Skip to main content
NCBI home page
Protein Engineering, Design and Selection logoLink to Protein Engineering, Design and Selection
. 2014 Jun 19;27(10):339–349. doi: 10.1093/protein/gzu023

Identification of human single-chain antibodies with broad reactivity for noroviruses

Wanzhi Huang 1, Moumita Samanta 1, Sue E Crawford 2, Mary K Estes 2,3, Frederick H Neill 2, Robert L Atmar 2,3, Timothy Palzkill 1,2,4
PMCID: PMC4191442  PMID: 24946948

Abstract

Norovirus infections are a common cause of gastroenteritis and new methods to rapidly diagnose norovirus infections are needed. The goal of this study was to identify antibodies that have broad reactivity of binding to various genogroups of norovirus. A human scFv phage display library was used to identify two antibodies, HJT-R3-A9 and HJT-R3-F7, which bind to both genogroups I and II norovirus virus-like particles (VLPs). Mapping experiments indicated that the HJT-R3-A9 clone binds to the S-domain while the HJT-R3-F7 clone binds the P-domain of the VP1 capsid protein. In addition, a family of scFv antibodies was identified by elution of phage libraries from the GII.4 VLP target using a carbohydrate that serves as an attachment factor for norovirus on human cells. These antibodies were also found to recognize both GI and GII VLPs in enzyme-linked immunosorbent assay (ELISA) experiments. The HJT-R3-A9, HJT-R3-F7 and scFv antibodies identified with carbohydrate elution were shown to detect antigen from a clinical sample known to contain GII.4 norovirus but not a negative control sample. Finally, phages displaying the HJT-R3-A9 scFv can be used directly to detect both GI.1 and GII.4 norovirus from stool samples, which has the potential to simplify and reduce the cost of diagnostics based on antibody-based ELISA methods.

Keywords: antibody library, diagnostics, norovirus, phage display, single-chain antibody

Introduction

Noroviruses (NoVs) are recognized as a leading cause of acute gastroenteritis (Atmar and Estes, 2006; Glass et al., 2009; Scallan et al., 2011). They are associated with large outbreaks involving hospitals, nursing homes, cruise ships, schools and other institutional settings. The median infectious dose of virus has been estimated to be 18–1320 virions and transmission of the virus can be food-borne, person-to-person or from environmental sources (Atmar and Estes, 2006; Glass et al., 2009; Scallan et al., 2011; Atmar et al., 2014).

NoVs are small, round RNA viruses that belong to the family Caliciviridae. They contain a single-stranded, positive sense genome that encodes three open reading frames (ORFs). ORF1 encodes non-structural proteins such as the polymerase and protease while ORF2 encodes the major capsid protein (VP1) and ORF3 encodes a minor structural protein (VP2) (Glass et al., 2009). Expression of ORF2 from insect cells results in the self-assembly of VP1 into virus-like particles (VLPs) that are antigenically and morphologically similar to native virions (Jiang et al., 1992; Green et al., 1993). Structural studies of VLPs and VP1 indicate that the major capsid protein can be divided into three domains with the shell (S) domain located at the interior and the protruding domains P1 and P2 progressively more exposed on the surface (Prasad et al., 1999). VLPs bind to carbohydrates from histo-blood group antigens (HBGAs) that are found on the surface of intestinal epithelial cells and these sites are thought to serve as attachment factors for virus (Hutson et al., 2003, 2004; Tan and Jiang, 2010). The X-ray crystal structure of the P-domain of Norwalk virus (NV) in complex with A- and H-type HBGAs indicates that the P2 domain contains the binding site for carbohydrate (Bu et al., 2008; Choi et al., 2008). Structural studies further indicate that the position of the HBGA-binding site on the P2 domain varies between genogroups of NoV (Cao et al., 2007; Choi et al., 2008; Hansman et al., 2011; Kubota et al., 2012).

NoVs are a genetically diverse group of viruses that have been classified into six genogroups (I–VI) based on the major capsid sequence (Kroneman et al., 2013). The human NoVs include genogroups I, II and IV and these have been further subdivided into at least 9 GI and 19 GII genotypes (Zheng et al., 2006; Kroneman et al., 2011). A large amount of amino acid sequence diversity occurs in VP1, particularly in the P2 protruding domain (Chen et al., 2004; Donaldson et al., 2010; Debbink et al., 2012). The diversity of NoV strains creates a challenge in the development of diagnostic assays that can be used to broadly detect NoVs. Three methods that have been used to detect NoVs include electron microscopy (EM), reverse transcription-polymerase chain reaction (RT-PCR) or, more recently, real-time RT-PCR, as well as enzyme-linked immunosorbent assay (ELISA) (Atmar and Estes, 2001; Kageyama et al., 2004; Glass et al., 2009). EM can be used to directly detect virions in stool samples; however, this method is work intensive and is less sensitive than molecular methods (Richards et al., 2003). RT-PCR is the most widely used method of detection and involves the use of virus-specific primers that are complementary to conserved regions in the genome (Atmar and Estes, 2001). The sequence diversity of NoV prohibits the use of a single primer pair for broad detection; however, the inclusion of two primer pairs allows detection of >90% of GI and GII viruses (Blanton et al., 2006).

The detection of viral antigens in stool by ELISA has also been used as a tool to diagnose NoV infections (Gray et al., 2007). ELISA-based methods have the advantage of ease of use and do not require specialized equipment. The difficulty with this approach has been the specificity of the assay. Because of the extensive diversity of the P-domain of VP1, antibodies that recognize viral surface antigens are quite specific for genogroups or even genotypes of NoV (Jiang et al., 1995). For example, hyperimmune sera raised against genogroup I or II VLPs were found to react specifically with GI or GII samples, respectively (Jiang et al., 1995; Atmar and Estes, 2001). Monoclonal antibodies (mAbs) have been developed, however, that are more broadly reactive within a genogroup and some have been used for diagnostic assays (Kitamoto et al., 2002; Parker et al., 2005; Lindesmith et al., 2012; Sakamaki et al., 2012). These ELISA-based diagnostic assays exhibit modest sensitivity (38%, Dako; 36%, Ridascreen) but high specificity (96%, Dako; 88%, Ridascreen) (de Bruin et al., 2006). Therefore, there is a need for the development of antibodies that can bind tightly to a broad range of GI and GII samples that could be used to enhance the sensitivity of ELISA-based diagnostic assays.

The mAbs commonly used for NoV detection were obtained from mice following oral or intraperitoneal inoculation and with standard hybridoma procedures (Hardy et al., 1996; Kitamoto et al., 2002; Parker et al., 2005; Sakamaki et al., 2012). Another approach to obtain mAbs is to use phage display to isolate antibodies of interest from large combinatorial libraries (Sidhu and Fellhouse, 2006; Michnick and Sidhu, 2008; Miersch and Sidhu, 2012). Synthetic antibody libraries consist of a single framework with the molecular diversity created in antigen-binding sites by site-directed mutagenesis (Miersch and Sidhu, 2012). In this study, NoV VLPs were used as targets for biopanning of a monoclonal human single-chain antibody (scFv) library by phage display to identify antibody fragments that bind to GI and GII NoV VLPs. Several antibodies that target NoVs were obtained and characterized. These antibodies may be useful as detection reagents for ELISA-based diagnostic assays.

Materials and methods

Screening of Tomlinson I + J phage libraries

The recombinant scFv antibody libraries were provided by MRC Geneservice. The I and J libraries are both based on a single human framework for VH (V3-23/DP-47 and JH 4b) and Vκ (012/02/DPK9 and Jκ1). Biopanning was initially performed with both I and J libraries. The I library has 18 residue positions CDRH2, CDRH3, CDRL2 and CDRL3 regions randomized with DVT codons (de Wildt et al., 2000). The J library has 18 residue positions in the CDRH2, CDRH3, CDRL2 and CDRL3 regions randomized with NNK codons (de Wildt et al., 2000). Both libraries are present in the pIT2 vector and consist of ∼1.4 × 108 independent clones.

