Structure of the Core Ectodomain of the Hepatitis C Virus Envelope Glycoprotein 2

J6 eE2 expression

eE2, eE2ΔHVR1 and E2 core domain encompasses residues 384-656, 413-656, and 456-656 from the HCV J6 genome, respectively. Due to incomplete deglycosylation at N7 (542) with EndoH, the crystallization construct contained an asparagine to glutamine mutation at this position. The expression constructs consisted of a CMV promoter, a prolactin signal sequence, E2 fragment, PreScission Protease cleavage site and a carboxyl-terminal protein-A (ProtA) tag. The entire prolactin-E2-ProtA sequence was PCR amplified and cloned into the pJG lentiviral vector (Dr. John Shires, Emory University).

Wild type and GnTI-HEK293T cells ²⁴ (kindly provided by Dr. Davide Comoletti, Rutgers University Robert Wood Johnson Medical School) were maintained in Dulbecco's Modified Eagle Medium (DMEM) with 10% fetal bovine serum (FBS) at 37°C with 5% CO₂. One day prior to the planned transfection, a single T-225 monolayer flask was seeded with 6.0×10⁶ HEK293T cells. 90 μg pJG-E2, 60 μg psPAX2 (HIV Gag-Pol packaging vector), 30 μg pMD2.G (VSV glycoprotein vector), and 450 μL of 2 M CaCl₂ were mixed and brought to a final volume of 4.5 mL with ddH₂0. 4.5 mL of 2× HEPES buffered saline was added at room temperature. After two minute incubation, the mixture was added directly to HEK293T cells. After 6-8 hours, the media was replaced with DMEM with 10% FBS and 1% Antibiotic/antimycotic (A/A) media and incubated for another 70 hours.

Two days post transfection, 10,000 GnTI- HEK293T cells were seeded into a single well of a 96-well plate. The supernatant from the transfection, containing the recombinant lentiviruses, was harvested and centrifuged for 30 min at 4000 ×g at 4°C to pellet large cellular debris. Clarified supernatant was transferred to a Beckman Ultracentrifuge tube and virus was pelleted for 1.5 hours at 25,000 rpm (80,000 ×g) at 4°C in an SW28 rotor. Supernatant was discarded and the pellet resuspended in 120 μL of DMEM containing 20% FBS, 1% A/A, and 8 μg/mL of polybrene. 60 μL of virus suspension was added to the prepared GnTI-HEK293T cells and incubated overnight. Infected cells were expanded and ultimately seeded into an adherent cell bioreactor (Cesco Bioengineering) for long term growth and protein production.

eE2, eE2ΔHVR1, and E2 core purification

Cell supernatant containing E2-ProtA was centrifuged for 10 min at 7,000 ×g, filtered through a 0.22 μM membrane, and loaded onto an IgG FF column (GE Healthcare). The column was extensively washed with 20 mM sodium phosphate pH 7.0 and then equilibrated with 20 mM HEPES pH 7.5, 250 mM NaCl, and 5% glycerol. PreScission Protease was added into the column at approximately 400 μg per L of supernatant and incubated overnight at 4°C. For deglycosylation, the pH of the protein solution was adjusted using 1 M sodium citrate pH 5.5 to a final concentration of 100 mM. EndoH was added at a ratio of 1 mg per 2 mg of E2 and the reaction was incubated at room temperature for 3-4 hours. The deglycosylated proteins were desalted into 20 mM HEPES pH 7.5, 50 mM NaCl, and 5% glycerol and purified by heparin affinity followed by size exclusion chromatography over Superdex200 column. Final yields for all E2 proteins averaged 30 mg per liter of supernatant.

