Subscriber access provided by Karolinska Institutet, University Library
Article
A Robust and User-Friendly Alternative to Densitometry Using Origami Biosensors and Digital Logic Steven M. Russell, Alejandra Alba, Marcio Borges, and Roberto de la Rica ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.8b00452 • Publication Date (Web): 07 Aug 2018 Downloaded from http://pubs.acs.org on August 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 8 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
ACS Sensors
A Robust and User-Friendly Alternative to Densitometry Using Origami Biosensors and Digital Logic Steven M. Russell,1,‡ Alejandra Alba,1,‡ Marcio Borges,2 Roberto de la Rica.1,2* 1
Department of Chemistry, University of the Balearic Islands, 07122 Palma de Mallorca, Illes Balears Spain. Multidisciplinary Sepsis Group, Balearic Islands Health Research Institute (IdISBa) KEYWORDS augmented reality, gold, nanoparticles, paper-based analytical device, colorimetric, smartphone 2
ABSTRACT: Colorimetric detection with smartphones is ideal for point-of-care measurements because the signal reader is easily available. Densitometric detection schemes enable semi-quantitative measurements but require a light-proof box to control photographic conditions and/or extensive data treatment to extract information. Approaches based on pattern recognition are not so sensitive to light artifacts but can only yield a yes/no type of answer when the signal is above or below a certain threshold. Here we introduce a new method for detecting different concentrations of proteins as well as light artifacts with origami immunosensors and digital logic. The origami design consists of a folded piece of paper with three identical biorecognition sites so that one drop of sample generates three colorimetric signals simultaneously. The three colorimetric signals are then evaluated with an augmented reality app that generates a virtual semaphore that sequentially turns on its green, yellow and red lights depending on the concentration of analyte. These three Boolean variables pass through “and” and “not” logic gates in a 3-to-8 decoder that enable the semiquantitative detection of proteins and that adds a failsafe against erroneous results. The proposed method can detect the model analyte mouse IgG with a limit of detection and sensitivity comparable to densitometry performed under light-controlled conditions. It can also detect the sepsis biomarker procalcitonin (PCT) at clinically relevant concentrations. With our approach the detection is performed in real time and signal processing is not required, which makes it suitable for rapid analyses by non-specialists at the point of need.
Paper-based analytical devices are a popular choice for developing fully integrated biosensors due to their low cost, easy disposal, and intrinsic fluid transport properties through wicking.1-4 Among these, sensors that use colorimetric transduction mechanisms are particularly appealing because the signals can be read with the cameras integrated in smartphones, which is a technology that already has a high global market penetration.510 In many cases, detection with a smartphone requires taking a photograph of the colored paper and evaluating the tonality or optical density with image processing software.11,12 This approach, however, is susceptible to errors caused by variable photographic conditions, which makes it necessary to control the illumination, distance, and angle at which the colorimetric signals are photographed. These kinds of controls are often accomplished with a small light-proof box, or a cradle.13-15 Another method is to compensate for these photographic variables with software that detects reference markings on the paper substrate. For example, black and white color reference spots can be used to evaluate various ambient light conditions and automatically perform correction calculations to the reading of the colorimetric signals.16 Likewise, reference markings that are evenly spaced around the colorimetric signal can be used to detect distance and angle variables and compensate for them accordingly.17 However, this approach may still yield erroneous results if the operator casts a partial shadow over the assay or the reference markings. A more robust method to detect colorimetric signals is via pattern recognition, that is, when the colorimetric signal is incorporated into a pattern, such as a barcode or a QR code.18-21 In this method the signal
