Rapid, femtomolar bioassays in complex matrices combining microfluidics and magnetoelectronics
Introduction
Rapid advancements in biosensor technology over the past several decades has led to widespread implementation across a variety of application fields, ranging from biodefense (Hindson et al., 2005, Ivnitski et al., 2003, Lim et al., 2005, Peruski and Peruski, 2003), to medical diagnostics (Andreotti et al., 2003, Gao et al., 2004), to environmental monitoring (Baeumner, 2003, Farre and Barcelo, 2003). The diversity of applications and sample matrices results in a nonuniform set of operational requirements. Even if one only considers biodefense applications, the paradigms of detect-to-protect, detect-to-treat, and detect-to-monitor place different emphases on sensitivity, selectivity, portability, and sample-to-answer time. Meanwhile, there is an emerging need for a decentralized laboratory with simple, portable systems suitable for use by non-technical personnel, such as first-responders, infantry, and point-of-care medical staff. Within this vast operational landscape, it is unlikely that a single technology will satisfy all applications; therefore, specific end-user requirements will guide the selection among promising approaches.
The most promising biosensor technologies distinguish themselves across several operational characteristics, including but not limited to sensitivity, selectivity, sample-to-answer time, portability, operational complexity, and cost. For example, the bio-bar-code assay of Mirkin and co-workers has state-of-the-art analytical sensitivity and selectivity for the detection of nonamplified nucleic acids and proteins (Nam et al., 2004, Stoeva et al., 2006). Using this technology, attomolar target concentrations can be detected, without the complications of amplification (e.g. via polymerase chain reaction (PCR)), because the bar-code DNA provides signal amplification. In a competing approach, Lieber and co-workers have achieved highly sensitive, multiplexed, label-free detection of cancer markers in complex matrices using an elaborate nanowire-based device (Zheng et al., 2005). They report detection of prostate specific antigen in undiluted (but de-salted) serum at concentrations as low as 0.9 pg/ml. In contrast, extremely rapid (<5 min) and simple lateral flow immunoassay kits are available from a number of sources, but these kits perform poorly for pathogen detection and are not easily multiplexed (Lim et al., 2005).
The majority of biosensors incorporate solid-phase binding assays whereby target analytes are captured by biomolecular recognition and labeled with “reporters,” such as fluorophores, enzymes, radiolabels, nanoparticles, or electrochemically active species. In general, the means for detecting the reporter label is independent of the target capture and labeling assay. System performance (i.e. sensitivity, specificity, reproducibility) is rarely limited by the ability to detect the labels, but rather by the background signal associated with nonspecific adsorption in the assay. A classic example is nucleic assay hybridization microarrays (gene chips) incorporating fluorescence labeling and detection (Epstein et al., 2002, Michalet et al., 2003). Typically, the target oligonucleotides are fluorescently labeled during PCR amplification, and multiple surfactant-laden wash steps and/or temperature cycles are applied to remediate nonspecific binding to noncomplementary capture probe spots. However, the performance of microarrays is not usually limited by the ability to detect the fluorescence. A similar state of affairs exists for conventional enzyme-linked immunosorbent assays (ELISAs) and related solid-phase immunoassays, where performance is typically limited by the background label density, not label detection.
The ubiquity of reporter label-based bioassays arises from the wide choice of labels that can be applied to common assay schemes. Labels used in bioassays are generally molecular or nanoscale in size in order to match the size of the biomolecular recognition probes and analyte targets. In contrast, micrometer-scale labels – such as microbeads used for magnetic separation and as assay substrates – have been discounted as labels based on concerns about their large relative size (Graham et al., 2004, Wang et al., 2005). However, we find microbead labels offer two significant advantages over smaller labels that far outweigh any disadvantages from the size mismatch. First, it is far simpler to detect low numbers of microbeads than molecular fluorophores, chromophores, or nanoparticles, with individual microbeads readily counted with routine optical microscopy (Lee et al., 2000, Mulvaney et al., 2004) or magnetic detection (Rife et al., 2003, Rife and Whitman, 2004). Second, and more significantly, we find that if a controlled laminar flow is maintained at the capture surface, fluidic drag forces can be applied to the microbead labels to preferentially remove nonspecifically bound labels and thereby dramatically improve the assay performance (Rife and Whitman, 2004). We have leveraged these advantages and use magnetic microbeads to label nucleic acid hybridization assays and protein immunoassays performed on top of a microarray of magnetoelectronic sensors. This combination of microfluidics and magnetoelectronics enables highly sensitive and specific multiplexed detection in minutes in complex sample matrices.
