Article pubs.acs.org/ac
Tuning the Bacterial Detection Sensitivity of Nanostructured Microelectrodes Jagotamoy Das† and Shana O. Kelley*,†,‡ †
Department of Pharmaceutical Sciences, Leslie Dan Faculty of Pharmacy and ‡Department of Biochemistry, Faculty of Medicine, University of Toronto, Toronto, Canada ABSTRACT: Fast, sensitive nucleic acid sensors that enable direct detection of bacteria and diagnosis of infectious disease would offer significant advantages over existing approaches that employ enzymatic amplification of nucleic acids. We have developed chip-based microelectrodes that are highly effective for bacterial detection and have shown that they can capture and permit the analysis of large slow moving mRNA targets. Here, we explore new approaches to tune their analytical sensitivity and investigate the effect of sensor size, material composition, and probe density on the electrochemical signals obtained in the presence of bacteria. Sensor size can be varied from 10 to 100 μm, and this parameter can change detection limits obtained by a factor of 100. Changing the surface coating can also be used to tune sensitivity, with more nanostructured coatings yielding the most sensitive detectors. Moreover, we assessed performance of the sensors by tuning probe density. Varying the density of the immobilized probe had a dramatic effect on sensitivity, with sparse probe monolayers providing superior levels of performance. Overall, this study points to several factors that can be used to tune detection limits.
D
charges nucleic acids that accumulate at the sensor surface, and on the regeneration of electrochemically reduced Ru(II) by Fe(CN)63−.25 The use of probes with low levels of intrinsic change enhances the signal changes occurring upon target binding. This sensing system has been applied toward a variety of applications for the detection of cancer biomarkers and infectious pathogens.19−23,26−28 Coupled with an electrochemical method for cell lysis, this approach has shown to enable rapid, sample-to-answer bacterial detection and to enable detection of bacteria at clinically relevant concentrations.20,22,23,28 While clinically relevant performance for the NME platform has been demonstrated previously, the effect of a variety of important parameters remain untested. Systematic characterization of the NME sensing system and the demonstration of robustness and predictable behavior represents an important activity so this technology may advance closer to clinical use. The composition of the NMEs, the density of the probe monolayer, and the size of the sensors may all affect detection performance. Here, we carefully characterize the effect of these parameters. Using NMEs of a variety of sizes, we investigate the effect of different surface coatings and probe densities on the electrochemical response to crude lysates of Escherichia coli. The best detection limits are obtained with large sensors with high nanostructured surfaces and low probe densities. Probe density is a particularly important factor, as dense monolayers can decrease sensitivity significantly, but if made too sparse, detection sensitivity degrades. These studies highlight that
evelopment of nucleic acid sensors is of significant interest because of potential for genetic analysis, early diagnosis of diseases, food safety testing and environmental monitoring.1−4 The current gold standard for nucleic acids detection is the polymerase chain reaction (PCR), which amplifies target molecules until they are present in a sample at levels where they can be detected using fluorescence or other approaches.5−7 While PCR is very powerful because it can amplify very rare analytes, it is challenging to automate and can be prone to contamination. The development of microchipbased sensors as a substitute for PCR is very appealing because of the potential for straightforward integration into sample-toanswer testing devices. Systems based on microchannel resonators,8 microcantilevers,9 field effect transistors,10 and electrochemical detectors11−15 have been developed and show great promise for nucleic acids detection. However, few chipbased systems have demonstrated clinically relevant levels of sensitivity. We have developed a class of chip-based electrochemical sensors that are highly sensitive and have been proven to be able to detect the low levels of nucleic acids present in clinical samples.16−23 These nanostructured microelectrode (NME) sensors are produced by electrodeposition of noble metals into a template created by a passivation layer introduced on the surface of a silicon or glass chip patterned with gold leads. By producing large, three-dimensional sensors with nanostructured surfaces, the sensors are able to capture slow-moving mRNA molecules. It has been hypothesized that the presence of nanostructuring on the sensor surface enhances probe display and promotes binding of target molecules.24 These sensors are used in conjunction with an electrocatalytic reporter system that relies on the interaction of Ru(NH3)63+ with negatively © 2013 American Chemical Society
