1. Introduction
Illnesses and deaths caused by foodborne microbial pathogens have received major attention worldwide. Hence, food safety is a serious public health issue. In 2013, Foodborne Diseases Active Surveillance Network (FoodNet) reported 19,056 cases of infections, 4200 hospitalizations, and 80 deaths in 10 U.S. sites (affecting approximately 15% of the U.S. population) due to foodborne diseases [
1]. Failure to detect foodborne pathogens in contaminated food may not only lead to far-reaching consequences on human health but also causes a large economic burden on the food industry. Research on rapid and reliable method of foodborne pathogen detection is still on going.
Conventional methods available for the detection and identification of microbial pathogenic agents are based on selective microbiological media to isolate the viable bacterial cells present in food [
2,
3]. As these media depend on the ability of the organisms to multiply to become visible colonies, it requires 2–3 days to have probable results and up to 5–7 days for confirmation [
3,
4,
5,
6]. Moreover, it is labor intensive and cumbersome as the analysis of samples require several steps e.g., culture medium preparation, inoculation of plates and colony counting [
2].
For sensitive detection of pathogens with automated or semi-automated instruments in near real time, several biosensors and bio-based methods have recently been developed. These include bioluminescence sensors [
2], Surface Plasmon Resonance (SPR) sensors [
6,
7,
8], electrochemical immunosensors [
9], Fluorescence Resonance Energy Transfer (FRET) [
5,
10,
11,
12], piezoelectric biosensors [
13] and cell based sensors [
14,
15,
16]. In order to reduce analysis period and achieve confirmation results for detection, several methods like the Polymerase Chain Reaction (PCR) [
17,
18,
19,
20], immunoassay [
21,
22,
23] and microarray [
24] have been developed. However, these methods and techniques require complex manual steps, skilled personnel and high-end sophisticated equipment. These requirements limit the accessibility of these techniques especially in resource-poor areas. Besides, most centralized and modern laboratories are limited to large cities. Therefore, near-patient tests that utilize cheap and reliable Point of Care Test (POCT) devices have become increasingly important to provide an alternative rapid, sensitive, accurate and automated method for detection of foodborne pathogens.
Lab-on-a-Chip (LOC) device also called “Micro Total Analysis System” (µTAS) is a reliable choice as a POCT device. Miniaturized device size, low consumption of reagent and sample, precise microfluidic volume control, less manual handling errors and automated fast turn-around time for operation of biochemical assay enable the integration of all necessary processes and laboratory operations into a single chip [
25,
26]. These compact devices can analyze the samples at the point of need rather than in a centralized highly equipped laboratory. Various researches have been conducted on microfluidic chips for pathogens detection [
27,
28,
29,
30,
31,
32]. Easley
et al. [
29] reported a microfluidic genetic analysis system for detection of
Bacillus anthracis by solid-phase extraction (SPE), PCR and microchip electrophoresis (ME) in less than 30 min. Beyor
et al. [
30] developed a laboratory-on-a-chip system for the detection of
Escherichia coli K12 and
E. coli O157 integrating cell pre-concentration, purification, PCR and capillary electrophoretic (CE) analysis on one single platform. Even though steps for pathogen detection start from target separation to endpoint detection are integrated on a single chip, it requires complicated tube connections for pneumatic fluid control which limits the miniaturization of full system.
Comparing all the LOC platform, lab-on-a-CD platform has received significant attention for its potential to integrate all biochemical reactions onto a single disc-pattern microfluidic device and perform multiplexed operations using a single motor. It leads toward the development of microfluidic sample-to-answer systems or µTAS
in vitro diagnostics (IVD) [
33,
34]. The centrifugal, inertial and coriolis forces generated by the disk rotation can be employed for fluidic manipulation
i.e., pumping [
35], mixing [
36,
37], metering [
38,
39], decanting [
40], calibrating [
40], sample splitting [
40], fluid separating [
40] and valving [
41]. Several biochemical operations have been implemented successfully on microfluidic compact disk such as biochemical analysis and immunoassay [
42], detection of protein [
43], fluorescence immunoassays [
44], DNA extraction [
45], nutrients determination in water [
46], ELISA system [
47,
48] and foodborne pathogen detection [
49].