is read with a smartphone by means of pattern recognition software instead of photographic analysis. Pattern recognition approaches are less affected by illuminance than densitometry, but even so, they could still yield false signals under extreme light conditions, or when the patterns are partially covered by shadows.20 Another limitation of using pattern recognition as a type of transduction mechanism is that it can only determine whether the analyte concentration is above or below a threshold value, which is less quantitative than photographic approaches that measure optical density. In this context, a detection platform that combines the robustness of pattern recognition methods with the more quantitative information afforded by photographic analysis could solve these issues and enable the detection of different concentrations of analyte with confidence. In this manuscript we introduce a method to generate and interpret three colorimetric signals originated by a single drop of sample on a paper-based biosensor. This method enables a semi-quantitative evaluation of the analyte concentration based on the combination of Boolean values derived from the three signals via pattern recognition and augmentation, i.e. an app with an augmented reality (AR) user-interface.20, 21 Moreover, this method provides an inner control mechanism for the detection of light artifacts in order to safeguard against some false signals that can commonly occur in densitometric methods. We developed this approach with a model immunoassay for the detection of mouse immunoglobulin (mouse IgG) using gold nanoparticles as colorimetric probes (Figure 1a). The competitive immunoassay consists of mixing the sample with
ACS Paragon Plus Environment
ACS Sensors 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 8
the colorimetric signals are dark enough to block the recognition of each of the superimposed AR target patterns. The relationships among the three Boolean values are then represented with an augmented model of a traffic semaphore, which acts as a visually intuitive user-interface for the interpretation of the assay results (Figure 1c). In logic terms, the three Boolean variables pass through “and” and “not” logic gates in a 3-to-8 decoder (Fig. 2). Interpreting combinations of the three Boolean variables in this way not only allows the semi-quantitative detection of the analyte, but also adds a failsafe against erroneous results caused by light artifacts as shown in Figure 2 and Table 1. Previously logic gates were used to process multiple signals into a single output.22-25 Here we use them to determine whether the concentration of analyte is low, mid, or high by observing the sequential generation of green, yellow and red lights in the traffic semaphore. Furthermore, the absence of green, yellow or red lights in the predetermined order of the semaphore can be used as an inner alert to identify invalid tests (Fig. 2). This is particularly useful for detecting interfering shadows cast by the user. All in all, the simple origami design shown in Fig. 1b overcomes previous limitations associated with yes/no signal outputs and false signals in colorimetric transduction mechanisms. This, along with the userfriendly interpretation of the traffic semaphore, makes our approach useful for detecting different levels of protein biomarkers with paper-based immunosensors at the point of care. Inputs
Figure 1. Schematic representation of the immunosensor and colorimetric signal detection; A) the analyte (mouse IgG) competes with substrate-bound molecules for the interaction with biotinylated antibodies; the signal is generated subsequently with avidin-decorated gold nanoparticles; B) Folding the paper like an accordion and modifying the top, middle and bottom layers with the target results in the generation of three colored spots upon addition of a single drop of sample and a single drop of nanoparitcles; C) The colorimetric signals are simultaneously evaluated with and augmented reality (AR) app that generates traffic lights when the signals are not strong enough to block the recognition of three patterns specially designed for this purpose.
the antibody and subsequently adding it to a piece of paper modified with mouse IgG so that the antibodies that do not bind the analyte become attached to the substrate. The amount of antibody bound to the paper is then estimated with avidindecorated gold nanoparticles. The key step for detecting different concentrations of analyte is to perform the assay in a folded piece of paper modified with multiple biorecognition sites as shown in Figures 1a and 1b. With this biosensor design, a single drop of sample travels downward through the layers and generates three signals (Fig. 1b). After unfolding the paper, the three colored spots are evaluated simultaneously with AR. We have previously demonstrated that the recognition of AR target patterns can be blocked by colorimetric signals when superimposed on the substrate with a transparent sheet, thus generating different augmented messages for the user as a function of the presence of analyte above or below a certain concentration value.20 Here we adapt this concept and expand it by using not one, but three AR target patterns superimposed over three colorimetric signals. Each of the AR target patterns returns a Boolean value depending on whether or not
SI
Outputs
G
Y
R
S0
S1
S2
S3
S4
S5
S6
S7
0
0
0
1
0
0
0
0
0
0
0
I
0
0
1
0
1
0
0
0
0
0
0
I
0
1
0
0
0
1
0
0
0
0
0
I
0
1
1
0
0
0
1
0
0
0
0
I
1
0
0
0
0
0
0
1
0
0
0
V/L
1
0
1
0
0
0
0
0
1
0
0
I
1
1
0
0
0
0
0
0
0
1
0
V/M
1
1
1
0
0
0
0
0
0
0
1
V/H
Table 1. Truth table for 3-to-8 decoder with Boolean inputs green (G), yellow (Y), and red (R) and semaphore interpretations (SI) of invalid (I), valid/low (V/L), valid/mid (V/M), and valid/high (V/H)
Figure 2. Logic scheme for a 3-to-8 decoder with corresponding traffic semaphore output from the AR user interface.
