Development and Application of a Novel Rapid and Throughput Method for Broad-Spectrum Anti-Foodborne Norovirus Antibody Testing

Foodbone norovirus (NoV) is the leading cause of acute gastroenteritis worldwide. Candidate vaccines are being developed, however, no licensed vaccines are currently available for managing NoV infections. Screening for stimulated antibodies with broad-spectrum binding activities can be performed for the development of NoV polyvalent vaccines. In this study, we aimed to develop an indirect enzyme-linked immunosorbent assay (ELISA) for testing the broad spectrum of anti-NoV antibodies. Capsid P proteins from 28 representative NoV strains (GI.1–GI.9 and GII.1–GII.22 except GII.11, GII.18, and GII.19) were selected, prepared, and used as coating antigens on one microplate. Combined with incubation and the horseradish peroxidase chromogenic reaction, the entire process for testing the spectrum of unknown antibodies required 2 h for completion. The intra-assay and inter-assay coefficients of variation were less than 10%. The new method was successfully performed with monoclonal antibodies and polyclonal antibodies induced by multiple antigens. In conclusion, the indirect ELISA assay developed in this study had a good performance of reliability, convenience, and high-throughput screening for broad-spectrum antibodies.


INTRODUCTION
Foodbone Norovirus (NoV) is one of the leading causes of epidemic acute gastroenteritis (AGE) worldwide, accounting for one-fifth of all gastroenteritis cases globally (Lopman et al., 2016). Annually, NoV has been estimated to cause 677 million episodes of diarrheal disease (95% uncertainty interval [UI]: 468-1,153 million) and 213,515 deaths (95% UI: 171,783-266,561) for all ages and among all modes of transmission (Pires et al., 2015). NoV is extremely contagious, and transmission occurs directly from person to person via fecal-oral and vomit-oral routes, but can also be caused by food-borne, water-borne, or environmental factors (Verhoef et al., 2015). Vomitus and feces of infected patients contain a considerable number of virions, whereas as few as 10 infectious particles are sufficient to cause AGE (Teunis et al., 2010). Additionally, NoV shedding after infection usually lasts for several weeks, and prolonged shedding post-infection can also contribute to spreading (Sukhrie et al., 2010). Thus, NoV remains a major threat to public health. Owing to the significant social and economic burden associated with the disease, adequate preventive measures against this virus should be formulated.
NoV strains can be classified into 10 genogroups (GI-GX) and at least 48 genotypes based on the diversity of the VP1 amino acid sequence; GI, GII, and GIV have been shown to cause infections in humans (Chhabra et al., 2019). Human histo-blood group antigens (HBGAs) can be used as receptor or co-receptor for NoV infection (Lindesmith et al., 2003;Rockx et al., 2005). The binding ability between NoVs and HBGAs is diverse, resulting in different infectivity. The global pandemic GII.4 NoV can infect almost all secretory individuals (Nordgren and Svensson, 2019); GI.1 VLP binds to A, AB, and O type, but not to B type saliva (Morozov et al., 2018); individuals with O type saliva are relatively susceptible to NoV, while with B type are not (Hutson et al., 2002); some strains do not bind any phenotype of HGBAs (Almand et al., 2017).
Several possible barriers hinder the development of effective NoV vaccines (Cortes-Penfield et al., 2017). First, circulating NoV exhibits genetic diversity and high variation, which may limit the durability of protection conferred by a vaccine that does not elicit broadly neutralizing antibodies. Second, the lack of sufficient cell lines for virus culture and successful animal models further hinders the development of antiviral drugs and vaccines. The following three types of NoV vaccines have been developed: virus-like particles (VLPs) that resemble the organization and morphology of native virus, P particles that resemble the P domain of native virus, and recombinant adenoviruses (Esposito and Principi, 2020).
Enzyme-linked immunosorbent assay (ELISA) is a sensitive and specific analytical biochemistry assay utilized for the detection and quantitative or qualitative analysis of an analyte without the requirement for sophisticated or expensive equipment (Gan and Patel, 2013). Currently, ELISA is used as an ideal method for the detection of viruses and antibodies (Fan et al., 2013;Ledesma et al., 2017;Ding et al., 2019). However, few ELISA techniques demonstrate the ability to directly detect the broad-spectrum binding activities of antibodies against all NoV strains that infect humans. Therefore, it is necessary to develop an indirect ELISA technique for the detection of a wide range of antibodies.
In this study, we aimed to develop an ELISA technique for vaccine development via rapid, efficient, and accurate screening of a substantial number of broad-spectrum antibodies.

