Quantitative determination of target gene with electrical sensor

Integrating loop-mediated isothermal amplification (LAMP) with capacitively coupled contactless conductivity detection (C4D), we have developed an electrical sensor for the simultaneous amplification and detection of specific sequence DNA. Using the O26-wzy gene as a model, the amount of initial target gene could be determined via the threshold time obtained by monitoring the progression of the LAMP reaction in real time. Using the optimal conditions, a detection limit of 12.5 copy/μL can be obtained within 30 min. Monitoring the LAMP reaction by C4D has not only all the advantages that existing electrochemical methods have, but also additional attractive features including being completely free of carryover contamination risk, high simplicity and extremely low cost. These benefits all arise from the fact that the electrodes are separated from the reaction solution, that is C4D is a contactless method. Hence in proof of principle, the new strategy promises a robust, simple, cost-effective and sensitive method for quantitative determination of a target gene, that is applicable either to specialized labs or at point-of-care.

Where dsDNA is synthesized at the expense of consuming primers and dNTPs 8 . An insoluble salt, magnesium pyrophosphate precipitate [17][18][19] and protons 17,20 are also produced. The consumption of primers and dNTPs, plus the yield of precipitate, leads to a decrease of the overall ionic strength (as illustrated in Fig 1A), which is what we exploit for monitoring the LAMP reaction via the change in conductivity 21 . Based on this principle, herein an electrical sensor for monitoring the LAMP amplification reaction in real time using a capacitively coupled contactless conductivity detection (C 4 D) was developed (as illustrated in Fig. 1B,C). The new strategy not only has the same advantages that existing electrochemical methods have, but also solves all the challenges existing electrochemical methods face. We anticipate it will enable the creation of a high throughput, portable device for simple, cost-effective and rapid nucleic acid analysis that is suitable both for working in specialized labs and at points of care.

Results
Characterization of the electrical sensor. Using 0.1, 0.2, 0.3, 0.4 and 0.5 M KCl as probes, the custom built electrical sensor was characterized, focusing on the sensitivity and stability of the C 4 D (Supplementary Note 1 and Supplementary Fig. 1-3). At room temperature, a sensitivity of 871 mV/M was obtained with the optimal conditions of an excitation amplitude of 16 V and an excitation frequency of 2.0 MHz. Moreover, the temperature inside the sensor can be kept stable at 65 °C. At this temperature the C 4 D also shows high sensitivity and stability. Using a NMR sample tube as a reaction vessel, a series of LAMP reactions (200 μ L per sample) were implemented in the electrical sensor when the temperature was programmed to be 65 °C. We observed that high efficiency and specificity of LAMP reactions could be obtained with this set of primers 22 using the optimized conditions (Supplementary Note 2 and Supplementary Fig. 4).
LAMP reaction leading decrease of conductivity response. With 1.25 × 10 4 copy/μ L template DNA or herring sperm DNA, we collected the conductivity responses of the LAMP reaction solutions at room temperature before and after the incubation, respectively. As shown in Fig. 2, the mean output potential value of the negative post-reaction solutions is 1.478 ± 0.001 V. It is not significantly different from the mean value obtained from the same solution prior to performing the amplification reaction. By contrast, the mean value of the positive post-reaction solutions is 1.467 ± 0.002 V, which represents a decrease of ~10 mV compared to the pre-reaction solution, suggesting a significant decrease of conductivity. This result is attributed mainly to the consumption of reactants and the formation of the magnesium pyrophosphate precipitate 21 (Supplementary Note 3 and Supplementary Fig. 5). In addition, the phosphate backbone of the produced dsDNA in the reaction solution can adsorb positive ions (e.g., [K + ], [Na + ], [Mg 2+ ]), leading to a decrease of overall ionic mobility 23 . In conclusion, successful amplification leads to a detectable decrease of conductivity response of the LAMP reaction solution; at the endpoint the change of conductivity can be monitored rapidly (< 2 s) with the electrical sensor, being free of tube-open operation.
Real time monitoring the progression of the LAMP reaction. We prepared the positive (containing 1.25 × 10 7 copy/μ L target DNA) and the negative (containing 1.25 × 10 7 copy/μ L herring sperm DNA) LAMP reaction samples at room temperature, followed by loading them into two NMR sample tubes, respectively. Real time collection of the conductivity responses of the reaction solutions in turn was conducted once the reaction tubes were inserted into the electrical sensor, in which the temperature was kept 65 °C. As shown in Fig. 3, for both cases during the first 3 minutes the output potentials increase rapidly because the rise in temperature leads to an increase in ion mobility 24,25 . With regards to the negative sample, the output potential comes to a plateau in the following 2 min, and remains stable for all the rest period, indicating no reaction proceeded. In contrast, for the positive sample the  output potential begins to drop sharply at the approximately 294th second, indicating the beginning of the detectable decrease of conductivity, which is contributed to the performance of DNA amplification. This time point was defined as threshold time (T t ). It is significantly shorter than that observed with either electrochemical methods 11,14,15 or optical methods 19,26,27 , suggesting a faster response. The sharp decline in the conductivity continues for ~110 s, suggesting a rather high efficiency of the biochemical reaction. Thereafter the decrease of output potential slows down to a gentle decline, indicating that the amplification reaction slows considerably, which may be due to the inhibition of polymerase activity by the fall of pH 16 , or the decrease of the concentration of Mg 2+ 28 . Note, the usage of more ThermoPol ® reaction buffer than the commended dosage, e.g., ≥ 1.2× , benefits to obtain stable curves of conductivity responses (see Supplementary Note 4 and Supplementary Fig. 6). In conclusion, the outcomes show that the electrical sensor can not only provide the temperature condition required for the LAMP reaction, but also monitor the progression of the biochemical reaction in real time under the selected conditions. Performance of quantifying target gene. Figure 4 shows the typical conductivity responses recorded in real time of LAMP reactions with different concentration of initial template DNA. We observed that the larger the amount of initial template DNA is, the shorter the T t is, similar to that appears in real time electrochemical monitoring 11,14,29 , turbidity monitoring 18 and fluorescence monitoring 26,27 . The insert of Fig. 4 shows the plot of T t versus log 10 initial concentration of template DNA. Over the four orders of magnitude concentration range of template DNA investigated, from 1.25 × 10 7 copy/μ L to 1.25 × 10 4 copy/μ L, there is a linear correlation between T t versus log concentration. These results show the quantity of the template DNA of an unknown concentration can be determined by comparing the T t value with the T t values of the template DNA of known concentrations. The velocity of LAMP reaction may depend on the nature of template DNA such as G/C or A/T ratio in amplified region. However, the velocity of LAMP reaction may not affect the quantitative determination by this method since the linearity between T t and the initial amount of template DNA is independent of the velocity of LAMP reaction 30 . Using serial dilutions of the O26-wzy gene sample, the limit of detection was determined to be 12.5 copy/μ L with an incubation time of 30 min, which is lower than that by existing real-time electrochemical method 14 . The sensitivity is 10 times higher than that obtained by conventional PCR method (35 cycles), which is in agreement with previous report 20 .