For biopanning, GII.4 Houston virus (HOV—Hu/Houston/TCH186/2002/US, Genbank EU310927) VLPs in phosphate-buffered saline (PBS) at a concentration of 5 μg/ml were added to immunotubes in a volume of 4 ml and incubated overnight at 4°C. The immunotubes were then washed three times with PBS and blocked with MPBS (2% dry milk in PBS) at room temperature for 2 h followed by an additional three washes with PBS. The I and J phage libraries were then added separately to individual tubes at 1011 phage in 4 ml and incubated for 2 h at room temperature. Each immunotube was washed 10 times with PBST (PBS with 0.1% Tween 20). The bound phages were eluted either by the addition of 0.5 ml of 1 mg/ml trypsin in PBS or with 0.5 ml of 5 μg/ml Led (H-type 1-PAA-biotin) carbohydrate in PBS. The elution mixtures were incubated for 10 min and then transferred to 1.5 ml microcentrifuge tubes. For amplification, 0.25 ml of each elution was added to 1.25 ml of Escherichia coli TG1 cells and incubated without shaking at 37°C for 30 min. Ten microliters was taken, serially diluted, and spread on TYZ agar plates containing 100 μg/ml ampicillin and 1% glucose. The remaining mixture (∼1.49 ml) was spread on TYZ agar plates containing 100 μg/ml ampicillin and 1% glucose and, following overnight incubation at 37°C, the colonies were pooled. Fifty microliters of the pooled cells were added to 50 ml of 2YT + 100 μg/ml ampicillin + 1% glucose and grown at 37°C to an OD600 of 0.4. A total of 10 ml of this culture was incubated with 5 × 1010 KM13 helper phages at 37°C for 30 min without shaking. The culture was centrifuged at 3000 g for 10 min and the supernatant was removed. The cell pellet was suspended in 50 ml 2YT + 100 μg/ml ampicillin + 0.1% glucose + 25 μg/ml kanamycin and incubated overnight with shaking at 30°C. The culture was centrifuged at 3300 g for 15 min and the supernatant was collected. A total of 5 ml of PEG6000/2.5 M NaCl was added to 20 ml of supernatant and incubated on ice for 1 h. The mixture was centrifuged at 3300 g for 30 min to pellet the phage particles. The phages were suspended in 1 ml PBS, transferred to a 1.5 ml microcentrifuge tube and centrifuged at 11 600 g for 10 min to remove any remaining cells. The titer of phages in each amplification stock was determined by infecting E. coli TG1 cells. The second and third rounds of biopanning to enrich for antibody-phages that bind to HOV VLPs were performed as described above except that 20 PBST washes of bound phage were performed.

Single-point phage ELISA

High-throughput screening of phage clones was performed by single-point phage ELISA (Deshayes et al., 2002). For these experiments, phages obtained after the third round of biopanning were used to infect E. coli TG1 cells and individual colonies were obtained on TYE agar plates containing 100 μg/ml ampicillin and 1% glucose. Individual colonies were inoculated into 1 ml 2YT medium containing 100 μg/ml ampicillin and 1% glucose in 96-well 2 ml deep well plates and grown with shaking at 37°C for 4 h. A total of 109 KM13 helper phages were then added to each culture well and incubated at 37°C for 30 min followed by centrifugation of the 96-well plate at 3000 g for 15 min. The supernatants were removed and the cell pellets were suspended in 1 ml 2YT + 100 μg/ml ampicillin + 0.1% glucose and grown overnight at 30°C. The 96-well plate was centrifuged at 3000 g for 15 min and the supernatants were transferred to a fresh 96-well plate. For ELISA, the wells of a 96-well microtiter plate were coated with 5 μg/ml HOV or GI.1 NV VLPs in 100 μl total volume and incubated overnight at 4°C. The wells were washed three times with PBS and blocked with MPBS at room temperature for 2 h. The wells were then washed three times with PBS and 100 μl of each phage supernatant was added to each VLP-coated well and incubated for 1 h. The wells were washed 10 times with PBST (0.1% Tween 20 in PBS) and anti-M13 antibody conjugated to horseradish peroxidase (HRP) was added and incubated for 1 h at room temperature. The wells were washed six times with PBST and incubated with 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) ABTS substrate followed by absorbance data collection at OD405 in a microplate reader.

Site-directed mutagenesis

The TAG amber stop codons found in the HJT-R3-A9, HJT-R3-F7, HJL-R3-B4, D11 and F11 scFvs were changed to CAG by site-directed mutagenesis using the QuikChange Mutagenesis method (Stratagene) according to the manufacturer's instructions. The HJL-R3-B4, D11 and F11 mutagenesis reactions utilized the same oligonucleotide because the CDRH2 region is identical in these clones. The DNA sequence of each scFv clone was obtained to confirm the presence of the altered codon and to ensure that extraneous mutations were not present. The primers used for mutagenesis were:

hjtr3-A9-TAG-CAG, 5′-GTGGGTCTCATCTATTCAGACT AAGGGTCGTGGG-3′

hjtr3-F7-TAG-CAG, 5′-GTCCTAGGGGTAAGCAGACAGTGTACGCAGAC-3′

hjlr3-D11-TAG-CAG, 5′-GCTAAGTGGGGTCAGGATAC AGTTTACGC-3′.

scFv purification

The plasmids expressing the HJT-R3-A9, HJT-R3-F7, HJL-R3-B4, D11 and F11 scFvs were used to transform E. coli RB791 cells for protein expression and purification (Amann et al., 1983). The transformed cells were grown overnight at 37°C in 10 ml of 2YT + 100 μg/ml ampicillin + 0.1% glucose. Ten milliliters of the overnight culture was used to inoculate 1 l of 2YT + 100 μg/ml ampicillin + 0.1% glucose and the culture was grown at 37°C to an OD600 of 0.8–1.0. Isopropyl β-d-1-thiogalactopyranoside was then added to a final concentration of 1 mM and the culture was incubated at 30°C for 5 h. The cells were harvested by centrifugation and resuspended in 50 ml of lysis buffer (25 mM sodium phosphate buffer, pH 7.4, 500 mM NaCl, 10 mM imidazole, 60 μg/ml DNAse, 1 tablet EDTA-free protease inhibitor and 25 mM MgCl2). A whole cell protein lysate was obtained from the suspended cells using a French press. The resulting lysate was centrifuged for 15 min at 10 K and the supernatant was filtered using a 0.45 μm Millipore filter. The lysate was bound with 3 ml of Talon metal affinity resin (Clontech, Inc.) incubated for 1 h at room temperature. The resin was then packed into a column and washed with 10 bed volumes of Wash 1 buffer (25 mM sodium phosphate buffer, pH 7.4, 500 mM NaCl, 10 mM imidazole, EDTA-free protease inhibitor) and 10 bed volumes of Wash 2 buffer (25 mM sodium phosphate buffer, pH 7.4, 500 mM NaCl, 20 mM imidazole, EDTA-free protease inhibitor). Bound protein was eluted with 25 mM sodium phosphate buffer, pH 7.4, 500 mM NaCl, 50 mM imidazole, EDTA-free protease inhibitor. Elution fractions were monitored for the presence of scFv by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE). Fractions of high purity were pooled, concentrated, adjusted to 15% glycerol and stored at −80°C. Protein concentrations were determined using dye-binding assays (Bradford, 1976).

Biacore surface plasmon resonance

The binding of purified scFv to VLPs was measured using surface plasmon resonance (SPR) on a Biacore 3000 instrument (GE Healthcare Biosciences) at 25°C. For these experiments, the HOV and Norwalk (NV) VLPs were immobilized to the surface of CM-5 sensor chips. To determine the binding parameters of the scFvs with HOV and NV VLPs, a sandwich method was used in which mAbs NS14 and 3912 were used to capture HOV VLP and NV VLPs, respectively. A pH scouting was first performed to determine the optimum pH for immobilization of the mAbs. The mAb NS14 at 53 μg/ml in acetate buffer (pH 5.0) was immobilized on the CM5 chip surface to an RU of 7900, using amine coupling chemistry following the manufacturer's protocol (activation: EDC and NHS at a 1 : 1 mixture were flowed at 10 μl/min for 7 min. Immobilization: NS14 was flowed at 10 μl/min for 7 min. Blocking: ethanolamine was flowed at 10 μl/min for 7 min). HOV VLPs at a concentration of 100 μg/ml in HBS-EP buffer were injected at 10 μl/min for 7 min to a ΔRU 2200 followed by injection of 100 nM of scFv at 10 μl/min for 5 min. Dissociation was allowed for ∼10 min after which regeneration was performed with two pulses of glycine pH 2.0 injections at 20 μl/min for 30 and 10 s. Each concentration of scFv binding to VLP was subjected to individual cycles with regeneration of the NS14 immobilized CM5 surface after each injection and fresh capture of HOV VLPs for binding measurements. The binding studies were performed at a concentration range of 5–400 nM for all the different scFvs including HJT-R3-A9, HJT-R3-F7, HJL-R3-B4, HJL-R3-D11 and HJL-R3-F11. Similar experiments were performed for NV VLPs, where mAb 3912 was used for capture in the sandwich method. Mass transfer was ruled out by checking binding parameters at higher flow rates of 20 and 30 μl/min. In the control flow channel, different concentrations of scFv were flowed directly over the immobilized mAb. The final subtracted binding curves were fitted to the classical 1 : 1 Langmuir binding model using BIAevaluation program (BIAcore version 3) to obtain the binding parameters.