ACS Paragon Plus Environment
Page 3 of 8 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
ACS Sensors
EXPERIMENTAL SECTION Materials. Whatman 41 ashless quantitative filter paper, Immunoglobulin from mouse serum (mouse IgG, technical grade, >95%), anti-mouse IgG (whole molecule)-biotin antibody produced in goat (biotin anti-mouse IgG, affinity isolated antibody) Tween-20, and bovine serum albumin (BSA, >98%) were purchased from SigmaAldrich. Procalcitonin monoclonal antibody (44d9) raised in mouse and human procalcitonin recombinant protein were purchased from Thermo Fisher Scientific. Phosphate buffer saline (PBS, pH 7.4) was prepared following standard protocols published elsewhere. PBST refers to PBS containing 0.1% Tween-20. PBS-BSA refers to a PBS solution containing 5 mg mL−1 BSA. 40 nm gold nanoparticles were synthesized with the Turkevich method.20 This nanoparticle size was chosen because large nanoparticles have larger extinction coefficients than their smaller counterparts. Competitive Immunoassay. Paper-based biosensors were prepared as follows. Whatman 41 ashless quantitative filter paper, which has a pore size of 20-25 µm, was used as the substrate for the assay. The pore size has a negligible impact in the assay, as shown in a previous study.27 The Filter paper was cut into 3 x 12 cm pieces and folded like an accordion to generate four 3 x 3 cm squares as shown in Figure 1b. Mouse IgG (5 µL, 100 µg⋅mL-1) was spotted in the center of the three top 3 x 3 cm squares. After drying for 5 min at room temperature, the paper substrate was folded and subsequently blocked by adding 1 mL of PBS-BSA in the center of the top layer and letting it dry unfolded. This blocking step was performed to reduce non-specific interactions in future biorecognition steps. The detection of mouse IgG with the as-prepared biosensors was accomplished with the following method. First, samples containing different concentrations of mouse IgG were obtained by diluting a stock solution in PBST. Biotinylated anti-mouse IgG was first diluted 1:10 in PBST. Then, 1 µL of diluted anti-mouse IgG was added to 10 µL of the mouse IgG solutions for 30 min. Next, 5 µL of the mix was added in the middle of the folded paper biosensor previously rehydrated with 1 ml of PBST. After 20 min, the strips were washed by adding 1 ml of PBST. Immediately after, 5 µl of avidin-modified gold nanoparticles was dripped in the center of each folded strip and allowed to react for 4 min. The 40 nm gold nanoparticles were stabilized with carboxylate-ending mercaptopolyethylene glycol ligands (Fig. S5) and covalently modified with avidin via EDC/sulfoNHS coupling as previously described.20 Then the biosensors were washed 3 times by adding 1 ml of PBST in order to remove the excess of gold nanoparticles. Finally, the paper strips were unfolded and dried at room temperature onto a piece of absorbent paper. PCT was detected with the same method but substituting mouse IgG for PCT and the biotinylated anti-mouse IgG for a mixture of anti-PCT (1:300 in PBST) and biotinylated anti-mouse IgG (1:10 in PBST). The mixture was incubated with PCT spiked into synthetic serum for 1h. The synthetic serum consisted in 140 mM NaCl, 4.5 mM KCl, 2.5 mM CaCl2, 0.8 mM MgCl2, 10 mg mL-1 BSA, 2.5 mM urea and 4.7 mM glucose in 10 mM phosphate buffer pH 7.4.29 Detection of colorimetric signals. Densitometric analysis was performed by imaging the paper biosensors with a flatbed scanner (Hewlett Packard color laser Jet Pro MFPM377dw). Images were then transformed into black and white images with Adobe Photoshop. After inverting
the color, the average pixel density inside each spot was obtained with the histogram function of Photoshop. The final densitometric signal displayed is the average of the 3 independent assays. Error bars are the standard deviation. Signal detection with augmented reality was achieved with an app developed using Unity2017.3 and its built-in augmented reality SDK from Vuforia. The app generates three augmented objects upon recognition of three different target patterns. The targets were designed following published procedures.21 The first target generates the semaphore structure and the green light, the second target generates a yellow light, and the third target generates the red light. The targets were printed on APLI 00860 high resistance polyester sheets with a Brother MFC-1910W printer. Colorimetric signals with high optical density were able to block pattern recognition, therefore preventing the generation of the digitally augmented media corresponding to each target pattern. Information about target calibration including the effect of the colorimetric signal’s diameter and optical density can be found in Figure 4 and Figures S1-S3 in the Supporting Information online. For the detection of mouse IgG the contrast of all the patterns was set by changing their transparency to 40% with Microsoft Word.20,21For the detection of PCT the pattern generating the green light was set to 30% transparency whereas the patterns generating the yellow and red lights were set to 35% and 40% transparency, respectively.
