Gold Nanoparticle Labels and Heterogeneous ... - ACS Publications

Subscriber access provided by GUANGXI UNIV

Article

Gold Nanoparticle Labels and Heterogeneous Immunoassays: The Case for the Inverted Substrate Alexis C. Crawford, Colin C. Young, and Marc D. Porter Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b02011 • Publication Date (Web): 15 Jun 2018 Downloaded from http://pubs.acs.org on June 15, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Gold Nanoparticle Labels and Heterogeneous Immunoassays: The Case for the Inverted Substrate Alexis C. Crawford,a,b Colin C. Young,b,c and Marc D. Portera,b,c* a Department of Chemistry, bNano Institute of Utah, and cDepartment of Chemical Engineering, University of Utah, Salt Lake City, UT 84112 *(E-mail [email protected]) Abstract: This paper examines how the difference in the spatial orientation of the capture substrate influences the analytical sensitivity and limits of detection for immunoassays that use gold nanoparticle labels (AuNPs) and rely on diffusion in quiet solution in the antigen capture and labeling steps. Ideally, the accumulation of both reactants should follow a dependence governed by the rate in which diffusion delivers reactants to the capture surface. In other words, the accumulation of reactants should increase with the square root of the incubation time, i.e., t1/2. The work herein shows, however, that this expectation is only obeyed when the capture substrate is oriented to direct the gravity-induced sedimentation of the AuNP labels away from the substrate. Using an assay for human IgG, the results show that circumventing the sedimentation of the gold nanoparticle labels by substrate inversion enables the dependence of the labeling step on diffusion, reduces nonspecific label adsorption, and improves the estimated detection limit by ~30×. High-density maps of the signal across the two types of substrates also demonstrate that inversion in the labeling step results in a more uniform distribution of AuNP labels across the surface, which translates to a greater measurement reproducibility. These results, which are supported by model simulations via the Mason-Weaver sedimentationdiffusion equation, and their potential implications when using other nanoparticle labels and related materials in diagnostic tests and other applications, are briefly discussed.

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 27

2 Introduction Nanoparticles (NPs) have unique optical,1-5 thermal,6-8 and magnetic9-11 properties of importance across a range of fundamental processes and technologies.12-16 As recently reviewed,17-22 a number of laboratories have been examining approaches that incorporate gold nanoparticles (AuNPs) as labels in sandwich immunoassays to detect disease markers and infectious agents.23-27 As overviewed in Figure 1, the assays are read out by surface-enhanced Raman scattering (SERS).28-29 Procedurally, an assay for a target is carried out by first using an antibody-modified gold substrate to selectively capture an antigen from a biofluid or other type of liquid sample. This step is followed by exposure of the substrate to a colloidal suspension of extrinsic Raman labels (ERLs), which are AuNPs modified with a layer of Raman reporter molecules (RRMs) and a coating of a secondary antibody to selectively tag the captured marker. The last step indirectly identifies the presence of captured antigen from the SERS spectrum of the RRM, with the strength of its most intense vibrational mode used for antigen quantification. In most of our past work with this methodology,30 the transport of antigen and ERLs to the capture surface takes place in quiet solution, meaning that the capture of antigen and the subsequent tagging of captured antigen should conceptually be controlled by the diffusional delivery of each reactant to the capture substrate. This point of view is supported by a number of studies31-36 that have shown the rate of antigen capture is usually limited in quiet solution by the diffusion of reactant to the capture substrate and not by the rate of antigen-antibody binding. By extension, we have routinely assumed that the rate of the ERL labeling step is diffusion controlled and that the use of long labeling times quantitatively tags the captured antigen. If correct, this perspective should translate to a SERS response that tracks antigen binding, at least at low levels of accumulation, by increasing linearly with the square root of the time (t1/2) in


Page 3 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3 which a substrate is exposed to sample in the antigen capture step. However, our repeated attempts to experimentally confirm the existence of such a dependence have proven unsuccessful. This discrepancy pointed to the need to re-examine the validity of the underlying assumption that we have applied to the transport and accumulation of ERLs in the labeling step. This paper describes the results from investigations aimed at identifying factors beyond diffusion that could contribute to ERL accumulation. We specifically focused on examining the possible impact of ERL sedimentation (i.e., gravity-driven ERL transport and accumulation) and aggregation, by using the capture and tagging of human IgG by anti-IgG antibodies immobilized on a smooth gold substrate and on the ERLs. The interest in these possibilities reflects some of our own experimental observations when running these experiments as well as reports on the effects of gravitational forces in other types of nanoparticle processing.37-44 For these purposes, the colloidal stability of the ERL suspensions with respect to aggregation was examined with ζpotential, dynamic light scattering (DLS), and spectrophotometric measurements. The possible contribution of sedimentation, which proved to be the key confounder to the expected diffusion controlled transport and accumulation of ERLs, was measured by optically monitoring settling rates that were then analyzed by a finite-element diffusion-sedimentation model. Experiments were also carried out that assessed the impact of sedimentation on the assay by examining the differences in the signal generated when positioning the capture substrate in an upright or inverted spatial orientation in order to separate the contributions of diffusion from gravity-driven sedimentation on ERL transport and accumulation. These results, along with those that demonstrate the improvements in the analytical figures of merit (e.g., limit of detection (LOD)) in assays not affected by sedimentation, are described. The possible implications of these findings to other areas of work with nanoparticles and other materials potentially susceptible to



Page 4 of 27

4 the effects of sedimentation are also briefly discussed.