Immunoblotting

The HOV VLP, NV VLP, CT303 VLP and NV P-domain proteins were boiled in sample buffer and loaded onto an SDS–PAGE gel at a concentration of 0.3 mg/ml with a total amount of 1.5 μg loaded into each well. The proteins were fractionated by electrophoresis and transferred to nitrocellulose using a trans-blot apparatus (Bio-Rad). The membrane was blocked with 1× TBST with 5% milk at 4°C overnight. The membrane was washed three times with 1× TBST with 1% milk for 10 min each wash. The membrane was incubated with biotinylated HJT-R3-A9 antibody at 0.65 μg/ml in 1× TBST with 1% milk for 1 h followed by three washes at 10 min each. Bound HJT-R3-A9 antibody was detected with avidin-HRP at 0.5 μg/ml in 1× TBST + 1% milk for 40 min. The membrane was washed three times with 1× TBST + 1% milk and two times with 1× TBST for 10 min each wash. The bands were visualized by addition of enhanced chemiluminesence (ECL) reagent (Amersham) and exposure of the membrane to X-ray film. The HJT-R3-F7 experiments were done similarly except samples were boiled and loaded at 2 µg/protein per well. HJT-R3-F7 was used at 5 µg/ml and bound antibody was detected with protein A-HRP. The immunoblotting experiments to map the binding site of the HJT-R3-F7 antibody were performed as described above except protein lysates of E. coli cultures expressing the glutathione-S-transferase (GST) fusion proteins shown in Fig. 5 were added to sample buffer, separated by SDS–PAGE, transferred to nitrocellulose and detected with 3 μg/ml HJT-R3-F7 antibody conjugated to HRP, washed and developed with ECL reagent (Parker et al., 2005). The immunoblots of HOV GST fusions were performed as described for the HJT-R3-F7 antibody only using an anti-GST antibody conjugated to HRP (GE Healthcare Life Sciences).

Figure 5.

Figure 5.

Open in a new tab

Schematic figure of the HOV P-domain indicating GST fusion proteins used to localize the binding site for the HJT-R3-F7 antibody on the P-domain. The sections of the P-domain that are present in each GST fusion protein are indicated by the numbers associated with each fusion. Reactivity of the GST fusion protein with HJT-R3-F7 antibody that was conjugated to HRP was determined by immunoblotting and positive or negative reactivity is indicated to the right of each fusion.

VLP capture and detection ELISA experiments with purified scFv antibodies

ELISA experiments to determine the binding profile of scFvs for GI and GII VLPs utilized capture antibodies mAb NS14 for GII VLPs and mAb 3912 for GI VLPs (Kitamoto et al., 2002). The antibodies were used at 10 μg/ml to coat ELISA wells (Immulon HB) in PBS overnight at 4°C. Wells were washed three times with 1× PBS and blocked with 10% milk in PBS at room temperature for 2 h. Wells were washed three times with PBS and bound with 5 μg/ml of either GI or GII VLP for 1 h at room temperature. Wells were washed six times with PBS-Tween and bound with anti-Myc-tag-HRP conjugated antibody for 45 min at room temperature. In the case of the bovine serum albumin (BSA) negative control, anti-BSA-HRP conjugated antibody was used for detection. Wells were washed six times with PBS-Tween and anti-Myc-HRP or anti-BSA-HRP was detected with the HRP substrate 3,3′,5′-tetramethylbenzidine (TMB). Stop solution was added after 15 min and reactions were read at OD450 in a microplate reader.

The VLP dose–response experiments for antibodies HJT-R3-A9 and F7 utilized anti-NV and anti-HOV polyclonal antibodies for capture of VLPs. A 1 : 1000 dilution in PBS of each polyclonal sera was used for coating wells overnight at 4°C. Wells were washed three times with 1× PBS and blocked with 10% milk in PBS at room temperature for 2 h. Wells were washed three times with PBS and bound with various concentrations of NV or HOV VLPs. Wells were washed three times with PBS and a 5 μg/ml solution of HJT-R3-A9 or F7 antibody was added and incubated for 2 h at room temperature. Wells were washed five times with PBS-Tween and a 1 : 5000 dilution of anti-Myc-tag antibody conjugated to HRP was added and incubated for 1 h at room temperature. Wells were washed six times with PBS-Tween, developed with TMB microwell peroxidase substrate, and signal was detected at OD450. Rotavirus antigen was used as a control for these experiments (Crawford et al., 2006). For this purpose, anti-VP6 6E7 mAb ascites (1 : 1000 dilution) was coated in wells overnight at 4°C (Crawford et al., 2006). The wells were washed three times with PBS and blocked with 10% milk in PBS at room temperature for 2 h. Wells were washed three times with PBS and RV antigen was added and incubated for 2 h at room temperature. Wells were washed five times with PBS-Tween and 5 μg/ml of HJT-R3-A9 or F7 antibody or, as a positive control, anti-RV GP511 antibody (1 : 2000 dilution), was added and incubated at room temperature for 2 h. Wells were washed six times with PBS-Tween and bound HJT-R3-A9 and F7 antibody were detected with an anti-Myc-tag antibody conjugated to HRP while the anti-RV GP511 antibody was detected with a goat a-gp IgG-HRP (1 : 4000 dilution) and TMB substrate.

Clinical sample ELISA using scFv antibodies or scFv-displaying phage for detection

Clinical samples were collected and tested under protocols reviewed and approved by the Institutional Review Board at Baylor College of Medicine (Atmar et al., 2008; Koo et al., 2013). The test of the scFv antibodies as detection reagents was performed by first immobilizing an anti-GII.4 antibody in the wells of HB microplates. For this purpose, the NS14 mAb and an anti-HOV polyclonal antibody were used to coat wells overnight at 4°C. Wells were washed three times with 1× PBS and then blocked with 10% dry milk in PBS (MPBS). Wells were washed three times with PBS and a 10% stool suspension was added for capture of virus. Wells were washed 10 times with PBS-Tween, and scFv antibody was added at 50 μg/ml in 2% milk-PBS and incubated for 2 h at room temperature. Wells were washed 10 times with PBS-Tween and anti-Myc antibody conjugated to HRP was added at a 1 : 5000 dilution in 2% MPBS and incubated 45 min at room temperature. Wells were washed eight times with PBS-Tween and the signal was detected by adding HRP substrate, TMB (KPL). The signal was read at OD450.

The test of the HJT-R3-A9-displaying phage as a detection reagent was performed by first coating HB microplate wells with a 1 : 3000 dilution of anti-NV or anti-HOV polyclonal antibody overnight at 4°C. Wells were washed three times with 1× PBS and then blocked with 10% dry milk in PBS (MPBS). Wells were washed three times with PBS and various dilutions of stool suspension were added for 2 h at room temperature for capture of virus. Wells were washed 10 times with PBS-Tween, and 1 × 1011 PFU of HJT-R3-A9-displaying phage was added in 1% milk-PBS and incubated for 2 h at room temperature followed by 10 washes with PBS-Tween. HRP-conjugated M13-antibody (Pharmacia Biotech) was diluted 1 : 5000 in 1% milk PBS and 100 μl was added to each well and incubated at room temperature for 45 min. After washing six times with PBS-T, HRP substrate was added to each well. Development was allowed to proceed for 10–20 min and stopped with 1 M H3PO4 (KPL). Optical density was measured at 450 nm on a Tecan infinite M200pro plate reader.