Figure 3. Photographs of unfolded origami biosensors biosensors showing the colorimetric signals generated by different concentrations of analyte in the top, middle and bottom layers as schematized in Fig. 1b.
Results and discussion The logical combinations of Boolean variables signals, as shown in Table 1 and Fig. 2, are based on generating 3 colorimetric signals with a single drop of sample. Figure 3 shows pictures of the unfolded origami biosensors used for the detection of mouse IgG with a competitive immunoassay. In these images, samples that contain high concentrations of analyte yield low signals because the antibodies bind their target in solution, and therefore they do not attach to the paper through biospecific interactions. Conversely, biotinylated antibodies are free to interact with substrate-bound molecules when the concentration of analyte is low, which results in highly colored spots upon addition of avidin-decorated nanoparticles. In these photographs it is also evident that the optical density of the colorimetric signals decreases in the middle and bottom layers as compared to the top layer. Furthermore, the diameter of the colored spots generated at the different layers decreases as the drop travels through the paper. The decrease in optical density and diameter is attributed to a decrease in sample concentration due to both biospecific interactions and radial diffusion. In summary, Figure 3 demonstrates that it is possible to gener-
ACS Paragon Plus Environment
ACS Sensors 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
ate three colorimetric signals with a single drop of sample followed by washing and adding a single drop of nanoparticle probes, and that the as-generated signal decreases in size and optical density as the liquids flow through the paper layers. Below we analyze the impact of these parameters in a signal transduction mechanism based on pattern recognition using augmented reality. Figure 4 shows the impact of optical density and colorimetric signal diameter in pattern recognition with augmented reality. In this Figure, the patterns printed on a transparent sheet act as colorimetric transducers. A sampling pattern with a diameter of 1.7 cm yields a positive (+) outcome when the underlying colored spot does not block pattern recognition whereas a control pattern yielding an asterisk (*) is used to ensure that the app is functioning properly. Guidelines for the fabrication of the pattern have been published elsewhere.20,21 In these experiments, the optical density was varied by changing the colloid concentration in the nanoparticle dispersions. Spots with diameters 13.4 ± 0.8, 9.8 ± 0.3, 7.8 ± 0.3 and 6.7 ± 0.3 were obtained by pipetting sample volumes 10, 5, 3 and 2 µL. In Figure 4, spots with low nanoparticle concentration do not block pattern recognition because their optical density is not high enough to change the local contrast sufficiently. Small spots have less impact in blocking pattern recognition because they do not cover enough of the target’s feature points to impede recognition. Discernibly, the size of the target pattern itself also plays a key role in pattern recognition, since 1.4 cm patterns were blocked by smaller and less intensely colored spots as shown in Figure S1 in the Supporting Information. These experiments demonstrate that smaller and less colored spots are less efficient at disrupting pattern recognition. This means that colorimetric signals generated in the middle and bottom layers of the origami biosensors are progressively less likely to block pattern recognition. Below we explore how to apply this concept to the semi-quantitative analysis of colorimetric signals by combining the three Boolean variables in a digital logic system.
