Biosensors and Bioelectronics 86 (2016) 346–352

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

journal homepage: www.elsevier.com/locate/bios

A novel, sensitive and label-free loop-mediated isothermal

amplification detection method for nucleic acids using luminophore
dyes
Sharmili Roy a, Sim Xiao Wei a, Jean Liew Zhi Ying a, Mohammadali Safavieh b,
Minhaz Uddin Ahmed a,n
a
Biosensors and Biotechnology Laboratory, Chemical Science Programme, Faculty of Science, Universiti Brunei Darussalam, Jalan Tungku Link, Gadong, BE
1410 Brunei Darussalam
b
Division of Biomedical Engineering, Division of Renal Medicine, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston,
MA, USA

art ic l e i nf o a b s t r a c t

Article history: Electrochemiluminescence (ECL) has been widely rendered for nucleic acid testing. Here, we integrate
Received 11 April 2016 loop-mediated isothermal amplification (LAMP) with ECL technique for DNA detection and quantifica-
Received in revised form tion. The target LAMP DNA bound electrostatically with [Ru(bpy)3] þ 2 on the carbon electrode surface,
21 June 2016
and an ECL reaction was triggered by tripropylamine (TPrA) to yield luminescence. We illustrated this
Accepted 21 June 2016
Available online 22 June 2016
method as a new and highly sensitive strategy for the detection of sequence-specific DNA from different
meat species at picogram levels. The proposed strategy renders the signal amplification capacities of
Keywords: TPrA and combines LAMP with inherently high sensitivity of the ECL technique, to facilitate the detection
Electrochemiluminescence of low quantities of DNA. By leveraging this technique, target DNA of Sus scrofa (pork) meat was detected
Loop-mediated isothermal amplification
as low as 1 pg/mL (3.43  10  1 copies/mL). In addition, the proposed technique was applied for detection
Nucleic acid detection
of Bacillus subtilis DNA samples and detection limit of 10 pg/mL (2.2  103 copies/mL) was achieved. The
Sus scrofa
Bacillus subtilis advantages of being isothermal, sensitive and robust with ability for multiplex detection of bio-analytes
makes this method a facile and appealing sensing modality in hand-held devices to be used at the point-
of-care (POC).
& 2016 Elsevier B.V. All rights reserved.

1. Introduction amplification (PSR), loop-mediated isothermal amplification

(LAMP). Among these isothermal amplifications, LAMP has been
The development of rapid, sensitive nucleic acid based point of used more than any other isothermal technique due to being
care diagnostics is of paramount importance for improving global sensitive, specific, robust and reliable.
health. These techniques can be utilized as an alternative to re- LAMP amplification is widely performed because of its high
place complementary and conventional cell culturing as well as specificity and sensitivity. It can detect very minimum or very few
immunological assays. Most of nucleic acid assays are difficult to target DNA copies per analysis. Despite the requirement of ex-
perform and requires bulky and costly equipments. For example, tensive effort in designing six sets of primers, LAMP can be easily
polymerase chain reaction (PCR) requires three different thermal integrated with lab on a chip devices to perform the detection
process, which clearly applicable in the development of biosensor
cycling, electrophoresis post processing or fluorescent labeling. All
and POC monitoring devices (Ahmed et al., 2014, Mori et al., 2001,
these limitations, have led nucleic acid amplification toward re-
Notomi et al., 2000). LAMP amplicon were detected utilizing tur-
moving thermal cycling and performing amplification iso-
bidity (Mori et al., 2001), electrochemical (Ahmed et al., 2009,
thermally. In this regard, various isothermal amplifications such as Safavieh et al., 2012a), gel electrophoresis (Le et al., 2012), fluor-
nucleic acid sequence base amplification (NSABA), Rolling poly- escence (Hong et al., 2012; Safavieh et al., 2012b, 2014a), lateral
merase amplification (RPA), Helicase dependent amplification, flow test (Safavieh et al., 2016) and magnetic beads (Roy et al.,
Smart amplification process 2 (SMAP), polymerase spiral 2016a). All these detection modalities have a major drawback in
providing significant difference between positive and negative
n
Corresponding author.
samples, which yields difficulties in quantitative analysis of nucleic
E-mail addresses: minhaz.ahmed@ubd.edu.bn, acids (Jiang et al., 2013). Consequently, effective and easy quanti-
minhazua@gmail.com (M.U. Ahmed). tative detectable processes are required (Table S1).

