Development and evaluation of novel one-step TaqMan realtime RT-PCR assays for the detection and direct genotyping of genogroup I and II noroviruses

Abstract

Background:

Current detection and genotyping methods of genogroup (G) I and II noroviruses (NoVs) consist of a 2-step approach including detection of viral RNA by TaqMan realtime RT-PCR (RT-qPCR) followed by conventional RT-PCR and sequencing of partial regions of ORF1 or ORF2.

Objective:

To develop novel long-template one-step TaqMan assays (L-RT-qPCR) for the rapid detection and direct genotyping of GI and GII NoVs and to evaluate the sensitivity and specificity of the assays.

Study design:

GI and GII-specific broadly reactive L-RT-qPCR assays were developed by combining existing NoV primers and probes targeting the open reading frame (ORF)1–ORF2 junction as well as region C at the 5′–ORF2. The assays were validated using GI and GII RNA transcripts and a coded panel of 75 stool samples containing NoV strains representing 9 GI genotypes and 12 GII genotypes, as well as sapoviruses, astroviruses, polioviruses, and rotaviruses. L-RT-qPCR products were typed by sequencing.

Results:

The novel GI and GII L-RT-qPCR assays detected and typed all but one of the NoV positive panel samples. As few as 5–500 RNA copies could be accurately typed by sequencing of amplicons.

Conclusions:

We developed novel one-step TaqMan RT-qPCR assays for the sensitive detection and direct genotyping of GI and GII NoVs from clinical and environmental matrices.

1. Background

Noroviruses (NoVs) cause up to 95% of viral gastroenteritis outbreaks worldwide in humans of all age groups^1,2 and have been recognized as the most frequent cause of foodborne outbreaks.^3,4 Outbreaks often occur in closed settings such as hospitals, cruise ships, hotels and long-term care facilities. The ID₅₀ of NoV is as low as 18 virus particles,⁵ which allows these viruses to spread rapidly through airborne droplets, person-to-person contact, and environmental contamination. Foodborne outbreaks caused by NoVs are often linked to infected food handlers who contaminate foods that are eaten raw or minimally processed (ready-to-eat foods) or by contaminated shellfish.⁶ NoVs have also been associated with outbreaks after exposure to contaminated drinking water^7–9 and irrigation water (e.g. soft fruits and vegetables).^10–12

Since NoVs cannot be grown in cell culture,¹³ detection of genomic RNA by reverse transcriptase-PCR (RT-PCR) and more recently TaqMan realtime RT-PCR (RT-qPCR) has become the gold standard for laboratory diagnosis.^14,15 Strains can be genotyped by conventional RT-PCR targeting small regions of either ORF1¹⁶ or ORF2.¹⁷

2. Objective

We developed and evaluated novel RT-qPCR assays for the detection and direct genotyping of GI and GII NoVs.

3. Study design

3.1. Virus strains and RNA transcripts

A panel of 75 archived stool samples representing strains of most NoV genotypes (Fig. 3) were used to evaluate the long template (L) RT-qPCR (L-RT-qPCR) assays developed in this study. The panel included stool samples from outbreaks of acute gastroenteritis that had previously been tested positive for NoV by RT-qPCR¹⁵ as well as other enteric viruses including two rotaviruses gp A (serotype 1 and 3), three sapoviruses (all G.IV), two astroviruses (both type 4) and two enteroviruses (Sabin type 1 and type 3). NoV positive samples had been genotyped by sequencing of region C and D RT-PCR products.^18,19 The stool samples had been stored at 4 °C. For quantification, GI.3b, GI.4, GII.1 and GII.4 RNA transcripts²⁰ at 1 × 10⁶ copies/μl were used to create standard curves. In addition, two water samples (3959 and 3963) that had been concentrated by ultrafiltration²¹ and previously tested positive for GI NoV²² were included.

Fig. 3.

Evaluation of the S-RT-qPCR, M-RT-qPCR and L-RT-qPCR assays on a panel of 66 NoV positive stool samples including 28 GI samples (A) and 38 GII (B) samples plotted against copy numbers calculated from the standard curve data obtained with GI.4 or GII.1 RNA transcripts, respectively. Stool panel samples are listed on the X-axis by genotype and panel ID between brackets.

