Development, evaluation of the PNA RT-LAMP assay for rapid molecular detection of SARS-CoV-2

Dual-labeled PNA probe used RT-LAMP molecular rapid assay targeting SARS-CoV-2 ORF1ab and N genes was developed, and the analytical, clinical performances for detection of SARS-CoV-2 RNA extracted from clinical nasopharyngeal swab specimens were evaluated in this study. Data showed that this assay is highly specific for SARS-CoV-2, and the absolute detection limit is 1 genomic copy per microliter of viral RNA which can be considered to be comparable to gold-standard molecular diagnostic method real-time reverse transcriptase PCR. Both clinical sensitivity and specificity against a commercial real-time RT-PCR assay were determined as identical. In conclusion, the PNA RT-LAMP assay showed high analytical and clinical accuracy which are identical to real-time RT-PCR which has been routinely used for the detection of SARS-CoV-2.

The analytical specificity of the assay. The analytical specificity of the test was evaluated using 35 microorganisms shown in Table 2 which are frequently found in the human respiratory tract spiked in clinical negative nasopharyngeal (NP) swab specimen at concentrations of 10 6 CFU/mL or higher for bacteria and 10 5 pfu/mL or higher for viruses. In addition, RNA isolate from the SARS-CoV-2 negative human nasal wash was tested for specificity against the human normal nasal microflora. No detectable amplification curve was observed in FAM detection channel for SARS-CoV-2 ORF1ab and N genes, whereas the internal control RNase P in HEX detection channel did show 100% detection rate as expected in all three (3) test replicates for all organisms as well as for the nasal wash. Those results showed that the exclusivity of the assay against the microorganisms tested in this study is 100% (0% of false positivity) for both ORF1ab and N gene amplicon sets. Results are summarized in Table 2.

Clinical performance of the assay against commercial Real-time PCR test. A clinical evaluation
of the PNA RT-LAMP assay was performed that evaluating a total of 270 blinded clinical NP swab specimens including 70 SARS-CoV-2 positive and 200 negative individual, leftover, de-identified specimens collected in the Chungnam National University Hospital which were previously tested using commercially available FDA EUA authorized real-time PCR test targeting SARS-CoV-2 specific RdRp and E genes (PowerCheck 2019-nCoV realtime PCR kit, Kogene Biotech). Positive samples were divided into 2 groups based on the Ct values exhibited on the real-time PCR: 1) High positives: a total of 44 samples which both targets showed Ct ≤ 30; 2) Low positives: a total of 26  This result showed that the clinical performances of the 2 assays are identical, whereas the analysis time of RT-LAMP is quite shorter than the real-time PCR. The results of the clinical evaluation are summarized in Table 3.
Comparative sensitivity of the PNA RT-LAMP assay and other EUA authorized SARS-CoV-2 molecular tests. The analytical sensitivity of the PNA RT-LAMP assay was evaluated using RNA extracts from heat-inactivated SARS-CoV-2 (USA-WA1/2020, ZeptoMetrix, USA) at tenfold dilution series spiked in clinical negative NP swabs comparing with commercially available FDA EUA authorized real-time RT-PCR test Table 1. Summary of oligonucleotide sequence analysis. Sequences of all primer and probes showed no mismatch against the target queries (SARS-CoV-2 whole genomic sequences).  Fig. S1). For further evaluation, positive detection rates of PNA RT-LAMP and Colorimetric LAMP assays were evaluated using 15 clinical individual positive NP swabs including five (5) high positives which exhibited Ct values up to 30 cycles, five (5) moderate positives which exhibited Ct values between 31 to 34 cycles, and five (5) low positives which exhibited Ct values higher than 35 cycles for both SARS-CoV-2 ORF1ab and N genes that were previously identified using the Real-time RT PCR test (SS-9930, Seasun Biomaterials). The PNA RT-LAMP assay successfully detected all 15 samples from the three positive groups whereas the colorimetric LAMP test has missed 2 low positives which exhibited Ct values over 37 cycles for both ORF1ab and N genes on the real-time PCR assay (Table 5, Supplementary Fig. S2).
Those results show that the sensitivity of the PNA RT-LAMP assay is higher than the colorimetric LAMP assay and identical to the real-time PCR method even testing the low positive samples showed late amplification rates on the real-time PCR method.
Respiratory syncytial virus The device is fully portable and compatible with tablet computers with an easy-to-use operating system that can be applicable at POC testing ( Fig. 1). All 15 samples with various viral loads were detectable 100% within 15 min when testing with the same run condition as in the real-time PCR instrument (Table 5). This data shows that the PNA RT-LAMP assay can be applicable at POC testing even further evaluations with increasing clinical sample numbers are required.

