Abstract
Rabies is diagnosed postmortem in animals, based on tests prescribed by the World Organization for Animal Health (OIE), such as the fluorescent antibody test, the direct rapid immunohistochemistry test, or pan-lyssavirus PCR assays. Several reverse-transcription real-time PCR (RT-rtPCR) methods have been developed and validated for rapid and accurate detection of lyssaviruses. We evaluated the performance of 6 TaqMan RT-rtPCR kits using different commercial master mixes and 2 real-time thermocyclers. Changing the master mix overall did not influence the TaqMan RT-rtPCR performance, regardless of the thermocycler used. The limits of detection at the 95% confidence level were 18.1–25.8 copies/µL for the Rotor-Gene Q MDx thermocycler and 16.7–21.5 for the Mx3005P thermocycler. Excellent repeatability was demonstrated for rabies virus (RABV) RNA samples of 100, 50, and 25 copies/µL regardless of the thermocycler used. RABV field samples (n = 35) isolated worldwide gave positive results using the most efficient of the 6 kits tested, with a copy number of 6.03 × 102 to 6.78 × 107 RNA copies per reaction. The TaqMan RT-rtPCR assay provides sensitive and rapid amplification of RABV RNA.
Keywords: Lyssavirus, rabies virus, TaqMan RT-rtPCR
Introduction
Rabies, an acute progressive fatal encephalitis, occurs on all continents apart from Antarctica. The disease is caused by viruses of the Lyssavirus genus (family Rhabdoviridae). The taxonomy of lyssaviruses includes 16 species recognized worldwide, including the prototype species of the Lyssavirus genus, rabies virus (RABV), which displays broad genetic diversity, mainly depending on hosts and geographic origin. RABV is responsible for the vast majority of human and animal rabies cases worldwide. Although rabies can be prevented by rabies pre-exposure vaccination or by post-exposure prophylaxis, tens of thousands of people die each year from rabies (http://who.int/rabies/en/). This estimated annual human incidence is probably underestimated15 given that this zoonosis is likely under-reported in many regions because of a lack of surveillance and laboratory infrastructure.10
Diagnosis of rabies is routinely conducted postmortem on brain tissue. The primary tests prescribed in 201833 by the World Organization for Animal Health (OIE) are the direct fluorescent antibody test (FAT), the direct rapid immunohistochemistry test (dRIT), or pan-lyssavirus PCR assays. FAT, dRIT, and PCR provide a reliable diagnosis in 98–100% of cases for all lyssavirus strains if an appropriate conjugate or primer–probe is used for the detection of RABV antigen.33
Molecular methods, such as the conventional or the reverse-transcription real-time PCR (RT-rtPCR) and other amplification techniques, are increasingly used in national reference laboratories (NRLs) for rabies as confirmatory tests or for epidemiologic analysis by typing the isolated rabies strains. RT-rtPCR assays for the detection of RABV have been published using TaqMan,4,14,19,30,31 SYBR Green,12,28 or a combination of SYBR Green and TaqMan RT-rtPCR.5,8
The evolution on virology tools clearly suggests that PCR techniques will in the near future supersede many of the classical direct methods of infectious agent detection, sensitivity and specificity of PCR generally being greater than viral isolation or antigen detection.18,22 PCR has been gradually replacing virus isolation or bacterial culture for the detection of agents that are difficult or impossible to culture.25,27 The development of various rtPCR methods, nucleic acid extraction robots, and rtPCR platforms has resulted in a large arsenal of high-throughput, robust, and reliable rtPCR assays. With the development of new molecular tools in the past 10 y associated with the increase of available kits, numerous studies have been performed3,9,23,26 on the evaluation of master mixes and rtPCR assays.
We previously demonstrated the pivotal influence of the master mix and the real-time platform for the detection of lyssavirus RNA by SYBR Green RT-rtPCR.21 Few comparison studies have been reported comparing the chemistries (TaqMan probes vs. SYBR Green) for detection of infectious agents by RT-rtPCR, particularly for the detection of RABV.5,8 We therefore evaluated 6 one-step probe RT-rtPCR kits routinely used in NRLs participating in inter-laboratory tests for rabies in order to identify those that produced the best sensitivity and repeatability for RABV detection. Our analysis of assay performance included analytical sensitivity with the limit of detection at 95%, repeatability, and the costs of kits, for 2 real-time thermocyclers.
Materials and methods
Preparation of in vitro RNA transcript
We selected the fixed RABV challenge virus standard strain (CVS-27, Anses 13-12, GenBank AY245851.1) for our study because this strain, which is commonly used as positive control for rabies diagnosis, belongs to the cosmopolitan RABV lineage found in domestic and wild animals worldwide. The generation of an in vitro RNA transcript was performed as follows: briefly, a conventional RT-PCR was performed on 5 µL of viral RNA with specific pan-lyssavirus primers13 [JW12 (forward): 5’-ATGTAACACCYCTACAATG, JW6 (reverse): 5’-CARTTVGCRCACATYTTRTG] for the amplification of a fragment of 606-bp of the nucleoprotein (N) gene. PCR products were purified and inserted into a vector (pGEM T easy vector, Promega France, Charbonnières-les-Bains, France) according to the manufacturer’s instructions for the production of recombinant plasmids. Plasmids were bi-directionally sequenced (Beckman Coulter Genomics, Takeley, UK) to check the sequence and the sense of the target insert. The plasmids were purified (PerfectPrep EndoFree plasmid maxi kit, 5Prime, VWR, Fontenay-sous-Bois, France), then concentrated using a step of sodium acetate–ethanol precipitation. Plasmids were linearized with the enzyme PstI and purified by acetate–ethanol precipitation.
In vitro transcription was performed on linearized DNA and T7 RNA polymerase according to the manufacturer’s instructions (Ambion MaxiScript SP6/T7 kit, Thermo Fisher Scientific, Illkirch, France). The resulting RNA transcripts were processed using 2 µL of DNase I (2 units/µL) for 15 min at 37°C, followed by a final purification step (RNeasy mini kit, Qiagen, Courtaboeuf, France).
The RNA concentration was determined (Qubit fluorometer, Invitrogen, Illkirch, France). The RNA molecule copy number was calculated following the manufacturer’s instructions (MaxiScript SP6/T7 kit, Thermo Fisher Scientific) using the following formula: Y molecules/µL = (X g/µL RNA/[transcript length in nucleotides × 330 g/mol] × 6.022 × 1023/mol). The estimated number of copies was 2.32 × 1011/µL for the CVS-27 RNA transcript.
Quality control of the synthesized RNA (diluted from 106 to 101 copies/µL) was checked by SYBR Green RT-rtPCR assay, as described previously.21 RNA transcripts were diluted in Tris–EDTA buffer and aliquoted in tubes of 5 µL at a concentration of 107 copies/µL. All aliquoted RNA transcripts were stored at −80°C to ensure the same storage conditions for the determination of analytical sensitivity.
