- Email: [email protected]
Smartphone-based portable biosensing system using cell viability biosensor for okadaic acid detection
Accepted Manuscript Title: Smartphone-based portable biosensing system using cell viability biosensor for okadaic acid detection Authors: Kaiqi Su, Yuxiang Pan, Zijian Wan, Longjie Zhong, Jiaru Fang, Quchao Zou, Hongbo Li, Ping Wang PII: DOI: Reference:
S0925-4005(17)30636-6 http://dx.doi.org/doi:10.1016/j.snb.2017.04.036 SNB 22122
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
31-8-2016 11-3-2017 7-4-2017
Please cite this article as: Kaiqi Su, Yuxiang Pan, Zijian Wan, Longjie Zhong, Jiaru Fang, Quchao Zou, Hongbo Li, Ping Wang, Smartphone-based portable biosensing system using cell viability biosensor for okadaic acid detection, Sensors and Actuators B: Chemicalhttp://dx.doi.org/10.1016/j.snb.2017.04.036 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.
Smartphone-based portable biosensing system using cell viability biosensor for okadaic acid detection Kaiqi Su1, 2, Yuxiang Pan1, Zijian Wan1, Longjie Zhong1, Jiaru Fang1, Quchao Zou1, 3, Hongbo Li1, Ping Wang1, 2, *
1. Biosensor National Special Laboratory, Key Laboratory for Biomedical Engineering of Education Ministry,
Department of Biomedical Engineering, Zhejiang University, Hangzhou310027, PR China
2. State Key Laboratory of Transducer Technology, Chinese Academy of Sciences, Shanghai 200050, China
3. Department of clinical engineering, the Second Affiliated Hospital of Zhejiang University School of Medicine,
Hangzhou 310009, Zhejiang Province, PR China
* Corresponding author: Tel.: +86 571 87952832; Fax: +86 571 87952832. E-mail address: [email protected] (Ping Wang)
Smartphone-based portable biosensing system using cell viability biosensor for okadaic acid detection
Highlights
A smartphone-based biosensing system using CVBS was developed for OA detection.
The system presented the good robustness in long-term image capture and analysis.
The system achieved label-free, non-invasive and long-term monitoring of cell viability.
The cost-effective CVBS was constructed by HepG2 cells, MTP and CCK-8 kit for OA detection.
The iPlate Monitor used traversal algorithm to obtain the best detection point-in-time.
Abstract
Okadaic acid (OA), as a diarrheic shellfish poisoning toxin, had wide distribution and frequent occurrence. Therefore, low-cost, high-throughput, wide-range and portable detection of OA was in high demand for food safety and environmental monitoring. In this study, a novel and portable smartphone-based system using cell viability biosensor (CVBS) was developed for label-free, non-invasive and long-term monitoring of cell viability. The variation of cell viability reflected the changes of cell morphology, cell count and cell proliferation indirectly. And this system applied the combination of image analysis and cell counting kit-8 assay (CCK-8) to monitor the reflection. The biosensing system chose HepG2 cells as sensing elements to build CVBS and used it in OA detection. Results showed this system could synchronously detect OA in 96 channels. And this biosensor presented a good performance to various OA concentrations, with a wide linear detection range (10 - 800 μg/L). Moreover, the point-in-time having best detection performance could be located by the traversal algorithm in the monitoring duration. Thus, this cell-based biosensor system provided a convenient and efficient approach in seafood safety testing such as OA screening.
Keyword: Cell viability biosensor (CVBS); Smartphone; Portable colorimetric reader; Shellfish toxins; Okadaic acid (OA)
1. Introduction OA, as a kind of diarrheic shellfish poisoning toxin, is produced by some unicellular algae from plankton and benthic microalgae and accumulates in the digestive glands of shellfish [1]. As one inhibitors of serine/threonine protein phosphatases type 1 (PP1) and 2A
(PP2A) [2], OA can cause some diarrheic symptoms including diarrhea, nausea, vomiting and abdominal pain [3]. Moreover, OA has been identified as a tumor promoter and has been proved that it had mutagenic and immunotoxin effects [4]. Currently, there are some biochemical methods including mouse bioassay (MBA) [5], pre-column oxidation liquid chromatography with fluorescence detection [6], high performance liquid chromatographymass spectrometry [7, 8], enzyme linked immunosorbent assay [9, 10] and cell-based biosensors [11, 12] for the detection of marine shellfish toxins. Thereinto, the research using cell-based biosensors in shellfish toxin detection is beginning to attract the people’s attention due to their abilities of observing the impact of toxin to cells and the quantitative analysis to toxin. However, shellfish toxins such as OA have wide distribution and frequent occurrence. Hence, the field of shellfish toxin detection requires a low-cost, high-throughput, wide-range and portable biosensing system for generalization, while the above cell-based biosensors can hardly meet these requirements.
How to establish a cell-based biosensing system, which can achieve low-cost, highthroughput, wide-range and portable detection? It is a great choice to combine image analysis with smartphone-based application. For biomedical research, the image analysis is utilized in the construction of low-cost cell-based biosensor [13, 14]. On the other hand, many biochemical analysis platforms or methods combined with smartphone appear one after another in point-of-care test domain due to the high integration of smartphone to processor and diverse sensors [15, 16]. These smartphone-based applications distribute in different fields including smartphone-based microscopy [17, 18], fluorescent imaging [19], imaging cytometry [20], electrocardiography [21, 22], lateral flow assays [23, 24], surface plasmon
resonance-based sensing [25], electrochemical sensing [26, 27], immunoassays [28-30], and other applications [31-33]. Also, the accessories of smartphone-based system appear one after another to improve detection performance [34-36].
In this study, a novel and portable smartphone-based system called CVBS system was developed in accordance with the combination of image analysis and smartphone-based application. Cooperating with CCK-8 kit that had no cytotoxicity, this biosensing system could achieve label-free, non-invasive and long-term monitoring of cell viability. Besides, HepG2 cell lines were chosen as sensing elements to specifically detect OA. This smartphone-based system presented the features of low cost, high throughput, wide detection range and portability in OA detection. Also, the cell culture process used in this system was more common than that of other cell-based biosensors, which could promote the generalization of this system in the field of shellfish toxin measurement. All the details will be discussed in the following sections.
2. Materials and methods 2.1 Reagent and Setup
The 96-well polystyrene plates are bought from Thermo Fisher Scientific, Germany. HepG2 cell line is bought from American Type Culture Collection and all of the reagents of cell culture are purchased from Gibico, USA. DMEM medium (Dulbecco’s modified eagle medium) with high glucose, fetal calf serum (FBS), 0.25% Trypsin–EDTA, and phosphate buffered saline (PBS) are obtained from Gibico, USA. Gonyautoxin2&3 (GTX2&3) and brevetoxin-2 (PbTx-2) are purchased from National Research Council (Canada). OA (Sigma,
USA) stock solution is prepared in DMSO and filtered with 0.22 μm membrane filter unit (Millipore, USA). OA is diluted by DMEM medium with high glucose when it is used to treat with cells. CCK-8 kit is obtained from Shanghai 7 sea biotech Co., LTD., China. The morphological changes of cells are observed by a stereo microscope (Sharp Inc., Japan).
2.2 Action principle of CCK-8 kit
For label-free, non-invasive and long-term monitoring of cell viability, CCK-8 kit is chosen as the developer of system. Comparing with other detection kits of cell proliferation and cytotoxicity such as MTT, XTT and WST-1, CCK-8 has excellent stability, a wider detection range, higher sensitivity and is easy to use. More importantly, it has no cytotoxicity. Figure 1A shows the action principle of CCK-8 kit. CCK-8 kit is based on water-soluble tetrazolium salt-8 (WST-8) [37]. And WST-8 can be reduced by dehydrogenases in cells to give a water-soluble formazan dye (WST-8 formazan) in the presence of electron mediator (1Methoxy PMS). The WST-8 formazan is an orange colored product and reflects the living cell status. Hence, this system monitors the cell viability by the variation of orange intensity over time.
2.3 The portable smartphone-based biosensing system
This system consists of CVBS, illumination provider and smartphone. The CVBS includes living cells, microtiter plate (MTP) and CCK-8 kit (Figure 1B). The basic structure design of illumination provider has been described elsewhere [38] and the contour structure has been upgraded according to industrial design in this study (Figure 1C). The smartphone installed with homemade iOS APP - iPlate Monitor (designed by Swift 1.0 and Object-C in
Xcode 7) undertakes the image acquisition, image analysis, data storage and transmission (Figure 1D). The iPlate Monitor introduces cell viability index (CVI) and normalized cell viability index (NCVI) to evaluate cell viability monitoring and analyze the response of CVBS. CVI and NCVI are calculated by the equ1 and equ2, respectively. CVI = BBlank − BTest equ1
Where CVI, BBlank and BTest stand for cell viability index, blue channel value of blank MTP and blue channel value of CVBS, respectively. CVI
NCVI = CVI t
0
equ2
Where CVIt and CVI0 are the CVI at any point-in-time and the CVI at 0 h, respectively. To confirm the best detection point-in-time, iPlate Monitor applies the traversal method to perform linear fitting at all time-points of monitoring process. Afterwards, the point-in-time of fitting curve having best performance, which is the best detection point-in-time, is determined by the fitting goodness and sensitivity. The supplementary material (Figure S1) shows the workflow of this APP and some actual smartphone screen picture during OA detection experiments.
