A self-amplifying RNA vaccine against COVID-19 with long-term room-temperature stability

saRNA expression plasmid design, cloning, and production

Three saRNA plasmids each with a unique candidate SARS-CoV-2 spike open reading frame were created, along with a fourth saRNA plasmid expressing SEAP instead of a vaccine antigen as an appropriate vector control. The SARS-CoV-2 spike open reading frame sequence from GenBank MT246667.1 incorporating the additional D614G mutation was used as the “baseline” sequence, with additional modifications to create two additional vaccine candidates: D614G-2P represents the baseline sequence with a substitution of PP for KV at amino acid positions 987–988 and an addition of nine N-terminal codons from the reference genome encoding amino acid sequence MFLLTTKRT; D614G-2P-3Q refers to the baseline sequence with the diproline substitution and an additional substitution of QQAQ for RRAR at the furin cleavage site at amino acid positions 683–686. These sequences were then codon-optimized for mammalian (human) expression by Codex DNA (San Diego, CA) using a proprietary algorithm, synthesized by BioXp (Codex DNA), and inserted into AAHI’s backbone saRNA expression vector by Gibson cloning. The SEAP-expressing plasmid was created by a similar process to insert the SEAP expression sequence in place of the vaccine antigen. Plasmid sequences were all confirmed by Sanger sequencing. Template plasmids were amplified in E. coli and extracted using Qiagen (Germantown, MD) maxi- or gigaprep kits, followed by NotI linearization (New England Biolabs [NEB], Ipswich, MA). Linearized DNA was purified by Qiagen DNA purification kits.

RNA manufacture

Vaccine saRNA was generated by T7 polymerase-mediated in vitro transcription (IVT) using NotI-linearized DNA plasmids as templates. An in-house optimized IVT protocol was used with commercially available rNTP mix (NEB) and commercially available T7 polymerase, RNase inhibitor, and pyrophosphatase enzymes (Aldevron, Fargo, ND). DNA plasmids were digested away (DNase I, Aldevron), and Cap 0 structures were added to the transcripts by treatment with guanylyltransferase (Aldevron), GTP, and S-adenosylmethionine (NEB). RNA was chromatographically purified using Capto Core 700 resin (GE Healthcare, Chicago, IL) followed by diafiltration and concentration through tangential flow filtration using a 750 kDa molecular weight cut-off modified polyethersulfone membrane (Repligen, Waltham, MA). The final bulk RNA contained 10 mM Tris-HCl pH 8. Terminal filtration of the saRNA material was done using a 0.22 µm PES filter, and the saRNA materials were stored at −80 °C until use/complexation. Agarose gel electrophoresis was used to characterize saRNA size and integrity. All gels that derive from the same experimental timepoint were processed in parallel. RNA concentration was quantified by UV absorbance (NanoDrop 1000) and RiboGreen assay (Thermo Fisher Scientific, Waltham, MA).

RNA for versions of saRNA with modified nucleosides was made with n1-methylpseudouridine-5’-triphosphate (TriLink BioTechnologies, San Diego, CA) substituted for UTP. The other three NTPs were used as normal, and production proceeded normally as described above.

NLC manufacture

Nanostructured lipid carrier (NLC) formulation was produced as follows38,39,44. Trimyristin (Sigma-Aldrich, St. Louis, MO), squalene (SAFC Supply Solutions, St. Louis, MO), sorbitan monostearate (Spectrum Chemical Mfg. Corp., New Brunswick, NJ), and the cationic lipid DOTAP (Lipoid, Ludwigshafen, Germany) were mixed and heated at 60 °C in a bath sonicator to create the oil phase. Polysorbate 80 (MilliporeSigma, Burlington, MA) was separately diluted in 10 mM sodium citrate trihydrate and heated to 60 °C in a bath sonicator to create the aqueous phase. After dispersion of components in each phase, a high-shear mixer (Silverson Machines, East Longmeadow, MA) was used at ~5000 rpm to mix the oil and aqueous phases. Particle size of the mixture was then further decreased by high-pressure homogenization by processing at 30,000 psi for ten discrete passes using an M110P Microfluidizer (Microfluidics, Westwood, MA). The NLC product was then filtered through a 0.22 µm PES filter and stored at 2–8 °C until use.

