Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Norovirus Extraction from Frozen Raspberries Using Magnetic Silica Beads

Norovirus Extraction from Frozen Raspberries Using Magnetic Silica Beads Human noroviruses (HuNoV) are among the main causes of acute gastroenteritis worldwide. Frozen raspberries have been linked to several HuNoV food-related outbreaks. However, the extraction of HuNoV RNA from frozen raspberries remains challenging. Recovery yields are low, and real-time quantitative reverse transcriptase PCR (RT-qPCR) inhibitors limit the sensitivity of the detection methodologies. A new approach using fine magnetic silica beads was developed for the extraction of HuNoV spiked on frozen raspberries. Relatively low recovery yields were observed with both the magnetic silica bead and the reference ISO 15216-1:2017 methods. High RT-qPCR inhibition was observed with the ISO 15216-1:2017 recommended amplification kit but could be reduced by using an alternative kit. Reducing RT-qPCR inhibition is important to limit the number of inconclusive HuNoV assays thus increasing the capacity to assess the HuNoV prevalence in frozen raspberries. Keywords Norovirus · Raspberries · RNA extraction · Silica · RT-qPCR Introduction with contaminated frozen product between 2008 and 2018 (Nasheri et al. 2019). Human Norovirus (HuNoV) is one of the leading causes of Noroviruses are small (27–40 nm) non-enveloped single- food-related illnesses in developed countries. HuNoV rep- stranded RNA viruses that are transmitted mainly via the resents 58% of all reported foodborne outbreaks of known fecal–oral route. Noroviruses belong to a genetically diverse etiology in the United States (Vinje 2015). About 300–400 group of viruses of the Caliciviridae family. There are 10 outbreaks of HuNoV are reported to the National Enteric distinct norovirus genetic groups (Chhabra et  al. 2019). Surveillance Program of the Public Health Agency of Can- HuNoV GI and GII are the most prevalent genogroups asso- ada each year (Government of Canada 2018). Forty-six berry ciated with outbreaks. Noroviruses can persist in an infec- outbreaks associated to HuNoV contamination with 15,827 tious state for prolonged periods of time in the environment, cases were reported globally between 1983 and 2018 (Boz- in water, and in food (reviewed in Cook et al. (2016)). They kurt et al. 2020). In the European Union, contaminated fro- can also withstand a broad pH range (pH 2–9). In cold tem- zen red fruits were shown to represent an important cause peratures, noroviruses can stay infectious for years. How- of HuNoV outbreaks (Boqvist et al. 2018). Contaminated ever, they are inactivated by cooking. The greatest risk of frozen raspberries are the most common sources (Bozkurt a foodborne HuNoV infection arises from the consumption et  al. 2020). Frozen raspberries were implicated as food of contaminated food such as fresh or frozen fruits, leafy vehicles in 33 of the 40 reported outbreak events associated vegetables, oysters, and drinking water. Detection of HuNoV relies on the extraction of RNA and real-time reverse transcriptase PCR amplification method- ologies (Vinje 2015). Prior to detection, minute amounts of * Philippe Raymond viruses must be extracted from food matrices. Methods for philippe.raymond@canada.ca concentrating noroviruses extracted from contaminated food Canadian Food Inspection Laboratory (CFIA), St. Hyacinthe are based mainly on ultrafiltration, ultracentrifugation, cati- Laboratory, Food Virology, Saint Hyacinthe, QC, Canada onic separation, and polyethylene glycol (PEG) precipita- Agriculture and Agri-Food Canada (AAFC), St. Hyacinthe tion. These methods are associated with high variability and Research and Development Centre, Saint Hyacinthe, QC, relatively low recovery of HuNoV (Summa and Maunula Canada Vol:.(1234567890) 1 3 Food and Environmental Virology (2021) 13:248–258 249 2018). The ISO/TS 15216-1:2013 technical specification and for Disease Control (BCCDC). The preparation of HuNoV the subsequent ISO 15216-1:2017 standard were published from clarified 10% stool samples was adapted from Houde to provide a reference method for the quantitative extrac- et al. (2006). Briefly, the stool samples were diluted in 1× tion and detection of HuNoV from various food products. phosphate-buffered saline (PBS), pH 7.4 (Gibco, Canada) to The extraction from soft fruit, leaf, stem, and bulb vegeta- obtain a 10% suspension and were homogenized by vigor- bles is based on PEG precipitation of the virus. However, ous agitation. Suspensions were clarified by centrifugation PCR inhibitors are frequently reported when PEG-based (20,000×g, 15 min, 4 °C) and kept frozen at − 80 °C. approaches are applied to soft fruits and can lead to false- negative results (Summa and Maunula 2018). PCR inhibi- Frozen Raspberry Samples tors are also reported when using ultrafiltration, Cat-Floc precipitation, and immunomagnetic bead-based extraction Frozen raspberries, from bags labeled as whole individu- methods as well, reducing the norovirus recovery yields ally quick frozen (IQF) collected at local stores, were sub- substantially (Summa et al. 2012). Extraction methods with divided in 25 g samples and used for artificial contamination low recovery yields increase the probability of false-negative experiments. results. Assays with more than 75% real-time quantitative reverse transcriptase PCR (RT-qPCR) inhibition should be Artificial Contamination of Frozen Raspberries considered inconclusive (ISO 2017). Other groups have used cationic beads or filters at high Aliquots of clarified 10% stools of HuNoV were vortexed for pH for norovirus adsorption and elution and have had vari- 2 s and diluted in PBS to the genomic equivalent copy (gEq) ous success rates (Morales-Rayas et al. 2010; Scherer et al. level needed in 100 µl per sample. To estimate the recovery 3 6 2010; Stals et al. 2012). The recombinant HuNoV capsid yields, between 10 and 10 gEq of HuNoV GI.5, GII.4, and protein VPI isoelectric point was reported to be pH 5 (Da MNV were inoculated. To estimate the recovery yields in the Silva et al. 2011). Accordingly, norovirus should be posi- presence of a competitor strain, 10 gEq HuNoV GII.4 and tively charged below its capsid isoelectric point. Relatively 10 gEq GI.5 were inoculated simultaneously. To estimate high recovery yields (68%) for HuNoV GII from wastewa- the limit of detection (LOD), the HuNoV GII.4 stocks were 5 2 ter were reported using celite, a siliceous rock powder, at serially diluted to 10 –10 gEq per 100 µl. pH 4 (Brinkman et al. 2013). Silica surfaces are covered Frozen raspberries were spiked with the diluted viral by silanol groups which can exist in different states. From suspensions on the surface of the food matrices (25 g) in a previous studies, it is expected that crystalline and vitreous Whirl–Pak filter bag (VWR, Canada), then left to air dry silica surfaces should maintain a negative charge above pH 30 min in a biosafety cabinet. A frozen raspberries sample, 2 and 3, respectively (Júnior and Baldo 2014). Andrade et al. with no virus added, was included in each extraction batch as (2009) reported an isoelectric point of 2.3 for silica-coated a negative control. The amount of virus in the 100 µl inocu- magnetic beads in suspension in KCl 1 mM. lum was assessed in parallel by extracting the total RNA In this study, we present the performance of a new mag- using the RNeasy kit (QIAGEN) followed by a RT-qPCR netic silica bead (MSB) methodology for the extraction of assay as described below. norovirus. We compared the performance of this approach to the ISO 15216-1:2017 method with a focus on the RT-qPCR Virus Elution and RNA Extraction inhibition associated to the extraction and detection. Reduc- ing RT-qPCR inhibition is important to limit the number of Viral RNA was extracted from frozen raspberries using inconclusive assays. the MSB method by proceeding with the following steps (Fig. 1): Step 1 The virus was eluted from the matrix by adding Methods 40 ml of elution buffer made of 150 mM Bis–Tris-Propane (Sigma-Aldrich, Canada) pH 8 to the sample in the filter Virus Stocks bag then closing the bag and shaking it at about 55 RPM for 30 min at room temperature (RT) using an orbital shaker. Murine norovirus-1 (MNV) was provided by Dr. H. Virgin Using a 25 ml borosilicate glass pipet, the eluate was trans- from Washington University (St. Louis, MO, USA). MNV ferred to a 50-ml conical centrifuges tubes. was propagated in the RAW 264.7 cell line as previously Step 2 The eluate was clarified by centrifugation at described and viral stocks were titrated by plaque assays 3500×g for 10 min. Thirty units of Aspergillus niger pec- (Gonzalez-Hernandez et al. 2012). HuNoV-positive stool tinase (Sigma-Aldrich) were added to the clarified eluate samples HuNoV GI.5 (CFIA-FVR-022) and GII.4 (CFIA- followed by a 30 min incubation at 37 °C with shaking at FVR-019) were provided by the British Columbia Center approximately 70 RPM. Meanwhile, a volume of 100 µl per 1 3 250 Food and Environmental Virology (2021) 13:248–258 Fig. 1 Norovirus extraction and concentration method using the magnetic silica beads (MSB) approach. Viruses (blue dots) are purified from matrix and soil contaminants (brown and red dots) using magnetic silica beads (gray dots) prior to the RNA extraction sample of AccuNanobeads (Bioneer, CA, USA, average size the elution buffer. The first and second elutions were range 300–700 nm) was vortexed for 3 min. The AccuNano- performed by agitation at about 1050 RPM in a Thermo- beads storage buffer was removed using a magnet, and the mixer R (Eppendorf, Canada) at RT for 10 min and 1 min, beads were resuspended in 1 ml per sample of the elution respectively. Virus eluates were immediately transferred buffer containing 100 µg/ml Pluronic F-127 (Anatrace, OH, to a 1.5 ml microtube and combined with 500 µl buffer USA) and vortexed for a few seconds. RLT plus β-mercaptoethanol from the RNeasy QIAcube Step 3 The bead suspension was added to the clarified kit (QIAGEN) or 10  µl of 2  M Dithiothreitol (Sigma- eluate and vortexed 3 s. Two milliliters of 200 mM ascor- Aldrich). Five microliters of RNA carrier (1 µg/ml) (QIA- bic acid (Sigma-Aldrich) and 3.5 ml of 200 mM malic acid GEN) and 140 µl of the polyvinylpolypyrrolidone (PVPP; (Sigma-Aldrich) were added to each sample, and the tubes Sigma-Aldrich) suspension (2% v/v final) were added to were vortexed another 3 s. The pH was lowered between pH each sample, and the samples were vortexed briefly. Sam- 2 and 3 with the addition of approximately 850 µl of 6 N ples were centrifuged at 10,000×g minimum for 5 min to HCl. The tubes were then mixed using a Dynal rotary shaker remove the PVPP. 94701 (Thermo Fisher, Canada) set between 19 and 22, for Step 6 Total RNA was extracted using the RNeasy QIA- 10 to 60 min at RT. cube kit supplemented with DNase I in the QIAcube plat- Step 4 The solution was separated from the beads using a form as described by the manufacturer (QIAGEN). The magnetic rack and removed by decantation. The remaining RNA was eluted from the spin column membrane with 50 µl beads were washed with 1 ml of washing bue ff r composed of of RNase-free water from the kit to which 40 units of the 17.4 mM Bis–Tris-Propane buffer pH 7 with 8.7 mM NaCl, RNasin Plus RNase inhibitor (Fisher Scientific, Canada) 0.87 mM CaCl , 10.9 mM ascorbic acid, and 15.2 mM malic were added. It was then stored at − 80 °C. acid. As a reference method, the ISO 15216-1:2017 method for Step 5 The viruses were eluted twice from the beads soft fruit samples (ISO 2017) was also used for the extrac- with 100 µl of the bead elution buffer composed of 45 mM tion of the viruses from spiked frozen raspberries. The pec- Bis–Tris-propane pH 9, 0.01% Tween 20, and 50  mM tinase from Aspergillus aculeatus (1140 U) was