Fig. 1
2.4 Cell culture
HepG2 cell line is cultured in DMEM medium which includes 10% heat inactivated FBS and 0.5% antibiotic solution (10 mg/mL streptomycin and 1000 U/mL penicillin). Then it is incubated at 37 °C in humidified air with 5% CO2 in an incubator (Thermo, USA). When the confluent cells reach 80%, 0.25%, Trypsin-EDTA is used to dislodge cells from the flask to 96-well plates.
2.5 Shellfish extracts preparation
Mytilus edulis are selected for actual sample testing and the non-toxin samples are purchased from the market. The preparation of shellfish extracts are carried out according to the protocol offered by Ledreux et al. [39].
2.6 Experiment setup
2.6.1 Cell viability monitoring of smartphone-based system HepG2 cells are detached from the culture flask and 100 μL cell suspensions with different cell seeding densities are prepared. Then HepG2 are inoculated in the 96-well MTP at different seeding densities. After cell inoculation, HepG2 are cultured for 4 h to adhere. Then 10 μL of CCK-8 solution is added to each well of the MTP and the MTP with smartphone-based system are placed inside a humidified incubator (at 37 °C, 5% CO2). Meanwhile, the detection system starts to monitor the cell viability curves. 2.6.2 Real-time monitoring of CVBS’s response to OA 100 μL of HepG2 cell suspensions with the same cell seeding density are added onto the MTP to build HepG2-CVBS. Before exposure to toxins, the cells are cultured for 24 hours to
achieve cell adherence and cell status stability. Then the medium of treated groups is removed and replaced by 100 μL fresh medium and 10 μL different concentrations of OA. Subsequently, 10 μL of CCK-8 solution and the CVBS with the smartphone-based system are placed in the incubator (at 37 °C, 5% CO2). Meanwhile, the detection system starts to monitor the responses of HepG2-CVBS to OA.
3. Results and discussion 3.1 Stability of image monitoring using the portable smartphone-based system
Stability was the core of real-time monitoring and pixel intensity was the initial output of system. The system captured image and calculated pixel intensity once each 2 minutes. As shown in Figure 2, the performance of system was respectively tested with RGB color model. The monitoring time was about 20 h. The system showed stable pixel intensity measurement, although having slight fluctuations in monitoring process. Table S1, Table S2 and Table S3 further illustrated the detecting stability of system. For all 96 wells of MTP, the coefficients of variation (CV) of red, green and blue channel were within 1.08%, 0.66% and 0.71%, respectively. The detecting differences of pixel intensity measurements were often less than 2.71% (red < 2.71%, green < 1.79% and blue < 1.99%).
Fig. 2
3.2 Dynamically monitoring of cell viability and CVBS construction
HepG2 cells were inoculated onto the MTP at 0, 5000, 10,000, 20,000, 30,000, 40,000, 50,000 and 60,000 cells/well. CVI curves were recorded for about 5 h after CCK-8 addition
(Figure 3A) and the monitoring process could be reappeared in the Video S1. The CVI had positive correlation with the orange depth. And the growth speed of CVI was positively associated with the cell density. The range of CVI was from 0 to 255 according to the range of pixel intensity. Hence, the CVI remained stable after it reached the maximum.
Figure 3B shows the variations of R2 and sensitivity of linear fitting curve over time. Commonly, two time points were picked out, which were the highest R2 and the highest sensitivity, respectively. In this experiment (Figure 3C), the best point-in-time was determined at 1.07 h (R2 was 0.9956 and slope was 169.4), which had both the highest R2 and sensitivity. The limit of detection (LOD) was calculated by 3δ/slope. At the best point-intime, the LOD of cell density detection was 3684 cells/well.
Fig. 3
To verify that CCK-8 kit had no cytotoxicity, the cellular morphology of two experimental groups with and without CCK-8 kit was observed by the stereo microscope. Figure 4 shows the cell growth situation under the views of 40 and 100 magnifications at the best detection point-in-time above. It was apparent that the number of cells was positively correlated with the initial seeding density. And the cells revealed their great growth situation and attachment status whether if CCK-8 was added or not. The high density cells were more resistant to toxin treatment. Hence, 5,000 cells/well was the suitable seeding density that CVBS could obtain relatively high sensitivity.
Fig. 4
3.3 Real-time response of CVBS to OA detection
Before adding marine toxins, HepG2 cells (5,000 cells/well) were seeded to build HepG2-CVBS. And different concentrations of OA (10, 25, 50, 100, 200, 400 and 800 µg/L) were tested. Figure 5A shows the real-time NCVI curves, which reflects the cell activity changes under OA treatment. The increase rate of NCVI was negatively linked with OA concentration. And the higher concentration of OA, the faster NCVI curve reached the plateau.
Figure 5B shows the variations of R2 and sensitivity of linear fitting curve over time. Two time points (2.85 h and 10.55 h) were picked out, which were the highest R2 and the highest sensitivity, respectively. Since the R2 of standard curve at 10.55 h was too low (R2 was 0.7772), 2.85 h was determined as the best detection point-in-time. Figure 5C shows the standard curve at 2.85 h (R2 was 0.9536 and slope was -2.569). At the best detection point-intime, the equation of standard curve was y = -2.569x + 13.8 (R2=0.9536), where y was NCVI, and x was logarithmic concentration of OA. The LOD was calculated by 3δ/slope, namely, the triple standard deviation (STD) was used to work out the LOD (33.9532 µg/L). Since the maximum concentration of OA in shellfish was permitted as 160 µg/kg by European Commission (EC) (EC 853/2004 15), the proposed method could reach this level. Moreover, MBA was used as the reference method for OA analysis proposed by AOAC and the detection limit was 0.05 MU/g equaling to 220 µg/kg. And about 12MU (52.8µg) was the
lowest amount of OA inducing mild form of illness (e.g., nausea, diarrhea and vomit) to an adult [40]. Suppose a maximum intake of mussels was 250 g, the lowest concentration of OA to start toxic symptoms was 211.2 µg/kg. This level could be detected by this biosensing system.
Fig. 5
To verify the response of CVBS to OA, the OA-induced morphological changes in HepG2 cells was observed by the stereo microscope at the best detection point-in-time. Figure 6 shows the cell growth situation under the views of 40 and 100 magnifications after continuous treatment with different concentrations of OA. It was apparent that the HepG2 cells had distinct morphological changes after the OA treatment. Meanwhile, the depth of morphological change was related to the OA concentration and treatment duration.
Fig. 6
For the real sample detection and the comparison between MBA and CVBS, nontoxic shellfish extracts were spiked with the concentrations of OA (300 – 700 µg/L). The concentration setup was in accordance with the detection limit (220 µg/kg) of MBA for OA analysis proposed by AOAC. As shown in Table 1, the results of CVBS were highly correlated to those of MBA. The percentage recoveries measured by CVBS and MBA were in the ranges of 93.54 – 104.03 and 93.92 – 107.32 for OA-spiked shellfish extracts,
respectively. And the corresponding average percentage recoveries were 99.06 and 100.16, respectively.
Tab. 1
3.4 Specificity of CVBS
The specificity of HepG2-CVBS to OA could be deduced by the previous study which adopted similar cell-based biosensor for marine toxin detection. OA could strongly influence the cell morphology, cell attachment and cell viability, which was the methodology basis of some cell-based biosensors, such as cell-based impedance biosensor (CIB), Love Wave biosensor and CVBS, to detect OA. The cell-based Love Wave biosensor using HepG2 cells compared the detection results in saxitoxin (STX) and brevetoxin 3 (PbTx-3). Then it was certified that this cell-based Love Wave biosensor presented a good specificity for OA detection. Similarly, the high selectivity of CIB using HepG2 cells was proved by the comparison results in STX and brevetoxin 2 (PbTx-2). Since the mechanism action of these marine neural toxins was different from that of OA, the above cell-based biosensors had a good specificity for OA detection.
Fig. 7
The specificity of HepG2-CVBS was evaluated by gonyautoxin2&3 (GTX2&3) and PbTx-2. GTX2&3 was one of the paralytic shellfish poisoning (PSP) toxins, and it could affect the propagation of the action potential by blocking sodium channels, which would
change the membrane potential on a cellular level and cause neurological symptoms in human [41]. PbTx-2 was a kind of neurologic shellfish poisoning (NSP) toxins which induced the enhancement of sodium inflow into cells [4]. The concentrations of GTX2&3 and PbTx-2 applied in this specificity assay were 400 μg/L and 800 μg/L. From the results shown in Figure 7A, it was obvious that the NCVI variations caused by the two toxins were not much different from those of control groups. According to the above best detection point-in-time (2.85 h), the responses of CVBS to three toxins were compared (Figure 7B). The NCVI caused by OA was lower than those which were induced by GTX2&3 and PbTx-2. And the OA experiment group had significant difference from the control group (levels of significance: ***p < 0.001; ****p < 0.0001) while the GTX2&3 and PbTx-2 groups didn’t have. Therefore, HepG2-CVBS had a good specificity for OA detection.