Among several isothermal nucleic acid amplification techniques, LAMP is one of most established methods [
50,
51]. LAMP provides a secure reaction temperature range of 60–65 °C as compared to other low temperature methods, for example—Nucleic Acid Sequence-Based Amplification (NASBA), Helicase-Dependent Amplification (HDA), Strand Displacement Amplification (SDA) and so on—because in all these methods there is a risk that the reaction is initialized before the completion of the reaction preparation [
50]. Therefore LAMP would be very beneficial as a POCT diagnostic device with its high sensitivity and specificity. In recent years, Ajima Muangchuen
et al. [
52] were able to develop a colorimetric endpoint detection technique based on DNA amplification by LAMP for
Ehrlichia canis, Wu
et al. [
53] have developed an integrated glass microdevice for LAMP, Lee
et al. [
54] have developed an integrated micro-reactor system that is able to detect Hepatitis B virus (HBV) DNA using LAMP reaction, Feiwu Li
et al. [
55] have also developed a visual and rapid LAMP assay for detection of the
cry2Ab and
cry3A genes in GM crops, these require either naked eye observation or UV-Visible Spectroscopy interfaced with a computer for obtaining confirmative results. Liang
et al. [
56] have reported a close-tube method using a wax-sealed fluorescent intercalator to detect LAMP product. Changchun Liu
et al. [
57] developed a single-chamber LAMP cassette to detect HIV-1 in oral fluid which requires a portable ESE optical detection system interfaced with a computer to get the confirmative results from generated graph. The developed system called micro-LAMP (μLAMP) by Fang
et al. [
58] required naked eye observation or compact real-time absorbance detection device for end-point detection which is expensive and complex system. Analyzing the current techniques for endpoint detection of LAMP reaction, it is necessary to develop an endpoint detection system which can provide automatic and digital confirmative results.
In this study, LAMP was performed in tubes with Bst DNA polymerase and a set of specially designed six primers which were incubated in 63 °C constant temperature for 60 min, followed by enzyme inactivation at 80 °C for 2 min to complete the reaction. We developed a low cost portable system equipped with an ultraviolet (UV) emitter and a color sensor for the purpose of automation and digitization for the endpoint detection technique of LAMP amplicons.
3. Results and Discussion
An Epoch Microplate Spectrophotometer and Synergy H1 Hybrid Multi-Mode Microplate Reader were utilized to detect absorbance and fluorescence of the resulting dye of LAMP amplicons with SYBR Green I, respectively.
Figure 5 shows the resulting dye absorbs UV light (λ
max = 262 nm) and emits green light (λ
max = 524 nm). A UV emitter of 365 nm peak wavelength was utilized for excitation of LAMP amplicons rather than 262 nm because of the cost and availability of the component. Moreover, the level of excitation from LAMP amplicons using 365 nm UV emitter is detectable using the color sensor. The group of green photodiodes in color sensor was enabled, because its relative spectral sensitivity is more suitable with the emission spectrum of LAMP amplicons dye as compare to other groups of photodiodes.
Figure 5.
Spectrogram of LAMP Amplicons with Salmonella bacteria.
Figure 5.
Spectrogram of LAMP Amplicons with Salmonella bacteria.
A total of seven LAMP reactions were performed with seven different concentrations of DNA templates (10-fold serial dilutions from 2.5 × 10
0 ng/µL to 2.5 × 10
−6 ng/µL) of
Salmonella bacteria and one reaction with deionized water (DW) as a negative control (
Figure 6A) and then the dye was transferred in the detection chamber of microfluidic CD (
Figure 6B). According to
Figure 6A, the initial orange color of SYBR Green I changed to yellowish green in tube number 1 to 4 (positive result), the color turned to faded orange in tube number 5 to 7 (negative result) and the orange color remained unchanged for tube number 8 (negative result). These results indicate the lower limit of detection is 2.5 × 10
−3 ng/µL for the DNA of
Salmonella bacteria. As the LAMP detection with SYBR Green I is a qualitative detection process, it gives same intensity of color visually for either positive or negative results.
Table 1 and
Figure 7 show the sensitivity test results of the detection system when the resulting dye is loaded in the microfluidic CD. The processed sensor output value (pulse duration) as described in the methodology section was normalized with feature scaling method.