Figure 4. Augmented reality-based detection of colorimetric signals originated by drops of gold nanoparticles at different concentrations and with different volumes (pattern diameter 1.7 cm)
The simultaneous detection of the 3 colorimetric signals generated in each origami biosensor was accomplished by aligning 3 patterns printed on a transparent sheet on top of the immunoassays as shown in Figure 1c. The calibration experiments shown in Figure 4 were performed for each pattern as shown in Figures S2 and S3. The pattern transducing the colorimetric signal from the bottom layer was programmed to yield the semaphore frame and the green light, whereas the patterns transducing the middle and top signals were programmed to generate the yellow and red light, respectively. The green light sensor is not as easily blocked than the yellow and red light sensors respectively (Figs. S1-S3), which is the basis for detecting different optical densities with the augmented semaphore. Figure 5 shows snapshots of the resulting augmented semaphores for the detection of mouse IgG in 3 independent experiments (see also Videos S1-S3 in the Supporting Information). In these experiments, when the concentration of analyte is 0 or 10-3 µg mL-1 the app generates a semaphore with a green light because the top and middle layers generate colorimetric signals whose optical density and diameter is large enough to block pattern recognition. In other words, in these experiments the red and yellow augmented signals are not generated. When the concentration of analyte is found in the range between 10-2 µg mL-1 and 100 µg mL-1 the middle signal is also not strong enough to change the local contrast, and thus the green and yellow lights are visible. Finally, when the concentration of analyte is 101 µg mL-1 or higher then none of the signals can be detected, and the green, yellow and red lights are visible. These experiments demonstrate that it is possible to detect different concentrations of analyte with a single drop of sample using the proposed origami biosensors. The limit of detection, expressed here as the sample that yields a signal different from the signal of the blank in 3 independent experiments, is 10-2 µg mL-1.
Figure 5. Snapshots of the augmented semaphore generated as function of the concentration of mouse IgG by means of a com-
ACS Paragon Plus Environment
Page 4 of 8
Page 5 of 8 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
ACS Sensors petitive immunoassay on origami biosensors. Each series of images corresponds to an independent series of experiments.
In Figure 6, densitometric analysis of the top signals under controlled light conditions with a flatbed scanner yields a comparable limit of detection of 10-2 µg mL-1 following the 3σ criterion (99% confidence). Samples with a concentration of 10-2, 10-1 and 100 µg mL-1 yield signals that are not significantly different among them, which is the same result obtained with the augmented reality app. Furthermore, it should be noted that the optical density in densitometric analysis is expressed as a relative value, that is, as the increment from the blank experiment in the same series of experiments, which greatly reduces the inter-series variability. Contrarily, the detection based on augmented reality analyzes absolute grayscale values in real time without any posterior signal processing steps. All in all, the experiments shown in Figures 5 and 6 indicate that, in the proposed experimental conditions, our method detects colorimetric signals with a sensitivity that is comparable to densitometry, but without the need for strictly controlled photographic conditions. Furthermore, the digitally augmented objects are generated in real time and do not require the additional step of using image processing software to extract data. Finally, the traffic semaphore system can be easily understood by anyone across ages and language or educational barriers, which makes this approach accessible to non-specialists performing analyses at the point of need.
inadequate conditions that may impede reading the assay confidently. In Figure 7b a defective biosensor that does not generate the expected signals can be detected as well thanks to the logic scheme shown in Figure 2. Artifacts partially or totally disrupting the generation of traffic lights in Figs 7c-7e also generate invalid tests and force the user to re-evaluate light conditions in accordance to Table 1. These experiments demonstrate that the proposed logic evaluation of colorimetric signals can reduce inaccurate interpretations caused by some of the most common artifacts in colorimetric detection, which is a unique feature compared to other colorimetric transduction mechanisms. Although the proposed control mechanism cannot detect light artifacts affecting only the red or red and yellow lights, it should be noted that such artifacts are less common because the light source should be placed in front of the user, and it is more difficult to project shadows on the assay under this scenario. Nevertheless, these problems could also be detected by using additional targets surrounding the assay that emit control signals depending on the presence of light artifacts originated from different angles (see Figure S4 in the Supporting Information).