Results

Identification of scFv antibodies that bind GI and GII VLPs

In order to obtain mAbs that bind to NoV GI and GII VLPs, the Tomlinson I + J svFc phage display libraries (de Wildt et al., 2000) were screened for binding to GII.4 HOV (GII) VLPs as described in the Materials and Methods section. Two methods of elution of bound phages were performed for each library. These included elution with trypsin protease, which cleaves between the scFv and the phage g3p to release the bound phages, and elution by the addition of a HBGA carbohydrate (H-type 1-PAA-biotin), which binds VLPs and is thought to serve as an attachment factor for NoVs (Hutson et al., 2003, 2004). The H-type 1 carbohydrate has been shown to bind NV VLPs and also some genogroup II VLPs (Huang et al., 2005; Tan and Jiang, 2005; Choi et al., 2008). Therefore, the carbohydrate elution procedure may displace and thereby enrich for antibody-phages that bind the HOV VLPs at or near the carbohydrate receptor binding site.

After the three rounds of binding enrichment, the amplified, pooled phages from each round were tested for binding to immobilized HOV VLPs by ELISA. The highest signal was obtained with phages from the Tomlinson J libraries eluted using either trypsin or carbohydrate after three rounds of binding enrichment (data not shown). The signal from the pooled phages from the Tomlinson I library after each round of panning was significantly lower than the J library for both the trypsin and carbohydrate elutions, suggesting that relatively few phages from these enrichments bind HOV VLPs. Therefore, the remainder of the study focused on the antibody-phages enriched from the Tomlinson J library.

In order to identify individual scFv phage clones that bind to HOV VLPs, single-point phage ELISA was performed using 90 randomly chosen J library clones obtained after three rounds of binding enrichment using trypsin as the elution agent (Deshayes et al., 2002). In addition, 90 J library clones obtained after three rounds of enrichment using Led carbohydrate elution were also examined (Deshayes et al., 2002). It was found that a large percentage of the clones obtained using the trypsin elution procedure bound to HOV VLPs as indicated by a high ELISA signal compared with a negative control protein (data not shown). In order to determine if any of the antibody-phage clones also bound to a genogroup I NoV capsid, the set of 90 clones was also tested for binding NV VLPs by ELISA. The results indicated that 40% (36/90) of the clones produced a high ELISA signal (>0.9 OD405) for both HOV and NV VLPs (data not shown). DNA sequence analysis of the scFv region of 10 clones exhibiting high signals for both VLP types revealed that they encode the same nucleotide and amino acid sequence, indicating they represent a single clonal population (Fig. 1A). The HJT-R3-A9 clone was used in subsequent studies as representative of this clonal population. DNA sequencing of phage clones that bound both HOV and NV VLPs with a moderate signal revealed an additional unique clone named HJT-R3-F7 (Fig. 1A).

Figure 1.

Figure 1.

Open in a new tab

(A) Amino acid sequences of scFv clones obtained after three rounds of binding enrichment on HOV VLPs followed by trypsin elution and ELISA screening for clones that also bound to NV VLPs. The CDR positions that are randomized in the Tomlinson J library are highlighted in red. The asterisk indicates the amber TAG stop codon, which is suppressed to glutamine by the supE amber suppressor gene present in the E.coli strain used to propagate phages. The number of clones with the same sequence is shown in parenthesis. For example, 10 clones that were picked and DNA sequenced were identical to clone A-9. (B) Amino acid sequences of scFv clones obtained after three rounds of binding enrichment on HOV VLPs followed by carbohydrate elution ELISA screening.

The single-point ELISA results from the 90 clones from the carbohydrate elution revealed a number of clones that bound to HOV VLPs. The pattern of binding among clones was different than that observed for the trypsin elution. For example, although numerous clones exhibited an ELISA signal significantly above background levels, no phage clones were identified that displayed extremely high ELISA signals (>1.0 OD405) for binding HOV VLPs. In addition, the signals observed for binding to NV VLPs were, in general, lower, suggesting modest cross-reactivity of the scFvs between HOV and NV VLPs. DNA sequence analysis of the scFv regions of 20 phages with the highest ELISA signals for binding HOV VLPs revealed a number of different sequences that could be placed into seven families (Fig. 1B). Interestingly, several of the families possess the same heavy chain sequence; five families encompassing 15 of the 20 sequenced clones utilize the same heavy chain sequence (Fig. 1B). In addition, the seven families are represented by six different light chain sequences due to the repeat of two light chain sequences in separate families. The use of a limited number of heavy chain and light chain CDR sequences that are combined in different ways to make up the seven families suggests the antibodies bind to a similar region on the HOV VLP. Binding of the clones to a single site would be consistent with the biopanning elution procedure with carbohydrate which is designed to displace phages from the carbohydrate-binding site. The use of limited heavy chains with diverse light chains also suggests the majority of the binding affinity for the target arises from the heavy chain.

Characterization of scFv antibodies that bind NoV VLPs

The scFv phages eluted with trypsin that bound to both HOV and NV VLPs are represented by HJT-R3-A9 and HJT-R3-F7 antibodies as described above. A broad spectrum scFv would be a useful diagnostic tool for NoV infections and therefore these scFv clones were characterized further. As indicated in Fig. 1A, the CDR H2 sequences of both HJT-R3-A9 and F7 contain a TAG stop codon which is suppressed to glutamine in the E. coli TG1 strain used for phage propagation. To facilitate protein expression and purification, the TAG codon was converted to the CAG glutamine codon by site-directed mutagenesis. The pIT2 plasmid encoding HJT-R3-A9 and HJT-R3-F7 were transferred to E. coli RB791 cells that do not contain a nonsense suppressor in order to express soluble scFv antibody protein. The HJT-R3-A9 and F7 proteins were purified by affinity chromatography and tested for binding to HOV and NV VLPs by ELISA (Fig. 2). For each antibody, a titration of concentrations of both HOV and NV VLPs were tested. For these experiments, the VLPs at the various concentrations were immobilized into microtiter wells that were first coated with anti-NV or anti-HOV polyclonal antibodies. A clear dose–response of binding was observed with increasing VLP concentrations for HJT-R3-A9 and F7 with both NV and HOV VLPs. In addition, rotavirus antigen was immobilized and used as a negative control antigen to rule out polyreactivity of the antibodies (Fig. 2C and F). The results indicate that the HJT-R3-A9 and HJT-R3-F7 antibodies do not bind the rotavirus antigen despite the high signal obtained when an anti-rotavirus antibody was used to detect the rotavirus antigen. Therefore, the signal observed with the NoV VLPs is not due to non-specific binding (Fig. 2).

Figure 2.

Figure 2.

Open in a new tab

ELISA experiments to examine the dose–response for binding of HJT-R3-A9 (A–C) and HJT-R3-F7 (D–F) to NV, HOV and rotavirus VLPs. (A, D) Binding to immobilized NV VLPs. Microtiter wells were coated with anti-NV antibody and NV VLPs were added and captured at the concentrations shown on the X-axis. scFv antibody was added at 175 nM, washed and detected with anti-Myc-tag antibody. (B, E) Binding to HOV VLPs that were added at the concentrations indicated and captured with anti-HOV antibody. (C, F) In column at left, rotavirus antigen was captured with anti-RVP6 6E7 mAb and scFv antibody was added at 175 nM, washed and detected with anti-Myc-tag antibody. In column at right, rotavirus antigen was captured and detected with anti-rotavirus antibody GP511.

The reactivity of the HJT-R3-A9 and F7 antibodies was then tested for additional GI.1 and GII.4 VLPs. For these experiments, various NoV GI VLPs including GI.1, GI.6 and GI.7, were immobilized by binding to microtiter wells first coated with the 3912 mAb that recognizes GI VLPs (Hale et al., 2000). The immobilized 3912 antibody then captured the GI VLPs and unbound VLPs were washed away. The purified HJT-R3-A9 and F7 scFvs were added and bound antibody was detected with an anti-Myc tag antibody which recognizes the Myc-tag present on the scFvs (Fig. 3A). The results indicate that the HJT-R3-A9 antibody bound the GI.1 as well as GI.6 and GI.7 VLPs. The HJT-R3-F7 bound to captured GI.1 and also, but to a lesser extent, the GI.6 and GI.7 VLPs (Fig. 3A). Therefore, the HJT-R3-A9 and F7 antibodies display a broad specificity towards GI genogroup VLPs.

Figure 3.

Figure 3.

