1. Introduction
Smoking is a widespread social problem, and there are 1.1 billion smokers worldwide. According to a recent report by the World Health Organization, over 8 million people died from smoking in 2019 [
1]. It is associated with various diseases such as cardiovascular diseases, cancer, diabetes, and lung diseases [
2,
3]. In the form of passive smoking, smoking also affects non-smokers, especially children and pregnant women [
4]. Due to its negative effects, smoking in public places has been prohibited by law in many countries. Moreover, social awareness to encourage smoking cessation is emerging. Therefore, immediate monitoring and testing of smoking is essential not only to maintain public health but also to help smokers who already know the harmful effects and want to quit.
Cotinine, a biomarker to detect smoking, is the main metabolite of nicotine, a highly addictive substance in cigarettes. It is a good biomarker with a long half-life of about 18 h after smoking; it is produced by the liver, kidneys, and lungs within 1 to 2 h after smoking and can be detected in blood, saliva, and urine samples. Traditional cotinine detection methods typically include high-performance chromatography (HPLC) and gas chromatography-mass spectroscopy (GC-MS). Although these chromatographic methods have relatively high sensitivity and accuracy, they are time-consuming and require expensive analytic instruments and personnel training. Like these methods, enzyme-linked immunosorbent assay (ELISA), a commonly used method, is also used at laboratory scale and is not easily accessible to general users. (
Table 1) [
5,
6,
7,
8,
9].
For immediately on-site and portable cotinine detection, colorimetric lateral flow immunoassay (LFIA) is used with gold conjugates. Because the cotinine has an insufficient size for a common sandwich immunoassay, competitive substance and antibodies were immobilized on paper strips to detect small molecules by competitive immunoassay [
10]. When cotinine is present in the sample, fewer antibodies bind to the immobilized competitive substance in the test line. As a sample for detecting cotinine, most studies use urine or saliva, which are non-invasive and more accessible than other biological liquids. Although saliva is more advantageous for testing than urine, which is limited in time and space for field diagnosis, most commercially available cotinine kits are urine-based, not saliva-based, because of the low sensitivity of kits using saliva with high viscosity and impurities [
11,
12,
13]. The composition of saliva differs from person to person, and it depends on food intake, hydration status, smoking, and saliva collection method. Moreover, viscosity of mucin from saliva could affect downstream analysis [
14]. Therefore, highly sensitive studies require oral hygiene procedures to avoid the contamination of samples by food. The pretreatments make it difficult to set up proper rapid saliva testing conditions applicable in on-site detection. Therefore, in the cases of commercial on-site saliva-based detection kits, they require filtering or dilution after collection of saliva by spitting or chewing sponge for using whole saliva, in order to overcome the background effect (
Table 2). Moreover, they require more volume of saliva and time to collect a sufficient amount of sample for testing. In this study, a simpler and more accurate saliva-based detection method for cotinine that can harness the convenience of saliva is developed, using a cotton swab. Moreover, since a relatively identical amount of sample is applied using the swab, the individual differences between users can be reduced compared to conventional cotinine kits.
A cotton swab is a useful tool to collect a saliva sample, thus making it non-invasive, easy-to-handle, and better accessible for users. Further, it can acquire similar sample volumes without using quantitative instruments. Thus, many on-site detection kits include swabs to collect samples such as saliva, nasal fluid, or environmental samples. In many studies, cotton swabs have been used for sample collection, similar to nasal fluids for respiratory virus tests, but the samples collected by swabs require a resuspension step in another buffer to apply the detection process, without it being applied to the whole saliva.
