Multiplexed Reverse Transcription Loop-Mediated Isothermal Amplification Coupled with a Nucleic Acid-Based Lateral Flow Dipstick as a Rapid Diagnostic Method to Detect SARS-CoV-2

Due to the high reproduction rate of COVID-19, it is important to identify and isolate infected patients at the early stages of infection. The limitations of current diagnostic methods are speed, cost, and accuracy. Furthermore, new viral variants have emerged with higher rates of infectivity and mortality, many with mutations at various primer binding sites, which may evade detection via conventional PCR kits. Therefore, a rapid method that is sensitive, specific, and cost-effective is needed for a point-of-care molecular test. Accordingly, we developed a rapid molecular SARS-CoV-2 detection kit with high specificity and sensitivity, RT-PCR, taking advantage of the loop-mediated isothermal amplification (LAMP) technique. Four sets of six primers were designed based on conserved regions of the SARS-CoV-2 genome: two outer, two inner and two loop primers. Using the optimized protocol, SARS-CoV-2 genes were detected as quickly as 10 min but were most sensitive at 30 min, detecting as little as 100 copies of template DNA. We then coupled the RT-LAMP with a lateral flow dipstick (LFD) for multiplex detection. The LFD could detect two genic amplifications on a single strip, making it suitable for multiplexed detection. The development of a multiplexed RT-LAMP-LFD reaction on crude VTM samples would be suitable for the point-of-care diagnosis of COVID-19 in diagnostic laboratories as well as in private homes.


Introduction
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was first identified in December 2019 during an outbreak originating from a market in Wuhan, China [1]. This novel coronavirus is the cause of coronavirus disease 2019  and is only the seventh known coronavirus that has been infectious to humans [2,3]. The rapid spread of COVID-19 has resulted in the World Health Organization (WHO) declaring a global pandemic. The rising number of infected individuals has caused the collapse of hospital systems as they cannot be accommodated [4,5]. This led to countries deciding to close their borders to manage the spread of the disease. However, with many countries adopting national vaccination policies, international travel opened again. Therefore, variants from all around the world were allowed to spread worldwide [6,7]. However, due to the rapidly mutating nature of the viral RNA, many new variants emerged with the ability to evade vaccine-induced immunity. Mutations on common PCR primer binding sites cause the misdiagnosis of COVID-19 via RT-PCR tests [8].
The spread of viral disease can be controlled by quickly identifying and isolating infected individuals. Therefore, it is important to have rapid and accurate diagnostic assays. Throughout the pandemic, various techniques were described and employed for the mass screening of COVID-19. This included real-time reverse transcription polymerase chain For each gene, a set of primers was designed, consisting of two inner primers (FIP and BIP), two outer primers (F3 and B3) and two loop primers (LF and LB). The forward inner primer (FIP) was designed by a combination of the complementary sequence of F1 (F1c) and F2, linked by a poly-T linker. Additionally, and similarly, the backward inner primer (BIP) was a combination of B1c and B2 with a poly-T linker as well (Table 1). Table 1. The four sets of primers were designed for different gene targets consisting of the inner, outer and loop primers, along with the size of the target amplicon. The "automatic judgement" feature and default parameters from Primer Explorer Version 5 (http://primerexplorer.jp/lampv5e/index.html, accessed on 3 January 2021) were used to design each set of primers.

Primer Set
Target Gene Primer Name Sequence (

Preparation of DNA Template
The target N, M and E genic regions were amplified via PCR using the outer primers of each primer set. The PCR reactions were as described [30] with modifications. The reaction of 25 µL consisted of a 1X PCR buffer, 1.5 mM MgCl 2 , 0.2 mM dNTP mix (Promega, Madison, WI, USA), 20 pmol of each outer primer and 0.2 U Taq DNA polymerase (Promega, Madison, WI, USA). The thermocycler protocol included an initial denaturation of 95 • C for 5 min, followed by 40 amplification cycles of 95 • C for 30 s, 55 • C for 30 s and 72 • C for 30 s, and a final extension at 72 • C for 10 min. PCR products were viewed on agarose gel, which were then excised and purified using a QIAquick Gel Extraction Kit (Qiagen, Germantown, MD, USA).