Primers and probe
We selected the universal pan-lyssavirus primers (JW12: 5’-ATGTAACACCYCTACAATG-3’ and N165-146: 5’-GCAGGGTAYTTRTACTCATA-3’) and the TaqMan RABV probe (Lys-Gt1: 5’-ACAAGATTGTATTCAAAGTCAATAATCAG-3’) for our study.31 Two technical questionnaires accompanying the inter-laboratory test, respectively undertaken in 2014 and 2015 by the EU Reference Laboratory for Rabies, demonstrated that 9 of 15 NRLs used pan-lyssavirus primers and 8 of 16 NRLs used the TaqMan Lys-Gt1 probe, confirming the selection of these primers and TaqMan probe. Primers JW12 and N165-146 used in combination with the TaqMan Lys-Gt1 probe were previously designed to specifically detect the classical RABV.31
Commercial RT-rtPCR kits studied
Based on answers to the technical questionnaires by NRLs participating in the inter-laboratory proficiency tests, 6 commercial one-step RT-rtPCR kits were selected for the study: QuantiTect probe RT-PCR kit, Qiagen; QuantiTect virus RT-PCR kit, Qiagen; SuperScript III Platinum one-step qRT-PCR kit, Invitrogen, Life Technologies; Verso 1-step RT-qPCR probe kit, Thermo Fisher Scientific; AgPath-ID one-step RT-PCR kit, Applied Biosystems; and RNA UltraSense one-step quantitative RT-PCR kit, Invitrogen, Life Technologies.
TaqMan RT-rtPCR assays
RT-rtPCR protocol
RT-rtPCR was performed using a total volume of 25 µL on 2 µL of each RNA dilution and 23 µL of the master mix components from each tested kit. The concentrations of the forward and reverse primers and the Lys-Gt1 probe were optimized for each tested kit.21
All commercial kits were optimized with regard to the reverse transcription duration according to the manufacturer’s instructions (Table 1). The same PCR thermal cycling conditions (Table 1) were applied for the 6 kits tested (i.e., 45 cycles of 30 s at 94°C, 30 s at 55°C, and 30 s at 72°C). Real-time PCRs were performed on 2 real-time thermocyclers: Rotor-Gene Q MDx (RG, Qiagen) and Mx3005P (Agilent Technologies, Les Ulis, France). PCR runs were performed in independent assays on multiple days by one trained laboratory technician. Preparation of PCR master mixes was performed by manual pipetting.
Table 1.
Characteristics of 6 commercial master mix kits tested for detection of rabies virus RNA.
| Test kit | RTase/polymerase supplied in the kit | Thermocycling conditions | Reaction mix* |
|---|---|---|---|
| QuantiTect probe RT-PCR kit (Qiagen) | RTases Omniscript and Sensiscript, and HotStarTaq DNA pol | 30 min at 50°C, 15 min at 95°C; 45 cycles of 30 s at 95°C; 30 s at 55°C; 30 s at 72°C | 1× QuantiTect probe master mix, 1 mM MgCl2, 1% QuantiTect RT mix, 0.8 µM [0.8 µM] each of pan-lyssavirus primers, 0.1 µM [0.2 µM] Lys-Gt1 probe |
| QuantiTect virus RT-PCR kit (Qiagen) | Sensiscript, and HotStarTaq Plus DNA pol | 30 min at 50°C, 5 min at 95°C; 45 cycles of 30 s at 95°C; 30 s at 55°C; 30 s at 72°C | 1× QuantiTect virus master mix, 1% QuantiTect virus RT mix, 0.5 µM [0.5 µM] each of pan-lyssavirus primers, 0.1 µM [0.2 µM] Lys-Gt1 probe |
| Verso 1-step RT-QPCR probe kit (Thermo Fisher Scientific) | Verso RTase (RNA-dependent DNA pol) and Thermo-Start DNA pol | 30 min at 50°C, 15 min at 95°C; 45 cycles of 30 s at 95°C; 30 s at 55°C; 30 s at 72°C | 1× Verso 1-step master mix, 5% RT enhancer, 1% Verso enzyme mix, 0.4 µM [0.4 µM] each of pan-lyssavirus primers, 0.1 µM [0.2 µM] Lys-Gt1 probe, 0.3% ROX |
| SuperScript III Platinum one-step RT-qPCR kit (Invitrogen) | SSIII RTase (=MMLV) and Platinum Taq DNA pol | 15 min at 50°C, 2 min at 95°C; 45 cycles of 30 s at 95°C; 30 s at 55°C; 30 s at 72°C | 1× SSIII Platinum master mix, 5 mM MgSO4, 2% enzyme mix, 0.4 µM [0.4 µM] each of pan-lyssavirus primers, 0.1 µM [0.2 µM] Lys-Gt1 probe |
| RNA UltraSense one-step quantitative RT-PCR kit (Invitrogen) | SSIII RT (=MMLV), Platinum Taq DNA pol, and RNaseOUT ribonuclease inhibitor | 15 min at 50°C, 2 min at 95°C; 45 cycles of 30 s at 95°C; 30 s at 55°C; 30 s at 72°C | 1× RNA UltraSense master mix, 5% enzyme mix, 0.2 µM [0.4 µM] each of pan-lyssavirus primers, 0.1 µM [0.2 µM] Lys-Gt1 probe |
| AgPath-ID one-step RT-PCR kit (Applied Biosystems) | ArrayScript RTase (=MMLV) and AmpliTaq Gold DNA pol | 10 min at 45°C, 10 min at 95°C; 45 cycles of 30 s at 95°C; 30 s at 55°C; 30 s at 72°C | 1× AgPath-ID master mix, 1× enzyme mix, 0.5 µM [0.4 µM] each of pan-lyssavirus primers, 0.1 µM [0.1 µM] Lys-Gt1 probe |
MMLV = Moloney murine leukemia virus reverse transcriptase; pol = polymerase; RTase = reverse transcriptase.
Total volume of reaction is 25 µL, containing 2 µL of RNA sample and 23 µL of reaction mix. The RT-rtPCR reaction protocol corresponded to RT-rtPCR assays performed on the Rotor-Gene Q MDx. The concentrations of primers–probe used on the real-time thermocycler Mx3005P are shown in brackets.
Data collection and analysis
Negative (no template control) and positive controls (106–101 copies/µL of CVS-27 RNA) were included in each assay for its validation. A threshold setting of 0.03 was used as the reference threshold for the RG thermocycler, regardless of the kit tested. For the Mx3005P thermocycler, the threshold setting was 0.03 (dRn) for the RNA UltraSense, SuperScript III Platinum, QuantiTect virus, and Ag Path-ID kits, and 500 (dR) for the Verso and QuantiTect probe kits.