Theory NP Transport Model. In the absence of aggregation, the transport of NPs in stagnant solution is governed by the interplay between gravitational forces and Brownian motion, which is described by the Mason-Weaver sedimentation-diffusion equation.45 We developed a finite simulation model along the same lines by starting with the reduced continuity equation and by assuming that the change in the concentration of AuNPs throughout the suspension results from the competition between gravity-induced particle sedimentation and the tendency of diffusion to restore the uniformity of the particle distribution throughout the liquid. We can then write

߲‫ܥ‬ ߲‫ܬ‬ =− ߲‫ݐ‬ ߲‫ݖ‬

(1)

where J is the flux of the nanoparticles in the suspension relative to the ‫ ݖ‬direction (i.e., the direction aligned with the gravitational force), C is the concentration of AuNPs, and t is time. Equation 1 can be rewritten and discretized (Equation 2) along the height of the liquid added to a cuvette for spectroscopic measurements, ∆z, as:

߲‫ܥ‬௘௟௘௠௘௡௧ ‫ܬ‬௜,௢௨௧ − ‫ܬ‬௜,௜௡ =− ߲‫ݐ‬ Δ‫ݖ‬

(2)

where i indicates the position of the ith volume element and in and out refer to the flux of nanoparticles into a discretized volume element. The suspension height (H) was then discretized with a computational cell-centered method, with the flux calculated at both the upper and lower faces of each volume element. The flux at each face can be written as the combination of the sedimentation flux and the diffusive flux:40


Page 5 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


5

‫ܬ‬௜ = −‫ݒܥ‬௦௘ௗ − ‫ܦ‬

߲‫ܥ‬ ߲‫ݖ‬

(3)

In Equation 3, D and ‫ݒ‬௦௘ௗ are the diffusion coefficient of the nanoparticle and the sedimentation velocity, respectively. The diffusion coefficient can be calculated with the Stokes-Einstein equation (Equation 4) and ‫ݒ‬௦௘ௗ can be calculated using Stoke’s Law (Equation 5).

‫=ܦ‬

ߥ௦௘ௗ =

݇஻ ܶ 6ߨߟ‫ݎ‬

2݃(ߩ௣ − ߩ௠ )‫ ݎ‬ଶ 9ߟ

(4)

(5)

In Equation 4, kB is Boltzmann’s constant, T is the temperature, η is the dynamic viscosity of the solution, and r is the particle radius. In Equation 5, ݃ is the acceleration due to gravity, ߩ௣ is the particle density, and ߩ௠ is the media density. Importantly, the rate of particle settling increases with both the density and the square of the radius of the particle. The concentration at the two faces of the element was determined using a linear interpolation between the cell centers. The differential in the diffusion term of Equation 3 is approximated using a central difference formula:

∂‫ܥ ܥ‬௜ାଵ − ‫ܥ‬௜ = ߲‫ݖ‬ Δ‫ݖ‬

(6)

The model was solved with no-flux boundary conditions of ‫ܬ‬௭ୀ଴,ு = 0 , and the discretized conservation equation (Equation 2) for each volume element at each time step is then given by: ௡ ௡ ‫ܬ‬௜ାଵ/ଶ − ‫ܬ‬௜ିଵ/ଶ ߲‫ܥ‬௜௡ =− ߲‫ݐ‬ Δ‫ݖ‬

(7)

where n indicates the time step being evaluated. Equation 7 can be numerically solved using a commercial software package (MATLAB



Page 6 of 27

6 R2012b, MathWorks, Inc.) and all the necessary experimental parameters. We then varied the values for the terminal sedimentation velocity and the absolute suspension height in an iterative process to determine the best fit parameters to match the experimental measurements via a least squares approximation. The results from these model studies are presented in a subsequent section.

