Antibody cloning, expression and purification
RAB1 and RAB2 were previously developed as anti-rabies glycoprotein human monoclonal antibodies isolated from human IgG transgenic mice immunized with rabies vaccine as previously described9. The heavy chain and light chain variable regions of RAB1 and RAB2 were amplified from hybridoma cells by RT-PCR and cloned into a pcDNA 3.1 (Life Technologies) vector containing the heavy or light constant regions of human IgG1. For MAb CR57, the variable region sequences of light chain (GenBank: CS239779), and heavy chain (GenBank: CS239777), were synthesized (GenScript Inc) and cloned into IgG1 expression vectors as described above.
Antibodies were expressed in CHO cells stably transfected with DNA encoding RAB1, RAB2, or CR57 IgG1. The cell supernatants were harvested by centrifugation, and then incubated with protein A Sepharose resin (GE Healthcare) for 2 h at room temperature. The beads were washed with phosphate buffered saline (PBS) and the antibody was eluted with 100 mM glycine pH 2.8. The antibody was then dialyzed into PBS storage buffer. Purified antibody was filter sterilized and the protein concentration was determined by spectrophotometry.
HEK-293 T/17 cells were grown in Dulbecco’s Modified Eagle Medium (DMEM) (ThermoFisher scientific) supplemented with 10% Fetal bovine serum (FBS) (Millipore Sigma) and 100 IU of penicillin–streptomycin (ThermoFisher scientific) at 37ºC with 5% CO2. The HEK293T/17 cells were harvested using PBS containing 5 mM EDTA (UltraPure™, pH 8.0, ThermoFisher) and then incubated for 5 min at room temperature. Human osteosarcoma (HOS) cells were grown in minimum essential medium (MEM) (ThermoFisher) supplemented with 10% FBS, 2% L-Glutamine, 2% MEM Vitamins, 1% penicillin–streptomycin (Gibco), and 1% sodium bicarbonate and then incubated at 37 °C with 5% CO2. Cells were harvested using trypsin–EDTA (0.25%) (ThermoFisher).
Generation of RAB2 resistant viruses
The method to generate RABV resistant viruses was described in our previous study9. Briefly, mouse neuroblastoma cells were plated at 1.5 × 105 cells/ml on Day 1. On Day 2, a titration of 1 × 101 to 108 FFU/ml/well of CVS-11 RABV was incubated with 50 ug/ml of RAB2 at 37 °C for 1 h. The virus/RAB2 mix was added to cells and incubated at 37 °C. 12 h later, cell media was exchanged with fresh media containing 50ug/mL RAB2. Cells were incubated for additional 3-days at 37 °C. On Day 5, supernatant containing potentially resistant virus was harvested and stored at 4 °C. The cells were fixed and stained for RFFIT. Virus taken from wells containing 1–5 fluorescent foci were further amplified on MNA cells for 3 days in the presence of RAB2. RNA was extracted from virus-infected cells for RT-PCR and sequencing analysis for the potential mutations in glycoprotein-encoding genes. The sequencing result was confirmed by two separate infections/RNA preparations to confirm the observation.
Cloning of rabies virus ERA glycoproteins
The gene sequence encoding glycoproteins of RABV Evelyn-Rokitnicki-Abelseth (ERA) was codon-optimized for mammalian expression as described previously9,15. The amino acid (a.a.) sequence of the RABV glycoprotein full length surface protein (a.a. 1–524), as well as its soluble fragment (a.a 1–439), were cloned into an expression vector pcDNA3.1Myc/His (ThermoFisher). Proteins were expressed in frame with the myc/his tags. Alanine scanning mutants were generated using site-directed mutagenesis. Overlapping primers containing the desired point mutations were used to amplify full-length mutant glycoprotein genes and the pcDNA3.1Myc/His vector from the previously cloned codon-optimized ERA RABV glycoprotein. PCR positive clones were selected according to the manufacturer’s recommendation (New England BioLabs® Site-Directed Mutagenesis Kit Protocol (E0554)).