PCR efficiency (E) and coefficient of determination (R2) values were calculated by the software Rotor-Gene Q Series and Mx3005Pro. For each tested kit, the standard curve was obtained from 4 dilutions (i.e., 106–103 copies/µL) according to the standard NF-U-47-600, which is the linear dynamic range for the 6 tested kits. Efficiency values outside of the given range of 90–110% are of concern and are considered unsatisfactory. R2 was considered suitable when >0.99. A cutoff value ≥45 was defined for negative results.
Analysis of assay performance
Assay performance was analyzed for each kit tested and for both real-time platforms. The performance analysis included limit of detection at 95% confidence (LOD95%), precision, and repeatability of the different assays.
Analytical sensitivity
All assays were performed in triplicate for the determination of the limit of detection of PCR (LDPCR), the slope, and the PCR efficiency using 7-fold serial dilutions of CVS-27 RNA (i.e. 106–1 copy/µL) prepared in RNase-free water before use.
LOD95%, which corresponds to the lowest RNA concentration that could be detected in the sample with a confidence of 95%, was calculated according to the NF-U-47-600-2 standard (French Association of Normalization [AFNOR], Animal health analysis methods—PCR—Part 2: requirements and recommendations for the development and the validation of veterinary PCR). Determination of LOD95% was assessed by testing 6 dilutions surrounding the estimated level of detection limit of rtPCR that was determined using 6 dilutions of triplicates of CVS-27 RNA (106–101 copies/µL). The 6 dilutions (50, 25, 10, 5, 2.5, and 1 copy/µL) were tested in 8 replicates in 3 independent runs by the same highly trained operator under identical laboratory conditions. The LOD95% was determined using Probit analysis following the recommendations of the NF-U-47-600-2 standard (AFNOR, Animal health analysis methods—PCR—Part 2: requirements and recommendations for the development and the validation of veterinary PCR).
Precision of the RT-rtPCR assays
Precision of the 6 assays was assessed by calculating the coefficient of variation (CV = SD/mean × 100) for the cycle threshold (Ct) values expressed in % and the standard deviation (SD) for 6 dilutions of CVS-27 RNA (106–101 copies/µL) tested in triplicate.
Analysis of cross-platform repeatability
Cross-platform repeatability was analyzed on 24 replicate Ct values for 3 weak-positive RNA controls (100, 50, and 25 copies/µL), showing 100% positive results. These 3 RNA controls were selected near the estimated detection limit of rtPCR for all 6 tested kits. The variation of CVs was compared for RNA dilutions of 100, 50, and 25 copies/µL.
The 24 replicate Ct values for the weak-positive sample of 100 copies/µL of RNA were analyzed to assess the variability in test results for all kits tested. The link between the 2 platforms for the 6 kits, and among the 6 kits for the same platform, was analyzed using the Spearman rank-correlation coefficient (rS) of the mean Ct values.23 Ct values obtained using the different kits for each platform were compared by ranks using the Kruskal–Wallis test followed by multiple comparison tests (http://www.R-project.org, using “stat” and “pgirmess” for multiple comparisons). The kruskalmc function in the pgirmess package implements Dunn post-hoc rank sum comparison using z-test statistics.24
For all tests, p ≤ 0.05 was considered statistically significant. The repeatability standard deviation, CV, the lowest and highest values, the mean, the median, and the various statistical tests used were computed (R software v.2.8.1, http://www.R-project.org).
Determination of diagnostic sensitivity
Viral samples were obtained from the collection of lyssavirus, representative of the RABV species, maintained by ANSES’s Nancy Laboratory for Rabies and Wildlife (Malzéville, France). The RABV field strains originated from South America, Africa, Asia, and Europe (Table 2). Brains were collected from naturally infected animals and brain homogenates from field strains after mouse inoculation testing. All samples analyzed were tested previously for the presence of lyssavirus antigen by FAT.6 Thirty-five animal brain samples (17 wild and 18 domestic animals) that had tested negative for lyssavirus antigen by FAT were included in the determination of diagnostic sensitivity (Table 3).
Table 2.
Results of testing the TaqMan RNA UltraSense one-step quantitative RT-rtPCR kit with a panel of representative rabies virus field samples isolated worldwide.
| Origin/Sample ID | Host | Average Ct | No. of copies/µL of RNA ± SD |
|---|---|---|---|
| Latvia | |||
| Lett2004-5 | Dog | 20.4 ± 0.21 | 1.58E + 06 ± 0.20 |
| Lett2004-6 | Dog | 25.4 ± 0.31 | 7.09E + 04 ± 1.54 |
| Lett2004-7 | Cat | 20.2 ± 0.04 | 1.75E + 06 ± 0.04 |
| Lett2004-8 | Cat | 17.0 ± 0.31 | 1.28E + 07 ± 0.25 |
| Romania | |||
| DR1017 | Cat | 25.2 ± 0.01 | 1.09E + 05 ± 0.00 |
| DR1021 | Wolf | 31.6 ± 0.30 | 1.95E + 03 ± 0.37 |
| DR1025 | Fox | 24.2 ± 0.37 | 2.17E + 05 ± 0.51 |
| DR1026 | Dog | 27.4 ± 0.30 | 2.71E + 04 ± 0.52 |
| Estonia | |||
| EST 4 | Raccoon dog | 25.4 ± 0.62 | 1.03E + 05 ± 0.40 |
| EST 7 | Fox | 22.3 ± 0.97 | 7.73E + 05 ± 4.46 |
| Bulgaria | |||
| BUL 1 | Fox | 31.2 ± 0.12 | 1.79E + 03 ± 0.13 |
| BUL 2 | Fox | 32.9 ± 0.02 | 6.03E + 02 ± 0.08 |
| BUL 5 | Cat | 27.9 ± 0.15 | 1.48E + 04 ± 0.14 |
| BUL 7 | Dog | 27.1 ± 0.46 | 2.48E + 04 ± 0.72 |
| Republic of Macedonia | |||
| DR-457 | Fox | 16.0 ± 0.18 | 2.89E + 07 ± 0.33 |
| DR-458 | Fox | 16.7 ± 0.80 | 1.97E + 07 ± 0.96 |
| DR-435 | Wolf | 18.9 ± 0.64 | 4.79E + 06 ± 0.19 |
| Greece | |||
| GR64C/12 | Fox | 23.3 ± 0.25 | 2.82E + 05 ± 0.45 |
| Slovenia | |||
| DR-579 | Fox | 23.8 ± 0.33 | 2.06E + 05 ± 0.43 |
| Serbia | |||
| DR-0802 | Fox | 31.2 ± 0.43 | 1.82E + 03 ± 0.49 |
| Tunisia | |||
| TUN 1 | ND | 25.4 ± 0.28 | 9.89E + 04 ± 1.69 |
| TUN 2 | ND | 24.7 ± 1.22 | 1.74E + 05 ± 1.23 |
| Mali | |||
| 5 | Dog | 21.3 ± 0.09 | 1.02E + 06 ± 0.06 |
| 6 | Dog | 18.0 ± 0.28 | 8.28E + 06 ± 1.49 |
| 11 | Dog | 17.2 ± 0.20 | 1.36E + 07 ± 0.16 |
| 57 | Dog | 14.7 ± 0.30 | 6.78E + 07 ± 1.29 |
| Tanzania | |||
| SER 1 | Dog | 18.0 ± 0.25 | 8.06E + 06 ± 1.31 |
| SER 2 | Cattle | 16.3 ± 0.48 | 2.55E + 07 ± 0.77 |
| SER 3 | Dog | 18.1 ± 0.16 | 7.78E + 06 ± 0.81 |
| SER 4 | African wild dog | 15.6 ± 0.47 | 3.89E + 07 ± 1.15 |
| China | |||
| CHINE 1 | ND | 20.8 ± 0.04 | 1.82E + 06 ± 0.04 |
| CHINE 2 | ND | 21.4 ± 1.17 | 1.44E + 06 ± 0.98 |
| Brazil | |||
| BRE 7 | Cattle | 21.7 ± 0.44 | 1.01E + 06 ± 0.28 |
| BRE 8 | Cattle | 23.2 ± 0.98 | 4.43E + 05 ± 2.60 |
| BRE 11 | Bat | 19.5 ± 0.49 | 4.30E + 06 ± 1.31 |
Ct = cycle threshold; ND = no data. The diagnostic sensitivity was performed with the most efficient of the 6 kits tested.
