Effect of high hydrostatic pressure on murine norovirus in Manila clams


Giuseppe Arcangeli, Istituto Zooprofilattico Sperimentale delle Venezie, V.le dell’Università, 10, 35020 Legnaro, PD, Italy. E-mail: garcangeli@izsvenezie.it


Aims:  Eating raw or insufficiently cooked bivalve molluscs contaminated with human noroviruses (NVs) can result in acute cases of gastroenteritis in humans. Manila clams (Ruditapes philippinarum) are particularly prone to exposure to NVs due to the brackish environment in which they are farmed which is known to be susceptible to human faecal contamination. High hydrostatic pressure processing (HHP) is a food treatment technique that has been shown to inactivate NV.

Methods and results:  In this study we investigated the ability of HHP to inactivate murine norovirus (MNV-1), a recognised surrogate for NV, in experimentally contaminated manila clams. Pools of contaminated live clams were subjected to hydrostatic pressure ranging from 300 to 500 MPa for different time intervals of between one and 10 min. The trial was repeated three times, at monthly intervals.

Conclusions:  Virus vitality post-treatment was assessed and the data obtained indicates that the use of high hydrostatic pressures of at least 500 MPa for 1 min was effective in inactivating MNV-1.

Significance and Impact of the Study:  HHP results to be an effective technique that could be applied to industrial process to obtain safe Manila clams ready to eat.


Bivalve molluscs are filtering organisms that can concentrate micro-organisms, algal toxins and chemical substances internally. The custom in various parts of world of eating raw or undercooked bivalve molluscs can result in viral gastroenteric diseases of varying gravity with some even reaching epidemic proportions (Lees 2000).

Many aetiological agents are behind such epidemics including bacteria naturally present in marine environments (e.g. Vibrionaceae) and salmonellas, usually as a result of pollution of water by sewage, and enteric viruses (e.g. noroviruses). Indeed, there are numerous clinical reports of cases linked to NVs in shellfish (Cheng et al. 2005; Le Guyader et al. 2006a, 2008; Le Guyader and Atmar 2007; Westrell et al. 2010).

The noroviruses, previously known as Norwalk-like viruses (NLV), are a group of nonenveloped viruses with a positive RNA strand of about 7.6 kbp containing three open reading frames (ORFs) that are a widespread cause of nonbacterial gastroenteritis. They belong to the family Caliciviridae, genus Norovirus, and are currently classified into five genogroups (GI, GII, GIII, GIV and GV), of which GI, GII and GIV are capable of infecting humans (Lopman et al. 2003; Ramirez et al. 2008).

NVs are unable to grow in cell culture and so, for this reason, ‘surrogate’ viruses are used in experimental trials, like FCV (feline calicivirus) and, more recently, MNV-1 (murine norovirus). Owing to their ability to growth in cell cultures, it is possible to verify and quantify the replication of these surrogate viruses (Li et al. 2009; Cannon et al. 2006; Hewitt et al. 2009; Wobus et al. 2006).

Presently, in Europe, the criteria for deciding whether molluscs are suitable for human consumption or not depend on the presence or absence of Escherichia coli. One problem with this system lies in the fact that the presence of E. coli does not correlate with the presence of enteric viruses such as noroviruses or hepatitis A virus given that these viruses are more persistent than faecal bacteria in brackish and sea water. Therefore, the absence of E. coli does not guarantee the absence of enteric viruses (Reg. CE 2073/2005: EEC 2005; FAO 2008; Suffredini et al. 2008).

Furthermore, the current depuration systems used for bivalve molluscs do not, in fact, ensure the complete elimination of NVs from the bivalves, given that truly effective treatment requires more than 3–4 days, a period simply not applicable in practice (Schwab et al.1998; Le Guyader et al. 2006b; Ueki et al. 2007).

In recent years, the number of cases of norovirus-associated gastroenteritis in humans because of the consumption of raw or undercooked bivalve molluscs has risen globally. For this reason, various groups have been investigating suitable treatment processes to remove NVs such as sterilization and high pressure treatment.

High hydrostatic pressure (HHP) treatment is a technique that has been used to inactivate bacteria and viruses in raw foods. In HHP, packaged food is placed in a pressure vessel and submitted to water pressures from 100 to 900 MPa. The pressure applied is isostatically transmitted inside the pressure vessel. Studies conducted on viruses belonging to different families indicate a range of variability in the efficacy of HHP in viral inactivation, which depends on the characteristics of the virus in question and the food matrix itself (Grove et al. 2006).

In particular, studies on bivalve molluscs show that HHP has the effect of reducing the infective capacity of Caliciviridae (Kingsley et al. 2002; Murchie et al. 2007; Buckow et al. 2008; Li et al. 2009). Furthermore, HHP at 600 MPa (6°C, 5 min) completely inactivated human NV genogroup 1.1 in seeded oysters, as determined in a randomized, double-blinded clinical trial (Leon et al. 2011).

In this study, the effect of HPP on the vitality of murine norovirus MNV-1, used as a surrogate for human NV, was assessed for the first time in Manila clams.

