Genomic versus sub-genomic RT-LAMP assaytarget validation.The performance of fluorescent and colorimetric end-pointRT-LAMP reactionstargeting the genomicOrf1a andsub-genomic N and E generegions were tested solo, in duplex (N+E) and in multiplex combination (Orf1a+N+E) against a single concentration oftwo synthetic positive control RNAs (Figure2). Fluorescent end-point data presented in Figure 2A demonstrates equivalent amplification of theTwist positive controls (T1 and T2)with no statistically significant difference. Reactions targeting the Orf1a were fastest to exceed the threshold(17 ± 0.17 and 16.7 ± 0.17 minutes,T1 andT2 respectively), followed by reactions targeting the E gene (12.1 ± 0.09 and 12.4 ± 0.10 minutes) and finally the N gene (17 ± 0.17 and 16.7 ± 0.17 minutes). Dual N+E reactions exceeded the amplification threshold at a mid-point between N and E alone (14.3 ± 0.10 and 14.2 ± 0.08 minutes). Finally, addition of the Orf1a primer set to the dual N+E reaction further enhanced dual velocity (13.4 ± 0.22 and 13.0 ± 0.21 minutes). A lack of cross-reactivity of all SARS-CoV-2 targets was confirmed by a lack of amplification in negative control wells (related human coronavirus OC43, non-related influenza A and total human RNA). Presence of viable RNA within these samples was confirmed by control ACTB amplification (8.9 ± 0.25, 13.0 ± 0.19 and 9.5 ± 0.09respectively).The ACTB total RNA control reaction failed to amplify the positive RNA controls consistent with these being synthetic material. Finally, no template controls (NTC) confirmed the absence of non-specific amplification in any reactions.Fluorescent data is summarised in Table 1. Summary melt curve data was also collated for each primer set (Orf1a 82.1 ± 0.058, N+E 89.4 ± 0.060, ACTB 92.7 ± 0.050), and used as an amplification product specificity check in subsequent analyses of patient samples. End-point colorimetric detection (Figure 2B) consistent with fluorescent findings, demonstrated clear amplification in under 30 minutes with a colour change from pink to yellow for all primers (Orf1a, N and E) and primer combinations (N+E & Orf1a+N+E) against positive RNA controls T1 andT2. SARS-CoV-2 targeting reactionsfailed to amplify negative controls (OC43,influenzaA and total human RNA), whilst the ACTB internal control reaction confirmed the presence of viable template RNA. ACTBinternal control reaction failed to amplify synthetic material (as expected) and a NTC confirmed the absence of non-specific amplification in any reactions.
Table 1. RT-LAMP primer investigation
SARS-CoV-2 target:
|
TTP (minutes):
|
Twist 1 (SKU 102019)
|
Twist 2 (SKU 102024)
|
OC43
|
Influenza A
|
Human RNA
|
NTC
|
Orf1a
|
9.1 ± 0.07
|
8.7 ± 0.16
|
na
|
na
|
na
|
na
|
N
|
17 ± 0.17
|
16.7 ± 0.17
|
na
|
na
|
na
|
na
|
E
|
12.1 ± 0.09
|
12.4 ± 0.10
|
na
|
na
|
na
|
na
|
N+E
|
14.3 ± 0.10
|
14.2 ± 0.08
|
na
|
na
|
na
|
na
|
Orf1a+N+E
|
13.4 ± 0.22
|
13.0 ± 0.21
|
na
|
na
|
na
|
na
|
ACTB
|
na
|
na
|
13.0 ± 0.19
|
9.5 ± 0.09
|
8.9 ± 0.25
|
na
|
Data summary of an average of 2 independent experiments each performed in quadruplicate, presented as mean TTP ± S.E.M. No amplification is noted as ‘na’.
