Oyster Heat Shock Protein 70 Plays a Role in Binding of Human Noroviruses

ABSTRACT

Human noroviruses (HuNoVs) are important foodborne pathogens causing acute gastroenteritis. Oysters are an important vehicle for the transmission of HuNoVs. Histo-blood group antigen (HBGA)-like substances are considered the primary ligands for bioaccumulation of HuNoVs in oyster tissues. In this study, proteinaceous ligands for specific binding of HuNoVs were mined from oyster tissues using a bacterial cell surface display system. The macromolecular target was captured and identified in proteomic analysis. The distribution of viral particles, oyster heat shock protein 70 (oHSP 70), and type A HBGA (positive control) in oyster tissue was investigated by multiplex immunofluorescence assays after artificial contamination with HuNoVs (GII.4). Our results demonstrated that oHSP 70 is a candidate vital ligand for specific binding of HuNoVs in oyster tissues. In addition, P proteins (GI.1 and GII.4) and viral particles (GI.1 and GII.4) were captured by recombinant oHSP 70 in an enzyme-linked immunosorbent assay with a sample signal/negative signal of 7.8, 6.3, 17.0, and 8.8, respectively. The findings suggested that oHSP 70 plays an important role in the binding of these foodborne viruses.

IMPORTANCE Human noroviruses (HuNoVs) are the most important pathogen for nonbacterial epidemic gastroenteritis cases. Foodborne transmission plays an important role in HuNoVs infection. Oysters, filter-feeding epibenthic bivalves, can be contaminated by fecal discharge in harvest water. A new proteinaceous ligand for HuNoVs other than HBGA is identified in oyster tissues. The significance of our research is in identifying and verifying the ligands in oyster tissues for HuNoV binding. Our data will allow a better understanding of HuNoV attachment in and transmission by oysters, leading to the control of undesired foodborne disease.

INTRODUCTION

Noroviruses (NoVs), which belong to the family Caliciviridae (1), are the major causative agents of acute nonbacterial gastroenteritis in humans. According to the latest classification, NoVs are classified into 10 genogroups (GI to GX) (2), among which NoVs GI, GII, GIV, GVIII, and GIX are capable of infecting humans and are termed human noroviruses (HuNoVs) (2). The genogroups are further divided into genotypes, and GII.4 has dominated over past decades (3).

Shellfish remain an important vehicle of transmission of HuNoVs to the human population, particularly species like oysters, which are commonly eaten raw (4). Pacific oysters (Crassostrea gigas) are filter-feeding epibenthic bivalves, which filter up to 19 liters of water per hour per gram of shucked body weight (5) and have been found to bioaccumulate different pathogens from seawater (6). Over the past 2 decades, several studies have investigated the presence of HuNoVs in bivalve shellfish, including oysters (7). The presence of HuNoVs in oysters has been reported in the range of 14.6% to 20.7% in China (8–10), 3.9% to 20.0% in the United States (11, 12), 9.0% in France (13), and 31.0% in Ireland (14).

There are currently no effective approaches available to eliminate HuNoVs from oysters without changing the characteristics of the product (15). Understanding the distribution, binding ligands, and attachment mechanism is a high-priority task. It has been demonstrated in immunohistochemistry assays that HuNoV particles and constructed virus-like particles can bind to different tissues of oysters (6, 16). Histo-blood group antigens (HBGAs) are becoming widely accepted as receptors/coreceptors in the recognition and specific binding of HuNoVs (3, 17). Investigators have found that some substances in oyster tissues are recognized by monoclonal antibodies to HBGAs (18, 19). Binding of HuNoVs to oyster tissues is generally considered to be associated with specific HBGA-mediated interactions (18, 20, 22, 32). Other potential candidates involved in the accumulation of HuNoVs in oysters have also been proposed (20, 23). Understanding the interaction of these viruses with easily contaminated food could help to control and reduce undesired outbreaks.