Selection of Representative Strains
The antigen used in this experiment was capsid P protein for each representative strain of NoV stored in our laboratory. We selected nine GI strains (GI.1-GI.9) and nineteen GII strains (GII.1-GII.22) as representative NoV strains; GII.11, GII.18, and GII.19, which were detected in swine (Vinjé, 2015), were excluded. In our FIGURE 1 | Step-by-step schematic representation of ELISA. Before addition of the antibody to be tested, the dilution corresponding to the OD value of 0.8-1.2 was determined. Broad-spectrum norovirus antibodies were determined within 2 h. previous study, GII.2,GII.3,GII.4,GII.6,GII.8,and GII.17 NoV were detected and preserved (Xue et al., 2015;Xue et al., 2016;Xue et al., 2019), which could be directly used for cloning. Others were synthesized.

Pre-coated of Antigen
For preparation of microplates, 96-well plates were coated with 0.2 µg of 28 NoV capsid P proteins after incubation at 4 • C overnight (three well per P protein). After subjecting the wells to washing steps with phosphate-buffered saline containing Tween 20 (PBST) three times, the wells were blocked using 5% skim milk for 2 h at 37 • C; then, the wells were washed three times using PBST, dried, and stored at 4 • C until further use. A combination of glutathione S-transferase (GST)-tagged protein and commercial anti-GST antibodies was used as a quality control. The negative serum was used as a negative control.

Enzyme Linked Immunosorbent Assay (ELISA)
The indirect ELISA method was used to detect the broad spectrum of antibodies (Figure 1). Briefly, the pre-coated microtiter plate was taken out. Serum samples were added at an appropriate dilution, which corresponded to the OD value in the range of 0.8-1.2, followed by incubation for 1 h at 37 • C. Subsequently, secondary horseradish peroxidase-conjugated goat anti-mouse IgG antibodies (diluted 1: 3,000) were added and the plates were incubated for 30 min at 37 • C. Tetramethylbenzidine was added as a peroxidase chromogenic substrate and reaction for 8 min at 37 • C was performed. After terminating the substrate reaction with 2 M H 2 SO 4 , the optical density (OD) was measured at 450 nm. Positive signal is defined as a mean OD 450 >0.

Intra-Assay Precision
Three replicates for each monoclonal antibody sample were analyzed using the same microplate. Precision was determined by calculating the mean, standard deviation, and coefficient of variation (CV%).

Inter-Assay Precision
For determination of inter-assay reproducibility, three replicates for each sample were analyzed in different microplates. Precision was determined by calculating the mean, standard deviation, and CV%. FIGURE 2 | Distribution of antigen coating in 96-well microplate. In the plate, each well was coated with 0.2 µg P protein, and each P protein was coated with three multiple wells. Wells coated with GST and coating buffer served as positive control and negative control, respectively.

Statistical Analysis
Statistical analyses were performed using Microsoft Excel 2016 and GraphPad Prism 8.0 software. Broad spectrum plot data are presented as mean ± standard deviation.

Assembly of Microplates
To perform this experiment, we selected representative strains for capsid P proteins of all genotypes in GI (GI.1-GI.9) and GII (GII.1-GII.22), except for GII.18, GII.19, and GII.20 (Table 1 and Supplementary Figure 1). As shown in Figure 2, the microplate was coated with 28 capsid P proteins obtained from representative strains, and each P protein was coated in three wells.