Open in a new tab

Test of scFvs as detection antibodies for NoV VLPs. (A) Binding of scFv antibodies to GI.1 VLPs was tested by coating the anti-GI mAb 3912 into ELISA wells followed by capture of GI VLPs. The various scFv antibodies were added to the captured GI VLPs, washed and bound scFv was detected with an anti-Myc-tag antibody. The scFv antibodies tested in the experiment are indicated in the inset. (B) Binding of scFv antibodies to GII VLPs was tested by coating the anti-GII mAb NS14 into ELISA wells followed by capture of GII VLPs. The various scFv antibodies were added to the captured GII VLPs, washed and bound scFv was detected with an anti-Myc-tag antibody as described for part (A) above.

The HJT-R3-A9 and F7 antibodies were also tested for binding to VLPs from genogroup II. For these experiments, the broad specificity anti-GII antibody NS14 was immobilized into microtiter wells and subsequently used to capture VLPs from a number of genotypes including GII.2, GII.3, GII.4, GII.6, GII.7, GII.12 and GII.17 (Fig. 3B). The HJT-R3-A9 and F7 scFvs were then added and bound protein was detected with the anti-Myc-tag antibody. The results indicated a broad binding specificity for both the A9 and F7 antibodies with ELISA signals above the BSA negative control for all of the GII VLPs tested with the highest reactivity observed for GII.4, which is consistent with the fact that the GII.4 HOV VLP was used as the target for biopanning with the phage library (Fig. 3B). Taken together, the results indicate that the HJT-R3-A9 and F7 antibodies bind to a broad range of VLPs from both the GI and GII genogroups.

Several clones were chosen for characterization from the biopanning experiment that utilized carbohydrate elution including HJL-R3-B4, HJL-R3-D11 and HJL-R3-F11. These clones also contained TAG stop codons, which were converted by site-directed mutagenesis to CAG glutamine-encoding codons. The antibodies were expressed and purified from E. coli and tested for binding to genogroup GI and GII VLPs as described for the HJT-R3-A9 and F7 clones. The scFv clones displayed a range of activity towards captured GI VLPs in the ELISA experiments (Fig. 3A). The HJL-R3-B4 clone exhibited high ELISA signals for GI.1, GI.6 and GI.7 VLPs while the D11 clone exhibited lower binding that was nevertheless well above the levels of the BSA negative control. In contrast, the HJL-R3-F11 did not bind GI VLPs (Fig. 3A).

The HJL-R3-B4, D11 and F11 antibodies were also tested for binding to captured GII VLPs as described for the HJT-R3-A9 and F7 clones (Fig. 3B). It was found that all of the clones gave strong ELISA signals for binding GII.4 VLP, which is consistent with the fact that they were isolated based by biopanning on immobilized HOV VLPs. The HJL-R3-D11 antibody exhibited strong ELISA signals for all of the GII VLPs tested. The HJL-R3-B4 clone also bound to other GII VLPs, albeit with a lower ELISA signal. The results indicate that the HJL-R3-D11 and, to a lesser extent, the B4 clone have a broad specificity that includes GI and GII genotypes. Therefore, scFv antibodies identified from both the trypsin (HJT-R3-A9, F7) and carbohydrate (HJL-R3-D11, B4) exhibit broad binding specificity that includes GI and GII VLPs.

Mapping the binding site of the HJT-R3-A9 and F7 antibodies on VLPs

Due to the broad binding specificity of the HJT-R3-A9 and F7 antibodies, it was of interest to determine where these reagents bound to VLPs. In order to gain information on the location of the HJT-R3-A9 scFv-binding epitope, immunoblotting was performed. For these experiments, HOV and NV VLPs were boiled in loading buffer and resolved by SDS–PAGE. In addition, boiled and unboiled samples of the CT303 VLP (Bertolotti-Ciarlet et al., 2002), which consists only of the NV S-domain, were separated by SDS–PAGE. Finally, a fusion protein of GST and the NV P-domain was run on SDS–PAGE with and without boiling (Parker et al., 2005). The results in Fig. 4A demonstrate that the HJT-R3-A9 antibody binds to the boiled HOV and NV VLPs. In addition, the antibody binds both the boiled and unboiled forms of the CT303 S-domain VLP but does not bind to either boiled or unboiled GST-P-domain fusion protein. Thus, the HJT-R3-A9 antibody appears to bind to a linear epitope in the S-domain. The S-domain is the most highly conserved region of VP1 between the different genogroups, which may explain the broad spectrum binding exhibited by the antibody (Chen et al., 2004). Finally, the HJT-R3-A9 antibody did not bind the negative control protein, TEM-1 β-lactamase, indicating it does not exhibit polyreactivity.

Figure 4.

Figure 4.

Open in a new tab

(A) Immunoblot to test binding of HJT-R3-A9 antibody to HOV, NV and CT303 VLPs and GST-GI-P-domain protein. Molecular weight marker positions are indicated by tic marks and the indicated molecular weights. The label P is to indicate P-domain. GST stands for glutathione-S-transferase. CT303 is a VLP consisting only of the S-domain of the major capsid protein. TEM-1 β-lactamase is a negative control protein. (B) Immunoblot to test binding of HJT-R3-F7 antibody to HOV and NV VLPs and P-domains. Molecular weight positions are labeled at left. GST stands for glutathione-S-transferase. CT303 is a VLP consisting only of the S-domain of the major capsid protein. TEM-1 β-lactamase is a negative control protein.

Immunoblotting was also performed to map the epitope for the HJT-R3-F7 antibody. For this purpose, NV and HOV VLPs as well and NV and HOV P-domain protein samples were boiled and separated by SDS–PAGE and immunoblotting was performed using the HJT-R3-F7 antibody. As seen in Fig. 4B, the F7 antibody bound to the NV and HOV VLPs as well as the P-domains but not to the negative control proteins indicating the binding site is located within the P-domain. Therefore, the results indicate that the HJT-R3-A9 and F7 antibodies utilize different epitopes to bind VLPs.

The binding site of the HJT-R3-F7 antibody was investigated further using a set of GST fusions to deletions of the P-domain from HOV NoV (Supplementary Fig. S2). The HJT-R3-F7 antibody was conjugated to HRP and binding to the GST fusions was evaluated by immunoblotting and the results are summarized in Fig. 5. Based on the fusion results, the binding site could be localized to the amino acid residues 417–488 region of the HOV P-domain. The positive signals from the 473–540 and 417–488 fusions suggest that the binding site is located between residues 473–488; however, a 473–488 peptide has not been directly tested for binding. The 417–488 region maps to the P1 region of the P-domain. The P1 subdomain is less accessible than P2 although it has been shown based on X-ray crystallography that the mouse 5B18 mAb binds to a buried portion of the GII.10 P1 subdomain and yet is able to capture VLPs suggesting considerable conformational flexibility of the capsid protein to facilitate antibody binding (Hansman et al., 2012).

SPR analysis of scFv binding to VLPs

The binding of the various single-chain antibodies to VLPs was examined by SPR experiments to obtain quantitative estimates of binding affinity (Supplementary Fig. S1, Table I). This was accomplished using a capture assay whereby the mAbs 3912 and NS14 were covalently attached to a Biacore CM5 chip for capture of NV and HOV VLPs, respectively. Binding of the HJT-R3-A9 and F7 antibodies was measured for both NV and HOV VLPs as was binding of the HJL-R3-B4, D11 and F11 clones (Table I). It was found that the HJT-R3-F7 antibody bound to NV and HOV VLPs with similar affinities of 25 and 45 nM Kd, respectively. The HJT-R3-A9 antibody, however, did not bind to either captured NV or HOV VLPs. This is somewhat surprising in that the HJT-R3-A9 antibody does bind to captured NV and HOV VLPs in the ELISA format. The lack of binding could be due to the binding epitope for HJT-R3A9 being located in the S-domain, which is buried below the P-domain in the structure of VLPs. It is possible that the antibody capture of VLPs on the surface of ELISA wells results in sufficient unfolding of VLPs to reveal the S-domain to allow HJT-R3-A9 binding. On the Biacore chip, the capture antibodies are attached to glycan strands that project away from the surface of the chip and this may preserve the structure of the VLP upon capture, thereby keeping the S-domain epitope masked from binding the HJT-R3-A9 antibody. It is of interest that when the HJT-R3-A9 antibody is directly coated into ELISA wells, it does not capture soluble NV or HOV VLPs, presumably because the S-domain epitope is masked in the soluble VLPs (data not shown).

Table I.