Table 3.
Results of testing the TaqMan RNA UltraSense one-step quantitative RT-rtPCR kit with a panel of samples negative by FAT.
| Origin/Year of isolation | Sample ID | Host | FAT | Ct | Number of copies/µL of RNA ± SD |
|---|---|---|---|---|---|
| Europe, France | |||||
| 2016 | 131535 | Red fox | Negative | No Ct | 0 |
| 2016 | 131593 | Red fox | Negative | No Ct | 0 |
| 2016 | 131662 | Red fox | Negative | No Ct | 0 |
| 2016 | 131708 | Red fox | Negative | No Ct | 0 |
| 2016 | 131756 | Badger | Negative | No Ct | 0 |
| 2016 | 131757 | Red fox | Negative | No Ct | 0 |
| 2016 | 131764 | Red fox | Negative | No Ct | 0 |
| 2016 | 131768 | Red fox | Negative | No Ct | 0 |
| 2016 | 131776 | Red fox | Negative | No Ct | 0 |
| 2016 | 131778 | Red fox | Negative | No Ct | 0 |
| 2016 | 131782 | Roe deer | Negative | No Ct | 0 |
| 2017 | 131807 | Cat | Negative | No Ct | 0 |
| 2017 | 131808 | Rat | Negative | No Ct | 0 |
| 2017 | 131811 | Cat | Negative | No Ct | 0 |
| 2017 | 131812 | Marten | Negative | No Ct | 0 |
| 2017 | 131813 | Cat | Negative | No Ct | 0 |
| 2017 | 131815 | Stone marten | Negative | No Ct | 0 |
| 2017 | 131816 | Dog | Negative | No Ct | 0 |
| 2017 | 131817 | Dog | Negative | No Ct | 0 |
| 2017 | 131819 | Cattle | Negative | No Ct | 0 |
| 2017 | 131821 | Cat | Negative | No Ct | 0 |
| 2017 | 131872 | Dog | Negative | No Ct | 0 |
| 2017 | 131880 | Dog | Negative | No Ct | 0 |
| 2017 | 131894 | Dog | Negative | No Ct | 0 |
| 2017 | 131897 | Stone marten | Negative | No Ct | 0 |
| 2017 | 131898 | Dog | Negative | No Ct | 0 |
| 2017 | 131919 | Red fox | Negative | No Ct | 0 |
| 2017 | 131969 | Dog | Negative | No Ct | 0 |
| 2017 | 131970 | Dog | Negative | No Ct | 0 |
| 2017 | 131973 | Cat | Negative | No Ct | 0 |
| 2017 | 131974 | Cat | Negative | No Ct | 0 |
| 2017 | 131975 | Cattle | Negative | No Ct | 0 |
| 2017 | 131979 | Cat | Negative | No Ct | 0 |
| 2017 | 131980 | Red squirrel | Negative | No Ct | 0 |
| 2017 | 131986 | Dog | Negative | No Ct | 0 |
Ct = cycle threshold; FAT = fluorescent antibody test; SD = standard deviation. The diagnostic sensitivity was performed with the most efficient of the 6 kits tested.
Viral RNA was extracted from 200 μL of supernatant of 10% (w/v) brain tissue suspensions (iPrep PureLink virus kit, Invitrogen) according to the manufacturer’s instructions. All extracted RNAs were stored at −80°C until use.
Classification of the 6 tested kits regarding technical and financial criteria
The kits were ranked from 1 to 6 in increasing order based on technical (i.e. LOD95%) and financial (costs of kits) criteria. The rank was determined by attributing a mixing ratio of 70% for LOD95% results (i.e. 35% for RG and 35% for Mx3005P), 20% for the costs of kits, and 10% for the run duration. The kit with the highest number of points was considered the best kit.
Results
Comparison of assay performance
All kits tested using thermocyclers RG and Mx3005P showed satisfactory efficiency of 93% and 102%, respectively, as well as a satisfactory coefficient of regression R2 (>0.99). No significant differences were shown between the 2 platforms.
Limit of detection at 95%
For RG, LOD95% was 18.1–25.8 copies/µL (Table 4). Of the 6 commercial master mixes tested, the RNA UltraSense kit yielded the highest detection with LOD95% of 18.1 copies/µL. The kits AgPath-ID and QuantiTect virus ranked second and third with detection of 19.2 and 22.4 copies/µL, respectively, followed by the SuperScript III Platinum, QuantiTect probe, and Verso probe kits, which demonstrated a comparable level of RNA detection with a LOD95% of 24.4, 24.6, and 25.8 copies/µL, respectively.
Table 4.