Experimental A detailed list of reagents and descriptions of experimental procedures are given in the Supplemental Information (SI). Note that all of the measurements with the as-received AuNPs were carried out directly in the vendor-supplied solution matrix and those with the ERLs were in phosphate buffered saline (pH 7.4) and 1% Tween 20 (PBST). NP-based SERS Immunoassay. The NP-based SERS immunoassay platform has been described previously.24, 30 Figure 1 details the preparation of the capture substrate and ERLs, along with the procedures for the capture and tagging of antigen. Procedurally, a smooth (~200 nm thick) gold substrate is first modified with a layer of capture human immunoglobulin G antibodies (α-IgG). The substrate is then exposed to an antigen solution of human immunoglobulin G (IgG), which is selectively extracted by the capture Abs. Next, a rinse step removes residual materials from the substrate. Finally, the substrate is exposed to a suspension of α-IgG-modified ERLs to tag captured antigens. Another rinse step removes excess labels. The ERLs were prepared with a 60-nm AuNP core modified with the RRM 5-5ʹdithiobis(succinimidyl-2-nitrobenzoate) (DSNB) and α-IgG. The ERLs produce a SERS signal proportional to the amount of antigen originally in the solution. The intensity of the symmetric nitro stretch, νs(NO2), at 1336 cm-1 of the DSNB-derived RRM was used for quantification.


Page 7 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


7 Dynamic Light Scattering (DLS) and ζ-potential Measurements. ζ-potentials and hydrodynamic diameters of the colloid suspensions were determined with a Malvern Zetasizer NanoZS. The instrument is equipped with a HeNe laser (632.8 nm) and avalanche photodetector. Each sample was analyzed at 25 °C with a DTS1060 folded capillary cell at a backscatter light collection angle of 90°. Five replicated measurements were carried out on both the as-received and completed ERLs each hour over a period of 24 h. The hydrodynamic diameters were determined by cumulants analysis and ζ-potentials were measured at an applied voltage oscillated between ±150 mV. The scattered data were collected and analyzed with resident NanoZS software. Spectrophotometric Measurements. Spectra (UV-Vis) of the particle suspensions were collected using a low-volume quartz cuvette [2-mm wide solution chamber and a 0.5 cm optical path length (0.500 ± 0.002 cm)] and a Cary UV-Vis-NIR 3000 (3 nm resolution). The cuvette was filled with 145 µL of a freshly prepared and evenly dispersed (vortexed) colloidal suspension, and then sealed with parafilm to limit evaporation. This volume of liquid resulted in a height of 1.45 cm. The cuvette was masked with flat black paint to create a small (0.2 × 0.2 cm) viewing window through the suspension (Figure 2a). The center of the window was positioned 0.2 cm below the top of the liquid in order to reduce the time required to complete an experiment. The spectra (440-580 nm) were collected every 15 min for ~80 h, the typical time required for the AuNPs to fully settle out of the observation window. The as-received AuNPs (dispersed in water with trace amounts of citrate, per manufacturer) were analyzed at the stock concentration of 2.1 x 1010 particles mL−1. The ERL suspensions were run at concentrations similar to the stock concentration. Raman Instrumentation. SERS spectra were collected with two different Raman



Page 8 of 27

8 spectrometers, both with 633 nm laser excitation. Measurements to quantify the dependence of ERL accumulation as a function of incubation time used a modified NanoRaman spectrometer (Concurrent Analytical),46 with ~30 µm diameter spot and a power of 3.0 mW at the sample. The response of each substrate was measured at 10 separate locations by repositioning the substrate under the laser with an x-y-z translation stage. High-density Raman spectral maps of the substrates were collected using a Thermo Scientific DXR Raman microscope. The instrument has a motorized sample stage, with a 1-µm step size for automated high-density data collection. Spectra were obtained at 50 µm steps across a sample at a laser power of 3 mW and 10 µm diameter spot size. All spectra were baseline corrected using a sixth-order polynomial algorithm.

Results and Discussion Preliminary Findings. The possible contributions of aggregation and sedimentation to the transport process were investigated for both as-received AuNPs and ERLs. We first examined the potential impact of aggregation by running hydrodynamic diameter and ζ-potential measurements on the suspensions for 24 h. These results are shown in Figure S1 and summarized in Table 1. The as-received AuNPs initially have a mean hydrodynamic diameter (z-average mean) of 59 nm, which agreed with the diameter of 58 nm determined by the spectrophotometric method of Haiss et al.47 and the strength of the extinction maximum at 535 nm. The Haiss method also yielded a AuNP concentration of 2.1 × 1010 particles mL−1. The diameter for the ERLs was 78 ± 0.1 nm. The diameters of both particles decrease slightly after 24 h, which is likely due to the sedimentation of larger particles (see below). The larger diameter of the ERLs compared to the as-received AuNPs is consistent with that expected for the overall thickness of