Materials and Methods

MNV-1 plaque assay

Virus stock and cells.  Murine norovirus-1 (MNV-1, kindly provided by Prof H.W. Virgin, Washington University School of Medicine, MO, USA) was replicated in murine monocytes/macrophages, RAW 264.7 (ATCC cat. no. TIB-71), cultured in Dulbecco’s modified Eagle medium (Sigma-Aldrich cat. no. D6429) supplemented with 10% foetal bovine serum (cat. no. ECS0180L; Euroclone, Pero, Milan, Italy) and 1% penicillin and streptomycin (Sigma P781; 10 000 IU penicillin and 10 mg streptomycin) and incubated at 37°C with 5% CO2. After cellular lysis, the lysate was frozen and thawed twice and then centrifuged at 1000 g for 10 min. The supernatant was filtered through a 0.45-μm filter, and the virus stock was stored at −80°C.

The number of MNV-1 plaque forming units PFU ml−1 was determined by plaque assay, as described previously by Li et al. 2009, with some modification. Briefly, serial 10-fold dilution of virus was used to infect confluent monolayer of RAW 264.7 in 6-well plates (Nunc, cat number 353046; Nunc cell culture microwell plates; Sigma Aldrich, St Louis, MO). Each dilution was tested in duplicate using 0.5 ml to infect each well. Virus inoculation was carried out for 1 h at 37°C, followed by overlay with 3 ml of a solution composed of 2 ×  Minimum Essential Medium (Sigma Aldrich) and Bacto Agar (Difco, Lawrence, KS) in a ratio of 1 : 1 (v/v). After a 72-h incubation, plaques were visualized by staining with a solution of 0.05% formalin (Sigma Aldrich) and 1% crystal violet (Sigma Aldrich). Formalin–crystal violet solution (4 ml) was added to each well. After a 3-h incubation at room temperature, the overlay was discarded and plates were washed with distilled water.

Sample contamination and experimental design

A total of 390 Manila clams were harvested from a lagoon located in the delta of the Po river (North-East of Italy) in April during a nonrainy period and with a E. coli load of <230 CFU per 100 g, analysed using the MPN microbiological method, ISO/TS 16649-3:2005.

Clams that had not undergone depuration were used as it was thought that they would have a better filtering ability with respect to depurated clams.

All of the clams were acclimatized for 1 day prior to experimental infection in an aerated tank containing 100 l of artificial seawater, prepared using synthetic marine salt (Instant Ocean; Aquarius System, Sarrebourg, France). The seawater conditions were salinity 3.5%; water temperature 15°C; and dissolved oxygen more than 90% of air saturation. The seawater was contaminated with MNV-1 at a final concentration of 104 PFU ml−1 of seawater. The shellfish weighing about 10 g per each were immersed in contaminated water for 24 h. The presence of excurrent siphons was observed, confirming that the clams had contact with the external water.

Three groups of 10 noncontaminated Manila clams were collected before contamination and used as negative controls. Another three groups of 10 Manila clams were collected at the end of the infection, were not treated with high pressure and were used as positive controls. Eleven groups of 10 Manila clams each were treated with high pressures for different exposure times (see Table 1).

Table 1.   Results of the hydrostatic pressure processing (HHP) treatments
Experimental conditions
PressureDuration (min)pool Ipool IIpool IIIMean
Untreated4.2 × 1043.8 × 1042.7 × 1043.6 × 104
300 MPa13.7 × 1041.5 × 1041.3 × 1042.2 × 104
300 MPa31.9 × 1041.9 × 1040.8 × 1041.5 × 104
300 MPa51.2 × 1041.8 × 1041.5 × 1041.5 × 104
400 MPa11.0 × 1041.4 × 1041.2 × 1041.2 × 104
400 MPa50.8 × 1041.2 × 1041.0 × 1041.0 × 104
400 MPa100.4 × 1040.4 × 1040.4 × 1040.4 × 104
500 MPa1negnegneg
500 MPa5negnegneg
500 MPa10negnegneg
600 MPa1negnegneg
600 MPa3negnegneg
Negative controlnegnegneg

HHP treatment

Preparation of samples.  A total of twelve experimental groups of contaminated shellfishes were prepared. Each pool was composed of 10 individuals (whole clams with intact shells), inserted into sterile Stomacher® 400 bags (International PBI srl, Milan, Italy), which were in turn placed inside heat-sealable polypropylene bags. Eleven of the groups were subjected to HHP, while the remaining group was used as positive/untreated controls. The pooled samples were stored at 4°C until analysis, which was carried out by 24 h. The trial was repeated three times, at monthly intervals

HHP treatment

Hyperbaric treatment was conducted using an Avure QFP 35® system produced by Avure SPA Vasteras (Avure Technologies, OH, USA). The system has a useful volume of 35 l and can reach a pressure of 600 MPa. The compression times to reach the maximum pressure starting with an empty pressurization chamber are around 2 min; decompression is practically instantaneous. The system permits the execution of treatments in adiabatic conditions under controlled temperatures of between 4 and 90°C. The starting temperature during the study was 10°C with an increase of 3°C for every 100 MPa, reaching a final temperature of 28°C at 600 MPa.