used. The EDTA (Sigma-Aldrich). Beads were vortexed 3  s with NucliSens miniMAG kit (Biomérieux, Canada) was used to 1 3 Food and Environmental Virology (2021) 13:248–258 251 extract RNA in 50 µl elution buffer following the manufac- QPCRs were performed as described below using the C1000 turer’s recommendations. touch Thermal cycler and ddPCR Multiple Supermix (Bio- Rad). Plates were read using the QX200 Droplet Reader RT‑qPCR (Bio-Rad). RT-qPCR assays were performed using 5 µl of diluted (1/10) Recovery Yield Calculation or non-diluted RNA extracts, using either the Mx3005P system (Stratagene, CA, USA) or the Quantstudio 6 system The recovery yields associated with the virus elution and (Thermo Fisher). Diluted RNA extracts were prepared using concentration steps were estimated using the cycle thresholds RNase-free water. Unless otherwise specified, HuNoV GII (ΔCt/m) (Ct) variation. The virus recovery yields = 10 × 100%, RT-qPCR was performed using QNIF2d and COG2R prim- where ΔC = Ct − Ct is the Ct value of t matrix inoculum matrix ers and the probe QNIFS (Table1) following the procedure extracted viral RNA from the matrix minus the Ct inoculum described in ISO 15216-1:2017 (ISO 2017). HuNoV GI RT- value of viral RNA extracted from the inoculum, and m is qPCR was performed using QNIF4 and NV1LCR primers the slope of the virus RNA transcript standard curve. (Da Silva et al. 2007; Svraka et al. 2007) with the TM9 For the MSB method, the inoculum viral RNA levels probes (Hoehne and Schreier 2006). MNV detection was were estimated from the extraction of 100 µl of the virus performed based on RT-qPCR ORF1/ORF2 primer system diluted with 100 µl of the bead elution buffer using the RNe- developed by Baert et al. (2008). In both cases, 5 µl from asy Qiacube kit and its analysis by RT-qPCR. the RNA extracts were tested using the TaqMan Fast Virus For the ISO 15216-1:2017 method, the inoculum viral 1-Step Master Mix (Thermo Fisher). The reverse transcrip- RNA levels were estimated from the extraction of 100 µl of tion was performed at 50 °C for 20 min, and the amplifica- the virus dilution using the NucliSens miniMAG kit and its tion profile included 20 s at 95 °C, and 45 cycles of 3 s at analysis by RT-qPCR. 95 °C and 30 s at 60 °C. Virus gEq quantification was determined using standard curves generated with in vitro RNA transcripts containing RT‑qPCR Inhibition target sequences for HuNoV GI, GII, or MNV with small sequence inserts to differentiate them from the circulating The RT-qPCR inhibition from the matrix was evaluated as strains. The HuNoV GII RNA transcript UV concentration recommended in ISO 15216-1:2017 using RNA transcripts was divided by a correction factor of 1.8, established by with insert as an external amplification control (EAC). droplet digital PCR (ddPCR) (Advance Analysis Centre Briefly, non-spiked raspberry samples were extracted using Guelph University, Canada). Briefly, reverse transcriptase the MSB or the ISO 15216-1:2017 protocol as described reactions were performed in triplicate as described above above. Five microliter of RNA extract was spiked with 625 using the COG2R primer. The Automated Droplet Gen- gEq of EAC and tested by RT-qPCR using either the RNA erator (Bio-Rad, Canada) was used to generate droplets. UltraSense or the TaqMan Fast Virus 1-Step Master Mix kit Table 1 Primers, probes, and RNA transcripts used in this study Methods Primer or probe Sequence 5′–3′ References Norovirus GI  RT-qPCR, qPCR QNIF4 CGC TGG ATG CGN TTC CAT Da Silva et al. (2007)  RT-qPCR, qPCR NV1LCR CCT TAG ACG CCA TCA TCA TTT AC Svraka et al. (2007)  RT-qPCR, qPCR FAM-TM9-MGBNFQ TGG ACA GGA GAT CGC Hoehne and Schreier (2006)  PCR GISKR CTG CCC GAA TTY GTA AAT GA Kojima et al. (2002)  PCR GISKF CCA ACC CAR CCA TTR TAC A Kojima et al. (2002) Norovirus GII  RT-qPCR, qPCR QNIF2d ATG TTC AGR TGG ATG AGR TTC TCW GA Loisy et al. (2005)  RT-qPCR, qPCR FAM-QNIFS-BHQ-1 AGC ACG TGG GAG GGC GAT CG Loisy et al. (2005)  RT-qPCR, qPCR COG2R TCG ACG CCA TCT TCA TTC ACA Kageyama et al. (2003) Murine norovirus  RT-qPCR FW-ORF1/ORF2 CAC GCC ACC GAT CTG TTC TG Baert et al. (2008)  RT-qPCR RV-ORF1/ORF2 GCG CTG CGC CAT CAC TC Baert et al. (2008)  RT-qPCR FAM-ORF1/ORF2-MGBNFQ CGC TTT GGA ACA ATG Baert et al. (2008) 1 3 252 Food and Environmental Virology (2021) 13:248–258 as described above. EAC spiked in RNase-free water was the pellet after the PEG precipitation step when 30 units of used as controls. pectinase from A. niger were used as indicated in the ISO/ (ΔCt/m) RT-qPCR inhibition rate = (1 − 10 ) × 100% where TS 15216-1:2013 (data not shown). When the pectinase ΔC = Ct − Ct is the Ct value of RNA transcript from A. aculeatus was used as indicated as an alterna- t matrix water matrix spiked in RNA extracted from the matrix minus Ct value tive in the ISO 15216-1:2017 protocol version, the pellet water of RNA transcript in water and m is the slope of the virus was resuspended more easily. Nevertheless, the HuNoV transcript RNA standard curve. GII.4 recovery yields from frozen raspberries using the MSB and ISO 15216-1:2017 methods were not statisti- Limit of Detection (LOD) Calculation cally different (p = 0.366) and were 2.6% and 1.8%, respec- tively. When high (5 × 10 gEq) or low amounts of viruses The PODLOD program (v9) (Wilrich and Wilrich 2009) was (1.7 × 10 gEq) were spiked and extracted using the MSB used to calculate the LOD and LOD . method, the HuNoV GII.4 recovery yields were again 50 95 not statistically different (p = 0.855). Recovery yields for Statistical Analyses HuNoV GI.5 were similar to HuNoV GII and calculated to be 3.6%. Repeated freezing and thawing of raspberry Unless otherwise specified, all statistical analyses were per - matrices before the virus elution appeared to release more formed on log-transformed values using the independent pectin and/or the formation of agglomerates that precluded samples t test (p > 0.05). The F-test was performed to evalu- the magnetic beads handling. Consequently, frozen rasp- ate the variance. A two-way analysis of variance (ANOVA) berry matrices were classified as unfit if they had been with the Bonferroni statistical correction was used to evalu- received thawed and were not used to evaluate recovery ate the impact of the extraction and detection method on the yields. RT-qPCR inhibition (p > 0.05) (MedCalc 17.5.5). As for MNV, the recovery yield from spiked frozen raspberries using the MSB method was 2.8%, similar to the HuNoV GI and GII recovery yields (One-way ANOVA, Results p = 0.431). The tissue culture infectious dose (T CID ) of the MNV production batch was titrated at 203 gEq/ Recovery Yields TCID (CI 95% 139–297; n = 7). With a recovery yield of 2.8% using MSB, it was extrapolated that 25 g of frozen The performance of the MSB method was evaluated by raspberries spiked with 100 µl of a MNV at 53 T CID / assessing its recovery yields in comparison to the ISO ml should allow 3 MNV genomic copies in the RT-qPCR 15216-1:2017 method using IQF frozen raspberries spiked tubes to be detected 94% of the time according to the Pois- with HuNoV GII.4, GI.5, or MNV (Table 2). In our labora- son distribution. tory setting, the virus elution and concentration methods Moreover, the addition of a competitor strain did not for norovirus using the MSB had a turnaround time of have any impact on norovirus recovery yields from spiked approximately 7 h, including 1 h required for the robotic frozen raspberries using the MSB method. Indeed, the RNA extraction. The ISO 15216-1:2017 method had a HuNoV GII recovery yields in the presence or absence turnaround time of 9 h, including the inoculation and the of HuNov GI.5 were in the same range 3.4% and 4.5%, manual NucliSens miniMAG RNA extraction performed respectively (p = 0.247) (Table 3). on the second day. We had some difficulty resuspending Table 2 Detection of norovirus a Extraction method VirusSpiking level n Undiluted Diluted (1/10) in spiked frozen raspberries b c b c CtRecovery yields CtRecovery yields MSB MNV 2.4 × 10 15 26.0 ± 0.6 2.8% (1.8–3.8) nt nt GII.4 5 × 10 15 29.7 ± 1.2 2.6% (1.7–5.0) 30.4 ± 2.5 6.2% (4.2–8.2) GII.4 1.7 × 10 5 34.0 ± 1.5 5.7% (0.8–6.6) nt nt GI.5 6 × 10 15 29.2 ± 0.8 3.6% (2.4–4.9) 32.8 ± 0.8 3.0% (2.1–3.9) 15216 GII.4 5 × 10 15 29.8 ± 1.9 1.8% (0.1–3.6) 30.8 ± 1.5 9.3% (7.1–11.5) n extraction assay, nt not tested, Ct cycle threshold, MSB magnetic silica beads, 15216 ISO 15216–1:2017 Genomic equivalent copies Average Ct ± standard deviation Geometric mean (95% confidence interval) 1 3 Food and Environmental Virology (2021) 13:248–258 253 Table 3 Norovirus recovery Extraction method n Virus Competitor virus yields from frozen raspberries a b a b in the presence of competition VirusSpiking levelRecovery yields VirusSpiking levelRecovery yields 5 5 MSB 5 GII.4 3 × 10 4.5% (3.5–5.8) GI.5 10 3.0% (2.0 to 5.0) 5 GII.4 4 × 10 3.4% (1.8–6.2) nt nt nt 5 nt nt nt GI.5 10 1.8% (− 0.4 to 4.0) Raspberries spiked with or without competitor virus were extracted and tested by RT-qPCR n extraction assay, nt not tested, MSB magnetic silica beads Genomic equivalent copies Geometric mean (95% confidence interval) frozen raspberry samples were tested. The MSB RT- RT‑qPCR Inhibition qPCR LOD and LOD were 2370 gEq per 25  g (95% 95 50 CI 1542–3642) and 548 gEq per 25 g (95% CI 357–843), The ratio of diluted (1/10) to non-diluted HuNoV GII RNA respectively. recovery yields indicated a fairly high relative RT-qPCR inhibition value of 5.2 when the HuNoV GII RNA was extracted from frozen raspberries using the ISO 15216- Discussion 1:2017 method compared to a ratio value of 2.4 observed with the MSB method (Table 2). Ratio values above 4 rep- Frozen raspberries have been associated with several noro- resent more than 75% RT-qPCR inhibition. In contrast, virus outbreaks and remain a challenging food matrix for the tenfold dilution of the HuNoV GI RNA from samples virus detection. We have experienced some difficulties with extracted using the MSB method on frozen raspberries had the workflow of the ISO/TS 15216-1:2013. The limited pH no impact on the recovery yields (p = 0.214). The average range during elution, the impact of PCR inhibitors, and a undiluted HuNoV GI RNA extract Ct was not lower than the tedious PEG pellet resuspension observed with frozen rasp- diluted ones after correction for the dilution factor. berry matrices make this method poorly adapted to testing Using the HuNoV GII and GI RNA transcripts with the large number of samples CFIA diagnostic labs need insert as EAC, a statistically significant impact of the to process every year (> 500). Some of these issues were RT-qPCR kit (p < 0.001) on the RT-qPCR inhibition was resolved in the 2017 version of the method. As an alterna- observed (Fig.  2). The average RNA UltraSense HuNoV tive, we developed a new food virus extraction method based GII RT-qPCR inhibition percentages measured using the on magnetic silica beads (MSB) which was better adapted RNA extracted with the ISO 15216-1:2017 and the MSB to our laboratory setting. The MSB approach to elute and approaches were 58% (95% CI 32–83) (n = 12 replicates) concentrate norovirus from raspberries is based on a stra- and 56% (95% CI 41–71) (n = 20), respectively. However, tegic use of electrostatic interactions at different steps of when testing the HuNoV GII EAC amplification using the the protocol. The variations of the virus surface charge at TaqMan Fast Virus 1-Step, the average RT-qPCR inhibi- different pH levels are illustrated in Fig.  1. According to the tion percentages calculated with the RNA extracts using the manufacturer, the fine magnetic silica beads have a small ISO 15216-1:2017, and the MSB protocols were decreased size (420 nm) and have a negative zeta potential in water to 24% (95% CI 7–42) (n = 12) and 1% (95% CI 7 to 9) (− 37 mV). On the other hand, silica particles are