Recombinant Plasmid Construction for Positive Control for PCR and LAMP Analysis
The purified PCR products were each ligated into a pJET.2 Blunt Cloning Vector in accordance with the CloneJET Blunt End PCR Cloning Kit (Thermo Fisher Scientific, Waltham, MA, USA) instructions. The recombinant plasmids were transformed into 50 µL of a chemically competent E. coli strain TOP10, which were then cultured on an ampicillin-Luria Bertani (LB) agar plate at 37 • C for 16 h. Single colonies were selected and grown in an LB broth at 37 • C for 16 h, followed by plasmid extraction using a GeneJET Plasmid Miniprep Kit (Thermo Fisher Scientific, USA) according to the manufacturer's protocol. The plasmid was then verified through DNA sequencing.

RNA Synthesis for RT-LAMP Protocol Verification
Synthetic RNA for each gene was synthesized from their respective plasmid using a HiScribe ® T7 Quick High Yield RNA Synthesis Kit (New England Biolab, Hitchin, UK). The protocol was conducted as per the manufacturer's instructions. Synthetic RNA was purified using the lithium chloride protocol [31].

Optimization of LAMP and RT-LAMP Reaction Condition with UV Analysis
Initially, the LAMP assay was conducted modified based on a previously described protocol with a 30 µL reaction mixture containing a 1X Isothermal Amplification Buffer II (New England Biolab, UK), 0.4 M of Betaine (Sigma-Aldrich, St. Louis, MO, USA), 8 mM of MgSO 4 , a 1.4 mM dNTP mix (Promega, USA), 10 U Bst of 3.0 DNA polymerase (New England Biolab, UK), 32 pmol of each inner primer, 8 pmol of each outer primer, 32 pmol of each loop primer, followed by 2 µL of template DNA or RNA [32]. Optimizations were performed by testing different ratios of outer, inner and loop primers, as well as the working concentrations of MgSO 4 .

Sensitivity Test for the Detection of SARS-CoV-2 Nucleocapsid, Envelope and Membrane Genes Using End Point-PCR and Quantitative PCR
The recombinant plasmid DNA with each gene was serially diluted 10-fold to achieve 10 8 copies to one copy number [30]. Both the endpoint and quantitative PCR were conducted on each dilution of recombinant plasmid using the outer primers stated in Table 1. For the endpoint PCR, the protocol was as mentioned in Section 2.2. Subsequently, the amplification products were visualized on 2.0% agarose gel electrophoresis, stained with ethidium bromide, and observed under UV light. In addition, quantitative PCR (qPCR) was conducted with the addition of an SYBR Green stain using a real-time PCR machine (Biorad CFX96, Hercules, CA, USA).

Sensitivity Test for the Detection of SARS-CoV-2 Nucleocapsid, Envelope and Membrane Genes Using LAMP-UV, LAMP-SYBR Green and LAMP-LFD Analyses
The optimized LAMP protocol was conducted on the same set of 10-fold serial dilution positive control recombinant plasmids for 30 min at 65 • C UV, and SYBR Green and LFD analyses were used to visualize the amplification products of the LAMP assays. The 1.5% agarose gel electrophoresis was conducted on amplification products, stained in ethidium bromide and observed under UV conditions. Colorimetry using SYBR Green was conducted by the addition of 2 µL of 1:10, which was diluted SYBR Green I nucleic acid gel stain to all tubes containing LAMP products, and observations on the color changes were immediate [30]. As for LFD, the LAMP protocol was conducted using the primers stated in Table 1, where inner primers were labeled with specific antigens for LFD detection. The PCRD Flex Nucleic Acid-Based Immunoassay (Abingdon Health, York, UK) was used according to the manufacturer's instructions.

Specificity Test of LAMP Assay
The specificity of each set of LAMP primers was evaluated in silico and in vitro. The sequences of closely related coronavirus, regardless of the hosts, were downloaded from Genbank (AY613950.1, KY352407.1, KY417144.1, NC_001451.1, NC_002645.1, NC_004718.3, NC_005831.2, NC_006213.1, NC_006577.2, NC_038294.1, NC_048213.1) and aligned on MEGA X [33]. Mismatches were observed on sequences of the primer binding regions on viral sequences that did not belong to SARS-CoV-2. For comparison, the synthetic DNA of SARS-CoV-1 and MERS-CoV genes, as well as the cDNA of Infectious Bronchitis Virus (IBV), were used for in vitro specificity tests as these were the ones available. The LAMP assays were performed at 65 • C for 30 min with 20 ng of each control.