Table 1 and
Figure 7 also illustrate the mean sensor readings with the standard deviation obtained for three different sets of data. These results indicate the normalized value range of sensor output was less than 0.1 (green triangle markers) for the positive control and higher than 0.8 (orange circle markers) for the negative control. As the frequency of sensor output pulse train is proportional to the incident light intensity on the sensor surface and the sensor was operated in green filter mode, which caused higher output pulse frequency and lower output pulse duration for positive control than negative control.
Figure 6.
Naked-eye inspection (A) Resulting dye (26 µL) of LAMP amplicons (25 µL) and SYBR Green I (1 µL) in ambient light. (B) Microfluidic CD loaded with the resulting dye.
Figure 6.
Naked-eye inspection (A) Resulting dye (26 µL) of LAMP amplicons (25 µL) and SYBR Green I (1 µL) in ambient light. (B) Microfluidic CD loaded with the resulting dye.
Table 1.
Sensitivity test results of the endpoint detection system at different concentration of Salmonella DNA template.
Table 1.
Sensitivity test results of the endpoint detection system at different concentration of Salmonella DNA template.
| Sample ID | Concentration of DNA Template (ng/µL) for LAMP Reaction | Color Changes of Resulting Dye ** | Visual Identification of Resulting Dye ** | Sensor Reading (Normalized Pulse Duration) Mean ± SD | Automatic Detection System Interpretation of Resulting Dye ** |
|---|
| 1 | 2.5E+00 | Yellowish Green | Positive | 0.067 ± 0.038 | Positive |
| 2 | 2.5E-01 | Yellowish Green | Positive | 0.025 ± 0.035 | Positive |
| 3 | 2.5E-02 | Yellowish Green | Positive | 0.045 ± 0.043 | Positive |
| 4 | 2.5E-03 | Yellowish Green | Positive | 0.065 ± 0.033 | Positive |
| 5 | 2.5E-04 | Faded Orange | Negative | 0.834 ± 0.042 | Negative |
| 6 | 2.5E-05 | Faded Orange | Negative | 0.832 ± 0.035 | Negative |
| 7 | 2.5E-06 | Faded Orange | Negative | 0.820 ± 0.037 | Negative |
| 8 | 0 * | Unchanged | Negative | 0.971 ± 0.037 | Negative |
Figure 7.
Sensitivity test results of the detection system with different concentrations of Salmonella DNA template. The target is classified as positive when the sensor reading is lower than 0.1 and as negative when the sensor reading is higher than 0.8.
Figure 7.
Sensitivity test results of the detection system with different concentrations of Salmonella DNA template. The target is classified as positive when the sensor reading is lower than 0.1 and as negative when the sensor reading is higher than 0.8.
As a parallel assessment, sensitivity test of LAMP assay and PCR assay were tested using 10-fold serial dilutions of pure Salmonella cultured DNA. The PCR can detect up to 3.4 × 106 CFU/mL of pure Salmonella culture. The detection limit of LAMP on the same set of serially diluted culture was 3.4 × 104 CFU/mL which is 100 times more sensitive than conventional PCR.
Analyzing the sensitivity test results of the detection system with different concentrations of
Salmonella bacteria DNA, the developed detection system can detect the LAMP amplicons successfully. The current endpoint detection techniques of LAMP amplicons are color changes observation of fluorescent dye with naked eye [
52,
53,
55,
56,
58] or spectrometer [
52,
54], accumulated white pellet visualization with turbidimeter after a brief centrifugation of LAMP amplicons [
55] and agarose gel electrophoresis showing ladder of DNA bands [
58,
61]. These methods take 2–5 min or more (depends on the number of LAMP reactions performed) to perform and require presence of personnel to make a confirmative decision (positive or negative). On other hand, this developed detection system can detect 16 samples of LAMP amplicons within 6 s automatically with one hundred percent accuracy. The detection results could be transferred from the detection system to a smart phone and remote places sequentially to improve the contamination monitoring and response time. The developed detection system has succeeded in improving the endpoint detection of LAMP amplicons in terms of time and labor. Moreover, the estimated manufacturing cost of the detection system is less than USD 160 which can be considered as low cost compared to existing detection techniques.