prone to (n = 20), respectively. Similar EAC inhibition levels were aggregation in the presence of salts (Metin et al. 2011). In observed using the HuNoV GI RT-qPCR. RT-qPCR inhi- the MSB method, a non-ionic buffer (Bis–Tris-Propane) bition values were above the 75% when tested with the was used, and salts were avoided in the buffer solutions to RNA UltraSense kit for 46% and 25% of the RNA samples reduce the ionic strength and to limit the aggregation of extracted with the ISO 15216-1:2017 and the MSB methods, magnetic silica particles. The effect of pH and charge of the respectively. There was no impact of the extraction method norovirus were also taken into account. According to Da on the RT-qPCR inhibition (p = 0.110). Silva et al. (2011), HuNoV GI Virus-like particles (VLP) are prone to adhesion onto silica below the VLP isoelectric Limit of Detection point whereas its attachment is reduced at higher pH at low salt concentrations. The limit of detection of the MSB method was evaluated It is difficult to compare method performance based by RT-qPCR using RNA extracted from frozen raspberries on the literature or laboratory reports (Li et al. 2018). In spiked with HuNoV GII.4 strain CFIA-FVR-019 (Supple- addition to the extraction protocols, variations in recovery mentary Fig. SF1). Overall, 70 spiked and 14 non-spiked 1 3 254 Food and Environmental Virology (2021) 13:248–258 Fig. 2 RT-qPCR inhibition (A) HuNoV GI percentages in frozen raspberry RNA extracts. A box plot of RT-qPCR inhibition evalu- ated using a HuNoV GI and b HuNoV GII external control RNA amplification is shown. The absence of RT-qPCR inhibition or enhancement should read as 0. The error bars represent the 95% confidence intervals of the mean. The horizontal dotted line represents 75% inhibition -20 -40 -60 -80 -100 RNA UltraSense TaqMan Fast Virus 1-Step RT-qPCR kit (B) HuNoV GII -20 -40 -60 -80 -100 RNA UltraSense TaqMan Fast Virus 1-Step RT-qPCR kit Extraction method ISO 15216-1:2017 MSB yields could be associated to the inoculated virus prepara- Norwalk strain was estimated at 1320 (95% CI 440–3760) tion, clarification, filtration, the virus strain, its integrity, as genomic equivalent (gEq) in serogroups O and A (Atmar well as the spiking conditions. The method should be fit for et al. 2014). The HuNoV GII L OD reported in this study 3 3 its intended purpose and detects HuNoV in frozen raspber- using the MSB approach was higher (2 × 10 vs. 0.7 × 10 ries at level equivalent to the 50% human infectious dose gEq per 25 g) than the ones reported by another study using (HID50). The HID50 of the HuNoV in susceptible healthy the ISO 15216-1:2017 (Li et al. 2018). On the other hand, adults varies with the serogroup. The HID50 of the HuNoV recovery yields obtained for undiluted HuNoV GI and 1 3 RT-qPCR inhibition(%) RT-qPCR Inhibition(%) Food and Environmental Virology (2021) 13:248–258 255 HuNoV GII using the MSB method were in the same range assays increases the burden of testing and could double (2.6–5.7%) as the ISO 15216-1:2017 reported by Fraisse reported prevalence estimates if this factor was not taken et al. (2017). While achieving the lowest detection limit is a into account. The ISO 15216-1:2017 recommends a larger major goal of extraction methodologies, its application also elution volume (100 µl vs 50 µl) and requires testing a ten- influences method selection. For instance, methodologies fold diluted RNA extract as well as testing EAC to avoid this that could discriminate between inactivated and infectious issue (ISO 2017). Previously, a European survey reported virus are required to avoid overestimating the viral infectiv- that positive frozen raspberries contaminated with HuNoV ity. Viability RT-qPCR based on viral integrity treatment is were only detected using tenfold diluted RNA when RNA a promising approach to improve risk assessment of positive was extracted with the ISO/TS 15216-1:2013 method RT-qPCR detection results (Chen et al. 2020). The integrity (Loutreul et al. 2014). Dilution of the soft fruit RNA extract of the virus following its elution and concentration might decreases the impact of PCR inhibitors in the RT-qPCR. varied between different viral RNA extraction methodolo- However, RNA extract dilution also impacts the capacity to gies and requires further investigation. detect the virus present at trace levels (Fraisse et al. 2017). Raspberries contain high levels of heteropolysaccha- Until improvements are shown regarding the recovery yields, rides, such as pectin, which have a major impact on the viral strategies that decrease the impact of PCR inhibitors from extraction process. The pectin appears to form a strong gel frozen raspberry RNA extracts should be encouraged. with calcium at a pH close to 5 which interferes with the Different PCR inhibitor removal kits or the use of digi- elution process (Han et al. 2017). Pectinase treatment, lower tal PCR have improved the ISO 15216-1: 2017 recovery pH, and low calcium conditions were required to avoid bead yields (Fraisse et al. 2017; Bartsch et al. 2016, 2018). For agglomeration when performing the MSB elution from fro- instance, an additional RNA purification step using either the zen raspberries. The current low pH extraction process was MobiSpin column or the OneS tep PCR Inhibitor Removal effective with a limited set of matrices. Kit (Zymo research) were added with the ISO/TS 15216- In addition to pectin, the extraction of HuNoV from fro- 1:2013 protocol to remove RT-PCR inhibitors from frozen zen raspberries presented other challenges. Raspberries strawberries and improve the detection limit (Bartsch et al. contain multiple components that are co-extracted with the 2016, 2018). Alternative extraction approaches could also viral genome and can impact its molecular detection. High reduce the amount of RT-PCR inhibitors. With Bovine Noro- levels of polyphenols (e.g., anthocyanin, flavonol, ellagi- virus spiked on frozen raspberries, Sun et al. (2019) have tannin, proanthocyanidin, phenolic acids, tannic acid) can reported the absence of RT-PCR inhibition using a direct act as PCR inhibitors (reviewed in Schrader et al. (2012)). lysis approach combined with either RNA filtration using the Heteropolysaccharides can disturb the RT and PCR enzy- MobiSpin column or digital PCR detection. The results in matic process by mimicking the structure of nucleic acids. this study indicated that the RT-qPCR inhibition is also influ- Phenolic compounds may cross‐link RNA under oxidizing enced by the selected detection method. The RNA UltraSense conditions and could degrade DNA polymerases. To reduce kit is described in the ISO 15216-2017 method, but it is not a the presence of PCR inhibitors, the MSB method includes a requirement for the method. The RNA UltraSense kit might treatment with insoluble polyvinylpolypyrrolidone that was be suitable for some food matrices included in the scope of reported to prevent polyphenol oxidation and subsequent this method that are associated to low RT-qPCR inhibition. binding to nucleic acids when extracting total RNA from However, the TaqMan Fast Virus 1-Step results indicate that raspberries (Jones et al. 1997). the UltraSense kit is not the most appropriate detection kit to High levels of RT-qPCR inhibition have been reported test RNA extracted from frozen raspberries. The impact of by several groups that evaluated PEG-derived extraction PCR inhibitors from frozen raspberry RNA extracts on other methods including ISO/TS 15216-1:2013 and ISO 15216- commercial RT-qPCR kit was not explored. Commercial 1:2017 with frozen raspberries (De Keuckelaere et al. 2015; RT-qPCR kits use die ff rent proprietary bue ff rs and additives Fraisse et al. 2017; Summa and Maunula 2018; Summa et al. to reduce the impact of PCR inhibitors. The extracted PCR 2012). With frozen raspberries RNA extracted using the inhibitors vary with the type of food matrix. Consequently, ISO 15216-1:2017, RT-qPCR inhibition was estimated at the impact of inhibitors on commercial RT-qPCR kits should 93.8% ± 2.5% using an external RNA control (Fraisse et al. be tested for each type of food matrix. 2017). According to the ISO 15216-1:2017, negative results Several groups have used various forms of the ISO obtained in presence of RT-qPCR inhibition levels > 75% 15216 method to recover HuNov from frozen raspberries. are not valid. The observed range of recovery yields from 1 to 6% range Meanwhile in this study, close to 50% of the undiluted could certainly impact the reported prevalence. Neverthe- samples extracted following the ISO 15216-1:2017 protocol less, Loutreul et al. (2014) reported a prevalence of 16.7% and tested using the RNA UltraSense detection kit presented (n = 162) for HuNov GI in frozen raspberries from Ser- more than 75% inhibition. Such a level of inconclusive bia, Chile, Bulgaria, Poland, and France. In the UK, 3.6% 1 3 256 Food and Environmental Virology (2021) 13:248–258 (n = 274) of frozen raspberry samples sold at retail were pos- Compliance with Ethical Standards itive for HuNoV (Cook et al. 2019). However, this group was Conflict of interest The authors have no conflict of interest to declare. not able to differentiate the sequence of the HuNoV detected from their EAC. Gao et al. (2019) reported that 9.2% and Open Access This article is licensed under a Creative Commons Attri- 13% of frozen raspberries from Heilongjiang Province in bution 4.0 International License, which permits use, sharing, adapta- China were positive to HuNoV in 2016 and 2017, respec- tion, distribution and reproduction in any medium or format, as long tively. Meanwhile, all export samples were negative. They as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes used a replicate Ct threshold for positive results, which could were made. The images or other third party material in this article are lower prevalence estimate, and did not analyze the RT-qPCR included in the article’s Creative Commons licence, unless indicated inhibition level. Maunula et al. (2013) did not detect HuNov otherwise in a credit line to the material. If material is not included in in frozen raspberry samples (0/39) from point of sale of four the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will European countries, but did nd fi some HuNov GII in irrigat - need to obtain permission directly from the copyright holder. To view a ing water from berry production sites (2/56). They used high copy of this licence, visit http://creativ ecommons .or g/licenses/b y/4.0/. pectinase concentration and RNA extract elution volume (300 µl) for berries but did not report any RT-qPCR inhibi- tion. The true prevalence as well as the extraction method recovery yields and the presence of PCR inhibitors could References play a role in resolving some of the discrepancies observed Andrade, A. L., Souza, D. M., Pereira, M. C., Fabris, J. D., & in terms of prevalence between those groups. Still, the con- Domingues, R. Z. (2009). Synthesis and characterization of mag- tamination levels in some of these surveys remains relatively netic nanoparticles coated with silica through a sol-gel approach. important from a risk analysis perspective. Cerâmica, 55, 420–424. Atmar, R. L., Opekun, A. R., Gilger, M. A., Estes, M. K., Crawford, S. E., Neill, F. H., et al. (2014). Determination of the 50% human infectious dose for Norwalk virus. Journal of Infectious Diseases, 209(7), 1016–1022. https ://doi.org/10.1093/infdi s/jit62 0. Conclusion Baert, L., Wobus, C. E., Van Coillie, E., Thackray, L. B., Debevere, J., & Uyttendaele, M. (2008). Detection of murine norovirus 1 by using plaque assay, transfection assay, and real-time reverse tran- A method based on magnetic silica beads to extract HuNoV scription-PCR before and after heat exposure. Applied and Envi- virus RNA from frozen raspberries was developed. The ronment Microbiology, 