Cross-platform comparison of the 6 RT-rtPCR assays for 7 dilutions of CVS-27 RNA ranging from 2.106 copies/µL to 1 copy/µL for the detection of rabies virus.
| No. of copies per sample | Testing of different RT-rtPCR kits (one-step RT-rtPCR) probe |
|||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| QuantiTect probe |
AgPath-ID |
Verso probe |
SSIII Platinum probe |
QuantiTect virus |
RNA UltraSense |
|||||||
| Average Ct | CV (%) | Average Ct | CV (%) | Average Ct | CV (%) | Average Ct | CV (%) | Average Ct | CV (%) | Average Ct | CV (%) | |
| Rotor-Gene Q MDx | ||||||||||||
| 106 | 20.9 ± 0.14 | 0.7 | 19.5 ± 0.09 | 0.5 | 20.9 ± 0.16 | 0.7 | 21.4 ± 0.11 | 0.5 | 19.9 ± 0.10 | 0.5 | 20.2 ± 0.10 | 0.5 |
| 105 | 24.6 ± 0.16 | 0.7 | 23.0 ± 0.18 | 0.8 | 24.7 ± 0.49 | 2.0 | 24.8 ± 0.42 | 1.7 | 23.4 ± 0.06 | 0.2 | 23.6 ± 0.02 | 0.1 |
| 104 | 27.9 ± 0.24 | 0.8 | 26.3 ± 0.09 | 0.4 | 27.5 ± 0.40 | 1.4 | 28.4 ± 0.05 | 0.2 | 26.3 ± 0.12 | 0.5 | 27.0 ± 0.07 | 0.3 |
| 103 | 31.2 ± 0.06 | 0.2 | 29.6 ± 0.28 | 0.9 | 31.2 ± 0.20 | 0.6 | 31.8 ± 0.25 | 0.8 | 30.1 ± 0.09 | 0.3 | 30.7 ± 0.34 | 1.1 |
| 102 | 34.2 ± 0.25 | 0.7 | 32.5 ± 0.21 | 0.7 | 35.5 ± 0.10 | 0.3 | 35.2 ± 0.22 | 0.6 | 33.2 ± 0.05 | 0.2 | 34.3 ± 0.24 | 0.7 |
| 101 | 38.4 ± 1.07 | 2.8 | 36.9 ± 0.36 | 1.0 | 39.8 ± 1.81 | 4.5 | 39.5 ± 2.22 | 5.6 | 37.1 ± 0.93 | 2.5 | 41.8 ± 0.86 | 2.0 |
| 1* | 1/3** | 1/3 | 0/3 | 1/3 | 1/3 | 0/3 | ||||||
| Efficiency | 96% | 98% | 98% | 93% | 100% | 94% | ||||||
| R² | 0.998 | 0.998 | 0.992 | 0.997 | 0.997 | 0.998 | ||||||
| Intercept | 41.5 | 39.8 | 41.3 | 42.3 | 39.9 | 41.1 | ||||||
| LOD95% (copies/µL) | 24.6 | 19.2 | 25.8 | 24.1 | 22.4 | 18.1 | ||||||
| Mx3005P | ||||||||||||
| 106 | 19.6 ± 0.20 | 1.1 | 19.3 ± 0.30 | 1.5 | 19.3 ± 0.32 | 1.7 | 20.6 ± 0.14 | 0.7 | 20.1 ± 0.08 | 0.4 | 21.3 ± 0.11 | 0.5 |
| 105 | 22.7 ± 0.24 | 1.1 | 22.4 ± 0.08 | 0.3 | 22.9 ± 0.50 | 2.2 | 23.9 ± 0.11 | 0.5 | 24.0 ± 0.23 | 1 | 25.0 ± 0.30 | 1.2 |
| 104 | 26.1 ± 0.29 | 1.1 | 25.9 ± 0.34 | 1.3 | 25.9 ± 0.33 | 1.3 | 27.5 ± 0.16 | 0.6 | 27.0 ± 0.17 | 0.6 | 28.3 ± 0.12 | 0.4 |
| 103 | 29.5 ± 0.25 | 0.9 | 29.2 ± 0.50 | 1.7 | 29.6 ± 0.11 | 0.4 | 31.1 ± 0.53 | 1.7 | 30.2 ± 0.38 | 1.2 | 31.5 ± 0.12 | 0.4 |
| 102 | 33.7 ± 0.22 | 0.6 | 32.2 ± 0.24 | 0.7 | 34.1 ± 0.11 | 0.3 | 34.8 ± 0.42 | 1.2 | 33.5 ± 0.29 | 0.9 | 37.8 ± 1.05 | 2.8 |
| 101 | 36.9 ± 0.99 | 2.7 | 35.6 ± 0.81 | 2.3 | 36.6 ± 1.50 | 4.1 | 37.2 ± 0.24 | 0.7 | 37.1 ± 0.94 | 2.5 | 35.7 ± 0.08 | 0.2 |
| 1* | 0/3 | 1/3 | 0/3 | 1/3 | 0/3 | 0/3 | ||||||
| Efficiency | 102% | 100% | 97% | 93% | 99% | 97.4% | ||||||
| R² | 0.996 | 0.994 | 0.992 | 0.996 | 0.995 | 0.998 | ||||||
| Intercept | 39.2 | 39.2 | 39.7 | 41.5 | 40.4 | 41.8 | ||||||
| LOD95% (copies/µL) | 20.9 | 20.8 | 21.5 | 20.8 | 20.3 | 16.7 | ||||||
Average cycle threshold (Ct) values = mean ± standard deviation (SD). CV (%) = SD/mean Ct × 100. CV was not calculated when some of the parallels failed to amplify. LOD95% = limit of detection at 95% confidence; SSIII = SuperScript III. PCR comparisons were performed in triplicate with pan-lyssavirus primers and Lys-Gt1 probe using Rotor-Gene Q MDx and Mx3005P thermocyclers.
Data is presented as number of positives/total number of experiments.
For Mx3005P, LOD95% was 16.7–21.5 copies/µL for the 6 probe kits tested (Table 4). The RNA UltraSense kit yielded the highest levels of detection with LOD95% of 16.7 copies/µL. The 5 others kits yielded comparable sensitivity with LOD95% of 20.3–21.5 copies/µL.
Precision of the RT-rtPCR assays
Regardless of the thermocycler used, all kits tested exhibited satisfactory repeatability, with a Ct CV <10% in the linearity range of rtPCR (6.3 Log–2.3 Log of copies/µL of RNA). CVs increased in the highest dilution tested (1.3 Log copies/µL) for all kits tested regardless of the machine used, with 0.97% (AgPath-ID) to 5.61% (SuperScript III Platinum probe) for RG, and 0.04% (RNA UltraSense) to 4.09% (Verso probe) for Mx3005P (Fig. 1).
Figure 1.
Cross-platform repeatability analysis of the 6 TaqMan RT-rtPCR assays. Six serial dilutions of rabies virus strain CVS-27 RNA (106–10 copies/µL of RNA) were tested 3 times on the same day in 3 independent runs. Coefficient of variation (CV) is shown for the A. Rotor-Gene Q MDx and B. Mx3005P thermocyclers.
Assay repeatability
The mean Ct values for the RNA control of 100 copies/µL were 33.0 (AgPath-ID) and 34.4 (QuantiTect virus) for Mx3005P, and 32.3 (AgPath-ID) to 34.9 (Verso probe) for RG. The SD ranged from 0.31 (RNA Ultrasense) to 0.70 (QuantiTect virus) for RG, and from 0.30 (QuantiTect virus) to 0.52 (SuperScript III Platinum and AgPath-ID) for Mx3005P (Table 3). SDs for the RNA UltraSense kit were 0.31 for RG and 0.50 for Mx3005P, and for the Verso probe kit were 0.46 for RG and 0.39 for Mx3005P.