In our previous studies, the protruding domain of the HuNoV capsid protein (P protein) was successfully displayed on the surface of Escherichia coli with the help of an N-terminal fragment of the ice nucleation protein (InaQn) (24, 25). The recognition of receptors by this bacterial cell surface-displayed HuNoV capsid protein system (BSDS) has been demonstrated (26–28). The recognized receptors/ligands can be separated with thrombin treatment of a “TB” (thrombin) sequence from the surface of the host E. coli for further study (26). Ligands for HuNoV attachment in lettuce have been identified with the help of a BSDS in our previous study (27). As an important HuNoV food-related vehicle, oyster tissues were analyzed to identify and verify the ligands for HuNoV binding.

In the present study, we aimed to gain a better understanding of the binding of HuNoV GII.4 in oyster tissues. Proteinaceous ligands of HuNoVs from oyster tissues were identified with the help of a BSDS (Fig. 1). Our results offer new insights into a novel ligand in the attachment mechanism of HuNoVs in oyster tissues.

FIG 1.

Brief strategy for the identification and verification of proteinaceous ligands for specific binding of human noroviruses (HuNoVs) in oyster tissues. The workflow for the separation and identification of HuNoV proteinaceous ligands in oyster tissues by bacterial cell surface-displayed HuNoV capsid protein system (BSDS) is shown in the left panel. Ways of verification of candidate ligands in this study are listed in the right panel. Abbreviations: InaQn, N-terminal fragment of ice nucleation protein; P domain, protruding domain; MS, mass spectrometry.

RESULTS

Proteins involved in the binding of HuNoVs.

E. coli BL21 strains with recombinant plasmids pET28a-inaQn-TB-P (HuNoVs GII.4) (P) and pET28a-inaQn-TB (T) were used as experimental and control BSDS in this study to identify HuNoVs ligands from oyster tissues. By identifying peptides in the thrombin-released supernatant of group P and group T, the major capsid protein of HuNoVs (GII.4) was found only in group P, which is consistent with the experimental design (data not shown). In order to eliminate nonspecific binding of BSDS, we hypothesize that the peptides found in the UniProt C. gigas database, which could be found only in group P and not in group T, were regarded as ligand candidates to bind HuNoVs in oyster tissues. According to this hypothesis, actin, coil-coil domain-containing proteins, dynein heavy chain, elongation factor, exosome complex exonuclease, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), heat shock protein 70 (HSP 70), and other 14 uncharacterized proteins were found to be ligand candidates. Among these candidates, three proteins belonging to the HSP 70 family were identified (UniProt accession numbers K1Q4I4, K1PJB0, and K1QGL3); the unique peptides for these three proteins were shown in Fig. 2. The identification of unique peptides for the corresponding protein indicated the existence of this protein.

FIG 2.

Tandem MS spectra of identified oyster heat shock protein 70 (oHSP 70) unique peptides. The identified peptides are marked on the top left of each MS spectra with fracture. The UniProt accession number of the corresponding protein is shown in the top right corner.

Distribution of HuNoVs (GII.4), HBGAs, and oHSP 70 in oyster tissue.

To verify the binding of HuNoV particles by oyster HSP 70 (oHSP 70) after contamination, we conducted multiplex immunofluorescence assays (29, 30). The type A HBGA monoclonal antibody was also involved, as it has been reported to be crucial for HuNoV binding in oyster tissues (31, 32). After labeling samples with different fluorescent dyes, antibodies to HuNoVs (GII.4), HSP 70, and type A HBGA displayed pink (CY5), green (fluorescein isothiocyanate [FITC]), and red (CY3) at the corresponding excitation and emission wavelengths (Fig. 3). In Fig. 3I, spots A, B, and C were recognized by an antibody against HuNoVs (GII.4). When merged with green (Fig. 3II), an obvious color change could be observed in these three spots, indicating the presence of HuNoVs (GII.4) and oHSP 70. However, when merged with red (Fig. 3III), the color change could be observed in spots B and C and not A, indicating that type A HBGA was not a necessary condition for HuNoV (GII.4) binding in oyster tissues. This situation was the same when the three colors were merged in that spot A could be recognized only by antibodies against HuNoVs (GII.4) and HSP 70 but not type A HBGA (Fig. 3IV). Spot D could be recognized by antibodies against HSP 70 and type A HBGA and not HuNoVs (GII.4).