Page 9 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


9 the layered coating [DSNB, α-IgG (~10 nm diameter),48 and BSA (~3.5 nm)49] The changes in the surface composition of the two types of particles are also reflected in the measured ζ-potentials. There was no apparent change (−57 ± 2 mV) in the ζ-potential for the as-received AuNPs over 24 h. The temporal evolution of the ζ-potential for the ERLs is different. In this case, the ζ-potential changes from ~-2 mV to ~-10 mV over 24 h. This change may arise from a gradual restructuring of the protein corona50 on the ERLs that we have yet to detect spectroscopically. The consistency of the two measurements supports the stability of the ERLs as individually suspended particles with respect to a detectable level of aggregation. We can, therefore, rule out the role of that ERL aggregation its terms of being a confounding factor to the inability to experimentally observe a diffusion-controlled dependence of the SERS response with respect to incubation time. The second preliminary assessment tested for the potential role of gravity and its tendency to drive nanoparticle sedimentation. To investigate the role of sedimentation, a spectrophotometric experiment was devised to monitor settling rates. This study placed a small volume of the as-received AuNPs in a masked cuvette (Figure 2a) such that the liquid level was slightly above that of the mask; this shortens the time required to measure the complete depletion of the NuNPS in the viewing windows within a few days. Spectra were then acquired every 15 min for 80 h. These results are shown in 1 h increments after normalization to the spectrum at time zero in Figure 2b. As evident, the extinction maximum at 535 nm decreases over the course of the experiment as particles settle out of solution and reach undetectable levels in ~60 h. These results begin to point to a contribution from sedimentation to the accumulation of ERLs on the capture substrate. Sedimentation Modeling and Experimental Comparisons. AuNPs. We next tested the



Page 10 of 27

10 accuracy of the finite-element sedimentation model by calculating settling rates up to 80 h for AuNPs with diameters from 45 to 65 nm in 1 nm increments and then drawing comparisons to the normalized spectrophotometric data. For this, the following input parameters were used: dynamic viscosity (8.9 × 10−4 kg m−1 s−1), AuNP density (1.92 × 104 kg m−3),38 density of water (9.9 × 102 kg m−3), temperature (298.15 K), suspension height (1.45 × 10−2 m), and the height of the viewing window in the masked cuvette. The normalized results from the simulations and the experimental measurements at the extinction maximum are plotted in Figure 3a. The shapes of the plots are strongly similar and can be broken up into three distinct regions with respect to time. There is little change in the normalized signal at short times (0-5 h). This is followed by a time-period (~5 to ~30 h) in which the response undergoes a notable decrease and then a period in which the decay in response slows and begins to approach an immeasurable level. The near-zero change in signal early in the experiment reflects the time required for a detectable difference in particle concentration to reach the top of the observation window of the cuvette. During this time, the number of AuNPs moving out of the bottom of the window roughly matches that entering the top of the window. Once the depletion due to sedimentation reaches the observation window, the drop in signal corresponds to the gradual loss of AuNPs across the zone of observation. The last stage of the process shows a gradual slowing in the signal decrease, which can be viewed to include a low-level resupply of AuNPs in the observation window via diffusion. The accuracy of the model was assessed using least squares fits to the experimental data. The experimental settling curve for the AuNPs was best fit by the simulation for a 56 nm diameter AuNP. This prediction is in good agreement with the earlier DLS and spectrophotometric extinction measurements, validating the interpretative utility of the model


Page 11 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


11 predictions. ERLs. Having established effectiveness of the model, we then applied it to examine how gravity may contribute to the accumulation of ERLs suspensions on the capture substrate. Figure 3b shows the best fit plot from the simulations and the measurements at the ERL extinction maximum (538 nm) using UV-vis spectrophotometry. Both plots have shapes much like that the AuNPs. However, the ERLs appear to settle more slowly than the AuNPs. This observation runs counter to the idea that a higher surface charge goes hand-in-hand with higher colloidal stability,51 in that the ERLs have lower overall charge than the AuNPs (Figure S1). However, the stabilizing effect of protein coatings on nanoparticle systems has been observed by others,5253

and is likely due to a combination of steric effects and electrostatic repulsion. ERL Transport to Capture Substrate. To gage the extent in which sedimentation

affects the actual labeling process in the assay, the ERL incubation step was carried out with the capture substrate positioned either in an upright or in an inverted orientation. An upright incubation directs sedimentation toward the substrate, whereas an inverted incubation directs sedimentation away from the substrate. These experiments, therefore, placed the capture substrates in an upright position for 12 h exposures to a blank buffer or to buffer spiked with 1.0 or 5.0 ng mL-1 of IgG (n = 3). After rinsing, the substrates were placed in either an upright or inverted orientation for ERL incubation from 30 s up to 16 h, rinsed again, and read out. Figure 4 summarizes these results by plotting the average and variance of the signal strength of the υs(NO2) at 1336 cm-1 of the ERLs as a function of t in Figures 4a-b and as a function of t1/2 in Figures 4c-d (see Figure 1 for an example of the observed spectral response). If labeling is controlled by diffusion, the accumulation of ERLs should be linearly dependent on t1/2. As such, the data in Figures 4c-d were fit using a linear least squares algorithm to determine