74(2), 543–546. https ://doi.org/10.1128/ MSB method performance was similar to the reference AEM.01039 -07. method ISO 15216-1:2017. The influence of RT-qPCR Bartsch, C., Hoper, D., Made, D., & Johne, R. (2018). Analysis of frozen strawberries involved in a large norovirus gastroenteri- inhibitors extracted using both methods was reduced using tis outbreak using next generation sequencing and digital PCR. an alternative RT-qPCR detection kit (TaqMan Fast Virus Food Microbiology, 76, 390–395. https ://doi.or g/10.1016/j. 1-Step Master Mix) and condition. In future, the reduction of fm.2018.06.019. the RT-qPCR inhibitors which impact the HuNoV detection Bartsch, C., Szabo, K., Dinh-Thanh, M., Schrader, C., Trojnar, E., & Johne, R. (2016). Comparison and optimization of detection meth- results should reduce the number of inconclusive assays and ods for noroviruses in frozen strawberries containing different influence prevalence estimates. amounts of RT-PCR inhibitors. Food Microbiology, 60, 124–130. https ://doi.org/10.1016/j.fm.2016.07.005. Supplementary Information The online version contains supplemen- Boqvist, S., Soderqvist, K., & Vagsholm, I. (2018). Food safety chal- tary material available at https:// doi.org/10.1007/s12560-02 1-09466-0 . lenges and one health within Europe. Acta Veterinaria Scandi- navica, 60(1), 1. https ://doi.org/10.1186/s1302 8-017-0355-3. Acknowledgements The authors acknowledge the work of analysts Bozkurt, H., Phan-Thien, K. Y., van Ogtrop, F., Bell, T., & McCo- and scientist from the CFIA Food Virology National Reference Center nchie, R. (2020). Outbreaks, occurrence, and control of noro- (FVNRC) for the production of RNA transcript standards (MO, VL, virus and hepatitis a virus contamination in berries: A review. YL). Names of specific vendors, manufacturers, or products are Critical Reviews in Food Science and Nutrition. https ://doi. included for informational purposes only and do not imply endorse- org/10.1080/10408 398.2020.17193 83. ment by authors. Brinkman, N. E., Haffler, T. D., Cashdollar, J. L., & Rhodes, E. R. (2013). Evaluation of methods using celite to concentrate noro- virus, adenovirus and enterovirus from wastewater. Journal of Author contributions Conceptualization: PR and LD. Funding acquisi- Virological Methods, 193(1), 140–146. https ://doi.org/10.1016/j. tion: PR and LD. Supervision: PR, Methodology: PR, SP and AP. Data jviro met.2013.05.014. curation: PR and SP. Formal analysis: PR. Writing original draft: PR. Chen, J., Wua, X., Sánchezb, G., & Randazzob, W. (2020). Viabil- Writing review & editing: PR, AP, SP and LD. ity RT-qPCR to detect potentially infectious enteric viruses on heat processed berries. Food Control, 107, 106818. https ://doi. Funding This work was supported by the Canadian Food Inspec- org/10.1016/j.foodc ont.2019.10681 8. tion Agency Research & Partnership Strategy (STH-F-1306 and Chhabra, P., de Graaf, M., Parra, G. I., Chan, M. C., Green, K., STH-F-1607). Martella, V., et al. (2019). Updated classification of norovirus 1 3 Food and Environmental Virology (2021) 13:248–258 257 genogroups and genotypes. Journal of General Virology, 100(10), Kageyama, T., Kojima, S., Shinohara, M., Uchida, K., Fukushi, 1393–1406. https ://doi.org/10.1099/jgv.0.00131 8. S., Hoshino, F. B., et  al. (2003). Broadly reactive and highly Cook, N., Knight, A., & Richards, G. P. (2016). Persistence and elimi- sensitive assay for Norwalk-like viruses based on real-time nation of human norovirus in food and on food contact surfaces: quantitative reverse transcription-PCR. Journal of Clini- A critical review. Journal of Food Protection, 79(7), 1273–1294. cal Microbiology, 41(4), 1548–1557. https ://doi.or g/10.1128/ https ://doi.org/10.4315/0362-028X.JFP-15-570. jcm.41.4.1548-1557.2003. Cook, N., Williams, L., & D’Agostino, M. (2019). Prevalence of noro- Kojima, S., Kageyama, T., Fukushi, S., Hoshino, F. B., Shinohara, M., virus in produce sold at retail in the United Kingdom. Food Micro- Uchida, K., et al. (2002). Genogroup-specific PCR primers for biology, 79, 85–89. https ://doi.org/10.1016/j.fm.2018.12.003. detection of Norwalk-like viruses. Journal of Virological Meth- Da Silva, A. K., Kavanagh, O. V., Estes, M. K., & Elimelech, M. ods, 100(1–2), 107–114. (2011). Adsorption and aggregation properties of norovirus GI Li, D., Butot, S., Zuber, S., & Uyttendaele, M. (2018). Monitoring of and GII virus-like particles demonstrate differing responses to foodborne viruses in berries and considerations on the use of RT- solution chemistry. Environmental Science and Technology, 45(2), PCR methods in surveillance. Food Control, 89, 235–240. https 520–526. https ://doi.org/10.1021/es102 368d.://doi.org/10.1016/j.foodc ont.2018.02.024. Da Silva, A. K., Le Saux, J. C., Parnaudeau, S., Pommepuy, M., Elime- Loisy, F., Atmar, R. L., Guillon, P., Le Cann, P., Pommepuy, M., & Le lech, M., & Le Guyader, F. S. (2007). Evaluation of removal of Guyader, F. S. (2005). Real-time RT-PCR for norovirus screening noroviruses during wastewater treatment, using real-time reverse in shellfish. Journal of Virological Methods, 123(1), 1–7. https :// transcription-PCR: Different behaviors of genogroups I and II. doi.org/10.1016/j.jviro met.2004.08.023. Applied and Environmental Microbiology, 73(24), 7891–7897. Loutreul, J., Cazeaux, C., Levert, D., Nicolas, A., Vautier, S., Le Sau- https ://doi.org/10.1128/AEM.01428 -07. vage, A. L., et al. (2014). Prevalence of human noroviruses in De Keuckelaere, A., Li, D., Deliens, B., Stals, A., & Uyttendaele, M. frozen marketed shellfish, red fruits and fresh vegetables. Food (2015). Batch testing for noroviruses in frozen raspberries. Inter- Environmental Virology, 6(3), 157–168. https ://doi.org/10.1007/ national Journal of Food Microbiology, 192, 43–50. https ://doi.s1256 0-014-9150-8. org/10.1016/j.ijfoo dmicr o.2014.09.024. Maunula, L., Kaupke, A., Vasickova, P., Soderberg, K., Kozyra, I., Fraisse, A., Coudray-Meunier, C., Martin-Latil, S., Hennechart- Lazic, S., et al. (2013). Tracing enteric viruses in the European Collette, C., Delannoy, S., Fach, P., et al. (2017). Digital RT-PCR berry fruit supply chain. International Journal of Food Micro- method for hepatitis A virus and norovirus quantification in soft biology, 167(2), 177–185. h t t ps : / / do i . o rg/ 1 0 . 1 0 16 / j . i j f oo d m i c r berries. International Journal of Food Microbiology, 243, 36–45. o.2013.09.003. https ://doi.org/10.1016/j.ijfoo dmicr o.2016.11.022. Metin, C. O., Lake, L. W., Miranda, C. R., & Nguyen, Q. P. (2011). Gao, X., Wang, Z., Wang, Y., Liu, Z., Guan, X., Ma, Y., et al. (2019). Stability of aqueous silica nanoparticle dispersions. Journal of Surveillance of norovirus contamination in commercial fresh/fro- Nanoparticle Research, 13(2), 839–850. https ://doi.org/10.1007/ zen berries from Heilongjiang Province, China, using a TaqMan s1105 1-010-0085-1. real-time RT-PCR assay. Food Microbiology, 82, 119–126. https Morales-Rayas, R., Wolffs, P. F., & Griffiths, M. W. (2010). Simultane- ://doi.org/10.1016/j.fm.2019.01.017. ous separation and detection of hepatitis A virus and norovirus in Gonzalez-Hernandez, M. B., Bragazzi Cunha, J., & Wobus, C. E. produce. International Journal of Food Microbiology, 139(1–2), (2012). Plaque assay for murine norovirus. Journal of Visualized 48–55. https ://doi.org/10.1016/j.ijfoo dmicr o.2010.02.011. Experiments, 66, e4297. https ://doi.org/10.3791/4297. Nasheri, N., Vester, A., & Petronella, N. (2019). Foodborne viral out- Government of Canada. (2018). National enteric surveillance program breaks associated with frozen produce. Epidemiology and Infec- annual summary 2017. Guelph: Public Health Agency of Canada. tion, 147, e291. https ://doi.org/10.1017/S0950 26881 90017 91. Han, W., Meng, Y., Hu, C., Dong, G., Qu, Y., Deng, H., et al. (2017). Scherer, K., Johne, R., Schrader, C., Ellerbroek, L., Schulenburg, J., Mathematical model of Ca2+ concentration, pH, pectin con- & Klein, G. (2010). Comparison of two extraction methods for centration and soluble solids (sucrose) on the gelation of low viruses in food and application in a norovirus gastroenteritis out- methoxyl pectin. Food Hydrocolloids, 66, 37–48. https ://doi. break. Journal of Virological Methods, 169(1), 22–27. https://doi. org/10.1016/j.foodh yd.2016.12.011.org/10.1016/j.jviro met.2010.06.008. Hoehne, M., & Schreier, E. (2006). Detection of Norovirus genogroup Schrader, C., Schielke, A., Ellerbroek, L., & Johne, R. (2012). PCR I and II by multiplex real-time RT-PCR using a 3’-minor groove inhibitors—Occurrence, properties and removal. Journal of binder-DNA probe. BMC Infectious Diseases, 6, 69. https ://doi. Applied Microbiology, 113(5), 1014–1026. https ://doi.org/10.11 org/10.1186/1471-2334-6-69.11/j.1365-2672.2012.05384 .x. Houde, A., Leblanc, D., Poitras, E., Ward, P., Brassard, J., Simard, C., Stals, A., Baert, L., Van Coillie, E., & Uyttendaele, M. (2012). Extrac- et al. (2006). Comparative evaluation of RT-PCR, nucleic acid tion of food-borne viruses from food samples: A review. Interna- sequence-based amplification (NASBA) and real-time RT-PCR tional Journal of Food Microbiology, 153(1–2), 1–9. https ://doi. for detection of noroviruses in faecal material. Journal of Viro-org/10.1016/j.ijfoo dmicr o.2011.10.014. logical Methods, 135(2), 163–172. https://doi.or g/10.1016/j.jviro Summa, M., & Maunula, L. (2018). Rapid detection of human noro- met.2006.03.001. virus in frozen raspberries. Food Environmental Virology, 10(1), ISO. (2017). Horizontal method for determination of Hepatitis A virus 51–60. https ://doi.org/10.1007/s1256 0-017-9321-5. and norovirus using real-time RT-PCR–Part 1: Method for quan- Summa, M., von Bonsdorff, C. H., & Maunula, L. (2012). Evalua- tification. In Microbiology of the food chain (Vol ISO 15216– tion of four virus recovery methods for detecting noroviruses on 1:2017). Geneva: International Organization for Standardization. fresh lettuce, sliced ham, and frozen raspberries. Journal of Viro- Jones, C. S., Iannetta, P. P., Woodhead, M., Davies, H. V., McNicol, R. logical Methods, 183(2), 154–160. https://doi.or g/10.1016/j.jviro J., & Taylor, M. A. (1997). The isolation of RNA from raspberry met.2012.04.006. (Rubus idaeus) fruit. Molecular Biotechnology, 8(3), 219–221. Sun, B., Bosch, A., & Myrmel, M. (2019). Extended direct lysis https ://doi.org/10.1007/BF027 60775 . method for virus detection on berries including droplet digital Júnior, J. A. A., & Baldo, J. B. (2014). The behavior of zeta potential of RT-PCR or real time RT-PCR with reduced influence from inhibi- silica suspensions. New Journal of Glass and Ceramics, 4, 29–37. tors. Journal of Virological Methods, 271, 113638. https ://doi. https ://doi.org/10.4236/njgc.2014.42004 .org/10.1016/j.jviro met.2019.04.004. 1 3 258 Food and Environmental Virology (2021) 13:248–258 Svraka, S., Duizer, E., Vennema, H., de Bruin, E., van der Veer, B., Wilrich, C., & Wilrich, P. T. (2009). Estimation of the POD func- Dorresteijn, B., et al. (2007). Etiological role of viruses in out- tion and the LOD of a qualitative microbiological measurement breaks of acute gastroenteritis in The Netherlands from 1994 method. Journal of AOAC International, 92(6), 1763–1772. through 2005. Journal of Clinical Microbiology, 45(5), 1389– 1394. https ://doi.org/10.1128/JCM.02305 -06. Publisher’s Note Springer Nature remains neutral with regard to Vinje, J. (2015). Advances in laboratory methods for detection and jurisdictional claims in published maps and institutional affiliations. typing of norovirus. Journal of Clinical Microbiology, 53(2), 373–381. https ://doi.org/10.1128/JCM.01535 -14. 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Food and Environmental Virology Springer Journals

Norovirus Extraction from Frozen Raspberries Using Magnetic Silica Beads

Loading next page...