3.2. RNA extraction

Viral RNA was extracted from 50 μl of 10% stool suspensions (in PBS pH 7.2) with the Ambion MagMax (Applied Biosystems, Foster City, CA) kit and the KingFisher automated extractor (Thermofisher, Waltham, MA) and eluted in 100 μl of TE. Ten-fold serial dilutions of transcript RNA were prepared in RNA storage solution (Ambion, Austin, TX). Viral RNA from 500 μl of concentrated water sample was extracted as described elsewhere.²¹

3.3. Development of long template TaqMan realtime reverse transcription-polymerase chain reaction (L-RT-qPCR)

Since the originally developed RT-qPCR assays¹⁴ have been replaced by more sensitive assays (S-RT-qPCR)^23–26 to detect NoV in environmental samples, we used primers and probes from these assays combined with modified reaction conditions (M-RT-qPCR) to develop the L-RT-qPCR assays. Both S- and M-RT-qPCR assays target a small region of the ORF1–ORF2 junction for GI²³ and GII^14,24 (Table 1). To develop the GI and GII L-RT-qPCR assays targeting the ORF1–ORF2 junction region as well as the 5′-end of ORF2 (Fig. 1), we replaced the oligonucleotide reverse primers of the M-RT-qPCR assay by primer G1SKR¹⁸ for GI and primer G2SKR¹⁸ for GII (Table 1). In addition, reverse primer JV23²⁷ was evaluated for the GI L-RT-qPCR assay.

Table 1.

Oligonucleotide primers and probes used in this study.

Genogroup	Oligonucleotides	Sequences (5′–3′)a	Orientations	Positionb	Reference
GI	JJVMF	CCA TGT TCC GTT GGA TGC	+	5283–5300	23
	JJV1R	TCC TTA GAC GCC ATC ATC AT	−	5358–5377	22
	JV23	CCN RCM YAA CCA TTR TAC AT	−	5653–5672	27
	G1SKR	CCA ACC CAR CCA TTR TAC A	−	5653–5671	18
	RING1(c)	FAMc-AGA TYG CGI TCI CCT GTC CA-BHQd	−	5321–5340	23
GII	QNIF2	ATG TTC AGR TGG ATG AGR TTC TCW GA	+	5012–5037	24
	COG2R	TCG ACG CCA TCT TCA TTC ACA	−	5080–5100	14
	G2SKR	CCR CCN GCA TRH CCR TTR TAC AT	−	5367–5389	18
	QNIFS	FAMc-AGC ACG TGG GAG GGC GAT CG-BHQd	+	5042–5061	24

^aIUPAC codes used to indicate degenerate positions.

^bNucleotide positions based on Norwalk virus, GI (GenBank accession no. M87661) and Lordsdale virus (GII) (GenBank accession no. X86557) genomic sequences

^cFAM: 6-carboxyfluorescein reporter dye.

^dBHQ: Black Hole Quencher.

Fig. 1.

Schematic presentation of the genomic location of the regions targeted by the L-RT-qPCR assays developed in this study (shaded box) compared to commonly used regions for norovirus RT-qPCR detection (black box) and region C genotyping (white box). Nucleotide positions based on Norwalk virus, GI (GenBank accession no. M87661) and Lordsdale virus, GII (GenBank accession no. X86557) genomic sequences.

The compatibility of the novel primer combinations used for the L-RT-qPCR assays (Table 1) was initially evaluated by conventional RT-PCR using the One-Step RT-PCR Kit (Qiagen, Valencia, CA) on 10-fold serial dilutions of GI.3, GI.4, GII.1 and GII.4 RNA transcripts. Reverse transcription was carried out at 50 °C (30 min), followed by 95 °C (15 min) and 40 cycles of PCR at 94 °C (15 s), 50 °C (30 s), 72 °C (45 s) and a final incubation at 72 °C (10 min). RT-PCR products were separated by electrophoresis on 2% agarose gels and visualized under UV light.