http://dx.doi.org/10.1016/j.bios.2016.06.065
0956-5663/& 2016 Elsevier B.V. All rights reserved.
 S. Roy et al. / Biosensors and Bioelectronics 86 (2016) 346–352 347

Electrochemiluminescence (ECL) is a sensitive detection mod- 2.3. ECL detection

ality, where high energy states are produced on the surface of
electrodes which facilitate electron transfer. This electron transfer For the analysis of the LAMP samples, 100 mL of [Ru(bpy)3]Cl2
phenomenon yields to excit species and subsequently emit light and 100 mL of TPrA were added in TE buffer, pH 7.5–5 mL of LAMP
which can be recorded as an intensity signal (McNaught and amplicon per analysis (ST 1 in Supplementary information). This
Wilkinson, 1997; Su et al., 2011). Various compound such as 9, 10- mixture was placed in an ECL cell comprising a 2 mL tubular bottle
diphenylanthracene (DPA) or ECL luminophores were discovered completely shielded with a silver-mirror film, leaving only the
(Santhanam and Bard, 1965). Tris(2, 2ʹ-bipyridyl) di- base (diameter 0.8 cm) to allow the transmission of light. The ECL
chlororuthenium(II) hexahydrate ([Ru(bpy)3]Cl2) was used as the cell was placed above the center of a PMT, which detected the ECL
ECL luminophore. [Ru(bpy)3]2 þ which showed strong lumino- intensity while the rest of the cell was shielded in a black box. The
phore characteristics, and provided high chemical stability in concentration range tested for [Ru(bpy)3]Cl2 was 200 mM to
aqueous solutions with high emission quantum yield and a long 1000 mM, while a concentration range of 10 mM to 50 mM was
excited state duration of 600 ns (Caspar and Meyer, 1983; Hou- tested for TPrA. These two parameters were tested individually to
ten and Watts, 1976). Various ECL biosensor requires nucleic acid optimize signal acquisition. The optimal combination of [Ru(bpy)3]
surface immobilization technique. Recently, immobilization-free Cl2 and TPrA concentrations was also tested over a wide range,
ECL was developed for micro-RNA detection by designing two from 1:2, 2:3, 3:2, 2:1, 3:1 to 1:3. The pH of the reaction mixture
auxiliary probe which yields the formation of long range self as- was similarly optimized following testing over pH 5.5–9.5.
sembly of long 1 dimensional DNA concatemer (Liu et al., 2014). In
addition, similar immobilization-free ECL sensing modality was
described to detect sequence-specific rolling-circle amplification 3. Results and discussion
(RCA) amplicon (Luo et al., 2015). However, RCA is not a suitable
amplification for variety of DNA with middle to long range (Nallur The main goal of this study was to develop a new LAMP-ECL
et al., 2001; Luo et al., 2015). In addition, due to relatively long sensor-based platform with high sensitivity and selectivity. This
amplification time to obtain a detectable signal (Marciniak et al., was achieved using LAMP DNA products directly without the need
2008) as well as requirement to design the padlock probes makes for immobilization and labeling. Compared with other studies
this technique cumbersome to be used for microbial and animal (Table S2) our method is simple, fast and easy to use, and produces
DNA amplification. On the other hand, LAMP provided the ability accurate and reproducible results. The method of detection via the
to amplify medium to long range nucleic acids within short am- quenching of the ECL signal with the target DNA was straightfor-
plification time (Notomi et al., 2000; Roy et al., 2016b). ward, using software that measures the ECL intensity. In the first
Herein, we present label and immobilization free ECL sensing step, reactive elements are formed in the electrochemical reaction
modality technique for LAMP amplicon detection. 1 pg/mL and and in the subsequent steps, the electrogenerated compounds
10 pg/mL of Sus scrofa and Bacillus subtilis DNA amplicon were were diffused from the electrode surface. In the solution, these