Binding constants for the interaction between VLPs and scFv antibodies

scFv VLP ka (1/M s) kd (1/s) KA (1/M) KD (M)
HJT-R3-A9 GII-4 HOV No binding
HJT-R3-F7 GII-4 HOV 1.1 × 105 5.1 × 10−3 2.2 × 107 4.5 × 10−8
HJL-R3-B4 GII-4 HOV 3.3 × 104 1.6 × 10−3 2.0 × 107 5.0 × 10−8
HJL-R3-D11 GII-4 HOV 1.9 × 105 1.2 × 10−3 1.6 × 108 6.1 × 10−9
HJL-R3-F11 GII-4 HOV 9.2 × 104 1.5 × 10−3 6.2 × 107 1.6 × 10−8
HJT-R3-A9 GI-1 NV No binding
HJT-R3-F7 GI-1 NV 2.5 × 105 6.2 × 10−3 4.0 × 107 2.5 × 10−8
HJL-R3-B4 GI-1 NV No binding
HJL-R3-D11 GI-1 NV No binding
HJL-R3-F11 GI-1 NV No binding
mAb NS14 GII-4 HOV 5.5 × 105 2.5 × 10−4 2.2 × 109 4.6 × 10−10
mAb 3912 GI-1 NV 7.0 × 105 5.0 × 10−4 1.4 × 109 7.1 × 10−10
Open in a new tab

The SPR experiments revealed that the HJL-R3-B4, D11 and F11 antibodies bound to captured HOV VLPs with binding constants (Kd) of 50, 6.1 and 16 nM, respectively (Table I). Therefore, consistent with the ELISA results, the antibody clones isolated by carbohydrate elution bind tightly to the HOV VLPs. However, the HJL-R3-B4, D11 and F11 antibodies did not exhibit detectable binding to NV VLPs. This result is in contrast to the ELISA experiments in Fig. 3B where the HJL-R3-B4, D11 and F11 antibodies gave a binding signal above background for antibody-captured NV VLPs. It is unclear why this is the case, however, it is possible that the binding epitope for these antibodies is less surface exposed on the NV versus the HOV VLP and therefore partial unfolding may be required for the antibodies to bind NV VLPs. As stated above, the capture antibodies are not directly on the SPR chip surface but rather are connected with glycan strands away from the surface, so the VLP with be in a different environment for the SPR versus the ELISA experiments. Finally, the SPR experiments indicate that the scFv antibodies bind to NV and HOV VLPs with ∼10-fold weaker affinity than the 3912 and NS14 mAbs (Table I). This likely reflects the fact that the scFv antibodies are synthetic and have not undergone affinity maturation.

Detection of GII.4 NoV from a clinical sample using purified scFv antibodies

The above experiments indicate that all of the antibodies tested bind to HOV GII.4 VLPs in the ELISA format and all but HJT-R3-A9 bind HOV VLPs with nanomolar affinity by SPR measurements. It was of interest, therefore, to determine if these antibodies could be used to detect GII.4 virus from a clinical sample. For this purpose, an ELISA experiment was performed whereby the anti-GII NS14 antibody as well as anti-HOV polyclonal antibodies were used as capture reagents and coated into microtiter wells. Suspensions from a known GII.4 positive and a negative stool sample were added as well as BSA as a negative control. The HJT-R3-A9 and F7 as well as the HJL-R3-B4, D11 and F11 were tested as detection antibodies for captured virus. The results indicate that all but the HJL-R3-B4 antibody gave strong ELISA signals for the positive stool sample and displayed little reactivity with the negative stool samples when either the NS14 or anti-HOV polyclonal antibodies were used for capture (Fig. 6). The positive result with the HJT-R3-A9 antibody is consistent with the results with captured VLPs in Fig. 3 and indicates this scFv can be used as a detection reagent in an ELISA format. The low reactivity of the HJL-R3-B4 antibody may be due to its relatively weak binding affinity for HOV (50 nM Kd) compared with the other antibodies as indicated in the SPR experiments (Table I). In total, the results indicate that this approach could be useful for discovering antibodies to be used as diagnostic tools for detection of NoV infections.

Figure 6.

Figure 6.

Open in a new tab

Test of scFv antibodies for detection of virus in clinical samples. The various scFv antibodies were tested for their ability to detect GII.4 virus from stool samples. A stool sample shown to be positive as well as a sample shown to be negative by RT-PCR was used. The virus was captured from samples using either the mAb NS14 or with polyclonal anti-HOV antibody that had been coated in ELISA wells. The various scFv antibodies were then added for detection of captured virus as indicated by the X-axis legend. The capture antibody and stool sample utilized are indicated by the key in the inset.

Detection of NoV from clinical samples with scFv phage

The above experiments demonstrate that purified scFv can be used to detect NoV VLPs and virus from a clinical sample. However, antibody purification requires extra time and expense in the development of a diagnostic reagent. In contrast, M13 bacteriophage are easy to produce and purify from E. coli and we have previously shown that phages displaying a peptide affinity reagent that recognizes NoV can be used to detect virus in clinical samples (Rogers et al., 2013). Therefore, phages displaying the HJT-R3-A9 scFv antibody were tested as detection reagents for GI.1 and GII.4 NoV from clinical samples. For this purpose, ELISA wells were coated with anti-GI.1 or anti-GII.4 polyclonal antibody and suspensions from several dilutions of known GI.1 and GII.4 as well as negative stool samples were added to the wells (Fig. 7A and B). The HJT-R3-A9 displaying phages were added to the captured NoV and bound phages were detected with an anti-M13 phage antibody (Rogers et al., 2013). The results indicate that phages displaying the HJT-R3-A9 antibody gave a strong ELISA signal with the GI.1 and GII.4 but not the negative stool samples (Fig. 7). For both GI.1 and GII.4 positive stool samples, the 1% dilution provided the optimal sensitivity. Based on these results, it may be possible to use the phages displaying the scFv antibodies directly as NoV detection reagents for diagnosing infections.

Figure 7.

Figure 7.

Open in a new tab

Test of HJT-R3-A9 scFv antibody-displaying phages for detection of GI.1 and GII.4 NoV in clinical samples. NoV-positive and a NoV-negative stool samples were diluted to the indicated concentrations and then captured with a GI.1 (A) or GII.4 (B) polyclonal antibody. Binding was tested for the M13 KM13 phage, which does not display a protein, with 20% NoV-positive stool samples as well as with HJT-R3-A9 scFv-displaying phage with the indicated dilutions. The key indicates NoV-positive and NoV-negative stool samples.

Discussion

Several methods have been used to detect NoV in diagnostic assays including EM, RT-PCR and ELISA (Atmar and Estes, 2001). The ease of use and lack of need for specialized equipment make ELISA-based methods powerful diagnostic tools. The extensive diversity of the NoV capsid protein between and within genogroups, however, makes the development of broadly cross-reacting antibodies a challenge and currently available ELISA-based diagnostics exhibit excellent specificity but modest sensitivity because of limited cross-reactivity (Richards et al., 2003; de Bruin et al., 2006). Therefore, there is a need to develop ELISA reagents with broad reactivity and high sensitivity.

In this study, a phage display library displaying human synthetic single-chain antibodies was used to identify reagents that bind to NoV genogroup I and II VLPs (de Wildt et al., 2000). The strategy employed was to screen the Tomlinson I + J phage libraries for antibodies that bind to GII.4 HOV VLPs and then to test the positive phage clones for binding to Norwalk GI.1 VLPs. This approach yielded multiple candidate phages but DNA sequencing indicated that these phages all encoded a single scFv sequence. A purified, soluble form of this scFv, named HJT-R3-A9, was found to efficiently detect antibody-captured GI and GII VLPs by ELISA. In addition, examination of phage clones that gave a moderate ELISA signal for binding both HOV and NV VLPs yielded the HJT-R3-F7 antibody, which can also efficiently detect antibody-captured GI and GII VLPs in ELISA experiments. Mapping of the binding sites for the HJT-R3-A9 and F7 clones by immunoblotting indicated that the HJT-R3-A9 scFv binds a linear epitope in the S-domain while the F7 antibody binds within the P1 subdomain of the P-domain. The discovery of an antibody that binds the S-domain by screening for scFvs that bind HOV and NV VLPs is not surprising in that the S-domain is the most highly conserved region in the capsid among different genogroups and broadly reactive mAbs targeting this region have been described previously (Yoda et al., 2003; Hansman et al., 2006). The discovery of the HJT-R3-F7 antibody indicates it is possible to utilize phage display libraries to identify clones that bind the P-domain of different genogroups, which could have significant implications for development of diagnostic reagents.