LAMP and RT-LAMP Optimization Using UV Analyses
The optimization of the LAMP protocol was performed with the purpose of shortening the time required for amplification without compromising on sensitivity. Positive LAMP reactions were indicated by the formation of ladder-like bands after performing agarose gel electrophoresis, stained with ethidium bromide, and these were observed under UV light. Four ratios of Outer:Inner:Loop primers were used, 1:1:1, 2:1:2, 4:1:4 and 8:1:8, to perform the LAMP assay on the recombinant plasmid DNA. Thus, we observed that a 4:1:4 primer ratio produced DNA amplification with every set of primers ( Figure 1A). Subsequent to identifying the right ratio of the primers in use, the same protocol was repeated against synthetic RNA to test for the single enzyme one-step RT-LAMP protocol. To further optimize the reaction time, 4 mM, 6 mM, 8 mM and 10 mM of MgSO 4 was used for the assay and tested on an incubation time of 5 min intervals (5, 10, 15, 20, 25 and 30), which showed that 8 mM had the fastest reaction speed; this allowed for amplification without having amplifications at the non-template control (NTC). The results showed that the addition of 8 mM of MgSO 4 was able to produce an amplification within 10 min ( Figure 1B). This was also the fastest reaction speed in comparison to the other MgSO 4 concentrations.
LFD detection. The PCRD Flex Nucleic Acid-Based Immunoassay (Abingdon Health, York, UK) was used according to the manufacturer's instructions.

Specificity Test of LAMP Assay
The specificity of each set of LAMP primers was evaluated in silico and in vitro. The sequences of closely related coronavirus, regardless of the hosts, were downloaded from Genbank (AY613950.1, KY352407.1, KY417144.1, NC_001451.1, NC_002645.1, NC_004718.3, NC_005831.2, NC_006213.1, NC_006577.2, NC_038294.1, NC_048213.1) and aligned on MEGA X [33]. Mismatches were observed on sequences of the primer binding regions on viral sequences that did not belong to SARS-CoV-2. For comparison, the synthetic DNA of SARS-CoV-1 and MERS-CoV genes, as well as the cDNA of Infectious Bronchitis Virus (IBV), were used for in vitro specificity tests as these were the ones available. The LAMP assays were performed at 65 °C for 30 min with 20 ng of each control.

LAMP and RT-LAMP Optimization using UV Analyses
The optimization of the LAMP protocol was performed with the purpose of shortening the time required for amplification without compromising on sensitivity. Positive LAMP reactions were indicated by the formation of ladder-like bands after performing agarose gel electrophoresis, stained with ethidium bromide, and these were observed under UV light. Four ratios of Outer:Inner:Loop primers were used, 1:1:1, 2:1:2, 4:1:4 and 8:1:8, to perform the LAMP assay on the recombinant plasmid DNA. Thus, we observed that a 4:1:4 primer ratio produced DNA amplification with every set of primers ( Figure  1A). Subsequent to identifying the right ratio of the primers in use, the same protocol was repeated against synthetic RNA to test for the single enzyme one-step RT-LAMP protocol. To further optimize the reaction time, 4 mM, 6 mM, 8 mM and 10 mM of MgSO4 was used for the assay and tested on an incubation time of 5 min intervals (5, 10, 15, 20, 25 and 30), which showed that 8 mM had the fastest reaction speed; this allowed for amplification without having amplifications at the non-template control (NTC). The results showed that the addition of 8 mM of MgSO4 was able to produce an amplification within 10 min ( Figure 1B). This was also the fastest reaction speed in comparison to the other MgSO4 concentrations.