Vaccine complexing and characterization

Vaccine complexes for immediate in vivo injection were created by mixing aqueous RNA 1:1 by volume with NLC diluted in a buffer containing 10 mM sodium citrate and 20% w/v sucrose. All vaccines were prepared at nitrogen:phosphate (N:P) ratio of 15, representing the ratio of amine groups on the NLC DOTAP to phosphate groups on the RNA backbone. This produced a vaccine solution containing the intended dose of complexed saRNA/NLC in an isotonic 10% w/v sucrose, 5 mM sodium citrate solution (with small (<4 mM) amounts of Tris buffer present from the bulk RNA material). Vaccine was incubated on ice for 30 min after mixing to ensure complete complexing.

Hydrodynamic diameter (particle size) of the NLC and vaccine nanoparticles was determined using dynamic light scattering (Zetasizer Nano ZS, Malvern Panalytical, Malvern, UK) on triplicate 1:100 dilutions in nuclease-free water. Sizing was done in a disposable polystyrene cuvette using previously established parameters39. Zeta potential (particle charge) of the vaccine nanoparticles was also measured by dynamic light scattering in a disposable folded capillary cell using the same instrument with the same material and dispersant parameters.

Free and complexed RNA integrity and NLC-provided protection against RNases were evaluated by visualizing RNA integrity after agarose gel electrophoresis. RNA samples were diluted to 40 ng/μL RNA in nuclease-free water. For RNase treatment, this diluted RNA was incubated with RNase A (10 mg/mL, Thermo Scientific, Waltham, MA) at a mass ratio of 1:500 RNase:RNA for 30 min at room temperature, followed by incubation with proteinase K (~20 mg/mL, Thermo Scientific) at a mass ratio of 1:100 RNase A:proteinase K for 10 minutes at 55 °C. For complexed samples, RNA was extracted from complexes prior to gel electrophoresis by phenol:chloroform extraction. All RNA samples were mixed with glyoxal loading dye (Invitrogen, Waltham, MA) 1:1 by volume, incubated at 50 °C for 20 min, loaded on a denatured 1% agarose gel in NorthernMax-Gly running buffer (Invitrogen) alongside Millennium RNA Markers (Thermo Fisher Scientific), and run at 120 V for 45 minutes before imaging on a ChemiDoc MP Imaging System (Bio-Rad Laboratories, Hercules, CA). All gels that derive from the same experiment timepoint were processed in parallel.

Vaccine lyophilization

Vaccine complexes intended for lyophilization and storage were prepared similarly to above by mixing aqueous RNA 1:1 by volume with NLC (in 10 mM sodium citrate) diluted in either 20% w/v sucrose only, 20% w/v sucrose and 5 mM sodium citrate, or 20% w/v sucrose with both 1 mM sodium citrate and 10 mM Tris. All vaccines with all excipient conditions contained approximately 2.6 mM Tris buffer from the bulk RNA material and were prepared at an N:P ratio of 15 and incubated on ice for 30 min after mixing to ensure complete complexing. After complexing, vaccine was lyophilized using a VirTis AdVantage 2.0 EL-85 (SP Industries, Warminster, PA) benchtop freeze dryer controlled by the microprocessor-based Wizard 2.0 software. The lyophilization cycle began with a freezing step at −50 °C, followed by primary drying at −30 °C and 50 mTorr, and finishing with secondary drying at 25 °C and 50 mTorr. When the cycle was complete, the samples were brought to atmospheric pressure, blanketed with high-purity nitrogen, and stoppered before being removed from the freeze-dryer chamber. At each timepoint, lyophilized samples were reconstituted using nuclease-free water to their original concentration.

Mouse studies

All animal studies in this work were approved by the Bloodworks Northwest Research Institute’s Institutional Animal Care and Use Committee (IACUC) (Seattle, WA) and conducted at the Bloodworks Northwest Research Institute Animal Facility. All animal work was in compliance with all applicable sections of the Final Rules of the Animal Welfare Act regulations (9 CFR Parts 1, 2, and 3) and the Guide for the Care and Use of Laboratory Animals, Eighth Edition77. This work complied with all pertinent ethical regulations for animal testing and research.

C57BL/6J mice obtained from The Jackson Laboratory (Harbor, ME) were used for all animal studies in this work. Mice were between 6 and 8 weeks of age at study onset. Equal numbers of male and female mice were used for all studies, with the sole exception of the Fig. 6 stability study in which all female mice were used. Mice were immunized by intramuscular injection bilaterally in the rear quadriceps muscle (50 µL/leg, 100 µL total). Survival blood samples were taken by retro-orbital bleed; terminal blood samples were taken by cardiac puncture.