 
/lp/springer-journals/norovirus-extraction-from-frozen-raspberries-using-magnetic-silica-tE4tkes0L0
Publisher
Springer Journals
Copyright
Copyright © Crown 2021
ISSN
1867-0334
eISSN
1867-0342
DOI
10.1007/s12560-021-09466-0
Publisher site
See Article on Publisher Site

Abstract

Human noroviruses (HuNoV) are among the main causes of acute gastroenteritis worldwide. Frozen raspberries have been linked to several HuNoV food-related outbreaks. However, the extraction of HuNoV RNA from frozen raspberries remains challenging. Recovery yields are low, and real-time quantitative reverse transcriptase PCR (RT-qPCR) inhibitors limit the sensitivity of the detection methodologies. A new approach using fine magnetic silica beads was developed for the extraction of HuNoV spiked on frozen raspberries. Relatively low recovery yields were observed with both the magnetic silica bead and the reference ISO 15216-1:2017 methods. High RT-qPCR inhibition was observed with the ISO 15216-1:2017 recommended amplification kit but could be reduced by using an alternative kit. Reducing RT-qPCR inhibition is important to limit the number of inconclusive HuNoV assays thus increasing the capacity to assess the HuNoV prevalence in frozen raspberries. Keywords Norovirus · Raspberries · RNA extraction · Silica · RT-qPCR Introduction with contaminated frozen product between 2008 and 2018 (Nasheri et al. 2019). Human Norovirus (HuNoV) is one of the leading causes of Noroviruses are small (27–40 nm) non-enveloped single- food-related illnesses in developed countries. HuNoV rep- stranded RNA viruses that are transmitted mainly via the resents 58% of all reported foodborne outbreaks of known fecal–oral route. Noroviruses belong to a genetically diverse etiology in the United States (Vinje 2015). About 300–400 group of viruses of the Caliciviridae family. There are 10 outbreaks of HuNoV are reported to the National Enteric distinct norovirus genetic groups (Chhabra et  al. 2019). Surveillance Program of the Public Health Agency of Can- HuNoV GI and GII are the most prevalent genogroups asso- ada each year (Government of Canada 2018). Forty-six berry ciated with outbreaks. Noroviruses can persist in an infec- outbreaks associated to HuNoV contamination with 15,827 tious state for prolonged periods of time in the environment, cases were reported globally between 1983 and 2018 (Boz- in water, and in food (reviewed in Cook et al. (2016)). They kurt et al. 2020). In the European Union, contaminated fro- can also withstand a broad pH range (pH 2–9). In cold tem- zen red fruits were shown to represent an important cause peratures, noroviruses can stay infectious for years. How- of HuNoV outbreaks (Boqvist et al. 2018). Contaminated ever, they are inactivated by cooking. The greatest risk of frozen raspberries are the most common sources (Bozkurt a foodborne HuNoV infection arises from the consumption et  al. 2020). Frozen raspberries were implicated as food of contaminated food such as fresh or frozen fruits, leafy vehicles in 33 of the 40 reported outbreak events associated vegetables, oysters, and drinking water. Detection of HuNoV relies on the extraction of RNA and real-time reverse transcriptase PCR amplification method- ologies (Vinje 2015). Prior to detection, minute amounts of * Philippe Raymond viruses must be extracted from food matrices. Methods for philippe.raymond@canada.ca concentrating noroviruses extracted from contaminated food Canadian Food Inspection Laboratory (CFIA), St. Hyacinthe are based mainly on ultrafiltration, ultracentrifugation, cati- Laboratory, Food Virology, Saint Hyacinthe, QC, Canada onic separation, and polyethylene glycol (PEG) precipita- Agriculture and Agri-Food Canada (AAFC), St. Hyacinthe tion. These methods are associated with high variability and Research and Development Centre, Saint Hyacinthe, QC, relatively low recovery of HuNoV (Summa and Maunula Canada Vol:.(1234567890) 1 3 Food and Environmental Virology (2021) 13:248–258 249 2018). The ISO/TS 15216-1:2013 technical specification and for Disease Control (BCCDC). The preparation of HuNoV the subsequent ISO 15216-1:2017 standard were published from clarified 10% stool samples was adapted from Houde to provide a reference method for the quantitative extrac- et al. (2006). Briefly, the stool samples were diluted in 1× tion and detection of HuNoV from various food products. phosphate-buffered saline (PBS), pH 7.4 (Gibco, Canada) to The extraction from soft fruit, leaf, stem, and bulb vegeta- obtain a 10% suspension and were homogenized by vigor- bles is based on PEG precipitation of the virus. However, ous agitation. Suspensions were clarified by centrifugation PCR inhibitors are frequently reported when PEG-based (20,000×g, 15 min, 4 °C) and kept frozen at − 80 °C. approaches are applied to soft fruits and can lead to false- negative results (Summa and Maunula 2018). PCR inhibi- Frozen Raspberry Samples tors are also reported when using ultrafiltration, Cat-Floc precipitation, and immunomagnetic bead-based extraction Frozen raspberries, from bags labeled as whole individu- methods as well, reducing the norovirus recovery yields ally quick frozen (IQF) collected at local stores, were sub- substantially (Summa et al. 2012). Extraction methods with divided in 25 g samples and used for artificial contamination low recovery yields increase the probability of false-negative experiments. results. Assays with more than 75% real-time quantitative reverse transcriptase PCR (RT-qPCR) inhibition should be Artificial Contamination of Frozen Raspberries considered inconclusive (ISO 2017). Other groups have used cationic beads or filters at high Aliquots of clarified 10% stools of HuNoV were vortexed for pH for norovirus adsorption and elution and have had vari- 2 s and diluted in PBS to the genomic equivalent copy (gEq) ous success rates (Morales-Rayas et al. 2010; Scherer et al. level needed in 100 µl per sample. To estimate the recovery 3 6 2010; Stals et al. 2012). The recombinant HuNoV capsid yields, between 10 and 10 gEq of HuNoV GI.5, GII.4, and protein VPI isoelectric point was reported to be pH 5 (Da MNV were inoculated. To estimate the recovery yields in the Silva et al. 2011). Accordingly, norovirus should be posi- presence of a competitor strain, 10 gEq HuNoV GII.4 and tively charged below its capsid isoelectric point. Relatively 10 gEq GI.5 were inoculated simultaneously. To estimate high recovery yields (68%) for HuNoV GII from wastewa- the limit of detection (LOD), the HuNoV GII.4 stocks were 5 2 ter were reported using celite, a siliceous rock powder, at serially diluted to 10 –10 gEq per 100 µl. pH 4 (Brinkman et al. 2013). Silica surfaces are covered Frozen raspberries were spiked with the diluted viral by silanol groups which can exist in different states. From suspensions on the surface of the food matrices (25 g) in a previous studies, it is expected that crystalline and vitreous Whirl–Pak filter bag (VWR, Canada), then left to air dry silica surfaces should maintain a negative charge above pH 30 min in a biosafety cabinet. A frozen raspberries sample, 2 and 3, respectively (Júnior and Baldo 2014). Andrade et al. with no virus added, was included in each extraction batch as (2009) reported an isoelectric point of 2.3 for silica-coated a negative control. The amount of virus in the 100 µl inocu- magnetic beads in suspension in KCl 1 mM. lum was assessed in parallel by extracting the total RNA In this study, we present the performance of a new mag- using the RNeasy kit (QIAGEN) followed by a RT-qPCR netic silica bead (MSB) methodology for the extraction of assay as described below. norovirus. We compared the performance of this approach to the ISO 15216-1:2017 method with a focus on the RT-qPCR Virus Elution and RNA Extraction inhibition associated to the extraction and detection. Reduc- ing RT-qPCR inhibition is important to limit the number of Viral RNA was extracted from frozen raspberries using inconclusive assays. the MSB method by proceeding with the following steps (Fig. 1): Step 1 The virus was eluted from the matrix by adding Methods 40 ml of elution buffer made of 150 mM Bis–Tris-Propane (Sigma-Aldrich, Canada) pH 8 to the sample in the filter Virus Stocks bag then closing the bag and shaking it at about 55 RPM for 30 min at room temperature (RT) using an orbital shaker. Murine norovirus-1 (MNV) was provided by Dr. H. Virgin Using a 25 ml borosilicate glass pipet, the eluate was trans- from Washington University (St. Louis, MO, USA). MNV ferred to a 50-ml conical centrifuges tubes. was propagated in the RAW 264.7 cell line as previously Step 2 The eluate was clarified by centrifugation at described and viral stocks were titrated by plaque assays 3500×g for 10 min. Thirty units of Aspergillus niger pec- (Gonzalez-Hernandez et al. 2012). HuNoV-positive stool tinase (Sigma-Aldrich) were added to the clarified eluate samples HuNoV GI.5 (CFIA-FVR-022) and GII.4 (CFIA- followed by a 30 min incubation at 37 °C with shaking at FVR-019) were provided by the British Columbia Center approximately 70 RPM. Meanwhile, a volume of 100 µl per 1 3 250 Food and Environmental Virology (2021) 13:248–258 Fig. 1 Norovirus extraction and concentration method using the magnetic silica beads (MSB) approach. Viruses (blue dots) are purified from matrix and soil contaminants (brown and red dots) using magnetic silica beads (gray dots) prior to the RNA extraction sample of AccuNanobeads (Bioneer, CA, USA, average size the elution buffer. The first and second elutions were range 300–700 nm) was vortexed for 3 min. The AccuNano- performed by agitation at about 1050 RPM in a Thermo- beads storage buffer was removed using a magnet, and the mixer R (Eppendorf, Canada) at RT for 10 min and 1 min, beads were resuspended in 1 ml per sample of the elution respectively. Virus eluates were immediately transferred buffer containing 100 µg/ml Pluronic F-127 (Anatrace, OH, to a 1.5 ml microtube and combined with 500 µl buffer USA) and vortexed for a few seconds. RLT plus β-mercaptoethanol from the RNeasy QIAcube Step 3 The bead suspension was added to the clarified kit (QIAGEN) or 10  µl of 2  M Dithiothreitol (Sigma- eluate and vortexed 3 s. Two milliliters of 200 mM ascor- Aldrich). Five microliters of RNA carrier (1 µg/ml) (QIA- bic acid (Sigma-Aldrich) and 3.5 ml of 200 mM malic acid GEN) and 140 µl of the polyvinylpolypyrrolidone (PVPP; (Sigma-Aldrich) were added to each sample, and the tubes Sigma-Aldrich) suspension (2% v/v final) were added to were vortexed another 3 s. The pH was lowered between pH each sample, and the samples were vortexed briefly. Sam- 2 and 3 with the addition of approximately 850 µl of 6 N ples were centrifuged at 10,000×g minimum for 5 min to HCl. The tubes were then mixed using a Dynal rotary shaker remove the PVPP. 94701 (Thermo Fisher, Canada) set between 19 and 22, for Step 6 Total RNA was extracted using the RNeasy QIA- 10 to 60 min at RT. cube kit supplemented with DNase I in the QIAcube plat- Step 4 The solution was separated from the beads using a form as described by the manufacturer (QIAGEN). The magnetic rack and removed by decantation. The remaining RNA was eluted from the spin column membrane with 50 µl beads were washed with 1 ml of washing bue ff r composed of of RNase-free water from the kit to which 40 units of the 17.4 mM Bis–Tris-Propane buffer pH 7 with 8.7 mM NaCl, RNasin Plus RNase inhibitor (Fisher Scientific, Canada) 0.87 mM CaCl , 10.9 mM ascorbic acid, and 15.2 mM malic were added. It was then stored at − 80 °C. acid. As a reference method, the ISO 15216-1:2017 method for Step 5 The viruses were eluted twice from the beads soft fruit samples (ISO 2017) was also used for the extrac- with 100 µl of the bead elution buffer composed of 45 mM tion of the viruses from spiked frozen raspberries. The pec- Bis–Tris-propane