Sensitivity Test of End Point-PCR, Quantitative PCR, LAMP-UV and LAMP-SYBR for the Detection of SARS-CoV-2 Genes
The sensitivity test was successfully conducted using a set of SARS-CoV-2-positive control plasmids which were serially diluted 10-fold, from 10 8 copies to just one copy. The visualization methods of the PCR (end-point and quantitative) and LAMP (UV, SYBR Green and LFD) were observed to not have had any effect on the sensitivity. The sensitivity of PCR and the optimized LAMP protocol varied between the different target gene types. However, ultimately, the optimized LAMP protocol proved to be equally sensitive and, on certain genes, more sensitive when compared to the PCR.
For the detection of the nucleocapsid (N) gene, two sets of LAMP primers were designed to target two different regions, namely N1 and N2. However, following the sensitivity test, it was revealed that the sensitivity of both primers set across all assays was identical. As revealed in Figure 2A,B, the detection limit for both end-points and quantitative PCR was 10 2 copies of plasmids using an N1 primer set, with the corresponding LAMP-UV, LAMP-SYBR Green, and LAMP-LFD assays providing a detection limit of 10 2 copies as well ( Figure 2D,F,H). Conversely, for the N2 primer set, the results showed that the detection limit across all assays (quantitative PCR, end-point PCR, LAMP-UV, LAMP-SYBR Green and LAMP-LFD) was 10 2 copies, respectively (  The results for the sensitivity test of the M gene showed that the LAMP protocol was able to detect as little as 10 3 copies of plasmids, as visualized with both UV and colorimetry by SYBR Green staining and LFD ( Figure 3C-E). This is similar to that of the PCR tests, which were able to detect 10 3 copies, as indicated by the graph and agarose gel photo in Figure 3A,B.  The sensitivity test showed that the detection limit of LAMP with regard to Envelope (E) gene primers was lower than that of the PCR. It can be observed that both qPCR and end-point PCR had a detection limit of 10 4 copies ( Figure 4A,B). However, as depicted in Figure 4C-E, the corresponding photos showed that the detection limit of LAMP was 10-fold lower at 10 3 copies.

Specificity Test of LAMP-UV, LAMP-SYBR Green and LAMP-LFD
Prior to the development of LAMP, in silico screening was performed to design primers not only with conserved regions within SARS-CoV-2 variants but also with a low

Specificity Test of LAMP-UV, LAMP-SYBR Green and LAMP-LFD
Prior to the development of LAMP, in silico screening was performed to design primers not only with conserved regions within SARS-CoV-2 variants but also with a low affinity toward the genes of closely related viral species. Specificity tests were successfully conducted on control plasmids with the gene inserts of various coronaviruses which had close genetic ties to SARS-CoV-2 and were available on hand, SARS-CoV-1, MERS-CoV and IBV.
Positive results indicated by ladder-like bands were observed only against the plasmids containing genes of SARS-CoV-2 ( Figure 5). The DNA samples with genes of other coronaviruses produced negative results, as shown by the absence of ladder-like bands. Thus, the results from this specificity test for each set of LAMP primers implied that it was specific only to SARS-CoV-2. bands. Thus, the results from this specificity test for each set of LAMP primers implied that it was specific only to SARS-CoV-2.

Visualization of Multiplexed LAMP on a Single Strip of LFD
Four different combinations were made from the designed primers ( Figure 6). This was because the differentiating labels were only digoxigenin (N1 and E) and fluorescein (N2 and M). Therefore, primers with the same labels could not be used in combination. As observed, in a multiplexed reaction, two test lines were observed on a single LFD.

Visualization of Multiplexed LAMP on a Single Strip of LFD
Four different combinations were made from the designed primers ( Figure 6). This was because the differentiating labels were only digoxigenin (N1 and E) and fluorescein (N2 and M). Therefore, primers with the same labels could not be used in combination. As observed, in a multiplexed reaction, two test lines were observed on a single LFD.

Visualization of Multiplexed LAMP on a Single Strip of LFD
Four different combinations were made from the designed primers ( Figure 6). This was because the differentiating labels were only digoxigenin (N1 and E) and fluorescein (N2 and M). Therefore, primers with the same labels could not be used in combination. As observed, in a multiplexed reaction, two test lines were observed on a single LFD.