Another advantage of this study is that it does not include an additional dilution step. The gap-LFIA is a simple device consisting of a sample pad having a 3 mm gap, modified from conventional LFIA strips, so that it can be applied directly to the sensor without dilution by using a cotton swab as a collection tool. The saliva-absorbed cotton swab can be easily inserted into the gap on the divided pad, and then, the addition of buffer solution on the sample pad dilutes saliva samples in the swab (
Figure 1). The saliva does not flow all at once, but the buffer flows, and the saliva dissolves little by little. It reduces the possibility that the undiluted saliva flow first and the membrane be clogged with impurities by the viscous material in the saliva. In this study, the gap-LFIA was designed and tested as a salivary cotinine detection method for its rapid, convenient, and cost-effective use for smoking detection. Compared to the conventional LFIA with saliva, the method allows the direct use of the whole saliva collected using a cotton swab from mouth to be applied directly to a cotinine sensor. In addition, the accuracy was improved by reducing the deviation without additional preparation or equipment.
2. Materials and Methods
2.1. Chemicals and Instruments
Free cotinine, anti-mouse immunoglobulin G (IgG) antibody, and skim milk powder were purchased from Sigma-Aldrich (St. Louis, MO, USA). Gold nanoparticles (AuNPs) were purchased from BB International (Cardiff, UK). Anti-cotinine antibodies and cotinine-bovine serum albumin (BSA) were purchased from CalBioreagent (Foster City, CA, USA). Phosphate-buffered saline (20× PBS, pH 7.4) was obtained from Biosesang Co. (Seongnam, Korea). Borate buffer (20×, 1 M, pH 8.5) was purchased from Thermo Scientific (Waltham, MA, USA). Nitrocellulose membranes (FF80HP) were purchased from GE Healthcare (Seoul, Korea). Absorbent (Grade222), conjugate (Glass fiber 6613), and sample pads (Grade222) were purchased from Alshtrom Munksjö (Helsinki, Finland). A salivary cotinine quantitative enzyme immunoassay kit was purchased from Salimetrics. (Carlsbad, CA, USA). A microcentrifuge (Hanil, Kimpo, Korea) was used to separate free antibody and skim milk from the AuNP conjugates. A dispenser (DCI100, Zeta Co., Gunpo, Korea) was used to immobilize the cotinine-BSA and anti-cotinine antibodies on test and control lines of the FF80HP membrane. A drying oven (KO-100, LK Lab Co., Namyangju, Korea) was used to incubate the samples and to dry the membranes. A cutting device (TBC-50Ts, Taewoo Co., Namyangju, Korea) was used for making the test strips. All visible measurements were performed using a ChemiDoc XPS+ imaging system (Bio-Rad, Hercules, CA, USA) and analyzed with the Image Lab software (ver. 5.2.1) (Bio-Rad). The absorbance of ELISA was detected using Cytation 5 Cell Imaging Multi-Mode Reader (BioTek, Winooski, VT, USA).
2.2. Construction of Gold Conjugates with Antibody
An AuNP solution (40 nm in diameter, 1× AuNP, λmax optical density = 1.0) (1 mL) was added to 100 μL of 100 mM borate buffer (pH 8.5), to which 10 μg of the anti-mouse IgG antibodies was added. After incubation for 30 min at room temperature, 100 μL of 0.1% skim milk in 10 mM borate buffer (pH 8.5) was added to block the surface of AuNP conjugates. After incubation for 30 min, the conjugates were washed with 10 mM borate buffer (pH 8.5) three-times by centrifugation (6500 rpm, 20 min, 10 °C) and resuspended in a 50 μL storage buffer (5% trehalose, 0.5% skim milk, 0.2% Tween 20, and 1% Triton X-100 in 1× PBS) to maintain the affinity of the conjugates. The 20× AuNP conjugate stock was stored at 4 °C.
