biocd mutianalyte detection

High-Speed Label-Free Multi-Analyte Detection through Micro-interferometry

Manoj M. Varma*a, David D. Nolte**a, Halina D. Inerowiczb, Fred E Regnierb

a Dept. of Physics, Purdue University, West Lafayette, IN 47907 b Dept. of Chemistry, Purdue University, West Lafayette, IN 47907

ABSTRACT

Interferometers can detect optical path changes down to a billionth-lambda at the half intensity point at quadrature, (defined when the signal and reference waves are out of phase by ninety degrees). We have fabricated interferometric microstructures on silicon all operating at quadrature. The ultimate capability of this approach is the fabrication of over a billion interferometric biosensors on a single spinning disk having the capacity for mega-samples per second sampling speed. As an initial proof of principle of this technique, we have detected the presence of immobilized anti-mouse IgG and the specific binding of mouse IgG at a sampling rate of 100kiloSamp/sec, while non-specific binding observed was low. We will demonstrate that this technique provides a label-free method that may rapidly screen thousands of proteins per assay. Keywords: multi-analyte assays, Interferometry, bio-sensors, proteomics

1. INTRODUCTION

The need for high though-put multi-analyte biosensors in the emerging areas of genomics and proteomics cannot be overemphasized. Some of the issues involved in the design of such biosensors are the high multiplicity of the biological binding sites that are needed for a complete assay, which can number many thousands depending on the application. Even when these biological receptors are exposed to the target molecules only a few may bind and it is important to identify these in as short a time as possible. In such applications, it is also important to screen for many target molecules at the same time. The techniques to perform such multi-analyte immunoassays have been drawn extensively from the technology of gene chips. These gene chips involve complex biochemistry performed using micro-fluidics and structured chips. They also rely on the ability to print micro-arrays with many receptors on a single chip [1]. Imaging detection of multi-analyte micro array immunoassay has been performed [2] and micro spot assays have been developed. The approaches described above rely on the optical properties of the immunoassay [3]. The throughput of such sensors is limited by the speed of the optical read-out even in the case of micro spot assays, which rely on thousands of separate spots of printed antibodies. In the case of micro spot assays, fluorescence of the tagged molecules is detected using confocal microscopes and sensitive photo-detectors. But the low intensity of fluorescent radiation requires long integration times. Some sensors have been developed on the basis of interferometry, as in the case of waveguides and grating couplers [4-8]. These approaches use static structures which prevent repetitetive scanning and thus lack the high-speed capabilities of the technique that we are going to describe here. We propose to use the well established technology of the * [email protected] Phone: 1-765-494-9203, Fax: 1-765-494-0706, Dept. of Physics, Purdue University, West Lafayette, IN, 47907. ** [email protected] Phone: 1-765-494-4033, Fax: 1-765-494-0706, Dept. of Physics, Purdue University, West Lafayette, IN, 47907.

Microarrays and Combinatorial Technologies for Biomedical Applications: Design, Fabrication,and Analysis, Dan V. Nicolau, Ramesh Raghavachari, Editors, Proceedings of SPIE Vol. 4966

(2003) © 2003 SPIE · 1605-7422/03/$15.00

58

optical CD to develop ultra high-throughput immunological assays performed on a single spinning disc with sampling rates as high as a MegaSamples/sec and sampling potentially as many as a billion separate micro-diffraction elements on a single Bio-CD. This would represent an improvement, by many orders of magnitude, in the throughput of immunological assays. Commercially available compact discs (CD’s) consist of concentrically arranged tracks of pits that are half a micron wide and separated by about 1.6 microns. The information on a CD is read by focusing a laser spot onto these pits and observing the far-field diffraction. The depth of the pits, is a quarter of the wavelength of the laser used for read-out. This depth difference between the pit and the land results in a phase difference of p between the light reflected off the pit and the light reflected off the land and leads to destructive interference at the detector, placed along the optic axis at the far-field plane. Thus each pit acts as a micro-interferometer, and an optical CD can be viewed as a device with a billion such micro-interferometers. Interferometers have a universal response curve with a half-intensity point defined by quadrature, when the signal and the reference waves are out of phase by ninety degrees. Maximum linear sensitivity to small optical perturbations is achieved by operating the interferometer near quadrature. Shot noise-limited detection of optical path length changes down to a billionth-lambda is achievable under these conditions. In this paper we describe the fabrication of microstructures on silicon that act as microscopic interferometers, as on an optical CD, but operating in quadrature with maximum linear sensitivity, unlike an optical CD.