Received: April 24, 2013 Accepted: June 25, 2013 Published: June 25, 2013 7333
dx.doi.org/10.1021/ac401221f | Anal. Chem. 2013, 85, 7333−7338
Analytical Chemistry
Article
aureus (S. aureus) was obtained from ATCC. E. coli and S. aureus were grown in appropriate growth medium in an incubating shaker at 33 °C. Approximate quantification of bacteria was performed by measuring the OD at 600 nm with an Agilent 8453 UV−vis spectrometer. After growth to the desired population the growth media was replaced with 1× PBS. Lysis of bacteria was performed utilizing a Claremont BioSolutions OmniLyse rapid cell lysis kit. Functionalization and Hybridization Protocol. Sensors were functionalized with PNA probes and MCH. To generate different probe densities on Pd-coated Au sensors, they were incubated with a solution containing 50 nM of the PNA probe and 450 nM MCH (low 3), 100 nM of the PNA probe and 900 nM MCH (low 2), 1 μM of the PNA probe and 9 μM MCH (low 1), 1 μM of the PNA probe, 9 μM MCH, and 5 mM MgCl2 (med), or 9 μM of the PNA probe, 1 μM MCH, and 5 mM MgCl2 (high) for 30 min at room temperature. Au sensors and Au-coated Au sensors were functionalized by incubating with a solution containing 5 μM of the PNA probe, 15 μM MCH, and 50 mM MgCl2 for 30 min at room temperature. Chips were washed twice for 5 min with 1 x PBS buffer after probe deposition. After washing, sensors were incubated with different concentration of E. coli lysates and S. aureus lysate for 15 min at 37 °C followed by washing with 2 × 5 min with 1× PBS. After washing, DPV measurements were performed following probe deposition and sample hybridization in the electrocatalytic solution. Determination of Probe Densities. For measuring the surface density of PNA probe on electrodes, chronocoulometry was performed in 0.1× PBS in the presence and absence of 100 μM [Ru(NH3)6]3+ after purging it thoroughly with nitrogen gas with a pulse period of 250 ms and a pulse width of 0.7 V (from +0.20 to −0.50 V). The surface density was derived from the charge of the surface confined [Ru(NH3)6]3+, which was obtained from the difference between the intercepts in chronocoulograms in the absence and presence of [Ru(NH3)6]3+. Following the published protocol, the calculated surface densities of PNA probe with different deposition conditions were 1.8 × 1013 molecules/cm2 (low 1), 2.7 × 1013 molecules/cm2 (medium), and 8.8 × 1013 molecules/cm2 (high). The surface area of Pd-coated Au sensors was calculated by integrating the Pd oxide reduction peak area obtained from cyclic voltammogram in the 50 mM H2SO4. In the forward scan, a monolayer of chemisorbed oxygen is formed and then it is reduced in the reverse scan. The reduction charge per microscopic unit area has been experimentally determined as 424 μC/cm2.24 The surface area was calculated by integrating the reduction peak (0.39 V vs Ag/AgCl) to obtain the reduction charge, and dividing this by 424 μC/cm2. Electrochemical Analysis and Scanning Electron Microscopy (SEM). All electrochemical experiments were carried out using a Bioanalytical Systems Epsilon potentiostat with a three-electrode system featuring a Ag/AgCl reference electrode and a platinum wire auxiliary electrode. Electrochemical signals were measured in a 0.1× PBS containing 10 μM [Ru(NH3)6]Cl3 and 4 mM K3[Fe(CN)6]. DPV signals were obtained with a potential step of 5 mV, pulse amplitude of 50 mV, pulse width of 50 ms and a pulse period of 100 ms. Signal changes that corresponded to target hybridization were calculated with background-subtracted currents: ΔI% = (Iafter − Ibefore)/Ibefore × 100 (where Iafter = current after target hybridization and Ibefore = current before target hybridization,
several variables can tune the sensitivity of NME sensors, offering a means to use sensors for different applications with differing detection thresholds.
