- Email: [email protected]
Fast analysis of ketamine using a colorimetric immunosorbent assay on a paper-based analytical device
Accepted Manuscript Title: Fast Analysis of Ketamine Using a Colorimetric Immunosorbent Assay on a Paper-Based Analytical Device Authors: Chung-An Chen, Peng-Wei Wang, Yu-Chun Yen, Hsin-Lan Lin, Yao-Chung Fan, Shou-Mei Wu, Chien-Fu Chen PII: DOI: Reference:
S0925-4005(18)32026-4 https://doi.org/10.1016/j.snb.2018.11.071 SNB 25669
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
18 July 2018 30 October 2018 14 November 2018
Please cite this article as: Chen C-An, Wang P-Wei, Yen Y-Chun, Lin H-Lan, Fan Y-Chung, Wu S-Mei, Chen C-Fu, Fast Analysis of Ketamine Using a Colorimetric Immunosorbent Assay on a Paper-Based Analytical Device, Sensors and amp; Actuators: B. Chemical (2018), https://doi.org/10.1016/j.snb.2018.11.071 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Fast Analysis of Ketamine Using a Colorimetric Immunosorbent Assay on a Paper-Based Analytical Device
SC R
Fanc,*, Shou-Mei Wud, and Chien-Fu Chena,*
IP T
Chung-An Chena,1, Peng-Wei Wangb,1, Yu-Chun Yena, Hsin-Lan Lina, Yao-Chung
Institute of Applied Mechanics, National Taiwan University, Taipei, Taiwan
b
Department of Psychiatry, Kaohsiung Medical University Chung Ho Memorial
U
a
A
Department of Computer Science and Engineering, National Chung Hsing
ED
University, Taichung, Taiwan
M
c
N
Hospital, Kaohsiung, Taiwan
Food and Drug Administration, Ministry of Health and Welfare, Taipei, Taiwan
1
These authors contribute equally to this work.
CC E
PT
d
* Corresponding Author
A
Chien-Fu Chen, E-mail: [email protected] Yao-Chung Fan, E-mail: [email protected]
1
SC R
IP T
Graphical abstract
A rapid colorimetric sensing system using competitive ELISA test on a
N
•
U
Highlights:
Oral fluid was selected for the test to facilitate the collection of samples based on
ED
•
M
ketamine, a frequently abused drug.
A
microfluidic paper-based analytical device was investigated for the detection of
PT
its advantages of low infection risk, noninvasiveness of sample collection, and decreased chance of sample adulteration.
CC E
•
An Android smartphone app is developed to analyze and measure the results of
A
the test, further enhancing the test’s portability, and image recording and data
•
transmission capabilities. After optimization of the operation parameters, including reaction temperature, vibration washing time, reaction time, and antibody concentration, the resulting assay can be completed in as little as 6 min with a detection limit of 0.03 ng/mL. 2
•
The paper-based platform features 90% sensitivity (confidence interval (CI): 76.34-97.21%) and 92% specificity (CI: 80.77-97.78%) for ketamine analysis of
IP T
90 oral fluid samples from drug abuse patients.
SC R
Abstract
In this study, we designed a rapid colorimetric sensing system using competitive
U
ELISA test on a microfluidic paper-based analytical device for the detection of
N
ketamine, a frequently abused drug. Oral fluid was selected for the test to facilitate the
A
collection of samples based on its advantages of low infection risk, noninvasiveness of
M
sample collection, and decreased chance of sample adulteration. After optimization of
ED
the operation parameters, including reaction temperature, vibration washing time,
PT
reaction time, and antibody concentration, the resulting assay can be completed in as
CC E
little as 6 min with a detection limit of 0.03 ng/mL. Moreover, we developed an Android smartphone app to analyze and measure the results of the test, further enhancing the test’s portability, and image recording and data transmission capabilities. In order to
A
test the feasibility and performance of the optimized colorimetric assay, 90 oral fluid samples from drug abuse patients were tested. The paper-based platform features 90% sensitivity (confidence interval (CI): 76.34-97.21%) and 92% specificity (CI: 80.7797.78%) for ketamine analysis. This competitive paper-based ELISA sensing system 3
provides a rapid, convenient, sensitive, and high-throughput approach for drug monitoring.
Keywords: Abused drug analysis; paper‐based analytical device; competitive ELISA;
IP T
colorimetric assay 1. Introduction
SC R
Ketamine is an N-methyl-ᴅ-aspartate receptor antagonist that was first synthesized
in the 1960s as an anesthetic drug with dissociative, analgesic, and psychedelic
U
properties [1, 2]. Ketamine has been one of the most commonly found illicit drugs in
A
N
Asia [3-5]. In 2012, the Food and Drug Administration of Taiwan reported that 60% of
M
drug abusers younger than 19 years old were addicted to ketamine [6]. Using ketamine
ED
in subanesthetic doses can provoke imaginative, dissociative states and psychotic symptoms resembling schizophrenia, as well as severely impairing semantic and
PT
episodic memory. In order to defeat and stop the spread of drug abuse in the teenage, it
CC E
is critical to establish a rapid, precise, and accurate method of ketamine detection. Conventional ketamine analytical methods in medical centers and research labs
A
are based on pretreatment of human fluids and subsequent instrument analysis. These samples are first pretreated by solid phase extraction [7, 8], solid phase microextraction [9, 10], liquid-liquid extraction [11, 12], liquid-phase microextraction [13], and dispersive liquid-liquid microextraction [14, 15], followed by analysis using gas 4
chromatography-mass spectrometry (GC-MS) [15-17], liquid chromatography-mass spectrometry [3, 18, 19], capillary electrophoresis [20, 21], and the sensors [22, 23]. However, the complex sample preparation required when using liquid-liquid or solid-
IP T
phase extraction, the bulky size of the instrumentation, the power source requirements, and the need for well-trained technicians to operate the instruments and perform data
SC R
analysis limits the possibility using this established technique of drug analysis. There
are several lateral flow formatted drug abuse tests available on the market [24].
U
However, they still suffer from low sensitivity in detection of markers. Therefore, the
A
N
development of a high sensitivity, rapid, and high-throughput ketamine analytical
ED
and drug abuse investigations.
M
platform that could be performed in the field would greatly benefit criminal, forensic,
In order to improve the sampling experience, we chose saliva as our testing
PT
material. Conventionally, human hair, blood, urine, and oral fluid have been used for
CC E
identifying ketamine and its metabolites [2]. Among them, oral fluid is an adequate alternative medium to urine and blood for drug-testing based on its advantages of low
A
infection risk, noninvasiveness of sample collection, and decreased chance of sample adulteration [25]. Additionally, private facilities and well-trained technicians are not necessary for sampling [26]. Moreover, oral fluid reflects the recent concentration of drug usage in human plasma, which provides a better correlation with the 5
pharmacodynamic effects on the patients [27]. In order to obtain ketamine testing results in a quantitative manner, we adapted a competitive enzyme-linked immunosorbent assay (c-ELISA), which is based on a
IP T
limited number of binding sites of the chemical or pharmaceutical target [28-30]. In cELISA tests, a competitive antibody is labeled with a protein-enzyme that upon
SC R
successful binding to the antigen results in a chemical signal that can be detected and
measured, the results of which are inversely proportional to the targeted antibody in the
U
testing medium. ELISA provides good specificity, sensitivity, and stability, which has
A
N
made it a commercial (semi) quantitative tool for real sample analysis [31, 32].
M
Additionally, the signal response of c-ELISA tests can be made more sensitive in a
ED
targeted concentration range by simply modifying the dose-response curves [33]. However, the turn-around time for obtaining test results is 2–4 hours.
PT
In order to shorten the detection time for sample analysis, we combined the c-
CC E
ELISA test with a microfluidic paper ‐ based analytical device (µPAD) and a colorimetric sensing app for complex, real-time analysis of biological samples [34].
A
Such paper-ELISA (P-ELISA) sensing systems have been successfully applied on many biological samples, including cancer biomarkers [35, 36], bacterium [37], and viruses [34, 38]. By adopting c-ELISA for drug test on PADs, the cP-ELISA possess dramatic advantages for use in resource-limited settings based on the small size of the test, its 6
cost effectiveness, disposability, ease of fabrication, low sample volume requirements, wettability (which helps eliminate the need for external flow control systems), and ease of storage and delivery [39-41]. Furthermore, compared to other paper-based analytical
IP T
devices, such as dipstick or lateral/vertical flow assays, the P-ELISA test can readily obtain multiplexed and semi-quantitative colorimetric or electrochemical results that
SC R
can be observed by the naked eye, a portable microscope, or a smartphone [42-44]. Low sample volume requirement is also advantageous for persons who are not able to deliver
U
sufficient amounts of oral fluid for the test due to physiology and health reasons [45,
A
N
46].
M
In this study, we report a cP-ELISA test capable of detecting ketamine for use in
ED
road-side drug analysis in the fight against drug abuse in the teenage. We used ketamine conjugated horseradish peroxidase (HRP) to compete with ketamine found in clinical
PT
oral samples, adding 3, 3’, 5, 5’-tetramethylbenzidine (TMB) to this paper-based
CC E
platform to induce a color change that is inversely proportional to the concentration of ketamine in the sample, enabling us to generate semi-quantitative results. The cP-
A
ELISA test can be easily printed and stored prior to use, making it an ideal test for ketamine analysis. In order to further enhance liquid diffusion through the paper-fibers of the platform to hasten the reaction processes and eliminate the environmental temperature variance, 7
we also explored the use of continuous heating using a portable heater to enhance the cP-ELISA test. To further save time during analysis, we also employed ultrasonic vibration function to wash the paper-based device to help improve non-specific binding
IP T
in the testing process. Simultaneously, we developed an Android-based app for capturing and analyzing the colorimetric results of the platform, further enhancing the
SC R
test’s portability, image recording and data transmission capabilities by smartphone. As a result of these optimizations, the cP-ELISA sensing system for ketamine detection
U
features a detection limit of 0.03 ng/mL, and a single assay can be completed in as little
A
N
as 6 min. It also provides a time-saving, sensitive, and user-friendly assay for assisting
ED
2. Materials and methods
M
ketamine detection in a variety of settings.
2.1 Reagents and Materials
PT
Ketamine hydrochloride, Ketamine antibody (K-Ab), and Ketamine-HRP (K-HRP)
CC E
1x were purchased from Fitzgerald (Concord, MA). Bovine serum albumin (BSA), phosphate buffered saline (PBS), phosphate buffered saline with Tween 20 (PBS-T),
A
and TMB were purchased from Sigma-Aldrich (St. Louis, MO). All of the reagents were prepared with 1x PBS buffer and stored at 4 °C. Ultrapure water (18.2 mΩ•cm) was used throughout the experiments, which was filtered through a Milli-Q system (Millipore, Milford, MA). 8
2.2 Instrumentation The μPADs were printed on a 4 cm by 3 cm piece of cellulose paper using a Xerox colorQube 8560dn Solid Ink Color Wax Printer (Fuji, Japan). Solid wax was printed on
IP T
the paper to form 4 mm wide circular hydrophobic barriers that contained the testing spots, the interior of which remain unmodified and therefore serve as hydrophilic wells
SC R
for the samples. The printed device was then heated to 120 °C for 10 min to help the
wax penetrate the cellulose fibers. In this study, we have employed an Android
U
smartphone (HTC, Taiwan) as a portable optical imaging device for taking pictures of
A
N
the colorimetric results of the μPADs under ambient lighting without flash. Optical
M
microscopy was used to record the results for calibration using a USB digital
ED
microscope (UPG650, Upmost, Taiwan) in the lab under a controlled lighting condition. The RGB values of the images were analyzed by a self-developed Android smartphone
PT
app and ImageJ software (National Institutes of Health, Bethesda, MD).
