(7a) Portable and Label-Free Quantitative Loop-Mediated Isothermal Amplification (qLAMP) for Reliable COVID-19 Diagnostics in 3 Minutes: An Arduino-Based System Assisted By a pH Microelectrode

COVID-19 has clearly exhibit the severe limitations in our capacities to respond to epidemic emergencies and has made clear that we must expand and strengthen the portfolio of available tools for effective diagnostics of infectious diseases. The retro-transcriptase quantitative polymerase chain reaction (RT-qPCR), currently the gold standard methodology for SARS-CoV-2 detection, possesses unquestionable accuracy and robustness, but also some serious limitations (e.g., lack of portability, dependence on centralized facilities, need for technical expertise to conduct RT-qPCR testing, and high infrastructural and operative costs). Such limitations have prevented the provision of cost-effective and massive diagnostics during this time of COVID-19. Colorimetric Loop-mediated isothermal amplification (colorimetric LAMP) has been recently studied as an alternative method for cost-effective diagnostics in the context of the current COVID-19 pandemic. Recent reports document that colorimetric LAMP-based methods have a comparable sensitivity and specificity to that of RT-qPCR. However, colorimetric LAMP also has some limitations. One is that the adequate performance of most colorimetric LAMP protocols (those mediated by phenol red) depend on the initial pH of the sample, which limits their applicability. For example, acidic saliva samples may prematurely shift the color of the reaction mix, even in non-infected samples, thereby possibly rendering false positives. In addition, LAMP-based methods also use four or six primer sets that may interact among themselves during amplification. This can promote non-specific amplifications that may also result in false positives. Moreover, most LAMP-based technique are final point and quantitative; this is a serious disadvantage when they are compared to RT-qPCR.

We report the use of a portable Arduino-based and label-free and quantitative LAMP-based amplification system (LF-qLAMP) assisted by pH microelectrodes for the accurate and reliable diagnosis of SARS-CoV-2 during the first 3 minutes of the amplification reaction. We fabricated a simple and portable prototype for the isothermal incubation of samples and online monitoring of the LAMP amplification process. The system was comprised of a ad hoc heating system that was assembled with components acquired over the internet (a 12V eliminator, a 10 mL vessel with a silicon lead, a heating mat (electric resistance), and an Arduino UNO–based (Arduino, Italy) proportional integral derivative (PID) temperature controller (Figure 1 A-E). The evolution of the amplification process was followed using commercially available pH electrodes (inLab nano, from Mettler Toledo, Switzerland; and Lab Sen 241-3, from Aprea Instruments, available at Amazon.com). These electrodes were interfaced through an Arduino UNO microprocessor using a pH sensor module (Gaohou pH 0-14 detect sensor module, available at Amazon.com). The InLab microelectrode, with a tip diameter of 1.3 mm, can be easily placed inside a conventional 200 µL Eppendorf tube and enables the use of small reaction volumes (25–50 µL of reaction mix and sample). Alternatively, we have run online experiments using a lower cost electrode from Apera Instruments (~150 USD) with a tip diameter of 3 mm, which can also be used with similar measuring performance. In our experiments, the insertion of the electrode within the reaction mix impedes closure of the Eppendorf tube during incubation. To avoid evaporation during heating, two drops of mineral oil were dispensed into each Eppendorf tube before electrode placement to allow the development of a thin lid layer of mineral oil at the liquid surface of the reaction mix (Figure 1A,C).

Using LF-qLAMP that we have described, we incubated a series of LAMP reactions and monitored the progress of the amplification reaction by measuring the electric potential in the reaction mix in real time. We successfully discriminated between COVID-19(+) and COVID-19(-) samples in a set of synthetic samples and in a set of saliva samples spiked with synthetic SARS-CoV-2 genetic material. In brief, we first characterized the baseline corresponding to negative samples, where no primers were added. As expected, the signal associated with these samples was relatively steady (average standard deviations ~10.7; n=3). We then monitored the evolution of the LAMP reactions in the negative samples containing LAMP reaction mix and primers, but no synthetic SARS-CoV-2 genetic material. The shape of the progression of the electric potential in the negative samples was highly consistent. The profile, characteristic of negative samples (gray curve in Figure 1F), is distinguishable from that inherent to the specific amplification observed in samples that do contain SARS-CoV-2 genetic material. Figure 1F also shows the trajectories of the electric potential associated with positive samples containing different quantities of SARS-CoV-2 synthetic genetic material (color curves). The potential exhibits a distinctive, practically immediate, and sharp increase at the incubation conditions that is associated with a high initial rate of amplification. The initial slope of the potential curve is steeper in positive samples than in negative samples.