Serum IgG, IgG1, and IgG2a titers by ELISA

SARS-CoV-2 spike protein-binding IgG antibodies in mouse serum were measured by ELISA. Plates (384-well high-binding plates, Corning, Corning, NY) were coated with 1 µg/mL of Recombinant SARS-CoV-2 Spike His Protein, Carrier Free, (R&D Systems, Minneapolis, MN; #10549-CV) in phosphate-buffered saline (PBS) and incubated overnight at 4 °C. The coating solution was removed, and blocking buffer (2% dry milk, 0.05% Tween 20, and 1% goat serum) was applied for at least 1 h. On a separate, low-binding plate, each sample was diluted 1:40 and then serially 1:2 to create a 14-point dilution curve for each sample. Naïve mouse serum, used as the negative control, was diluted identically to the samples. A SARS-CoV-2 neutralizing monoclonal antibody (mAb; GenScript, Piscataway, NJ; #A02057), was used as a positive control at a known starting concentration of 3.2 ng/µL followed by serial 1:2 dilutions similarly to each sample and negative control. SARS-CoV-2 spike protein-coated and blocked assay plates were washed, and serially diluted samples were then transferred onto the coated plates followed by a 1-h incubation. Plates were then washed, and spike protein-bound antibodies were detected using an Anti-Mouse IgG (Fc Specific)-Alkaline Phosphatase antibody (Sigma-Aldrich, #A2429) at a 1:4000 dilution in blocking buffer. Plates were washed and then developed using phosphatase substrate (Sigma-Aldrich, #S0942) tablets dissolved in diethanolamine substrate buffer (Fisher Scientific, Waltham, MA; #PI34064) at a final concentration of 1 mg/mL. After a 30-minute development, plates were read spectrophotometrically at 405 nm. A 4-point logistic curve was used to fit the antibody standard titration curve. Sample concentrations were interpolated off the linear region of each sample dilution curve using the standard curve for absolute quantification of antibody titers. For Supplementary Fig. 4 only, antibody level was quantified by endpoint titer, with the endpoint titer defined as the dilution at which each sample dilution curve rises to above 3 standard deviations above assay background.

For IgG1 and IgG2a isotype-specific ELISAs, the identical plate coating and blocking procedures were conducted. For the IgG1 assay, the standard curve was run using an IgG1 SARS-CoV-2 neutralizing antibody (GenScript #A02055) or an IgG2a SARS-CoV-2 neutralizing antibody (GenScript #BS-M0220) for full quantification. Sample dilution and incubation were identical to the total IgG curve, and plates were probed with IgG1- and IgG2a-specific secondary alkaline phosphatase-conjugated detection antibodies prior to development, reading, and quantification as described above.

Pseudovirus neutralization assay

The SARS-CoV-2 pseudovirus-neutralizing antibody titers of immunized mouse sera were measured by pseudovirus neutralization assays using procedures adapted from Crawford et al.78. Lentiviral SARS-CoV-2 spike protein pseudotyped particles were prepared by following the Bioland Scientific (Paramount, CA) BioT plasmid transfection protocol. Briefly, HEK-293 cells (American Type Culture Collection, Manassas, VA; #CRL-3216) were plated at 4 × 105 cells/mL in six-well plates 18–24 hours before the assay to achieve 50–70% confluency at assay start. Cellular growth medium was then replaced with serum-free Gibco Dulbecco’s Modified Eagle Medium (DMEM) with GlutaMAX immediately prior to transfection. The HEK-293 cells were co-transfected with several plasmids: a plasmid containing a lentiviral backbone expressing luciferase and ZsGreen (BEI Resources, Manassas, VA; #NR-52516), plasmids containing lentiviral helper genes (BEI Resources; #NR-52517, NR-52518, and NR-52519), and a plasmid expressing a delta19 cytoplasmic tail-truncated SARS-CoV-2 spike (Wuhan strain, B.1.1.7, and B.1.351 plasmids from Jesse Bloom, Fred Hutchinson Cancer Center, Seattle, WA; and B.1.617.2 plasmid from Thomas Peacock, Imperial College London, UK). BioT transfection reagent (Bioland Scientific) was used to mediate the co-transfection. The assay plates were incubated for 24 h at standard cell culture conditions (37 °C and 5% CO2). Then the serum-free media was aspirated off, and fresh growth medium (Gibco DMEM with GlutaMAX and 10% fetal bovine serum [FBS]) was added to the plates before returning them to the incubator for an additional 48 h. Pseudovirus stocks were collected by harvesting the supernatants from the plates, filtering them through a 0.2 μm PES filter (Thermo Scientific), and freezing at −80 oC until titration and use.

