Characterisation of a household norovirus outbreak occurred in Valencia (Spain)

Subjects and samples

Eight individuals were enrolled in this study, none of whom had taken any medication at or around the time of sample collection. The biosafety and ethics committees of the University of Valencia, Spain, approved the present study. Appropriate informed written consents were obtained from all individuals. Of these, at least 4 subjects were in contact with the suspected source of infection, a previously infected individual who was considered the index case, and three of them showed typical symptoms of NoV infection (nausea, vomiting, diarrhoea, abdominal pain and fever). The family members were living together and the outbreak was considered to be person-to-person transmitted. In order to determine the etiological agent of this outbreak, stool samples were collected from involved individuals with and without clinical symptoms. Serum samples from 3 household members were collected at 14 days (considered convalescent serum) and 1 year (considered memory serum) post-infection to analyze NoV-specific IgG antibodies and their blocking capacity. Blood samples were also collected to determine the histo-blood group antigens of each individual by hemagglutination. Saliva samples from all the household members were collected. The freshly collected saliva samples were centrifuged at 10,000 × g for 5 min to remove particulate material, host and microbial cells and boiled at 100 °C to inactivate antibodies. The supernatants were collected, divided into several aliquots and stored at -80 °C until their use. To minimize the effects of the circadian rhythm, saliva samples were consistently collected in the morning hours (between 8 and 10 a.m.). Also, participants were instructed not to smoke, eat, drink, or brush their teeth in the 2 h before saliva collection. For genotyping and NoV-specific IgA antibody detection, saliva was centrifuged, divided into several aliquots and stored directly at -80 °C without boiling to preserve IgA integrity.

Norovirus detection

Viral RNA was extracted from 20 % stool suspensions in PBS using the Trizol LS reagent (Invitrogen, Paisley, Scotland), eluted in diethyl pyrocarbonate-treated water containing RNasin (Promega, Madison, WI) and stored at -80 °C. Reverse transcriptase-polymerase chain reaction (RT-PCR) for NoV was performed as described previously [16]. The genotype of NoV strains was determined by partial sequence analysis of the polymerase gene and the genotype was identified and classified using the typing tool at www.rivm.nl/mpf/norovirus/typingtool based on the similarity of the sequences to reference strains representing known genotypes [17].

Secretor status analysis

The secretor (FUT2+) and non-secretor (FUT2-) status was investigated by genotyping the FUT2 gene. Genomic DNA was extracted from saliva samples using the Qiagen QIAamp DNA Mini Kit. Genotyping for FUT2 was performed by PCR-RFLP as described previously by Marionneau et al. [18]. A fragment of the FUT2 gene was amplified with primers 5′-GAGGAATACCGCCACATCCCGGGGGAGTAC-3′ (forward) and 5′-ATGGACCCCTACAAAGGTGCCCGGCCGGCT-3′ (reverse) and was digested with Ava II (Fermentas, Life Technologies, Alcobendas, Spain) for 2.5 h at 37 °C.

ABO blood type and Lewis antigen determination

The presence of HBGAs was determined by detection of group A and B glycans by hemagglutination assay (ALBAclone® Monoclonal ABO Antisera, Alpha Laboratories, Eastleigh, England) and Lewis antigens were phenotyped by ELISA with monoclonal antibodies against Le^a, Le^b, Le^x and Le^y (Covance, Dedham, MA, USA). Briefly, plates were coated with saliva samples at 1/1,000 dilution in carbonate/bicarbonate buffer (pH 9.6) at 4 °C overnight. After blocking with PBS containing 3 % (w/v) bovine seroalbumin (PBS-BSA), plates were incubated with the different anti-Lewis antigen monoclonal antibodies (anti-Le^a BG-5, Le^b BG-6, Le^x BG-7 or Le^y BG-3, Covance) at 1/100 during 1 h at 37 °C and detected by a secondary mouse antibody mix (anti-IgG, anti-IgM and anti-IgA) (Sigma, Madrid, Spain) conjugated to horseradish peroxidase (HRP). After each step, plates were washed with PBS containing 0.05 % of Tween-20 (PBS-T). The reaction was developed by the addition of OPD Fast (Sigma) and stopped after 10 min incubation with 3 M H₂SO₄. Absorbance was measured at 492 nm in a microplate reader (Multiskan FC, Thermo Scientific, Alcobendas, Madrid).

Norovirus VLPs

In the present study, VLPs corresponding to GII.4-Den Haag_2006b genotype were used. The recombinant baculovirus expressing the GII.4-2006b was already available in the laboratory [19]. The NoV VLPs were produced and purified as previously described [19]. Because VLPs from the GII.4-New Orleans_2009 variant causing the NoV gastroenteritis outbreak were not available for patients’ antibody testing, analyses were performed with VLPs of the GII.4-Den Haag_2006b variant, which share a 93.66 % homology at the amino acid level with the GII.4-2009 variant.