We then evaluated the S-, M- and L-RT-qPCR assays in duplicate on an ABI Prism^® 7500 Sequence Detection System (Applied Biosystems). Reaction conditions (25 μl) for the GI and GII assays included the use of the UltraSense^™ One-step qRT-PCR kit (Invitrogen, Carlsbad, CA), 5 μl of template RNA, and oligonucleotide primers and probes for each S-, M-, and L assay (Table 1). In each experiment, GI.3b or GII.4 RNA transcripts were included as positive controls and water as a negative control. Cycling conditions for the S-RT-qPCR assays were performed as described previously^15,28 while the M-RT-qPCR assays included an RT-step at 55 °C (60 min), inactivation of RT at 95 °C (5 min), followed by 5 cycles at 95 °C (15 s), 55 °C (1 min), 65 °C (1 min) and 40 cycles at 95 °C (15 s), 60 °C (1 min), 65 °C (1 min). For the L-RT-qPCR assays, the cycling conditions included RT at 50 °C (GI assay) or 55 °C (GII assay) (60 min), inactivation of the RT at 95 °C (5 min) followed by 5 cycles at 95 °C (15 s), 50 °C (GI assay) or 55 °C (GII assay) (1 min), 65 °C (1 min) and further 40 cycles at 95 °C (15 s), 55 °C (GI assay) or 60 °C (GII assay) (1 min), 65 °C (32 s) using the default ramp-rate of 1.5 °C s⁻¹.

3.4. Validation of the L-RT-qPCR assays

The detection limit was determined by testing 10-fold serial dilutions of GI.3b, GI.4 or GII.1, GII.4 RNA transcripts representing 0.5–50,000 RNA copies/reaction (Fig. 2), and on a GI.1 stool extract. The specificity and broad reactivity of the L-RT-qPCR assays were compared with the S- and M-RT-qPCR assays on RNA extracted from the panel (Fig. 3).

Fig. 2.

Comparison of standard curves for the GI and GII norovirus S-RT-qPCR, M-RT-qPCR and L-RT-qPCR assays using 10-fold serial diluted norovirus 3 kb RNA transcripts GI.3b (A), GI.4 (B), GII.1 (C), and GII.4 (D). The Log₁₀ of the RNA copy numbers per reaction is plotted on the X-axis and the cycle number crossing point values (Ct values) are plotted on the Y-axis.

3.5. DNA sequencing and genotyping

All L-RT-qPCR products of appropriate size (389 bp for GI and 378 bp for GII) were gel-purified and cycle sequenced using BigDye^® chemistry (Applied Biosystems). Samples were analyzed on an ABI Prism^® 3130xl Genetic Analyzer (Applied Biosystems) and genotyped by local BLAST against GI and GII NoV reference sequence databases at CDC.

4. Results

4.1. Development and optimization of L-RT-qPCR assays

The GI and GII L-RT-qPCR assays were developed based on the M-RT-qPCR assays by replacing reverse primer JJV1R with G1SKR or JV23, and reverse primer COG2R with G2SKR, respectively. Since the assay with reverse primer JV23 detected fewer strains, the G1 L-RT-qPCR assay with G1SKR was further validated. Several different concentrations of primers and probes were evaluated using the four RNA transcripts and best results were obtained with 500/900/500 nM of the forward primer/reverse primer/probe (data not shown). Lowering the reverse transcription temperature from 55 to 50 °C improved the sensitivity of the GI L-RT-qPCR assay by 1 Log₁₀ but had no effect on the GII L-RT-qPCR assay (data not shown). In addition, shortening the elongation time of the GI and GII L-RT-qPCR assays from 60 s (as used in the S- and M-RT-qPCR assays) to 32 s did not affect the sensitivity of the assays. The L-RT-qPCR assays did detect all NoV positive panel samples except for one GI.1 sample, and all amplicons could be genotyped successfully after sequencing.

4.2. Detection limit of the L-RT-qPCR assays

Using ten-fold serial dilutions of the GI.3b and GI.4 RNA transcripts, the GI L-RT-qPCR standard curves were linear over a range of 50–50,000 RNA copies/reaction (Fig. 2A and B), while the end point detection of GI.3b was 5 copies (data not shown). The linear range for the GII L-RT-qPCR standard curves on GII.1 and GII.4 transcripts was 5–50,000 RNA copies/reaction (Fig. 2C and D). Lower Ct values were observed for the M- and L-RT-qPCR assays compared to the S-RT-qPCR assays (Fig. 2). The amplification efficiency for the S/M/L-RT-qPCR assays was 93/79/100%, 95/85/79%, 90/97/87%, and 69/90/87% for the GI.3b, GI.4, GII.1, and GII.4 transcript RNAs, respectively. The correlation coefficients (R²) ranged from 0.980 to 1.000 for the S-RT-qPCR standard curves, and 0.992 to 1.000 for the M- and L-RT-qPCR standard curves (Fig. 2). The detection limit for the GI.1 RNA stool extract was 100 RNA copies/reaction for the S- and L-RT-qPCR GI assays. The detection limit of the L-RT-qPCR assays for typing by sequencing of the amplicons was 500 and 5 for GI and GII, respectively.