Crystallization

A 1.1:1 molar ratio of E2 core domain to Fab was incubated for 1-2 hours at 4°C. The complex was purified over a Superdex200 column equilibrated with 20 mM HEPES pH 7.5 and 100 mM NaCl. The complex was concentrated to 5-7 mg/mL and crystals were grown by the hanging drop vapor diffusion method. Briefly, 2.5 μL of complex was mixed with an equal volume of reservoir solution, comprising of 22% (w/v) PEG 3350, 0.5 M MgCl₂, 0.1 M HEPES pH 7.5, and 15% (v/v) dioxane. Initially, clusters of plate-like crystals grew in 3-4 days. Single, plate-like crystals were obtained via microseeding using a similar reservoir solution supplemented with 2% (v/v) formamide. Crystals were cryoprotected using reservoir solution with 24% (v/v) ethylene glycol and flash cooled in liquid nitrogen. Data was collected at a wavelength of 0.979 Å using beam line X25 of the National Synchrotron Light Source (NSLS), Brookhaven National Laboratory.

Structure determination and refinement

The crystals belong to space group P2₁2₁2 with cell parameters a= 85.96 Å, b= 194.57 Å, c= 37.92 Å. Phases were determined by the molecular replacement method using PHENIX ²⁷ and the coordinates from chains A and B from PDB entry 2GSI. Unambiguous placement of the Fab heavy and light chains provided the necessary phases to extend the map to cover E2 core domain using iterative rounds of model building and density modification by COOT ²⁸, PHENIX, REFMAC ²⁹, and PARROT ³⁰. The final model was built to a resolution of 2.40 Å, comprising residues 492-522, 538-571, and 596-649 of E2 from the J6 genome, 1-217 of 2A12 light chain, and 1-133, and 136-218 of 2A12 heavy chain with two N-linked, N-acetylglucosamine, six molecules of formamide and 141 solvents molecules. The model coordinates were refined to R_work 0.217 and R_free 0.269. Model validation demonstrated 95.0% of the residues located in the most favorable region of the Ramachandran plot with 4.8% in the generously allowed regions ²⁷. Statistics of the data processing and structure refinement are summarized in Supplementary Table 1.

Small angle X-ray scattering (SAXS)

Glycosylated E2 proteins were purified over IgG and anion exchange columns. The proteins were equilibrated with either pH 7.5 buffer (50 mM HEPES pH 7.5, 250 mM NaCl and 1% glycerol) or pH 5.0 buffer (50 mM sodium citrate pH 5.0, 250 mM NaCl and 1% glycerol) by Superdex200 gel filtration column. Glycosylated eE2ΔHVR1 alone or complex with CD81-LEL (1:2 molar ratios) was purified using pH 7.5 buffer by gel filtration chromatography. Three concentrations of each protein were prepared along with their respective buffers as background control. SAXS data was collected on the SIBYLS beam line at the Advanced Light Source, Lawrence Berkeley National Laboratory. Sample analysis and processing was performed using BioSAXS RAW ²⁵ and GNOM ³¹. The ab initio models were calculated using the application DAMMIF ³². Consensus models and the normalized spatial discrepancy (NSD) values were calculated by averaging 10 ab initio models using the application DAMAVER ³³. X-ray structures of CD81 (PDB ID 1G8Q), HCV E2 core and ab initio models were aligned using the application SUPCOMB ³⁴.

Hydrogen deuterium exchange

HD exchange experiments were conducted as described by Sharma et al (2009) ³⁵. Briefly, 5 μl of deglycosylated eE2 (1.5 mg/mL), in 200 mM NaCl, 20 mM HEPES pH 7.5, was incubated with 15 μl of the same buffer made with 99.96% ²H₂O (Cambridge Isotope Laboratories) for 10, 100, or 1000 seconds and quenched in 30 μl of 2 M urea, 0.8% formic acid and 50 mM tris(2-carboxyethyl)phosphine (TCEP). The reaction mixture was immediately frozen on dry ice until injection. For the zero time point experiment, the protein was incubated in the buffer made with ¹H₂O and then quenched and frozen. To correct background exchange, a completely deuterated sample was produced by incubating the protein with 100 mM TCEP in 99.96% ²H₂O overnight before being quenched and frozen. Dionex RSLC with a C18 column (2.1×50 mm, 3 μm, Q-C18, 150A, CMP Scientific) and LTQ Velos Orbitrap pro were used for LC-MS analysis. The mass was measured using Orbitrap with resolution of 60,000 and mass range from 300-2000 with top10 MSMS performed. The LC-MS data were analyzed using HDexaminer 1.2.0 (Serra Analytics) with manual checking of each peptide afterwards.