Limit of detection of fluorescent and colorimetric end-point RT-LAMP reactions targeting the Orf1a, dual N+E gene and multiplex Orf1a+N+E genes of SARS-CoV-2: Parallel fluorescent and colorimetric end-point reactions were performed against synthetic RNA (Twist control 2)serially diluted to 10,000, 1,000, 500, 100, 50 and 10 copies of viral sequence. Representative fluorescent end-point amplification curves and linear regression analysis of primer setsOrf1a, N+E duplex and multiplex Orf1a+N+E targetare presented in Figure 3 with summary data presented in Table 2.RT-LAMP targeting Orf1a was the fastest to exceed amplification threshold at 8.9 ± 0.12 minutes (10,000 viral RNA copies),capable of reproducibly detecting500 copies synthetic viral RNA. Lower viral loads down to 10 copies were detectable within 20 minutes although not reproducibly. RT-LAMP dual targetingN+E and multiplex Orf1a+N+E were~5 minutes slower to exceed amplification threshold for equivalent viral loads,analogously capable of reproducibly detecting 500 copies of synthetic viral RNA. All assays demonstrated good linearity and efficiency, 102.1%, 99.5% and 105.5% respectively.Colorimetric reactions (Additional file 1) augmented with EvaGreen intercalating dye and guanidine hydrochloride were equally able to reproducibly detect 500 viral RNA copies, exceeding the amplification threshold slightly faster that their fluorescent counterpart (2.7, 5.6 and 4.5 minutes faster for Orf1a, N+E and Orf1a+N+E respectively for 500 viral copies). All reactions that exceeded the amplification threshold (indicated byintercalating dye associated TTP) also showed a clear visual colorimetric change from pink to yellow in under 30 minutes.
Table 2. Limit of detection of fluorescent end-point RT-LAMP reactions
Synthetic positive control RNA
|
Fluorescent RT-LAMP
Mean TTP (min) ± SEM (N)
|
Copies / reaction
|
Orf1a
|
N+E
|
Orf1a+N+E
|
10,000
|
8.9 ± 0.12 (8/8)
|
14.4 ± 0.16 (8/8)
|
14.1 ± 0.12 (8/8)
|
1,000
|
11.6 ± 0.49 (8/8)
|
17.1 ± 0.77 (7/8)
|
17.2 ± 0.50 (8/8)
|
500
|
13.3 ± 0.65 (8/8)
|
21.4 ± 2.51 (8/8)
|
17.2 ± 0.29 (8/8)
|
100
|
14.1 ± 0.0 (1/8)
|
22.0 ± 1.44 (5/8)
|
23.2 ± 1.66 (5/8)
|
50
|
16.4 ± 3.91 (3/8)
|
na
|
29.2 ± 0.0 (1/8)
|
10
|
11.6 ± 0.0 (1/8)
|
na
|
21.1 ± 0.85 (2/8)
|
NTC
|
na
|
na
|
na
|
Reactions targeting the Orf1a, dual N+E gene and multiplex Orf1a+N+E of SARS-CoV-2 using a serially diluted Twist Bioscience synthetic positive control RNA.Data summary is an average of 2 independent experiments, performed in quadruplicate and presented as mean TTP ± S.E.M. Numeration in parentheses indicates the number of repeat reactions achieving the amplification threshold required to report a TTP. No amplification (‘na’).
Diagnostic validation of fluorescent end-point RT-LAMP assays targeting Orf1a and dual N+E gene regions.Residual RNA extracted from patient ON swab samples originally processed through the diagnosticRT-PCR service in theUHL Trustwere processedthrough parallel RT-LAMP reactions targeting the Orf1a, dual N+E gene and ACTB internal control.Two PCR platforms wereassessed:the ABI StepOnePlus PCR platform (114 UHL positive samples & 88 UHL negative samples) andthe QiagenRotor-Gene Q(39 UHL positive samples & 41 UHL negative samples).RT-LAMP reactions performed on the StepOnePlus platformwithpatient samples reporting an original RT-PCR of Ct<20 were equally detected by Orf1a and N+E RT-LAMP at 100% / 100% (DSe/DSp). DSe / DSp decreased slightly to 93.6% / 100% and 92.6% / 100% for theOrf1a and N+E targets respectively for patient samples reporting an RT-PCR Ct <30 and then to79.8%/100% and 78.1%/100% DSe / DSp respectively upon processing all patient samples reporting an RT-PCR Ct<40 (Table 3A).For the Rotor-Gene Q platform,RT-LAMP reactions of samples reporting an original RT-PCR of Ct<20 were detected equivalently by Orf1a and dual N+E RT-LAMP,reporting 100% / 100 % DSe / DSp respectively. This decreased to 93.9 % / 100% and97.0% / 100 % DSe / DSp for samples reporting an RT-PCR Ct<30 and finally to 84.3 % / 100% and 82.4% / 100 % DSe / DSp for samples reporting an RT-PCR Ct< 40 (Table 3B).Additional product specificity check provided by the melt curve (Tm) confirmed selective amplification of the product.
Table 3:DSe and DSp of single genomic versus dual sub-genomic SARS-CoV-2 fluorescent end-point RT-LAMP assays.