Recombinant glycoprotein expression and purification
All constructs were transfected into HEK-293 /17(ATCC) cells using a 1:4 DNA-Polyethylenimine (PEI) (Sigma) ratio, as described by the manufacturer. Cells were grown to 95% confluence in 150 mm tissue culture dishes in 30 mL of DMEM-10% (FBS) (Sigma). 30 µg of DNA mixed with 120 µl of PEI was added to the cells, and the plates were incubated overnight at 37 °C. The media was removed and stored at 24-, 48- and 72-h post-transfection for secreted soluble proteins or was discarded for membrane bound proteins.
Soluble rabies glycoproteins containing Myc and His epitope tags were purified from the cell culture supernatant by incubation with nickel-nitrilotriacetic acid (Ni–NTA) beads (ThermoFisher) for 2 h, followed by column filtration using 10 mM imidazole to washg non-specific binding off of the column, and 250 mM imidazole for the final elution of the protein. The eluted proteins were then dialyzed for two hours in PBS at pH 7.
Affinity determination for RAB2
Bio-layer interferometry (BLI) with an Octet HTX (PALL/ForteBio) was used to determine the affinity of RAB2 IgG for the soluble glycoprotein of ERA. RAB2 was added to 96 well plates at 1000 nM and titrated 1:2 to 62 nM in PBS. Purified RAB2 antibody was immobilized on anti-human IgG Fc Capture (AHC) Biosensors (ForteBio) for 120 s at 333 nM. After a baseline step, the RAB2-antigen binding rate was determined when the biosensors with immobilized antibody were exposed to ERA glycoprotein at different concentrations for 120 s. Following the association step, the soluble RAB2-ERA glycoprotein complex was exposed to PBS and the rate of the ERA glycoprotein dissociation from RAB2 was measured. Each assay was performed in triplicate. The binding affinity of RAB2 for the ERA glycoprotein was calculated using association and dissociation rates with ForteBio Data analysis software v8.1 (PALL).
Cell surface staining
HEK-293T/17 cells were transfected with constructs encoding the full-length wild type glycoprotein of ERA or with individual alanine mutants. Cells were harvested 48 h post-transfection and incubated with varying concentrations of HuMabs. The binding activities of HuMAb to glycoproteins was detected by phycoerythrin labeled anti-human IgG (Jackson) in flow cytometry (MACSQuant), and analyzed with MACSQuant® Analyzer 10 (Miltenyi Biotec®).
ELISA plates were coated with ERA soluble glycoproteins (a.a. 1–439). Plates were washed and then blocked with ELISA blocking buffer (EBB; PBS, 1% BSA, 0.05% Tween). Antibody samples were added to the plate and then diluted 1:2 in EBB. Samples were detected with a goat anti-human IgG conjugated with Alkaline Phosphatase (1:500, Jackson ImmunoResearch). The plates were developed with p-Nitrophenyl phosphate disodium salt at 1 mg/mL in 1 M diethanolamine (ThermoFisher) for 20 min, and then analyzed at 405 nm with a Vmax plate reader (Molecular Devices).
Production of pseudotyped viruses and neutralization assay
A plasmid containing a replication defective lentivirus backbone (Env-, Vpr- HIV; HIV resources) fused with the firefly luciferase gene, pNL4-3.Luc.R-E-, was co-transfected with a plasmid encoding RABV glycoprotein ERA into 293 T cells using PEI as described previously16,17. Following a media change 24 h post-transfection, pseudoviral particles were harvested 48–72 h post-transfection, concentrated, and stored at − 80 °C. An infection assay was performed to determine the luciferase counts per second (cps) of the pseudovirus preparations. 24 h post-infection, the media was changed to DMEM without phenol red (ThermoFisher) with 10% FBS and 1% Penicillin–Streptomycin. Luciferase activity was detected 72 h post-infection using the Bright-Glo Luciferase Assay System (Promega) and analyzed by Victor3 de nivo multilabel plate reader (PerkinElmer). Neutralization assays were performed with approximately 50,000 cps of pseudovirus and incubated for 1 h at room temperature with a 1:5 serial dilution of antibody mix starting at 37 ng/ml. The antibody/virus suspension was then applied to HOS cells (ATCC# CRL-1543) which were then incubated at 37ºC with 5% CO2. Luciferase activity was assayed 72 h post-infection using the Bright-Glo reagent (Promega), according to the manufacturer’s protocol. Neutralization results are expressed as the percent inhibition of the baseline luciferase activity.