CC E
2.3 Optimization of ketamine detection on the cP-ELISA sensing system The standard protocol of ketamine detection on the cP-ELISA sensing system
A
involved first preheating the preprinted paper-platform on a heater at 40 °C for 10 min, followed by the succesive addition of 3 μL of 88.9 μg/mL K-Ab, 3 μL of 1 wt% BSA, 3 μL of different concentrations of ketamine standard solution (10-6-102 μg/ mL), and 3 μL of 0.033x diluted K-HRP to the hydrophilic wells of the μPAD, incubating for 3 min 9
between each step at 40 °C on the heater. After the hydrophilic wells dried completely, the paper was rinsed with PBS-T buffer followed by 30 s vibration to wash the μPAD and remove non-specific binding, followed by drying at 40 °C for 3 min. Finally, 3 μL
IP T
of TMB solution was added to the hydrophilic wells of the μPAD at room temperature in order to induce the color change. After 3 min of TMB reaction, a portable optical
SC R
imaging device (microscope or smartphone) recorded the results, which we also uploaded to a cloud computing system. The images were analyzed in terms of the
U
red/blue (R/B) ratios of the testing spots based on the images’ red-green-blue (RGB)
A
N
values as determined by ImageJ software and a developed Android smartphone app.
M
The error bars correspond to the standard deviation of the sample mean.
ED
2.3.1 Determining the optimized concentrations of K-Ab, ketamine standard, and K-HRP
PT
The cP-ELISA procedure was optimized using 3 μL of different K-Ab
CC E
concentrations (44.5 μg/mL, 88.9 μg/mL, and 177.8 μg/mL), 3 μL of 1 wt% BSA to avoid non-specific binding, 3 μL of different concentrations of ketamine standard (10 -102 μg/ mL), and 3 μL of different K-HRP concentrations (diluted 0.1x (10 times
A
6
diluted), 0.033x (30 times diluted), and 0.017x (60 times diluted)) to compete with KAb in the reaction. Between each step, all of the reagents were added in succession to the hydrophilic wells of the μPADs and incubated for 3 min at 40 °C on a heater. After 10
the hydrophilic wells dried completely, the paper was rinsed with the PBS-T buffer followed by 30 s vibration to wash the μPAD platforms, followed by drying at 40 °C for 3 min. Finally, 3 μL of TMB solution was added to the μPADs at room temperature
IP T
in order to observe the color change. 2.3.2 Determining the working temperature, vibration washing, and TMB reaction
SC R
times
The cP-ELISA procedure was accompanied by 3 μL of 88.9 μg/mL K-Ab, 3 μL of
U
1 wt% BSA, 3 μL of different concentrations of ketamine standard (10-6-102 μg/ mL),
A
N
and 3 μL of 0.033x diluted K-HRP to compete with K-Ab in the reaction. Between each
M
successive addition of these components to the hydrophilic wells of the μPAD, the
ED
incubation time was set at 3 min and the platform was heated on a heater at different temperatures (30, 40, 50, and 60 °C). After the hydrophilic wells dried completely, the
PT
paper was rinsed and vibration washed in PBS-T buffer for different times (20, 25, 30,
CC E
and 40 s), and then dried on a heater for 3 min at the respective heating temperature. Finally, 3 μL of TMB solution was added to each testing well of the μPAD at room
A
temperature to observe the color change and recorded at different reaction times (1, 3, 5, and 7 min). 2.4 Real oral fluid samples analyzed for ketamine using the cP-ELISA sensing system 2.4.1 Institutional Review Board Approval 11
Human oral fluid samples used in this study were approved and issued by the institutional review board of Kaohsiung Medical University Hospital, Kaohsiung, Taiwan.
IP T
2.4.2 Clinical oral fluid samples for ketamine analysis by the cP-ELISA sensing system and GC-MS
SC R
For the μPAD sensing system, we utilized the optimized parameters determined
previously for this experiment. The clinical oral fluid samples were provided by
U
Kaohsiung Medical University Hospital and stored at -20 °C, and were defrosted before
A
N
sample pretreatment. To run the test, 3 μL of 88.9 μg/mL K-Ab, 3 μL of 1 wt% BSA, 3
M
μL of the unknown, unfiltered oral fluid sample, and 3 μL of the 0.033x diluted K-HRP
ED
were added in sequence to the hydrophilic wells of the μPAD, followed by incubation for 3 min on a 40 °C heater. After the hydrophilic wells dried completely, the paper was
PT
rinsed and vibration washed in PBS-T buffer for 30 s. The μPAD was then dried at 40
CC E
°C for 3 min. Finally, 3 μL TMB solution was added to the dried μPAD to observe the color change at room temperature, and then the images were analyzed. To compare the
A
results from the cP-ELISA sensing system, we also used a GC-MS to measure the concentration of ketamine in the oral fluid samples. 3. Results and discussion Here we present the results of a 3-year special program of drug tests in Taiwan. 12
Most commercially available ketamine tests that can be collected or performed in the field are based on lateral flow immunoassays (LFIAs) or ELISA tests, which are highly sensitive and specific for the detection of a wide variety of target analytes. However,
IP T
ELISA is time consuming, usually taking 2–4 hours to complete the assay, and while LFIA can rapidly obtain detection results, it can only provide a binary yes/no response
SC R
with limited sensitivity [47, 48]. In this study, our paper-based colorimetric sensing platform with modified detection processes provides a rapid, sensitive, specific, and
U
cost-effective approach for ketamine detection that we are proposing to overcome these
A
N
issues.
M
We have chosen to develop a cP-ELISA test system that can be analyzed using a
ED
smartphone and an app in order to obtain quasi-quantitative results in resource-limited settings, such as on the road [49, 50]. The app mainly serves two purposes: (1) taking
PT
image results of the μPADs; and (2) performing image processing for rapid testing.
CC E
Adopting smartphones as the optical imaging device for μPADs enables high portability and detection flexibility capabilities, but comes with some color display issues caused
A
by varying ambient light on the captured images. To address these color display issues, we conduct a white balance correction process, which is an essential and commonly used algorithm in image processing to eliminate color casts due to illumination [51]. The goal of this process is to adjust the intensities of the colors such that they are 13
rendered to reflect the colors viewed by the naked-eye [52]. As a result, the proposed cP-ELISA sensing system not only provides a qualitative result for the presence of ketamine, but also indicates different concentrations of the analyte via the relative color
IP T
change of the hydrophilic sample wells printed on the paper surface based on the amount of analyte present in the testing solution.
SC R
A schematic illustration of the ketamine detection system using the paper-based analytical device is shown in Scheme 1. Oral fluid samples from a suspect can be
U
collected in one of two ways. The first involves the established method of sending the
A
N
sample to a central lab for GC-MS confirmation, a timely and expensive process. The
M
alternative is our proposed ketamine detection system using the µPAD and recording
ED
the results via a smartphone or portable microscope system. The advantages of immediate quantification of the suspect’s ketamine levels are obvious.
PT
The µPAD platform utilizes a chemical test that is based on the cP-ELISA assay,
CC E
which uses an enzyme reaction to display a color change that is inversely proportional to the concentration of the target analyte. The amount of K-Ab, K-HRP, and sample
A
solution were all limited to a small volume of just 3 µL each. This makes the sensing platform more cost-effective, safe, and rapid compared to traditional ELISA and UVvisible absorption analysis, which typically requires a sample volume of 200 μL. In addition, the small sample volume requirement also results in a shorter diffusion length 14
during the color development, so the total assay time can be reduced from hours to minutes. The color of the cellulose paper, the hydrophobic barriers, and printed reference color bars are utilized to adjust lighting conditions.
IP T
The detection results are then recorded using an app for immediate and automatic analysis. The Android smartphone app was implemented based on the Linux Kernel
SC R
mobile operating system and Android Marshmallow version 6.0. Upon binding to increasing concentration of ketamine, the color of the sample wells changes from deep
U
blue to light blue. We obtained the RGB values of these images and calculated the ratios
A
N
of R/G, R/B, and G/B in order to determine which value could be used to most
M
effectively monitor the color change of the platform. Our results showed that the R/B
ED
ratio was both concentration-dependent and produced the most significant color change in our µPAD cP-ELISA sensing system with normalized deviation (Fig. 1). Therefore,
PT
we selected the R/B value for the detection of ketamine in all subsequent experiments.
CC E
3.1 The effects of K-Ab and K-HRP concentration in the cP-ELISA sensing system For c-ELISA, the fixed antibody and competitive antibody concentration are the
A
two most important factors in the colorimetric assay. The purpose of optimizing the antibody concentration was to obtain low background interference and high sensitivity for the immunoassay. When the antibody concentration is too high, the low-affinity non-specific antibody cannot be removed completely by the washing step if its 15
specificity to the target analyte is limited. This causes a higher background noise. If the antibody concentration is too low, the sensitivity of the ELISA test will not be sufficient based on the limited epitopes and signal molecules [47, 53]. In order to optimize these
IP T
conditions, we tested different K-Ab concentrations (44.5 μg/mL, 88.9 μg/mL, and 177.8 μg/mL) on the paper-based colorimetric assay (Fig. 2A-B). Our results showed
SC R
that the lowest signal variability was observed using a concentration of 88.9 μg/mL KAb. Lower or higher antibody concentration produced greater signal variation without
U
offering better binding.
A
N
We also tested the K-HRP concentration, diluted by 0.1x, 0.033x, and 0.017x (Fig.
M
2C-D). Our results showed that the 0.033x diluted K-HRP sample presented the most
ED
stable and minimal variability in the data. For the 0.1x diluted K-HRP sample, the cELISA reaction saturated rapidly, resulting in a color change that was not significant.
PT
Alternatively, the 0.017x diluted K-HRP sample demonstrated a color change in the
CC E
paper-based colorimetric assay. However, there was large variation for the sample wells that featured a low ketamine concentration (10-6 to 10-4 μg/mL) and insignificant color
A
change for the medium ketamine concentration (10-3 to 10-1 μg/mL). Overall, this cPELISA sensing system could be quantified using the R/B ratio of the acquired images to generate linear results as a function of ketamine concentration. Based on these findings, for all subsequent experiments we used the 88.9 μg/mL K-Ab concentration 16
and the 0.033x diluted K-HRP solution. 3.2 The effects of working temperature and washing time on the cP-ELISA sensing system
IP T
To increase the rate and efficiency of the sensing system, we heated the paper platform and washed it by vibration washed in PBS-T buffer to reduce the effects of
SC R
non-specific binding. Traditional ELISA tests utilize elevated incubation temperatures and time to improve the efficiency of antibody binding, however, excessive heating is
U
also known to increase the background signal [54]. The effective diffusivity is an
A
N
important factor in the immunoassay, which is also related to the incubation time used
M
to achieve the saturated limited of the test [55-57]. Based on these reasons, heating
ED
should shorten the time to process the cP-ELISA test with the potential for rapid screening of ketamine. In addition, due to the low reagent volume required for the cP-
PT
ELISA test, heating could enhance the diffusion of the antigen-antibody reaction and
CC E
enhance reagent evaporation. To find the optimum condition, we tested different working temperatures,
A
including 30, 40, 50, and 60 °C (Fig. 3A-B). Our results showed that the most stable data with the lowest standard deviation occurred at a temperature of 40 °C. At 30 °C, the test results were not linear with ketamine concentration, which we surmised was due to the antigen-antibody interaction not being stable at this lower temperature. And 17
at higher temperatures, the color change was not significant and the variability between sample replicates increased, possibly due to denaturing of the antibody. Using 40 °C for the working temperature is also consistent with the baseline temperature of the
IP T
human immune system. Additionally, according to the Stokes-Einstein equation, incubating at the appropriate temperature increases the diffusion of intermolecular
SC R
collision in the ELISA reaction, thus providing a rapid immunoassay [55].
U
where D is the diffusion constant, KB is the Boltzmann's constant, T is the absolute
A
N
temperature, η is the dynamic viscosity and r is the radius of the spherical particle.