■
EXPERIMENTAL SECTION Materials. HAuCl4 solution, potassium ferricyanide (K3[Fe(CN) 6 ]), potassium ferrocyanide trihydrade (K 2 [Fe(CN)6·3H2O), mercaptohexanol (MCH) from Sigma Aldrich. ACS-grade acetone and isopropyl alcohol (IPA) were obtained from EMD (U.S.A.); 6 N hydrochloric acid was purchased from VWR (U.S.A.). Phosphate-buffered saline (PBS, pH 7.4, 1×) was obtained from Invitrogen. PNA monomers were obtained from Link Technologies (Lanarkshire, Scotland), Fmoc-Lys(Boc)-OH from Advanced ChemTech (Louisville, Kentucky), Knorr resin from NovaBioChem and HATU (O-(7-azabenzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate) from Protein Technologies, Inc. (Tucson, Arizona). Chip Fabrication. Chips were fabricated at Advanced Micro Systems (Ottawa, Canada). Silicon wafers (six inches (15.24 cm)) were passivated using a thick layer of thermally grown silicon dioxide. A 25 nm Ti layer was deposited on the silicon dioxide. A 350 nm Au layer was deposited on the chip using electron beam assisted Au evaporation. The Au film was patterned using a standard photolithography and lift-off process. A 5 nm Ti layer was deposited on the Au film. A 500 nm layer of insulating Si3N4 was deposited using chemical vapor deposition; 5 μm apertures were imprinted on the electrodes using standard photolithography and 0.4 mm ×2 mm bond pads were exposed using standard photolithography. Fabrication of Microelectrodes. Chips were cleaned by sonication in acetone for 5 min, rinsed with isopropyl alcohol, and DI water, and dried with a flow of nitrogen. Electrodeposition was performed at room temperature; 5 μm apertures on the fabricated electrodes were used as the working electrode and were contacted using the exposed bond pads. All three different sized Au sensors were made using a deposition solution containing 50 mM solution of HAuCl4 and 0.5 M HCl. The 10, 35, and 100 μm sensors Au structures were formed using DC potential amperometry at 0 mV for 10, 30, and 120 s, respectively. After washing with deionized water and drying, the Au sensors were coated with a thin layer of either Au or Pd to form nanostructures by replating in a solution of 20 mM HAuCl4 and 0.5 M HClO4 or 5 mM H2PdCl4 and 0.5 M HClO4, respectively. Au structures were replated by Au using DC potential amperometry at −400 mV for 3 (for 10 μm structure), 5 (for 35 μm structure), and 12 s (for 100 μm structure), or Au structures were replated by Pd using DC potential amperometry at −250 mV for 3 (for 10 μm structure), 5 (for 35 μm structure), and 12 s (for 100 μm structure). Synthesis and Purification of Peptide Nucleic Acid. Inhouse synthesis of peptide nucleic acid (PNA) probe was carried out using a Protein Technologies Prelude peptide synthesizer. The E. coli probe sequence specific to mRNA of E. coli was utilized for detection: NH2-Cys-Gly-Asp-ATC TGC TCT GTG GTG TAG TT-Asp-CONH2. After synthesis the probe was stringently purified by reverse phase high performance liquid chromatography. Probe sequences were quantified by measuring absorbance at 260 nm with a NanoDrop and excitation coefficients were calculated from http://www. panagene.com. Bacterial Samples and Lysis. Escherichia coli (E. coli) was acquired from Invitrogen (18265-017) and Staphlophyticus 7334
dx.doi.org/10.1021/ac401221f | Anal. Chem. 2013, 85, 7333−7338
Analytical Chemistry
Article
Figure 1. (A) Schematic of a microfabricated chip that possesses 20 sensors and 5 μm openings for the electrochemical deposition of sensors at the end of gold leads. (B) Sensors of variable sizes were electrodeposited on that 5 μm openings. The size of the sensor is controlled by varying the plating time as shown in the chronoamperometry traces at right. (C) Schematic representation of E. coli detection. After electrodeposition, sensors were modified by PNA probe (left) complementary to the E. coli gene. After target hybridization (middle), sensors were interrogated using the electrocatalytic reporter system (right). (D) Differential pulse voltammetry (DPV) showing signal increase observed in the presence of a lysate of E. coli.