3.5 Performance of smartphone-based biosensing system using CVBS for OA detection
Living cell status is useful to monitor the changes of chemicals in microenvironment. In this study, cell viability was used as sensor response in OA detection. The biosensing system utilized the optical image analysis for a label-free, non-invasive and long-term monitoring of cell viability and constructed CVBS to detect OA.
In the previous report [38], bionic electronic eye presented inconvenience and inaccuracy in the kinetics assay due to the lack of automation. For instance, since cells needed stable cultivation environment, the kinetics assay of CCK-8 employed multiple MTPs to compare the performance of microtiter plate reader and bionic electronic eye at different point-intimes. Therefore, the system in this work was developed to a longtime and real-time
monitoring system by the homemade software – iPlate Monitor. The iPlate Monitor achieved longtime and real-time image acquisition and analysis, which met the requirements of monitoring to CVBS having living cells in the incubator. Moreover, the iPlate Monitor used traversal algorithm to obtain the best detection point-in-time, which could enhance the detection accuracy.
CVBS consisted of living cells, MTP and CCK-8 kit, which was a simple construction comparing with some cell-based biosensors such as CIB and Love Wave biosensor. CCK-8 could be reduced by dehydrogenases in cells to give an orange water-soluble formazan dye. The orange intensity reflected the viability of cells. Thus CVBS used CCK-8 to transform the cell viability to optical signal and this signal could be detected by the portable biosensing system. The durability of CVBS related to its structure. Thereinto, living cells and CCK-8 kit belonged to disposable supplies. MTP could be reused about three times for adherent cell cultivation after a series of operations including abstersion, overnight soak with acid, distilled water wash and ultraviolet radiation. However, MTP was also regarded as disposable supplies due to the guarantee of experiment effect and its low cost (about US$ 0.02/well). Therefore, CVBS (US$ 0.3/well) could be employed as disposable consumables according to some commercially disposable cell-based biosensors (about US$ 3.2/well). Moreover, CVBS was easy-to-use due to its good biocompatibility.
At the point-in-time, the LOD of OA detection was 33.9532 µg/L and the working range was 10 – 800 µg/L. Furthermore, the iPlate Monitor had the ability of data storage and transmission, which provided other people with a convenient way for further analysis via the internet. Therefore, the biosensing system using CVBS provided a low-cost, high-throughput,
wide-range and portable method to detect OA, which could satisfy the requirement of huge samples detection to OA in food safety and environmental monitoring fields.
4. Conclusions In this paper, the smartphone-based biosensing system using CVBS was proposed for the OA detection. This system presented the good robustness in long-term image capture and analysis. Then the system cooperating with CCK-8 kit achieved label-free, non-invasive and long-term monitoring of cell viability, and explored the construction of CVBS. CVBS chose HepG2 cell lines as sensing element to respond to the stimulation of OA. The homemade APP – iPlate Monitor used traversal algorithm to obtain the best detection point-in-time. At the point-in-time, CVBS had a wide working range (10 – 800 µg/L). As a consequence, the smartphone-based biosensing system using CVBS provided a low-cost, high-throughput, portable and efficient platform to detect marine shellfish toxin such as OA.
Acknowledgments This work was supported by National 973 Project of China (No. 2015CB352101), Public Welfare Project of Zhejiang Province (No. 2015C34010) and Natural Science Foundation of China (No. 31571004，61320106002) and National Marine Public Welfare Project of China (No. 201305010).
References [1] A. Sassolas, A. Hayat, G. Catanante, J.-L. Marty, Detection of the marine toxin okadaic acid: Assessing seafood safety, Talanta, 105(2013) 306-16. [2] T. Sassa, W.W. Richter, N. Uda, M. Suganuma, H. Suguri, S. Yoshizawa, et al., Apparent “activation” of protein kinases by okadaic acid class tumor promoters, Biochem Bioph Res Co, 159(1989) 939-44. [3] T. Aune, M. Yndestad, Diarrhetic shellfish poisoning, Academic Press: London, UK1993, pp. 87-104. [4] M. Campas, B. Prieto-Simón, J.-L. Marty, Biosensors to detect marine toxins: Assessing seafood safety, Talanta, 72(2007) 884-95. [5] K.D. Cusick, G.S. Sayler, An overview on the marine neurotoxin, saxitoxin: Genetics, molecular targets, methods of detection and ecological functions, Marine drugs, 11(2013) 991-1018. [6] J. Nicolas, P.J. Hendriksen, A. Gerssen, T.F. Bovee, I.M. Rietjens, Marine neurotoxins: State of the art, bottlenecks, and perspectives for mode of action based methods of detection in seafood, Molecular nutrition & food research, 58(2014) 87-100. [7] X. Li, Z. Li, J. Chen, Q. Shi, R. Zhang, S. Wang, et al., Detection, occurrence and monthly variations of typical lipophilic marine toxins associated with diarrhetic shellfish poisoning in the coastal seawater of Qingdao City, China, Chemosphere, 111(2014) 560-7. [8] A. Braña-Magdalena, J.M. Leão-Martins, T. Glauner, A. Gago-Martínez, Intralaboratory validation of a fast and sensitive UHPLC/MS/MS method with fast polarity switching for the analysis of lipophilic shellfish toxins, Journal of AOAC International, 97(2014) 285-92. [9] M. Campàs, J.-L. Marty, Enzyme sensor for the electrochemical detection of the marine toxin okadaic acid, Analytica chimica acta, 605(2007) 87-93. [10] T. Mouratidou, I. Kaniou-Grigoriadou, C. Samara, T. Kouimtzis, Detection of the marine toxin okadaic acid in mussels during a diarrhetic shellfish poisoning (DSP) episode in Thermaikos Gulf, Greece, using biological, chemical and immunological methods, Science of the total environment, 366(2006) 894-904. [11] X. Zhang, J. Fang, L. Zou, Y. Zou, L. Lang, F. Gao, et al., A novel sensitive cell-based Love Wave biosensor for marine toxin detection, Biosensors and Bioelectronics, 77(2016) 573-9. [12] L. Zou, Q. Wang, M. Tong, H. Li, J. Wang, N. Hu, et al., Detection of diarrhetic shellfish poisoning toxins using high-sensitivity human cancer cell-based impedance biosensor, Sensors and Actuators B: Chemical, 222(2016) 205-12. [13] S. BokáKim, S. JunáMoon, A cell-based biosensor for real-time detection of cardiotoxicity using lensfree imaging, Lab on a Chip, 11(2011) 1801-7. [14] M.P. Walzik, V. Vollmar, T. Lachnit, H. Dietz, S. Haug, H. Bachmann, et al., A portable low-cost long-term live-cell imaging platform for biomedical research and education, Biosensors and Bioelectronics, 64(2015) 639-49. [15] S.K. Vashist, P.B. Luppa, L.Y. Yeo, A. Ozcan, J.H. Luong, Emerging technologies for nextgeneration point-of-care testing, Trends in biotechnology, 33(2015) 692-705. [16] S.K. Vashist, O. Mudanyali, E.M. Schneider, R. Zengerle, A. Ozcan, Cellphone-based devices for bioanalytical sciences, Analytical and bioanalytical chemistry, 406(2014) 3263-77. [17] D.N. Breslauer, R.N. Maamari, N.A. Switz, W.A. Lam, D.A. Fletcher, Mobile phone based clinical microscopy for global health applications, Plos One, 4(2009) e6320. [18] D. Tseng, O. Mudanyali, C. Oztoprak, S.O. Isikman, I. Sencan, O. Yaglidere, et al., Lensfree microscopy on a cellphone, Lab on a Chip, 10(2010) 1787-92.