Page 12 of 27

12 the slope and residuals from the fit. Deviations from linearity would then be indicative of contributions from sedimentation or other factors to the accumulation process. As evident in Figures 4a-b, the responses for both orientations increase with time. The responses for the substrates in the upright orientation, however, clearly tend to be larger than those in the inverted orientation. This difference is particularly apparent for the two sets of blank samples; the impact on detection will be examined shortly. The differences in the temporal dependencies of the responses for the two sets of substrates become more evident in the plots shown in Figures 4c-d. The responses for the IgGcontaining buffers more closely follow a t1/2 dependence with the inverted substrates than those in the upright orientation. The improvement in conformity to linearity is more readily drawn out by comparing the residuals of the linear least squares fits, as presented in Figure 4e for the upright substrates and in Figure 4f for the inverted substrates. We can again conclude that the accumulation of ERLs, and thus the tagging of captured antigen, more closely follows a diffusional delivery process when circumventing contributions from sedimentation by inverting the substrate when labeling. As examined in the next section, substrate inversion also improves the precision of the assay. Impact of Sedimentation on Detection. The most significant impact of sedimentation on the assay becomes evident by more closely examining the time dependence of the responses for the blank measurements. The responses from the blank samples for the two orientations, which are shown with an expanded scaling of the y-axis in Figure 5a, are comparable at short times (t < 30 min). At longer times, however, the blank response for upright substrates undergoes a much larger increase than the inverted substrates. The larger blank responses for the upright substrates show how sedimentation increases the overall level of accumulation (i.e., nonspecific


Page 13 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


13 adsorption) of ERLs beyond that from expected from a diffusion only scenario, which is more closely reflected by the temporal dependence of responses for the inverted substrates. The lower level of nonspecific ERL adsorption for the inverted substrates has a notable effect on the ability to detect IgG at low levels. This can be gauged by replotting, for example, the data in Figures 4a-b at 16-h incubations as shown by the dose-response plots in Figure 5b. Interestingly, the slopes of the plots are comparable, differing only by ~5%. This indicates that sedimentation does not have an observably detrimental impact on the analytical sensitivity of this assay. The more important observation comes from estimating the LOD, defined as the sample concentration that corresponds to the average blank signal plus three times its standard deviation, for the two difference substrate orientations. This yields an LOD of 0.02 ng mL-1 (~0.12 pM) when using an inverted substrate and 0.58 ng mL-1 (~4 pM) for the upright substrates. This analysis, while based on plots constructed with only three data points, shows that the inversion of the capture substrate during ERL incubation results in a marked improvement (~30x) in LOD. This improvement is directly tied to the much lower amounts of nonspecific ERL adsorption realized by circumventing the contribution of ERL accumulation by sedimentation. The impact of sedimentation on the response was also examined by generating high density maps of the SERS signal across an upright and inverted substrate, which provides further insight into the difference in the accuracy and precision of the assay when labeling in the two different substrate orientations. This analysis mapped the samples from Figure 4 that were prepared by upright exposures of antigen (5 ng mL-1 IgG for 12 h) and then ERL incubations (16 h) in the upright and inverted orientations. The maps were collected by mounting the samples on the x-y translation stage of a Raman microscope and measuring the signal across each address at



Page 14 of 27

14 locations offset (center-to-center) by 50 µm with a 10 µm laser spot size (633 nm at 3 mW at the sample). These maps are shown in Figures 6a (upright) and 6b (inverted), each of which are composed of ~1600 data points and extend slightly beyond the border of the 2 mm circular addresses. To aid visualization and simplify the analysis, the responses were normalized to the average signal intensity within a 0.7 mm diameter circle positioned at the center of the address (~850 data points) of the inverted sample. The maps are color scaled, with yellow representing a normalized intensity of 1.0. Cooler colors (blue, green) represent normalized intensities below 1.0 and warmer colors (red, orange) represent normalized intensities greater than 1.0. The differences in the homogeneity of the signal across the two substrates is readily apparent: the upright substrate has a much more pronounced level of variability. Note that the signal in the centermost regions of the inverted substrate is slightly stronger than those for the inverted substrate. This difference is consistent with the higher levels of nonspecific adsorption found in the dose-response plot for the inverted substrate in Figure 5b. The impact of the difference in the variation for the signal across the two substrates becomes more tractable by converting the maps from Cartesian to polar coordinates. This conversion can then be used to plot the normalized signal with respect to the radial displacement from the center of the disk-shaped address. Figures 6c-d show the resulting plots. In both cases, the signal undergoes a large drop to background levels near the edge of the address. The improvement in the quality of the measurement when using an inverted substrate is obvious. The improvement is more clearly quantified by the plots of the radial dependence of the average and standard deviation of the signal shown in Figures 6e-f. With the inverted substrate, the average signal undergoes a gradual decrease and drops below unity at a radial displacement


Page 15 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


15 of ~7000 µm. The average signal for the inverted substrate is, in contrast, much more consistent and reproducible over the same range of radial displacement. In other words, the relative average deviation and relative standard deviation of the signal for the radially bounded area (0 to 7000 µm) for the upright substrate is ~8% and ~17%, respectively, but only ~1% and 5% for the inverted substrate. These results begin to demonstrate that the platform described in Figure 1 has the potential to reach the same level of reliability as ELISA,54 one of the real workhorses in diagnostic testing arena.

