Anal. Chem. 2003, 75, 1141-1146
Immobilized Particle Arrays: Coalescence of Planar- and Suspension-Array Technologies Priscilla Wilkins Stevens,† C. H. Jeffrey Wang,‡ and David M. Kelso*,†
Department of Biomedical Engineering, Robert R. McCormick School of Engineering and Applied Science, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208-3107, and Spherotech, Inc., 1840 Industrial Drive, Suite 270, Libertyville, Illinois 60048-9467
Combining positive attributes of planar arrays and suspension arrays, immobilized particle arrays offer a new format in which immobilized submicrometer particles are arrayed on hydrogel-coated slides, providing 100+ assay replicates within each spot. This research describes how to prepare immobilized protein arrays and how to assay the binding of labeled target molecules to the arrayed capture probes. The assay system exhibits an intrinsic dynamic range of two to three decades, with coefficients of variation from 5 to 10%. For antibody-antigen binding, target capture appears to be reaction rate limited. For labeled antibody binding to antigen on the immobilized particles, the detection limit is ∼0.5 ng/mL. When antibodies on the immobilized particles exhibit multivalent binding of target molecules, the detection limit is ∼0.01 ng/mL. For protein arrays, potential advantages of this format are improved coating of the capture reagent, an increased number of options for protein presentation, reduced mass transport effects, and higher density multiplexing. Protein arrays offer the capability of simultaneously interrogating multiple bimolecular interactions.1 Applications include protein interaction mapping, proteomics, functional genomics, drug development, immunodiagnostics, hybridoma- and phage-library screening, and protein profiling of cells and tissues.2,3 Two different array formats, planar and suspension, have been developed for these applications. With the planar array format, capture reagents are spotted onto a slide, filter, well, or other planar surface, and each array element is identified by its location within the array’s grid. Target molecules captured by an element of a planar array are generally detected by laser scanning,4 although other methods have also been utilized, including mass spectrometry5,6 and surface plasmon resonance imaging.7 In a * To whom correspondence should be addressed. Phone: 847-467-2167. Fax: 847-491-4928. E-mail:
[email protected]. † Northwestern University. ‡ Spherotech, Inc. (1) Wilson, D. S.; Nock, S. Curr. Opin. Chem. Biol. 2002, 6, 81-85. (2) Zhu, H.; Bilgin, M.; Bangham, R.; Hall, D.; Casamayor, A.; Bertone, P.; Lan, N.; Jansen, R.; Bidlingmaier, S.; Houfek, T.; Mitchell, T.; Miller, P.; Dean, R. A.; Gerstein, M.; Snyder, M. Science 2001, 293, 2101-2105. (3) Templin, M. F.; Stoll, D.; Schrenk, M.; Traub, P. C.; Vohringer, C. F.; Joos, T. O. Trends Biotechnol. 2002, 20, 160-166. (4) Ramdas, L.; Wang, J.; Hu, L.; Cogdell, D.; Taylor, E.; Zhang, W. BioTechniques 2001, 31, 546-552. 10.1021/ac020580d CCC: $25.00 Published on Web 01/30/2003
© 2003 American Chemical Society
suspension array, however, capture reagents are immobilized on coded microparticle surfaces. Most often flow cytometry is applied to detect targets captured by the various suspension array elements, each of which is identified by the particle’s code.8,9 Other methods for reading particle-based array technologies include fiber-optic microsensing of particles that adhere to etched microwells,10,11 fluorescence detection of microbead arrays constrained in flow cells12 or microfluidic channels,13,14 thermal lens microscopy of particles in a microchannel,15,16 and optical imaging of striped metallic particles.17 In this paper, we describe a hybrid array format with attributes of both planar arrays and suspension arrays: the immobilized particle array (IPA). Microparticles coated with capture probes are immobilized in an array on a thin film applied to a microscope slide or other planar surface. A single array spot, therefore, contains a large number of submicrometer particles widely dispersed to minimize mass transport limitations. In the research described in this paper, the particles in an arrayed spot are all coated with the same capture reagent, providing multiple assay replicates. Coded particles, however, provide another alternative, since they offer the potential for a single spot of the array to (5) Fung, E. T.; Enderwick, C. BioTechniques Suppl. 2002, 32, S34-S41. (6) Rodi, C. P.; Darnhofer-Patel, B.; Stanssens, P.; Zabeau, M.; van den Boom, D. BioTechniques 2002, 32, S62-S69. (7) Nelson, B. P.; Grimsrud, T. E.; Liles, M. R.; Goodman, R. M.; Corn, R. M. Anal. Chem. 2001, 73, 1-7. (8) Nolan, J. P.; Sklar, L. A. Trends Biotechnol. 