pH 9, 0.01% Tween 20, and 50  mM tinase from Aspergillus aculeatus (1140 U) was used. The EDTA (Sigma-Aldrich). Beads were vortexed 3  s with NucliSens miniMAG kit (Biomérieux, Canada) was used to 1 3 Food and Environmental Virology (2021) 13:248–258 251 extract RNA in 50 µl elution buffer following the manufac- QPCRs were performed as described below using the C1000 turer’s recommendations. touch Thermal cycler and ddPCR Multiple Supermix (Bio- Rad). Plates were read using the QX200 Droplet Reader RT‑qPCR (Bio-Rad). RT-qPCR assays were performed using 5 µl of diluted (1/10) Recovery Yield Calculation or non-diluted RNA extracts, using either the Mx3005P system (Stratagene, CA, USA) or the Quantstudio 6 system The recovery yields associated with the virus elution and (Thermo Fisher). Diluted RNA extracts were prepared using concentration steps were estimated using the cycle thresholds RNase-free water. Unless otherwise specified, HuNoV GII (ΔCt/m) (Ct) variation. The virus recovery yields = 10 × 100%, RT-qPCR was performed using QNIF2d and COG2R prim- where ΔC = Ct − Ct is the Ct value of t matrix inoculum matrix ers and the probe QNIFS (Table1) following the procedure extracted viral RNA from the matrix minus the Ct inoculum described in ISO 15216-1:2017 (ISO 2017). HuNoV GI RT- value of viral RNA extracted from the inoculum, and m is qPCR was performed using QNIF4 and NV1LCR primers the slope of the virus RNA transcript standard curve. (Da Silva et al. 2007; Svraka et al. 2007) with the TM9 For the MSB method, the inoculum viral RNA levels probes (Hoehne and Schreier 2006). MNV detection was were estimated from the extraction of 100 µl of the virus performed based on RT-qPCR ORF1/ORF2 primer system diluted with 100 µl of the bead elution buffer using the RNe- developed by Baert et al. (2008). In both cases, 5 µl from asy Qiacube kit and its analysis by RT-qPCR. the RNA extracts were tested using the TaqMan Fast Virus For the ISO 15216-1:2017 method, the inoculum viral 1-Step Master Mix (Thermo Fisher). The reverse transcrip- RNA levels were estimated from the extraction of 100 µl of tion was performed at 50 °C for 20 min, and the amplifica- the virus dilution using the NucliSens miniMAG kit and its tion profile included 20 s at 95 °C, and 45 cycles of 3 s at analysis by RT-qPCR. 95 °C and 30 s at 60 °C. Virus gEq quantification was determined using standard curves generated with in vitro RNA transcripts containing RT‑qPCR Inhibition target sequences for HuNoV GI, GII, or MNV with small sequence inserts to differentiate them from the circulating The RT-qPCR inhibition from the matrix was evaluated as strains. The HuNoV GII RNA transcript UV concentration recommended in ISO 15216-1:2017 using RNA transcripts was divided by a correction factor of 1.8, established by with insert as an external amplification control (EAC). droplet digital PCR (ddPCR) (Advance Analysis Centre Briefly, non-spiked raspberry samples were extracted using Guelph University, Canada). Briefly, reverse transcriptase the MSB or the ISO 15216-1:2017 protocol as described reactions were performed in triplicate as described above above. Five microliter of RNA extract was spiked with 625 using the COG2R primer. The Automated Droplet Gen- gEq of EAC and tested by RT-qPCR using either the RNA erator (Bio-Rad, Canada) was used to generate droplets. UltraSense or the TaqMan Fast Virus 1-Step Master Mix kit Table 1 Primers, probes, and RNA transcripts used in this study Methods Primer or probe Sequence 5′–3′ References Norovirus GI  RT-qPCR, qPCR QNIF4 CGC TGG ATG CGN TTC CAT Da Silva et al. (2007)  RT-qPCR, qPCR NV1LCR CCT TAG ACG CCA TCA TCA TTT AC Svraka et al. (2007)  RT-qPCR, qPCR FAM-TM9-MGBNFQ TGG ACA GGA GAT CGC Hoehne and Schreier (2006)  PCR GISKR CTG CCC GAA TTY GTA AAT GA Kojima et al. (2002)  PCR GISKF CCA ACC CAR CCA TTR TAC A Kojima et al. (2002) Norovirus GII  RT-qPCR, qPCR QNIF2d ATG TTC AGR TGG ATG AGR TTC TCW GA Loisy et al. (2005)  RT-qPCR, qPCR FAM-QNIFS-BHQ-1 AGC ACG TGG GAG GGC GAT CG Loisy et al. (2005)  RT-qPCR, qPCR COG2R TCG ACG CCA TCT TCA TTC ACA Kageyama et al. (2003) Murine norovirus  RT-qPCR FW-ORF1/ORF2 CAC GCC ACC GAT CTG TTC TG Baert et al. (2008)  RT-qPCR RV-ORF1/ORF2 GCG CTG CGC CAT CAC TC Baert et al. (2008)  RT-qPCR FAM-ORF1/ORF2-MGBNFQ CGC TTT GGA ACA ATG Baert et al. (2008) 1 3 252 Food and Environmental Virology (2021) 13:248–258 as described above. EAC spiked in RNase-free water was the pellet after the PEG precipitation step when 30 units of used as controls. pectinase from A. niger were used as indicated in the ISO/ (ΔCt/m) RT-qPCR inhibition rate = (1 − 10 ) × 100% where TS 15216-1:2013 (data not shown). When the pectinase ΔC = Ct − Ct is the Ct value of RNA transcript from A. aculeatus was used as indicated as an alterna- t matrix water matrix spiked in RNA extracted from the matrix minus Ct value tive in the ISO 15216-1:2017 protocol version, the pellet water of RNA transcript in water and m is the slope of the virus was resuspended more easily. Nevertheless, the HuNoV transcript RNA standard curve. GII.4 recovery yields from frozen raspberries using the MSB and ISO 15216-1:2017 methods were not statisti- Limit of Detection (LOD) Calculation cally different (p = 0.366) and were 2.6% and 1.8%, respec- tively. When high (5 × 10 gEq) or low amounts of viruses The PODLOD program (v9) (Wilrich and Wilrich 2009) was (1.7 × 10 gEq) were spiked and extracted using the MSB used to calculate the LOD and LOD . method, the HuNoV GII.4 recovery yields were again 50 95 not statistically different (p = 0.855). Recovery yields for Statistical Analyses HuNoV GI.5 were similar to HuNoV GII and calculated to be 3.6%. Repeated freezing and thawing of raspberry Unless otherwise specified, all statistical analyses were per - matrices before the virus elution appeared to release more formed on log-transformed values using the independent pectin and/or the formation of agglomerates that precluded samples t test (p > 0.05). The F-test was performed to evalu- the magnetic beads handling. Consequently, frozen rasp- ate the variance. A two-way analysis of variance (ANOVA) berry matrices were classified as unfit if they had been with the Bonferroni statistical correction was used to evalu- received thawed and were not used to evaluate recovery ate the impact of the extraction and detection method on the yields. RT-qPCR inhibition (p > 0.05) (MedCalc 17.5.5). As for MNV, the recovery yield from spiked frozen raspberries using the MSB method was 2.8%, similar to the HuNoV GI and GII recovery yields (One-way ANOVA, Results p = 0.431). The tissue culture infectious dose (T CID ) of the MNV production batch was titrated at 203 gEq/ Recovery Yields TCID (CI 95% 139–297; n = 7). With a recovery yield of 2.8% using MSB, it was extrapolated that 25 g of frozen The performance of the MSB method was evaluated by raspberries spiked with 100 µl of a MNV at 53 T CID / assessing its recovery yields in comparison to the ISO ml should allow 3 MNV genomic copies in the RT-qPCR 15216-1:2017 method using IQF frozen raspberries spiked tubes to be detected 94% of the time according to the Pois- with HuNoV GII.4, GI.5, or MNV (Table 2). In our labora- son distribution. tory setting, the virus elution and concentration methods Moreover, the addition of a competitor strain did not for norovirus using the MSB had a turnaround time of have any impact on norovirus recovery yields from spiked approximately 7 h, including 1 h required for the robotic frozen raspberries using the MSB method. Indeed, the RNA extraction. The ISO 15216-1:2017 method had a HuNoV GII recovery yields in the presence or absence turnaround time of 9 h, including the inoculation and the of HuNov GI.5 were in the same range 3.4% and 4.5%, manual NucliSens miniMAG RNA extraction performed respectively (p = 0.247) (Table 3). on the second day. We had some difficulty resuspending Table 2 Detection of norovirus a Extraction method VirusSpiking level n Undiluted Diluted (1/10) in spiked frozen raspberries b c b c CtRecovery yields CtRecovery yields MSB MNV 2.4 × 10 15 26.0 ± 0.6 2.8% (1.8–3.8) nt nt GII.4 5 × 10 15 29.7 ± 1.2 2.6% (1.7–5.0) 30.4 ± 2.5 6.2% (4.2–8.2) GII.4 1.7 × 10 5 34.0 ± 1.5 5.7% (0.8–6.6) nt nt GI.5 6 × 10 15 29.2 ± 0.8 3.6% (2.4–4.9) 32.8 ± 0.8 3.0% (2.1–3.9) 15216 GII.4 5 × 10 15 29.8 ± 1.9 1.8% (0.1–3.6) 30.8 ± 1.5 9.3% (7.1–11.5) n extraction assay, nt not tested, Ct cycle threshold, MSB magnetic silica beads, 15216 ISO 15216–1:2017 Genomic equivalent copies Average Ct ± standard deviation Geometric mean (95% confidence interval) 1 3 Food and Environmental Virology (2021) 13:248–258 253 Table 3 Norovirus recovery Extraction method n Virus Competitor virus yields from frozen raspberries a b a b in the presence of competition VirusSpiking levelRecovery yields VirusSpiking levelRecovery yields 5 5 MSB 5 GII.4 3 × 10 4.5% (3.5–5.8) GI.5 10 3.0% (2.0 to 5.0) 5 GII.4 4 × 10 3.4% (1.8–6.2) nt nt nt 5 nt nt nt GI.5 10 1.8% (− 0.4 to 4.0) Raspberries spiked with or without competitor virus were extracted and tested by RT-qPCR n extraction assay, nt not tested, MSB magnetic silica beads Genomic equivalent copies Geometric mean (95% confidence interval) frozen raspberry samples were tested. The MSB RT- RT‑qPCR Inhibition qPCR LOD and LOD were 2370 gEq per 25  g (95% 95 50 CI 1542–3642) and 548 gEq per 25 g (95% CI 357–843), The ratio of diluted (1/10) to non-diluted HuNoV GII RNA respectively. recovery yields indicated a fairly high relative RT-qPCR inhibition value of 5.2 when the HuNoV GII RNA was extracted from frozen raspberries using the ISO 15216- Discussion 1:2017 method compared to a ratio value of 2.4 observed with the MSB method (Table 2). Ratio values above 4 rep- Frozen raspberries have been associated with several noro- resent more than 75% RT-qPCR inhibition. In contrast, virus outbreaks and remain a challenging food matrix for the tenfold dilution of the HuNoV GI RNA from samples virus detection. We have experienced some difficulties with extracted using the MSB method on frozen raspberries had the workflow of the ISO/TS 15216-1:2013. The limited pH no impact on the recovery yields (p = 0.214). The average range during elution, the impact of PCR inhibitors, and a undiluted HuNoV GI RNA extract Ct was not lower than the tedious PEG pellet resuspension observed with frozen rasp- diluted ones after correction for the dilution factor. berry matrices make this method poorly adapted to testing Using the HuNoV GII and GI RNA transcripts with the large number of samples CFIA diagnostic labs need insert as EAC, a statistically significant impact of the to process every year (> 500). Some of these issues were RT-qPCR kit (p < 0.001) on the RT-qPCR inhibition was resolved in the 2017 version of the method. As an alterna- observed (Fig.  2). The average RNA UltraSense HuNoV tive, we developed a new food virus extraction method based GII RT-qPCR inhibition percentages measured using the on magnetic silica beads (MSB) which was better adapted RNA extracted with the ISO 15216-1:2017 and the MSB to our laboratory setting. The MSB approach to elute and approaches were 58% (95% CI 32–83) (n = 12 replicates) concentrate norovirus from raspberries is based on a stra- and 56% (95% CI 41–71) (n = 20), respectively. However, tegic use of electrostatic interactions at different steps of when testing the HuNoV GII EAC amplification using the the protocol. The variations of the virus surface charge at TaqMan Fast Virus 1-Step, the average RT-qPCR inhibi- different pH levels are illustrated in Fig.  1. According to the tion percentages calculated with the RNA extracts using the manufacturer, the fine magnetic silica beads have a small ISO 15216-1:2017, and the MSB protocols were decreased size (420 nm) and have a negative zeta potential in water to 24% (95% CI 7–42) (n = 12) and 