detected, using the integration of LAMP and ECL based quantita- compounds react with each other and produce light at the elec-
tive detection within 13.5 min and 17.2 min, respectively. trodes (Chang and Zheng, 2006).
For this study, we chose to investigate species-specific detec-
tion of one model animal organism and one bacterial organism
2. Materials and methods which both have a significant impact on food purity and safety.
The presence of Sus scrofa in food products and its use to adulte-
2.1. DNA preparation and collection rate other species’ meat are vital issues in terms of the food in-
dustry's compliance with religious codes, as Islam, Hinduism and
Pork meat samples were collected from different markets with Judaism restricts followers from consuming any food containing
different origins and following genomic DNA extraction with the pork or its derivatives (Ahmed et al., 2010; Regenstein et al., 2003;
DNeasy extraction method, the purified DNA concentrations were Roy et al., 2016a; Safavieh et al., 2014a, 2014b). Apart from re-
measured with a NanoPhotometer (Implen GmbH, Germany) and ligious issues, the consumption of pork poses the risk of infection
these samples were then stored at  20 °C until further use. The with dangerous parasites such as the Trichina worm, which can
Bacillus subtilis genomic DNA was obtained from the Leibniz In- cause several internal diseases. Bacillus subtilis was chosen as the
stitute (DSMZ, Germany) and stored at  20 °C. second target species to confirm that the method described in this
study can be adapted to different species in a species-specific
2.2. Preparation of LAMP reactions manner. It is also an important food contaminant as this bacterium
is a very common cause of food poisoning and damage to the
Six different sets of Sus scrofa primers and five different sets of human immune system (Richard et al., 1988).
Bacillus subtilis primers were tested for the best primer combina- We successfully illustrated the detection of species-specific
tions for LAMP amplification and ECL-based detection. These no- target DNA of these two organisms with a solution-based ECL
vel, optimized primer sets for Sus scrofa and Bacillus subtilis de- assay without the need for immobilization, using [Ru(bpy)3]Cl2 as
tection are listed in Table S1, and comprise regions of the Cyto- the luminophore and TPrA as a co-reactant with the target DNA.
chrome b gene (GenBank accession #AFO34253.1) and rpoB gene We used this ECL-based method to detect the LAMP products
(GenBank accession #NC_000964.3), respectively. Primers were specifically amplified from pork or Bacillus subtilis genomic DNA
designed using PrimerExplorer version 4.0 (Eiken Chemical Co., (gDNA). For this purpose, we designed new and highly sensitive
Ltd., Japan). LAMP reactions were carried out at temperatures primers for LAMP amplification, rather than resorting to other
ranging from 60 °C to 65 °C and for different time periods of up to conventional and expensive amplification methods (Wu et al.,
60 min (ST 1 in Supplementary information). The total volume of 2013). The electrostatic interactions and binding affinity of [Ru
the LAMP reaction was 25 mL each and 5 mL of gDNA template was (bpy)3]Cl2 and DNA were observed to slow down the diffusion of
added to each reaction. To avoid evaporation, 10 mL of mineral oil the luminophore on the carbon electrode surface, resulting in a
was used to overlay the reaction solutions. The gels were run at reduction in ECL intensity in the presence of LAMP DNA products
80 V for 2 h and subsequently viewed with an Ultra Violet Product compared with the signal obtained in the absence of these pro-
(UVP machine, Upland, CA, USA). ducts (Fig. 1). Firstly, LAMP amplifications were performed with
348 S. Roy et al. / Biosensors and Bioelectronics 86 (2016) 346–352