2. FABRICATION AND EXPERIMENTAL SETUP The Bio-CD is fabricated by evaporating gold lines on a 3” silicon wafer in a radial pattern as shown in Fig 1. There are 1024 gold lines arranged in a radial fashion like spokes. These gold spokes are 20 microns wide and are evaporated to a thickness of 79.1 nm, putting them in the quadrature condition for read-out with a 633 nm He-Ne laser.

Fig.1. Schematic of the arrangement of the gold lines on the Bio-CD

The primary advantage of this design over a design based on the pits of an optical CD is that the gold ridges provide a simple means to immobilize of macromolecules such as antibodies on the interferometric structures. By using alkanethiols as a bridge between the gold and the macromolecule, we immobilize antibodies on the gold region alone, thus making a selective pattern of antibodies on the wafer. By using poly di-methoxy silane (PDMS) stamps we immobilize antibodies in the gold pattern only in a selected annuli of interest. Thus a single wafer can have regions with and without antibodies, providing a means of control as we test for antigen binding events in the regions with antibody.

Fig. 2 shows the fluorescence image from Bovine Serum Albumin (BSA) immobilized on a Bio-CD. The BSA molecules were tagged with fluorescein. In the procedure for immobilizaztion, the wafer was treated with a 10mM solution of Hexa decane thiol in ethanol resulting in the deposition of a thiol layer on the gold spokes. The thiolated

Proc. of SPIE Vol. 4966 59

wafer was then treated with fluorescein-conjugated Bovine Serum Albumin (BSA) (200 µg/ml), resulting in the immoblisation of BSA on the gold spokes on the wafer. As the fluorescence image shows, the pattern of BSA is very well defined and demonstrates the effectiveness of the gold-ridge approach in immobilizing antibodies.

Fig.2. Fluorescence image of immobilized BSA on the gold spokes of the Bio-CD. 2.1 EXPERIMENTAL SETUP

A schematic of the experimental setup is shown in Fig 3. The laser beam is focussed by a 4x objective on to the gold microstrips on the silicon wafer placed on a photoresist spinner. A 10 cm focal length lens was used to perform a Fourier transform of the image of the gold microstrip to obtain the far-field signal at the back focal plane. A photodetector was placed at the Fourier plane to monitor the far-field signals. A 25 micron pinhole was used to aperture the diffraction peak. The signal from the photodetector was sent to a preamplifier and then to an oscillioscope.

Fig. 3 Schematic of the experimental setup One of the significant goals of the Bio-CD project is to achieve high sampling rates of the order of 100kSamp/sec. To achieve this, the Bio-CD must be mounted on a spinning platform capable of achieving speeds of the order of 6000 rpm or more, and detection must be done while the device is spinning. We use a conventional photo-resist spinner (Model

He-Ne Laser

f = 5 cm f = 20 cm

f = 10 cm fourier transform

lens

Spinner

Bio-CD

25 micron pinhole

Pre-amplifierOscilloscope

60 Proc. of SPIE Vol. 4966

P6204, SCS) to achieve this goal. Fig. 4 shows a Bio-CD mounted on the photoresist spinner and the objective used for read-out, the spinner is capable of rotating up-to 7500 rpm, which represents a sampling rate of up-to 128kSamples/sec.

Fig. 4. Bio-CD mounted on a photoresist spinner.

3. EXPERIMENTAL METHODOLOGY AND RESULTS The beam is focused to a waist diameter of approximately 80 microns. As the Bio-CD rotates at 6000 rpm, quadrature is achieved when the beam symmetrically straddles the gold spoke, at all other times, there is an off-quadrature condition. The quadrature condition results in a peak in the signal and thus the raw signal is composed of several peaks from the different spokes. When a macromolecule is immobilized on the spoke, it modulates the intensity detected at quadrature thus altering the peaks. This modulation is a function of the amplitude and phase effects caused by the immobilized bio-layer. Fig 5 shows the raw signal captured by the oscilloscope in the presence of

Fig. 5 Modulation of the far-field signal in the presence of an immobilized macro-molecule film. BSA. We can clearly see the decrease in the peak-valley difference of the signal in the presence of BSA. As the spinner rotates at 6000 rpm, the chuck wobble causes slight optical misalignments resulting in DC shifts in the signal. However, this is not a fundamental problem and can be addressed very simply by the use of high-stability spinners. We extract the

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

No-AbAb

Sig

nal (

V)

Spokes

Quadrature

flat


quadrature peaks from the raw data with a computer algorithm and measure the modulation of the far-field signal to detect the binding of antigens. 3.1 DETECTING ANTIGENS We conducted an experiment to test the ability of the bio-CD to detect the presence of antigens in an analyte. We immobilized part of the bio-CD with anti-mouse IgG using the procedure described earlier. We exposed part of the Bio-CD to a solution of the specific mouse antigen, namely mouse IgG. Fig 6 shows the sequence of the Bio-CD with the different species.