M
Washing is another important procedure in the ELISA test because it helps to
ED
decrease signal interference and raises the sensitivity of the assay by reducing nonspecific binding. Therefore, we tested different washing times by vibration washed the
PT
paper platform for 20, 25, 30, and 40 s in PBS-T buffer (Fig. 3C-D). Our results show
CC E
that an optimum washing time of 30 s achieves the most stable and minimal standard deviation in the colorimetric data. The results were similar for the 20 and 25 s washed
A
samples, but there was less significant color change in the cP-ELISA test. In addition, the results of washing for 40 s resulted in low color change and did not demonstrate good linearity for ketamine detection. We hypothesize that too long of a washing time could result in removal of the antigen-antibody complex, resulting in poor signal 18
detection. Based on these results, for all subsequent experiments we used a 40 °C working temperature and 30 s washing time for the cP-ELISA test. 3.3 Effect of TMB reaction time in the cP-ELISA sensing system
IP T
The enzyme-substrate reaction is the final step in the ELISA test, which is used to produce a color change for measurement. We studied various TMB reaction times,
SC R
including 1, 3, 5, and 7 min, in order to determine which produced the largest and most
linear color development as a function of ketamine concentration (Fig. 3E-F). Our
U
results showed the enzyme-substrate reaction at 3 min resulted in a significant color
A
N
change. The R/B ratio did not change significantly at 1 min and 7 min, suggesting the
M
reaction was just beginning at 1 min and became saturated at 7 min. When the reaction
ED
time reaches 3 min, a highest color variance can be observed for a lower detection limit. Based on these results, for all subsequent experiments we selected 3 min for the TMB
PT
reaction time.
CC E
3.4 Standard curve for ketamine detection in the cP-ELISA sensing system Based on the prior experiments, we determined the optimum parameters for the
A
cP-ELISA sensing system consisted of 88.9 μg/mL K-Ab, K-HRP 0.033x diluted, a 40 °C working temperature, 30 s washing time, and 3 min TMB reaction. Using these conditions, we established a standard curve of ketamine (blank, 10-6-102 μg/mL) in an oral fluid sample that had been pretreated by filtration through a 0.22 μm membrane, 19
as well as a corresponding dataset in oral fluid without pretreatment, in order to more closely model clinical trials with limited processes (Fig. 4). The linear range of the oral fluid sample without filtering pretreatment was from 10-3–100 μg/mL (R/B ratio from
IP T
0.62 to 0.94), which corresponded to a limit of detection (LOD) of 0.3 ng/mL and an R2 value of 0.98. On the other hand, using the 0.22 μm membrane to filter the oral fluid
SC R
sample resulted in a linear range of 10-4–10-1 μg/mL (R/B ratio from 0.59 to 0.84), and
LOD of 0.03 ng/mL with an R2 value of 0.98 (based on the concentration of the test that
U
produced a signal that was three times greater than the baseline noise signal of the
A
N
control; LOD = yblank + 3SDblank). We surmised that the unfiltered oral fluid sample had
M
some interference or non-specific binding to cause the R/B ratio to increase. However,
ED
in order to achieve the convenience of fast screening, using an unfiltered oral fluid sample would help accelerate sample analysis and save time.
PT
3.5 Clinical application for cP-ELISA sensing system
CC E
In order to demonstrate the cP-ELISA sensing system has the potential for analysis of ketamine, we analyzed 90 clinical samples, 40 of which were positive for ketamine
A
and 50 that were negative, though we were unaware of these values at the time of analysis. In each test, the clinical samples were compared to a blank oral fluid sample given by a volunteer. The cut-off point and receiver operating characteristic (ROC) curve of our cP-ELISA sensing system found 36 samples that tested positive for 20
ketamine and 46 that tested negative, which is 90% sensitivity (Confidence Interval (CI): 76.34-97.21%) and 92% specificity (CI: 80.77-97.78%) (Fig. 5). GC-MS testing of the samples also supported our cP-ELISA sensing results.
IP T
The traditional ELISA procedure requires waiting for antigen-antibody binding, incubation, and washing time, which takes at least 2–4 hours in total. Herein, our µPAD
SC R
c-ELISA sensing system, which can measure multiple samples in parallel, requires just
minutes to complete the test. This cP-ELISA sensing system also conforms to the point-
U
of-care testing standards, and provides a rapid, cheap, and simple method of ketamine
A
N
analysis [58].
M
4. Conclusions
ED
In this study, we used a microfluidic paper-based analytical device with the modified detection processes to test saliva for ketamine in an ultrafast manner. One of
PT
the advantages of using oral fluid is that the process of sampling saliva is non-invasive,
CC E
which makes testing in the field easier and faster. A key potential advantage of the proposed technology is the ability to simultaneously and effectively measure low
A
concentrations of multiple drug compounds in oral fluid without the need for a large amount of samples, complex equipment, or long processing times. This research provides a new drug testing platform for high-throughput drug abuse monitoring. Preliminary results show that the LOD of this cP-ELISA sensing system 21
for ketamine is 0.03 ng/mL, which is lower than the 100 ng/mL legal limit in Taiwan [59]. This cP-ELISA sensing system lowers the uncertainty of testing samples by heating and washing the platform. Comparing to other P-ELISA tests, the detection
IP T
time can be effectively shortened from ~ 30 min to 6 min [34, 37]. It was also only requires a few μL of oral fluid samples. Compared with the traditional ELISA test, this
SC R
sensing system improves the sensitivity and specificity for ketamine analysis. In
addition, the analysis can be performed using a smartphone app under resource-
U
restricted conditions and does not require expensive instrumentation. The cP-ELISA
A
N
sensing system offers a high quality and fast testing platform that could be expanded to
M
an array of drugs beyond ketamine, including amphetamines, cocaine, opioids, etc. We
ED
expect the findings from this work can be used to build cost-effective and portable approaches for highly sensitive and specific drug analysis and other biomedical ELISA
CC E
PT
tests in resource limited environments.
Acknowledgment
A
This research was supported by the Ministry of Science and Technology, Taiwan (1062221-E-002-139-MY2 & 104-2218-E-002-039), Kaohsiung Medical University Chung-Ho Memorial Hospital (KMUH105-5T58), and the Higher Education Sprout Program at National Taiwan University. 22
References
IP T
[1] L. Li, P.E. Vlisides, Ketamine: 50 Years of Modulating the Mind, Frontiers in Human Neuroscience, 10(2016) 612. [2] E.F. Göktaş, F. Arıöz, A review of chromatographic methods for ketamine and its metabolites norketamine and dehydronorketamine, Biomedical chromatography : BMC, (2017) e4014. [3] K.W. Leung, Z.C.F. Wong, J.Y.M. Ho, A.W.S. Yip, J.S.C. Ng, S.P.H. Ip, et al., Determination of hair ketamine cut-off value from Hong Kong ketamine users by LC–
SC R
MS/MS analysis, Forensic Science International, 259(2016) 53-8. [4] G.L. Yang, A.D. Zhang, Y. Yu, H. Liu, F.Y. Long, J. Yan, Drug use and its associated factors among money boys in Hunan Province, China, Public Health, 140(2016) 213-
M
A
N
U
20. [5] C.-A. Yang, H.-C. Liu, D.-L. Lin, R.H. Liu, Y.-Z. Hsieh, S.-P. Wu, Simultaneous Quantitation of Methamphetamine, Ketamine, Opiates and their Metabolites in Urine by SPE and LC–MS-MS, Journal of Analytical Toxicology, 41(2017) 679-87. [6] D. Favretto, S. Vogliardi, M. Tucci, I. Simoncello, R. El Mazloum, R. Snenghi, Occupational exposure to ketamine detected by hair analysis: a retrospective and prospective toxicological study, Forensic Science International, 265(2016) 193-9.
CC E
PT
ED
[7] M. Sergi, D. Compagnone, R. Curini, G. D’Ascenzo, M. Del Carlo, S. Napoletano, et al., Micro-solid phase extraction coupled with high-performance liquid chromatography–tandem mass spectrometry for the determination of stimulants, hallucinogens, ketamine and phencyclidine in oral fluids, Analytica Chimica Acta, 675(2010) 132-7. [8] A.S. Barreto, V.F. Brant, E. Spinelli, S.V. Rodrigues, Validation of a SPE-LC–MS/MS method for the determination of ketamine and norketamine in micropulverized hair after a single IV dose, Journal of Chromatography B, 1033(2016) 200-9. [9] F. Bianchi, F. Bisceglie, S. Dugheri, G. Arcangeli, V. Cupelli, E. del Borrello, et al.,
A
Ionic liquid-based solid phase microextraction necklaces for the environmental monitoring of ketamine, Journal of Chromatography A, 1331(2014) 1-9. [10] S. Gentili, C. Mortali, L. Mastrobattista, P. Berretta, S. Zaami, Determination of different recreational drugs in sweat by headspace solid-phase microextraction gas chromatography mass spectrometry (HS-SPME GC/MS): Application to drugged drivers, Journal of Pharmaceutical and Biomedical Analysis, 129(2016) 282-7. [11] F.A. Sandbaumhüter, R. Theurillat, R.N. Bektas, A.P.N. Kutter, R. BettschartWolfensberger, W. Thormann, Pharmacokinetics of ketamine and three metabolites 23
in Beagle dogs under sevoflurane vs. medetomidine comedication assessed by
IP T
enantioselective capillary electrophoresis, Journal of Chromatography A, 1467(2016) 436-44. [12] E. Bonfanti, F. Cosnier, L. Wathier, P. Campo, Measurement of ketamine and xylazine in rat brain by liquid–liquid extraction and gas chromatography–mass spectrometry, Journal of Pharmacological and Toxicological Methods, 77(2016) 6-9. [13] Y. Yamini, S. Seidi, R. Feizbakhsh, T. Baheri, M. Rezazadeh, Liquid-phase microextraction based on two immiscible organic solvents followed by gas chromatography with mass spectrometry as an efficient method for the preconcentration and determination of cocaine, ketamine, and lidocaine in human urine samples, Journal of Separation Science, 37(2014) 2364-71.
SC R
[14] L. Meng, B. Wang, F. Luo, G. Shen, Z. Wang, M. Guo, Application of dispersive liquid–liquid microextraction and CE with UV detection for the chiral separation and determination of the multiple illicit drugs on forensic samples, Forensic Science
M
A
N
U
International, 209(2011) 42-7. [15] Z. Lin, J. Li, X. Zhang, M. Qiu, Z. Huang, Y. Rao, Ultrasound-assisted dispersive liquid-liquid microextraction for the determination of seven recreational drugs in human whole blood using gas chromatography–mass spectrometry, Journal of Chromatography B, 1046(2017) 177-84. [16] P.-S. Cheng, C.-Y. Fu, C.-H. Lee, C. Liu, C.-S. Chien, GC–MS quantification of ketamine, norketamine, and dehydronorketamine in urine specimens and
CC E
PT
ED
comparative study using ELISA as the preliminary test methodology, Journal of Chromatography B, 852(2007) 443-9. [17] I. Moreno, M. Barroso, A. Martinho, A. Cruz, E. Gallardo, Determination of ketamine and its major metabolite, norketamine, in urine and plasma samples using microextraction by packed sorbent and gas chromatography-tandem mass spectrometry, Journal of Chromatography B, 1004(2015) 67-78. [18] F. Vaiano, F.P. Busardò, D. Palumbo, C. Kyriakou, A. Fioravanti, V. Catalani, et al., A novel screening method for 64 new psychoactive substances and 5 amphetamines in blood by LC–MS/MS and application to real cases, Journal of Pharmaceutical and
A
Biomedical Analysis, 129(2016) 441-9. [19] M. Hasan, R. Hofstetter, G.M. Fassauer, A. Link, W. Siegmund, S. Oswald, Quantitative chiral and achiral determination of ketamine and its metabolites by LC– MS/MS in human serum, urine and fecal samples, Journal of Pharmaceutical and Biomedical Analysis, 139(2017) 87-97. [20] Y.-H. Ho, C.-C. Wang, Y.-T. Hsiao, W.-K. Ko, S.-M. Wu, Analysis of ten abused drugs in urine by large volume sample stacking–sweeping capillary electrophoresis with an experimental design strategy, Journal of Chromatography A, 1295(2013) 136-41. 24
[21] N. Porpiglia, G. Musile, F. Bortolotti, E.F. De Palo, F. Tagliaro, Chiral separation
IP T
and determination of ketamine and norketamine in hair by capillary electrophoresis, Forensic Science International, 266(2016) 304-10. [22] A. Argente-García, N. Jornet-Martínez, R. Herráez-Hernández, P. Campíns-Falcó, A passive solid sensor for in-situ colorimetric estimation of the presence of ketamine in illicit drug samples, Sensors and Actuators B: Chemical, 253(2017) 1137-44. [23] B. Deiminiat, G.H. Rounaghi, Fabrication of a new electrochemical imprinted sensor for determination of ketamine based on modified polytyramine/sol-gel/fMWCNTs@AuNPs nanocomposite/pencil graphite electrode, Sensors and Actuators B: Chemical, 259(2018) 133-41. [24] G.A. Posthuma-Trumpie, J. Korf, A. van Amerongen, Lateral flow (immuno)assay:
SC R
its strengths, weaknesses, opportunities and threats. A literature survey, Analytical and Bioanalytical Chemistry, 393(2009) 569-82. [25] R. Dams, R.E. Choo, W.E. Lambert, H. Jones, M.A. Huestis, Oral fluid as an
M
A
N
U
alternative matrix to monitor opiate and cocaine use in substance-abuse treatment patients, Drug & Alcohol Dependence, 87(2007) 258-67. [26] O.R. Idowu, B. Caddy, A Review of the Use of Saliva in the Forensic Detection of Drugs and Other Chemicals, Journal of the Forensic Science Society, 22(1982) 123-35. [27] A.G. Verstraete, Oral fluid testing for driving under the influence of drugs: history, recent progress and remaining challenges, Forensic Science International, 150(2005) 143-50.
