© Copyright 2006 American Chemical Society
AUGUST 29, 2006 VOLUME 22, NUMBER 18
Letters Lateral Spread of an Amplification Signal Using an Enzymatic System on a Conductive Surface Melissa S. Hasenbank,* Elain Fu, and Paul Yager UniVersity of Washington, Department of Bioengineering, Box 355061, Seattle, Washington 98195 ReceiVed May 8, 2006. In Final Form: June 16, 2006 In this letter, we report a novel signal amplification phenomenon that rapidly and dramatically increases both the magnitude and the lateral extent of the original signal. This phenomenon utilizes an enzyme immobilized on a conductive surface to generate amplified signals at locations remote from the original site of enzyme activity. The result is demonstrated on a microfluidic platform using the established precipitating enzyme-substrate system of horseradish peroxidase (HRP) and 3,3′,5,5′-tetramethylbenzidine (TMB) on a surface plasmon resonance (SPR) imaging system.
Amplification of binding events remains the primary method of enhancing the sensitivity of biosensors. Signal amplification is of particular importance for detection methods with limited inherent sensitivity and for applications in which small sample volumes restrict the total number of analyte molecules present, such as microfluidic chemical sensors in diagnostic lab-on-achip systems. The transduction of binding events into a measurable signal requires that the change in some property at the site of binding be substantial. To make the local property change as large as possible, a number of signal amplification schemes have been investigated, including enzyme precipitation.1-5 As indicated in Figure 1a, this change in property is typically restricted to the immediate area in which the binding event and subsequent amplification reaction occurs. In contrast, we describe a novel phenomenon that is capable of rapidly increasing both the magnitude and the lateral extent * To whom correspondence should be addressed. E-mail: hasenm@ u.washington.edu. (1) Alfonta, L.; Willner, I.; Throckmorton, D. J.; Singh, A. K. Anal. Chem. 2001, 73, 5287-5295. (2) Kim, M.-G.; Shin, Y.-B.; Jung, J.-M.; Ro, H.-S.; Chung, B. H. J. Immunol. Methods 2005, 297, 125-132. (3) Lyon, L. A.; Musick, M. D.; Natan, M. J. Anal. Chem. 1998, 70, 51775183. (4) Speijer, H.; Laterveer-Vreeswijk, R. H.; Glatz, J. F. C.; Nieuwenhuizen, W.; Hermens, W. T. Anal. Biochem. 2004, 326, 257-261. (5) Wink, T.; van Zuilen, S. J.; Bult, A.; van Bennekom, W. P. Anal. Chem. 1998, 70, 827-832.
of the original enzyme-mediated signal. A small number of binding events can radically change the properties of a large sensor area, as indicated in the schematic of Figure 1b. This result is demonstrated on a gold-coated surface plasmon resonance (SPR) sensor surface, which has been selectively patterned with the enzyme horseradish peroxidase (HRP). The labeling solution containing 3,3′,5,5′-tetramethylbenzidine (TMB), obtained from US Biological, is then introduced to initiate the precipitate-forming signal amplification reaction, as illustrated in the set of SPR images before and after amplification in Figure 1c. Note that, in contrast to the traditional amplification format, this lateral signal amplification is not restricted to the sites of enzyme activity. A simple microfluidic SPR sensor incorporating an isolated gold region was utilized to further demonstrate this novel signal amplification phenomenon. Soda lime glass microscope slides were selectively patterned with 1 nm chromium for adhesion and 45 nm gold by applying a shadow mask during deposition via electron beam evaporation (Washington Technology Center Microfabrication Lab). This custom mask, designed to generate an electrically isolated “island” of gold within the main goldcoated region, was cut from dicing tape using a CO2 laser cutting system. The gold-coated slides were cleaned in a solution of deionized (DI) water, NH4OH, and 30% H2O2 in a ratio of 5:1:1 and stored in DI water until further use. HRP was then patterned onto the gold-coated slides using a piezoelectric inkjet printing system from MicroFab Technologies,
10.1021/la061289m CCC: $33.50 © 2006 American Chemical Society Published on Web 07/28/2006
7452 Langmuir, Vol. 22, No. 18, 2006
Letters
Figure 1. (a) Traditional signal amplification involving enzyme precipitation. (b) Lateral signal amplification demonstrated in this letter. The primary distinction between (a) and (b) involves immobilization of an enzyme directly on a conductive surface (gold), which enables the enzyme precipitation reaction to spread laterally and provide significant signal amplification over a large area. The lateral signal amplification scheme is demonstrated in the set of SPR images in (c). Here, the addition of the TMB labeling solution to a gold-coated SPR sensor surface containing a patch of immobilized enzyme causes significant precipitate formation (corresponding to the light regions of the image) that is not restricted to the patch of immobilized enzyme, i.e., the site of enzyme activity.