We observed that the discrimination between negative samples and positive samples with medium to high viral loads (i.e., at least 100 copies) is possible by comparison of the potential trajectories. However, samples containing a low copy number (i.e., less than 10 gene copies) cannot be distinguished from negative samples by only comparing trajectories. Indeed, we found that the behavior of the trajectory of the amplification reaction in positive and negative samples can be better discriminated at this first stage of the amplification. For instance, the value of the area under the curve or of ΔP versus t is distinguishably higher in positive samples than in negative samples during the first stage of the amplification. Therefore, we defined IP3, the integral of the potential with respect to time for the first 3 minutes of reaction, as an indicator of specific amplification of genetic material in the LAMP reactions (equation 1,2).

IP₃= ∫ΔP^* dt│_3min equation(1)

ΔP^* = (ΔP_i –ΔP_o) equation (2)

Here, ΔP_i is the value of electric potential measured at every sampling point and ΔP_o is the value of electric potential at the beginning of the experiment. Therefore ΔP^* is the increment in potential measured at each sampling event (i.e., in our case, every second) with respect to the initial value of electric potential in the reaction mix at the initial point of the amplification. Equation 3 provides a step by step approximation of IP₃.

IP₃ ~ Σ_{ito n} (ΔP*)_i (Δt) equation (3)

Here, n is the number of sampling points within the time frame from 0 to 3 minutes. Indeed, we found that IP₃ is a reliable and robust predictor that enables the consistent and rapid identification of samples containing (or not containing) SARS-CoV-2 genetic material. Figure 1G shows the evolution of the integral of IP₃ for the first 3 minutes of a set of LAMP reactions. Progressions corresponding to samples added with different quantities of synthetic genetic material (i.e., 10, 100, 1000, and 10000 copies of the N gene of SARS-CoV-2) were compared versus three independent repeats of negative samples (i.e., samples without SARS-CoV-2 genetic material). Negative samples (plotted in yellow) exhibit integrals significantly lower than those calculated from the analysis of positive samples.

Overall, we show that LF-qLAMP enables a straightforward discrimination between samples containing or not containing artificial SARS-CoV-2 genetic material in the range of 10 to 10,000 copies per 50 µL of reaction mix. We also spiked saliva samples with SARS-CoV-2 synthetic material and corroborated that the LAMP reaction can be successfully monitored in real time using microelectrodes in saliva samples as well. These results may have profound implications for the design of real-time and portable quantitative systems for the reliable detection of viral pathogens including SARS-CoV-2.

Figure 1. Experimental setup and results. (A) A micro pH electrode is inserted into the Loop-mediated Isothermal Amplification (LAMP) colorimetric reaction mix and (B) a thin layer of oil is added to avoid evaporation during incubation (C) in an oil bath. (D) Image of the actual system; a heating mat connected to an Arduino-based proportional integral derivative (PID) controller is used to control the temperature at 62.5 +/- 1.2 °C. (E) Actual image of the Arduino-based proportional integral derivative (PID) temperature controller and the online pH/electric potential monitoring system used in our experiments. (F) Comparison of the online potential signal in samples containing 10,000 (magenta), 1000 (red), 100 (pink), and 10 (yellow) copies of synthetic SARS‑CoV-2 genetic material. The average of four negative samples is depicted in gray. The inset shows a close-up of the first five minutes of the amplification reaction. (G) Plot of the integral of the differential of potential with respect to time for the first 3 minutes of amplification in samples containing different copy numbers of synthetic SARS-CoV-2 genetic material: 10,000 (dark blue), 1000 (brown), 100 (light blue), and 10 (red) copies of synthetic SARS-CoV-2 genetic material. Trends associated with negative samples are presented in blue, orange, green, and yellow.