[19] H. Zhu, S.O. Isikman, O. Mudanyali, A. Greenbaum, A. Ozcan, Optical imaging techniques for point-of-care diagnostics, Lab on a Chip, 13(2013) 51-67. [20] H. Zhu, I. Sencan, J. Wong, S. Dimitrov, D. Tseng, K. Nagashima, et al., Cost-effective and rapid blood analysis on a cell-phone, Lab on a Chip, 13(2013) 1282-8. [21] W.-J. Yi, W. Jia, J. Saniie, Mobile sensor data collector using Android smartphone, Circuits and Systems (MWSCAS), 2012 IEEE 55th International Midwest Symposium on, IEEE2012, pp. 956-9. [22] K.B. Sneed, Integration of a Team Approach to Hypertension Treatment, Hypertension in High Risk African Americans, Springer2015, pp. 199-212. [23] O. Mudanyali, S. Dimitrov, U. Sikora, S. Padmanabhan, I. Navruz, A. Ozcan, Integrated rapiddiagnostic-test reader platform on a cellphone, Lab on a Chip, 12(2012) 2678-86. [24] D.J. You, T.S. Park, J.-Y. Yoon, Cell-phone-based measurement of TSH using Mie scatter optimized lateral flow assays, Biosensors and Bioelectronics, 40(2013) 180-5. [25] P. Preechaburana, M.C. Gonzalez, A. Suska, D. Filippini, Surface plasmon resonance chemical sensing on cell phones, Angewandte Chemie, 124(2012) 11753-6. [26] P.B. Lillehoj, M.-C. Huang, N. Truong, C.-M. Ho, Rapid electrochemical detection on a mobile phone, Lab on a Chip, 13(2013) 2950-5. [27] D. Zhang, J. Jiang, J. Chen, Q. Zhang, Y. Lu, Y. Yao, et al., Smartphone-based portable biosensing system using impedance measurement with printed electrodes for 2, 4, 6-trinitrotoluene (TNT) detection, Biosensors and Bioelectronics, 70(2015) 81-8. [28] A.F. Coskun, R. Nagi, K. Sadeghi, S. Phillips, A. Ozcan, Albumin testing in urine using a smartphone, Lab on a Chip, 13(2013) 4231-8. [29] C.M. McGeough, S. O'Driscoll, Camera phone-based quantitative analysis of C-reactive protein ELISA, Biomedical Circuits and Systems, IEEE Transactions on, 7(2013) 655-9. [30] S.-Y. Lu, C. Lin, Y.-S. Li, Y. Zhou, X.-M. Meng, S.-Y. Yu, et al., A screening lateral flow immunochromatographic assay for on-site detection of okadaic acid in shellfish products, Analytical biochemistry, 422(2012) 59-65. [31] A. Muaremi, B. Arnrich, G. Tröster, Towards Measuring Stress with Smartphones and Wearable Devices During Workday and Sleep, BioNanoScience, 3(2013) 172-83. [32] P. Preechaburana, A. Suska, D. Filippini, Biosensing with cell phones, Trends in biotechnology, (2014). [33] J. Jiang, X. Wang, R. Chao, Y. Ren, C. Hu, Z. Xu, et al., Smartphone based portable bacteria preconcentrating microfluidic sensor and impedance sensing system, Sensors and Actuators B: Chemical, 193(2014) 653-9. [34] Y. Chen, Y. Zilberman, P. Mostafalu, S.R. Sonkusale, Paper based platform for colorimetric sensing of dissolved NH3 and CO2, Biosensors and Bioelectronics, (2014). [35] M. Zangheri, L. Cevenini, L. Anfossi, C. Baggiani, P. Simoni, F. Di Nardo, et al., A simple and compact smartphone accessory for quantitative chemiluminescence-based lateral flow immunoassay for salivary cortisol detection, Biosensors and Bioelectronics, 64(2015) 63-8. [36] B. Berg, B. Cortazar, D. Tseng, H. Ozkan, S. Feng, Q. Wei, et al., Cellphone-based hand-held microplate reader for point-of-care testing of enzyme-linked immunosorbent assays, ACS nano, 9(2015) 7857-66. [37] H. Tominaga, M. Ishiyama, F. Ohseto, K. Sasamoto, T. Hamamoto, K. Suzuki, et al., A water-soluble tetrazolium salt useful for colorimetric cell viability assay, Analytical Communications, 36(1999) 47-50. [38] K. Su, Q. Zou, J. Zhou, L. Zou, H. Li, T. Wang, et al., High-sensitive and high-efficient biochemical
analysis method using a bionic electronic eye in combination with a smartphone-based colorimetric reader system, Sensors and Actuators B: Chemical, 216(2015) 134-40. [39] A. Ledreux, A.-L. Sérandour, B. Morin, S. Derick, R. Lanceleur, S. Hamlaoui, et al., Collaborative study for the detection of toxic compounds in shellfish extracts using cell-based assays. Part II: Application to shellfish extracts spiked with lipophilic marine toxins, Analytical and bioanalytical chemistry, 403(2012) 1995-2007. [40] T. Yasumoto, Y. Oshima, M. Yamaguchi, Occurrence of a new type of shellfish poisoning in the Tohoku district, Bulletin of the Japanese Society of Scientific Fisheries, 44(1978) 1249-55. [41] M. Wiese, P.M. D’agostino, T.K. Mihali, M.C. Moffitt, B.A. Neilan, Neurotoxic alkaloids: saxitoxin and its analogs, Marine drugs, 8(2010) 2185-211.
Biographies Kaiqi Su received his B.E. degree and M.S. degree of biomedical engineering in Zhejiang University, PR China in 2011 and 2014. Now he is a Ph.D. candidate of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, cell-based biosensor, cell metabolic physiology, biosensor instrument establishing and signal processing.
Yuxiang Pan received his B.E. degree of biomedical engineering in Sichuan University, PR China in 2015. Now he is a Ph.D. candidate of biomedical engineering of Zhejiang University. His work includes research of flow cytometry, cell-based biosensor and cell metabolic physiology.
Zijian Wan received his B.E. degree of biomedical engineering in Southeast University, PR China in 2015. Now he is a master student of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, biosensor instrument establishing and signal processing.
Jielong Zhong received his B.E. degree of biomedical engineering in Zhejiang University, PR China in 2016. Now he is a master student of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, biosensor instrument establishing and signal processing.
Jiaru Fang received her BS degree of biomedical engineering, PR China in 2014 in Jinan University. Now she is a master student of biomedical engineering of Zhejiang University.
Her work includes research of electrochemical sensors, cell-based biosensor and cell metabolic physiology.
Quchao Zou received his BS degree of biomedical engineering, PR China in 2013 in University of Shanghai for Science and Technology. Now he is a master student of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, cell-based biosensor, cell metabolic physiology, biosensor instrument establishing and signal processing.
Hongbo Li received his B.E. degree of biomedical engineering in Zhejiang University, PR China in 2013. Now he is a Ph.D. candidate of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, cell-based biosensor, cell metabolic physiology, biosensor instrument establishing and signal processing.
Ping Wang received his BS degree, MS degree and PhD degree of electrical engineering in Harbin Institute of Technology, Harbin, PR China in 1984, 1987 and 1992, respectively. He is currently a professor of Biosensors National Special Lab, Department of Biomedical Engineering of Zhejiang University. His research interests include biomedical sensors, electrochemical sensors and measurement technique.
Illustration of the figures Figure 1 (A) The action principle of CCK-8 kit. The CCK-8 kit contains WST-8 and 1Methoxy PMS. The dehydrogenases catalyze the conversion from NAD+ to NADH in cells while the inverse conversion from NADH to NAD+ is accompanied by the conversion from 1Methoxy PMS to 1-Methoxy PMS reduced form. In the extracellular microenvironment, WST-8 reacts with 1-Methoxy PMS reduced form to generate the WST-8 formazan (orange water-soluble product) and 1-Methoxy PMS. Thus the color depth can be detected to reflect the living cell count or cell status. (B) Construction of CVBS. (C) The picture of portable smartphone-based system and (D) the main interface of the homemade software iPlate Monitor.
Figure 2 Image monitoring test of the portable smartphone-based system. The system monitored the 96 wells of blank MTP for about 20 h. And the captured image sequence was analyzed by (A) red channel, (B) green channel and (C) blue channel. There were 96 curves in every subgraph.
Figure 3 Real-time monitoring of various cell seeding densities. (A) The real-time CVI curves of HepG2 cells were measured by the system with different cell densities for about 5 h. (B) Rsquared value and sensitivity variations over time and (C) the standard curve of best detection point-in-time. The R-squared (0.9956) and slope (169.4) both reached the maximum at 1.07 h during the monitoring process. The CVI was calculated once each 2 minutes and the image was saved once every 4 minutes. For the clarity of figure, the data in (A) were resampled and
the time interval was 6 min. All the data in (A) and (C) were performed in 3 repeated test and the error bars represented the standard deviation (SD).
Figure 4 The microscopic images of HepG2 cells at best detection point-in-time with different initial cell seeding densities. The cellular morphology of the group with CCK-8 addition was consistent with that of the group without CCK-8 addition. The magnification scales of all cell images were 40 and 100. Figure 5 (A) Real-time NCVI curves of the test groups (10 – 800 µg/L OA) and control group (fresh medium) during the detection process. (B) R2 and sensitivity variations over time and (C) the standard curves of the highest R2 and the highest sensitivity. The R-squared (0.9536) reached the maximum while the slope was 2.569 at 2.85 h. All the data in (A) and (C) were performed in 3 repeated test and the error bars represented the SD.
Figure 6 The microscopic images of HepG2 cells after treatment with different concentrations of OA at the 2.85 h. The magnification scales of all cell images were 40 and 100.
Figure 7 Real-time monitoring of the response to control, 800 µg/L GTX, 400 µg/L GTX, 800 µg/L PbTx-2 and 400 µg/L PbTx-2 by HepG2-CVBS (A). The NCVIs of HepG2-CVBS after 2.85 h treatments of three toxins (B). All the data were performed in 3 repeated test and the error bars represented the SD. Levels of significance: ***p < 0.001; ****p < 0.0001.
Fig 1
Fig 2
Fig 3
Fig 4
Fig 5
Fig 6
Fig 7
Table 1 The results of detecting marine toxins in non-toxic mussel extracts spiked with standard solution by CVBS and MBA CVBS
MBA
Added conc. Determined conc.
Percentage
Determined conc.