Fig. 1. Schematic illustration of the mechanism of the LAMP-ECL system in the presence or absence of DNA. A. In absence of dsDNA, [Ru(bpy)3]Cl2 and TPrA emit high ECL
intensities on the surface of the carbon electrodes. B. When dsDNA strands are produced by the LAMP reaction these DNA molecules bind to [Ru(bpy)3]Cl2 and TPrA, resulting
in lower photon emission and lower ECL intensity being detected.

Sus scrofa and Bacillus subtilis gDNA, followed by quick and simple The LAMP reactions were performed with 100–0.0001 ng/mL of
ECL detection of these LAMP amplicon. pork and Bacillus input genomic DNA to determine the lowest limit
As shown in Fig. 1, in the absence of double-stranded DNA of detection obtainable for each using our method, with a resulting
(dsDNA) on the surface of the carbon electrode, we observed that limit of detection for pork and Bacillus subtilis species detection of
the ECL intensity was higher than when in presence of dsDNA. The 1 pg/mL and 10 pg/mL, respectively. In 1 fg/mL (0.0001 ng/mL), LAMP
main explanation for this reduction in ECL intensity is that the did not amplify the target and no amplicon was produced.
diffusion of the ECL luminophore on the carbon electrode is slow Therefore, the ECL intensity was higher compare to positive LAMP
(Ahmed et al., 2013), due to bonding between positively charged (100–0.0001 ng/mL) product. To check for species cross-reactivity
[Ru(bpy)3]Cl2 and negatively charged dsDNA, which in turn re- (i.e. assay non-specificity), the LAMP assays for both species were
duces the amount of light emission detected, causing a low ECL also tested with genomic DNA input from various other species
intensity signal. (Fig. 2, Supplementary Tables S3 and S4). It was observed that in
both cases the loop primers did not cross-react with other species’
3.1. Optimization of LAMP assay and [Ru(bpy)3]Cl2/TPrA experi- genomic DNA (Fig. 2). A DNA template input concentration of
mental conditions 10 pg/mL was used for all cross-reactivity experiments.
In addition, the ECL sensing strategy was optimized with [Ru
LAMP was optimized and standardized for Sus scrofa and Ba- (bpy)3]Cl2-TPrA in the presence and absence of the targeted species’
cillus subtilis species detection using species-specific primers (Ta- dsDNA (i.e. LAMP amplicon). As [Ru(bpy)3]Cl2 is a critical component
ble S1). First, DNA was extracted from different types of processed of this label-free detection method, its concentration was the first to
foods (Supplementary Table S3). The food samples purchased were be optimized over a range of 200 mM to 1000 mM (Fig. 3A). The highest
of different origins and most were imported from neighboring ECL signal was obtained with 800 mM [Ru(bpy)3]Cl2, while increasing
countries. For Sus scrofa, LAMP amplifications were tested with six concentrations led to a steady reduction of ECL intensity. This was
different sets of primers and the combinations with the best most likely due to the increasing concentrations resulting in saturation
specificity and sensitivity were used for further experiments. For of [Ru(bpy)3]Cl2 in solution, leading to the stacking of excess [Ru
Bacillus subtilis, a similar optimization was carried out with five (bpy)3]Cl2 on the surface of carbon electrodes hindering the trans-
different sets of primers. The LAMP amplification temperature was mission of light signal (Li and Macdonald, 2015; Tang et al., 2013).
next optimized, by testing amplification at 60 °C, 63 °C and 65 °C. The concentration of TPrA was also an important factor in
It was observed that for S. scrofa, optimal LAMP amplification oc- achieving optimal ECL intensity. A range of TPrA concentrations of
curred at 65 °C whereas 63 °C was optimal for B. subtilis. At each of 10–50 mM was tested and 20 mM was found to be optimal (Fig. 3B).
the different temperatures tested, various reagent concentrations Increasing concentrations of TPrA above 20 mM led to reduced ECL
were simultaneously tested over different ranges to find the op- response. This effect may similarly be caused by excess TPrA
timal concentrations for high LAMP product yield. Betaine was stacking with [Ru(bpy)3]Cl2 on the surface of the electrodes (Zu
optimized from 0.4 to 1.4 M, dNTPs from 0.4 to 1.4 mM, and MgSO4 and Bard, 2000). While optimizing the concentration of [Ru(bpy)3]
from 3 to 8 mM. Optimal results for the pork samples were ob- Cl2, which was varied from 200 mM to 1000 mM the concentration
tained with betaine at 1.4 M, dNTPs at 1.4 mM and MgSO4 at of TPrA was maintained at 20 mM. Similarly, for TPrA optimization
6 mM. For the Bacillus samples, optimal results were obtained with which was varied from 10 mM to 50 mM and the concentration of
betaine at 0.64 M, dNTPs at 0.4 mM and MgSO4 at 3 mM. For op- [Ru(bpy)3]Cl2 was maintained at 800 mM. A series of tests were
timization of the reagent concentrations, all experiments were performed with different combinations of concentrations before
carried out in triplicate. The resulting LAMP amplification products concluding these particular concentrations as the optimum ECL
of each experiment were subjected to 2% agarose gel electro- signals. Therefore, all further experiments included 800 mM [Ru
phoresis for analysis. (bpy)3]Cl2 and 20 mM TPrA.
 S. Roy et al. / Biosensors and Bioelectronics 86 (2016) 346–352 349

Fig. 2. Cross-reactivity of Sus scrofa and Bacillus subtilis primers. A. Cross-reactivity of the pork primers was measured for the negative control (NTC) and chicken, pork
(positive control) and beef genomic DNA. B. Cross-reactivity of the Bacillus subtilis primers was measured for Bacillus subtilis (positive control), Bacillus cereus, Staphylococcus
aureus and Legionella pneumophila genomic DNA. To check the specificity of this platform, ECL detection was also performed with various other samples with the pork-
specific primers (Fig. S1).