Overall, assay repeatability was satisfactory for the 3 tested low-positive RNA controls of 100, 50, and 25 copies/µL (Table 5). For the 6 kits tested, the inter-assay CVs were <10%, demonstrating good reproducibility for all tested kits.20 For the positive RNA control of 100 copies/µL, CVs were 0.86% (i.e., QuantiTect) and 1.58% (i.e., AgPath-ID) for Mx3005P, and 0.93% (i.e., RNA UltraSense) and 2.13% (i.e., QuantiTect) for RG (Table 5). Regardless of the thermocycler or the kit used, all tested assays performed with the RNA concentration of 50 copies/µL exhibited satisfactory repeatability with CVs <10% (1.21–2.75% for RG and 1.16–2.20% for Mx3005P). CVs for the lowest RNA concentration tested (i.e., 25 copies/µL) were 1.14–3.09% for RG and 1.03–3.67% for Mx3005P.
Table 5.
Results of analyses of the 3 weak RNA positive controls (25, 50, and 100 copies/µL) tested 24 times with 6 commercial master mixes using the 2 thermocyclers: Rotor-Gene Q MDx and Mx3005P.
| Test kit/No. of copies/µL | Rotor-Gene Q MDx |
Mx3005P |
||||||
|---|---|---|---|---|---|---|---|---|
| Mean Ct | Median Ct | SD | CV (%) | Mean Ct | Median Ct | SD | CV (%) | |
| QuantiTect probe | ||||||||
| 100 | 34.7 | 34.7 | 0.4 | 1.2 | 33.4 | 33.4 | 0.4 | 1.2 |
| 50 | 35.7 | 35.7 | 0.7 | 1.9 | 33.9 | 33.9 | 0.7 | 2.2 |
| 25 | 36.6 | 36.4 | 0.7 | 2.0 | 35.0 | 35.0 | 0.6 | 1.8 |
| AgPath-ID | ||||||||
| 100 | 32.3 | 32.3 | 0.4 | 1.1 | 33.0 | 33.1 | 0.5 | 1.6 |
| 50 | 33.2 | 33.2 | 0.5 | 1.4 | 34.1 | 34.0 | 0.4 | 1.3 |
| 25 | 34.3 | 34.4 | 0.8 | 2.3 | 34.6 | 34.6 | 0.5 | 1.6 |
| Verso probe | ||||||||
| 100 | 34.9 | 34.9 | 0.5 | 1.3 | 32.8 | 32.8 | 0.4 | 1.2 |
| 50 | 36.1 | 36.0 | 1.0 | 2.7 | 33.8 | 33.9 | 0.5 | 1.5 |
| 25 | 38.1 | 38.1 | 1.2 | 3.1 | 35.1 | 35.1 | 1.3 | 3.7 |
| SSIII Platinum probe | ||||||||
| 100 | 34.7 | 34.7 | 0.5 | 1.4 | 33.0 | 33.0 | 0.5 | 1.6 |
| 50 | 36.3 | 36.2 | 0.8 | 2.1 | 33.8 | 33.7 | 0.4 | 1.2 |
| 25 | 37.0 | 36.8 | 0.9 | 2.4 | 34.7 | 34.7 | 0.9 | 2.6 |
| QuantiTect virus | ||||||||
| 100 | 33.1 | 33.2 | 0.7 | 2.1 | 34.4 | 34.5 | 0.3 | 0.9 |
| 50 | 34.4 | 34.4 | 0.6 | 1.7 | 35.1 | 35.1 | 0.5 | 1.4 |
| 25 | 34.9 | 35.1 | 0.7 | 2.1 | 36.4 | 36.4 | 0.4 | 1.0 |
| RNA UltraSense | ||||||||
| 100 | 33.9 | 34.0 | 0.3 | 0.9 | 33.6 | 33.5 | 0.5 | 1.5 |
| 50 | 35.1 | 35.0 | 0.4 | 1.2 | 34.5 | 34.5 | 0.5 | 1.4 |
| 25 | 35.9 | 35.9 | 0.4 | 1.1 | 35.7 | 35.7 | 0.5 | 1.4 |
Ct = cycle threshold; CV = coefficient of variation; SD = standard deviation; SSIII = SuperScript III. PCR comparisons were run on the Rotor-Gene Q MDx and Mx3005P thermocyclers.
Regardless of the machine or the kit used, the different assays exhibited overall excellent repeatability results for the RNA concentration of 100 copies/µL based on CVs, rS values, and p value (Table 6). The rS values were high regardless of the machine used or the kit tested, with rS values >0.998. The rS values were excellent and consistent within and between the RG and Mx3005P platforms tested (Table 6).
Table 6.
Spearman rank correlation coefficients for the different assays performed with the weak-positive control at 100 copies/µL across the thermocyclers Rotor-Gene Q MDx and Mx3005P.
| Rotor-Gene Q MDx | Mx3005P |
|||||
|---|---|---|---|---|---|---|
| QuantiTect probe | AgPath-ID | Verso probe | SSIII Platinum probe | QuantiTect virus | RNA UltraSense | |
| QuantiTect probe | ||||||
| rS | / | 0.999 | 0.998 | 0.999 | 0.999 | 0.999 |
| p | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| AgPath-ID | ||||||
| rS | 0.999 | / | 0.999 | 0.999 | 0.999 | 0.998 |
| p | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| Verso probe | ||||||
| rS | 0.999 | 0.998 | / | 0.999 | 0.999 | 0.998 |
| p | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| SSIII Platinum probe | ||||||
| rS | 0.999 | 0.999 | 0.999 | / | 0.999 | 0.999 |
| p | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| QuantiTect virus | ||||||
| rS | 0.999 | 0.999 | 0.999 | 0.999 | / | 0.999 |
| p | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| RNA UltraSense | ||||||
| rS | 0.999 | 0.999 | 0.999 | 0.999 | 0.999 | / |
| p | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
SSIII = SuperScript III. Spearman correlation coefficient (rS) was calculated for the 6 kits across the 2 platforms used. Comparisons were performed for the Rotor-Gene Q MDx and Mx3005P thermocyclers.
Classification of the kits by cost and performance
Of the 6 kits tested, AgPath-ID was the least expensive, followed by QuantiTect probe, QuantiTect virus, RNA UltraSense, Verso probe, and SuperScript III Platinum probe, which was the most expensive of the 6 kits tested.
When the 6 kits were ranked based on technical (performance and run duration) and financial (cost per PCR reaction) criteria, RNA Ultra Sense (LOD95% of 18.1 and 16.7) was ranked 1, followed by AgPath-ID, QuantiTect virus, QuantiTect probe, SuperScript III Platinum probe, and Verso probe (Table 7).