Percentage
(µg L-1)
recovery
(µg L-1)
recovery
300
298.05±27.42
99.35
295.23±54.32
98.41
400
386.36±37.47
96.59
382.68±75.01
95.67
500
520.15±44.21
104.03
536.60±54.73
107.32
600
561.24±34.79
93.54
563.52±71.56
93.92
700
712.46±37.76
101.78
738.36±57.59
105.48
(µg L-1)
S0925-4005(17)30636-6 http://dx.doi.org/doi:10.1016/j.snb.2017.04.036 SNB 22122
To appear in:
Sensors and Actuators B
Received date: Revised date: Accepted date:
31-8-2016 11-3-2017 7-4-2017
Please cite this article as: Kaiqi Su, Yuxiang Pan, Zijian Wan, Longjie Zhong, Jiaru Fang, Quchao Zou, Hongbo Li, Ping Wang, Smartphone-based portable biosensing system using cell viability biosensor for okadaic acid detection, Sensors and Actuators B: Chemicalhttp://dx.doi.org/10.1016/j.snb.2017.04.036 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.
Smartphone-based portable biosensing system using cell viability biosensor for okadaic acid detection Kaiqi Su1, 2, Yuxiang Pan1, Zijian Wan1, Longjie Zhong1, Jiaru Fang1, Quchao Zou1, 3, Hongbo Li1, Ping Wang1, 2, *
1. Biosensor National Special Laboratory, Key Laboratory for Biomedical Engineering of Education Ministry,
Department of Biomedical Engineering, Zhejiang University, Hangzhou310027, PR China
2. State Key Laboratory of Transducer Technology, Chinese Academy of Sciences, Shanghai 200050, China
3. Department of clinical engineering, the Second Affiliated Hospital of Zhejiang University School of Medicine,
Hangzhou 310009, Zhejiang Province, PR China
* Corresponding author: Tel.: +86 571 87952832; Fax: +86 571 87952832. E-mail address: [email protected] (Ping Wang)
Smartphone-based portable biosensing system using cell viability biosensor for okadaic acid detection
Highlights
A smartphone-based biosensing system using CVBS was developed for OA detection.
The system presented the good robustness in long-term image capture and analysis.
The system achieved label-free, non-invasive and long-term monitoring of cell viability.
The cost-effective CVBS was constructed by HepG2 cells, MTP and CCK-8 kit for OA detection.
The iPlate Monitor used traversal algorithm to obtain the best detection point-in-time.
Abstract
Okadaic acid (OA), as a diarrheic shellfish poisoning toxin, had wide distribution and frequent occurrence. Therefore, low-cost, high-throughput, wide-range and portable detection of OA was in high demand for food safety and environmental monitoring. In this study, a novel and portable smartphone-based system using cell viability biosensor (CVBS) was developed for label-free, non-invasive and long-term monitoring of cell viability. The variation of cell viability reflected the changes of cell morphology, cell count and cell proliferation indirectly. And this system applied the combination of image analysis and cell counting kit-8 assay (CCK-8) to monitor the reflection. The biosensing system chose HepG2 cells as sensing elements to build CVBS and used it in OA detection. Results showed this system could synchronously detect OA in 96 channels. And this biosensor presented a good performance to various OA concentrations, with a wide linear detection range (10 - 800 μg/L). Moreover, the point-in-time having best detection performance could be located by the traversal algorithm in the monitoring duration. Thus, this cell-based biosensor system provided a convenient and efficient approach in seafood safety testing such as OA screening.
Keyword: Cell viability biosensor (CVBS); Smartphone; Portable colorimetric reader; Shellfish toxins; Okadaic acid (OA)
1. Introduction OA, as a kind of diarrheic shellfish poisoning toxin, is produced by some unicellular algae from plankton and benthic microalgae and accumulates in the digestive glands of shellfish [1]. As one inhibitors of serine/threonine protein phosphatases type 1 (PP1) and 2A
(PP2A) [2], OA can cause some diarrheic symptoms including diarrhea, nausea, vomiting and abdominal pain [3]. Moreover, OA has been identified as a tumor promoter and has been proved that it had mutagenic and immunotoxin effects [4]. Currently, there are some biochemical methods including mouse bioassay (MBA) [5], pre-column oxidation liquid chromatography with fluorescence detection [6], high performance liquid chromatographymass spectrometry [7, 8], enzyme linked immunosorbent assay [9, 10] and cell-based biosensors [11, 12] for the detection of marine shellfish toxins. Thereinto, the research using cell-based biosensors in shellfish toxin detection is beginning to attract the people’s attention due to their abilities of observing the impact of toxin to cells and the quantitative analysis to toxin. However, shellfish toxins such as OA have wide distribution and frequent occurrence. Hence, the field of shellfish toxin detection requires a low-cost, high-throughput, wide-range and portable biosensing system for generalization, while the above cell-based biosensors can hardly meet these requirements.
How to establish a cell-based biosensing system, which can achieve low-cost, highthroughput, wide-range and portable detection? It is a great choice to combine image analysis with smartphone-based application. For biomedical research, the image analysis is utilized in the construction of low-cost cell-based biosensor [13, 14]. On the other hand, many biochemical analysis platforms or methods combined with smartphone appear one after another in point-of-care test domain due to the high integration of smartphone to processor and diverse sensors [15, 16]. These smartphone-based applications distribute in different fields including smartphone-based microscopy [17, 18], fluorescent imaging [19], imaging cytometry [20], electrocardiography [21, 22], lateral flow assays [23, 24], surface plasmon
resonance-based sensing [25], electrochemical sensing [26, 27], immunoassays [28-30], and other applications [31-33]. Also, the accessories of smartphone-based system appear one after another to improve detection performance [34-36].
In this study, a novel and portable smartphone-based system called CVBS system was developed in accordance with the combination of image analysis and smartphone-based application. Cooperating with CCK-8 kit that had no cytotoxicity, this biosensing system could achieve label-free, non-invasive and long-term monitoring of cell viability. Besides, HepG2 cell lines were chosen as sensing elements to specifically detect OA. This smartphone-based system presented the features of low cost, high throughput, wide detection range and portability in OA detection. Also, the cell culture process used in this system was more common than that of other cell-based biosensors, which could promote the generalization of this system in the field of shellfish toxin measurement. All the details will be discussed in the following sections.
2. Materials and methods 2.1 Reagent and Setup
The 96-well polystyrene plates are bought from Thermo Fisher Scientific, Germany. HepG2 cell line is bought from American Type Culture Collection and all of the reagents of cell culture are purchased from Gibico, USA. DMEM medium (Dulbecco’s modified eagle medium) with high glucose, fetal calf serum (FBS), 0.25% Trypsin–EDTA, and phosphate buffered saline (PBS) are obtained from Gibico, USA. Gonyautoxin2&3 (GTX2&3) and brevetoxin-2 (PbTx-2) are purchased from National Research Council (Canada). OA (Sigma,
USA) stock solution is prepared in DMSO and filtered with 0.22 μm membrane filter unit (Millipore, USA). OA is diluted by DMEM medium with high glucose when it is used to treat with cells. CCK-8 kit is obtained from Shanghai 7 sea biotech Co., LTD., China. The morphological changes of cells are observed by a stereo microscope (Sharp Inc., Japan).
2.2 Action principle of CCK-8 kit
For label-free, non-invasive and long-term monitoring of cell viability, CCK-8 kit is chosen as the developer of system. Comparing with other detection kits of cell proliferation and cytotoxicity such as MTT, XTT and WST-1, CCK-8 has excellent stability, a wider detection range, higher sensitivity and is easy to use. More importantly, it has no cytotoxicity. Figure 1A shows the action principle of CCK-8 kit. CCK-8 kit is based on water-soluble tetrazolium salt-8 (WST-8) [37]. And WST-8 can be reduced by dehydrogenases in cells to give a water-soluble formazan dye (WST-8 formazan) in the presence of electron mediator (1Methoxy PMS). The WST-8 formazan is an orange colored product and reflects the living cell status. Hence, this system monitors the cell viability by the variation of orange intensity over time.
2.3 The portable smartphone-based biosensing system
This system consists of CVBS, illumination provider and smartphone. The CVBS includes living cells, microtiter plate (MTP) and CCK-8 kit (Figure 1B). The basic structure design of illumination provider has been described elsewhere [38] and the contour structure has been upgraded according to industrial design in this study (Figure 1C). The smartphone installed with homemade iOS APP - iPlate Monitor (designed by Swift 1.0 and Object-C in
Xcode 7) undertakes the image acquisition, image analysis, data storage and transmission (Figure 1D). The iPlate Monitor introduces cell viability index (CVI) and normalized cell viability index (NCVI) to evaluate cell viability monitoring and analyze the response of CVBS. CVI and NCVI are calculated by the equ1 and equ2, respectively. CVI = BBlank − BTest equ1
Where CVI, BBlank and BTest stand for cell viability index, blue channel value of blank MTP and blue channel value of CVBS, respectively. CVI
NCVI = CVI t
0
equ2
Where CVIt and CVI0 are the CVI at any point-in-time and the CVI at 0 h, respectively. To confirm the best detection point-in-time, iPlate Monitor applies the traversal method to perform linear fitting at all time-points of monitoring process. Afterwards, the point-in-time of fitting curve having best performance, which is the best detection point-in-time, is determined by the fitting goodness and sensitivity. The supplementary material (Figure S1) shows the workflow of this APP and some actual smartphone screen picture during OA detection experiments.