Conclusions This work has demonstrated the impact of gravity-induced sedimentation on the delivery (i.e., the case for the inverted substrate) and accumulation of ERLs with 60 nm AuNP cores to the capture substrates in an immunoassay through a series of spectrophotometric measurements and modeling studies of settling rates and immunoassays. When the capture substrates were oriented in an upright orientation, sedimentation was directed towards the substrate, and the corresponding SERS signal failed to follow a t1/2 dependence expected for an accumulation process controlled only by diffusion. The upright position also resulted in capture substrates with a much greater heterogeneity of signal across the surface. The dependence of accumulation for the substrate inverted during the ERL incubation, which circumvented the settling of ERLs onto the capture substrate and more closely followed a t1/2 dependence expected for a diffusioncontrolled process. Surface scans show that the signal is decreased relative to the upright orientation and is much more homogeneous across the surface with significantly reduced standard deviation. These results suggest that sedimentation of nanoparticles can significantly affect the resulting



Page 16 of 27

16 measured signal from SERS immunoassays, and has potential implications in assays and other areas in the analytical sciences that seek to take advantage of signaling phenomena with these materials. Indeed, we are presently designing experiments to determine the potential impact of settling immunoassays based on magnetic nanoparticles and on quantum dots. We are also working to develop a more in depth model for both substrate orientations of the transport and accumulation of antigen and labels that increases the dimensionality of the model in order to examine smaller sized capture addresses and assays at much lower antigen and label concentrations.

Associated Content Supporting Information. Expanded details of the experimental procedures, descriptions of Raman instrumentation, and dynamic light scattering and ζ-potential measurements are available free of charge via the internet at http://pubs.acs.org.

Acknowledgments This research was supported by the National Institutes of Health (UG3CA211551 and R01AI111495).

References 1. Novotny, L.; Hecht, B., Principles of Nano-optics. Cambridge University Press: Cambridge, 2012. 2. Kelly, K. L.; Coronado, E.; Zhao, L. L.; Schatz, G. C., J. Phys. Chem. B 2003, 107, 668-677. 3. Daniel, M.-C.; Astruc, D., Chem. Rev. 2004, 104, 293-346. 4. Rogach, A. L.; Eychmüller, A.; Hickey, S. G.; Kershaw, S. V., Small 2007, 3, 536-557.


Page 17 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


17 5. Reiss, P.; Protiere, M.; Li, L., Small 2009, 5, 154-168. 6. Jain, P. K.; Huang, X.; El-Sayed, I. H.; El-Sayed, M. A., Acc. Chem. Res. 2008, 41, 15781586. 7. Huang, X.; El-Sayed, I. H.; Qian, W.; El-Sayed, M. A., J. Am. Chem. Soc. 2006, 128, 21152120. 8. Loo, C.; Lowery, A.; Halas, N.; West, J.; Drezek, R., Nano Lett. 2005, 5, 709-711. 9. Pankhurst, Q. A.; Connolly, J.; Jones, S.; Dobson, J., J. Phys. D: Appl. Phys. 2003, 36, R167R168. 10. Hyeon, T., Chem. Comm. 2003, 927-934. 11. Lin, X.-M.; Samia, A. C., J. Magn. Magn. Mater. 2006, 305, 100-109. 12. Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A., Science 1997, 277, 1078-1081. 13. De, M.; Rana, S.; Akpinar, H.; Miranda, O. R.; Arvizo, R. R.; Bunz, U. H.; Rotello, V. M., Nat. Chem. 2009, 1, 461-465. 14. Park, J.-W.; Shumaker-Parry, J. S., ACS Nano 2015, 9, 1665-1682. 15. Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M., Chem. Rev. 2005, 105, 1103-1170. 16. Vericat, C.; Vela, M.; Benitez, G.; Carro, P.; Salvarezza, R., Chem. Soc. Rev. 2010, 39, 18051834. 17. Cialla-May, D.; Zheng, X.-S.; Weber, K.; Popp, J., Chem. Soc. Rev. 2017, 46, 3945-3961. 18. Wang, Z.; Zong, S.; Wu, L.; Zhu, D.; Cui, Y., Chem. Rev. 2017, 117, 7910-7963. 19. McNay, G.; Eustace, D.; Smith, W. E.; Faulds, K.; Graham, D., Appl. Spectrosc. 2011, 65, 825-837. 20. Fan, M.; Andrade, G. F.; Brolo, A. G., Anal. Chim. Acta 2011, 693, 7-25. 21. Schlücker, S., Angew. Chem. Int. Ed. 2014, 53, 4756-4795. 22. Granger, J. H.; Schlotter, N. E.; Crawford, A. C.; Porter, M. D., Chem. Soc. Rev. 2016, 45, 3865-3882. 23. Ni, J.; Lipert, R. J.; Dawson, G. B.; Porter, M. D., Anal. Chem. 1999, 71, 4903-4908. 24. Grubisha, D. S.; Lipert, R. J.; Park, H.-Y.; Driskell, J.; Porter, M. D., Anal. Chem. 2003, 75, 5936-5943.