Fig. 3. Optimization of the [Ru(bpy)3]Cl2 and TPrA concentrations for ECL signal detection. (A) A range of [Ru(bpy)3]Cl2 concentrations of 200–1000 mM was tested. (B) A
range of TPrA concentrations of 10–50 mM was tested.

Fig. 4. Optimization of the ratio of [Ru(bpy)3]Cl2 to TPrA for ECL signal detection. (A) The tested ratios of [Ru(bpy)3]Cl2 to TPrA were 1:2, 2:3, 3:2, 2:1, 3:1 and 1:3. The highest
ECL intensity was obtained at a ratio of 2:1. (B). The optimal pH of the reaction solution was tested over the pH range of 5.5–9.5 in (a) the presence of a LAMP-negative DNA
sample and (b) in the presence of a LAMP-positive DNA sample.
350 S. Roy et al. / Biosensors and Bioelectronics 86 (2016) 346–352

We next investigated the optimal ratio of [Ru(bpy)3]Cl2 to TPrA testing 2–10 mL LAMP amplicon. The addition of a 5 mL volume of
by testing ratios of 1:2, 2:3, 3:2, 2:1, 3:1 and 1:3. When the ratio LAMP DNA was observed to yield significant and reproducible ECL
reached 2:1 (200 mL of [Ru(bpy)3]Cl2 and 100 mL of TPrA in solu- intensities.
tion) the ECL intensity was at its highest, while increasing ratios The sensitivity of the ECL-LAMP assay was then determined for
produced lower ECL intensities (Fig. 4A). The 2:1 ratio was both test species using all of the above optimized parameters. For
therefore used for all further experiments. The reduced ECL in- Sus scrofa, the limit of detection was determined to be 1 pg/mL
tensities from increasing ratios probably occurred due to excess (Fig. 5A), while for Bacillus subtilis the limit of detection was 10 pg/
luminophore and co-reactant stacking onto the surface of the mL (Fig. 5B). These experiments were carried out with a 2:1 ratio of
electrodes, thereby reducing light transmission (Zu and Bard, [Ru(bpy)3]Cl2 to TPrA inside an ECL cell containing 5 mL of LAMP
2000). DNA at the tested concentrations. The total volume inside the ECL
The combination of [Ru(bpy)3]Cl2 and TPrA in the presence of cell was maintained at 305 mL. The carbon chip was attached to the
TE buffer leads to light emission as a result of an irreversible connector and immersed in the solution. The parameters during
anodic wave that causes direct oxidation of both compounds (Zu sample scanning were sensitivity ¼1.e  005 A/V with amplifying
and Bard, 2000). Determining the optimal pH of this solution was series 3, with the voltage of the PMT tube set to 700 V. Maximum
therefore a vital issue, especially for optimizing DNA detection. A signal intensity was observed  2 s after starting each scan.
range of pH values was evaluated from pH 5.5 to 9.5 (Fig. 4B). The As per the principle of ECL detection on carbon electrodes, it
highest signal both in the presence and absence of LAMP DNA was observed that while testing LAMP amplicon concentrations of
products was obtained at  pH 5.5, with a steady decrease in 1000 pg/mL to 1 pg/mL the ECL intensity remained low due to
signal obtained with increasing pH. As explained previously, when charge compensation on the electrode surface (Ahmed et al.,
[Ru(bpy)3]Cl2 and TPrA bind with the positive LAMP amplicon, 2013). The use of even lower concentrations of 100 fg/mL LAMP
lower ECL signals are obtained compared with those obtained in amplicon resulted in high ECL intensity similar to that for the
the absence of such amplicon (Fig. 4BA). Since pH 7.5 is close to the negative control.
physiological pH therefore, this pH value was selected and used We next investigated the ECL intensity-potential curves ob-
throughout the study. tained with carbon electrodes with [Ru(bpy)3]Cl2 and TPrA only, or
Over most of the pH range tested, the ECL signal was lower on in the presence of LAMP-negative samples or LAMP-positive
the surface of the carbon electrodes when in the presence of LAMP samples (Supplementary Figs. S7 and S8). The [Ru(bpy)3]Cl2 and
DNA (see e.g. signal at pH 7.5, Fig. 4B). pH 7.5 is physiological pH at TPrA-only reaction produced the highest ECL emission peak due to
which DNA remains stable in solution and would therefore pro- the absence of any DNA in the reaction (curve a). However, after
vide a more stable ECL signal over time; we therefore chose TE the addition of LAMP-negative sample (only template gDNA pre-
buffer pH 7.5 for all future experiments. sent), the ECL intensity was reduced (curve b), while the addition
of a LAMP-positive sample (i.e. both template gDNA and LAMP
3.2. Sensitivity and specificity of the LAMP-ECL sensor amplicon present) reduced the intensity of the curve even further
(curve c) due to the [Ru(bpy)3]Cl2 and TPrA forming complexes
After LAMP-positive amplicon were confirmed by gel electro- with the LAMP DNA, resulting in slower diffusion of the lumino-
phoresis, all positive samples were analyzed with the ECL detec- phore across the electrode surface (Wang et al., 2012).
tion system. LAMP amplification with a set of several specific We also wished to evaluate the changes in the current resulting
primers produces highly specific amplicon, thereby improving the from the addition of [Ru(bpy)3]Cl2 -TPrA and the LAMP amplicon
ease and specificity of ECL-based detection. For negative control under electrochemical conditions. From this investigation, it was
samples, only [Ru(bpy)3]Cl2 and TPrA were used, while for the noticed that in the presence of LAMP amplicon, the peak height
above optimization experiments, 5 mL of LAMP amplicon was ad- was lower than when only [Ru(bpy)3]Cl2 and TPrA were present
ded to [Ru(bpy)3]Cl2 and TPrA in solution. Once the other para- (i.e. in the absence of any LAMP amplicon; Supplementary Fig. S2).
meters described above were optimized, the volume of LAMP To observe the surface charge in the presence of [Ru(bpy)3]Cl2
amplicon was added, the detection solution was also optimized by -TPrA and LAMP amplicon, CC tests were also performed