Fig. 1
2.4 Cell culture
HepG2 cell line is cultured in DMEM medium which includes 10% heat inactivated FBS and 0.5% antibiotic solution (10 mg/mL streptomycin and 1000 U/mL penicillin). Then it is incubated at 37 °C in humidified air with 5% CO2 in an incubator (Thermo, USA). When the confluent cells reach 80%, 0.25%, Trypsin-EDTA is used to dislodge cells from the flask to 96-well plates.
2.5 Shellfish extracts preparation
Mytilus edulis are selected for actual sample testing and the non-toxin samples are purchased from the market. The preparation of shellfish extracts are carried out according to the protocol offered by Ledreux et al. [39].
2.6 Experiment setup
2.6.1 Cell viability monitoring of smartphone-based system HepG2 cells are detached from the culture flask and 100 μL cell suspensions with different cell seeding densities are prepared. Then HepG2 are inoculated in the 96-well MTP at different seeding densities. After cell inoculation, HepG2 are cultured for 4 h to adhere. Then 10 μL of CCK-8 solution is added to each well of the MTP and the MTP with smartphone-based system are placed inside a humidified incubator (at 37 °C, 5% CO2). Meanwhile, the detection system starts to monitor the cell viability curves. 2.6.2 Real-time monitoring of CVBS’s response to OA 100 μL of HepG2 cell suspensions with the same cell seeding density are added onto the MTP to build HepG2-CVBS. Before exposure to toxins, the cells are cultured for 24 hours to
achieve cell adherence and cell status stability. Then the medium of treated groups is removed and replaced by 100 μL fresh medium and 10 μL different concentrations of OA. Subsequently, 10 μL of CCK-8 solution and the CVBS with the smartphone-based system are placed in the incubator (at 37 °C, 5% CO2). Meanwhile, the detection system starts to monitor the responses of HepG2-CVBS to OA.
3. Results and discussion 3.1 Stability of image monitoring using the portable smartphone-based system
Stability was the core of real-time monitoring and pixel intensity was the initial output of system. The system captured image and calculated pixel intensity once each 2 minutes. As shown in Figure 2, the performance of system was respectively tested with RGB color model. The monitoring time was about 20 h. The system showed stable pixel intensity measurement, although having slight fluctuations in monitoring process. Table S1, Table S2 and Table S3 further illustrated the detecting stability of system. For all 96 wells of MTP, the coefficients of variation (CV) of red, green and blue channel were within 1.08%, 0.66% and 0.71%, respectively. The detecting differences of pixel intensity measurements were often less than 2.71% (red < 2.71%, green < 1.79% and blue < 1.99%).
Fig. 2
3.2 Dynamically monitoring of cell viability and CVBS construction
HepG2 cells were inoculated onto the MTP at 0, 5000, 10,000, 20,000, 30,000, 40,000, 50,000 and 60,000 cells/well. CVI curves were recorded for about 5 h after CCK-8 addition
(Figure 3A) and the monitoring process could be reappeared in the Video S1. The CVI had positive correlation with the orange depth. And the growth speed of CVI was positively associated with the cell density. The range of CVI was from 0 to 255 according to the range of pixel intensity. Hence, the CVI remained stable after it reached the maximum.
Figure 3B shows the variations of R2 and sensitivity of linear fitting curve over time. Commonly, two time points were picked out, which were the highest R2 and the highest sensitivity, respectively. In this experiment (Figure 3C), the best point-in-time was determined at 1.07 h (R2 was 0.9956 and slope was 169.4), which had both the highest R2 and sensitivity. The limit of detection (LOD) was calculated by 3δ/slope. At the best point-intime, the LOD of cell density detection was 3684 cells/well.
Fig. 3
To verify that CCK-8 kit had no cytotoxicity, the cellular morphology of two experimental groups with and without CCK-8 kit was observed by the stereo microscope. Figure 4 shows the cell growth situation under the views of 40 and 100 magnifications at the best detection point-in-time above. It was apparent that the number of cells was positively correlated with the initial seeding density. And the cells revealed their great growth situation and attachment status whether if CCK-8 was added or not. The high density cells were more resistant to toxin treatment. Hence, 5,000 cells/well was the suitable seeding density that CVBS could obtain relatively high sensitivity.
Fig. 4
3.3 Real-time response of CVBS to OA detection
Before adding marine toxins, HepG2 cells (5,000 cells/well) were seeded to build HepG2-CVBS. And different concentrations of OA (10, 25, 50, 100, 200, 400 and 800 µg/L) were tested. Figure 5A shows the real-time NCVI curves, which reflects the cell activity changes under OA treatment. The increase rate of NCVI was negatively linked with OA concentration. And the higher concentration of OA, the faster NCVI curve reached the plateau.
Figure 5B shows the variations of R2 and sensitivity of linear fitting curve over time. Two time points (2.85 h and 10.55 h) were picked out, which were the highest R2 and the highest sensitivity, respectively. Since the R2 of standard curve at 10.55 h was too low (R2 was 0.7772), 2.85 h was determined as the best detection point-in-time. Figure 5C shows the standard curve at 2.85 h (R2 was 0.9536 and slope was -2.569). At the best detection point-intime, the equation of standard curve was y = -2.569x + 13.8 (R2=0.9536), where y was NCVI, and x was logarithmic concentration of OA. The LOD was calculated by 3δ/slope, namely, the triple standard deviation (STD) was used to work out the LOD (33.9532 µg/L). Since the maximum concentration of OA in shellfish was permitted as 160 µg/kg by European Commission (EC) (EC 853/2004 15), the proposed method could reach this level. Moreover, MBA was used as the reference method for OA analysis proposed by AOAC and the detection limit was 0.05 MU/g equaling to 220 µg/kg. And about 12MU (52.8µg) was the
lowest amount of OA inducing mild form of illness (e.g., nausea, diarrhea and vomit) to an adult [40]. Suppose a maximum intake of mussels was 250 g, the lowest concentration of OA to start toxic symptoms was 211.2 µg/kg. This level could be detected by this biosensing system.
Fig. 5
To verify the response of CVBS to OA, the OA-induced morphological changes in HepG2 cells was observed by the stereo microscope at the best detection point-in-time. Figure 6 shows the cell growth situation under the views of 40 and 100 magnifications after continuous treatment with different concentrations of OA. It was apparent that the HepG2 cells had distinct morphological changes after the OA treatment. Meanwhile, the depth of morphological change was related to the OA concentration and treatment duration.
Fig. 6
For the real sample detection and the comparison between MBA and CVBS, nontoxic shellfish extracts were spiked with the concentrations of OA (300 – 700 µg/L). The concentration setup was in accordance with the detection limit (220 µg/kg) of MBA for OA analysis proposed by AOAC. As shown in Table 1, the results of CVBS were highly correlated to those of MBA. The percentage recoveries measured by CVBS and MBA were in the ranges of 93.54 – 104.03 and 93.92 – 107.32 for OA-spiked shellfish extracts,
respectively. And the corresponding average percentage recoveries were 99.06 and 100.16, respectively.
Tab. 1
3.4 Specificity of CVBS
The specificity of HepG2-CVBS to OA could be deduced by the previous study which adopted similar cell-based biosensor for marine toxin detection. OA could strongly influence the cell morphology, cell attachment and cell viability, which was the methodology basis of some cell-based biosensors, such as cell-based impedance biosensor (CIB), Love Wave biosensor and CVBS, to detect OA. The cell-based Love Wave biosensor using HepG2 cells compared the detection results in saxitoxin (STX) and brevetoxin 3 (PbTx-3). Then it was certified that this cell-based Love Wave biosensor presented a good specificity for OA detection. Similarly, the high selectivity of CIB using HepG2 cells was proved by the comparison results in STX and brevetoxin 2 (PbTx-2). Since the mechanism action of these marine neural toxins was different from that of OA, the above cell-based biosensors had a good specificity for OA detection.
Fig. 7
The specificity of HepG2-CVBS was evaluated by gonyautoxin2&3 (GTX2&3) and PbTx-2. GTX2&3 was one of the paralytic shellfish poisoning (PSP) toxins, and it could affect the propagation of the action potential by blocking sodium channels, which would
change the membrane potential on a cellular level and cause neurological symptoms in human [41]. PbTx-2 was a kind of neurologic shellfish poisoning (NSP) toxins which induced the enhancement of sodium inflow into cells [4]. The concentrations of GTX2&3 and PbTx-2 applied in this specificity assay were 400 μg/L and 800 μg/L. From the results shown in Figure 7A, it was obvious that the NCVI variations caused by the two toxins were not much different from those of control groups. According to the above best detection point-in-time (2.85 h), the responses of CVBS to three toxins were compared (Figure 7B). The NCVI caused by OA was lower than those which were induced by GTX2&3 and PbTx-2. And the OA experiment group had significant difference from the control group (levels of significance: ***p < 0.001; ****p < 0.0001) while the GTX2&3 and PbTx-2 groups didn’t have. Therefore, HepG2-CVBS had a good specificity for OA detection.
3.5 Performance of smartphone-based biosensing system using CVBS for OA detection
Living cell status is useful to monitor the changes of chemicals in microenvironment. In this study, cell viability was used as sensor response in OA detection. The biosensing system utilized the optical image analysis for a label-free, non-invasive and long-term monitoring of cell viability and constructed CVBS to detect OA.