Figure 2. SPR images before (a) and after (b) the addition of the TMB labeling solution and the subsequent rinse with PBS. The outlined regions in (a) denote the printed enzyme patches, while the outlined regions in (b) indicate the location of the analyzed ROIs. The rightmost ROI is adjacent to the printed enzyme; the middle ROI is upstream of the enzyme in the electrically isolated region; and the leftmost ROI is located further upstream of the printed enzyme patches. Following the addition of the TMB labeling solution to the microchannel, the electrically isolated region of the gold substrate remained free of precipitate. In contrast, precipitate formed in regions devoid of enzyme that were electrically connected to the region containing enzyme.
Inc. Specifically, a 0.1 mg/mL solution of HRP was deposited in discrete rectangular regions using the instrument’s drop-ondemand mode, with individual droplet volumes on the order of 50 pL. Three arrays of approximate dimensions 320 × 480 µm were generated, with a center-to-center droplet spacing of 40 µm. Based on these values, it was estimated that more than one monolayer of HRP was deposited onto the gold surface. As depicted in Figure 2a, the three HRP patterns were deposited downstream of the isolated gold region. Layers of plain and adhesive-coated Mylar were then adhered to the slide, such that the HRP-patterned gold surface formed the bottom layer of a microfluidic flow cell. Approximate channel dimensions were 100 µm × 3.2 mm × 55 mm. The microfluidic
flow cell was installed in a custom manifold and mounted on a custom-built SPR microscope that operates by wavelength interrogation in the near-infrared at a fixed incident angle of approximately 65°.6,7 For each stage of the experiment, the imaging wavelength was set to a position near the base of the linear region on the SPR curve for the initial sample, between 850 and 900 nm, such that an increase in refractive index (e.g., due to precipitate formation) produced an increase in intensity. The linear range of the SPR instrument used in this work was measured to be approximately 20% change in reflectivity at an imaging wavelength of 855 nm (data not shown). The gold surface was then rinsed from left to right with an excess of phosphate buffered saline (PBS) using positive displacement syringe pumps at a volumetric flow rate of 10 µL/s to remove any excess or loosely bound enzyme molecules. A Reynolds number of approximately 20 (laminar flow conditions predominated) and a Peclet number of greater than 10 000 (diffusion of HRP was negligible) indicated that the transport of excess enzyme was restricted to regions downstream of the original HRP patterns, i.e., no enzyme was present upstream of or lateral to the original HRP patterns. After this initial rinse, approximately 250 µL of the TMB labeling solution was introduced into the entire microchannel at a volumetric flow rate of 1 µL/s. The TMB labeling solution was then rinsed from the channel with an excess of PBS at the same flow rate. Sequences of SPR images were captured every 1 or 2 s throughout the PBS rinse and TMB precipitate development. Representative SPR images of the sensor surface before and after the addition of the TMB labeling solution and the subsequent rinse with PBS are shown in Figure 2. The outlined regions in Figure 2a indicate the general location of the three rectangular patterned enzyme regions downstream of the electrically isolated gold patch. The outlined regions in Figure 2b indicate the location of the analyzed regions of interest (ROIs). The addition of TMB to the microchannel resulted in precipitate formation in regions of the surface that contained no enzyme but were electrically connected to the regions that contained enzyme (e.g., regions several millimeters upstream of the HRP patches). In contrast, there was no detectable precipitate formation in regions that did not contain enzyme and were electrically isolated from the regions that contained enzyme. This result was verified visually, as the blue-colored precipitate was present on the gold surface both upstream and downstream of the patterned enzyme, but not in the electrically isolated gold region, i.e., in the same pattern suggested by the SPR images. (6) Fu, E.; Chinowsky, T.; Foley, J.; Weinstein, J.; Yager, P. ReV. Sci. Instrum. 2004, 75, 2300-2304. (7) Fu, E.; Foley, J.; Yager, P. ReV. Sci. Instrum. 2003, 74, 3182-3184.