Norovirus-specific IgA antibody in saliva samples

Microtiter plates were coated GII.4-2006b VLPs at 2 μg/ml in carbonate/bicarbonate buffer (pH 9.6) and incubated at 4 °C overnight. Plates were washed three times with PBS-T and blocked for 1 h at 37 °C with PBS-T 3 % BSA. After blocking, plates were incubated with serial dilutions of saliva samples (from 1/20 to 1/160) in PBS-T 1 % BSA for 1.5 h at 37 °C and NoV-specific IgA antibody was detected with a secondary anti-IgA human antibody conjugated with HRP (Sigma) at a dilution of 1/4,000 in PBS-T 1 % BSA for 1 h at 37 °C. As controls, all saliva samples were tested in parallel in wells coated with BSA at 2 μg/ml. To rule out nonspecific binding (false positives), boiled saliva samples were also used as negative controls on wells coated with VLPs. Wells were washed four times with PBS-T, and the bound antibody was detected by the addition of 50 μl of o-phenylenediamine (Sigma). The reaction was stopped at 10 min with 3 M H₂SO₄, and the absorbance was read at 492 nm (Multiskan FC spectrophotometer, Thermo Scientific).

Norovirus-specific IgG antibody titers in serum samples

Purified GII.4-2006b VLPs were used to coat 96-well polystyrene microtiter plates as described above. After blocking with PBS-BSA 3 %, plates were incubated with serial serum dilutions from 1/200 to 1/12,800 diluted in PBS-T containing 1 % BSA for 1 h at 37 °C. Then HRP-conjugated anti-human IgG antibody (Santa Cruz Biotechnology) diluted 1/4,000 was added for 1 h and the reaction was developed and measured as described above. A sample was considered positive when its OD was three times higher than the absorbance value of the same sample in the specificity control (wells coated with 2 μg/ml of BSA). The titer of each sample was considered the inverse of the last dilution showing a positive reaction.

Binding blocking assays

Microtiter plates were coated with saliva at 1/500 in carbonate/bicarbonate buffer (pH 9.6) overnight at 4 °C. Plates were washed three times with PBS-T and blocked for 1 h at 37 °C with PBS-T 3 % BSA. Plates were then incubated for 1.5 h with GII.4-2006b VLPs (2 μg/ml) that had been pre-incubated during 1 h at 37 °C in the presence of the different sera at serial 2-fold dilutions from 1/100 to 1/3.200, or with PBS in the negative control. The mixtures were then added to the saliva-coated plates and incubated for 1 h at 37 °C. The detection was performed with the anti-NoV rabbit polyclonal serum at 1/1,000 followed by a secondary HRP-conjugated anti-rabbit IgG antibody diluted at 1/2,000.

Blocking of VLP binding by the tested sera was determined by comparing the OD values obtained in wells in duplicate containing potential blocking reagents against the control wells (without the blocking steps). Blocking was considered when the OD value was less than 50 % of the positive control OD value (ranging from 0.8 to 1.2 OD₄₉₂ units).

Gastroenteritis outbreak was caused by norovirus GII.4-New Orleans_2009 variant

In January 2012 three members of a household in Valencia showed typical symptoms of acute gastroenteritis caused by NoV 24 h after having been in contact with the index case. Clinical symptoms were nausea, vomiting, diarrhoea and fever, which lasted approximately 48 h. One of the patients had to be taken care of dehydration at the emergency room.

RT-PCR assay performed on stool samples collected (n = 3) from two symptomatic patients and from one asymptomatic individual revealed the presence of NoV GII.4, specifically the GII.4-New Orleans_2009 variant. As no other enteric viruses (rotavirus or adenovirus) were detected by molecular analyses neither enteropathogenic bacteria were isolated, it was concluded that NoV was the etiological agent of this outbreak. The most relevant result was that the asymptomatic individual also shed NoVs as his faecal sample was also positive.

Non-secretor individual (sese) showed no symptoms after norovirus GII.4-New Orleans_2009 infection

Titers of salivary anti-norovirus IgA antibodies

Salivary IgA antibodies to the GII.4-2006b were positive in all samples. Although the number of samples was too small to perform statistical analyses, no differences were found in secretory IgA levels from the two analyzed groups: secretors and non-secretors (Fig. 1). These results suggest that non-secretors are also susceptible to NoV infection even if they do not show clinical symptoms, as it was also confirmed by RT-PCR.

Norovirus-specific IgG antibody titers

Three sera were collected and compared with a convalescent serum from a sporadic case of NoV infection, which was used as the positive control. Of these sera, two were from secretor individuals that showed symptoms and the other one was from an infected non-secretor without symptoms.

All tested convalescent serum samples showed antibody titers higher than 1/100 to the GII.4-2006b variant. IgG serum titers showed significant differences (P < 0.01) between secretors and the non-secretor (Fig. 2). Titers of NoV-specific IgG antibody of 1/12,800 were the highest titers found, corresponding to the two symptomatic infected individuals in their convalescence (14 days p.i.) and memory (1 year p.i.) serum samples, independently whether they were homo- or heterozygous for the FUT2 mutation. The serum from the non-secretor individual showed no changes in its antibody titers, despite having been in contact with the virus and showing the lowest antibody titer (Fig. 2).

Blockade of the binding of VLPs to saliva

We examined the blocking capacity of the different serum samples (both convalescent and memory sera) on the binding of GII.4-2006b VLPs to saliva from a secretor, blood type O individual. As expected, the binding to receptors present in the saliva was blocked efficiently by secretor individual sera, up to 60 % (Fig. 2). On the other hand, the non-secretor serum sample did not block the binding at all. The results obtained with the memory serum showed the same blocking rate as the corresponding convalescent serum samples.