RNA is bound to the sensor (Figure 1C). To detect this binding event, a previously developed [Ru(NH3)6]3+/[Fe(CN)6]3− catalytic reporter system25 is used to generate a signal that can be monitored by differential pulse voltammetry (DPV) (Figure 1D). This reporter strategy is effective because the Ru(III) ions are electrostatically attracted to the sensor surface, reduced to Ru(II), and the Fe(III) ions are present to regenerate Ru(III) and enable many electrons to flow to each ruthenium ion. High catalytic currents are observed only when mRNA is released from bacteria containing a target sequence complementary to the PNA probes, whereas the signal does not increase when the sensor is challenged with bacteria lysates that do not contain the target mRNA. Size Variation. Since the sizes of the templated sensors we electrodeposit on our chips can be varied, we explored how the sensitivity and detection limit of the E. coli lysate detection approach would be affected by sensor footprint. Three different Au structures were fabricated on chips using different deposition times. Sensors with diameters of 10, 35, and 100 μm were generated by varying the electrodeposition time. Gold electrodeposition curves of these Au sensors are shown in right of Figure 1B and SEM images are shown in Figure 2A. The variation in size is reflected in the electrochemical signals produced as measured by differential pulse voltammetry (DPV) (Figure 3A). As expected, current height and charging currents increased from with the size and surface areas of the NMEs. To evaluate the detection limits of the sensors formed on the three differently sized structures, we measured differential pulse voltammograms (DPVs) of catalytic solutions before and after incubation with different concentrations of E. coli lysates for 15 min. These structures were made of gold with no additional
that is, current with only probe). The SEM images were obtained using a Hitachi S-3400 SEM.
■
RESULTS AND DISCUSSION Production of NME Sensors. The NME sensors are produced using a silicon chip as a base. A schematic of the microelectronic chip used for this work, which features apertures for the production of 20 electrodes, is shown in Figure 1A. Using photolithographic patterning, an array of sensor leads and contacts is produced on the surface of a microelectronic chip. A layer of Si3N4 is then used to passivate the surface of the chip. To provide a template for the growth of electrodeposited sensors, photolithography is used to open 5 μm apertures in the Si3N4. The electrodeposition of Au is then used to deposit metal in these apertures to form sensors with different sizes by varying deposition time (Figure 1B). The growth process first fills the pore with metal, and then the sensor grows onto the surface of the chip to form a threedimentional structure that reaches into solution both vertically and horizontally until the deposition is ceased. Deposition time, potential and supporting electrolyte can be used to vary the size and surface morphology of the chip.16,26 The amount of time used for electrodepostion is the most straightforward way to control the size of the sensor. Sensors with approximate diameters of 10, 35, and 100 μm could be generated by using deposition times of 10, 30, and 120 s, respectively. To make the sensors specific for a particular bacterial target, a self-assembled monolayer of PNA probes is immobilized on the sensors. Mercaptohexanol is used as a coligand and spacer to prevent nonspecific binding and control probe density. Upon incubation with a lysate of bacteria, in this case E. coli, a specific 7335
dx.doi.org/10.1021/ac401221f | Anal. Chem. 2013, 85, 7333−7338
Analytical Chemistry
Article
Figure 2. SEM images of sensors with different footprints and surface coatings. (A) Au sensors. Electrodeposition of Au was performed using a voltage of 0 mV for 10, 30, and 120 s (from left to right respectively). (B) Au-coated Au sensors. After electrodeposition of Au sensors, these structures were coated with a thin layer of Au using a voltage of −400 mV for 3, 5, and 12 s (from left to right respectively). (C) Pd-coated Au sensors. After electrodeposition of Au sensors, these structures were coated with a thin layer of Pd using a voltage of −250 mV for 3, 5, and 12 s (from left to right respectively). (D) Cyclic voltammograms collected for Au only (Au), Au/Au, and Au/Pd 10 μm NMEs in 50 mM H2SO4.