Fig. 5. ECL detection of LAMP amplicon produced with Sus scrofa and Bacillus subtilis loop primers. A) The limit of detection of the LAMP-ECL assay for pork species of LAMP
DNA. B) The limit of detection of the LAMP-ECL assay for Bacillus subtilis. A negative control containing no DNA was also tested (NTC).
 S. Roy et al. / Biosensors and Bioelectronics 86 (2016) 346–352 351

(Supplementary Fig. S3). An increased ECL intensity was observed the LAMP method and the easy fabrication of the ECL system
in the presence of [Ru(bpy)3]Cl2 and TPrA alone, while lower signal combined hold significant potential for incorporation into point-
was obtained with the mixture of [Ru(bpy)3]Cl2, TPrA and LAMP of-care devices for food and clinical safety in low-resource areas,
amplicon. which could incorporate chip types constructed from different
materials such as paper and plastics to make it even more cost-
3.3. Time-dependence and stability testing of ECL sensor effective (Maxwell et al., 2013; Liana et al., 2013; Hu et al., 2013;
Nahar et al., 2015; Tlili et al., 2013; Ahmed et al., 2016; Safavieh
We next evaluated the stability testing over time in the ECL et al., 2016). Overall, the main advantage of the system described
system for LAMP DNA detection. In comparison with conventional here is the faster detection, minimal instruments and consump-
methods such as gel electrophoresis, the ECL sensor was shown to tion of less reagents. The combination of ECL sensing modality
be more sensitive (Fig. S4). In general, LAMP amplification takes with high sensitivity and specificity and rapid amplification of
15–60 min, while the post-amplification detection process, which LAMP makes this technology an appealing for robust quantifica-
is commonly carried out with gel electrophoresis, takes  40 to tion of nucleic acids. Fast, sensitive, specific, easy-to-operate as-
60 min. In contrast, our ECL method can detect positive LAMP says combined with the miniaturization of instruments have be-
amplicon after only 15 min of amplification (Fig. S6A). In total come increasingly important in research (Ahmed et al., 2008), and
therefore, our ECL sensor platform takes less than 16 min instead our LAMP-ECL approach can fulfill all these requirements making
of the total time of at least 55 min for conventional methods. This this system ideal for incorporation in POC devices for various
faster detection time is made possible by the upstream use of the applications.
LAMP reaction which is highly sensitive and specific.
The effect of incubation time of LAMP DNA with [Ru(bpy)3]Cl2-
TPrA on the ECL intensity was considered another critical para- Acknowledgments
meter. The LAMP DNA was shown to bind to [Ru(bpy)3]Cl2-TPrA
and produce ECL signal within a few seconds, with increasing Minhaz Uddin Ahmed would like to acknowledge financial
signal observed with increasing incubation periods (Supplemen- support for the project provided by a Universiti Brunei Darussalam
tary Fig. S5). In addition, no signal reduction was observed after (UBD) grant, Grant# UBD/PNC2/2/RG/1 (255). Mohammadali Sa-
6 cycles over a period of 2 h, with clear distinction remaining favieh wishes to acknowledge financial support from Natural Sci-
between signal generated in the presence or absence of LAMP DNA ences and Engineering Research Council of Canada (NSERC)
(Fig. S6B). In previous research, it was shown that when the redox through NSERC postdoctoral fellowship. Sharmili Roy wishes to
component binds to the LAMP amplicon it produces less signal on thank the UBD Graduate Research Scholarship (GRS) for her Ph. D.
the carbon surface. This signal decrease is due to the interaction of fellowship. The authors also acknowledge Mohammad Rizwan for
these compounds with the carbon electrodes, especially redox- his useful comments.
DNA complexes which have slow diffusion rates (Ahmed et al.,
2010). Similarly in our study, when the LAMP amplicon bound to