Limited proteolysis

8-10 μg of deglycosylated eE2 protein was mixed with trypsin, chymotrypsin or GluC at 1:120 (w/w) ratio (endopeptidase:E2) and incubated at room temperature. Samples were taken at noted time points and analyzed by reducing SDS-PAGE, mass spectrometry and N-terminal sequencing.

Production of monoclonal antibody 2A12

BALB/c mice were immunized intraperitoneally with 50 μg eE2 in either Complete Freund's Adjuvant (first immunization only), or incomplete Freund's Adujvant bi-weekly for eight weeks. A final immunization with 50 μg of eE2 was given intravenously four days prior to collection of splenocytes. Hybridomas were generated using a cloned HAT-sensitive mouse myeloma cell line as a fusion partner. Proliferating hybridomas were screened for their ability to bind eE2 via ELISA, at which point 2A12 was positively identified. Monoclonal antibodies were generated in the laboratory of Dr. Arash Grakoui (IACUC protocol number YER-2002369-070816GN, Emory University School of Medicine, principal investigator Dr. Arash Grakoui).

Generation and purification of 2A12 Fab

Hybridoma cells were expanded to a final volume of 2 L in spinner flasks at 100 rpm using Iscove's Modified Dulbecco's Medium, 10% ultra-low IgG FBS, 1% A/A, and 10 mM HEPES (Life Technologies). Cells were harvested at 2-3×10⁶ cells/mL, centrifuged for 10 min at 7,000 ×g, filtered through a 0.22 μM membrane, and loaded onto a Protein G column (GE Healthcare Life Sciences). After completion, the column was washed with 20 mM sodium phosphate (pH 7.0) followed by phosphate buffered saline (PBS). The antibody was eluted with 0.05% TFA in 2 mL fractions into tubes containing 100 μL of 1 M Tris pH 7.5 for immediate pH neutralization. The eluted antibody was dialyzed into 20 mM sodium phosphate pH 7.0 and 10 mM EDTA. Insoluble papain was added to the antibody at 0.15 mg per 1 mg of antibody. Freshly prepared L-cysteine was added to the reaction to a final concentration of 20 mM and mixed at 37°C for 2 hours. The papain was removed by centrifugation at 3,500 ×g for 2 minutes and filtration through a 0.22 μM membrane. Fab was purified by subtractive chromatography over Protein A FF column and desalted into 20 mM Tris pH 8.0.

Sequencing Ig H and L chain gene segments of 2A12 antibody

Total RNA isolated from 2A12 hybridoma cells was reverse transcribed into cDNA using random hexamers. Expressed heavy (H) and light (L) chains were amplified using standard primers that are complimentary to all murine H and L chain gene segments ³⁶. The PCR products were sequenced either directly or following cloning into pCR 2.1-TOPO vector (Life Technologies).

CD81 purification and binding assays

Human CD81 LEL (residues 112-202) was produced as a fusion with carboxyl-terminal protA tag in HEK293T cells using the same lentiviral expression system described for eE2. Cell culture supernatants were loaded onto an IgG FF column, washed with 20 mM sodium phosphate pH 7.0, eluted with 100 mM sodium citrate pH 3.0 containing 20 mM KCl and immediately neutralized with 1 M Tris pH 9.0. The protA tag was cleaved by PreScission Protease (GE Healthcare Life Sciences) in a ratio of 1:50 (w/w) followed by overnight dialysis in 20 mM HEPES pH 7.5, 250 mM NaCl, and 5% glycerol. High purity CD81 protein was obtained by anion exchange and size exclusion chromatography.