CC E
PT
ED
[28] X. Pei, B. Zhang, J. Tang, B. Liu, W. Lai, D. Tang, Sandwich-type immunosensors and immunoassays exploiting nanostructure labels: A review, Analytica Chimica Acta, 758(2013) 1-18. [29] A. Makaraviciute, A. Ramanaviciene, Site-directed antibody immobilization techniques for immunosensors, Biosensors and Bioelectronics, 50(2013) 460-71. [30] Z. Wang, S. Zong, L. Wu, D. Zhu, Y. Cui, SERS-Activated Platforms for Immunoassay: Probes, Encoding Methods, and Applications, Chemical Reviews, 117(2017) 7910-63. [31] R. Hu, T. Liu, X.-B. Zhang, Y. Yang, T. Chen, C. Wu, et al., DLISA: A DNAzyme-Based
A
ELISA for Protein Enzyme-Free Immunoassay of Multiple Analytes, Analytical Chemistry, 87(2015) 7746-53. [32] Y. Li, X. Ma, Z. Xu, M. Liu, Z. Lin, B. Qiu, et al., Multicolor ELISA based on alkaline phosphatase-triggered growth of Au nanorods, Analyst, 141(2016) 2970-6. [33] B.M. Murphy, X. He, D. Dandy, C.S. Henry, Competitive Immunoassays for Simultaneous Detection of Metabolites and Proteins Using Micromosaic Patterning, Analytical Chemistry, 80(2008) 444-50. [34] C.-M. Cheng, A.W. Martinez, J. Gong, C.R. Mace, S.T. Phillips, E. Carrilho, et al., 25
Paper-Based ELISA, Angewandte Chemie International Edition, 49(2010) 4771-4.
IP T
[35] S. Wang, L. Ge, X. Song, J. Yu, S. Ge, J. Huang, et al., Paper-based chemiluminescence ELISA: Lab-on-paper based on chitosan modified paper device and wax-screen-printing, Biosensors and Bioelectronics, 31(2012) 212-8. [36] M.-Y. Hsu, Y.-C. Hung, D.-K. Hwang, S.-C. Lin, K.-H. Lin, C.-Y. Wang, et al., Detection of aqueous VEGF concentrations before and after intravitreal injection of anti-VEGF antibody using low-volume sampling paper-based ELISA, Scientific Reports, 6(2016) 34631. [37] C.-M. Shih, C.-L. Chang, M.-Y. Hsu, J.-Y. Lin, C.-M. Kuan, H.-K. Wang, et al., Paperbased ELISA to rapidly detect Escherichia coli, Talanta, 145(2015) 2-5. [38] H.-K. Wang, C.-H. Tsai, K.-H. Chen, C.-T. Tang, J.-S. Leou, P.-C. Li, et al., Cellulose-
SC R
Based Diagnostic Devices for Diagnosing Serotype-2 Dengue Fever in Human Serum, Advanced Healthcare Materials, 3(2014) 187-96. [39] A.W. Martinez, S.T. Phillips, M.J. Butte, G.M. Whitesides, Patterned paper as a
M
A
N
U
platform for inexpensive, low-volume, portable bioassays, Angewandte ChemieInternational Edition, 46(2007) 1318-20. [40] A.W. Martinez, S.T. Phillips, G.M. Whitesides, E. Carrilho, Diagnostics for the Developing World: Microfluidic Paper-Based Analytical Devices, Analytical Chemistry, 82(2010) 3-10. [41] G.H. Chen, W.Y. Chen, Y.C. Yen, C.W. Wang, H.T. Chang, C.F. Chen, Detection of Mercury(II) Ions Using Colorimetric Gold Nanoparticles on Paper-Based Analytical
CC E
PT
ED
Devices, Analytical Chemistry, 86(2014) 6843-9. [42] E.M. Spyrou, D.P. Kalogianni, S.S. Tragoulias, P.C. Ioannou, T.K. Christopoulos, Digital camera and smartphone as detectors in paper-based chemiluminometric genotyping of single nucleotide polymorphisms, Analytical and Bioanalytical Chemistry, 408(2016) 7393-402. [43] S.M. Russell, A. Domenech-Sanchez, R. de la Rica, Augmented Reality for RealTime Detection and Interpretation of Colorimetric Signals Generated by Paper-Based Biosensors, ACS sensors, 2(2017) 848-53. [44] J. Narang, N. Malhotra, C. Singhal, A. Mathur, D. Chakraborty, A. Anil, et al., Point
A
of care with micro fluidic paper based device integrated with nano zeolite–graphene oxide nanoflakes for electrochemical sensing of ketamine, Biosensors and Bioelectronics, 88(2017) 249-57. [45] Y.Y. Xia, J. Si, Z.Y. Li, Fabrication techniques for microfluidic paper-based analytical devices and their applications for biological testing: A review, Biosensors & Bioelectronics, 77(2016) 774-89. [46] Y. Yang, E. Noviana, M.P. Nguyen, B.J. Geiss, D.S. Dandy, C.S. Henry, Paper-Based Microfluidic Devices: Emerging Themes and Applications, Analytical Chemistry, 26
89(2017) 71-91.
IP T
[47] Y. Hayashi, R. Matsuda, T. Maitani, K. Imai, W. Nishimura, K. Ito, et al., Precision, Limit of Detection and Range of Quantitation in Competitive ELISA, Analytical Chemistry, 76(2004) 1295-301. [48] M.-H. Huang, M.-Y. Wu, C.-H. Wu, J.-L. Tsai, H.-H. Lee, R.H. Liu, Performance characteristics of ELISAs for monitoring ketamine exposure, Clinica Chimica Acta, 379(2007) 59-65. [49] Q. Mei, H. Jing, Y. Li, W. Yisibashaer, J. Chen, B. Nan Li, et al., Smartphone based visual and quantitative assays on upconversional paper sensor, Biosensors and Bioelectronics, 75(2016) 427-32. [50] M.M. Gong, D. Sinton, Turning the Page: Advancing Paper-Based Microfluidics
SC R
for Broad Diagnostic Application, Chemical Reviews, 117(2017) 8447-80. [51] G.D. Finlayson, S.D. Hordley, P.M. Hubel, Color by correlation: A simple, unifying framework for color constancy, IEEE Transactions on Pattern Analysis and Machine
M
A
N
U
Intelligence, 23(2001) 1209-21. [52] J.Y. Huo, Y.L. Chang, J. Wang, X.X. Wei, Robust automatic white balance algorithm using gray color points in images, Ieee Transactions on Consumer Electronics, 52(2006) 541-6. [53] M. Palmieri, M. Vagnini, L. Pitzurra, P. Rocchi, B.G. Brunetti, A. Sgamellotti, et al., Development of an analytical protocol for a fast, sensitive and specific protein recognition in paintings by enzyme-linked immunosorbent assay (ELISA), Analytical
CC E
PT
ED
and Bioanalytical Chemistry, 399(2011) 3011-23. [54] E.T. Thean, B.H. Toh, Western immunoblotting: Temperature-dependent reduction in background staining, Analytical Biochemistry, 177(1989) 256-8. [55] M. Stenberg, H. Nygren, A diffusion limited reaction theory for a solid-phase immunoassay, Journal of Theoretical Biology, 113(1985) 589-97. [56] M. Stenberg, M. Werthén, S. Theander, H. Nygren, A diffusion limited reaction theory for a microtiter plate assay, Journal of Immunological Methods, 112(1988) 239. [57] D. Rath, S. Panda, Contribution of rotational diffusivity towards the transport of
A
antigens in heterogeneous immunosensors, Analyst, 140(2015) 6579-87. [58] C.P. Price, Point of care testing, BMJ : British Medical Journal, 322(2001) 1285-8. [59] H.-Y. Liao, J.-H. Chen, J.-J. Shyue, C.-T. Shun, H.-W. Chen, S.-W. Liao, et al., Rapid label-free determination of ketamine in whole blood using secondary ion mass spectrometry, Talanta, 143(2015) 50-5.
27
Figure captions: Scheme 1. A schematic illustration of the proposed cP-ELISA sensing system for drug abuse tests. The paper-based platform should enhance the speed and lower the cost of
IP T
ketamine analysis.
SC R
Fig. 1. Image analysis of the cP-ELISA sensing system using the red (R), green (G), and blue (B) color values of images taken of the paper platform (N = 5). Colorimetric
U
results of ketamine analysis on (A) the paper platform, (B) the corresponding R, G, B,
A
N
and gray values, and (C) the R/G, R/B, and G/B ratios of ketamine analysis were
M
determined by ImageJ software. The concentration of K-Ab and K-HRP were 88.9
ED
μg/mL and 0.033x diluted, respectively. The heating temperature was 40 °C and the
PT
washing time was 30 s. TMB reacted for 3 min to induce the color change.
CC E
Fig. 2. Effect of the concentration of K-Ab and K-HRP in the cP-ELISA sensing platform (N = 5). The colorimetric results of ketamine analysis based on (A) the paper
A
platform and (B) the corresponding R/B ratios for samples using 44.5 μg/mL, 88.9 μg/mL, and 177.8 μg/mL of K-Ab and 0.033x diluted K-HRP. Different K-HRP concentrations were also tested, ranging from 0.1x, 0.033x, and 0.017x diluted, on (C) paper and (D) the µPAD cP-ELISA sensing system. The heating temperature was set at 28
40 °C. The paper platforms were also washed by vibration washed for 30 s, and TMB was allowed to react for 3 min to induce the color change.
IP T
Fig. 3. Effect of the heating temperature, washing time and TMB reaction time of the cP-ELISA sensing system (N = 5). The colorimetric results of ketamine analysis based
SC R
on (A) the paper platform and (B) the corresponding R/B ratios for 30, 40, 50, and 60 °C antibody incubation temperatures. Different washing times ranging from 20, 25, 30,
U
and 40 s were tested on (C) the paper platform, with (D) the corresponding R/B ratios
A
N
of the µPAD cP-ELISA sensing system also shown. The concentration of K-Ab and K-
M
HRP were set at 88.9 μg/mL and 0.033x diluted, respectively. TMB was reacted for 3
ED
min to induce the color change. (E)(F) The colorimetric results of ketamine analysis of the corresponding R/B ratios and the paper platform for reaction times of 1, 3, 5, and 7
PT
min. The concentration of K-Ab and K-HRP were set at 88.9 μg/mL and 0.033x diluted,
CC E
respectively. The heating temperature was 40 °C and the washing time was 30 s.
A
Fig. 4. The standard ketamine curve in oral fluid with and without pre-treatment by a 0.22 μm membrane using the cP-ELISA sensing system (N = 5). The linear range of the oral fluid sample without filtering pretreatment was from 10-3–100 mg/mL, which corresponded to a LOD of 0.3 ng/mL. On the other hand, using the 0.22 μm membrane 29
to filter the oral fluid sample resulted in a linear range of 10-4–10-1 mg/mL and an LOD of 0.03 ng/mL.
IP T
Fig. 5. Statistical results of (A) the cut-off point and (B) ROC curve from the cP-ELISA sensing system for clinical ketamine analysis. The cut-off value was 0.04, which
SC R
resulted in a sensitivity and specificity for the platform of 90% (CI: 76.34-97.21%) and
A
CC E
PT
ED
M
A
N
U
92% (CI: 80.77-97.78%), respectively.