FIG 3.

Multiplex immunofluorescence assays of human noroviruses (HuNoVs) (GII.4) with oyster heat shock protein 70 (oHSP 70) and type A HBGA. Cell nuclei are shown in blue by 4’,6-diamidino-2-phenylindole (DAPI). (I) Shows the artificially contaminated HuNoVs (GII.4) in pink by CY5; spots A, B, and C suggest presence of HuNoVs. (II) Shows oHSP 70 in green by FITC merged HuNoVs in pink by CY5; spots A, B, and C suggest presence of oHSP 70 and HuNoVs, and spot D suggests presence of oHSP 70. (III) Shows type A HBGA in red by CY3; spots B, C, and D suggest presence of type A HBGA, and spots B and C suggest presence of type A HBGA and HuNoVs. (IV) Shows the merged image of the four channels.

One of the identified oHSP 70s (UniProt accession number K1Q4I4) was expressed and purified in a prokaryotic system (recombinant oHSP 70 [roHSP 70]). A polyclonal antibody against roHSP 70 was used to explore the expression and distribution in oyster tissues by immunohistochemistry (Fig. 4). By comparing a with negative control (see Fig. S1 in the supplemental material), oHSP 70 was found to be widely distributed on the surface of oyster tissues, including the intestine, stomach and gills.

FIG 4.

Distribution of oyster heat shock protein 70 (oHSP 70) in oyster tissues. Cell nuclei are shown in blue by hematoxylin. oHSP 70 is shown in brown by 3, 3-diaminobenzidine (DAB). Digestive tissues along with gills are shown in the middle. Images showing the surfaces of the intestine, stomach, and gills are zoomed in and distributed around. The presence of oHSP 70 is indicated by red arrows.

Binding affinity of roHSP 70 to virus and P protein.

Enzyme-linked immunosorbent assay (ELISA) was used to further illustrate the binding affinity of roHSP 70 to HuNoV GI.1 and GII.4 (Fig. 5). When wells were coated with 1.0% bovine serum albumin (BSA), the optical density at 450 nm (OD₄₅₀) of all tested proteins or viruses was less than 0.4, and no nonspecific binding was observed. In Fig. 5, results were shown for sample signal/negative signal (S/N). Taking the criterion of an S/N of ≥2.0, type III porcine gastric mucin (PGM) (1.0 mg/ml) showed positive binding with GI.1 P protein, GII.4 P protein, GI.1 virus, and GII.4 virus with an S/N of 4.0, 2.5, 2.1, and 2.1, respectively. When the coated concentration of PGM was adjusted to 20.0 μg/ml, no positive signal was observed. However, roHSP 70 (20.0 μg/ml) showed positive binding with GI.1/GII.4 P protein and GI.1/GII.4 virus with an S/N of 7.8, 6.3, 17.0, and 8.8, respectively. The binding of the GI.1 P protein and virus was significantly higher than that of GII.4. However, for wells coated with human-derived recombinant HSP 70 (hHSP 70) (20.0 μg/ml), only the GI.1 P protein and GI.1 clinical sample showed positive binding, with an S/N of 3.8 and 2.1, respectively.

FIG 5.

Binding ability of recombinant heat shock protein 70 (roHSP 70), human-derived recombinant heat shock protein 70 (hHSP 70), and type III porcine gastric mucin (PGM) to P proteins and clinical samples of human noroviruses (HuNoVs). Samples were considered positive when the sample signal/negative signal (S/N) was equal to 2.0 or above. Heat scale for the S/N ratio is shown on the right. A higher S/N ratio indicates a higher binding ability of tested samples, represented by a darker blue in the figure. S/N ratios are inserted in corresponding boxes.

DISCUSSION

In this study, HuNoV ligand candidates in oyster tissues were identified with the help of BSDS. The basic principle is similar to the pulldown assay, which immobilizes the bait protein on immobilized beads to capture ligand candidates from a complex sample matrix by centrifugation for purification and then releasing the bait protein and ligand from the immobilized beads for further identification (27). The BSDS used in this study has some advantages over the pulldown assay in that the recombinant cells are easy to construct. There is no need to purify recombinant proteins or to immobilize on beads. Once the recombinant cells have been constructed, only culture and induction of the recombinant bacterial cells are needed, which represents a cost-effective and less variable alternative to pulldown assays.