Page 18 of 27

18 25. Driskell, J. D.; Kwarta, K. M.; Lipert, R. J.; Porter, M. D.; Neill, J. D.; Ridpath, J. F., Anal. Chem. 2005, 77, 6147-6154. 26. Granger, J. H.; Granger, M. C.; Firpo, M. A.; Mulvihill, S. J.; Porter, M. D., Analyst 2013, 138, 410-416. 27. Crawford, A. C.; Laurentius, L. B.; Mulvihill, T. S.; Granger, J. H.; Spencer, J. S.; Chatterjee, D.; Hanson, K. E.; Porter, M. D., Analyst 2016, 142, 186-196. 28. Stiles, P. L.; Dieringer, J. A.; Shah, N. C.; Van Duyne, R. P., Annu. Rev. Anal. Chem. 2008, 1, 601-626. 29. Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R. R.; Feld, M. S., J. Phys.: Condens. Matter 2002, 14, R597-R624. 30. Porter, M. D.; Lipert, R. J.; Siperko, L. M.; Wang, G.; Narayanan, R., Chem. Soc. Rev. 2008, 37, 1001-1011. 31. Stenberg, M.; Stiblert, L.; Nygren, H., J. Theor. Biol. 1986, 120, 129-140. 32. Stenberg, M.; Werthén, M.; Theander, S.; Nygren, H., J. Immunol. Methods 1988, 112, 2329. 33. Nygren, H.; Werthen, M.; Stenberg, M., J. Immunol. Methods 1987, 101, 63-71. 34. Frackelton, A.; Weltman, J. K., J. Immunol. 1980, 124, 2048-2054. 35. Wang, G.; Driskell, J. D.; Porter, M. D.; Lipert, R. J., Anal. Chem. 2009, 81, 6175-6185. 36. Driskell, J. D.; Uhlenkamp, J. M.; Lipert, R. J.; Porter, M. D., Anal. Chem. 2007, 79, 41414148. 37. Allouni, Z.; Cimpan, M.; Høl, P.; Skodvin, T.; Gjerdet, N., Colloids Surf. B Biointerfaces 2009, 68, 83-87. 38. Cho, E.; Zhang, Q.; Xia, Y., Nat. Nanotechnol. 2011, 6, 385-391. 39. Park, M. S.; Park, J.; Jeon, S. K.; Yoon, T. H., J. Nanosci. Nanotechnol. 2013, 13, 72647270. 40. Alexander, C. M.; Dabrowiak, J. C.; Goodisman, J., J. Colloid Interface Sci. 2013, 396, 5362. 41. Ganguly, S.; Chakraborty, S., Phys. Lett. A 2011, 375, 2394-2399. 42. Alexander, C. M.; Goodisman, J., J. Colloid Interface Sci. 2014, 418, 103-112. 43. Phenrat, T.; Saleh, N.; Sirk, K.; Tilton, R. D.; Lowry, G. V., Environ. Sci. Technol. 2007, 41, 284-290.


Page 19 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


19 44. Wijenayaka, L. A.; Ivanov, M. R.; Cheatum, C. M.; Haes, A. J., J. Phys. Chem. C 2015, 119, 10064-10075. 45. Mason, M.; Weaver, W., Phys. Rev. 1924, 23, 412-426. 46. Park, H.-Y.; Driskell, J. D.; Kwarta, K. M.; Lipert, R. J.; Porter, M. D.; Schoen, C.; Neill, J. D.; Ridpath, J. F., Ultrasensitive Immunoassays Based on Surface-enhanced Raman Scattering by Immunogold Labels. Springer: Berlin, 2006; p 427-446. 47. Haiss, W.; Thanh, N. T.; Aveyard, J.; Fernig, D. G., Anal. Chem. 2007, 79, 4215-4221. 48. Reth, M., Nat. Immunol. 2013, 14, 765-767. 49. Axelsson, I., J. Chromatogr. A 1978, 152, 21-32. 50. Wang, W.; Ding, X.; Xu, Q.; Wang, J.; Wang, L.; Lou, X., Colloids Surf. B Biointerfaces 2016, 148, 541-548. 51. Hirtzel, C.; Rajagopalan, R., Colloidal phenomena: Advanced topics. Noyes Pubns: 1985. 52. Alarcon, E. I.; Udekwu, K.; Skog, M.; Pacioni, N. L.; Stamplecoskie, K. G.; Gonzalez-Bejar, M.; Polisetti, N.; Wickham, A.; Richter-Dahlfors, A.; Griffith, M.; Scaiano, J. C., Biomaterials 2012, 33, 4947-4956. 53. Rajput, S.; Werezuk, R.; Lange, R. M.; McDermott, M. T., Langmuir 2016, 32, 8688-8697. 54. Semenova, V.; Schiffer, J.; Steward-Clark, E.; Soroka, S.; Schmidt, D.; Brawner, M.; Lyde, F.; Thompson, R.; Brown, N.; Foster, L., J. Immunol. Methods 2012, 376, 97-107.