30
A ED
PT
CC E
IP T
SC R
U
N
A
M
Scheme 1.
31
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 1.
32
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 2.
33
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 3.
34
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 4.
35
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 5.
36
S0925-4005(18)32026-4 https://doi.org/10.1016/j.snb.2018.11.071 SNB 25669
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
18 July 2018 30 October 2018 14 November 2018
Please cite this article as: Chen C-An, Wang P-Wei, Yen Y-Chun, Lin H-Lan, Fan Y-Chung, Wu S-Mei, Chen C-Fu, Fast Analysis of Ketamine Using a Colorimetric Immunosorbent Assay on a Paper-Based Analytical Device, Sensors and amp; Actuators: B. Chemical (2018), https://doi.org/10.1016/j.snb.2018.11.071 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Fast Analysis of Ketamine Using a Colorimetric Immunosorbent Assay on a Paper-Based Analytical Device
SC R
Fanc,*, Shou-Mei Wud, and Chien-Fu Chena,*
IP T
Chung-An Chena,1, Peng-Wei Wangb,1, Yu-Chun Yena, Hsin-Lan Lina, Yao-Chung
Institute of Applied Mechanics, National Taiwan University, Taipei, Taiwan
b
Department of Psychiatry, Kaohsiung Medical University Chung Ho Memorial
U
a
A
Department of Computer Science and Engineering, National Chung Hsing
ED
University, Taichung, Taiwan
M
c
N
Hospital, Kaohsiung, Taiwan
Food and Drug Administration, Ministry of Health and Welfare, Taipei, Taiwan
1
These authors contribute equally to this work.
CC E
PT
d
* Corresponding Author
A
Chien-Fu Chen, E-mail: [email protected] Yao-Chung Fan, E-mail: [email protected]
1
SC R
IP T
Graphical abstract
A rapid colorimetric sensing system using competitive ELISA test on a
N
•
U
Highlights:
Oral fluid was selected for the test to facilitate the collection of samples based on
ED
•
M
ketamine, a frequently abused drug.
A
microfluidic paper-based analytical device was investigated for the detection of
PT
its advantages of low infection risk, noninvasiveness of sample collection, and decreased chance of sample adulteration.
CC E
•
An Android smartphone app is developed to analyze and measure the results of
A
the test, further enhancing the test’s portability, and image recording and data
•
transmission capabilities. After optimization of the operation parameters, including reaction temperature, vibration washing time, reaction time, and antibody concentration, the resulting assay can be completed in as little as 6 min with a detection limit of 0.03 ng/mL. 2
•
The paper-based platform features 90% sensitivity (confidence interval (CI): 76.34-97.21%) and 92% specificity (CI: 80.77-97.78%) for ketamine analysis of
IP T
90 oral fluid samples from drug abuse patients.
SC R
Abstract
In this study, we designed a rapid colorimetric sensing system using competitive
U
ELISA test on a microfluidic paper-based analytical device for the detection of
N
ketamine, a frequently abused drug. Oral fluid was selected for the test to facilitate the
A
collection of samples based on its advantages of low infection risk, noninvasiveness of
M
sample collection, and decreased chance of sample adulteration. After optimization of
ED
the operation parameters, including reaction temperature, vibration washing time,
PT
reaction time, and antibody concentration, the resulting assay can be completed in as
CC E
little as 6 min with a detection limit of 0.03 ng/mL. Moreover, we developed an Android smartphone app to analyze and measure the results of the test, further enhancing the test’s portability, and image recording and data transmission capabilities. In order to
A
test the feasibility and performance of the optimized colorimetric assay, 90 oral fluid samples from drug abuse patients were tested. The paper-based platform features 90% sensitivity (confidence interval (CI): 76.34-97.21%) and 92% specificity (CI: 80.7797.78%) for ketamine analysis. This competitive paper-based ELISA sensing system 3
provides a rapid, convenient, sensitive, and high-throughput approach for drug monitoring.
Keywords: Abused drug analysis; paper‐based analytical device; competitive ELISA;
IP T
colorimetric assay 1. Introduction
SC R
Ketamine is an N-methyl-ᴅ-aspartate receptor antagonist that was first synthesized
in the 1960s as an anesthetic drug with dissociative, analgesic, and psychedelic
U
properties [1, 2]. Ketamine has been one of the most commonly found illicit drugs in
A
N
Asia [3-5]. In 2012, the Food and Drug Administration of Taiwan reported that 60% of
M
drug abusers younger than 19 years old were addicted to ketamine [6]. Using ketamine
ED
in subanesthetic doses can provoke imaginative, dissociative states and psychotic symptoms resembling schizophrenia, as well as severely impairing semantic and
PT
episodic memory. In order to defeat and stop the spread of drug abuse in the teenage, it
CC E
is critical to establish a rapid, precise, and accurate method of ketamine detection. Conventional ketamine analytical methods in medical centers and research labs
A
are based on pretreatment of human fluids and subsequent instrument analysis. These samples are first pretreated by solid phase extraction [7, 8], solid phase microextraction [9, 10], liquid-liquid extraction [11, 12], liquid-phase microextraction [13], and dispersive liquid-liquid microextraction [14, 15], followed by analysis using gas 4
chromatography-mass spectrometry (GC-MS) [15-17], liquid chromatography-mass spectrometry [3, 18, 19], capillary electrophoresis [20, 21], and the sensors [22, 23]. However, the complex sample preparation required when using liquid-liquid or solid-
IP T
phase extraction, the bulky size of the instrumentation, the power source requirements, and the need for well-trained technicians to operate the instruments and perform data
SC R
analysis limits the possibility using this established technique of drug analysis. There
are several lateral flow formatted drug abuse tests available on the market [24].
U
However, they still suffer from low sensitivity in detection of markers. Therefore, the
A
N
development of a high sensitivity, rapid, and high-throughput ketamine analytical
ED
and drug abuse investigations.
M
platform that could be performed in the field would greatly benefit criminal, forensic,
In order to improve the sampling experience, we chose saliva as our testing
PT
material. Conventionally, human hair, blood, urine, and oral fluid have been used for
CC E
identifying ketamine and its metabolites [2]. Among them, oral fluid is an adequate alternative medium to urine and blood for drug-testing based on its advantages of low
A
infection risk, noninvasiveness of sample collection, and decreased chance of sample adulteration [25]. Additionally, private facilities and well-trained technicians are not necessary for sampling [26]. Moreover, oral fluid reflects the recent concentration of drug usage in human plasma, which provides a better correlation with the 5
pharmacodynamic effects on the patients [27]. In order to obtain ketamine testing results in a quantitative manner, we adapted a competitive enzyme-linked immunosorbent assay (c-ELISA), which is based on a
IP T
limited number of binding sites of the chemical or pharmaceutical target [28-30]. In cELISA tests, a competitive antibody is labeled with a protein-enzyme that upon
SC R
successful binding to the antigen results in a chemical signal that can be detected and
measured, the results of which are inversely proportional to the targeted antibody in the
U
testing medium. ELISA provides good specificity, sensitivity, and stability, which has
A
N
made it a commercial (semi) quantitative tool for real sample analysis [31, 32].
M
Additionally, the signal response of c-ELISA tests can be made more sensitive in a
ED
targeted concentration range by simply modifying the dose-response curves [33]. However, the turn-around time for obtaining test results is 2–4 hours.
PT
In order to shorten the detection time for sample analysis, we combined the c-
CC E
ELISA test with a microfluidic paper ‐ based analytical device (µPAD) and a colorimetric sensing app for complex, real-time analysis of biological samples [34].
A
Such paper-ELISA (P-ELISA) sensing systems have been successfully applied on many biological samples, including cancer biomarkers [35, 36], bacterium [37], and viruses [34, 38]. By adopting c-ELISA for drug test on PADs, the cP-ELISA possess dramatic advantages for use in resource-limited settings based on the small size of the test, its 6
cost effectiveness, disposability, ease of fabrication, low sample volume requirements, wettability (which helps eliminate the need for external flow control systems), and ease of storage and delivery [39-41]. Furthermore, compared to other paper-based analytical
IP T
devices, such as dipstick or lateral/vertical flow assays, the P-ELISA test can readily obtain multiplexed and semi-quantitative colorimetric or electrochemical results that
SC R
can be observed by the naked eye, a portable microscope, or a smartphone [42-44]. Low sample volume requirement is also advantageous for persons who are not able to deliver
U
sufficient amounts of oral fluid for the test due to physiology and health reasons [45,
A
N
46].
M
In this study, we report a cP-ELISA test capable of detecting ketamine for use in
ED
road-side drug analysis in the fight against drug abuse in the teenage. We used ketamine conjugated horseradish peroxidase (HRP) to compete with ketamine found in clinical
PT
oral samples, adding 3, 3’, 5, 5’-tetramethylbenzidine (TMB) to this paper-based
CC E
platform to induce a color change that is inversely proportional to the concentration of ketamine in the sample, enabling us to generate semi-quantitative results. The cP-
A
ELISA test can be easily printed and stored prior to use, making it an ideal test for ketamine analysis. In order to further enhance liquid diffusion through the paper-fibers of the platform to hasten the reaction processes and eliminate the environmental temperature variance, 7
we also explored the use of continuous heating using a portable heater to enhance the cP-ELISA test. To further save time during analysis, we also employed ultrasonic vibration function to wash the paper-based device to help improve non-specific binding
IP T
in the testing process. Simultaneously, we developed an Android-based app for capturing and analyzing the colorimetric results of the platform, further enhancing the
SC R
test’s portability, image recording and data transmission capabilities by smartphone. As a result of these optimizations, the cP-ELISA sensing system for ketamine detection
U
features a detection limit of 0.03 ng/mL, and a single assay can be completed in as little
A
N
as 6 min. It also provides a time-saving, sensitive, and user-friendly assay for assisting
ED
2. Materials and methods
M
ketamine detection in a variety of settings.
2.1 Reagents and Materials
PT
Ketamine hydrochloride, Ketamine antibody (K-Ab), and Ketamine-HRP (K-HRP)
CC E
1x were purchased from Fitzgerald (Concord, MA). Bovine serum albumin (BSA), phosphate buffered saline (PBS), phosphate buffered saline with Tween 20 (PBS-T),
A
and TMB were purchased from Sigma-Aldrich (St. Louis, MO). All of the reagents were prepared with 1x PBS buffer and stored at 4 °C. Ultrapure water (18.2 mΩ•cm) was used throughout the experiments, which was filtered through a Milli-Q system (Millipore, Milford, MA). 8
2.2 Instrumentation The μPADs were printed on a 4 cm by 3 cm piece of cellulose paper using a Xerox colorQube 8560dn Solid Ink Color Wax Printer (Fuji, Japan). Solid wax was printed on
IP T
the paper to form 4 mm wide circular hydrophobic barriers that contained the testing spots, the interior of which remain unmodified and therefore serve as hydrophilic wells
SC R
for the samples. The printed device was then heated to 120 °C for 10 min to help the
wax penetrate the cellulose fibers. In this study, we have employed an Android
U
smartphone (HTC, Taiwan) as a portable optical imaging device for taking pictures of
A
N
the colorimetric results of the μPADs under ambient lighting without flash. Optical
M
microscopy was used to record the results for calibration using a USB digital
ED
microscope (UPG650, Upmost, Taiwan) in the lab under a controlled lighting condition. The RGB values of the images were analyzed by a self-developed Android smartphone
PT
app and ImageJ software (National Institutes of Health, Bethesda, MD).
CC E
2.3 Optimization of ketamine detection on the cP-ELISA sensing system The standard protocol of ketamine detection on the cP-ELISA sensing system
A
involved first preheating the preprinted paper-platform on a heater at 40 °C for 10 min, followed by the succesive addition of 3 μL of 88.9 μg/mL K-Ab, 3 μL of 1 wt% BSA, 3 μL of different concentrations of ketamine standard solution (10-6-102 μg/ mL), and 3 μL of 0.033x diluted K-HRP to the hydrophilic wells of the μPAD, incubating for 3 min 9
between each step at 40 °C on the heater. After the hydrophilic wells dried completely, the paper was rinsed with PBS-T buffer followed by 30 s vibration to wash the μPAD and remove non-specific binding, followed by drying at 40 °C for 3 min. Finally, 3 μL
IP T
of TMB solution was added to the hydrophilic wells of the μPAD at room temperature in order to induce the color change. After 3 min of TMB reaction, a portable optical
SC R
imaging device (microscope or smartphone) recorded the results, which we also uploaded to a cloud computing system. The images were analyzed in terms of the
U
red/blue (R/B) ratios of the testing spots based on the images’ red-green-blue (RGB)
A
N
values as determined by ImageJ software and a developed Android smartphone app.