Table 7.
Comparison of the 6 commercial master mixes tested with regard to the LOD95% results across the 2 thermocyclers Rotor-Gene Q MDx and Mx3005P, the cost for each kit, and the practicality of the different assays in terms of run duration.
| Test kit | Rank/LOD95% |
Score for cost per PCR reaction | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| RG |
Mx |
Run duration (min) |
|||||||||
| LOD95% | Score | LOD95% | Score | RG | Mx | Mean | Score | Final score | Final rank | ||
| QuantiTect probe | 24.6 | 25.8 | 20.9 | 28.0 | 17.6 | 170 | 152 | 161 | 7.9 | 79.2 | 4 |
| AgPath-ID | 19.2 | 33.0 | 20.8 | 28.0 | 20.0 | 145 | 127 | 136 | 9.3 | 90.3 | 2 |
| Verso probe | 25.8 | 24.6 | 21.5 | 27.2 | 13.5 | 158 | 152 | 155 | 8.2 | 73.5 | 6 |
| SSIII Platinum probe | 24.1 | 26.3 | 20.8 | 28.0 | 11.2 | 130 | 124 | 127 | 10.0 | 75.6 | 5 |
| QuantiTect virus | 22.4 | 28.3 | 20.3 | 28.8 | 15.8 | 148 | 142 | 145 | 8.8 | 81.7 | 3 |
| RNA UltraSense | 18.1 | 35.0 | 16.7 | 35.0 | 14.9 | 130 | 124 | 127 | 10 | 94.9 | 1 |
RG = Rotor-Gene Q MDx; Mx = Mx3005P. The cost per reaction was calculated by dividing the cost for each commercial kit by the total number of PCR reactions that could be performed with each kit, based on the negotiated price of 2016. The final rank was determined by attributing a score of 70% for limit of detection at 95% confidence (LOD95%) results (i.e. 35% for RG and 35% for Mx3005P), 20% for the costs of the kits, and 10% for the run duration.
Diagnostic sensitivity
The RNA UltraSense kit, which was the most efficient of the 6 kits tested, was selected for the determination of diagnostic sensitivity. Of 35 samples tested, representative of Europe (n = 20), America (n = 3), Asia (n = 2), and Africa (n = 10), the 4 genetic lineages of RABV tested gave positive results. The mean Ct value was 22.4 (SD ± 0.39), with maximum and minimum Ct values of 14.7 and 32.9. Copy numbers were 6.03 × 102 to 6.78 × 107 RNA copies/µL (Table 2). All negative samples shown negative by FAT (n = 35) tested negative with the RNA UltraSense kit (Table 3).
Discussion
We evaluated the efficiency and repeatability of 6 commercial probe kits used, to date, in NRLs for the detection of lyssavirus RNA by TaqMan RT-rtPCR. In contrast to our previous evaluation of SYBR Green RT-rtPCR kits that showed variability of the SYBR Green master mix in rtPCR performance and the pivotal influence of the thermocycler,21 all TaqMan kits tested in our current study showed very satisfactory sensitivity and repeatability within and between thermocyclers. The analytical sensitivity of 18.1–25.8 copies/µL for Rotor-Gene Q MDx and 16.7–21.5 copies/µL for Mx3005P were acceptable. These results are comparable to the limit of detection of the LN34 pan-lyssavirus real-time RT-PCR assay published in 201811 estimated as 8 RNA copies (95% confidence interval: 0–18 copies). The CVs used to evaluate the reliability of thermocyclers were acceptable. CVs were <5% for all assays, and the coefficient of correlation was >0.998 for all tested kits within and between platforms. The same low variations were also reported8 in the evaluation of the pan-Lyssa RT-rtPCR assays showing the good reliability of the TaqMan RT-PCR for the detection of lyssavirus RNA.
We showed that the thermocycler used did not influence the TaqMan RT-rtPCR performance, in contrast to previous studies3,16,17 that indicated significant variability in the analytical sensitivity and robustness of PCR among different types of thermocyclers. The majority of studies undertaken for testing the robustness and the analytical sensitivity of rtPCR were based on the comparison of thermocyclers equipped with Peltier-based heating and cooling systems.3,16,17 The 2 thermocyclers selected in our present study, RG and Mx3005P, showed different throughput capacities. Few studies were performed with RG, which differs from Mx3005P on thermal engine (heating block technology based on the Peltier effect vs. heat exchange technology, which permits more rapid thermal ramp rates than blocks). The performance characteristics of 3 TaqMan assays were shown to be different for the detection of Borrelia17 on 3 thermocyclers, and appeared to perform best on the RG thermocycler.
Comparison of kits using RG and Mx3005P showed equivalent LOD95% results, demonstrating that the change of instrument and/or platform and the TaqMan kits constituted a minor change for a detection assay, according to the OIE validation guideline.32
We tested a panel of 35 field animal samples representative of a broad range of RABV field samples with the most efficient kit, RNA UltraSense. Ct values ranged from 14.7 for positive samples (high RNA level) and 32.9 for weak-positive samples (low RNA level), with an average of 22.4 ± 0.4 (7.19 × 106 RNA copies/µL). Such results are consistent with previous studies showing a diagnostic sensitivity >99% with the same method (one-step technique).5,8,30,31 The high diagnostic sensitivity coupled to the high analytical sensitivity of the TaqMan RT-rtPCR (100%) confirmed the robustness of the method tested with the pan-lyssavirus primers.
We previously showed that the LOD95% varied 15–981 copies/µL for SYBR Green RT-rtPCR following optimization of the detection method.21 The cross-platform evaluation of TaqMan RT-rtPCR kits in our present study showed comparable analytical sensitivity results. We demonstrated that the LOD95% was <26 copies/µL using thermocyclers RG and Mx3005P. These results are consistent with other studies demonstrating that the performance of SYBR Green rtPCR can be comparable to the TaqMan performance after optimization of the SYBR Green rtPCR.1,29
The advantages of molecular testing for viral diagnosis are widely recognized.7 Our study demonstrates that PCR assays are reliable and sensitive tools for the detection of lyssavirus RNA within suspect specimens with the advantage that they do not require the presence of live virus.33 The TaqMan RT-rtPCR assay tested in our ongoing study represents a useful method for the sensitive and rapid amplification of RABV RNA. To date, the most widely used test in NRL for RABV detection is the FAT, as both specificity and sensitivity approach 100%.2 With the advent of PCR methods that are increasingly used for the detection of pathogens, and the data accumulated showing the combined sensitivity and robustness of the TaqMan assay, the new gold standard in RABV detection could become, in the short-to-medium term, the RT-rtPCR.
Acknowledgments
We thank Alexandre Servat, Estelle Litaize, Sebastien Kempff, and Valère Brogat for proficient technical support.
Footnotes
Declaration of conflicting interests: The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: This work was supported by the European Commission and ANSES.