A number of scFv antibodies were also discovered from the phage display libraries by eluting phages bound to HOV VLPs with H-type 1-carbohydrate (Fig. 2). The rationale for this approach was to specifically displace phages bound to VLPs by competition with carbohydrate so as to bias the panning experiments to enrich for scFvs that bind VLPs at or near the carbohydrate-binding site. It is of interest that, although several scFvs were identified, a common heavy chain was used in 15 of the 20 clones sequenced. In addition, two light chains were found in more than one scFv (Fig. 2). Therefore, the families of scFv sequences obtained are related, which suggests that the antibodies may bind to a similar site on the VLP. This site could correspond to a carbohydrate-binding site. Studies are in progress to localize the binding sites for these antibodies on HOV VLPs. One of the goals of obtaining antibodies that bind VLPs at the carbohydrate-binding site is the potential for using these reagents as prophylactics to block cell binding by NoVs. An scFv that recognizes the carbohydrate-binding site has previously been constructed from a mAb that bound recombinant NV VLPs (Ettayebi and Hardy, 2008). It was shown that this scFv could block binding of NV VLPs to CHO cells but the binding spectrum to various genogroups was not discussed (Ettayebi and Hardy, 2008). In the present study, several scFv antibodies were enriched from a phage library by carbohydrate elution after binding GII.4 HOV VLPs. The GII.4 strains are currently the predominant NoV strains circulating worldwide and therefore these scFvs warrant further study (Fankhauser et al., 2002; Bull et al., 2006).

Interestingly, the antibodies identified by carbohydrate elution proved to bind both GI and GII genogroup VLPs in ELISA experiments where the VLP was first captured by an unrelated antibody. In SPR experiments, however, the carbohydrate eluted HJL-R3-B4, D11 and F11 antibodies did not bind GI.1 NV VLPs but did bind GII.4 HOV VLPs. These results indicate that scFv binding is sensitive to the environment in which the VLP is immobilized. We hypothesize that conformational changes occur upon immobilization of the VLP in the capture assays and the extent of the change is sensitive to the environment of the capture event. For example, the HJT-R3-A9 antibody binding site was clearly mapped to the S-domain and yet this antibody can be used to detect whole VLPs that are captured by NS14 or 3912 mAbs in ELISA wells and yet when the same capture antibodies are used for SPR on a platform of glycan strands rather than an ELISA well, the conformational change apparently does not occur because the HJT-R3-A9 antibody does not bind VLPs in these experiments.

It was also shown in this study that the scFv antibodies identified by phage display can be used to detect GII.4 NoV from clinical samples. In ELISA experiments using the NS14 mAb or anti-HOV polyclonal antibody as the capture reagent, the scFv antibodies efficiently detect virus from a sample shown by RT-PCR to contain GII.4 virus but do not react with negative control stool samples. Therefore, the scFv antibodies have potential as broadly reactive detection reagents for NoV diagnostic assays.

We also demonstrated that M13 bacteriophages displaying the HJT-R3-A9 scFv antibody can be used directly to detect both GI.1 and GII.4 genogroup NoV from stool samples. As we have discussed previously, the use of phage directly as a detection reagent has several advantages including the ease and low cost of phage propagation and purification (Rogers et al., 2013). A total of 1 × 1013 PFU of phage can be obtained from 20 ml of infected E. coli culture and purified from supernatants using a PEG precipitation (Rogers et al., 2013). An additional advantage of the scFv phage display system is that the clones obtained can be further developed by directed evolution strategies to increase binding affinity and modify specificity (Rajpal et al., 2005; Laffly et al., 2008; Michnick and Sidhu, 2008). A potential concern is the possible interaction of phages with E. coli present in stool samples. However, the experiments performed here on stool samples with phage displaying antibody as well as previous studies using phage displaying a NoV-binding peptide suggest that the background binding to E. coli is low (Rogers et al., 2013). Therefore, the scFv phage clones described here represent first-generation scFvs whose properties can be improved by random or site-directed mutagenesis followed by phage display selections.

Supplementary data

Supplementary data are available at PEDS online.

Funding

This research was funded by Public Health Service Grants [NIH P01 AI057788], [NIH P30DK56338], Agriculture and Food Research Initiative Competitive Grant [2011-68003-30395] from the USDA National Institute of Food and Agriculture and the John S. Dunn Research Foundation.

Supplementary Material

Supplementary Data
supp_27_10_339__index.html (1.1KB, html)