Discussion
Due to the expensive, time-consuming, and tedious nature of RT-PCR tests, healthcare professionals are opting for the less reliable but fast antigen-based rapid detection tests. Therefore, a simple and fast yet accurate detection method is needed for the detection of SARS-CoV-2 at early stages of infection. Taking this into consideration, a multiplexed LAMP-based approach was optimized as fast but also specific and sensitive.
The genes selected for this study were the N, M and E genes (Table 1). This was due to the high read coverage among coronavirus genes when RNA was sequenced from cultured tissues infected with HCoV-299E coronavirus [34]. Others reported LAMP-based SARS-CoV-2 detection using N, RdRp, S, ORF1ab, ORF8 and E genes [35][36][37][38][39][40][41]. In addition to that, through GISAID, we obtained genomic sequences for screening with the conserved regions within these genes. This was to ensure that the primers designed were able to detect SARS-CoV-2 across all variants. Therefore, false negative diagnoses were avoided. It was well described that to avoid false negatives, amplicons should be selected from conserved regions or multiple regions at the same time [42]. This was especially difficult as the viral genome is constantly mutating. In addition, even the primers from the gold standard, RT-PCR, were found to produce false negatives [8]. This was due to the ever so rapid occurrence of mutations at the common commercially used primer binding site.
A wide range of genomic sequences of SARS-CoV-2 was downloaded from the GI-SAID database (Appendix A Table A1), along with the reference genome from GenBank (NC_045512.2), as mentioned in Section 2.1. Using multiple alignment tools, we managed to obtain two conserved regions in the N gene and a region from the M and E genes, respectively, ranging from 200 to 220 bp in length. Six primers were designed for each set of primers, a pair for the inner and outer primers, respectively, as well as a pair of loop primers. Loop primers were found to increase the amplification speed of LAMP [19,30,43] as well as increase its specificity [44].
The LAMP protocol described In this study was able to detect SARS-CoV-2 control plasmids and synthetic RNA in as little as 10 min (Figure 1). With the use of Bst, 3.0 DNA polymerase (NEB), the time required for a one-step RT-LAMP reaction was reduced. This is due to the high reverse transcription activity of Bst 3.0 DNA polymerase: a single enzyme reaction could be performed using a single temperature (65 • C) [45]. Therefore, simple apparatus could be used to conduct the test (for example, a water bath or heat block). For optimum sensitivity, 30 min of incubation is ideal. Other LAMP-based detection methods have been described using various temperatures in the range of 60-65 • C with an extensive incubation time of up to 60 min. Two-step RT-LAMP protocols, however, take away the rapidness, thus making it unsuitable for point-of-care tests.
Plasmid DNA and synthetic RNA were used for the in vitro testing in the development of this multiplexed RT-LAMP-based LFD. The approach to amplify the genes using the outer primers was subsequently cloned for use as a template for the test, as previously described by others [30,46]. This provided a more accurate quantitative approach compared to repeated RNA extraction from SARS-CoV-2 virions. Furthermore, repeated exposure to SARS-CoV-2 brought a risk of infection during the experimentations.
The colorimetry LAMP visualization method used in this study was the addition of SYBR Green. Color changes were observed to indicate positive (green) and negative (orange) results ( Figure 1A Bottom). However, as seen in lane 6 ( Figures 3D and 4D, respectively), the color changes we not as vivid. SYBR Green is a DNA intercalating dye with double-stranded DNA and showed color changes from orange to green [47,48]. This method is rapid in the sense that positive detection was observed with the naked eye without the need for agarose gel electrophoresis. However, at a low copy number of templates, the intermediate colors varied from individual observers as observed by others [49,50]. Therefore, colorimetric visualization is not the best option for LAMP. Other colorimetry dyes were documented for LAMP visualizations with varying success rates. This included phenol red [51,52], leuco crystal violet [53], calcein [54], and hydroxy-naphthol blue [55].
Another approach to minimize errors from calorimetry dyes is through the use of a lateral flow dipstick (LFD) in combination with the RT-LAMP, as described in this study. These clear distinct lines on LFD indicate the successful amplification of the N1, N2, M and E genes by LAMP. We used a carbon nanoparticle-based LFD with two distinct test lines targeting Biotin-fluorescein and Biotin-DIG complexes, respectively. The LFD presents tremendous prospects for point-of-care testing because it is straightforward, rapid and visual [56,57]. The genes amplified via LAMP with the primers listed in Table 1 exponentially increased the number of amplicons carrying Biotin-FAM or Biotin-DIG. Through capillary actions, the amplicons migrated through the LFD, where they bound to anti-biotin antibodies bonded with carbon nanoparticles, which were then mobilized [58]. Migrating amplicons carrying the carbon nanoparticle were immobilized at the test lines (coated with neutravidin or anti-DIG antibody). Thus, leaving the black lines observed on the test lines, a control line was formed due to the excess amplicons and/or biotinlabeled primers (no LAMP reaction) that were immobilized at the control line by unspecific antibodies. The use of RT-LAMP coupled with LFD reduced the need for potentially harmful carcinogens (for example, ethidium bromide stain in agarose gel electrophoresis), increased the accuracy (eliminates the use of colorimetric dyes), and even the use of expensive machinery (real-time PCR machine) [59][60][61].