ORCID iD: Evelyne Picard-Meyer 
https://orcid.org/0000-0001-9122-9027
References
- 1. Arikawa E, et al. Cross-platform comparison of SYBR Green real-time PCR with TaqMan PCR, microarrays and other gene expression measurement technologies evaluated in the MicroArray Quality Control (MAQC) study. BMC Genomics 2008;9:328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Bourhy H, et al. Comparative field evaluation of the fluorescent-antibody test, virus isolation from tissue culture, and enzyme immunodiagnosis for rapid laboratory diagnosis of rabies. J Clin Microbiol 1989;27:519–523. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Buzard GS, et al. Multi-platform comparison of ten commercial master mixes for probe-based real-time polymerase chain reaction detection of bioterrorism threat agents for surge preparedness. Forensic Sci Int 2012;223:292–297. [DOI] [PubMed] [Google Scholar]
- 4. Coertse J, et al. Improved PCR methods for detection of African rabies and rabies-related lyssaviruses. J Clin Microbiol 2010;48:3949–3955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Dacheux L, et al. Dual combined real-time reverse transcription polymerase chain reaction assay for the diagnosis of lyssavirus infection. PLoS Negl Trop Dis 2016;10:e0004812. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Dean DJ, et al. The fluorescent antibody test. In: Meslin FX, et al., eds. Laboratory Techniques in Rabies. 4th ed. Geneva: World Health Organization, 1996:88–95. [Google Scholar]
- 7. Espy MJ, et al. Real-time PCR in clinical microbiology: applications for routine laboratory testing. Clin Microbiol Rev 2006;19:165–256. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Fischer M, et al. Molecular double-check strategy for the identification and characterization of European lyssaviruses. J Virol Methods 2014;203:23–32. [DOI] [PubMed] [Google Scholar]
- 9. Fischer M, et al. Perspectives on molecular detection methods of lyssaviruses. Berl Munch Tierarztl Wochenschr 2012;125:264–271. [PubMed] [Google Scholar]
- 10. Fooks AR, et al. Current status of rabies and prospects for elimination. Lancet 2014;384:1389–1399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Gigante CM, et al. Multi-site evaluation of the LN34 pan-lyssavirus real-time RT-PCR assay for post-mortem rabies diagnostics. PLoS One 2018;13:e0197074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Hayman DT, et al. A universal real-time assay for the detection of lyssaviruses. J Virol Methods 2011;177:87–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Heaton PR, et al. Heminested PCR assay for detection of six genotypes of rabies and rabies-related viruses. J Clin Microbiol 1997;35:2762–2766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Hoffmann B, et al. Improved safety for molecular diagnosis of classical rabies viruses by use of a TaqMan real-time reverse transcription-PCR “double check” strategy. J Clin Microbiol 2010;48:3970–3978. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Knobel DL, et al. Re-evaluating the burden of rabies in Africa and Asia. Bull World Health Organ 2005;83:360–368. [PMC free article] [PubMed] [Google Scholar]
- 16. Luque-Perez E, et al. Testing the robustness of validated methods for quantitative detection of GMOs across qPCR instruments. Food Anal Methods 2013;6:343–360. [Google Scholar]
- 17. Maes L, et al. Technical and clinical validation of three commercial real-time PCR kits for the diagnosis of neuroborreliosis in cerebrospinal fluid on three different real-time PCR platforms. Eur J Clin Microbiol Infect Dis 2017;36:273–279. [DOI] [PubMed] [Google Scholar]
- 18. McElhinney LM, et al. Effects of carcase decomposition on rabies virus infectivity and detection. J Virol Methods 2014;207:110–113. [DOI] [PubMed] [Google Scholar]
- 19. Nadin-Davis SA, et al. Development of real-time reverse transcriptase polymerase chain reaction methods for human rabies diagnosis. J Med Virol 2009;81:1484–1497. [DOI] [PubMed] [Google Scholar]
- 20. Pfaffl M. Quantification strategies in real-time PCR. In: Bustin S, ed. A-Z of Quantitative PCR. La Jolla, CA: International University Line, 2004,87–112. [Google Scholar]
- 21. Picard-Meyer E, et al. Cross-platform evaluation of commercial real-time SYBR green RT-PCR kits for sensitive and rapid detection of European bat Lyssavirus type 1. Biomed Res Int 2015;2015:839518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Romeiro MF, et al. Evaluation and optimization of SYBR Green real-time reverse transcription polymerase chain reaction as a tool for diagnosis of the Flavivirus genus in Brazil. Rev Soc Bras Med Trop 2016;49:279–285. [DOI] [PubMed] [Google Scholar]
- 23. Sato F, et al. Intra-platform repeatability and inter-platform comparability of microRNA microarray technology. PLoS One 2009;4:e5540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Siegel S, Castellan NJ. Non Parametric Statistics for the Behavioural Sciences. New York: MacGraw Hill, 1988:213–214. [Google Scholar]
- 25. Speers DJ. Clinical applications of molecular biology for infectious diseases. Clin Biochem Rev 2006;27:39–51. [PMC free article] [PubMed] [Google Scholar]
- 26. Stephens KW, et al. Cross-platform evaluation of commercial real-time reverse transcription PCR master mix kits using a quantitative 5’nuclease assay for Ebola virus. Mol Cell Probes 2010;24:370–375. [DOI] [PubMed] [Google Scholar]
- 27. Storch GA. Diagnostic virology. Clin Infect Dis 2000;31:739–751. [DOI] [PubMed] [Google Scholar]
- 28. Suin V, et al. A two-step lyssavirus real-time polymerase chain reaction using degenerate primers with superior sensitivity to the fluorescent antigen test. Biomed Res Int 2014;2014:256175. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Tajadini M, et al. Comparison of SYBR Green and TaqMan methods in quantitative real-time polymerase chain reaction analysis of four adenosine receptor subtypes. Adv Biomed Res 2014;3:85. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Wadhwa A, et al. A pan-lyssavirus Taqman real-time RT-PCR assay for the detection of highly variable rabies virus and other lyssaviruses. PLoS Negl Trop Dis 2017;11:e0005258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Wakeley PR, et al. Development of a real-time, TaqMan reverse transcription-PCR assay for detection and differentiation of lyssavirus genotypes 1, 5, and 6. J Clin Microbiol 2005;43:2786–2792. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. World Organization for Animal Health (OIE). Chapter 3.6.3. Development and optimisation of nucleic acid detection assays. In: OIE Manual of Diagnostic Tests and Vaccines for Terrestrial Animals. Paris, France: OIE, 2018:1–11. [Google Scholar]
- 33. World Organization for Animal Health (OIE). Chapter 2.1.17. Rabies (infection with rabies and other lyssaviruses). In: OIE Manual of Diagnostic Tests and Vaccines for Terrestrial Animals. Paris, France: OIE, 2018:1–35. [Google Scholar]