Page 20 of 27

20

Figure 1. SERS-based heterogeneous immunoassay: (a) preparation of extrinsic Raman labels (ERLs), which are gold nanoparticles modified with a RRM (DSNB), tracer antibody, and BSA; (b) preparation of the capture substrate; (c) antigen capture; subsequent labeling by ERLs, and Raman read out via a focused laser on the capture surface. The analyte is quantified using the symmetric nitro stretch [νs(NO2)] of the DSNB-based layer at 1336 cm-1. Further details of the procedures are given in the Supporting Information (S1-S2).


Page 21 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


21

Table 1. Summary of the DLS and ζ-potential measurements on as-received AuNPs and ERLs. Mean Hydrodynamic Size (nm) Mean ζ-potential (mV) t=0h t = 24 h t=0h t = 24 h As-Received AuNPs

59 ± 0.2

56 ± 0.1

-53 ± 0.9

-56 ± 2.7

Extrinsic Raman Labels

82 ± 0.5

78 ± 0.1

-1.8 ± 0.5

-10 ± 0.3



Page 22 of 27

22

Figure 2. (a) Schematic of the masked (black paint) quartz cuvette for the spectroscopic monitoring of the settling of as-received AuNPs. The optical window was ~0.2 × 0.2 cm in lateral size and was positioned 0.2 cm below the top of the AuNP suspension. The optical pathlength was 0.5 cm. (b) UV-Vis spectra of stagnant suspension of as-received 60-nm AuNPs (2.1×1010 particles/mL) using the cuvette in Figure 2(a). The spectra are shown at 1 h intervals for an 80-h duration. The spectra were baseline corrected and normalized to the extinction measurement at time zero. based on final and initial suspension measurements, respectively.


Page 23 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


23

Figure 3. (a) Experimental data collected for sedimentation of as-received 60 nm AuNPs by UVvis spectrometry. The data are normalized to the measured extinction of the as-received AuNPs. Black circles represent the experimental data collected while the grey lines indicate simulation results for the sedimentation model. The red line indicates model results for AuNPs with a 56-nm diameter. (b) Measured experimental settling for as-received AuNPs and completed ERLs with sedimentation model best-fit lines. The as-received AuNPs (black) show an increased sedimentation velocity compared to the completed ERLs (red).



Page 24 of 27

24

Figure 4. SERS-based immunoassay for human IgG in PBS with Tween 20. Dose-response plot for five separate calibration runs for human IgG (1 and 5 ng/mL) and a negative control sample with ERLs incubated in the (a) upright and (b) inverted incubation orientations. Dose-response plot for human IgG calibration versus t1/2 for (c) upright and (d) inverted ERL incubations. Residuals for linear fit to a t1/2 dependence for the (e) upright and (f) inverted ERL incubations.


Page 25 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


25

Figure 5. (a) Rescaling of the blank substrate data from the upright (blue circles) and inverted (red squares) shown in Figures 4c-d. (b) Dose-response plots for the blank, 1 ng mL-1, and 5 ng mL-1 of both the upright (circles) and inverted (squares) assay substrates after a 16 h incubation.



Page 26 of 27

26

Figure 6. High density Raman surface analysis for 5 ng/mL of human IgG immunoassay substrates in upright and inverted ERL incubation position. The high-density map consists of 50 ± 1 µm steps on the surface taken at 3 mW of power with a 5× Olympus objective with a 5 µm spot size. (a) Representative sample substrate with a SERS signal intensity normalized to the average SERS signal from the inner 1.6 mm diameter radius of the substrate from an upright and (b) inverted ERL incubation. (c) Normalized signal intensities versus the radial component of polar coordinates showing differences in the signal variation for an upright and (d) inverted ERL incubation. (e) Average and standard deviation of the average normalized SERS intensity for concentric and separate rings for upright and (f) inverted ERL incubation.


Page 27 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


27

TOC Graphic


Gold Nanoparticle Labels and Heterogeneous ... - ACS Publications

Recommend Documents