2002, 20, 9-12. (9) Taylor, J. D.; Briley, D.; Nguyen, Q.; Long, K.; Iannone, M. A.; Li, M.-S.; Ye, F.; Afshari, A.; Lai, E.; Wagner, M.; Chen, J.; Weiner, M. P. BioTechniques 2001, 30, 661-669. (10) Michael, K. L.; Taylor, L. C.; Schultz, S. L.; Walt, D. R. Anal. Chem. 1998, 70, 1242-1248. (11) Ferguson, J. A.; Steemers, F. J.; Walt, D. R. Anal. Chem. 2000, 72, 56185624. (12) Brenner, S.; Johnson, M.; Bridgham, J.; Golda, G.; Lloyd, D. H.; Johnson, D.; Luo, S.; McCurdy, S.; Foy, M.; Ewan, M.; Roth, R.; George, D.; Eletr, S.; Albrecht, G.; Vermaas, E.; Willams, S. R.; Moon, K.; Burcham, T.; Pallas, M.; DuBridge, R. B.; Kirchner, J.; Fearon, K.; Mao, J.; Corcoran, K. Nat. Biotechnol. 2000, 18, 630-634. (13) Hayes, M. A.; Polson, N. A.; Phayre, A. N.; Garcia, A. A. Anal. Chem. 2001, 73, 5896-5902. (14) Buranda, T.; Huang, J.; Perez-Luna, V. H.; Schrayer, B.; Sklar, L. A.; Lopez, G. P. Anal. Chem. 2002, 74, 1149-1156. (15) Sato, K.; Tokeshi, M.; Odake, T.; Kimura, H.; Ooi, T.; Nakao, M.; Kitamori, T. Anal. Chem. 2000, 72, 1144-1147. (16) Sato, K.; Tokeshi, M.; Kimura, H.; Kitamori, T. Anal. Chem. 2001, 73, 12131218. (17) Walton, I. D.; Norton, S. M.; Balasingham, A.; He, L.; Oviso, D. F., Jr.; Gupta, D.; Raju, P. A.; Natan, M. J.; Freeman, R. G. Anal. Chem. 2002, 74, 22402247.
Analytical Chemistry, Vol. 75, No. 5, March 1, 2003 1141
a protein assay that utilizes this array format. In the following paper, we detail the imaging and analysis of IPAs.
Figure 1. Immobilized particle array. (A) This array contains 25 spots of microparticles arrayed on a planar surface. (B) Each spot within the array contains microparticles coated with capture reagent. In this example, particle codes distinguish the five different types of capture reagents coated on particles within the spot.
present particles coated with a variety of capture reagents. This alternative means that IPAs may potentially afford additional multiplexing capacity for ultrahigh-density array applications (Figure 1). The IPA format combines the best features of both suspension array and planar array technologies. Advocates of suspension arrays8,18 point out that each array element is prepared in bulk, a variety of attachment chemistries are available, hundreds of replicates are read for each assay, and detection limits are on the level of a few hundred fluors per particle. Proponents of planar arrays1,3 stress the advantages of printing tens of thousands of features on a single microscope slide, high throughput, and low cost per assay. Both sets of advantages apply to IPAs. A significant attribute of IPAs is the small size of the capture domain. A typical spot on a conventional planar array has a diameter of 150-200 µm, while a suspension array uses 100 5-6µm-diameter particles. The IPA format, however, typically employs 100 0.8-µm-diameter particles for each array element. Since mass transport effects are inversely proportional to the diameter of the capture region,19 IPAs have more than a 6-fold advantage over suspension arrays and a 200-fold advantage over planar arrays. To retain the advantage of a small capture domain upon immobilization, IPA geometry mimics ultramicroelectrode arrays, where electrodes of 100 particles are printed per spot (data not shown). The results in Figure 5 demonstrate that MPF is not significantly affected by a small change in particle number per spot. Since all of the particles represent identical reaction sites, results are similar, for example, even for 2-fold differences in particle number. Printing 100-200 particles/spot corresponds to an average interparticle distance of 15-20 µm. Theoretically, since this is more than 10 times the radius of the capture site, the diffusion fields should not overlap, and the response will closely follow the kinetics of particles in suspension.20 Spots on planar arrays may exhibit irregularities of size, shape, or quality due to the imperfections of slide surfaces and the contact printing processes.33,34 IPA capture probes, however, are coated onto the surface of precisely sized particles while the particles are in suspension, thus ensuring uniformity of capture-probe presentation on each particle. Printing ∼50-500 microparticles in each array spot enhances assay precision by providing a large number of replicate values for each assay. In addition, for IPAs, precise dispense volume and spot size are not particularly important, since each particle represents an independent assay. Antibody Capture by IPA Antigens. A typical configuration