Discussion
The portable electrical sensor is composed of two key components. As been characterized, the electrical heater with a programmable thermostat allows us to keep the temperature inside at a desired stable isothermal level; the sensitive C 4 D system allows us to monitor the conductivity response of the solution in the reaction tubes in real time. Hence, simultaneous amplification and detection sequence-specific DNA can be implemented by LAMP. Moreover, it also opens the door to investigate the thermodynamic and kinetic mechanisms of many chemical and biochemical reactions, in which change of ionic activity are involved. Note, the associated electronics could be easily miniaturised to a thumb nail size or less 31 , promising the development of portable and high throughput instrumentations.
In LAMP reaction four or six primers are used to recognize six or eight distinct regions of the target gene sequence, so that the specificity is extremely high 4,7,8 . Thus, even indirect methods for monitoring the reaction can be employed to perform the determination of target gene 32 . Among the real-time  Fig. 3. monitoring methods, turbidity shows relatively low sensitivity and slow response 19 . With the help of fluorescence, both the sensitivity and the response speed can be improved significantly at the expense of higher running cost, complex handling procedures 7,27 and non-negligible inhibitory potential from the probes employed 32 . Furthermore, the necessity of optical-electrical signal transferring components increases the complexity in miniaturizing instrumentations and hence manufacturing costs 7,33 . These optical-based methods have an outstanding advantage, however, they are free of carryover contamination risk due to the absence of any tube-open operation. By contrast, electrochemical methods, including voltammetry [11][12][13][14] , conductivity 21 and potentiometry 15,16 , have the advantages of not only the inherent miniaturization and portability, but also the independence from sample turbidity, low-cost/low-power requirements and compatibility with microfabrication technology. However, even under optimal conditions there are still several challenges for monitoring the LAMP reaction with electrochemical methods, e.g., inhibition from the indicators and high risk of carryover contamination 7 .
The C 4 D method employed here in the electrical sensor shares all the merits that these existing electrochemical methods have. Moreover, the unique nature, i.e. the separation between the electrode and reaction solution 34 , highlights dramatic advances by solving completely all the problems the electrochemical methods faced, leading to 1) capacity of realizing successive monitoring non-invasively; 2) free of any probes, indicators or labels; 3) complete elimination of carryover contamination risk; 4) extremely simplicity of operation; 5) extremely low cost. Among these advantages listed above, it is worth a special emphasis on the elimination of carryover contamination because the LAMP reaction may lead to incorrect results upon contamination of even a small quantity of amplification product 4,8 . In conclusion, the electrical sensor has the advantages that optical and electrochemical methods have, meanwhile eliminates their disadvantages, though at present the sensitivity is a little lower than some other schemes 14,32,33 (≤ 1 copy/μ L). The fluctuation of base-line in the present conductivity outputs is another challenge. It is probably be overcome by selecting more suitable reaction vessel. Hence, as a proof of principle, the new strategy promises a superior quantitative determination of target gene, applying either in specialized labs or at the point of care.