1% (95% CI 7 to 9) (− 37 mV). On the other hand, silica particles are prone to (n = 20), respectively. Similar EAC inhibition levels were aggregation in the presence of salts (Metin et al. 2011). In observed using the HuNoV GI RT-qPCR. RT-qPCR inhi- the MSB method, a non-ionic buffer (Bis–Tris-Propane) bition values were above the 75% when tested with the was used, and salts were avoided in the buffer solutions to RNA UltraSense kit for 46% and 25% of the RNA samples reduce the ionic strength and to limit the aggregation of extracted with the ISO 15216-1:2017 and the MSB methods, magnetic silica particles. The effect of pH and charge of the respectively. There was no impact of the extraction method norovirus were also taken into account. According to Da on the RT-qPCR inhibition (p = 0.110). Silva et al. (2011), HuNoV GI Virus-like particles (VLP) are prone to adhesion onto silica below the VLP isoelectric Limit of Detection point whereas its attachment is reduced at higher pH at low salt concentrations. The limit of detection of the MSB method was evaluated It is difficult to compare method performance based by RT-qPCR using RNA extracted from frozen raspberries on the literature or laboratory reports (Li et al. 2018). In spiked with HuNoV GII.4 strain CFIA-FVR-019 (Supple- addition to the extraction protocols, variations in recovery mentary Fig. SF1). Overall, 70 spiked and 14 non-spiked 1 3 254 Food and Environmental Virology (2021) 13:248–258 Fig. 2 RT-qPCR inhibition (A) HuNoV GI percentages in frozen raspberry RNA extracts. A box plot of RT-qPCR inhibition evalu- ated using a HuNoV GI and b HuNoV GII external control RNA amplification is shown. The absence of RT-qPCR inhibition or enhancement should read as 0. The error bars represent the 95% confidence intervals of the mean. The horizontal dotted line represents 75% inhibition -20 -40 -60 -80 -100 RNA UltraSense TaqMan Fast Virus 1-Step RT-qPCR kit (B) HuNoV GII -20 -40 -60 -80 -100 RNA UltraSense TaqMan Fast Virus 1-Step RT-qPCR kit Extraction method ISO 15216-1:2017 MSB yields could be associated to the inoculated virus prepara- Norwalk strain was estimated at 1320 (95% CI 440–3760) tion, clarification, filtration, the virus strain, its integrity, as genomic equivalent (gEq) in serogroups O and A (Atmar well as the spiking conditions. The method should be fit for et al. 2014). The HuNoV GII L OD reported in this study 3 3 its intended purpose and detects HuNoV in frozen raspber- using the MSB approach was higher (2 × 10 vs. 0.7 × 10 ries at level equivalent to the 50% human infectious dose gEq per 25 g) than the ones reported by another study using (HID50). The HID50 of the HuNoV in susceptible healthy the ISO 15216-1:2017 (Li et al. 2018). On the other hand, adults varies with the serogroup. The HID50 of the HuNoV recovery yields obtained for undiluted HuNoV GI and 1 3 RT-qPCR inhibition(%) RT-qPCR Inhibition(%) Food and Environmental Virology (2021) 13:248–258 255 HuNoV GII using the MSB method were in the same range assays increases the burden of testing and could double (2.6–5.7%) as the ISO 15216-1:2017 reported by Fraisse reported prevalence estimates if this factor was not taken et al. (2017). While achieving the lowest detection limit is a into account. The ISO 15216-1:2017 recommends a larger major goal of extraction methodologies, its application also elution volume (100 µl vs 50 µl) and requires testing a ten- influences method selection. For instance, methodologies fold diluted RNA extract as well as testing EAC to avoid this that could discriminate between inactivated and infectious issue (ISO 2017). Previously, a European survey reported virus are required to avoid overestimating the viral infectiv- that positive frozen raspberries contaminated with HuNoV ity. Viability RT-qPCR based on viral integrity treatment is were only detected using tenfold diluted RNA when RNA a promising approach to improve risk assessment of positive was extracted with the ISO/TS 15216-1:2013 method RT-qPCR detection results (Chen et al. 2020). The integrity (Loutreul et al. 2014). Dilution of the soft fruit RNA extract of the virus following its elution and concentration might decreases the impact of PCR inhibitors in the RT-qPCR. varied between different viral RNA extraction methodolo- However, RNA extract dilution also impacts the capacity to gies and requires further investigation. detect the virus present at trace levels (Fraisse et al. 2017). Raspberries contain high levels of heteropolysaccha- Until improvements are shown regarding the recovery yields, rides, such as pectin, which have a major impact on the viral strategies that decrease the impact of PCR inhibitors from extraction process. The pectin appears to form a strong gel frozen raspberry RNA extracts should be encouraged. with calcium at a pH close to 5 which interferes with the Different PCR inhibitor removal kits or the use of digi- elution process (Han et al. 2017). Pectinase treatment, lower tal PCR have improved the ISO 15216-1: 2017 recovery pH, and low calcium conditions were required to avoid bead yields (Fraisse et al. 2017; Bartsch et al. 2016, 2018). For agglomeration when performing the MSB elution from fro- instance, an additional RNA purification step using either the zen raspberries. The current low pH extraction process was MobiSpin column or the OneS tep PCR Inhibitor Removal effective with a limited set of matrices. Kit (Zymo research) were added with the ISO/TS 15216- In addition to pectin, the extraction of HuNoV from fro- 1:2013 protocol to remove RT-PCR inhibitors from frozen zen raspberries presented other challenges. Raspberries strawberries and improve the detection limit (Bartsch et al. contain multiple components that are co-extracted with the 2016, 2018). Alternative extraction approaches could also viral genome and can impact its molecular detection. High reduce the amount of RT-PCR inhibitors. With Bovine Noro- levels of polyphenols (e.g., anthocyanin, flavonol, ellagi- virus spiked on frozen raspberries, Sun et al. (2019) have tannin, proanthocyanidin, phenolic acids, tannic acid) can reported the absence of RT-PCR inhibition using a direct act as PCR inhibitors (reviewed in Schrader et al. (2012)). lysis approach combined with either RNA filtration using the Heteropolysaccharides can disturb the RT and PCR enzy- MobiSpin column or digital PCR detection. The results in matic process by mimicking the structure of nucleic acids. this study indicated that the RT-qPCR inhibition is also influ- Phenolic compounds may cross‐link RNA under oxidizing enced by the selected detection method. The RNA UltraSense conditions and could degrade DNA polymerases. To reduce kit is described in the ISO 15216-2017 method, but it is not a the presence of PCR inhibitors, the MSB method includes a requirement for the method. The RNA UltraSense kit might treatment with insoluble polyvinylpolypyrrolidone that was be suitable for some food matrices included in the scope of reported to prevent polyphenol oxidation and subsequent this method that are associated to low RT-qPCR inhibition. binding to nucleic acids when extracting total RNA from However, the TaqMan Fast Virus 1-Step results indicate that raspberries (Jones et al. 1997). the UltraSense kit is not the most appropriate detection kit to High levels of RT-qPCR inhibition have been reported test RNA extracted from frozen raspberries. The impact of by several groups that evaluated PEG-derived extraction PCR inhibitors from frozen raspberry RNA extracts on other methods including ISO/TS 15216-1:2013 and ISO 15216- commercial RT-qPCR kit was not explored. Commercial 1:2017 with frozen raspberries (De Keuckelaere et al. 2015; RT-qPCR kits use die ff rent proprietary bue ff rs and additives Fraisse et al. 2017; Summa and Maunula 2018; Summa et al. to reduce the impact of PCR inhibitors. The extracted PCR 2012). With frozen raspberries RNA extracted using the inhibitors vary with the type of food matrix. Consequently, ISO 15216-1:2017, RT-qPCR inhibition was estimated at the impact of inhibitors on commercial RT-qPCR kits should 93.8% ± 2.5% using an external RNA control (Fraisse et al. be tested for each type of food matrix. 2017). According to the ISO 15216-1:2017, negative results Several groups have used various forms of the ISO obtained in presence of RT-qPCR inhibition levels > 75% 15216 method to recover HuNov from frozen raspberries. are not valid. The observed range of recovery yields from 1 to 6% range Meanwhile in this study, close to 50% of the undiluted could certainly impact the reported prevalence. Neverthe- samples extracted following the ISO 15216-1:2017 protocol less, Loutreul et al. (2014) reported a prevalence of 16.7% and tested using the RNA UltraSense detection kit presented (n = 162) for HuNov GI in frozen raspberries from Ser- more than 75% inhibition. Such a level of inconclusive bia, Chile, Bulgaria, Poland, and France. In the UK, 3.6% 1 3 256 Food and Environmental Virology (2021) 13:248–258 (n = 274) of frozen raspberry samples sold at retail were pos- Compliance with Ethical Standards itive for HuNoV (Cook et al. 2019). However, this group was Conflict of interest The authors have no conflict of interest to declare. not able to differentiate the sequence of the HuNoV detected from their EAC. Gao et al. (2019) reported that 9.2% and Open Access This article is licensed under a Creative Commons Attri- 13% of frozen raspberries from Heilongjiang Province in bution 4.0 International License, which permits use, sharing, adapta- China were positive to HuNoV in 2016 and 2017, respec- tion, distribution and reproduction in any medium or format, as long tively. Meanwhile, all export samples were negative. They as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes used a replicate Ct threshold for positive results, which could were made. The images or other third party material in this article are lower prevalence estimate, and did not analyze the RT-qPCR included in the article’s Creative Commons licence, unless indicated inhibition level. Maunula et al. (2013) did not detect HuNov otherwise in a credit line to the material. If material is not included in in frozen raspberry samples (0/39) from point of sale of four the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will European countries, but did nd fi some HuNov GII in irrigat - need to obtain permission directly from the copyright holder. To view a ing water from berry production sites (2/56). They used high copy of this licence, visit http://creativ ecommons .or g/licenses/b y/4.0/. pectinase concentration and RNA extract elution volume (300 µl) for berries but did not report any RT-qPCR inhibi- tion. The true prevalence as well as the extraction method recovery yields and the presence of PCR inhibitors could References play a role in resolving some of the discrepancies observed Andrade, A. L., Souza, D. M., Pereira, M. C., Fabris, J. D., & in terms of prevalence between those groups. Still, the con- Domingues, R. Z. (2009). Synthesis and characterization of mag- tamination levels in some of these surveys remains relatively netic nanoparticles coated with silica through a sol-gel approach. important from a risk analysis perspective. Cerâmica, 55, 420–424. Atmar, R. L., Opekun, A. R., Gilger, M. A., Estes, M. K., Crawford, S. E., Neill, F. H., et al. (2014). Determination of the 50% human infectious dose for Norwalk virus. Journal of Infectious Diseases, 209(7), 1016–1022. https ://doi.org/10.1093/infdi s/jit62 0. Conclusion Baert, L., Wobus, C. E., Van Coillie, E., Thackray, L. B., Debevere, J., & Uyttendaele, M. (2008). Detection of murine norovirus 1 by using plaque assay, transfection assay, and real-time reverse tran- A method based on magnetic silica beads to extract HuNoV scription-PCR before and after heat exposure. Applied and