Figure 7. Detection of artifacts and false signals by means of logic evaluation of AR-generated signal outputs.
Figure 6. Scanned images (A) and densitometric analysis (B) of the top layer of the paper biosensors after the competitive immunoassays. Error bars are the standard deviation (n=3).
The semi-quantitative evaluation of colorimetric signals with AR shown in Figure 5 is based in the sequential generation of traffic lights, as depicted in Table 1. This type of signal output can also be used as an inner control to detect false signals (invalid tests) as exemplified in Figure 7. For example, one of the most common artifacts in colorimetric detection is originated when a light source behind the user casts a partial shadow on the assay. This has a higher effect on signals that are closer to the user, in our case the green light. In Figure 7a shadows blocking the assay are easily recognized because the green signal is not present, therefore alerting the user about
Figure 8. Detection of procalcitonin (PCT) at different concentrations with a competitive immunoassay and the augmented semaphore. Each row corresponds to a different series of experiments.
ACS Paragon Plus Environment
ACS Sensors 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
Finally, we sought to apply the proposed detection system for quantifying different levels of the sepsis biomarker procalcitonin (PCT). Patients with serum levels above 2 ng⋅mL-1 are at risk of suffering from sepsis, and those with serum levels equal or higher than 10 ng⋅mL-1 are very likely to suffer from sepsis with bacterial infection.28 With this in mind we adapted the proposed detection system to display a yellow light when the concentration of PCT is higher than 2 ng⋅mL-1 and a red light when the concentration was higher than 10 ng⋅mL -1. Samples were prepared by spiking recombinant human PCT into synthetic serum to simulate a real matrix. In Figure 8, the competitive immunoassay yields a yellow light in three independent experiments when the concentration of PCT is 2 ng⋅mL-1, and the semaphore turns red when the concentration of PCT is 10 ng⋅mL-1 or higher, which indicates that the proposed method could be used for the rapid and easy evaluation of PCT in septic patients.
Conclusions In conclusion we have introduced a new method for semiquantitative immunodetection through pattern recognition. It consists of simultaneously generating three signals with a single drop of sample thanks to the origami biosensor design, and to evaluate the intensity of the three signals simultaneously with AR as a function of the concentration of the target molecule. This concept was demonstrated with an augmented semaphore that sequentially generates green, yellow or red lights as a function of the concentration of analyte, for example, at different clinically relevant concentrations of PCT. This approach also enabled the detection of artifacts when the traffic lights did not appear following the expected sequence. Our method is as sensitive as densitometry performed with a flatbed scanner, and it does not require any data treatment to extract semi-quantitative information. This, along with the straightforward origami design, makes our approach promising for developing mobile biosensors for decentralized healthcare.26
ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Supporting Information document (PDF) Videos S1-S3 (.avi)
AUTHOR INFORMATION Corresponding Author *
[email protected] Author Contributions The manuscript was written through contributions of all authors. / All authors have given approval to the final version of the manuscript. / ‡These authors contributed equally.
Notes Any additional relevant notes should be placed here.
ACKNOWLEDGMENT RR acknowledges financial support from grant CTQ2017-82432R (MINECO/AEI/FEDER, UE) as well as a Ramón y Cajal con-
tract from Ministerio de Economía, Industria y Competitividad, Agencia estatal de investigación, Universitat de les Illes Balears, Conselleria d’Innovació, Recerca i Turisme and the European Social Fund. AA acknowledges a scholarship from Fundación Carolina.
ABBREVIATIONS AR augmented reality; IgG immunoglobulin; PCT procalcitonin.