DPVs measured with the different sensors were collected before and after incubation with different concentrations of E. coli lysates for 15 min. With lysates containing 200 cfu/μL E. coli lysates, significant current changes were observed for all the Au-coated and Pd-coated sensors (Figure 3B), irrespective of the sensor size. No signal change was observed when a similar level of S. aureus was used. For lysates containing 20 cfu/μL E. coli, significant signal changes were observed with the largest (100 μm) NMEs, but not the 10 or 35 μm Au/AuNMEs. However, the Au/Pd structures of all sizes successfully detected E. coli at this concentration. When lysates containing 2 cfu/μL were used, only the 100 μm Au/Au and Au/Pd sensors produced a detectable response. The enhancement in sensitivity observed for the Au/Au and Au/Pd sensors relative to the uncoated Au sensors likely relates to the display of the probes on the nanostructured surface and an increase in productive collisions with RNA molecules. 100 μm sensors are predicted to have many collisions within the 15 min incubation time, but these collisions may have different outcomes depending on the conformation of the PNA probes.
surface coating. A statistically significant change is current was only observed for 100 μm sensors with 200 cfu/μL E. coli lysates (Figure 3B), and no significant increases were observed with the smaller structures (35 and 10 μm). This difference indicates that at this concentration, not enough successful collisions with target molecules arise at small sensors with less surface area. No signal change was observed when a similar level of S. aureus was used. Effects of Coating Material. Previous work in our laboratory suggested that introducing nanostructures on the surface of a microelectrode increases sensitivity significantly.16,20 However, as seen in Figure 2, the Au sensors are substantially smooth on length scale 100 nm. To test whether the sensitivity improved with nanostructuring, we introduced a fine layer of gold or palladium nanoparticles using electrodeposition. Characteristic cyclic voltammograms in H2SO4 of these electrodes are shown in Figure 2D. The gold layer produced slightly larger features approaching 50 nm, while the use of palladium produced small nanostructures that measured 10−20 nm. 7336
dx.doi.org/10.1021/ac401221f | Anal. Chem. 2013, 85, 7333−7338
Analytical Chemistry
Article
Figure 3. (A) Differential pulse voltammograms for 10, 35, and 100 μm NMEs after probe deposition with no coating (Au), a nanostructured Au coating (Au/Au), and a nanostructured Pd coating (Au/Pd). Scans were collected in solution containing Ru/Fe electrocatalytic reporter system. (B) Comparison of the sensitivities and detection limit of different sensors when challenged with E. coli lysates. The sensors were generated with 3 different sizes and with no coating (Au), a nanostructured Au coating (Au/Au), and a nanostructured Pd coating (Au/Pd). Sensors were challenged with with 2, 20, and 200 cfu/μL E. coli. A control trial was also run with 200 cfu/μL S. aureus (inset of rightmost plot).
Figure 4. (A) DPVs showing qualitative probe densities with different probe deposition conditions at 100 μm Pd-coated Au sensors. (B) Hybridization efficiency as a function of PNA probe coverage or different probe deposition conditions on the Pd- coated Au sensors with different footprints and 200 cfu/μL of E. coli lysates. Deposition conditions used to vary probe coverage were: (high) 9 μM PNA probe, 1 μM MCH, and 5 mM MgCl2, (med) 1 μM PNA probe, 9 μM MCH, and 5 mM MgCl2, or 1 μM PNA probe, 9 μM MCH (low). (C) Further investigation of effect of probe density on response to different levels of E. coli with 100 μm Pd-coated Au sensors. Conditions for high, med and low 1 are the same as for the center panel, and two additional conditions were added: 100 nM PNA probe, 900 nM MCH (low 2), and 50 nM PNA probe, 450 nM MCH (low 3). These sensors were challenged with different concentrations of E. coli lysates.