MNV-1 extraction and plaque assays

The virus extraction procedure was performed as described previously (Baert et al. 2008) with some modifications. Briefly, digestive glands of the Manila clams were removed by dissection, diluted 1 : 10 (w/v) in phosphate buffer saline (0.05 mol l−1 to pH 7.0–7.4) with antibiotics (penicillin: 10 000 IU ml−1, streptomycin: 10 mg ml−1, nystatin: 5000 IU ml−1, gentamycin sulfate: 250 μg ml−1) and homogenized with sterile quartz sand. The hepatopancreas was chosen as it is known to be an organ that concentrates NV in molluscs (Le Guyader et al. 2006a,b). Samples were then stored at 4°C for 1 h and centrifuged for 10 min at 1500 g to allow clarification of supernatant; 10-fold serial dilutions were made in DMEM, and plaque assays were performed using the same method as that used for virus stock titration (Baert et al. 2008).

Statistical analysis

Pooled samples were tested, and reductions were calculated as log (Nt/N0). Statistical analysis to evaluate the possible influence of time and pressure on the viral titre obtained by HHP experiment was performed using the bootstrap regression model (Efron and Tibshirani 1993). This provided a parameter estimation of each variable in the model without assumption on data distribution. The normal, percentile and bias-corrected confidence interval were calculated for each parameter. Twenty thousand bootstrap replications were considered in the regression to guarantee robust results. STATA software was used to carry out the analysis (Statacorp LP, College Station, TX).


The results of the HHP treatment are shown in Table 1. The bootstrap regression model relating the viral titre with pressure and exposure time shows that the titre statistically decreases when pressure and exposure time increase (P < 0.001). The viral titre depends on pressure and exposure time, but the interaction between the two does not seem significant (P > 0.10). This means that with the passing of time the titre decreases in the same way at any pressure value.

Table 1 reports the progressive decrease of the viral titre when increasing the intensity and the duration of pressure application, to the point of complete inactivation of MNV-1 after a treatment of 500 MPa for at least 1 min.


In summary, the consumption of raw or undercooked clams exposes the consumer to certain risks of infection. The objective of this study was to find an effective sterilization treatment for norovirus in clams defining the optimal balance between exposure time and pressure. Our results show that the vitality of MNV-1 significantly decreases after treatment at high HHP. In particular, exposing the clams to 500 MPa for 1 min at 20°C allowed us to obtain a virus-free product without altering the visual impact of the clam and the consistency of the flesh. Moreover, by using HHP treatment, it is possible to obtain sucked clams, which have lost less weight than the pasteurized products.

In other similar works, feline calicivirus was reduced to undetectable levels in shellfish by treatment at pressures above 300 MPa (Murchie et al. 2007), while a treatment of 450 MPa for 15 min at 45°C was sufficient to inactivate 6.5 log of infectious MNV in culture medium (Sanchez et al., 2011).

The value of 500 MPa, here identified as the minimum value of hydrostatic pressure able to guarantee a virus-free product, is also higher than that reported by Kingsley et al. (2007), who determined a treatment of 400 MPa for 5 min at 5°C to be sufficient to inactivate MNV-1 in contaminated oysters. Some differences between these and our study exist, such as the species of shellfish or the material tested, the equipment used for the hyperbaric treatment and the procedure used for the identification of the infecting virus, which could explain the different results. Furthermore, the differences could be also be related to the temperature during the hyperbaric treatment, a factor that deserves further investigation.

The conditions of hyperbaric treatment reported herein are very similar to those already used in Italy by some producers of clams and therefore could easily be applied to industrial processes.

Indeed, in Italy, the largest producer of Manila clams in Europe, some companies are already applying HHP (300 MPa, for 2–3 min) in the de-shelling process (A. Brutti, personal communication), which although shown to be capable of removing Gram-negative bacteria such a Vibrio parahaemolyticus in C. gigas (Kural et al. 2008; Ma and Su 2011) does not seem to be sufficient in the inactivation of NV. Hence, a modification of industrial procedures to obtain the inactivation of this important human pathogen is required.

Finally, it must be remembered that the procedures recommended herein will provide a product, which although safe regarding the presence of NV, is still exposed to other potential risks, for example, the presence of sporing bacteria (i.e. Clostridium botulinum), which are resistant to these pressure values (Cheftel 1995; San Martin et al. 2002; Margosch et al. 2006). Indeed, the last step of the commercial processing of shellfish is the vacuum-packaging of the drained clams, a product that must be kept refrigerated to avoid the growth of anaerobic bacteria.


This study was conducted using funds provided by the Health Ministry for IIZZSS (current IZSVe research study no. 02/2007). The authors would like to thank Fabio Borghesan and Marco Penzo for their technical advice and Dr William Dundon for critical reading and editing of the manuscript.

Conflict of Interests

All the authors have no conflict of interest to declare.