For binding studies, a 96-well plate (Nalgene Nunc, Thermo Fisher Scientific) was coated with 50 μg of CD81-LEL overnight at 4°C. All experiments were duplicated against BSA as a negative control. Plates were washed 3× with PBS containing 0.05% Tween 20 (PBS-T) and blocked with 3% (w/v) BSA in PBS-T for 1 hour at room temperature. 50 μL of eE2 or E2 core at different concentrations was added to appropriate wells and incubated overnight at 4°C. On day 3, the wells were washed 3× with PBS-T and incubated with mAb 2A12 cell supernatant for 1 hour at room temperature. Plates were washed 3× with PBS-T and incubated with anti-mouse-HRP conjugated antibody for 1 hour at room temperature. Finally, the plate was washed 5× with PBS-T. 50 μL of TMB substrate (ThermoFisher Scientific) was added to each well and incubated for 5 min, followed by the addition of 50 μL of 2 M sulfuric acid to stop the reaction. Absorbance readings were acquired at 450 nm using Softmax Pro software on a Spectra Max 250 (Molecular Devices).

Neutralization assay

Huh-7.5 cells were maintained in DMEM containing 10% FBS (Hyclone,) and 100 μg/mL of penicillin and streptomycin (Cellgro) at 37°C in 5% CO₂. Naive Huh-7.5 cells were seeded at 6,000 cells per well in a 96-well plate. The following day, 100 μl of 2C1, 2A12, or H113 serially diluted in complete media were added per well at various concentrations. In parallel, 100 μL of eE2, E2 core, gp140, or CD81-LEL serially diluted in complete DMEM were added at varying concentrations beginning at 100 μg/mL. Cells were then infected with 100 μL of genotype 2a virus Cp7 encoding the Renilla luciferase gene ³⁷. 72 hours post-infection, relative light units were measured on a Clarity 4.0 luminometer (Biotek) using the Renilla Luciferase Assay System (Promega).

Assessment of cellular cytotoxicity

Huh-7.5 cells were incubated with varying concentrations of protein as described above, beginning at 100 μg/mL. After 72 hours, cells were washed once with PBS, treated with trypsin, and harvested in 100 μL PBS. Cells were stained with 7-AAD according to the manufacturer's instructions (BD Biosciences) and analyzed using a BD LSR II and FlowJo software (Tree Star).

Human plasma ELISA

96 well enzyme immunoassay plates (ThermoFisher Scientific) were coated overnight at 4° with 50 μL of a 1 μg/mL solution of eE2 or E2 core diluted in 0.1 M Na₂CO₃. Plates were washed twice with PBS-T and then blocked for one hour at 37 °C in PBS-T containing 10% fetal calf serum (HyClone). Blood samples were collected in heparin tubes (Becton Dickinson) and plasma was isolated and frozen at -80 °C. Plasma was serially diluted in a binding solution composed of 0.1% (v/v) normal goat serum in PBS-T (Jackson ImmunoResearch Laboratories). 100 μL of sample were added per well and incubated at room temperature for 90 minutes. Following eight washes with PBS, 100 μL of mouse anti-human IgG biotin antibody (Mabtech) diluted 1:20,000 in binding solution were added per well and incubated 1 hour at room temperature. Following five additional washes with PBS, 100 μL streptavidin-horseradish peroxidase (HRP) conjugate was added to each well at a 1:2,000 dilution in binding buffer and incubated for 45 minutes at room temperature (Mabtech). Absorbance was measured and analyzed using a VersaMax microplate reader and SoftMax Pro software (Molecular Devices) following five washes and the addition of tetramethylbenzidine substrate solution (Ebioscience). Human sera were isolated from whole-blood samples informed consent was obtained for all subjects. (IRB no. 1358-2004, Emory University School of Medicine, principal investigator Dr. Arash Grakoui).

Alignment

Secondary structures were assigned using the program DSSP ³⁸. Sequences were obtained from the National Center for Biotechnology Information (NCBI) using the following accession numbers: J6 ADV40003.1, H77 ACA53555.1, J8 P26661.3, S52 AEB71616.2, ED43 AEB71617.2, SA13 AEB71618.2, HK6a AEB71619.2, QC69 ACM69041.1. The E2 sequences were aligned with multiple alignment using fast fourier transform (MAFFT) ³⁹ and edited for figure generation using JalView version 2 ⁴⁰.