4.3. Specificity of L-RT-qPCR assays

None of the other enteric viruses in the panel tested positive in any of the assays. The M- and L-RT-qPCR assays had lower Ct values for almost all GI and GII genotypes (data not shown). The ΔCt_{(S-RT-qPCR–M-RT-qPCR)} ranged from 7.6 to 20.6 (mean 11.3) for GI and from 4.1 to 13.2 (mean 7.7) for GII, while ΔCt_{(S-RT-qPCR–L-RT-qPCR)} ranged from −8.1 to 10.8 (mean 4.4) for GI and from −6.2 to 8.2 (mean 2.2) for GII (data not shown).

The number of RNA copies that could be detected per reaction demonstrated that the performance of the S-, M- and L-RT-qPCR assays was comparable for most strains (Fig. 3). The ΔLog₁₀ RNA copies/reaction(S-RT-qPCR–M-RT-qPCR) ranged from −3.5 to 0.4 (mean − 0.7) for GI and from −1.5 to 1.2 (mean 0.2) for GII, while ΔLog₁₀ RNA copies/reaction_{(S-RT-qPCR–L-RT-qPCR)} ranged from −1.1 to 4.2 (mean 0.7) for GI and from −1.4 to 2.6 (mean 0.3) for GII (Fig. 3). The differences were genotype specific although there was some variation within single genotypes. For example, four GII.3 strains had a ΔLog₁₀ RNA copies/reaction_{(S-RT-qPCR–L-RT-qPCR)} less than 0.45 whereas the remaining two GII.3 samples differed with 0.73 and 0.98. For GII.4 strains, the smallest and largest differences were 0.02 and 1.41, respectively.

Noteworthy discrepancies were found for a few strains (Fig. 3A), where the GI L-RT-qPCR assay either failed to detect (for GI.1) or for a few strains (one GI.4 and one GI.6 strain) detected much lower amounts of NoV than the S- and M-RT-qPCR assays. To explore this further, we compared end-point titrations for the S-, M- and L-RT-qPCR assays using a GI.1 sample and found that the S- and L-RT-qPCR was one log less sensitive compared to the M-RT-qPCR assays, respectively.

RNA extracted from water sample 3963 tested positive in the GI L-RT-qPCR assay and contained an estimated 6000 NoV GI RNA copies per 0.5 ml of water concentrate. The L-RT-qPCR product was sequenced and typed as GI.4.

5. Discussion

We developed novel broadly reactive long-template TaqMan assays for the detection and direct typing of GI and GII NoV targeting a region encompassing the ORF1–ORF2 junction region and the 5′-end of ORF2. These assays allow the detection of as few as 5–50 RNA copies per reaction as well as direct genotyping of 5–500 RNA copies by DNA sequencing of the L-RT-qPCR products. Most GII strains tested positive in the L-RT-qPCR assays with <1 Log loss of sensitivity compared to an assay that has been proposed as a European standard assay for the detection of NoV GII in food samples.^28,29

Since optimal qPCR efficiency requires an amplicon length between 50 and 150 bases, it was not surprising that the L-RT-qPCR assays which generate amplicons of 389 bp for GI and 378 bp for GII were less sensitive. Using a GII.4 RNA transcript as template and template-specific reverse primers producing amplicons of 250, 200 and 150 bases, we confirmed that loss of sensitivity of the GII L-RT-qPCR assay was directly related to the size of the amplicons. We also evaluated several different probe quenchers (BHQ, TAMRA, and MGB) or Locked Nucleic Acids (LNA) primers, but none of these modifications resulted in a notable improvement.

Direct sequencing of the L-RT-qPCR amplicons will speed up an outbreak investigation when the link between patient and source needs to be verified rapidly making re-amplifying RNA by conventional RT-PCR redundant which not only saves time in the laboratory but also minimizes the risk of cross contamination between samples in case a nested PCR reaction is required.

In summary, we successfully developed and validated novel RT-qPCR assays for the detection and direct genotyping of GI and GII NoVs. These assays may be used for rapid confirmation and typing of NoV in clinical and environmental matrices.

Abbreviations:

G

genogroup

NoV

norovirus

ORF

open reading frame

PBS

phosphate buffered saline

R ²

correlation coefficient

RNA

ribonucleic acid

RT-PCR

reverse transcriptase polymerase chain reaction

RT-qPCR

TaqMan realtime RT-PCR

S

short template

M

modified short template

L

long-template