M
The error bars correspond to the standard deviation of the sample mean.
ED
2.3.1 Determining the optimized concentrations of K-Ab, ketamine standard, and K-HRP
PT
The cP-ELISA procedure was optimized using 3 μL of different K-Ab
CC E
concentrations (44.5 μg/mL, 88.9 μg/mL, and 177.8 μg/mL), 3 μL of 1 wt% BSA to avoid non-specific binding, 3 μL of different concentrations of ketamine standard (10 -102 μg/ mL), and 3 μL of different K-HRP concentrations (diluted 0.1x (10 times
A
6
diluted), 0.033x (30 times diluted), and 0.017x (60 times diluted)) to compete with KAb in the reaction. Between each step, all of the reagents were added in succession to the hydrophilic wells of the μPADs and incubated for 3 min at 40 °C on a heater. After 10
the hydrophilic wells dried completely, the paper was rinsed with the PBS-T buffer followed by 30 s vibration to wash the μPAD platforms, followed by drying at 40 °C for 3 min. Finally, 3 μL of TMB solution was added to the μPADs at room temperature
IP T
in order to observe the color change. 2.3.2 Determining the working temperature, vibration washing, and TMB reaction
SC R
times
The cP-ELISA procedure was accompanied by 3 μL of 88.9 μg/mL K-Ab, 3 μL of
U
1 wt% BSA, 3 μL of different concentrations of ketamine standard (10-6-102 μg/ mL),
A
N
and 3 μL of 0.033x diluted K-HRP to compete with K-Ab in the reaction. Between each
M
successive addition of these components to the hydrophilic wells of the μPAD, the
ED
incubation time was set at 3 min and the platform was heated on a heater at different temperatures (30, 40, 50, and 60 °C). After the hydrophilic wells dried completely, the
PT
paper was rinsed and vibration washed in PBS-T buffer for different times (20, 25, 30,
CC E
and 40 s), and then dried on a heater for 3 min at the respective heating temperature. Finally, 3 μL of TMB solution was added to each testing well of the μPAD at room
A
temperature to observe the color change and recorded at different reaction times (1, 3, 5, and 7 min). 2.4 Real oral fluid samples analyzed for ketamine using the cP-ELISA sensing system 2.4.1 Institutional Review Board Approval 11
Human oral fluid samples used in this study were approved and issued by the institutional review board of Kaohsiung Medical University Hospital, Kaohsiung, Taiwan.
IP T
2.4.2 Clinical oral fluid samples for ketamine analysis by the cP-ELISA sensing system and GC-MS
SC R
For the μPAD sensing system, we utilized the optimized parameters determined
previously for this experiment. The clinical oral fluid samples were provided by
U
Kaohsiung Medical University Hospital and stored at -20 °C, and were defrosted before
A
N
sample pretreatment. To run the test, 3 μL of 88.9 μg/mL K-Ab, 3 μL of 1 wt% BSA, 3
M
μL of the unknown, unfiltered oral fluid sample, and 3 μL of the 0.033x diluted K-HRP
ED
were added in sequence to the hydrophilic wells of the μPAD, followed by incubation for 3 min on a 40 °C heater. After the hydrophilic wells dried completely, the paper was
PT
rinsed and vibration washed in PBS-T buffer for 30 s. The μPAD was then dried at 40
CC E
°C for 3 min. Finally, 3 μL TMB solution was added to the dried μPAD to observe the color change at room temperature, and then the images were analyzed. To compare the
A
results from the cP-ELISA sensing system, we also used a GC-MS to measure the concentration of ketamine in the oral fluid samples. 3. Results and discussion Here we present the results of a 3-year special program of drug tests in Taiwan. 12
Most commercially available ketamine tests that can be collected or performed in the field are based on lateral flow immunoassays (LFIAs) or ELISA tests, which are highly sensitive and specific for the detection of a wide variety of target analytes. However,
IP T
ELISA is time consuming, usually taking 2–4 hours to complete the assay, and while LFIA can rapidly obtain detection results, it can only provide a binary yes/no response
SC R
with limited sensitivity [47, 48]. In this study, our paper-based colorimetric sensing platform with modified detection processes provides a rapid, sensitive, specific, and
U
cost-effective approach for ketamine detection that we are proposing to overcome these
A
N
issues.
M
We have chosen to develop a cP-ELISA test system that can be analyzed using a
ED
smartphone and an app in order to obtain quasi-quantitative results in resource-limited settings, such as on the road [49, 50]. The app mainly serves two purposes: (1) taking
PT
image results of the μPADs; and (2) performing image processing for rapid testing.
CC E
Adopting smartphones as the optical imaging device for μPADs enables high portability and detection flexibility capabilities, but comes with some color display issues caused
A
by varying ambient light on the captured images. To address these color display issues, we conduct a white balance correction process, which is an essential and commonly used algorithm in image processing to eliminate color casts due to illumination [51]. The goal of this process is to adjust the intensities of the colors such that they are 13
rendered to reflect the colors viewed by the naked-eye [52]. As a result, the proposed cP-ELISA sensing system not only provides a qualitative result for the presence of ketamine, but also indicates different concentrations of the analyte via the relative color
IP T
change of the hydrophilic sample wells printed on the paper surface based on the amount of analyte present in the testing solution.
SC R
A schematic illustration of the ketamine detection system using the paper-based analytical device is shown in Scheme 1. Oral fluid samples from a suspect can be
U
collected in one of two ways. The first involves the established method of sending the
A
N
sample to a central lab for GC-MS confirmation, a timely and expensive process. The
M
alternative is our proposed ketamine detection system using the µPAD and recording
ED
the results via a smartphone or portable microscope system. The advantages of immediate quantification of the suspect’s ketamine levels are obvious.
PT
The µPAD platform utilizes a chemical test that is based on the cP-ELISA assay,
CC E
which uses an enzyme reaction to display a color change that is inversely proportional to the concentration of the target analyte. The amount of K-Ab, K-HRP, and sample
A
solution were all limited to a small volume of just 3 µL each. This makes the sensing platform more cost-effective, safe, and rapid compared to traditional ELISA and UVvisible absorption analysis, which typically requires a sample volume of 200 μL. In addition, the small sample volume requirement also results in a shorter diffusion length 14
during the color development, so the total assay time can be reduced from hours to minutes. The color of the cellulose paper, the hydrophobic barriers, and printed reference color bars are utilized to adjust lighting conditions.
IP T
The detection results are then recorded using an app for immediate and automatic analysis. The Android smartphone app was implemented based on the Linux Kernel
SC R
mobile operating system and Android Marshmallow version 6.0. Upon binding to increasing concentration of ketamine, the color of the sample wells changes from deep
U
blue to light blue. We obtained the RGB values of these images and calculated the ratios
A
N
of R/G, R/B, and G/B in order to determine which value could be used to most
M
effectively monitor the color change of the platform. Our results showed that the R/B
ED
ratio was both concentration-dependent and produced the most significant color change in our µPAD cP-ELISA sensing system with normalized deviation (Fig. 1). Therefore,
PT
we selected the R/B value for the detection of ketamine in all subsequent experiments.
CC E
3.1 The effects of K-Ab and K-HRP concentration in the cP-ELISA sensing system For c-ELISA, the fixed antibody and competitive antibody concentration are the
A
two most important factors in the colorimetric assay. The purpose of optimizing the antibody concentration was to obtain low background interference and high sensitivity for the immunoassay. When the antibody concentration is too high, the low-affinity non-specific antibody cannot be removed completely by the washing step if its 15
specificity to the target analyte is limited. This causes a higher background noise. If the antibody concentration is too low, the sensitivity of the ELISA test will not be sufficient based on the limited epitopes and signal molecules [47, 53]. In order to optimize these
IP T
conditions, we tested different K-Ab concentrations (44.5 μg/mL, 88.9 μg/mL, and 177.8 μg/mL) on the paper-based colorimetric assay (Fig. 2A-B). Our results showed
SC R
that the lowest signal variability was observed using a concentration of 88.9 μg/mL KAb. Lower or higher antibody concentration produced greater signal variation without
U
offering better binding.
A
N
We also tested the K-HRP concentration, diluted by 0.1x, 0.033x, and 0.017x (Fig.
M
2C-D). Our results showed that the 0.033x diluted K-HRP sample presented the most
ED
stable and minimal variability in the data. For the 0.1x diluted K-HRP sample, the cELISA reaction saturated rapidly, resulting in a color change that was not significant.
PT
Alternatively, the 0.017x diluted K-HRP sample demonstrated a color change in the
CC E
paper-based colorimetric assay. However, there was large variation for the sample wells that featured a low ketamine concentration (10-6 to 10-4 μg/mL) and insignificant color
A
change for the medium ketamine concentration (10-3 to 10-1 μg/mL). Overall, this cPELISA sensing system could be quantified using the R/B ratio of the acquired images to generate linear results as a function of ketamine concentration. Based on these findings, for all subsequent experiments we used the 88.9 μg/mL K-Ab concentration 16
and the 0.033x diluted K-HRP solution. 3.2 The effects of working temperature and washing time on the cP-ELISA sensing system
IP T
To increase the rate and efficiency of the sensing system, we heated the paper platform and washed it by vibration washed in PBS-T buffer to reduce the effects of
SC R
non-specific binding. Traditional ELISA tests utilize elevated incubation temperatures and time to improve the efficiency of antibody binding, however, excessive heating is
U
also known to increase the background signal [54]. The effective diffusivity is an
A
N
important factor in the immunoassay, which is also related to the incubation time used
M
to achieve the saturated limited of the test [55-57]. Based on these reasons, heating
ED
should shorten the time to process the cP-ELISA test with the potential for rapid screening of ketamine. In addition, due to the low reagent volume required for the cP-
PT
ELISA test, heating could enhance the diffusion of the antigen-antibody reaction and
CC E
enhance reagent evaporation. To find the optimum condition, we tested different working temperatures,
A
including 30, 40, 50, and 60 °C (Fig. 3A-B). Our results showed that the most stable data with the lowest standard deviation occurred at a temperature of 40 °C. At 30 °C, the test results were not linear with ketamine concentration, which we surmised was due to the antigen-antibody interaction not being stable at this lower temperature. And 17
at higher temperatures, the color change was not significant and the variability between sample replicates increased, possibly due to denaturing of the antibody. Using 40 °C for the working temperature is also consistent with the baseline temperature of the
IP T
human immune system. Additionally, according to the Stokes-Einstein equation, incubating at the appropriate temperature increases the diffusion of intermolecular
SC R
collision in the ELISA reaction, thus providing a rapid immunoassay [55].
U
where D is the diffusion constant, KB is the Boltzmann's constant, T is the absolute
A
N
temperature, η is the dynamic viscosity and r is the radius of the spherical particle.
M
Washing is another important procedure in the ELISA test because it helps to
ED
decrease signal interference and raises the sensitivity of the assay by reducing nonspecific binding. Therefore, we tested different washing times by vibration washed the
PT
paper platform for 20, 25, 30, and 40 s in PBS-T buffer (Fig. 3C-D). Our results show
CC E
that an optimum washing time of 30 s achieves the most stable and minimal standard deviation in the colorimetric data. The results were similar for the 20 and 25 s washed
A
samples, but there was less significant color change in the cP-ELISA test. In addition, the results of washing for 40 s resulted in low color change and did not demonstrate good linearity for ketamine detection. We hypothesize that too long of a washing time could result in removal of the antigen-antibody complex, resulting in poor signal 18
detection. Based on these results, for all subsequent experiments we used a 40 °C working temperature and 30 s washing time for the cP-ELISA test. 3.3 Effect of TMB reaction time in the cP-ELISA sensing system
IP T
The enzyme-substrate reaction is the final step in the ELISA test, which is used to produce a color change for measurement. We studied various TMB reaction times,
SC R
including 1, 3, 5, and 7 min, in order to determine which produced the largest and most
linear color development as a function of ketamine concentration (Fig. 3E-F). Our
U
results showed the enzyme-substrate reaction at 3 min resulted in a significant color
A
N
change. The R/B ratio did not change significantly at 1 min and 7 min, suggesting the
M
reaction was just beginning at 1 min and became saturated at 7 min. When the reaction
ED
time reaches 3 min, a highest color variance can be observed for a lower detection limit. Based on these results, for all subsequent experiments we selected 3 min for the TMB
PT
reaction time.