Letters
Figure 3. Comparison of the lateral signal amplification kinetics, presented as the change in SPR reflectivity versus time, in the ROIs outlined in Figure 2b. The three ROIs correspond to regions upstream of and electrically connected to the regions of patterned enzyme (0); adjacent to and electrically connected to the regions of patterned enzyme (2); and upstream of and electrically isolated from the regions of patterned enzyme ([). The zero time point corresponds to the approximate arrival of TMB in the microchannel.
The three darker rectangular regions in Figure 2b correspond to the location of the patterned enzyme, where significantly less precipitate formation was observed. This result is hypothesized to be linked to the mechanism behind this novel signal amplification phenomenon and will be addressed in a future manuscript. To quantify this lateral signal amplification, several 15 × 15 pixel ROIs, both within the microchannel and within the adhesive (to serve as a static reference region), were selected for analysis. The average SPR signal, denoted by the percent reflectivity (%R), in each of these N × N pixel regions was then calculated and normalized using custom LabVIEW programs.8 The data in Figure 3 demonstrate the magnitude and time scale of this lateral signal amplification, where the change in SPR signal for the three ROIs highlighted in Figure 2b is plotted versus time. The zero time point corresponds to the approximate N TM dark TE (8) The SPR signal was calculated as % Rsample ) 1/N2 ∑i,j ((Ii,j - Ii,j )/(Ii,j dark θ TM TE dark - Ii,j ))/Ci,j , where Ii,j , Ii,j , and Ii,j denote the intensity of pixel i,j in the TM (transverse-magnetically polarized), TE (transverse-electrically polarized), and θ TM,w/oAu dark TE,w/oAu dark dark images, respectively. The factor Ci,j ) ((Ii,j - Ii,j )/(Ii,j - Ii,j )) is a correction for the polarization-dependent transmission of the filter at a tilt 6 angle of θ, and is specific to the operation of an instrument of our design. For a given sequence of images, an additional correction for intensity fluctuations in the system was performed using the static reference region in the image, % sample sample Rt,reference - (% Rreference - % Rreference ). corrected ) % Rt t 1
Langmuir, Vol. 22, No. 18, 2006 7453
arrival of TMB into the microchannel, and the changes in percent reflectivity with time are due to precipitate formation. For all electrically connected regions, including those lacking enzyme, the signal amplification was substantial, with increases in reflectivity on the order of 50% or greater. In contrast, the change in SPR intensity following TMB addition was negligible in the isolated gold region. Furthermore, the signal amplification was rapid, with the majority of the precipitate formation in local and remote regions completed within two to three minutes. In summary, we have demonstrated a novel and versatile signal amplification phenomenon that rapidly and dramatically enhances the magnitude and the lateral extent of the original enzymemediated signal. This phenomenon was demonstrated using SPR imaging and a well-known enzyme precipitation reaction on gold. This lateral signal amplification has been shown to produce significant amplification of the original enzyme-mediated signal, with changes in percent reflectivity approaching 70% in areas that ordinarily would have experienced a negligible signal change. (Note that this signal change is outside of the linear range of the instrument.) The total area of the sensor surface experiencing amplification was dramatically increased, with precipitate formation occurring in a region without enzyme that measured several square millimeters in area. The precipitate formation also occurred rapidly, with the majority of the signal change, even at remote sites, completed within two to three minutes. The only apparent requirements are that the region devoid of enzyme contains the precipitating substrate solution and is electrically connected to the original site containing enzyme. We are not aware of a physical limit of the extent to which the precipitate could spread. This novel signal amplification format has the potential to improve overall detection sensitivity by translating the signal due to binding events restricted to a small area into a largemagnitude signal spread over a much larger “reporter” region. In addition, the use of this lateral signal amplification format presents opportunities for significant savings in assay times and reagent costs. A demonstration of an assay based on this format as well as a complete description of the underlying mechanism will be provided in subsequent manuscripts. Acknowledgment. The authors thank Ken Hawkins for helpful discussions. Financial support was provided by NIDCR (Grant Numbers UO1-DE014971 and T32-DE07023). Microfabrication facilities were provided by the Washington Technology Center. LA061289M