[Ru(bpy)3]Cl2 through electrostatic interactions at the carbon Appendix A. Supplementary material
electrode surface, these complexes showed slower diffusion on the
carbon electrode surface resulting in decreased ECL signal. More- Supplementary data associated with this article can be found in
over, the [Ru(bpy)3]Cl2 complex binds DNA through intercalation the online version at doi:http://dx.doi.org/10.1016/j.bios.2016.06.
with the double helix, stacking with the DNA without disturbing 065.
the base pairs (Lim et al., 2009). Based on calculations of the DC,
reactions in the presence of DNA (0.0425 cm2/s) produce slower
diffusion than in the absence of DNA (0.876 cm2/s). The DC cal- References
culations prove that on the surface of the carbon electrodes the
presence of DNA would indeed result in lower ECL signal in- Ahmed, M.U., Hasan, Q., Hossain, M.H., Masato, S., Tamiya, E., 2010. Food Control 21,
tensities than those produced in its absence. 599–605.
Ahmed, M.U., Nahar, S., Safavieh, M., 2013. Analyst 138, 907–915.
Ahmed, M.U., Saaem, I., Wu, P.C., Brown, A.S., 2014. Crit. Rev. Biotechnol. 34,
180–196.
4. Conclusions Ahmed, M.U., Hossain, M.M., Tamiya, E., 2008. Electroanalysis 20, 616–626.
Ahmed, M.U., Hossain, M., Safavieh, M., Wong, Y.L., Rahman, I,A., Zourob, M., Ta-
miya, E., 2016, 36, pp. 495–505.
A new label-free and immobilization-free ECL-based DNA Caspar, J.V., Meyer, T., 1983. J. Am. Chem. Soc. 105, 5583–5590.
sensor was simply constructed based on complex formation of [Ru Chang, Z., Zheng, X., 2006. J. Electroanal. Chem. 587, 161–168.
(bpy)3]Cl2 and TPrA in the presence or absence of LAMP amplicon. Hong, M., Zha, L., Fu, W., Zou, M., Li, W., Xu, D., 2012. Microbiol. Biotechnol. 28,
523–531.
Various parameters combined with newly designed species-spe- Houten, J.V., Watts, R.J., 1976. J. Am. Chem. Soc. 98, 4853–4858.
cific primers improved the ECL signal and yielded robust, re- Hu, J., Wang, S., Wang, L., Li, F., Murphy, B.P., 2013. Biosens. Bioelectron. 54,
producible, and quick detection, with high sensitivity and speci- 585–597.
Jiang, Y., Li, B., Milligan, J.N., Bhadra, S., Ellington, A.D., 2013. J. Am. Chem. Soc. 135,
ficity. This sensor was shown to detect LAMP amplicon from the 7430–7433.
11.3 min of amplification, with less than 1 min of incubation with Li, J., Macdonald, J., 2015. Biosens Bioelectron. 64, 196–211.
[Ru(bpy)3]Cl2-TPrA and a few seconds of scanning with the ECL Le, T.H., Nguyen, N.T.B., Truong, N.H., 2012. Microbiol. 50 (4), 1178–1184.
Liana, D.D., Raguse, B., Gooding, J.J., Chow, E., 2013. Sensors 12, 11505–11526.
sensor required for detection of these amplicon, in contrast with Lim, M.H., Song, H., Olmon, E.D., Dervan, E.E., Barton, J.K., 2009. Inorg. Chem. 48,
the 60 min total required for other combinations of amplification 5392–5397.
and post-amplification detection strategies. With this system we Liu, T., Chen, X., Hong, C., Xu, X., Yang, H., 2014. Microchim. Acta 181, 731–736.
Luo, F., Xiang, G., Pu, X., Yu, J., Chen, M., Chen, G., 2015. Sensors 15, 2629–2643.
could detect LAMP amplicon at the picogram level within few
Marciniak, J.Y., Kummel, A.C., Esener, S.C., Heller, M.J., Messmer, B.T., 2008. Bio-
minutes. For pork sample, the limit of detection was determined to techniques 45, 275–280.
be 1 pg/mL (3.43  10  1 copies/mL), while for B. subtilis the limit Maxwell, E.J., Mazzeo, A.D., Whitesides, G.M., 2013. MRS Bull. 38, 309–314.
was 10 pg/mL (2.2  103 copies/mL). Therefore, this method is likely McNaught, A.D., Wilkinson, A., 1997. IUPAC. Compendium of Chemical Terminol-
ogy, second ed. Blackwell Scientific Publications, Oxford, UK.
to facilitate various new and improved biosensor applications, Mori, Y., Nagamine, K., Tomita, N., Notomi, T., 2001. Biochem. Biophys. Res. Com-
both in biomedical and molecular diagnostics. The simplicity of mun. 289, 150–154.
352 S. Roy et al. / Biosensors and Bioelectronics 86 (2016) 346–352