CC E
3.4 Standard curve for ketamine detection in the cP-ELISA sensing system Based on the prior experiments, we determined the optimum parameters for the
A
cP-ELISA sensing system consisted of 88.9 μg/mL K-Ab, K-HRP 0.033x diluted, a 40 °C working temperature, 30 s washing time, and 3 min TMB reaction. Using these conditions, we established a standard curve of ketamine (blank, 10-6-102 μg/mL) in an oral fluid sample that had been pretreated by filtration through a 0.22 μm membrane, 19
as well as a corresponding dataset in oral fluid without pretreatment, in order to more closely model clinical trials with limited processes (Fig. 4). The linear range of the oral fluid sample without filtering pretreatment was from 10-3–100 μg/mL (R/B ratio from
IP T
0.62 to 0.94), which corresponded to a limit of detection (LOD) of 0.3 ng/mL and an R2 value of 0.98. On the other hand, using the 0.22 μm membrane to filter the oral fluid
SC R
sample resulted in a linear range of 10-4–10-1 μg/mL (R/B ratio from 0.59 to 0.84), and
LOD of 0.03 ng/mL with an R2 value of 0.98 (based on the concentration of the test that
U
produced a signal that was three times greater than the baseline noise signal of the
A
N
control; LOD = yblank + 3SDblank). We surmised that the unfiltered oral fluid sample had
M
some interference or non-specific binding to cause the R/B ratio to increase. However,
ED
in order to achieve the convenience of fast screening, using an unfiltered oral fluid sample would help accelerate sample analysis and save time.
PT
3.5 Clinical application for cP-ELISA sensing system
CC E
In order to demonstrate the cP-ELISA sensing system has the potential for analysis of ketamine, we analyzed 90 clinical samples, 40 of which were positive for ketamine
A
and 50 that were negative, though we were unaware of these values at the time of analysis. In each test, the clinical samples were compared to a blank oral fluid sample given by a volunteer. The cut-off point and receiver operating characteristic (ROC) curve of our cP-ELISA sensing system found 36 samples that tested positive for 20
ketamine and 46 that tested negative, which is 90% sensitivity (Confidence Interval (CI): 76.34-97.21%) and 92% specificity (CI: 80.77-97.78%) (Fig. 5). GC-MS testing of the samples also supported our cP-ELISA sensing results.
IP T
The traditional ELISA procedure requires waiting for antigen-antibody binding, incubation, and washing time, which takes at least 2–4 hours in total. Herein, our µPAD
SC R
c-ELISA sensing system, which can measure multiple samples in parallel, requires just
minutes to complete the test. This cP-ELISA sensing system also conforms to the point-
U
of-care testing standards, and provides a rapid, cheap, and simple method of ketamine
A
N
analysis [58].
M
4. Conclusions
ED
In this study, we used a microfluidic paper-based analytical device with the modified detection processes to test saliva for ketamine in an ultrafast manner. One of
PT
the advantages of using oral fluid is that the process of sampling saliva is non-invasive,
CC E
which makes testing in the field easier and faster. A key potential advantage of the proposed technology is the ability to simultaneously and effectively measure low
A
concentrations of multiple drug compounds in oral fluid without the need for a large amount of samples, complex equipment, or long processing times. This research provides a new drug testing platform for high-throughput drug abuse monitoring. Preliminary results show that the LOD of this cP-ELISA sensing system 21
for ketamine is 0.03 ng/mL, which is lower than the 100 ng/mL legal limit in Taiwan [59]. This cP-ELISA sensing system lowers the uncertainty of testing samples by heating and washing the platform. Comparing to other P-ELISA tests, the detection
IP T
time can be effectively shortened from ~ 30 min to 6 min [34, 37]. It was also only requires a few μL of oral fluid samples. Compared with the traditional ELISA test, this
SC R
sensing system improves the sensitivity and specificity for ketamine analysis. In
addition, the analysis can be performed using a smartphone app under resource-
U
restricted conditions and does not require expensive instrumentation. The cP-ELISA
A
N
sensing system offers a high quality and fast testing platform that could be expanded to
M
an array of drugs beyond ketamine, including amphetamines, cocaine, opioids, etc. We
ED
expect the findings from this work can be used to build cost-effective and portable approaches for highly sensitive and specific drug analysis and other biomedical ELISA
CC E
PT
tests in resource limited environments.
Acknowledgment
A
This research was supported by the Ministry of Science and Technology, Taiwan (1062221-E-002-139-MY2 & 104-2218-E-002-039), Kaohsiung Medical University Chung-Ho Memorial Hospital (KMUH105-5T58), and the Higher Education Sprout Program at National Taiwan University. 22
References
IP T
[1] L. Li, P.E. Vlisides, Ketamine: 50 Years of Modulating the Mind, Frontiers in Human Neuroscience, 10(2016) 612. [2] E.F. Göktaş, F. Arıöz, A review of chromatographic methods for ketamine and its metabolites norketamine and dehydronorketamine, Biomedical chromatography : BMC, (2017) e4014. [3] K.W. Leung, Z.C.F. Wong, J.Y.M. Ho, A.W.S. Yip, J.S.C. Ng, S.P.H. Ip, et al., Determination of hair ketamine cut-off value from Hong Kong ketamine users by LC–
SC R
MS/MS analysis, Forensic Science International, 259(2016) 53-8. [4] G.L. Yang, A.D. Zhang, Y. Yu, H. Liu, F.Y. Long, J. Yan, Drug use and its associated factors among money boys in Hunan Province, China, Public Health, 140(2016) 213-
M
A
N
U
20. [5] C.-A. Yang, H.-C. Liu, D.-L. Lin, R.H. Liu, Y.-Z. Hsieh, S.-P. Wu, Simultaneous Quantitation of Methamphetamine, Ketamine, Opiates and their Metabolites in Urine by SPE and LC–MS-MS, Journal of Analytical Toxicology, 41(2017) 679-87. [6] D. Favretto, S. Vogliardi, M. Tucci, I. Simoncello, R. El Mazloum, R. Snenghi, Occupational exposure to ketamine detected by hair analysis: a retrospective and prospective toxicological study, Forensic Science International, 265(2016) 193-9.
CC E
PT
ED
[7] M. Sergi, D. Compagnone, R. Curini, G. D’Ascenzo, M. Del Carlo, S. Napoletano, et al., Micro-solid phase extraction coupled with high-performance liquid chromatography–tandem mass spectrometry for the determination of stimulants, hallucinogens, ketamine and phencyclidine in oral fluids, Analytica Chimica Acta, 675(2010) 132-7. [8] A.S. Barreto, V.F. Brant, E. Spinelli, S.V. Rodrigues, Validation of a SPE-LC–MS/MS method for the determination of ketamine and norketamine in micropulverized hair after a single IV dose, Journal of Chromatography B, 1033(2016) 200-9. [9] F. Bianchi, F. Bisceglie, S. Dugheri, G. Arcangeli, V. Cupelli, E. del Borrello, et al.,
A
Ionic liquid-based solid phase microextraction necklaces for the environmental monitoring of ketamine, Journal of Chromatography A, 1331(2014) 1-9. [10] S. Gentili, C. Mortali, L. Mastrobattista, P. Berretta, S. Zaami, Determination of different recreational drugs in sweat by headspace solid-phase microextraction gas chromatography mass spectrometry (HS-SPME GC/MS): Application to drugged drivers, Journal of Pharmaceutical and Biomedical Analysis, 129(2016) 282-7. [11] F.A. Sandbaumhüter, R. Theurillat, R.N. Bektas, A.P.N. Kutter, R. BettschartWolfensberger, W. Thormann, Pharmacokinetics of ketamine and three metabolites 23
in Beagle dogs under sevoflurane vs. medetomidine comedication assessed by
IP T
enantioselective capillary electrophoresis, Journal of Chromatography A, 1467(2016) 436-44. [12] E. Bonfanti, F. Cosnier, L. Wathier, P. Campo, Measurement of ketamine and xylazine in rat brain by liquid–liquid extraction and gas chromatography–mass spectrometry, Journal of Pharmacological and Toxicological Methods, 77(2016) 6-9. [13] Y. Yamini, S. Seidi, R. Feizbakhsh, T. Baheri, M. Rezazadeh, Liquid-phase microextraction based on two immiscible organic solvents followed by gas chromatography with mass spectrometry as an efficient method for the preconcentration and determination of cocaine, ketamine, and lidocaine in human urine samples, Journal of Separation Science, 37(2014) 2364-71.
SC R
[14] L. Meng, B. Wang, F. Luo, G. Shen, Z. Wang, M. Guo, Application of dispersive liquid–liquid microextraction and CE with UV detection for the chiral separation and determination of the multiple illicit drugs on forensic samples, Forensic Science
M
A
N
U
International, 209(2011) 42-7. [15] Z. Lin, J. Li, X. Zhang, M. Qiu, Z. Huang, Y. Rao, Ultrasound-assisted dispersive liquid-liquid microextraction for the determination of seven recreational drugs in human whole blood using gas chromatography–mass spectrometry, Journal of Chromatography B, 1046(2017) 177-84. [16] P.-S. Cheng, C.-Y. Fu, C.-H. Lee, C. Liu, C.-S. Chien, GC–MS quantification of ketamine, norketamine, and dehydronorketamine in urine specimens and
CC E
PT
ED
comparative study using ELISA as the preliminary test methodology, Journal of Chromatography B, 852(2007) 443-9. [17] I. Moreno, M. Barroso, A. Martinho, A. Cruz, E. Gallardo, Determination of ketamine and its major metabolite, norketamine, in urine and plasma samples using microextraction by packed sorbent and gas chromatography-tandem mass spectrometry, Journal of Chromatography B, 1004(2015) 67-78. [18] F. Vaiano, F.P. Busardò, D. Palumbo, C. Kyriakou, A. Fioravanti, V. Catalani, et al., A novel screening method for 64 new psychoactive substances and 5 amphetamines in blood by LC–MS/MS and application to real cases, Journal of Pharmaceutical and
A
Biomedical Analysis, 129(2016) 441-9. [19] M. Hasan, R. Hofstetter, G.M. Fassauer, A. Link, W. Siegmund, S. Oswald, Quantitative chiral and achiral determination of ketamine and its metabolites by LC– MS/MS in human serum, urine and fecal samples, Journal of Pharmaceutical and Biomedical Analysis, 139(2017) 87-97. [20] Y.-H. Ho, C.-C. Wang, Y.-T. Hsiao, W.-K. Ko, S.-M. Wu, Analysis of ten abused drugs in urine by large volume sample stacking–sweeping capillary electrophoresis with an experimental design strategy, Journal of Chromatography A, 1295(2013) 136-41. 24
[21] N. Porpiglia, G. Musile, F. Bortolotti, E.F. De Palo, F. Tagliaro, Chiral separation
IP T
and determination of ketamine and norketamine in hair by capillary electrophoresis, Forensic Science International, 266(2016) 304-10. [22] A. Argente-García, N. Jornet-Martínez, R. Herráez-Hernández, P. Campíns-Falcó, A passive solid sensor for in-situ colorimetric estimation of the presence of ketamine in illicit drug samples, Sensors and Actuators B: Chemical, 253(2017) 1137-44. [23] B. Deiminiat, G.H. Rounaghi, Fabrication of a new electrochemical imprinted sensor for determination of ketamine based on modified polytyramine/sol-gel/fMWCNTs@AuNPs nanocomposite/pencil graphite electrode, Sensors and Actuators B: Chemical, 259(2018) 133-41. [24] G.A. Posthuma-Trumpie, J. Korf, A. van Amerongen, Lateral flow (immuno)assay:
SC R
its strengths, weaknesses, opportunities and threats. A literature survey, Analytical and Bioanalytical Chemistry, 393(2009) 569-82. [25] R. Dams, R.E. Choo, W.E. Lambert, H. Jones, M.A. Huestis, Oral fluid as an
M
A
N
U
alternative matrix to monitor opiate and cocaine use in substance-abuse treatment patients, Drug & Alcohol Dependence, 87(2007) 258-67. [26] O.R. Idowu, B. Caddy, A Review of the Use of Saliva in the Forensic Detection of Drugs and Other Chemicals, Journal of the Forensic Science Society, 22(1982) 123-35. [27] A.G. Verstraete, Oral fluid testing for driving under the influence of drugs: history, recent progress and remaining challenges, Forensic Science International, 150(2005) 143-50.