Section snippets
Assay fluidics
Two flow cells were used for this work. The data from Fig. 1 was collected in straight channels 2 cm long × 250 μm high × 800 μm wide with all capture spots located in the center of a channel. All other assays were performed in an acrylic assembly incorporating multiple flow cells mounted on a microscope slide (see Supplementary Fig. 1 online). Each flow cell had central dimensions of 2.8 mm long × 2.2 mm wide × 100 μm high, with tapered entrance and exits to assure uniform, laminar flow of reagents across
Results
Atomic force microscopy studies have demonstrated that the binding strengths of specific biological ligand-receptor interactions are at least an order of magnitude greater than those of the ligand nonspecifically interacting with receptors or nonfouling surfaces (Lee et al., 1994, Metzger et al., 1999). Several bioassay methods have been developed that exploit this differential in binding by applying magnetic forces to microparticle labels, a technique known as magnetic force discrimination
Discussion
As the diversity of biosensing applications continues to expand, no single technology currently meets all required performance criteria, and relatively few are capable of addressing multiple application areas. The most successful systems aim to achieve high assay sensitivity and specificity while minimizing sample preparation requirements, operational complexity, and sample-to-answer time. The tradeoffs between these operational characteristics will ultimately guide end-user selection.
Assay
Conclusions
Utilizing fluidic force discrimination assays, we have demonstrated femtomolar detection of both DNA and proteins in a rapid, multiplexed format using as few as two reagents. Our magnetic, microbead labels provide us with several advantages for biosensing. Because the beads are used as a physical label, the assays can directly detect targets out of complex matrices with minimal (and in some cases no) sample processing; the only requirement is that the matrix not interfere with biomolecular
Acknowledgments
This work was supported in part by a Cooperative Research and Development Agreement with Seahawk Biosystems Inc. (NCRADA-NRL-04-341). The authors would also like to thank Seahawk and SAE Magnetics (H.K. Ltd.) for providing the BARC® chip used in this work. Authors SPM, MDK, MM, and MWS are employees of Nova Research Inc., 1900 Elkins St. Suite 230, Alexandria, VA 22308, USA.
References (40)
- et al.
Biosens. Bioelectron.
(2005) - et al.
Biosens. Bioelectron.
(1998) - et al.
J. Virol. Meth.
(2006) - et al.
Biosens. Bioelectron.
(2000) - et al.
Anal. Chim. Acta
(2002) - et al.
Anal. Chim. Acta
(2004) - et al.
Trac-Trend. Anal. Chem.
(2003) - et al.
Chem. Eng. Sci.
(1967) - et al.
Trends Biotechnol.
(2004) - et al.
Biosens. Bioelectron.
(2005)
J. Virol. Meth.
Biosens. Bioelectron.
Anal. Biochem.
Anal. Chim. Acta
Biosens. Bioelectron.
Anal. Biochem.
Sens. Actuat. A: Phys.
Biosens. Bioelectron.
J. Mag. Mag. Mater.
Anal. Chim. Acta
Cited by (119)
Magnetic properties-based biosensors for early detection of cancer
2021, Biosensor Based Advanced Cancer Diagnostics: From Lab to ClinicsBiomarkers detection with magnetoresistance-based sensors
2020, Biosensors and BioelectronicsDevelopment of magnetic sensor technologies for point-of-care testing: Fundamentals, methodologies and applications
2020, Sensors and Actuators, A: PhysicalAn integrated model for bead-based immunoassays
2020, Biosensors and BioelectronicsCitation Excerpt :Bead-based assays can take various formats, and in particular, surface coverage assays have gained considerable research interest, because they can be ultra-sensitive, rapid, and readily implemented in microfluidics for point-of-care (PoC) applications (Bruls et al., 2009; Dittmer et al., 2008; Koets et al., 2009; Tekin et al., 2013; Wu et al., 2018). Various surface coverage assays are developed, and they use different parameters and manifest different performance (Bruls et al., 2009; Mulvaney et al., 2007; Otieno et al., 2014; Wu et al., 2018). For example, an optomagnetic sensor achieved sub-picomolar LOD using 500 nm nanoparticles (Bruls et al., 2009), while attomolar LOD was obtained using fluidic force discrimination assays and 2.8 μm microbeads (Mulvaney et al., 2009).
Nanomaterial- and micromaterial-based immunoassays
2018, Handbook of Immunoassay Technologies: Approaches, Performances, and Applications
- 1
Current address: National Air and Space Museum, Smithsonian Institution, Washington, DC 20013, USA.