HSP 70 is a class of functionally related proteins, which have essential roles in the cellular response to stress conditions (33). HSP 70 has also been found to be involved in immune responses and participates in the viral life cycle through folding, transporting, positioning, assembling, or degrading activities (34). Using pulldown assays, HSP 70 has been found to interact with structural proteins of Chinese sacbrood virus (35). Guerrero et al. (36) demonstrated that antibodies against HSP 70 blocked rotavirus infectivity in MA104 cells. Thus, HSP 70 has been identified as a receptor candidate for rotavirus. The expression of HSP 70 has been found to be significantly upregulated in Dojo Loach when artificially infected with bacteria (Flavobacterium columnare G4), parasites (Ichthyophthirius multifiliis), and fungus (Saprolegnia sp.) (37). After heat stress and injection with Vibrio parahaemolyticus, the expression levels of HSP 70 in the gills and hemocytes of golden and brown scallops were both significantly increased, indicating that the gene is involved in resistance or immune responses (38). As for NoVs, the upregulation of some immune and disease-related genes, including HSP 70, has been observed, although the NoVs do not cause disease in the oysters (23). In this study, oHSP 70 was captured, identified, and verified as a ligand candidate of HuNoVs, which might be involved in the specific binding of HuNoVs in oyster tissues.

According to the results of the multiplex immunofluorescence assays (Fig. 3), spot D could be recognized by antibodies against HSP 70 and type A HBGA but not HuNoVs, demonstrating that the titer of HuNoVs (GII.4) for artificial contamination was less saturated and indicating that nonspecific binding of HuNoVs (GII.4) was avoided. In spot A, the position of contaminated HuNoVs (GII.4) and oHSP 70 remained the same and was not overlapped with type A HBGA. HuNoVs (GII.4), oHSP 70, and type A HBGA could be observed in spots B and C. After being verified by a commercial human-derived recombinant HSP 70 antibody, an antibody against roHSP 70 was also prepared and used to investigate the distribution of oHSP 70 in oyster tissues. Our findings suggested that oHSP 70 is widely distributed on the surface of oyster tissues, including the intestine, stomach, and gills, providing suitable sites for HuNoVs attachment (Fig. 4).

The binding affinity of roHSP 70 was further compared with PGM. PGM is considered to contain HBGAs and the ability to bind various genotypes of HuNoVs (39). In Fig. 5, roHSP 70 identified in this study could bind to the P protein (GI.1 and GII.4) and viral particles (GI.1 and GII.4), and the binding ability was significantly higher than that of PGM. The affinity of HuNoVs with HBGAs and roHSP 70 could be revealed in further study. However, roHSP 70 used in this study was not in a native folding or modification since it was expressed by a prokaryotic system; thus, further verification is needed. Furthermore, roHSP 70 and hHSP 70 showed different binding patterns with different genotypes of HuNoVs; the conserved region and active domain of HuNoVs binding need to be studied in the future.

Sequencing and assembly of the Pacific oyster genome have been conducted previously (40), providing valuable information on the transcriptome and proteome of oysters. However, owing to a lack of sufficient information on the oyster genome, many proteins remain uncharacterized in the database. Among the identified candidate proteins, 14 out of 31 were marked as uncharacterized. Thus, no antibody and commercial purified protein could be obtained, and no further information of these proteins could be predicted, which are great challenges for characterizing these proteins and determining their binding abilities with HuNoVs. Therefore, further studies need to be conducted on these uncharacterized proteins.

MATERIALS AND METHODS

Bacterial strains and culture.

Recombinant cells P and T were stored in our lab (27). The recombinant cells were cultured in Luria-Bertani (LB) liquid medium (Hope Bio, Qingdao, China) containing 100.0 μg/ml kanamycin with shaking at 37°C. Then, 200.0 μl of overnight culture was transferred into 200.0 ml of fresh LB medium (100.0 μg/ml kanamycin) with shaking at 37°C. When the culture reached an OD₆₀₀ of approximately 0.6, isopropyl β-d-1-thiogalactopyranoside (IPTG; Merck, Germany) was added to a final concentration of 0.5 mM, and this mixture was shaken at 25°C for 12 h at 150 rpm. The cells were then washed with phosphate-buffered saline (PBS) for further use.