CC E
PT
ED
[28] X. Pei, B. Zhang, J. Tang, B. Liu, W. Lai, D. Tang, Sandwich-type immunosensors and immunoassays exploiting nanostructure labels: A review, Analytica Chimica Acta, 758(2013) 1-18. [29] A. Makaraviciute, A. Ramanaviciene, Site-directed antibody immobilization techniques for immunosensors, Biosensors and Bioelectronics, 50(2013) 460-71. [30] Z. Wang, S. Zong, L. Wu, D. Zhu, Y. Cui, SERS-Activated Platforms for Immunoassay: Probes, Encoding Methods, and Applications, Chemical Reviews, 117(2017) 7910-63. [31] R. Hu, T. Liu, X.-B. Zhang, Y. Yang, T. Chen, C. Wu, et al., DLISA: A DNAzyme-Based
A
ELISA for Protein Enzyme-Free Immunoassay of Multiple Analytes, Analytical Chemistry, 87(2015) 7746-53. [32] Y. Li, X. Ma, Z. Xu, M. Liu, Z. Lin, B. Qiu, et al., Multicolor ELISA based on alkaline phosphatase-triggered growth of Au nanorods, Analyst, 141(2016) 2970-6. [33] B.M. Murphy, X. He, D. Dandy, C.S. Henry, Competitive Immunoassays for Simultaneous Detection of Metabolites and Proteins Using Micromosaic Patterning, Analytical Chemistry, 80(2008) 444-50. [34] C.-M. Cheng, A.W. Martinez, J. Gong, C.R. Mace, S.T. Phillips, E. Carrilho, et al., 25
Paper-Based ELISA, Angewandte Chemie International Edition, 49(2010) 4771-4.
IP T
[35] S. Wang, L. Ge, X. Song, J. Yu, S. Ge, J. Huang, et al., Paper-based chemiluminescence ELISA: Lab-on-paper based on chitosan modified paper device and wax-screen-printing, Biosensors and Bioelectronics, 31(2012) 212-8. [36] M.-Y. Hsu, Y.-C. Hung, D.-K. Hwang, S.-C. Lin, K.-H. Lin, C.-Y. Wang, et al., Detection of aqueous VEGF concentrations before and after intravitreal injection of anti-VEGF antibody using low-volume sampling paper-based ELISA, Scientific Reports, 6(2016) 34631. [37] C.-M. Shih, C.-L. Chang, M.-Y. Hsu, J.-Y. Lin, C.-M. Kuan, H.-K. Wang, et al., Paperbased ELISA to rapidly detect Escherichia coli, Talanta, 145(2015) 2-5. [38] H.-K. Wang, C.-H. Tsai, K.-H. Chen, C.-T. Tang, J.-S. Leou, P.-C. Li, et al., Cellulose-
SC R
Based Diagnostic Devices for Diagnosing Serotype-2 Dengue Fever in Human Serum, Advanced Healthcare Materials, 3(2014) 187-96. [39] A.W. Martinez, S.T. Phillips, M.J. Butte, G.M. Whitesides, Patterned paper as a
M
A
N
U
platform for inexpensive, low-volume, portable bioassays, Angewandte ChemieInternational Edition, 46(2007) 1318-20. [40] A.W. Martinez, S.T. Phillips, G.M. Whitesides, E. Carrilho, Diagnostics for the Developing World: Microfluidic Paper-Based Analytical Devices, Analytical Chemistry, 82(2010) 3-10. [41] G.H. Chen, W.Y. Chen, Y.C. Yen, C.W. Wang, H.T. Chang, C.F. Chen, Detection of Mercury(II) Ions Using Colorimetric Gold Nanoparticles on Paper-Based Analytical
CC E
PT
ED
Devices, Analytical Chemistry, 86(2014) 6843-9. [42] E.M. Spyrou, D.P. Kalogianni, S.S. Tragoulias, P.C. Ioannou, T.K. Christopoulos, Digital camera and smartphone as detectors in paper-based chemiluminometric genotyping of single nucleotide polymorphisms, Analytical and Bioanalytical Chemistry, 408(2016) 7393-402. [43] S.M. Russell, A. Domenech-Sanchez, R. de la Rica, Augmented Reality for RealTime Detection and Interpretation of Colorimetric Signals Generated by Paper-Based Biosensors, ACS sensors, 2(2017) 848-53. [44] J. Narang, N. Malhotra, C. Singhal, A. Mathur, D. Chakraborty, A. Anil, et al., Point
A
of care with micro fluidic paper based device integrated with nano zeolite–graphene oxide nanoflakes for electrochemical sensing of ketamine, Biosensors and Bioelectronics, 88(2017) 249-57. [45] Y.Y. Xia, J. Si, Z.Y. Li, Fabrication techniques for microfluidic paper-based analytical devices and their applications for biological testing: A review, Biosensors & Bioelectronics, 77(2016) 774-89. [46] Y. Yang, E. Noviana, M.P. Nguyen, B.J. Geiss, D.S. Dandy, C.S. Henry, Paper-Based Microfluidic Devices: Emerging Themes and Applications, Analytical Chemistry, 26
89(2017) 71-91.
IP T
[47] Y. Hayashi, R. Matsuda, T. Maitani, K. Imai, W. Nishimura, K. Ito, et al., Precision, Limit of Detection and Range of Quantitation in Competitive ELISA, Analytical Chemistry, 76(2004) 1295-301. [48] M.-H. Huang, M.-Y. Wu, C.-H. Wu, J.-L. Tsai, H.-H. Lee, R.H. Liu, Performance characteristics of ELISAs for monitoring ketamine exposure, Clinica Chimica Acta, 379(2007) 59-65. [49] Q. Mei, H. Jing, Y. Li, W. Yisibashaer, J. Chen, B. Nan Li, et al., Smartphone based visual and quantitative assays on upconversional paper sensor, Biosensors and Bioelectronics, 75(2016) 427-32. [50] M.M. Gong, D. Sinton, Turning the Page: Advancing Paper-Based Microfluidics
SC R
for Broad Diagnostic Application, Chemical Reviews, 117(2017) 8447-80. [51] G.D. Finlayson, S.D. Hordley, P.M. Hubel, Color by correlation: A simple, unifying framework for color constancy, IEEE Transactions on Pattern Analysis and Machine
M
A
N
U
Intelligence, 23(2001) 1209-21. [52] J.Y. Huo, Y.L. Chang, J. Wang, X.X. Wei, Robust automatic white balance algorithm using gray color points in images, Ieee Transactions on Consumer Electronics, 52(2006) 541-6. [53] M. Palmieri, M. Vagnini, L. Pitzurra, P. Rocchi, B.G. Brunetti, A. Sgamellotti, et al., Development of an analytical protocol for a fast, sensitive and specific protein recognition in paintings by enzyme-linked immunosorbent assay (ELISA), Analytical
CC E
PT
ED
and Bioanalytical Chemistry, 399(2011) 3011-23. [54] E.T. Thean, B.H. Toh, Western immunoblotting: Temperature-dependent reduction in background staining, Analytical Biochemistry, 177(1989) 256-8. [55] M. Stenberg, H. Nygren, A diffusion limited reaction theory for a solid-phase immunoassay, Journal of Theoretical Biology, 113(1985) 589-97. [56] M. Stenberg, M. Werthén, S. Theander, H. Nygren, A diffusion limited reaction theory for a microtiter plate assay, Journal of Immunological Methods, 112(1988) 239. [57] D. Rath, S. Panda, Contribution of rotational diffusivity towards the transport of
A
antigens in heterogeneous immunosensors, Analyst, 140(2015) 6579-87. [58] C.P. Price, Point of care testing, BMJ : British Medical Journal, 322(2001) 1285-8. [59] H.-Y. Liao, J.-H. Chen, J.-J. Shyue, C.-T. Shun, H.-W. Chen, S.-W. Liao, et al., Rapid label-free determination of ketamine in whole blood using secondary ion mass spectrometry, Talanta, 143(2015) 50-5.
27
Figure captions: Scheme 1. A schematic illustration of the proposed cP-ELISA sensing system for drug abuse tests. The paper-based platform should enhance the speed and lower the cost of
IP T
ketamine analysis.
SC R
Fig. 1. Image analysis of the cP-ELISA sensing system using the red (R), green (G), and blue (B) color values of images taken of the paper platform (N = 5). Colorimetric
U
results of ketamine analysis on (A) the paper platform, (B) the corresponding R, G, B,
A
N
and gray values, and (C) the R/G, R/B, and G/B ratios of ketamine analysis were
M
determined by ImageJ software. The concentration of K-Ab and K-HRP were 88.9
ED
μg/mL and 0.033x diluted, respectively. The heating temperature was 40 °C and the
PT
washing time was 30 s. TMB reacted for 3 min to induce the color change.
CC E
Fig. 2. Effect of the concentration of K-Ab and K-HRP in the cP-ELISA sensing platform (N = 5). The colorimetric results of ketamine analysis based on (A) the paper
A
platform and (B) the corresponding R/B ratios for samples using 44.5 μg/mL, 88.9 μg/mL, and 177.8 μg/mL of K-Ab and 0.033x diluted K-HRP. Different K-HRP concentrations were also tested, ranging from 0.1x, 0.033x, and 0.017x diluted, on (C) paper and (D) the µPAD cP-ELISA sensing system. The heating temperature was set at 28
40 °C. The paper platforms were also washed by vibration washed for 30 s, and TMB was allowed to react for 3 min to induce the color change.
IP T
Fig. 3. Effect of the heating temperature, washing time and TMB reaction time of the cP-ELISA sensing system (N = 5). The colorimetric results of ketamine analysis based
SC R
on (A) the paper platform and (B) the corresponding R/B ratios for 30, 40, 50, and 60 °C antibody incubation temperatures. Different washing times ranging from 20, 25, 30,
U
and 40 s were tested on (C) the paper platform, with (D) the corresponding R/B ratios
A
N
of the µPAD cP-ELISA sensing system also shown. The concentration of K-Ab and K-
M
HRP were set at 88.9 μg/mL and 0.033x diluted, respectively. TMB was reacted for 3
ED
min to induce the color change. (E)(F) The colorimetric results of ketamine analysis of the corresponding R/B ratios and the paper platform for reaction times of 1, 3, 5, and 7
PT
min. The concentration of K-Ab and K-HRP were set at 88.9 μg/mL and 0.033x diluted,
CC E
respectively. The heating temperature was 40 °C and the washing time was 30 s.
A
Fig. 4. The standard ketamine curve in oral fluid with and without pre-treatment by a 0.22 μm membrane using the cP-ELISA sensing system (N = 5). The linear range of the oral fluid sample without filtering pretreatment was from 10-3–100 mg/mL, which corresponded to a LOD of 0.3 ng/mL. On the other hand, using the 0.22 μm membrane 29
to filter the oral fluid sample resulted in a linear range of 10-4–10-1 mg/mL and an LOD of 0.03 ng/mL.
IP T
Fig. 5. Statistical results of (A) the cut-off point and (B) ROC curve from the cP-ELISA sensing system for clinical ketamine analysis. The cut-off value was 0.04, which
SC R
resulted in a sensitivity and specificity for the platform of 90% (CI: 76.34-97.21%) and
A
CC E
PT
ED
M
A
N
U
92% (CI: 80.77-97.78%), respectively.
30
A ED
PT
CC E
IP T
SC R
U
N
A
M
Scheme 1.
31
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 1.
32
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 2.
33
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 3.
34
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 4.
35
A ED
PT
CC E
IP T
SC R
U
N
A
M
Fig. 5.
36