qualitatively using DPV (Figure 4A), and differences were also approximated using a method that quantitates Ru(III) binding using chronocoulometry.31 The surface areas of the sensors were measured by scanning in sulfuric acid and measuring the amount of oxide formed and stripped from the surface. This analysis yielded estimates of probe density for the three different deposition conditions ranging from 20 to 30 pmol/cm2 and up to 90 pmol/cm2. With increasing probe densities, hybridization efficiency decreased significantly (Figure 4B). At the lower probe densities, detection was successful with lysates containing 200 cfu/μL, while at the highest density, no significant signal changes were observed. It appears that significant probe crowding has a dramatic effect on detection sensitivity, and the performance of nanostructured surfaces is negatively affected by densely packed monolayers even though the
Probes displayed on nanostructured surfaces experience a deflection angle29 that creates more room for a target molecule to bind, thereby increasing the probability that a collision is converted to a binding event.24 While the nanostructuring created with Au is effective, the best results are obtained with Pd, which creates smaller nanostructures. These smaller features would permit a larger deflection angle to be created, lending support to the theory that this underlies more efficient complex formation. Effect of Probe Density. Probe densities can have a strong influence on hybridization efficiency.30 The nanostructured NMEs might be better more suited to the use of high probe densities, and so this was explored using three different sets of conditions. The ratio of PNA to mercaptohexanol was varied, and a cationic salt (MgCl2) was also used to modulate probe coverage. Differences in probe density were observed 7337
dx.doi.org/10.1021/ac401221f | Anal. Chem. 2013, 85, 7333−7338
Analytical Chemistry
■
deflection-promoting nanostructures should allow some room to be created around individual probe strands.24 We investigated whether there was a limit to how sparse a monolayer could be while still providing high levels of performance, tested two additional sets of probe deposition conditions, and also performed titrations of E. coli lysates (Figure 4C). Similar results were obtained with high and medium surface coverages, with lowered detection limits observed. Interestingly, when probe density was lowered below the first level tested, no further enhancements to the signal were observed, and when the density was lowered further, the sensitivity decreased further. This may reflect that there is an optimal probe spacing that facilitates efficient target binding, where steric effects are overcome but proximal probes can help recruit a target molecule and promote the formation of a bound complex. While a similar effect was observed previously with peptide-based probes,32 this is the first study to document this type of behavior with nucleic acids probes.
REFERENCES
(1) Audrey Sassolas, A.; Leca-Bouvier, B. D; Blum, L. J. Chem. Rev. 2008, 108, 109−139. (2) Erdem, A.; Papakonstantinou, P.; Murphy, H. Anal. Chem. 2006, 78, 6656−6659. (3) Drummond, T. G.; M., G.; Barton, J. K. Nat. Biotechnol. 2003, 21, 1192−1199. (4) Terry, L. A; White, S. F.; Tingwell, L. J. J. Agric. Food Chem. 2005, 53, 1309−1316. (5) Saiki, R. K.; Scharf, S.; Faloona, F.; Mullis, K. B.; Horn, G. T.; Erlich, H. A.; Arnheim, N. Science 1985, 230, 1350−1354. (6) Kopp, M. U.; de Mello, A. J.; Manz, A. Science 1998, 280, 1046− 1048. (7) Clinical and Laboratories Standards Institute. CLSI document M100-S21; Clinical and Laboratories Standards Institute, PA, 2011. (8) Burg, T. P.; Godin, M.; Knudsen, S. M.; Shen, W.; Carlson, G.; Foster, J. S.; Babcock, K.; Manalis, S. R. Nature 2007, 446, 1066−1069. (9) Fritz, J.; Baller, M. K.; Lang, H. P.; Rothuizen, H.; Vettiger, P.; Meyer, E.; Guntherodt, H.-J.; Gerber, C.; Gimzewski, J. K. Science 2000, 288, 316−318. (10) Tian, B.; Cohen-Karni, T.; Qing, Q.; Duan, X.; Xie, P.; Lieber, C. M. Science 2010, 13, 830−834. (11) Wang, J. Chem.Eur. J. 1999, 5, 1681−1685. (12) Xiao, Y.; Lubin, A. A.; Baker, B. R.