The detection limit using N1, N2 and M primer sets was equal to that of PCR detection (Figures 2 and 3), with the exception of primers for the E gene ( Figure 4). This was because the sensitivity test for the E gene indicated that LAMP was more sensitive than the PCR. Furthermore, our sensitivity test was conducted with just 30 min of incubation in an isothermal condition (65 • C). With regard to LAMP, several publications were reportedly able to detect lower concentrations of the template material but required more time or the conduction of the reverse transcription process separately [62][63][64][65]. We used the isothermal enzyme Bst 3.0 DNA polymerase (NEB) for both reverse transcription and LAMP. The Bst 3.0 DNA polymerase has a dual activity of reverse transcriptase and polymerase in a single temperature incubation [32]. The use of this single enzyme here was efficient and made the one-step RT-LAMP more economical.
It is also important to consider the fact that the detection of low copies of SARS-CoV-2 did not indicate that a person was currently infected with COVID-19 [66]. A higher viral load was required for the body to show symptoms and severity increase with viral load [67]. It has previously been reported that positive qPCR results were observed weeks after infection, the majority of which had a high C t value, which was an indication of a low viral load [68][69][70][71].
For the specificity test, the designed primers were indeed specific only toward the intended targets. Only the SARS-CoV-2 positive control plasmids and synthetic RNA revealed ladder-like bands, and none of the negative controls (SARS-CoV-1, MERS-CoV and IBV) were amplified ( Figure 5). However, the control plasmids used in this study were limited due to the difficulty in procuring the genetic material of infectious pathogens. A more comprehensive screening was needed to further validate the results. Therefore, the in silico screening conducted earlier during primer design played an important role in the specificity of these primers.
The detection of two genes with two test lines on a single LFD is extremely relevant in diagnosis as this reduces the chances of false negatives ( Figure 6). LFD is useful in multiplex LAMP (mLAMP) reactions as the amplification products could not be differentiated when using gel electrophoresis or visualized with SYBR Green. Therefore, a binary (positive or negative) interpretation of results may be flawed when it comes to multiplexed reactions. The different labels modified on the inner primers play an important role in LFD detection. A single tube assay along with the LFD is convenient for diagnosing COVID-19. Typically, two genes are required to further increase the specificity by avoiding false negatives. This is especially useful as the risk of mutation at one of the primer binding sites is everpresent. Others have described multiplexed molecular amplification when diagnosing SARS-CoV-2, notably using RT-PCR. In RT-PCR, these amplicons could be differentiated using different labels on the probes used. A multiplexed RT-PCR-based protocol [72] was recommended by WHO, which utilized the RdRp and E genes. This highly accurate multiplex RT-PCR approach, which has been utilized in many commercial kits, is still time-consuming and expensive.
While we acknowledge the need for clinical testing, our data show that the RT-LAMP-LFD protocol described here was accurate and rapid. Efforts are currently ongoing to utilize clinical samples to provide a more comprehensive examination of the RT-LAMP-LFD protocol developed from this study. Ongoing and not yet presented results on limited positive samples on hand revealed great promises. Due to varying geographical outbreaks, it was not possible to collect clinical samples of all the variants. Therefore, we decided it would be of interest to other researchers to release LAMP primers and protocols. Therefore, they could begin to test the samples available to them.
To further validate the protocol, trials using samples collected from swabs or saliva could be encouraged. Trials using crude saliva could generate high usefulness in point-ofcare settings, preferably with variability in the variants. The convenience of direct testing from crude samples (saliva [73,74] and nasopharyngeal swabs [75]) was documented as before. Additionally, Bst 3.0 DNA polymerase was robust and capable of sustaining its activities in the presence of inhibitors [76]. This was especially useful for saliva and other samples which are known to carry amplification inhibitors [77]. This would ensure that the RT-LAMP-LFD was useful in a clinical setting.

Conclusions
The current protocol that we have developed is indeed rapid, sensitive, and specific in regard to the study procedures. Therefore, the development of a multiplexed RT-LAMP-LFD reaction on crude VTM samples would be suitable for point-of-care diagnosis of COVID-19 in diagnostic laboratories as well as in private homes.  Acknowledgments: We thank Zarina Amin for providing assistance with the funding and sample material.

Conflicts of Interest:
The authors declare no conflict of interest.
Appendix A Table A1. Full list of genome sequences of SARS-CoV-2 collected from the GISAID database, together with the countries of origin of samples from which the sequences were derived.