Noroviruses (NoVs) belong to the Caliciviridae family and are envelope-free positive-sense single-stranded RNA viruses [1]. NoVs contain three open reading frames (ORFs): ORF1, which encodes the primary nonstructural protein of the virus; ORF2, which encodes the primary structural protein of the virus (capsid protein, VP1); and ORF3, which encodes the secondary structural protein (VP2) [2]. Based on the complete VP1 gene, NoVs can be divided into five genomes (GI–GV) [3]. Chhabra et al. further expanded the number of NoV genomes to 10 (GI–GX) based on the diversity of VP1 amino acid sequences [4]. NoVs are the most common pathogenic cause of global outbreaks of acute gastroenteritis, with common symptoms, such as vomiting, diarrhea and stomach cramps ^[5][6][5,6]. NoVs are mainly transmitted via the fecal–oral route and have a high incidence in autumn and winter ^[7][8][7,8]. Human NoVs (HuNoVs) include GI, GII and GIV, with GII being the most common [9]. NoVs cause acute gastroenteritis outbreaks as they are usually associated with foodborne and waterborne transmission and have high environmental resistance and pathogenicity (only 18 virus particles are required to cause the disease) due to the lack of vaccines and specific drugs against them [10].

In recent years, cases of foodborne disease caused by NoVs have been reported in many countries, including the United States, Japan and the United Kingdom, mostly due to consumption of raw or inadequately cooked NoV-contaminated shellfish ^[11][12][13][11,12,13]. Japan experienced a large outbreak of NoVs in 2006 [14]. After several investigations, it was concluded that this outbreak was caused by the consumption of NoV-contaminated shellfish. Thebault et al. modeled and analyzed NoV transmission events in France and found that they were highly associated with oyster consumption [15]. Oysters are typically filter feeders, and plankton, microalgae and viruses in seawater enter their filter-feeding system during feeding. NoVs often bioaccumulate in oysters through filter feeding, and this accumulation can increase the concentration of NoVs in oysters by tens to thousands of times that in the environment ^[16][17][16,17]. However, the NoVs that accumulate in oysters are often difficult to excrete during metabolic processes or purification [18].

Histo-blood group antigens (HBGAs) are the target receptors for NoVs. A complex carbohydrate present on the virus surface interacts with the fucose region of HBGAs [5]. Studies have confirmed the presence of multiple HBGA-like molecules in oyster intestinal tissues, to which various NoVs can specifically bind and bioaccumulate [19]. Further studies have confirmed that the seasonally dependent expression levels of NoV ligands in oysters were consistent with the bioaccumulation efficiency of NoVs in oysters ^[20][21][20,21], which may be one of the reasons for the high prevalence of NoV infections in autumn and winter. Lowther et al. investigated NoVs contamination in two oyster farming areas in the UK for 31 consecutive months and found that the detection rate of NoVs was approximately 17 times higher from November to March each year than that in other months [22]. Some researchers have suggested that this might be due to the fact that oysters secrete certain substances that favor the bioaccumulation of NoVs under low-temperature conditions [23].

Understanding the mechanisms of virus bioaccumulation in food matrices and developing more advanced assays to detect virus concentrations in matrices can provide prerequisites for subsequent efficient virus inactivation, thereby reducing the risk of viral transmission and maintaining food safety.

Oysters are typical filter-feeding bivalves, which draw seawater into their gills during feeding. During feeding, a large number of suspended particles are transported together into the digestive tract, with some of the particles being excreted and others being absorbed by phagocytic blood cells. NoVs usually exist in the water column as adsorbed particles and are absorbed along with other particles by blood cells ^[24][36]. Blood cells contain acidic vesicles, and whether viruses can survive in blood cells depends on their stability in acidic conditions ^[25][37]. NoVs are highly resistant to acidic environments and can therefore remain in blood cells for a long time and are subsequently transported to the peri-intestinal tissues of oysters ^[26][38]. NoVs then bind to HBGA-like molecules in the gastrointestinal tract epithelium, leading to their bioaccumulation ^[27][39].

Although all bivalves can be contaminated by NoVs during filter feeding, oysters play a more important role in the transmission of NoVs than other bivalves, possibly due to the mode of consumption (raw or lightly cooked), the proximity of their habitat to sewage outfalls, or the ability of NoVs to specifically bind to HBGA-like molecules, which are difficult to remove from oysters ^[28][29][30][40,41,42].