Envi- virus RNA from frozen raspberries was developed. The ronment Microbiology, 74(2), 543–546. https ://doi.org/10.1128/ MSB method performance was similar to the reference AEM.01039 -07. method ISO 15216-1:2017. The influence of RT-qPCR Bartsch, C., Hoper, D., Made, D., & Johne, R. (2018). Analysis of frozen strawberries involved in a large norovirus gastroenteri- inhibitors extracted using both methods was reduced using tis outbreak using next generation sequencing and digital PCR. an alternative RT-qPCR detection kit (TaqMan Fast Virus Food Microbiology, 76, 390–395. https ://doi.or g/10.1016/j. 1-Step Master Mix) and condition. In future, the reduction of fm.2018.06.019. the RT-qPCR inhibitors which impact the HuNoV detection Bartsch, C., Szabo, K., Dinh-Thanh, M., Schrader, C., Trojnar, E., & Johne, R. (2016). Comparison and optimization of detection meth- results should reduce the number of inconclusive assays and ods for noroviruses in frozen strawberries containing different influence prevalence estimates. amounts of RT-PCR inhibitors. Food Microbiology, 60, 124–130. https ://doi.org/10.1016/j.fm.2016.07.005. Supplementary Information The online version contains supplemen- Boqvist, S., Soderqvist, K., & Vagsholm, I. (2018). Food safety chal- tary material available at https:// doi.org/10.1007/s12560-02 1-09466-0 . lenges and one health within Europe. Acta Veterinaria Scandi- navica, 60(1), 1. https ://doi.org/10.1186/s1302 8-017-0355-3. Acknowledgements The authors acknowledge the work of analysts Bozkurt, H., Phan-Thien, K. Y., van Ogtrop, F., Bell, T., & McCo- and scientist from the CFIA Food Virology National Reference Center nchie, R. (2020). Outbreaks, occurrence, and control of noro- (FVNRC) for the production of RNA transcript standards (MO, VL, virus and hepatitis a virus contamination in berries: A review. YL). Names of specific vendors, manufacturers, or products are Critical Reviews in Food Science and Nutrition. https ://doi. included for informational purposes only and do not imply endorse- org/10.1080/10408 398.2020.17193 83. ment by authors. Brinkman, N. E., Haffler, T. D., Cashdollar, J. L., & Rhodes, E. R. (2013). Evaluation of methods using celite to concentrate noro- virus, adenovirus and enterovirus from wastewater. Journal of Author contributions Conceptualization: PR and LD. Funding acquisi- Virological Methods, 193(1), 140–146. https ://doi.org/10.1016/j. tion: PR and LD. Supervision: PR, Methodology: PR, SP and AP. Data jviro met.2013.05.014. curation: PR and SP. Formal analysis: PR. Writing original draft: PR. Chen, J., Wua, X., Sánchezb, G., & Randazzob, W. (2020). Viabil- Writing review & editing: PR, AP, SP and LD. ity RT-qPCR to detect potentially infectious enteric viruses on heat processed berries. Food Control, 107, 106818. https ://doi. Funding This work was supported by the Canadian Food Inspec- org/10.1016/j.foodc ont.2019.10681 8. tion Agency Research & Partnership Strategy (STH-F-1306 and Chhabra, P., de Graaf, M., Parra, G. I., Chan, M. C., Green, K., STH-F-1607). Martella, V., et al. (2019). Updated classification of norovirus 1 3 Food and Environmental Virology (2021) 13:248–258 257 genogroups and genotypes. Journal of General Virology, 100(10), Kageyama, T., Kojima, S., Shinohara, M., Uchida, K., Fukushi, 1393–1406. https ://doi.org/10.1099/jgv.0.00131 8. S., Hoshino, F. B., et  al. (2003). Broadly reactive and highly Cook, N., Knight, A., & Richards, G. P. (2016). Persistence and elimi- sensitive assay for Norwalk-like viruses based on real-time nation of human norovirus in food and on food contact surfaces: quantitative reverse transcription-PCR. Journal of Clini- A critical review. Journal of Food Protection, 79(7), 1273–1294. cal Microbiology, 41(4), 1548–1557. https ://doi.or g/10.1128/ https ://doi.org/10.4315/0362-028X.JFP-15-570. jcm.41.4.1548-1557.2003. Cook, N., Williams, L., & D’Agostino, M. (2019). Prevalence of noro- Kojima, S., Kageyama, T., Fukushi, S., Hoshino, F. B., Shinohara, M., virus in produce sold at retail in the United Kingdom. Food Micro- Uchida, K., et al. (2002). Genogroup-specific PCR primers for biology, 79, 85–89. https ://doi.org/10.1016/j.fm.2018.12.003. detection of Norwalk-like viruses. Journal of Virological Meth- Da Silva, A. K., Kavanagh, O. V., Estes, M. K., & Elimelech, M. ods, 100(1–2), 107–114. (2011). Adsorption and aggregation properties of norovirus GI Li, D., Butot, S., Zuber, S., & Uyttendaele, M. (2018). Monitoring of and GII virus-like particles demonstrate differing responses to foodborne viruses in berries and considerations on the use of RT- solution chemistry. Environmental Science and Technology, 45(2), PCR methods in surveillance. Food Control, 89, 235–240. https 520–526. https ://doi.org/10.1021/es102 368d.://doi.org/10.1016/j.foodc ont.2018.02.024. Da Silva, A. K., Le Saux, J. C., Parnaudeau, S., Pommepuy, M., Elime- Loisy, F., Atmar, R. L., Guillon, P., Le Cann, P., Pommepuy, M., & Le lech, M., & Le Guyader, F. S. (2007). Evaluation of removal of Guyader, F. S. (2005). Real-time RT-PCR for norovirus screening noroviruses during wastewater treatment, using real-time reverse in shellfish. Journal of Virological Methods, 123(1), 1–7. https :// transcription-PCR: Different behaviors of genogroups I and II. doi.org/10.1016/j.jviro met.2004.08.023. Applied and Environmental Microbiology, 73(24), 7891–7897. Loutreul, J., Cazeaux, C., Levert, D., Nicolas, A., Vautier, S., Le Sau- https ://doi.org/10.1128/AEM.01428 -07. vage, A. L., et al. (2014). Prevalence of human noroviruses in De Keuckelaere, A., Li, D., Deliens, B., Stals, A., & Uyttendaele, M. frozen marketed shellfish, red fruits and fresh vegetables. Food (2015). Batch testing for noroviruses in frozen raspberries. Inter- Environmental Virology, 6(3), 157–168. https ://doi.org/10.1007/ national Journal of Food Microbiology, 192, 43–50. https ://doi.s1256 0-014-9150-8. org/10.1016/j.ijfoo dmicr o.2014.09.024. Maunula, L., Kaupke, A., Vasickova, P., Soderberg, K., Kozyra, I., Fraisse, A., Coudray-Meunier, C., Martin-Latil, S., Hennechart- Lazic, S., et al. (2013). Tracing enteric viruses in the European Collette, C., Delannoy, S., Fach, P., et al. (2017). Digital RT-PCR berry fruit supply chain. International Journal of Food Micro- method for hepatitis A virus and norovirus quantification in soft biology, 167(2), 177–185. h t t ps : / / do i . o rg/ 1 0 . 1 0 16 / j . i j f oo d m i c r berries. International Journal of Food Microbiology, 243, 36–45. o.2013.09.003. https ://doi.org/10.1016/j.ijfoo dmicr o.2016.11.022. Metin, C. O., Lake, L. W., Miranda, C. R., & Nguyen, Q. P. (2011). Gao, X., Wang, Z., Wang, Y., Liu, Z., Guan, X., Ma, Y., et al. (2019). Stability of aqueous silica nanoparticle dispersions. Journal of Surveillance of norovirus contamination in commercial fresh/fro- Nanoparticle Research, 13(2), 839–850. https ://doi.org/10.1007/ zen berries from Heilongjiang Province, China, using a TaqMan s1105 1-010-0085-1. real-time RT-PCR assay. Food Microbiology, 82, 119–126. https Morales-Rayas, R., Wolffs, P. F., & Griffiths, M. W. (2010). Simultane- ://doi.org/10.1016/j.fm.2019.01.017. ous separation and detection of hepatitis A virus and norovirus in Gonzalez-Hernandez, M. B., Bragazzi Cunha, J., & Wobus, C. E. produce. International Journal of Food Microbiology, 139(1–2), (2012). Plaque assay for murine norovirus. Journal of Visualized 48–55. https ://doi.org/10.1016/j.ijfoo dmicr o.2010.02.011. Experiments, 66, e4297. https ://doi.org/10.3791/4297. Nasheri, N., Vester, A., & Petronella, N. (2019). Foodborne viral out- Government of Canada. (2018). National enteric surveillance program breaks associated with frozen produce. Epidemiology and Infec- annual summary 2017. Guelph: Public Health Agency of Canada. tion, 147, e291. https ://doi.org/10.1017/S0950 26881 90017 91. Han, W., Meng, Y., Hu, C., Dong, G., Qu, Y., Deng, H., et al. (2017). Scherer, K., Johne, R., Schrader, C., Ellerbroek, L., Schulenburg, J., Mathematical model of Ca2+ concentration, pH, pectin con- & Klein, G. (2010). Comparison of two extraction methods for centration and soluble solids (sucrose) on the gelation of low viruses in food and application in a norovirus gastroenteritis out- methoxyl pectin. Food Hydrocolloids, 66, 37–48. https ://doi. break. Journal of Virological Methods, 169(1), 22–27. https://doi. org/10.1016/j.foodh yd.2016.12.011.org/10.1016/j.jviro met.2010.06.008. Hoehne, M., & Schreier, E. (2006). Detection of Norovirus genogroup Schrader, C., Schielke, A., Ellerbroek, L., & Johne, R. (2012). PCR I and II by multiplex real-time RT-PCR using a 3’-minor groove inhibitors—Occurrence, properties and removal. Journal of binder-DNA probe. BMC Infectious Diseases, 6, 69. https ://doi. Applied Microbiology, 113(5), 1014–1026. https ://doi.org/10.11 org/10.1186/1471-2334-6-69.11/j.1365-2672.2012.05384 .x. Houde, A., Leblanc, D., Poitras, E., Ward, P., Brassard, J., Simard, C., Stals, A., Baert, L., Van Coillie, E., & Uyttendaele, M. (2012). Extrac- et al. (2006). Comparative evaluation of RT-PCR, nucleic acid tion of food-borne viruses from food samples: A review. Interna- sequence-based amplification (NASBA) and real-time RT-PCR tional Journal of Food Microbiology, 153(1–2), 1–9. https ://doi. for detection of noroviruses in faecal material. Journal of Viro-org/10.1016/j.ijfoo dmicr o.2011.10.014. logical Methods, 135(2), 163–172. https://doi.or g/10.1016/j.jviro Summa, M., & Maunula, L. (2018). Rapid detection of human noro- met.2006.03.001. virus in frozen raspberries. Food Environmental Virology, 10(1), ISO. (2017). Horizontal method for determination of Hepatitis A virus 51–60. https ://doi.org/10.1007/s1256 0-017-9321-5. and norovirus using real-time RT-PCR–Part 1: Method for quan- Summa, M., von Bonsdorff, C. H., & Maunula, L. (2012). Evalua- tification. In Microbiology of the food chain (Vol ISO 15216– tion of four virus recovery methods for detecting noroviruses on 1:2017). Geneva: International Organization for Standardization. fresh lettuce, sliced ham, and frozen raspberries. Journal of Viro- Jones, C. S., Iannetta, P. P., Woodhead, M., Davies, H. V., McNicol, R. logical Methods, 183(2), 154–160. https://doi.or g/10.1016/j.jviro J., & Taylor, M. A. (1997). The isolation of RNA from raspberry met.2012.04.006. (Rubus idaeus) fruit. Molecular Biotechnology, 8(3), 219–221. Sun, B., Bosch, A., & Myrmel, M. (2019). Extended direct lysis https ://doi.org/10.1007/BF027 60775 . method for virus detection on berries including droplet digital Júnior, J. A. A., & Baldo, J. B. (2014). The behavior of zeta potential of RT-PCR or real time RT-PCR with reduced influence from inhibi- silica suspensions. New Journal of Glass and Ceramics, 4, 29–37. tors. Journal of Virological Methods, 271, 113638. https ://doi. https ://doi.org/10.4236/njgc.2014.42004 .org/10.1016/j.jviro met.2019.04.004. 1 3 258 Food and Environmental Virology (2021) 13:248–258 Svraka, S., Duizer, E., Vennema, H., de Bruin, E., van der Veer, B., Wilrich, C., & Wilrich, P. T. (2009). Estimation of the POD func- Dorresteijn, B., et al. (2007). Etiological role of viruses in out- tion and the LOD of a qualitative microbiological measurement breaks of acute gastroenteritis in The Netherlands from 1994 method. Journal of AOAC International, 92(6), 1763–1772. through 2005. Journal of Clinical Microbiology, 45(5), 1389– 1394. https ://doi.org/10.1128/JCM.02305 -06. Publisher’s Note Springer Nature remains neutral with regard to Vinje, J. (2015). Advances in laboratory methods for detection and jurisdictional claims in published maps and institutional affiliations. typing of norovirus. Journal of Clinical Microbiology, 53(2), 373–381. https ://doi.org/10.1128/JCM.01535 -14. 1 3

Journal

Food and Environmental VirologySpringer Journals

Published: Mar 2, 2021

There are no references for this article.