Fig. 6. Schematic of the Bio-CD showing the position of the antibody and the antigen Fig. 7 illustrates the experimental results obtained. We see a difference in the signal level from the region of the wafer containing the antibody to the region with only thiolated gold indicating the presence of the antibody in that region. After we expose the Bio-CD to the antigen we see that there is a further drop in the signal from the region containing the antibody, demonstrating the binding of the specific antigen to the immobilized antibody.

Fig. 7. Detection of the binding of antigen, (mouse IgG)

In the figure, ‘Ab’ refers to anti-mouse IgG and ‘Ag’ refers to mouse IgG. The modulation caused by the binding of the antigens is about 25 %. This result clearly demonstrates the success of our interferometric detection technique.

3.2 NON-SPECIFIC BINDING

An important performance parameter for bio chips, is the frequency of non-specific binding events. Non-specific binding occurs when the signal level shows a binding event for antigens not specific to the antibody that is immobilized on the

0

0.02

0.04

0.06

0.08

0.1

0.12

0 50 100 150 200

NoAb-MeanAb-MeanAb+Ag-Mean

Vol

tage

(V

)

Spokes


device. We immobilized the Bio-CD with anti-mouse IgG and exposed part of it to mouse antigen, which is the specific antigen in this case, and the remaining part to BSA which was considered as a non-specific species, in the sequence shown in Fig 8.

Fig. 8 Schematic of the Bio-CD showing the position of the antibody and the specific and non-specific antigens, mouse IgG is the specific antigen and BSA is the non-specific antigen.

Fig. 9 shows the results of the non-specific binding experiments. The expected change in the signal from the region that was exposed to mouse IgG, the specific antigen, indicates a binding event. However, the region exposed to BSA, the non-specific antigen, shows no appreciable signal change, verifying a low non-specific binding. In the figure, ‘Ab’ refers to anti-mouse IgG and ‘Ag’ refers to mouse IgG.

Fig. 9 Results of the non-specific binding experiments showing small non-specific binding signal modulation.

4.CONCLUSION AND FUTURE WORK To summarize, we have successfully demonstrated the effectiveness of an interferometric approach for rapid and simultaneous multi-analyte detection. This approach would also be useful in applications other than immunoassays.

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0 50 100 150

AbAb+BSAAb+AgNo-Ab

Vol

tage

(V

)

Spokes


Potential applications may include proteomics, genomics, drug screening etc. We are studying performance parameters such as signal variability and novel methods of interferometric sensing including adaptive homodyne detection. This would enable us to determine the degree to which amplitude and phase trade off as single polarizable molecules coalesce into macroscopic dielectric films.

5.REFERENCES [1] Gavin MacBeath and Stuart L. Schreiber, "Printing Proteins as Microarrays for High-Throughput Function Determination," Science 289 (5485), 1760-1763 (2000). [2] J. W. Silzel, B. Cercek, C. Dodson, T. Tsay, and R. J. Obremski, "Mass-sensing, multianalyte microarray immunoassay with imaging detection," Clin. Chem. 44 (9), 2036-2043 (1998). [3] R. M. Ostroff, D. Maul, G. R. Bogart, S. Yang, J. Christian, D. Hopkins, D. Clark, B. Trotter, and G. Moddel, "Fixed polarizer ellipsometry for simple and sensitive detection of thin films generated by specific molecular interactions: applications in immunoassays and DNA sequence detection," Clin. Chem. 44 (9), 2031-2035 (1998). [4] H. Gao, M. Sanger, R. Luginbuhl, and H. Sigrist, "Immunosensing with photo-immobilized immunoreagents on planar optical wave guides," Biosensors and Bioelectronics 10, 317-328 (1995). [5] B. Maisenholder, H. P. Zappe, R. E. Kunz, P. Riel, M. Moser, and J. Edlinger, "A GaAs/AlGaAs-based refractometer platform for integrated optical sensing applications," Sensors and Actuators B 38-39, 324-329 (1997). [6] R. E. Kunz, "Miniature integrated optical modules for chemical and biochemical sensing," Sensors and Actuators B 38-39, 13-28 (1997). [7] J. Dübendorfer and R. E. Kunz, "Reference pads for miniature integrated optical sensors," Sensors and Actuators B 38-39, 116-121 (1997). [8] A. Brecht and G. Gauglitz, "recent developments in optical transducers for chemical or biochemical applications," Sensors and Actuators B 38-39, 1-7 (1997).


biocd mutianalyte detection

Documents