Nahar, S., Ahmed, M.U., Safavieh, M., Rochette, A., Toro, C., Zourob, M.A., 2015. Safavieh, M., Ahmed, M.U., NG, A., Zourob, M., 2014a. Analyst 139 (2), 482–487.
Analyst 140, 931–937. Safavieh, M., Ahmed, M.U., NG, A., Zourob, M., 2014b. Biosens. Bioelectron. 15,
Nallur, G., Luo, C., Fang, L., Cooley, S., Dave, V., Lambert, J., Kukanskis, K., Kingsmore, 101–106.
S., Lasken, R., Schweitzer, B., 2001. Nucleic Acids Res. 29, e118. Safavieh, M., Kanakasabathy, M.K., Tarlan, F., Ahmed, M.U., Zourob, M., Asghar, W.,
Notomi, T., Okayama, H., Masubuchi, H., Yonekawa, T., Watanabe, K., Amino, N., Shafiee, H., 2016. ACS Biomater. Sci. Eng. 2, 278–294.
2000. Nucleic Acids Res. 25, e63. Su, M., Wei, W., Liu, S., 2011. Anal. Chim. Acta 704, 16–32.
Regenstein, J.M., Chaudry, M.M., Regenstein, C.E., 2003. Compr. Rev. Food Sci. F.2, Tang, X., Zhao, D., He, J., Li, F., Peng, J., Zhang, M., 2013. Anal. Chem. 85 (3),
111–127. 1711–1718.
Richard, V., Van, D.P., Snoeck, R., Daneau, D., Meunier, F., 1988. Eur. J. Clin. Microbiol. Tlili, C., Sokullu, E., Safavieh, M., Tolba, M., Ahmed, M.U., Zourob, M., 2013. Anal.
7, 783–785. Chem. 85, 4893–4901.
Roy, S., Rahman, I.A., Santos, J.H., Ahmed, M.U., 2016a. Food Control 61, 70–78. Wang, Y., Chen, Z., Ye, Z., Huang, J.Y., 2012. J. Cryst. Growth 341, 42–45.
Roy, S., Rahman, I.A., Ahmed, M.U., 2016. Anal. Methods. 8, 2391–2399. Wu, Y., Wang, L., Wang, H.Y., Zhang, X.L., Li, L., Zhang, N., Pan, L., Xing, N.N., 2013.
Santhanam, K.S.V., Bard, A.J., 1965. J. Am. Chem. Soc. 87, 139–140. Biosens. Bioelectron. 40, 356–361.
Safavieh, M., Ahmed, M.U., Tolba, M., Zourob, M., 2012a. Biosens. Bioelectron. 31 (1), Zu, Y., Bard, A.J., 2000. Anal. Chem. 72, 3223–3232.
523–528.
Safavieh, M., Ahmed, M.U., Zourob, M., 2012b. Proc. SPIE 8412, Photonics North.
84120A-84120A-6.