control experiments with bare luminophore-doped nanoparticles (no antibody attached) did not show any labeling of the cells, as shown in Figure 6c,d (optical image and luminescent image). This clearly shows that the nanoparticle conjugated with antibody is able to perform as a biomarker for cancer cells through the antibody-antigen recognition. We also carried out control experiments using the antibody-labeled LDS nanoparticles and PTK2 cells.25 There was no binding of the antibody-labeled particles to the cell membrane surface, because the specific antigen did not exist on the cell membrane surface. With further development of this system, the nanoparticles can serve as an efficient way for cancerous cell identification. 4992

Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

DISCUSSION There are a few interesting points worth discussing. We have shown an excellent approach for the preparation of uniform size, spherical-luminophore-doped silica nanoparticles. The microemulsion route has been found to be an easy and efficient method26-29 for the preparation of such nanoparticles over other existing methodologies.30,31 Although the water pool size of the microemulsion predominantly determines the size of the resulting nanoparticles, the amount of TEOS and the duration of the polymerization reaction also affect their sizes. The size of the water pool is defined by the molar ratio of the water to surfactant, and this ratio is designated as the W0 value. By reducing the value of W0, the amount of TEOS, and the reaction time, smaller 32 nanoparticles can be prepared. The isolation of the LDS nanoparticles from the microemulsion in pure form was also easily achieved by choosing proper solvent and centrifugation. The TEM results (Figure 2) show that the particles are very uniform and the luminophores are distributed homogeneously throughout the silica network. The presence of the heavy atom ruthenium is also seen clearly as black dots in the silica network. The red shift of 7 nm in the emission spectra of the LDS nanoparticle is observed when it is compared with the pure luminophore in the solution (Figure 3). This is due to the presence of silica network surrounding the luminophores, and the red shift could be explained as a weak interaction between the silica and the ruthenium ion in the silica network. Fluorescent latex particles have also been used as labeling materials for biomarkers. However, silica nanoparticles possess several advantages over them when used as biomarkers: (i) Silica nanoparticles are easy to centrifuge during preparation, functionalization, and other solution treatment processes because of their (26) Li, T.; Moon, J.; Morrone, A. A.; Mecholsky, J. J.; Talham, D. R.; Adair, J. H. Langmuir 1999, 15, 4328. (27) Shiojiri, S.; Hirai, T.; Komasawa, I. Chem. Commun. 1998, 14, 1439. (28) Stathatos, E.; Lianos, P.; DelMonte, F.; Levy, D.; Tsiourvas, D. Langmuir 1997, 13, 4295. (29) Chang, S.-Y.; Liu, L.; Asher, S. A. J. Am. Chem. Soc. 1994, 116, 6739; 1994, 116, 6745. (30) Li, M.; Schnablegger, H.; Mann, S. Nature 1999, 402, 393. (31) Davis, S. S. Trends Biotechnol. 1997, 15, 217. (32) Santra, S.; Tapec, R.; Theodoropoulou, N.; Dobson, J.; Hebard, A.; Tan, W. Langmuir, 2001, 17, 2900.

higher density of silica (e.g., 1.96 g/cm3 versus 1.05 g/cm3 for polystyrene). (ii) latex particles are generally more hydrophobic than silica particles and, therefore, tend to agglomerate in aqueous medium. To prevent agglomeration, stabilizers (e.g., surfactants) are used in most applications. The presence of silanol groups on the silica surface makes silica nanoparticles more hydrophilic, and therefore, it is easy to suspend them in aqueous medium. In addition, silica is a nontoxic substance and can easily be modified with biomolecules. (iii) Dye-doped latex particles swell in organic solvents, resulting in dye molecule leakage. Latex particles are also soluble in many organic solvents. These problems do not exist in silica-based nanoparticles, because silica is resistant to both aqueous (except extremely high pH solution) and nonaqueous solvents. The present work shows that the Rubpy-doped antibodyfunctionalized nanoparticles are high-quality markers for leukemia cell labeling and cell membrane surface antigen elucidation. The ratio between the intensities of the bright and the dark areas in the fluorescence image (Figure 6b) is over 500. In a standard method using FITC-conjugated secondary antibody for leukemia cell detection, a ratio of 100 is considered as a significantly high signal. The LDS nanoparticle assay provides a much-improved sensitivity for the detection using a similar antibody-antigen recognition method. The signal-to-background ratio, that is, the sensitivity for antigen detection, will be further increased with the optimization of the imaging system and the immobilization process. This, therefore, should provide a very useful way for detecting trace amounst of antigen when antigen expression is at an early stage during disease development. Our approach of making LDS nanoparticles is a general one, and it is not restricted to a particular type of luminophore molecule, as selected in this study. There are several important advantages of these luminophore-doped nanoparticles with functionality for biomolecular recognition. The nanoparticles are easily prepared using relatively simple procedures. They have uniform size and can be made as small as a few nanometers.32 We have also optimized the amount of luminophore loading (∼18.6 wt %) in Rubpy LDS nanoparticles to obtain maximum emission intensity, especially for ultrasensitive detection. Their surfaces can be functionalized in different ways to meet the needs on the basis of a variety of biomolecular recognition mechanisms. The conjugation could be widely useful for many different biomolecules, because the silica surface linkage used in this work is based on a functionalized amine group from the biomolecule. The luminophores doped inside the nanoparticles are well-protected from the surrounding environments and are, thus, immune to the potential quenching and bleaching in solution, which may exist in a rather

complicated medium, such as cellular culture solution. In addition, these nanoparticles are well-dispersed in water and can be easily handled for aqueous-solution-based applications. CONCLUSION Uniform luminophore-doped silica nanoparticles have been used for the development of a general method for biomolecular conjugation with optically stable luminophores. The nanoparticles have been prepared by the water-in-oil microemulsion method. A controlled hydrolysis of tetraethyl orthosilicate in W/O microemulsion has led to the formation of monodisperse luminophoredoped silica nanoparticles. The nanoparticles have a size of ∼63 nm, and are uniform, with a size distribution of less than 7%. The luminophores are well-protected from the environmental oxygen when they are doped inside the silica network of the nanoparticle. Post silica coating of the luminophore-doped silica nanoparticles has basically eliminated photobleaching, even with intense laser excitation. We have used this method to conjugate an antibody, using a silica immobilization method, for leukemia cell recognition with the newly developed luminophore nanoparticles. The antibodyconjugated nanoparticles have been applied effectively for leukemia cell recognition. The LDS nanoparticles have shown unique advantages over existing dye molecules (e.g., Rh6G), quantum dots, and latex-based fluorescent particles for biomolecular recognition and cellular staining in the following three major areas: easy preparation, good photostability, and high sensitivity. Our new method demonstrates a general way to conjugate desired biomolecules using a sensitive signal transduction method. The conjugation strategy can be used for a variety of biomolecules, because our approach uses the well-developed silica surface immobilization chemistry for biomolecular linking. The new method will also promote the application of excellent luminophores, especially those based on photostable metalorganic luminophores, in biochemical analysis and biomolecular interaction studies. ACKNOWLEDGMENT We thank Melissa G. Chen of I.C.B.R. Flow Cytometry Core Laboratory at The University of Florida for providing the leukemia cells and antibody samples as well as valuable discussion during the course of this study. This work is supported by NIH NS3989101 and CHE 9733650. K.W. and W.T. thank Oversea Youth Scholar grant 20028506 and 29825110 of China.

Received for review April 6, 2001. Accepted July 18, 2001. AC010406+

Analytical Chemistry, Vol. 73, No. 20, October 15, 2001

4993