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 16677−16680. (13) Das, J.; Yang, H. J. Phys. Chem. C 2009, 113, 6093−6099. (14) Das, J.; Kim, H.; Jo, K.; Park, K. H.; Jon, S.; Lee, K.; Yang, H. Chem. Commun. 2009, 6394−6396. (15) Das, J.; Lee, J-A; Yang, H. Langmuir 2010, 26, 6804−6808. (16) Soleymani, L.; Fang, Z.; Sargent, E. H.; Kelley, S. O. Nat. Nanotechnol. 2009, 4, 844−848. (17) Soleymani, L.; Fang, Z.; Sun, X.; Yang, H.; Taft, B. J.; Sargent, E. H.; Kelley, S. O. Angew. Chem., Int. Ed. Engl. 2009, 48, 8457−8460. (18) Yang, H.; Hui, A.; Pampalakis, G.; Soleymani, L.; Liu, F.-F.; Sargent, E. H.; Kelley, S. O. Angew. Chem., Int. Ed. Engl. 2009, 48, 8461−8464. (19) Fang, Z.; Soleymani, L.; Pampalakis, G.; Yoshimoto, M.; Squire, J. A.; Sargent, E. H.; Kelley, S. O. ACS Nano 2009, 3, 3207−3213. (20) Soleymani, L.; Fang, Z.; Lam, B.; Bin, X.; Vasilyeva, E.; Ross, A.; Sargent, E. H.; Kelley, S. O. ACS Nano 2011, 5, 3360−3366. (21) Vasilyeva, E.; Fang, Z.; Minden, M.; Sargent, E. H.; Kelley, S. O. Angew. Chem., Int. Ed. 2011, 50, 4137−4141. (22) Lam, B.; Fang, Z.; Sargent, E. H.; Kelley, S. O. Anal. Chem. 2012, 84, 21−25. (23) Das, J.; Cederquist, K. B.; Zaragoza, A. A.; Lee, P. E.; Sargent, E. H.; Kelley, S. O. Nat. Chem. 2012, 4, 642−648. (24) Bin, X.; Sargent, E. H.; Kelley, S. O. Anal. Chem. 2010, 82, 5928−5931. (25) Lapierre, M. A.; O’Keefe, M. M.; Taft, B. J.; Kelley, S. O. Anal. Chem. 2003, 75, 6327−6333. (26) Das, J.; Kelley, S. O. Anal. Chem. 2011, 83, 1167−1172. (27) Ivanov, I.; Stojcic, J.; Stanimirovic, A.; Sargent, E.; Nam, R. K.; Kelley, S. O. Anal. Chem. 2013, 85, 398−403. (28) Lam, B.; Holmes, R. D.; Das, J.; Poudineh, M.; Sage, A.; Sargent, E. H.; Kelley, S. O. Lab Chip 2013, DOI: 10.1039/C3LC41416G. (29) Hill, H. D.; Millstone, J. E.; Banholzer, M. J.; Mirkin, C. A. ACS Nano 2009, 3, 418−424. (30) Peterson, A. W.; Heaton, R. J.; Georgiadis, R. M. Nucleic Acids Res. 2001, 29, 5163−5168. (31) Steel, A. B.; Herne, T. M.; Tarlov, M. J. Anal. Chem. 1998, 70, 4670−4677. (32) Bamdad, C. Biophys. J. 1998, 75, 1989−1996.
■
CONCLUSIONS We have tuned the detection limits of chip-based NME sensors using sensor footprints, different nanostructured coating materials, and probe densities. The size of an NME sensor affects sensitivity by controlling how many collisions will occur with a target molecule. The presence of a nanostructured Pd or Au coating enhanced the productivity of collisions, likely through enhancing display of the probes. The density of the probe monolayer immobilized on the NME sensors was shown to have a dramatic effect on detection sensitivity, with very high or very low coverages degrading sensitivity significantly. This study points to these factors as ones that can be used to tune sensitivity across a range of concentrations. The fact that different sensitivities can be realized with differently tuned sensors presents the possibility that an array of sensors could be used to perform quantitation using an approach different from the typical strategy of linking current magnitude to target concentration. Rather than putting the currents through an algorithm for quantitation, binary information about whether a sensor was “turned on” (i.e., currents rise over a certain threshold) could be used to determine bacterial load. This would be a useful approach in devices where limited signal processing can be done (e.g., a dipstick). In relationship to the prior work done on the NME sensing system, this study presents important evidence that these sensors are robust, reproducible, and that they can systematically be tuned. The role of sensor size and probe density in detection sensitivity illustrate general paradigms that could also be applied to other chip-based approaches.
■
Article
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This research was sponsored by the Defense Advanced Research Projects Agency through the Autonomous Diagnostics to Enable Prevention and Therapeutics: Diagnostics on DemandPoint-of-Care (ADEPT:DxOD−POC) program and by the Ontario Research Fund. 7338
dx.doi.org/10.1021/ac401221f | Anal. Chem. 2013, 85, 7333−7338