REFERENCES (1) Boehle, K. E.; Gilliand, J.; Wheeldon, C. R.; Holder, A.; Adkins, J. A.; Geiss, B. J.; Ryan, E. P.; Henry, C. S. Utilizing PaperBased Devices for Antimicrobial-Resistant Bacteria Detection. Angew. Chemie - Int. Ed. 2017, 56 (24), 6886–6890. (2) Xu, G.; Nolder, D.; Reboud, J.; Oguike, M. C.; van Schalkwyk, D. A.; Sutherland, C. J.; Cooper, J. M. Paper-Origami-Based Multiplexed Malaria Diagnostics from Whole Blood. Angew. Chemie - Int. Ed. 2016, 55 (49), 15250–15253. (3) Lee, M.; Oh, K.; Choi, H. K.; Lee, S. G.; Youn, H. J.; Lee, H. L.; Jeong, D. H. Subnanomolar Sensitivity of Filter Paper-Based SERS Sensor for Pesticide Detection by Hydrophobicity Change of Paper Surface. ACS Sensors 2018, 3 (1), 151–159. (4) Tsai, T. T.; Huang, C. Y.; Chen, C. A.; Shen, S. W.; Wang, M. C.; Cheng, C. M.; Chen, C. F. Diagnosis of Tuberculosis Using Colorimetric Gold Nanoparticles on a Paper-Based Analytical Device. ACS Sensors 2017, 2 (9), 1345–1354. (5) Meng, X.; Schultz, C. W.; Cui, C.; Li, X.; Yu, H. Z. On-Site Chip-Based Colorimetric Quantitation of Organophosphorus Pesticides Using an Office Scanner. Sens. Actuators, B Chem. 2015, 215, 577–583. (6) Dong, C.; Wang, Z.; Zhang, Y.; Ma, X.; Iqbal, M. Z.; Miao, L.; Zhou, Z.; Shen, Z.; Wu, A. High-Performance Colorimetric Detection of Thiosulfate by Using Silver Nanoparticles for Smartphone-Based Analysis. ACS Sensors 2017, 2 (8), 1152–1159. (7) Deraney, R. N.; Mace, C. R.; Rolland, J. P.; Schonhorn, J. E. Multiplexed, Patterned-Paper Immunoassay for Detection of Malaria and Dengue Fever. Anal. Chem. 2016, 88 (12), 6161–6165. (8) Huang, J. Y.; Lin, H. T.; Chen, T. H.; Chen, C. A.; Chang, H. T.; Chen, C. F. Signal Amplified Gold Nanoparticles for Cancer Diagnosis on Paper-Based Analytical Devices. ACS Sensors 2018, 3 (1), 174–182. (9) Zhang, D.; Gao, B.; Chen, Y.; liu, H. Converting Colour to Length Based on Coffee-Ring Effect for Quantitative Immunoassays Using a Ruler as Readout. Lab Chip 2017, 18, 271–275. (10) Pilavaki, E.; Demosthenous, A. Optimized Lateral Flow Immunoassay Reader for the Detection of Infectious Diseases in Developing Countries. Sensors 2017, 17 (11), 2673. (11) Wong, J. X. H.; Liu, F. S. F.; Yu, H. Z. Mobile App-Based Quantitative Scanometric Analysis. Anal. Chem. 2014, 86 (24), 11966– 11971. (12) Yu, H.; Le, H. M.; Kaale, E.; Long, K. D.; Layloff, T.; Lumetta, S. S.; Cunningham, B. T. Characterization of Drug Authenticity Using Thin-Layer Chromatography Imaging with a Mobile Phone. J. Pharm. Biomed. Anal. 2016, 125, 85–93. (13) Yetisen, A. K.; Jiang, N.; Tamayol, A.; Ruiz-Esparza, G. U.; Zhang, Y. S.; Medina-Pando, S.; Gupta, A.; Wolffsohn, J. S.; Butt, H.; Khademhosseini, A.; et al. Paper-Based Microfluidic System for Tear Electrolyte Analysis. Lab Chip 2017, 17 (6), 1137–1148. (14) Roda, A.; Michelini, E.; Cevenini, L.; Calabria, D.; Calabretta, M. M.; Simoni, P. Integrating Biochemiluminescence Detection on Smartphones: Mobile Chemistry Platform for Point-of-Need Analysis. Anal. Chem. 2014, 86, 7299-7304. (15) Gallegos, D.; Long, K. D.; Yu, H.; Clark, P. P.; Lin, Y.; George, S.; Nath, P.; Cunningham, B. T. Label-Free Biodetection Using a Smartphone. Lab Chip 2013, 13 (11), 2124. (16) Hong, J. Il; Chang, B.-Y. Development of the SmartphoneBased Colorimetry for Multi-Analyte Sensing Arrays. Lab Chip 2014, 14 (10), 1725–1732. (17) Jia, M. Y.; Wu, Q. S.; Li, H.; Zhang, Y.; Guan, Y. F.; Feng, L. The Calibration of Cellphone Camera-Based Colorimetric Sensor Array and Its Application in the Determination of Glucose in Urine. Biosens. Bioelectron. 2015, 74, 1029–1037. (18) Zhang, Y.; Sun, J.; Zou, Y.; Chen, W.; Zhang, W.; Xi, J. J.; Jiang, X. Barcoded Microchips for Biomolecular Assays. Anal. Chem. 2015, 87 (2), 900–906.