In the previous report [38], bionic electronic eye presented inconvenience and inaccuracy in the kinetics assay due to the lack of automation. For instance, since cells needed stable cultivation environment, the kinetics assay of CCK-8 employed multiple MTPs to compare the performance of microtiter plate reader and bionic electronic eye at different point-intimes. Therefore, the system in this work was developed to a longtime and real-time
monitoring system by the homemade software – iPlate Monitor. The iPlate Monitor achieved longtime and real-time image acquisition and analysis, which met the requirements of monitoring to CVBS having living cells in the incubator. Moreover, the iPlate Monitor used traversal algorithm to obtain the best detection point-in-time, which could enhance the detection accuracy.
CVBS consisted of living cells, MTP and CCK-8 kit, which was a simple construction comparing with some cell-based biosensors such as CIB and Love Wave biosensor. CCK-8 could be reduced by dehydrogenases in cells to give an orange water-soluble formazan dye. The orange intensity reflected the viability of cells. Thus CVBS used CCK-8 to transform the cell viability to optical signal and this signal could be detected by the portable biosensing system. The durability of CVBS related to its structure. Thereinto, living cells and CCK-8 kit belonged to disposable supplies. MTP could be reused about three times for adherent cell cultivation after a series of operations including abstersion, overnight soak with acid, distilled water wash and ultraviolet radiation. However, MTP was also regarded as disposable supplies due to the guarantee of experiment effect and its low cost (about US$ 0.02/well). Therefore, CVBS (US$ 0.3/well) could be employed as disposable consumables according to some commercially disposable cell-based biosensors (about US$ 3.2/well). Moreover, CVBS was easy-to-use due to its good biocompatibility.
At the point-in-time, the LOD of OA detection was 33.9532 µg/L and the working range was 10 – 800 µg/L. Furthermore, the iPlate Monitor had the ability of data storage and transmission, which provided other people with a convenient way for further analysis via the internet. Therefore, the biosensing system using CVBS provided a low-cost, high-throughput,
wide-range and portable method to detect OA, which could satisfy the requirement of huge samples detection to OA in food safety and environmental monitoring fields.
4. Conclusions In this paper, the smartphone-based biosensing system using CVBS was proposed for the OA detection. This system presented the good robustness in long-term image capture and analysis. Then the system cooperating with CCK-8 kit achieved label-free, non-invasive and long-term monitoring of cell viability, and explored the construction of CVBS. CVBS chose HepG2 cell lines as sensing element to respond to the stimulation of OA. The homemade APP – iPlate Monitor used traversal algorithm to obtain the best detection point-in-time. At the point-in-time, CVBS had a wide working range (10 – 800 µg/L). As a consequence, the smartphone-based biosensing system using CVBS provided a low-cost, high-throughput, portable and efficient platform to detect marine shellfish toxin such as OA.
Acknowledgments This work was supported by National 973 Project of China (No. 2015CB352101), Public Welfare Project of Zhejiang Province (No. 2015C34010) and Natural Science Foundation of China (No. 31571004，61320106002) and National Marine Public Welfare Project of China (No. 201305010).
References [1] A. Sassolas, A. Hayat, G. Catanante, J.-L. Marty, Detection of the marine toxin okadaic acid: Assessing seafood safety, Talanta, 105(2013) 306-16. [2] T. Sassa, W.W. Richter, N. Uda, M. Suganuma, H. Suguri, S. Yoshizawa, et al., Apparent “activation” of protein kinases by okadaic acid class tumor promoters, Biochem Bioph Res Co, 159(1989) 939-44. [3] T. Aune, M. Yndestad, Diarrhetic shellfish poisoning, Academic Press: London, UK1993, pp. 87-104. [4] M. Campas, B. Prieto-Simón, J.-L. Marty, Biosensors to detect marine toxins: Assessing seafood safety, Talanta, 72(2007) 884-95. [5] K.D. Cusick, G.S. Sayler, An overview on the marine neurotoxin, saxitoxin: Genetics, molecular targets, methods of detection and ecological functions, Marine drugs, 11(2013) 991-1018. [6] J. Nicolas, P.J. Hendriksen, A. Gerssen, T.F. Bovee, I.M. Rietjens, Marine neurotoxins: State of the art, bottlenecks, and perspectives for mode of action based methods of detection in seafood, Molecular nutrition & food research, 58(2014) 87-100. [7] X. Li, Z. Li, J. Chen, Q. Shi, R. Zhang, S. Wang, et al., Detection, occurrence and monthly variations of typical lipophilic marine toxins associated with diarrhetic shellfish poisoning in the coastal seawater of Qingdao City, China, Chemosphere, 111(2014) 560-7. [8] A. Braña-Magdalena, J.M. Leão-Martins, T. Glauner, A. Gago-Martínez, Intralaboratory validation of a fast and sensitive UHPLC/MS/MS method with fast polarity switching for the analysis of lipophilic shellfish toxins, Journal of AOAC International, 97(2014) 285-92. [9] M. Campàs, J.-L. Marty, Enzyme sensor for the electrochemical detection of the marine toxin okadaic acid, Analytica chimica acta, 605(2007) 87-93. [10] T. Mouratidou, I. Kaniou-Grigoriadou, C. Samara, T. Kouimtzis, Detection of the marine toxin okadaic acid in mussels during a diarrhetic shellfish poisoning (DSP) episode in Thermaikos Gulf, Greece, using biological, chemical and immunological methods, Science of the total environment, 366(2006) 894-904. [11] X. Zhang, J. Fang, L. Zou, Y. Zou, L. Lang, F. Gao, et al., A novel sensitive cell-based Love Wave biosensor for marine toxin detection, Biosensors and Bioelectronics, 77(2016) 573-9. [12] L. Zou, Q. Wang, M. Tong, H. Li, J. Wang, N. Hu, et al., Detection of diarrhetic shellfish poisoning toxins using high-sensitivity human cancer cell-based impedance biosensor, Sensors and Actuators B: Chemical, 222(2016) 205-12. [13] S. BokáKim, S. JunáMoon, A cell-based biosensor for real-time detection of cardiotoxicity using lensfree imaging, Lab on a Chip, 11(2011) 1801-7. [14] M.P. Walzik, V. Vollmar, T. Lachnit, H. Dietz, S. Haug, H. Bachmann, et al., A portable low-cost long-term live-cell imaging platform for biomedical research and education, Biosensors and Bioelectronics, 64(2015) 639-49. [15] S.K. Vashist, P.B. Luppa, L.Y. Yeo, A. Ozcan, J.H. Luong, Emerging technologies for nextgeneration point-of-care testing, Trends in biotechnology, 33(2015) 692-705. [16] S.K. Vashist, O. Mudanyali, E.M. Schneider, R. Zengerle, A. Ozcan, Cellphone-based devices for bioanalytical sciences, Analytical and bioanalytical chemistry, 406(2014) 3263-77. [17] D.N. Breslauer, R.N. Maamari, N.A. Switz, W.A. Lam, D.A. Fletcher, Mobile phone based clinical microscopy for global health applications, Plos One, 4(2009) e6320. [18] D. Tseng, O. Mudanyali, C. Oztoprak, S.O. Isikman, I. Sencan, O. Yaglidere, et al., Lensfree microscopy on a cellphone, Lab on a Chip, 10(2010) 1787-92.