for protein arrays has a solution-phase antibody target that probes a number of arrayed capture antigens. To evaluate IPA performance with a solution-phase antibody probing particle-immobilized antigen, we tested the binding of Cy3-rAbRHRP to arrayed HRP particles. Eight 3 × 3 grids of HRP particles were printed, 109 ( 33 particles/spot, with one grid at each well center. Seven concentrations of Cy3-rAbRHRP ranging from 1000 to 0.1 ng/mL were applied to the wells, while the final well was a blank that received dilution buffer containing no antibody target. Results, presented in Figure 6, are quantitative down to 1 ng/mL and demonstrate a linear range from 1.5 to 25 ng/mL. The particles are approximately half saturated at the 100 ng/mL target concentration and fully occupied with 1000 ng/mL target, where the maximum signal is 20 counts/ms. For each target concentration, the between-spot coefficient of variation (CV) was generally between 5 and 10%. With the 0.1 ng/mL target concentration, which is not plotted in Figure 6, signal (mean MPF ) 0.150 ( 0.009 counts/ms) was indistinguishable from the background signal of blank spots (mean MPF ) 0.145 ( 0.010 counts/ms). From the slope of the response between the 0.39 and 1.56 ng/mL target concentrations (slope 0.63) and the SD of signals from the 0 and 0.1 ng/mL target concentrations (SD ) 0.010), we calculated the theoretical limit (33) Eisen, M. B.; Brown, P. O. Methods Enzymol. 1999, 303, 179-205. (34) Dolan, P. L.; Wu, Y.; Ista, L. K.; Metzenberg, R. L.; Nelson, M. A.; Lopez, G. P. Nucleic Acids Res. 2001, 29, e107.
Figure 6. Cy3-rAbRHRP captured on HRP particles. Eight spots were analyzed for each target concentration. Each diamond plots the net MPF for a single spot.
of detection. The two-SD limit of detection for this direct-label assay is 0.32 ng/mL, or 2 pM antibody target. Other investigators report similar detection limits for systems where fluorescently labeled reporters bind to capture proteins covalently attached on activated slides.35-37 When an additional enzyme amplification step has been added, limits of detection in the 1-10 pg/mL range have been observed.38-40 Another system, which incorporated a waveguide into the slide for evanescent field excitation, also has a detection limit of ∼10 pg/mL.41 A particlebased format employing a fluorescently labeled reporter demonstrated detection limits in the range 10-100 pg/mL.42 Detection limit is dependent on the affinity of the binding partners as well as the sensitivity of the labeling and imaging systems.43 At equilibrium, which would obtain in ∼8 h for a Kd of 10-9 M and a dissociation rate constant of 10-4 s-1, a 1 ng/mL target concentration would result in a fractional occupancy of ∼0.01, which is ∼80 molecules/particle.22 The reaction would be approximately half completed in 2 h, yielding 40 molecules/ particle.31 Since under these assumptions the number of molecules captured is directly proportional to target concentration, one would expect less than 1 molecule captured per particle for target concentrations less than 0.01 ng/mL. High-Avidity Binding. In the previous experiment, the Cy3rAbRHRP target antibodies interacted with the HRP antigen through one or both of the two identical antibody-combining sites at the tips of the antibody arms. To test whether a system with multivalent binding would yield a lower theoretical limit of detection for the IPA, Cy3-rAbRHRP target was reacted with RrIgG (35) MacBeath, G.; Schreiber, S. L. Science 2000, 289, 1760-1763. (36) Haab, B. B.; Dunham, M. J.; Brown, P. O. Genome Biol. 2001, 2, 4.1-4.13. (37) Silzel, J. W.; Cercek, B.; Dodson, C.; Tsay, T.; Obremski, R. J. Clin. Chem. 1998, 44, 2036-2043. (38) Moody, M. D.; Van Arsdell, S. W.; Murphy, K. P.; Orencole, S. F.; Burns, C. BioTechniques 2001, 31, 186-194 (39) Schweitzer, B.; Roberts, S.; Grimwade, B.; Shao, W.; Wang, M.; Fu, Q.; Shu, Q.; Laroche, I.; Zhou, Z.; Tchernev, V. T.; Christiansen, J.; Velleca, M.; Kingsmore, S. F. Nat. Biotechnol. 2002, 20, 359-365. (40) Wiese, R.; Belosludtsev, Y.; Powdrill, T.; Thompson, P.; Hogan, M. Clin. Chem. 2001, 47, 1451-1457. (41) Pawlak, M.; Schick, E.; Bopp, M. A.; Schneider, M. J.; Oroszian, P.; Ehrat, M. Proteomics 2002, 2, 383-393. (42) Swartzman, E. E.; Miraglia, S. J.; Mellentin-Michelotti, J.; Evangelista, L.; Yuan, P.-M. Anal. Biochem. 1999, 271, 143-151. (43) At equilibrium, the equation for fractional occupancy (f) less than 0.1, is f ) (VC0)/(SΓc + VKd), where V is reaction volume, C0 target concentration, S total capture surface area, Γc capture probe density, and Kd equilibrium dissociation constant.