ACS Paragon Plus Environment
Page 6 of 8
Page 7 of 8 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
ACS Sensors (19) Yamada, K.; Suzuki, K.; Citterio, D. Text-Displaying Colorimetric Paper-Based Analytical Device. ACS Sensors 2017, 2 (8), 1247–1254. (20) Russell, S. M.; Doménech-Sánchez, A.; de La Rica, R. Augmented Reality for Real-Time Detection and Interpretation of Colorimetric Signals Generated by Paper-Based Biosensors. ACS Sensors 2017, 2 (6), 848–853. (21) Russell, S. M.; de la Rica, R. Paper Transducers to Detect Plasmon Variations in Colorimetric Nanoparticle Biosensors Sensors Actuators, B Chem. 2018, 270, 327-332. (22) Chen, J.; Fang, Z.; Lie, P.; Zeng, L. Computational Lateral Flow Biosensor for Proteins and Small Molecules: A New Class of Strip Logic Gates. Anal. Chem. 2012, 84 (15), 6321–6325. (23) Reuel, N. F.; McAuliffe, J. C.; Becht, G. A.; Mehdizadeh, M.; Munos, J. W.; Wang, R.; Delaney, W. J. Hydrolytic Enzymes as (Bio)Logic for Wireless and Chipless Biosensors. ACS Sensors 2016, 1 (4), 348–353. (24) Molinnus, D.; Poghossian, A.; Keusgen, M.; Katz, E.; Schöning, M. J. Coupling of Biomolecular Logic Gates with Electronic Transducers: From Single Enzyme Logic Gates to Sense/Act/Treat Chips. Electroanalysis 2017, 29 (8), 1840–1849. (25) Peng, Y.; Zhou, W.; Yuan, R.; Xiang, Y. Dual-Input Molecular Logic Circuits for Sensitive and Simultaneous Sensing of Multiple
microRNAs from Tumor Cells. Sens. Actuators B Chem. 2018, 264, 202– 207. (26) Russell, S. M.; de la Rica, R. Policy Considerations for Mobile Biosensors. ACS Sensors 2018, 3, 1059-1068. (27) Alba-Patiño, A.; Russell, S. M.; de la Rica, R. Origamienabled signal amplification for paper-based colorimetric biosensors Sens. Actuators B 2018, 273, 951-954.
(28) Meynaar, I. A.; Droog, W.; Bastra, M.; Vreede, R.; Herbrink, P. In Critically Ill Patients, Serum Procalcitonin Is More Useful in Differentiating between Sepsis and SIRS than CRP, Il-6, or LBP. Crit. Care Res. Pract. 2011, 2011, 594645. (29) Çevik, S. Xanthine Biosensor Based on XO/AuNP/PtNP/MWCNT Hybrid Nanocomposite Modified GCPE. Biotechnol. Bioprocess Eng. 2016, 21, 314-320.
ACS Paragon Plus Environment
ACS Sensors 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 8 of 8
SYNOPSIS TOC
ACS Paragon Plus Environment
8