Oyster source and acclimation.

Oysters (C. gigas) were purchased from a local seafood market which were freshly collected from a shellfish farm in Zhanjiang, Guangdong Province, China, within 1 day. The samples were placed on ice and sent to the laboratory immediately. After being surface cleaned, 25 oysters were transferred into 20 liter of clean water with 1.8% sea salt (Yier, Guangdong, China), dissolved to simulate a marine environment. An air pump (Sunsun, Zhejiang, China) was used to supply oxygen continuously. Oyster samples used in this study were treated as described in this section.

Protein extraction.

After 24 h of acclimatization at room temperature, the oysters were shucked. The entire heart (200 mg) was extracted and ground using a sterile glass homogenizer on ice. Protein was isolated by a membrane protein extraction kit (Sangon Biotech Co., Ltd., Shanghai, China) according to the manufacturer’s instructions. A total of 200 μl of membrane protein was collected.

Releasing thrombin-sensitive components.

Recombinant cells P and T from 200 ml of culture were resuspended in 50 ml of 20 mM Tris-HCl with 200 μl of oyster protein extracted as described above. After incubation at 37°C for 30 min with shaking at 80 rpm, the mixture was centrifuged at 3,000 × g for 5 min. The cell pellet was washed with PBS and resuspended in 5.0 ml of thrombin digestion buffer (20 mM Tris-HCl containing 150 mM NaCl [pH 8.0]). Thrombin (Yeason, Shanghai, China) was added (20 IU) and incubated with shaking at 120 rpm at 37°C for 3 h. The supernatant was collected after centrifugation at 12,000 × g for 2 min.

Analyzing proteinaceous ligand candidates.

Trypsin digestion. The enzyme-released supernatant of P and T were collected from a 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel. Bands were collected, then cut into small pieces, and rinsed with Millipore water twice. NH₄HCO₃ (25 mM) in 50% acetonitrile (ACN) was added for decolorization at 37°C. The liquid supernatant was discarded, and the pellet was washed twice with ACN; the ACN was discarded, and the pellet was dried using a vacuum centrifugal concentrator (Thermo Fisher Scientific, Waltham, MA, USA). Dithiothreitol (25 mM) was then added and the mixture incubated at 55°C for 45 min. Samples were then rinsed with ACN, and iodacetamide (25 mM) was added for processing in the dark for 45 min at room temperature. The gel pieces were rinsed with ACN again, and 12.5 ng/μl trypsin was added to the gel pieces for digestion at 37°C overnight. The supernatant was subsequently collected. Then, peptides were extracted by adding 50% ACN with 0.1% trifluoroacetic acid to the gel pieces and incubating at 37°C for 30 min, with a final step of mixing the supernatant. Following evaporation of ACN with a vacuum centrifugal concentrator, the sample was desalted with a C₁₈ ZipTip (Millipore), and the eluate was analyzed using nano-liquid chromatography-electrospray tandem mass spectrometry (nano-LC-MS/MS).

Nano-LC-MS/MS analysis.

Samples were resuspended with 30 μl solvent C (water with 0.1% formic acid), separated with nano-LC-MS/MS, and analyzed using online electrospray tandem mass spectrometry. The experiments were performed on a Nano Acquity ultraperformance liquid chromatography (UPLC) system (Waters Corporation, Milford, MA, USA) connected to a Q-Exactive mass spectrometer (Thermo Fisher Scientific) equipped with an online nano-electrospray ion source. The peptide sample was loaded onto the trap column (Thermo Scientific Acclaim PepMap C₁₈; 100 μm by 2 cm), with a flow of 10 μl/min for 3 min, and subsequently separated on the analytical column (Thermo Scientific Acclaim PepMap C₁₈; 100 μm by 2 cm) with a 90-min linear gradient, from 5% D (ACN with 0.1% formic acid) to 55% D. The column was re-equilibrated under initial conditions for 10 min. The column flow rate was maintained at 300 nl/min. The electrospray voltage of 2 kV versus the inlet of the mass spectrometer was used. The mass spectrometer was run under data-dependent acquisition mode and automatically switched under MS and MS/MS mode. MS1 mass resolution was set to 70 K with an m/z of 300 to 1800, and the MS/MS resolution was set to 17.5 K under higher-energy collisional dissociation (HCD) mode. The dynamic exclusion time was set to 10 s.