The bioaccumulation of NoVs in oysters is correlated with their genotypes. A study of NoV contamination in an oyster farming area in France reported that the percentage of NoV GI- and GII-positive samples was 59% and 70%, respectively, in 17 samples collected during the first week, and decreased to 41% and 17%, respectively, in 17 samples collected after 3 weeks ^[31][43]. The decrease in GI-positive samples was significantly smaller than that in GII-positive samples, which indicated that GI had a stronger persistence in oysters than GII, probably because the binding strength to HBGA-like molecules in vivo is higher for GI than that for GII. Another study based on RT-qPCR determined the binding of NoVs to oyster tissue by microwell plates coated with oyster gut homogenates, and the results showed that the highest binding rate was found with NoV GI.3 and oysters (97.3%), whereas 35.1% of NoV GII.4 bound to oysters [16]. Therefore, GI was detected frequently in NoV outbreaks triggered by oysters. Molecular dynamics membrane simulation data showed that the high binding strength of NoVs to HBGA-like molecules was due to fully exposed (1,2)-linked α-L-Fucp and (1,4)-linked α-L-Fucp residues, and weaker binding strengths or an inability to bind was due to the very limited accessibility of (1,3)-linked α-L-Fucp residues when glycolipids were embedded in phospholipid membranes ^[32][44]. Within the same genome, different genotypes of NoVs exhibit different degrees of bioaccumulation efficiency in oysters. For example, NoV GII.4 and GII.3 have similar binding patterns to digestive tissues, gills and mantle, but the bioaccumulation efficiency varies greatly. GII.3 was effectively bioaccumulated by oysters. In contrast, the bioaccumulation level of GII.4 was very poor, and the concentration rate in oysters was less than 0.01% [21].

Several studies investigating the presence of NoV-specific ligands in oysters have found that oysters can selectively bioaccumulate NoVs via the specific binding of HBGA-like molecules to NoVs in vivo. For example, NoV GI.1 accumulates in the midgut and digestive diverticulum of oysters, but not in other tissues ^[33][45]. GII.4 accumulates in multiple tissues, including the midgut, digestive diverticula, gills and mantle ^[28][34][40,46]. Type A HBGA-like molecules are prevalent in the digestive tissue, gills and mantle of oysters, whereas type H1 and Lewis b HBGA-like molecules are mainly expressed in digestive tissues ^[19][35][19,47]. Using validation tests with monoclonal antibodies and particles of NoVs with a mutant capsid, Le Guyader et al. confirmed that NoV GI.1 bound to oyster tissues via type A HBGA-like molecules, similar to the binding of NoVs to HBGAs in human epithelial cells ^[33][45]. The binding of NoV GII to oyster digestive tissues also occurs via type A HBGA-like molecules, whilst binding to the gills and coat membranes occurs via sialic acid residues ^[20][21][20,21]. To assess the effect of ligands on the bioaccumulation of NoVs in oysters, Maalouf et al. selected three strains of NoVs (GI.1, GII.3 and GII.4) and compared their bioaccumulation efficiencies, tissue distributions and seasonal influences [21]. The findings showed that GI.1 bound to digestive tissues via type A HBGA-like molecules but hardly bound to a few other tissues and the concentration was negligible. For example, the concentration of GI.1 in the gill and coat membrane after 24 h was 1000-fold different from that in digestive tissues. The bioaccumulation efficiency of GI.1 in the digestive tissues of oysters is consistent with the seasonally dependent expression levels of the ligand [20]. In specific months, the level of GI.1 in oysters is more active, and bioaccumulation efficiency also significantly increases. In contrast, the bioaccumulation efficiency of GII.4 in oysters is low regardless of the month and shows a different tissue distribution. After the GII.4 strain enters the oyster, it preferentially accumulates in the digestive tissues within 24 h. However, most of the virus is then transferred to the gills and mantle by binding to sialic acid residues, which may prevent subsequent viral particles from entering the oyster’s digestive tissues [21]. The concentration efficiency of NoV GI in oysters is superior to that of GII. To bioaccumulate 1 viral RNA copy/g oyster tissue, the concentration of NoV GI in water needs to be 30 viral RNA copies/L water, whereas it needs to be 1200 viral RNA copies/L water for GII ^[36][48]. The result provided further evidence for NoV GI to be a dominant strain in oysters.

Inhibiting the interactions between NoVs and HBGA-like molecules in oysters is a feasible method for preventing novel transmission ^[37][49]. Some bacteria exhibit strong anti-NoVs activity and can be used to reduce NoVs infection in oysters ^[38][39][50,51]. According to Gorji et al., Bifidobacterium bifidum JCM 1254 secreted a rockulosidase that could hydrolyze certain HBGAs bound to NoVs in the digestive tissue of oysters ^[40][52]. Therefore, it was proposed that the discovery or artificial modification of specific active rockulosidase-secreting strains and their application would have great potential for reducing the risk of NoVs outbreaks. Li et al. screened one of the 122 Lactobacillus isolates for the most potent murine NoV (MNV) antagonist, Limosilactobacillus fermentum PV22, which was observed to reduce viral titers by 2.23 ± 0.38 (logarithmic) in 5 min ^[39][51]. Genome mining revealed that the antiviral activity of this strain was due to the synthesis of γ-aminobutyric acid. The results of this research suggest that the use of bacteria with anti-NoVs activity could be an effective method for preventing NoVs outbreaks.