Discussion
Here, we developed and evaluated analytical, clinical performances of PNA based RT-LAMP assay targeting ORF1ab and N genes of SARS-CoV-2. However, the LAMP has been known as having high and rapid amplification efficiency which makes its analytical sensitivity comparable to real-time PCR, high rate of false positivity while testing field clinical samples has been reported on the strength of the result interpretation method based on pH dependent colorimetric visualization 28,29 . Since the colorimetric display of traditional LAMP is based on the principle of color change reaction of pH indicators such as phenol red 30 , the result is significantly affected by remnants from nucleic acid extraction reagents contained in the clinical sample elutes as well as the LAMP reaction buffers and enzyme contents 31 . To overcome those issues, we applied dual-labeled PNA as a detection probe in LAMP reaction for fluorescence detection of the amplification product in a real-time manner. PNA has been reported that having superior specificity against its template nucleic acid with its neutrally charged peptide backbone nature which does not have a nonspecific binding affinity with minus charged natural phosphate backbone of the template nucleic acid 22,32 . The un-cleavable peptide backbone also reduces the risk of non-specific signal production as a result of thermal degradation during long-term incubation at elevated temperatures. We confirmed that the PNA RP-LAMP assay does not cross-react with non-target microorganisms at high concentrations. Also, both clinical specificity and sensitivity against real-time RT PCR assay showed 100% of    www.nature.com/scientificreports/ accuracies. The above results showed that PNA RT-LAMP assay employs high enough analytical specificity as well as the clinical performances are comparable to the gold standard real-time PCR method. Results of the sensitivity testing using tenfold dilution series of SARS-CoV-2 inactivated isolate demonstrated the potential of the PNA RT-LAMP test could detect ~ 1 copy of template RNA per microliter of the sample within approximately 12 min, however, the next tenfold dilution series contain approximately 0.1 copy of the SARS-CoV-2 genomic material could not be detected even in reaction up to 30 min. Practically, every tenfold dilution series in real-time PCR exhibit the Ct values of 3.3 apart while PNA RT-LAMP (in this study) exhibited one (1) minute lateness for every tenfold dilution series. According to these practices, the dilution series contain ~ 0.1 genomic copies could be detected on PNA RT-LAMP assay although it couldn't. Our assumption on this was the most amount of Bst polymerase has already been consumed during amplification of an internal control RNase P, and not enough concentration of active enzymes remained for amplification of the SARS-CoV-2 in extremely low concentration. Because RNase P showed amplification signals at around 15-20 Tt on the RT LAMP while the SARS-CoV-2 targets near to the test LOD showed Tt around 12-13 ( Supplementary Fig. S1). We observed that once RNase P amplification has started SARS-CoV-2 targets were not amplified. To test this assumption, we did further tests by increasing the Bst concentration up to tenfold, reducing the primer/probe concentrations of the internal control, however the expected result has not been obtained.
The final goal of this study was the development of a rapid molecular detection method that can be applicable at POC testing while having a comparable clinical performance with the gold standard real-time PCR method. We have confirmed that the PNA RT-LAMP assay can detect low positive samples contain a few copies of target RNA that exhibiting Ct values > 35 on the real-time PCR although traditional LAMP has missed those samples. This result was reproducible on a both benchtop real-time PCR detection system and a portable isothermal amplification device.
However, the PNA RT-LAMP assay developed in this study is currently applicable with RNA extracts from clinical NP swabs, we have been working on optimization of the assay on use of saliva and direct NP swabs without an additional sample preparation step which is more suitable at POC testing and in low-resource settings.

Materials used in this study. 2 × LAMP Master Mix contains M-MLV Reverse Transcriptase and Bst
DNA polymerase were purchased from Elpis-Biotech, Korea. Oligonucleotides and dual labeled PNA probes were synthesized in Cosmo Genetech, Korea, and Panagene, Korea respectively. Heat inactivated SARS-CoV-2 isolate USA-WA1/2020 was purchased from ZeptoMetrix, USA. Strains of the microorganisms used in the specificity testing were purchased from Korean Bank of Pathogenic Viruses and the National Culture Collection for Pathogens, Korea. Sequence analysis. A total of 12 primer and 2 PNA probe sequences were analyzed against 56,303 SARS-CoV-2 genomic sequences contain whole genome information downloaded from GenBank (https:// www. ncbi. nlm. nih. gov/) on CLC Main Workbench 9.5.2 with molecular biology tool "Find binding sites and create fragment".
Sample preparation. For analytical sensitivity study, RNAs were purified from 300 µL of NP swabs spiked in tenfold dilutions series of SARS-CoV-2 inactivated isolate (ZeptoMetrix, USA-WA1/2020) using TOP Viral DNA/RNA extraction kit (Seasun Biomaterials, SS-1300) according to the manufacturer's instructions and eluted in 30 µL of elution buffer included in the kit. Genomic copies per µL were previously quantified using NanoDrop values of the nucleic acid extract of undiluted SARS-CoV-2 isolate as a formula in below: The size of the SARS-CoV-2 reference genome (NCBI Reference Sequence: NC_045512.2) is assumed to be 29,903 bp ss-RNA was used in the calculations. For comparative clinical sensitivity and clinical evaluation, 300 µL clinical positive NP swab specimens were processed using PANAMAX48 viral DNA/RNA extraction kit (Panagene, PNAK 1001) on PANAMAX nucleic acid automated extractor following the manufacturer's instructions. Each RNA isolate was used immediately after the extraction.
PNA RT-LAMP amplification and detection. PNA RT-LAMP test was performed with a total of 30 µL of total reaction volume using 15 µL of reaction buffer, 1 µL of enzyme mix, 4 µL of the reaction mix, and 10 µL of template RNA on CFX-96 real-time PCR detection system or SMARTAMP portable isothermal amplifier with the run condition of 60 degrees Celsius for 30 min with fluorescence signal collection at every 1 min. Samples that exhibited positive amplification signals within 30 min on FAM fluorescence channel in at least one reaction well was defined as positives. Samples that did not produce positive FAM signal while HEX detection channel regarding endogenous quality control RNase P produced amplification curve were defined as negatives.
Clinical evaluation. RNAs were extracted from 200 µL of individual, leftover, de-identified nasopharyngeal swab specimens collected in the Chungnam National University Hospital previously tested with FDA authorized under EUA SARS-CoV-2 real-time PCR assay PowerCheck 2019-nCoV real-time PCR kit (Kogene Biotech, Korea) targeting SARS-CoV-2 specific RdRp and E genes. 10 µL of RNA extracts were tested for each reaction mixtures of the PNA RT-LAMP assay in a blinded manner on the CFX-96 real-time PCR detection system. Clinical accuracies were calculated by using the standard method in the base of CI 95% 33 .