Database comparison.

Tandem mass spectra were processed using PEAKS Studio version 8.5 (Bioinformatics Solutions Artificial contamination Inc., Waterloo, Canada). A PEAKS database (DB) was set up to search the UniProt database assuming the digestion enzyme trypsin. Species including Norovirus sp. and C. gigas were included in this study. For the inevitable fragmentation of host cells, the UniProt E. coli database was also included for tandem mass spectra analysis. The PEAKS DB was searched with a fragment ion mass tolerance of 0.05 Da and a parent ion tolerance of 10 ppm. Carbamidomethylation was specified as a fixed modification. Oxidation (M), deamidation (NQ), and acetylation (protein N-term) were specified as variable modifications. Both protein and peptide false discovery rates (FDRs) were controlled at ≤1%, while a minimum of one unique peptide per protein was required.

Distribution of HuNoVs (GII.4) and viral ligands in artificially contaminated oyster tissue by multiplex immunofluorescence assays.

After 24 h of acclimation, five fresh oysters were immersed in 4 liters of water with 1.8% sea salt under oxygenation together with GII.4 clinical samples (final titer around 10⁶ RNA copies/liter) at room temperature for 5 h for bioaccumulation. The entire heart was extracted and fixed in a 4% paraformaldehyde fixing solution (Sangon Biotech Co., Ltd.), then embedded in paraffin, and sliced into thin sections. Paraffin sections were rinsed with xylene, ethanol, 85% ethanol, 75% ethanol, and distilled water. Microwave antigen retrieval was carried out as described previously (41). Then, the specimens were washed three times with PBS. Bovine serum albumin (BSA; 1.0%; Yeason) was added for blocking for 30 min. After a washing step, 1:3,000 diluted primary antibodies were used for incubation at 4°C overnight. Primary antibodies used in this study included mouse anti-type A HBGA monoclonal antibody (Covance, Emeryville, CA, USA), rabbit anti-hHSP 70 polyclonal antibody (Beyotime, Shanghai, China), and mouse anti-HuNoVs GII.4 monoclonal antibody. The slides were then washed with PBS three times; the 1:5,000-diluted corresponding secondary antibodies (Servicebio, Wuhan, China) were then added, and specimens were incubated for 50 min at room temperature. After three washes with PBS, 100 μl of fluorescent dyes was added, and the slides were incubated in the dark for 10 min. Slides were then washed with Tris-buffered saline containing 0.5‰ Tween 20 (TBST) three times. Microwave antigen retrieval was then carried out before the same processing was conducted for the other primary antibodies. After all antibodies were processed, 100 μl of a spontaneous fluorescence quenching agent (Servicebio) was added, and the specimens were incubated for 5 min and then washed with running water. Dye 4′,6-diamidino-2-phenylindole (DAPI) was used to visualize a hyperchromatic nucleus. The slides were then washed with PBS (pH 7.4) three times and sealed with an antifluorescence quenching sealer (Servicebio). Images were obtained using a fluorescence microscope (Nikon). The corresponding fluorescent dyes were CY3 for type A HBGA, with an excitation wavelength of 550 nm and emission wavelength of 590 nm; FITC for HSP 70, with an excitation wavelength of 488 nm and emission wavelength of 525 nm; and CY5 for HuNoVs GII.4, with an excitation wavelength of 647 nm and emission wavelength of 666 nm. Merged figures were visualized by the software CaseViewer.

Purification of roHSP 70.