2.3. Preparation of Gap-LFIA Strip Sensor
For cotinine LFIA detection, 1 mg/mL of cotinine-BSA (test line) and 100 μg/mL anti-mouse IgG antibody (control line) were immobilized on the FF80HP membrane with 1 μL/cm of each solution. The distance between the test line and control line in the LFIA strip was approximately 5 mm. After immobilization, the membranes were oven-dried at 37 °C for 1 h. An absorbance pad was attached to the upper part of the membrane with a 3 mm overlap, was cut to 3.8 cm width using the cutting device, and was stored in a humidity-controlled chamber (21~23% humidity) before use. Conjugate pads (3.8 × 4.0 mm), pre-blocked with the storage buffer, were applied with 3.8 μL of the anti-cotinine antibody (conjugate pad 1) and AuNP-antibody conjugates (conjugate pad 2), both diluted with 1× PBS having 1% polyvinylpyrrolidone (PVP) and 0.5% Tween 20. The conjugate pads were dried and attached to the bottom of the strip with a 3 mm overlap. A sample pad, divided into two segments with a 3 mm gap for inserting the cotton swab, was placed further below as depicted. The gap-LFIA strip sensor was constructed immediately before the experiments.
2.4. Gap-LFIA Testing of Cotinine in a Buffer and Spiked Saliva
A standard cotinine stock solution was serially diluted to 1, 10, and 100 ng/mL using 1× PBS and spiked at the same concentrations into real saliva. At least 120 μL of cotinine sample solution was taken to saturate each cotton swab, which was then inserted into the gap in the divided sample pads of the test strip. Then, the buffer (100 μL, 1× PBS containing 1% PVP and 0.5% Tween 20) was added to the bottom sample pad to flow along the strip through the cotton swab, carrying the sample toward the sensor region (
Figure 2). The added buffer mixes with the saliva contained in the cotton swab and moves toward the membrane. The strip was incubated for 15 min at room temperature, and the intensities of the control and test lines were observed. All images were obtained using the ChemiDoc XPS+ imaging system and analyzed using the Image Lab 6.0 program. First, the various concentrations of the anti-cotinine antibody (12.5, 25, and 50 ng/mL) and 1 mg/mL of cotinine-BSA, which is sufficient for competition with free cotinine in the saliva samples, were tested. To determine the optimal concentration of the anti-cotinine antibody, the ratio of changes of control to test (C/T) line density was compared by analyzing cotinine concentrations.
2.5. Test of Real Samples Collected from Smokers and Non-Smokers and Comparison with ELISA
Human saliva samples for cotinine detection were obtained from 14 volunteers including 6 smokers, and all participants aged 20 to 50 gave informed consent to the study approved by the Gwangju Institute of Science and Technology. The amount of smoking in the last 24 h before saliva collection was recorded for each volunteer. The saliva samples were collected by absorbing several cotton swabs and spitting in snap tubes from the mouth of the volunteers without a hygiene procedure. The cotton swabs and saliva samples were stored in each snap tube at 4 °C, and the collected samples were used within 24 h for the gap-LFIA tests. For the verification of the gap-LFIA, the results were compared with the results of a commercial salivary cotinine ELISA kit using the same saliva samples. Further, in this study, high concentrations of caffeine and nicotine with cotinine were tested using the sensor to confirm its selectivity. Caffeine is an alkaloid that can often be found in human saliva, and nicotine is the precursor of cotinine with a similar structure [
15].
4. Conclusions
Simple cotinine detection for smoking is important not only for individuals but also for public health. Although many cotinine assay kits have been commercialized, they involve various inconvenient processes to handle salivary samples. In this study, the gap-LFIA using cotton swab was shown as an easy-to-handle method for detecting cotinine from saliva. The gap-LFIA could apply the saliva directly to the strip without additional sample preparation steps and could dilute the saliva samples from the cotton swabs reducing background hindrance. As stated in the results, 1 ng/mL salivary cotinine was measured in 15 min by the image program. Moreover, the non-smokers and smokers could be more accurately distinguished, compared with the results obtained from the conventional LFIA without the gap. The results of gap-LFIA showed a high correlation with the conventional ELISA method, showing that it is reliable, fast, and easy to use for cotinine detection in saliva. In addition, the gap-LFIA can also be utilized for direct application of other different types of samples such as blood, seawater, or powder.