Methods
The electrical sensor. As illustrated in Fig. 1B, the electrical sensor was composed of a C 4 D and an electronic heater with thermostat. The C 4 D included two metal electrodes (an excitation electrode and a pick-up electrode). The practical equivalent circuit is shown in Fig. 1C. The two electrodes, the insulating tube and the electrolyte solution form two coupling capacitances C 1 and C 2 . And there was also a stray capacitance arising from direct capacitive coupling between the two electrodes through air (C 3 ) 34,35 . The solution in the reaction tube was equivalent to a resistor R. An ER225 C 4 D System (eDAQ Pty Ltd., Australia) was used to provide an AC source (maximum peak to peak amplitude of 40 V) and an AC current pick-up unit. Thus, an alternating current path was formed. The application of an AC voltage on the excitation electrode led to an AC current flowing through the AC path. From the AC current obtained by the AC current pick-up unit, the conductivity detection of the solution in the tube could be implemented. The property of the C 4 D was characterized referring to the method previous reported 24,35,36 , because the geometry and placement of the sensing electrodes play very important roles in the signal coupling and sensitivity 37 . The electronic heater with a programmable thermostat could provide an isothermal temperature condition (precision: ± 0.3 °C), over the range of room temperature − 120 °C. Commonly, the temperature could be stable in about 20 min after the appointment.

LAMP. We retrieved sequences of O-antigen gene clusters of Escherichia coli serogroups O26 from
GenBank using accession numbers AF529080 (http://www.ncbi.nlm.nih.gov/nuccore/AF529080). Within the cluster, serogroup-specific O26-wzy gene was selected as target to design LAMP primers. A dsDNA fragment related to O26-wzy gene (190 bp in length), which was inserted in pUC57-Amp (2710 bp in length), was synthesized by GENEWIZ Inc. (USA), and was used as template DNA. Primers were synthesized in Genework Pty Ltd. (Sydney, Australia) with the sequences referred to Wang et al. 22  frequency of 2.0 MHz and an excitation amplitude of 16 V, We collected the output potential of its conductivity response. After the reaction solution was incubated at 65 °C for 12 min, it cooled down to room temperature. Then we collected the output potential again with the same parameters.
Real time monitoring the progression of the LAMP reaction. The preparation and load of the reaction solution was at room temperature. Then we inserted the NMR sample tube loaded with reaction solution into the electrical sensor, in which the temperature was programmed to keep 65 °C. We began to collect the output potential with the ER225 C 4 D System in real time at a speed of 1 point per second from the beginning of the incubation. Unless otherwise stated, an excitation frequency of 2.0 MHz and an excitation amplitude of 16 V were selected; and the record lasted for 12 min. Another sample could be implemented as soon as the former was finished, without the step for renewing working electrodes.
Performance of quantifying target gene. Serial dilution method was used to study the performance of quantifying O26-wzy gene. We prepared a series of LAMP reaction samples, in which contained 1.25 × 10 0 , 1.25 × 10 1 , 1.25 × 10 2 , 1.25 × 10 3 , 1.25 × 10 4 , 1.25 × 10 5 , 1.25 × 10 6 , 1.25 × 10 7 and 1.25 × 10 8 copy/μ L, respectively. Then we performed the amplification in the electrical sensor by incubating at 65 °C. Meanwhile, we recorded the conductivity responses in real time, respectively. The reaction time, at which the output potential started to drop sharply, was defined as threshold time (T t ). Average T t from 5 samples for each concentration of template DNA was plotted against log 10 concentration of template DNA. Error bars represent the variation (RSD) between different samples. Generally, we continued to record the output potential for another 5 min after the appearance of T t . The PCR control experiments were performed by referencing to Wang et al. 22