Reproducibility of ELISA
Next, we evaluated intra-assay and inter-assay variabilities to test the reproducibility of the method. The monoclonal TABLE 2 | Intra-and inter-assay coefficients of variation (%) for the broad-spectrum detection of GII.4-4A21C1 monoclonal antibodies by ELISA.

DISCUSSION
Owing to the high frequencies of NoV outbreaks (Hall et al., 2011;Verhoef et al., 2013;Grytdal et al., 2016) and severe acute gastroenteritis (Atmar and Mullen, 2013;Cardemil et al., 2017), high risk of infection transmission (Sukhrie et al., 2010;Teunis et al., 2010;Verhoef et al., 2015), and enormous socioeconomic burden after infection (Pires et al., 2015), the development of NoV vaccines is critical for improving public health. However, the characteristics of NoV genetic diversity and high variation have markedly hindered vaccine development. At present, NoV vaccines rely on VLP and P particles prepared by bioengineering technology as subunit vaccines. The morphology and antigenicity of recombinant NoV VLP are similar to natural virus particles (Estes et al., 2000). Preclinical studies have shown that VLP is immunogenic when administered by intranasal, oral or parenteral routes, which can induce serum and mucosal immune responses, and co-administration with mucosal adjuvants induces stronger immune responses (Estes et al., 2000). Three surface loops per P domain are potential sites for foreign antigen presentation, which could enhance the antigenicity and immunogenicity of antigens (Tan et al., 2011). Su et al. (2015) have shown that immune P particles induced antiserum in both mice (1:245,600) and rabbits (1:145,700), and also induced neutralizing antibodies to block the binding of NoV variants and receptors. The results prove the immunogenicity of P particles. Therefore, both NoV VLP and P particles are suitable choices for vaccine preparation.
Several NoV vaccines using a variety of different technologies have been developed, including three in clinical trials. The first is the bivalent intramuscular GI/GII.4 vaccine developed TABLE 3 | Intra-and inter-assay coefficients of variation for the broad-spectrum detection of GII.4-5D41A12 monoclonal antibodies by ELISA.

GII.4-5D41A12
Intra-assay Precision CV/% Inter-assay Precision CV/%  by Takeda Pharmaceutical Company Limited (Treanor et al., 2014). The second, developed by Vaxart, Inc., is based on a G I.1 NoV sequence (Kim et al., 2018). The third is the NoV tetravalent vaccine developed by Institut Pasteur of Shanghai Chinese Academy of Sciences and Chongqing Zhifei Biological Products Co., Ltd. (XINHUANET, 2019). Therefore, analysis of cross-protection among different strains is a significant step for the development of a successful NoV vaccine. The immune cross-protection between different genotypes is complicated. Studies have shown GI.1 can induce blockade antibodies against other GI strains (Lindesmith et al., 2015); GII.4 antiserum and GII.17 antiserum have limited cross-reactivity or no cross-protection; Malm et al. (2015) demonstrated there is no cross-cluster protection between GI and GII, but the research of Czakó et al. (2015) showed GI.1 infection will cause an increase of GII.4 antiserum. In short, the development of effective vaccines must contain both GI and GII viruses.
In addition to vaccine development, detection technology is also an important part of NoV prevention and control. The immunoassay methods mainly include enzyme linked immunosorbent assay (ELISA) and immunochromatographic (ICG). In this study, capsid P proteins of 28 representative NoV strains were used as coating antigens for the development of an indirect ELISA to detect their binding ability with antibodies. The broad spectrum of NoV antisera could be quickly detected using this method. Furthermore, one genotype or a combination of genotypes with broad-spectrum detection could be screened for the design of multivalent vaccines. Owing to the genetic diversity of NoV and 5%-10% cross-reactivity observed among genotypes (lower cross-reactivity among genogroups), the development of a broad-spectrum multivalent vaccine platform is necessary to protect against as many NoV-prevalent strains as possible (LoBue et al., 2006).
In summary, in this study, we developed an indirect ELISA with the potential to serve as a rapid, reliable, and highthroughput method for screening broad-spectrum antibodies to expedite the development of NoV vaccines.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.