The entire nucleic acid sequence of oHSP 70 (UniProt accession number K1Q4I4) was analyzed for codon optimization and synthesized by Sangon Biotech, Co. The sequence was inserted into plasmid pET-28a (+) and transferred into competent E. coli BL21 cells. Recombinant cells were cultured as described in bacterial strains and culture. The overnight culture was then washed with PBS (pH 7.2) and disrupted using ultrasonication (Scientz, Ningbo, China). Purification of roHSP 70 was carried out by Ni-nitrilotriacetic acid (NTA) beads 6FF (Smart-Lifesciences, Changzhou, China) according to the manufacturer’s instructions. The P protein of HuNoVs GI.1 and GII.4 was purified in the same way (GI.1 P protein and GII.4 P protein).

Distribution of identified oHSP 70 in oyster tissues.

Fresh oyster samples were collected as described above and then shucked. After 24 h of acclimation, digested tissues including the gills were retained, extracted, fixed in 4% paraformaldehyde fixation solution (Sangon Biotech Co., Ltd.), and then embedded in paraffin and sliced into thin sections. The sections were rinsed in xylene, ethanol, 85% ethanol, 75% ethanol, and distilled water. The specimens were then washed three times with PBS. The activity of endoperoxidase was blocked with a 3.0% H₂O₂ solution in the dark at room temperature for 25 min, and specimens were then washed three times with PBS. BSA (1.0%) was then added for blocking for 30 min. The mouse anti-roHSP 70 polyclonal antibody was prepared by Friendbio Science & Technology (Wuhan, China), Co., Ltd., and used as the primary antibody. After a washing step, 1:3,000 of diluted primary antibody was applied and the specimens were incubated at 4°C overnight. After being washed three times with PBS, horseradish peroxidase (HRP)-conjugated secondary antibody (Servicebio, Wuhan, China) was added and incubated for 50 min at room temperature. The slides were then washed three times with PBS, and 3, 3-diaminobenzidine (DAB) was added to label the ligands. Finally, the slides were stained with hematoxylin and eosin (Servicebio) and observed under a microscope (Nikon, Japan).

Binding of P protein and HuNoVs to ligands measured by ELISA.

PGM (Sigma, St. Louis, MO), roHSP 70, and hHSP 70 (Sino Biological, Beijing, China) were diluted to 20.0 μg/ml and coated onto wells (100.0 μl in each well) of ELISA plates (Sangon Biotech Co., Ltd.) at 4°C overnight. PGM (1.0 mg/ml) was used as a positive control. After being washed twice with PBS, the wells were blocked with 1.0% BSA at 37°C for 1 h. After another three washes with PBS, 100 μl P protein (GI.1 or GII.4; 10.0 μg/ml) or HuNoV clinical samples (GI.1 or GII.4, stored in our laboratory [42]) were diluted to the final titer (10² to 10³ RNA copies/μL) and then added into each well. After incubation at 37°C for 1 h, wells were washed with TBST three times. Primary antibodies against the corresponding genotypes of P protein or viral particles stored in our lab were added to each well, and the plates were incubated for 1 h at 37°C (25). Unbound antibodies were removed by washing the wells with TBST four times. The secondary antibody peroxidase-conjugated goat anti-mouse IgG (Sangon Biotech Co., Ltd.) was added to the appropriate wells and then incubated at 37°C for 1 h. After three washes with TBST, 3,3′,5,5′-tetramethylbenzidine (TMB; Frdbio, Wuhan, China) was added to each well. After incubation in the dark for 10 min, the reaction was stopped using 2.0 mol/liter H₂SO₄. The plates were read at 450 nm with a Sunrise microplate reader (Tecan Sunrise, Switzerland). In the negative-control group, a 1.0% BSA solution was used to coat the wells, with other reagents treated in the same way. Samples were considered positive when the S/N was equal to 2.0 or above (43). Each experiment was repeated three times in parallel. Data were analyzed by GraphPad Prism 7. A P value of <0.05 was considered statistically significant.

Data availability.

A detailed description on the bacterial cell surface-displayed HuNoV capsid protein system (BSDS) utilized in this study could be obtained from references 25–28. Information on HSP 70 identified in this study could be fetched by UniProt accession numbers K1Q4I4, K1PJB0, and K1QGL3.