References

  1. Amann E., Brosius J., Ptashne M. Gene. 1983;25:167–178. doi: 10.1016/0378-1119(83)90222-6. [DOI] [PubMed] [Google Scholar]
  2. Atmar R.L., Estes M.K. Clin. Microbiol. Rev. 2001;14:15–37. doi: 10.1128/CMR.14.1.15-37.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Atmar R.L., Estes M.K. Gastroenterol. Clin. N. Am. 2006;35:275–290. doi: 10.1016/j.gtc.2006.03.001. [DOI] [PubMed] [Google Scholar]
  4. Atmar R.L., Opekun A.R., Gilger M.A., Estes M.K., Crawford S.E., Neill F.H., Graham D.Y. Emerg. Infect. Dis. 2008;14:1553–1557. doi: 10.3201/eid1410.080117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Atmar R.L., Opekun A.R., Gilger M.A., et al. J. Infect. Dis. 2014;209:1016–1022. doi: 10.1093/infdis/jit620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bertolotti-Ciarlet A., White L.J., Chen R., Prasad B.V.V., Estes M.K. J. Virol. 2002;76:4044–4055. doi: 10.1128/JVI.76.8.4044-4055.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Blanton L.H., Adams S.M., Beard R.S., Wei G., Bulens S.N., Widdowson M.A., Glass R.I., Monroe S.S. J. Infect. Dis. 2006;193:413–421. doi: 10.1086/499315. [DOI] [PubMed] [Google Scholar]
  8. Bradford M.M. Anal. Biochem. 1976;72:248–254. doi: 10.1016/0003-2697(76)90527-3. [DOI] [PubMed] [Google Scholar]
  9. Bu W., Mamedova A., Tan M., Xia M., Jiang X., Hedge R.S. J. Virol. 2008;82:5340–5347. doi: 10.1128/JVI.00135-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bull R.A., Tu E.T., McIver C.J., Rawlinson W.D., White P.A. J. Clin. Microbiol. 2006;44:327–333. doi: 10.1128/JCM.44.2.327-333.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cao S., Lou Z., Tan M., et al. J. Virol. 2007;81:5949–5957. doi: 10.1128/JVI.00219-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Chen R., Neill J.D., Noel J.S., Hutson A.M., Glass R.I., Estes M.K., Prasad B.V.V. J. Virol. 2004;78:6469–6479. doi: 10.1128/JVI.78.12.6469-6479.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Choi J.M., Hutson A.M., Estes M.K., Prasad B.V.V. Proc. Natl Acad. Sci. USA. 2008;105:9175–9180. doi: 10.1073/pnas.0803275105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Crawford S.E., Patel D.G., Cheng E., Berkova Z., Hyser J.M., Ciarlet M., Finegold M.J., Conner M.E., Estes M.K. J. Virol. 2006;80:4820–4832. doi: 10.1128/JVI.80.10.4820-4832.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. de Bruin E., Duizer E., Vennema H., Koopmans M.P.G. J. Virol. Methods. 2006;137:259–264. doi: 10.1016/j.jviromet.2006.06.024. [DOI] [PubMed] [Google Scholar]
  16. de Wildt R.M.T., Mundy C.R., Gorick B.D., Tomlinson I.M. Nat. Biotechnol. 2000;18:989–994. doi: 10.1038/79494. [DOI] [PubMed] [Google Scholar]
  17. Debbink K., Lindesmith L.C., Donaldson E.F., Baric R.S. PLoS Pathog. 2012;8:e1002921. doi: 10.1371/journal.ppat.1002921. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Deshayes K., Schaffer M.L., Skelton N.J., Nakamura G.R., Kadkhodayan S., Sidhu S.S. Chem. Biol. 2002;9:495–505. doi: 10.1016/s1074-5521(02)00129-1. [DOI] [PubMed] [Google Scholar]
  19. Donaldson E.F., Lindesmith L.C., Lobue A.D., Baric R.S. Nat. Rev. Microbiol. 2010;8:231–241. doi: 10.1038/nrmicro2296. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Ettayebi K., Hardy M.E. Virol. J. 2008;5:21. doi: 10.1186/1743-422X-5-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Fankhauser R.L., Monroe S.S., Noel J.S., Humphrey C.D., Bresee J.S., Parashar U.D., Ando T., Glass R.I. J. Infect. Dis. 2002;193:413–421. doi: 10.1086/341085. [DOI] [PubMed] [Google Scholar]
  22. Glass R.I., Parashar U.D., Estes M.K. N. Engl. J. Med. 2009;361:1776–1785. doi: 10.1056/NEJMra0804575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gray J.J., Kohli E., Ruggeri F.M., et al. Clin. Vaccine Immunol. 2007;14:1349–1355. doi: 10.1128/CVI.00214-07. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Green K.Y., Lew J.F., Jiang X., Kapikian A.Z., Estes M.K. J. Clin. Microbiol. 1993;31:2185–2191. doi: 10.1128/jcm.31.8.2185-2191.1993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Hale A.D., Tanaka T.N., Kitamoto N., Ciarlet M., Jiang X., Takeda N., Brown D.W.G., Estes M.K. J. Clin. Microbiol. 2000;38:1656–1660. doi: 10.1128/jcm.38.4.1656-1660.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Hansman G.S., Natori K., Shirato-Horikoshi H., et al. J. Gen. Virol. 2006;87:909–919. doi: 10.1099/vir.0.81532-0. [DOI] [PubMed] [Google Scholar]
  27. Hansman G.S., Biertumpfel C., Georgiev I., McLellan J.S., Chen L., Zhou T., Katayama K., Kwong P.D. J. Virol. 2011;85:6687–6701. doi: 10.1128/JVI.00246-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hansman G.S., Taylor D.W., McLellan J.S., et al. J. Virol. 2012;86:3635–3646. doi: 10.1128/JVI.06868-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Hardy M.E., Tanaka T.N., Kitamoto N., White L.J., Ball J.M., Jiang X., Estes M.K. Virology. 1996;217:252–261. doi: 10.1006/viro.1996.0112. [DOI] [PubMed] [Google Scholar]
  30. Huang P.W., Farkas T., Zhong W., Tan M., Thornton S., Morrow A.L., Jiang X. J. Virol. 2005;79:6714–6722. doi: 10.1128/JVI.79.11.6714-6722.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Hutson A.M., Atmar R.L., Marcus D.M., Estes M.K. J. Virol. 2003;77:405–415. doi: 10.1128/JVI.77.1.405-415.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Hutson A.M., Atmar R.L., Estes M.K. Trends Microbiol. 2004;12:279–287. doi: 10.1016/j.tim.2004.04.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Jiang X., Wang M., Graham D.Y., Estes M.K. J. Virol. 1992;66:6527–6532. doi: 10.1128/jvi.66.11.6527-6532.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Jiang X., Wang J., Estes M.K. Arch. Virol. 1995;140:363–374. doi: 10.1007/BF01309870. [DOI] [PubMed] [Google Scholar]
  35. Kageyama T., Shinohara M., Uchida K., et al. J. Clin. Microbiol. 2004;42:2988–2995. doi: 10.1128/JCM.42.7.2988-2995.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Kitamoto N., Tanaka T., Natori K., Takeda N., Nakata S., Jiang X., Estes M.K. J. Clin. Microbiol. 2002;40:2459–2465. doi: 10.1128/JCM.40.7.2459-2465.2002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Koo H.L., Neill F.H., Estes M.K., Munoz F.M., Cameron A., DuPont H.L., Atmar R.L. J. Pediatr. Infect. Dis. Soc. 2013;2:57–60. doi: 10.1093/jpids/pis070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Kroneman A., Vennema H., Deforche K., Avoort H.v.d., Penaranda S., Oberste M.S., Vinje J., Koopmans M. J. Clin. Virol. 2011;51:121–125. doi: 10.1016/j.jcv.2011.03.006. [DOI] [PubMed] [Google Scholar]
  39. Kroneman A., Vega E., Vennema H., et al. Arch. Virol. 2013 doi: 10.1007/s00705-013-1708-5. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Kubota T., Kumagai A., Ito H., et al. J. Virol. 2012;86:11138–11150. doi: 10.1128/JVI.00278-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Laffly E., Pelat T., Cedrone F., Blesa S., Bedouelle H., Thullier P. J. Mol. Biol. 2008;378:1094–1103. doi: 10.1016/j.jmb.2008.03.045. [DOI] [PubMed] [Google Scholar]
  42. Lindesmith L.C., Beltramello M., Donaldson E.F., Corti D., Swanstrom J., Debbink K., Lanzavecchia A., Baric R.S. PLoS Pathog. 2012;8:e1002705. doi: 10.1371/journal.ppat.1002705. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Michnick S.W., Sidhu S.S. Nat. Chem. Biol. 2008;4:326–329. doi: 10.1038/nchembio0608-326. [DOI] [PubMed] [Google Scholar]
  44. Miersch S., Sidhu S.S. Methods. 2012;57:486–498. doi: 10.1016/j.ymeth.2012.06.012. [DOI] [PubMed] [Google Scholar]
  45. Parker T.D., Kitamoto N., Tanaka T., Hutson A.M., Estes M.K. J. Virol. 2005;79:7402–7409. doi: 10.1128/JVI.79.12.7402-7409.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Prasad B.V.V., Hardy M.E., Dokland T., Bella J., Rossmann M.G., Estes M.K. Science. 1999;286:287–290. doi: 10.1126/science.286.5438.287. [DOI] [PubMed] [Google Scholar]
  47. Rajpal A., Beyaz N., Haber L., Cappuccilli G., Yee H., Bhatt R.R., Takeuchi T., Lerner R.A., Crea R. Proc. Natl Acad. Sci. USA. 2005;102:8466–8471. doi: 10.1073/pnas.0503543102. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Richards A.F., Lopman B., Gunn A., et al. J. Clin. Virol. 2003;26:109–115. doi: 10.1016/s1386-6532(02)00267-6. [DOI] [PubMed] [Google Scholar]
  49. Rogers J.D., Ajami N.J., Fryszczyn B.G., Estes M.K., Atmar R.L., Palzkill T. J. Clin. Microbiol. 2013;51:1803–1808. doi: 10.1128/JCM.00295-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Sakamaki N., Ohiro Y., Ito M., Makinodan M., Ohta T., Suzuki W., Takayasu S., Tsuge H. Clin. Vaccine Immunol. 2012;19:1949–1954. doi: 10.1128/CVI.00427-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Scallan E., Hoekstra R.M., Angulo F.J., Tauxe R.V., Widdowson M.-A., Roy S.L., Jones J.L., Griffin P.M. Emerg. Infect. Dis. 2011;17:7–15. doi: 10.3201/eid1701.P11101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Sidhu S.S., Fellhouse F.A. Nat. Chem. Biol. 2006;2:682–688. doi: 10.1038/nchembio843. [DOI] [PubMed] [Google Scholar]
  53. Tan M., Jiang X. Trends Microbiol. 2005;13:285–293. doi: 10.1016/j.tim.2005.04.004. [DOI] [PubMed] [Google Scholar]
  54. Tan M., Jiang X. PLoS Pathog. 2010;6:e1000983. doi: 10.1371/journal.ppat.1000983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Yoda T., Suzuki Y., Terano Y., Yamazaki K., Sakon N., Kuzuguchi T., Oda H., Tsukamoto T. J. Clin. Microbiol. 2003;41:2367–2371. doi: 10.1128/JCM.41.6.2367-2371.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zheng D.P., Ando T., Fankhauser R.L., Beard R.S., Glass R.I., Monroe S.S. Virology. 2006;346:312–323. doi: 10.1016/j.virol.2005.11.015. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Data
supp_27_10_339__index.html (1.1KB, html)
supp_gzu023_gzu023supp.doc (28KB, doc)
supp_gzu023_gzu023supp_fig1.docx (437.7KB, docx)
supp_gzu023_gzu023supp_fig2.docx (669.7KB, docx)

Articles from Protein Engineering, Design and Selection are provided here courtesy of Oxford University Press

RESOURCES