A)StepOnePlus; Applied Biosystems:
RT-LAMP target:
|
SARS-CoV-2 Orf1a
|
SARS-CoV-2 Dual N+E
|
Dse (95% CI)
|
100.0% (91.6% to 100.0%)
|
100.00% (91.6% to 100.0%)
|
DSp (95% CI)Ct < 20 (N42)
|
100.0% (95.9% to 100.0%)
|
100.00% (95.9% to 100.0%)
|
Dse (95% CI)
|
93.6% (86.6% to 97.6%)
|
92.6% (85.3% to 97.0%)
|
DSp (95% CI) Ct < 30 (N94)
|
100.0% (95.9% to 100.0%)
|
100.0% (95.9% to 100.0%)
|
Dse (95% CI)
|
79.8% (71.3% to 86.8%)
|
78.1% (69.4% to 85.3%)
|
DSp (95% CI) Ct < 40(N114)
|
100.0% (95.9% to 100.0%)
|
100.0% (95.9% to 100.0%)
|
B) Rotor-Gene Q; Qiagen
RT-LAMP target:
|
SARS-CoV-2 Orf1a
|
SARS-CoV-2 Dual N+E
|
Dse (95% CI)
|
100.0% (76.8% to 100.0%)
|
100.0% (76.8% to 100.0%)
|
DSp (95% CI) Ct < 20 (N14)
|
100.0% (91.4% to 100.0%)
|
100.0% (91.4% to 100.0%)
|
Dse (95% CI)
|
93.9% (79.8% to 99.3%)
|
97.0% (84.2% to 99.9%)
|
DSp (95% CI) Ct < 30 (N33)
|
100.0% (91.4% to 100.0%)
|
100.0% (91.4% to 100.0%)
|
Dse (95% CI)
|
84.4% (76.0% to 90.6%)
|
82.4% (73.9% to 89.1%)
|
DSp (95% CI)Ct < 40 (N39)
|
100.0% (95.9% to 100.0%)
|
100.00% (95.9% to 100.0%)
|
RT-LAMP reactions, performed using residual RNA extracted from clinical patient ON swab samples with comparator RT-PCR Ct values, were run at 65°C for 30 minutes on A) StepOnePlus PCR platform (114 UHL positive samples & 88 UHL negative samples) andB) Rotor-Gene Q Qiagen PCR platform (39 UHL positive samples & 41 UHL negative samples). Sensitivity and specificity with 95% CI are shown for samples with a corresponding RT-PCR Ct <20, Ct <30 and Ct < 40. Calculations performed using the MedCalc Scientific Software.
Diagnostic validation of Colorimetric RT-LAMP assays targeting Orf1a and dual N+E gene regions: Residual RNA extracted from patient ON swab samples were also processed through colorimetric end-point RT-LAMP reactions targeting the Orf1a, dual N+E geneplusACTB internal control reaction. Samples reporting an original RT-PCR of Ct<30 were detected equally by Orf1a and N+E RT-LAMP assays(21 of 21 RT-PCR positive samples and 40 of 40 negative samples) demonstrating 100% DSe and DSp. Assay performance decreased to 85.7%/ 100 % DSe / DSp for patient samples reporting an RT-PCR Ct<40 (18 of 21 RT-PCR positive samples).Additional performance of a triple target (Orf1a+N+E) assay demonstrated 100% DSe and DSp (41 of 41 RT-PCR positive samples and 40 of 40 negative samples) decreasing to 90.2% / 100% DSe / DSp for samples reporting an RT-PCR Ct<40 (37 of 41 RT-PCR positive samples).(Additional file 2).
University of Leicester RT-LAMP asymptomatic screening Programme: Screening was performed to ISO 15189:2012 standards, guided by the Leicester Molecular Diagnostic Lab. RNA extracted from throat and lower nasal cavity swabs from 1,673 symptom free individuals attending campus (autumn 2020) were processed through RT-LAMP targeting the Orf1a and internal control ACTB.During this period, a single repeat swab was requested due to inefficient sampling as identified by failure of amplification within the ACTB control reaction indicating absence of total RNA.In total, 9 RT-PCR confirmed RT-LAMP positive resultsfrom a total of 1,673 tests demonstrated a prevalence of asymptomatic infection of 0.54% (540 cases per 100,000) demonstrating the value and usability of RT-LAMP molecular diagnostic tool for the detection of SARS-CoV-2 in an asymptomatic population.