[19] H. Zhu, S.O. Isikman, O. Mudanyali, A. Greenbaum, A. Ozcan, Optical imaging techniques for point-of-care diagnostics, Lab on a Chip, 13(2013) 51-67. [20] H. Zhu, I. Sencan, J. Wong, S. Dimitrov, D. Tseng, K. Nagashima, et al., Cost-effective and rapid blood analysis on a cell-phone, Lab on a Chip, 13(2013) 1282-8. [21] W.-J. Yi, W. Jia, J. Saniie, Mobile sensor data collector using Android smartphone, Circuits and Systems (MWSCAS), 2012 IEEE 55th International Midwest Symposium on, IEEE2012, pp. 956-9. [22] K.B. Sneed, Integration of a Team Approach to Hypertension Treatment, Hypertension in High Risk African Americans, Springer2015, pp. 199-212. [23] O. Mudanyali, S. Dimitrov, U. Sikora, S. Padmanabhan, I. Navruz, A. Ozcan, Integrated rapiddiagnostic-test reader platform on a cellphone, Lab on a Chip, 12(2012) 2678-86. [24] D.J. You, T.S. Park, J.-Y. Yoon, Cell-phone-based measurement of TSH using Mie scatter optimized lateral flow assays, Biosensors and Bioelectronics, 40(2013) 180-5. [25] P. Preechaburana, M.C. Gonzalez, A. Suska, D. Filippini, Surface plasmon resonance chemical sensing on cell phones, Angewandte Chemie, 124(2012) 11753-6. [26] P.B. Lillehoj, M.-C. Huang, N. Truong, C.-M. Ho, Rapid electrochemical detection on a mobile phone, Lab on a Chip, 13(2013) 2950-5. [27] D. Zhang, J. Jiang, J. Chen, Q. Zhang, Y. Lu, Y. Yao, et al., Smartphone-based portable biosensing system using impedance measurement with printed electrodes for 2, 4, 6-trinitrotoluene (TNT) detection, Biosensors and Bioelectronics, 70(2015) 81-8. [28] A.F. Coskun, R. Nagi, K. Sadeghi, S. Phillips, A. Ozcan, Albumin testing in urine using a smartphone, Lab on a Chip, 13(2013) 4231-8. [29] C.M. McGeough, S. O'Driscoll, Camera phone-based quantitative analysis of C-reactive protein ELISA, Biomedical Circuits and Systems, IEEE Transactions on, 7(2013) 655-9. [30] S.-Y. Lu, C. Lin, Y.-S. Li, Y. Zhou, X.-M. Meng, S.-Y. Yu, et al., A screening lateral flow immunochromatographic assay for on-site detection of okadaic acid in shellfish products, Analytical biochemistry, 422(2012) 59-65. [31] A. Muaremi, B. Arnrich, G. Tröster, Towards Measuring Stress with Smartphones and Wearable Devices During Workday and Sleep, BioNanoScience, 3(2013) 172-83. [32] P. Preechaburana, A. Suska, D. Filippini, Biosensing with cell phones, Trends in biotechnology, (2014). [33] J. Jiang, X. Wang, R. Chao, Y. Ren, C. Hu, Z. Xu, et al., Smartphone based portable bacteria preconcentrating microfluidic sensor and impedance sensing system, Sensors and Actuators B: Chemical, 193(2014) 653-9. [34] Y. Chen, Y. Zilberman, P. Mostafalu, S.R. Sonkusale, Paper based platform for colorimetric sensing of dissolved NH3 and CO2, Biosensors and Bioelectronics, (2014). [35] M. Zangheri, L. Cevenini, L. Anfossi, C. Baggiani, P. Simoni, F. Di Nardo, et al., A simple and compact smartphone accessory for quantitative chemiluminescence-based lateral flow immunoassay for salivary cortisol detection, Biosensors and Bioelectronics, 64(2015) 63-8. [36] B. Berg, B. Cortazar, D. Tseng, H. Ozkan, S. Feng, Q. Wei, et al., Cellphone-based hand-held microplate reader for point-of-care testing of enzyme-linked immunosorbent assays, ACS nano, 9(2015) 7857-66. [37] H. Tominaga, M. Ishiyama, F. Ohseto, K. Sasamoto, T. Hamamoto, K. Suzuki, et al., A water-soluble tetrazolium salt useful for colorimetric cell viability assay, Analytical Communications, 36(1999) 47-50. [38] K. Su, Q. Zou, J. Zhou, L. Zou, H. Li, T. Wang, et al., High-sensitive and high-efficient biochemical
analysis method using a bionic electronic eye in combination with a smartphone-based colorimetric reader system, Sensors and Actuators B: Chemical, 216(2015) 134-40. [39] A. Ledreux, A.-L. Sérandour, B. Morin, S. Derick, R. Lanceleur, S. Hamlaoui, et al., Collaborative study for the detection of toxic compounds in shellfish extracts using cell-based assays. Part II: Application to shellfish extracts spiked with lipophilic marine toxins, Analytical and bioanalytical chemistry, 403(2012) 1995-2007. [40] T. Yasumoto, Y. Oshima, M. Yamaguchi, Occurrence of a new type of shellfish poisoning in the Tohoku district, Bulletin of the Japanese Society of Scientific Fisheries, 44(1978) 1249-55. [41] M. Wiese, P.M. D’agostino, T.K. Mihali, M.C. Moffitt, B.A. Neilan, Neurotoxic alkaloids: saxitoxin and its analogs, Marine drugs, 8(2010) 2185-211.
Biographies Kaiqi Su received his B.E. degree and M.S. degree of biomedical engineering in Zhejiang University, PR China in 2011 and 2014. Now he is a Ph.D. candidate of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, cell-based biosensor, cell metabolic physiology, biosensor instrument establishing and signal processing.
Yuxiang Pan received his B.E. degree of biomedical engineering in Sichuan University, PR China in 2015. Now he is a Ph.D. candidate of biomedical engineering of Zhejiang University. His work includes research of flow cytometry, cell-based biosensor and cell metabolic physiology.
Zijian Wan received his B.E. degree of biomedical engineering in Southeast University, PR China in 2015. Now he is a master student of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, biosensor instrument establishing and signal processing.
Jielong Zhong received his B.E. degree of biomedical engineering in Zhejiang University, PR China in 2016. Now he is a master student of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, biosensor instrument establishing and signal processing.
Jiaru Fang received her BS degree of biomedical engineering, PR China in 2014 in Jinan University. Now she is a master student of biomedical engineering of Zhejiang University.
Her work includes research of electrochemical sensors, cell-based biosensor and cell metabolic physiology.
Quchao Zou received his BS degree of biomedical engineering, PR China in 2013 in University of Shanghai for Science and Technology. Now he is a master student of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, cell-based biosensor, cell metabolic physiology, biosensor instrument establishing and signal processing.
Hongbo Li received his B.E. degree of biomedical engineering in Zhejiang University, PR China in 2013. Now he is a Ph.D. candidate of biomedical engineering of Zhejiang University. His work includes research of electrochemical sensors, cell-based biosensor, cell metabolic physiology, biosensor instrument establishing and signal processing.
Ping Wang received his BS degree, MS degree and PhD degree of electrical engineering in Harbin Institute of Technology, Harbin, PR China in 1984, 1987 and 1992, respectively. He is currently a professor of Biosensors National Special Lab, Department of Biomedical Engineering of Zhejiang University. His research interests include biomedical sensors, electrochemical sensors and measurement technique.
Illustration of the figures Figure 1 (A) The action principle of CCK-8 kit. The CCK-8 kit contains WST-8 and 1Methoxy PMS. The dehydrogenases catalyze the conversion from NAD+ to NADH in cells while the inverse conversion from NADH to NAD+ is accompanied by the conversion from 1Methoxy PMS to 1-Methoxy PMS reduced form. In the extracellular microenvironment, WST-8 reacts with 1-Methoxy PMS reduced form to generate the WST-8 formazan (orange water-soluble product) and 1-Methoxy PMS. Thus the color depth can be detected to reflect the living cell count or cell status. (B) Construction of CVBS. (C) The picture of portable smartphone-based system and (D) the main interface of the homemade software iPlate Monitor.
Figure 2 Image monitoring test of the portable smartphone-based system. The system monitored the 96 wells of blank MTP for about 20 h. And the captured image sequence was analyzed by (A) red channel, (B) green channel and (C) blue channel. There were 96 curves in every subgraph.
Figure 3 Real-time monitoring of various cell seeding densities. (A) The real-time CVI curves of HepG2 cells were measured by the system with different cell densities for about 5 h. (B) Rsquared value and sensitivity variations over time and (C) the standard curve of best detection point-in-time. The R-squared (0.9956) and slope (169.4) both reached the maximum at 1.07 h during the monitoring process. The CVI was calculated once each 2 minutes and the image was saved once every 4 minutes. For the clarity of figure, the data in (A) were resampled and
the time interval was 6 min. All the data in (A) and (C) were performed in 3 repeated test and the error bars represented the standard deviation (SD).
Figure 4 The microscopic images of HepG2 cells at best detection point-in-time with different initial cell seeding densities. The cellular morphology of the group with CCK-8 addition was consistent with that of the group without CCK-8 addition. The magnification scales of all cell images were 40 and 100. Figure 5 (A) Real-time NCVI curves of the test groups (10 – 800 µg/L OA) and control group (fresh medium) during the detection process. (B) R2 and sensitivity variations over time and (C) the standard curves of the highest R2 and the highest sensitivity. The R-squared (0.9536) reached the maximum while the slope was 2.569 at 2.85 h. All the data in (A) and (C) were performed in 3 repeated test and the error bars represented the SD.
Figure 6 The microscopic images of HepG2 cells after treatment with different concentrations of OA at the 2.85 h. The magnification scales of all cell images were 40 and 100.
Figure 7 Real-time monitoring of the response to control, 800 µg/L GTX, 400 µg/L GTX, 800 µg/L PbTx-2 and 400 µg/L PbTx-2 by HepG2-CVBS (A). The NCVIs of HepG2-CVBS after 2.85 h treatments of three toxins (B). All the data were performed in 3 repeated test and the error bars represented the SD. Levels of significance: ***p < 0.001; ****p < 0.0001.
Fig 1
Fig 2
Fig 3
Fig 4
Fig 5
Fig 6
Fig 7
Table 1 The results of detecting marine toxins in non-toxic mussel extracts spiked with standard solution by CVBS and MBA CVBS
MBA
Added conc. Determined conc.
Percentage
Determined conc.
Percentage
(µg L-1)
recovery
(µg L-1)
recovery
300
298.05±27.42
99.35
295.23±54.32
98.41
400
386.36±37.47
96.59
382.68±75.01
95.67
500
520.15±44.21
104.03
536.60±54.73
107.32
600
561.24±34.79
93.54
563.52±71.56
93.92
700
712.46±37.76
101.78
738.36±57.59
105.48
(µg L-1)