Analytical Chemistry, Vol. 75, No. 5, March 1, 2003
1145
Figure 7. Cy3-rAbRHRP captured on RrIgG particles. Eight spots were analyzed for each concentration of target. Each diamond plots the net MPF for a single spot.
particles. In this system, multiple gAbRrIgG antibodies on the particle surface might interact with the same Cy3-labeled target antibody, resulting in tighter binding of the target molecule to the particles printed in the IPA. A 3 × 3 grid of RrIgG particles was printed with 211 ( 43 particles/spot at each of the eight well centers. Seven 4-fold dilutions of Cy3-rAbRHRP ranging from 10 to 0.002 ng/mL were applied to the wells, and dilution buffer with no antibody target was applied to the eighth well. With this assay, the analytical region extended from 10 to 0.01 ng/mL (see Figure 7), while the 0.002 ng/mL target concentration yielded signal indistinguishable from the background signal of the blank well. Spot-to-spot CVs were generally 5-10%. The two-SD limit of detection calculated from low-end target concentrations was 8 pg/mL, or ∼55 fM. The avidity effect of solid-phase polyclonal RrIgG antibodies was most likely responsible for the lower detection limit, since the solution-phase target was the same as the one used with HRP particles (Figure 6), where the detection limit was 40-fold higher (320 versus 8 pg/mL). This effect has been observed by other investigators.44,45 CONCLUSION With IPA technology, arrays of dispersed, submicrometer capture domains can be fabricated with standard laboratory equipment and reagents. Our research demonstrates feasibility for immobilizing protein-coated, particles on activated, hydrogel slides using a quill pin arrayer, for capturing labeled targets on the immobilized particles, and for obtaining assay measurements (44) Soukka, T.; Harma, H.; Paukkunen, J.; Lovgren, T. Anal. Chem. 2001, 73, 2254-2260. (45) Kalinin, N. K.; Ward, L. D.; Winzor, D. J. Anal. Biochem. 1995, 228, 238244. (46) Schwab, C.; Bosshard, H. R. 1992. J. Immunol. Methods 1992, 147, 125134.
1146
Analytical Chemistry, Vol. 75, No. 5, March 1, 2003
with a CCD camera on an epifluorescence microscope. These proof-of-concept studies suggest that the analytical performance of IPAs is similar to conventional planar and suspension arrays in terms of precison and detection limits. Demonstrating feasibility for IPAs as a new array format means that it is now possible to explore the potential of IPAs for characterizing kinetic interactions of proteins. If, as theory predicts, mass transport effects have been eliminated, then it should be possible to determine kinetic rate constants from sparsely populated spots of submicrometer particles. It is now also possible to explore the potential of IPAs for providing solutions for difficulties traditionally encountered in fabricating planar protein arrays. Whereas nucleic acids are linear, have known directionality, and can easily be attached to a surface via either the 5′ or 3′ end, proteins are globular and chemically complex. When proteins are adsorbed or covalently attached onto solid phases, there is no assurance that the region of interest will retain its conformation or functionality.46 Standard planar arrays require that all capture proteins be attached to the surface with the same chemistry. The IPA format, however, allows for flexibility during protein array fabrication, since there is the possibility to optimize protein presentation for each individual capture-probe protein. Options for each protein include attachment to any of a variety of microparticle surfaces via a whole range of covalent or affinity interactions. Then all the members of such a diverse series of microparticles may be arrayed by printing on the same slide. Like suspension arrays, protein coating of microparticle surfaces can be performed in aqueous solutions under conditions optimal for the individual proteins being coated. Particles are coated in a batch and stored in liquid, thus providing a large stock of immobilized proteins ready for printing. Testing the functionality of proteins coated on the solid surface is likewise vastly simplified, because protein-coated microparticles can be tested prior to printing the array. For many applications, microparticles coated with a single type of capture reagent will be printed in each spot of the IPA. Thus, unlike suspension arrays, which require expensive, proprietary, fluorescently labeled particles, many IPAs can be prepared with inexpensive, unlabeled particles for all array components. However, by printing multiple types of coded particles within each arrayed spot, IPAs also offer the potential for generating ultrahighdensity arrays. Ultrahigh-density arrays will be important both for applications involving more than 104 capture probes and for situations where array size must be miniaturized to maximize the information potential from rare or limited samples. Received for review September 17, 2002. Accepted December 17, 2002. AC020580D