In vitro RABV RFFIT neutralization
The rapid fluorescent focus inhibition test (RFFIT) was performed as described previously18. MAbs were diluted to a standard starting concentration then serially diluted fivefold. The starting concentration of MAb was increased until neutralization was detected within the dilution range. Different RABV isolates or laboratory strains (including CVS-11) were added to the MAb dilutions. The starting concentration of each virus was empirically determined to result in 50 × 50% fluorescent foci doses (50 FFD50) with an acceptable range of 32–100 FFD50. For each virus, standard rabies immune globulin (SRIG, US FDA lot R-3) at 2 IU/mL, HyperRab (HRIG, Grifols, Barcelona, Spain) at > 150 IU/mL, each individual MAb, and a 1:1 mixture (cocktail) of the component MAb was tested.
In vivo efficacy in the Syrian hamster model
Experimental protocols were approved by the CDC’s Institutional Animal Care and Use Committee (#2622SMIHAMC-A18) prior to starting and strictly followed to ensure animal welfare. All methods were carried out in accordance with ARRIVE guidelines and regulations (https://arriveguidelines.org). Four-week-old female Syrian hamsters were purchased from Charles River Laboratory (Wilmington, MA, USA) or Envigo (Indianapolis, IN, USA) and arbitrarily assigned to groups containing 12 or 21 animals depending on the RABV variant used. Animals underwent a three-day quarantine/acclimation period and veterinary examination prior to starting. On day 0, animals were anesthetized with isoflurane (3–5%), and labeled with subcutaneous injection of a preprogrammed passive integrated transponder. While still under anesthesia, the animals were weighed, 0.15 mL of blood was collected using the subclavicle method, and 0.1 mL of RABV was injected intramuscularly (IM) in the gluteal region of the left hind leg. Virus titration data was previously determined in hamsters with various isolates. Each dose is titrated to cause > 80% mortality in untreated and vaccine alone groups. Challenge doses and groups sizes were as follows: canine RABV variant, TX coyote 323R, 103.5 MICLD50, n = 12; fox RABV isolate, AZ fox 2400L, 8.6 × 102 ffu, n = 21; Bat RABV variant, PA bat EF A06-3684, 1.4 × 103 ffu, n = 21. Animals were monitored every five minutes until they fully recovered from anesthesia and then once daily until 7 days post-infection (pi).
PEP was initiated 24 h pi and consisted of 20 IU/kg of HyperRab (HRIG, Grifols, and Barcelona, Spain) as a positive control. 1.1 mg/kg of an irrelevant HuMAb was used as a negative control. 1.1 mg/kg of HuMAb cocktail R172, 0.1 mg/kg of R172, or 0.01 mg/kg of R172 was injected IM in the same leg as the challenge virus.0.05 mL of the Imovax rabies vaccine (Sanofi Pasteur, Lyon, France) was injected IM in the opposite leg. One group did not receive any PEP. Rabies vaccine administration was repeated for the surviving animals 4, 8, 15, and 29 days pi (days 3, 7, 14, and 28 of PEP). Blood collection under anesthesia was repeated on days four and eight pi (days 3, and 7 of PEP). The PEP injection site was assessed daily until day 8 pi for any localized reactions.
From days 7 to 21 pi, animals were observed twice daily and were euthanized by isoflurane overdose at the first clinical signs of rabies, according to established euthanasia criteria, or as deemed necessary by the attending veterinarian. From day 22 pi until the experimental endpoint, animals were observed once daily. All surviving animals were euthanized 45 days pi.
All animals that were euthanized with clinical signs and approximately half of the animals surviving to day 45 were tested for RABV antigen in the brain stem by a standard direct fluorescent antibody (DFA) test19. For selected animals, RNA was extracted from the brain stem and tested for RABV RNA using the LN34 assay as described previously20 or converted to cDNA for RABV sequencing. Serum was separated from the collected blood and used to measure virus neutralizing antibodies in a standard RFFIT18.
All statistical calculations were performed using GraphPad Prism version 8.1.1 (GraphPad Software, La Jolla, CA). EC50 and IC50 values were calculated by sigmoidal curve fitting using nonlinear regression analysis. Survival analysis was calculated from a Kaplan–Meier curve using a Mantel-Cox test. p < 0.05 was considered significant.
Data availability and method declaration
The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request. All methods were performed in accordance with the relevant guidelines and regulations.