Mouse serum samples were diluted 1:10 in medium (Gibco DMEM with GlutaMAX and 10% FBS) and then serially diluted 1:2 for 11 total dilutions in 96-well V-bottom plates. Polybrene (Sigma-Aldrich) was then added at a concentration of 5 µg/mL to every well on the plate, and pseudovirus, diluted to a titer of 1 × 108 total integrated intensity units/mL as per titers determined independently for each pseudovirus batch, was added 1:1 to the diluted serum samples. The plates were incubated for 1 h at 37 °C and 5% CO2. Serum-virus mix was then added in duplicate to Human Angiotensin-Converting Enzyme 2-expressing HEK-293 cells (BEI Resources, #NR-52511) seeded at 4 × 105 cells/mL on a 96-well flat-bottom plate and incubated at 37 °C and 5% CO2 for 72 hours. To determine 50% inhibitory concentration (IC50) values, plates were scanned on a high-content fluorescent imager (ImageXpress Pico Automated Cell Imaging System, Molecular Devices, San Jose, CA) for ZsGreen expression. Total integrated intensity per well was used to calculate the percent of pseudovirus inhibition in each well. Neutralization data for each sample were fit with a four-parameter sigmoidal curve that was used to interpolate IC50 values.

A Wuhan-strain pseudoneutralization test using the World Health Organization (WHO) standard for neutralization assays with an official IC50 of 1000 resulted in an IC50 of 7800 using our assay, suggesting that our test is more sensitive than the WHO assay but may overreport IC50 values—by less than one log10—an effect likely most pronounced for strongly neutralizing samples.

Splenocyte harvest, intracellular cytokine staining, and flow cytometry

Spleens were dissociated in 4 mL of RPMI medium by manual maceration through a cell strainer using the end of a syringe plunger. Homogenized samples were briefly centrifuged at 400 × g (15–20 s) to pellet fat cells. Samples were then carefully resuspended and fat clumps were removed by pipette. The supernatants containing lymphocytes were transferred to 5-mL mesh-cap tubes to strain out any remaining tissue debris or were lysed with ammonium-chloride-potassium buffer and washed. Cell counts for each sample were obtained on a Guava easyCyte (Luminex, Austin, TX). Each spleen sample was seeded in 96-well round-bottom plates at 1–2 × 106 cells per well in RPMI medium containing 10% FBS, 50 μM beta-mercaptoethanol, CD28 costimulatory antibody (0.4 μL/test, BD Biosciences #553294), and brefeldin A. Cells were stimulated with one of three stimulation treatments: 0.0475% dimethyl sulphoxide (DMSO) as a negative stimulation control, 0.2 μg/well (1 μg/mL) per peptide of spike peptide pool (JPT Peptide Technologies, Berlin, Germany; #PM-WCPV-S-1) in an equivalent amount of DMSO, or 10 μg/well of phorbol myristate acetate (PMA)/ionomycin solution. After 6 h of incubation at 37 °C with 5% CO2, plates were centrifuged at 400 × g for 3 min, the supernatants were removed by pipetting, and cells were resuspended in PBS. Plates were centrifuged, the supernatants were removed, and cells were stained for flow cytometry. Splenocytes were stained for viability with Zombie Green (BioLegend, San Diego, CA) in 50 μL of PBS, and then Fc receptors were blocked with CD16/CD32 antibody (0.25 μL/test, Invitrogen #14–0161–86). Cells were then surface stained with fluorochrome-labeled mAbs specific for mouse CD4 (0.3 μL/test, PerCP-Cy5.5, eBioscience #45–0042–82), CD8 (2 μL/test, BV510, BD Biosciences #563068), CD44 (0.2 μL/test, APC-Cy7, BD Biosciences #560568), and CD107a (2 μL/test, APC, BioLegend #121614) in 50 μL of staining buffer (PBS with 0.5% bovine serum albumin and 0.1% sodium azide). Cells were washed twice, permeabilized using the Fixation/Permeabilization Kit (BD Biosciences, Franklin Lakes, NJ), and stained with fluorochrome-labeled mAbs specific for mouse TNFα (2 μL/test, BV421, BioLegend #506327), IL-2 (0.4 μL/test, PE-Cy5, BioLegend #503824), IFNγ (0.2 μL/test, PE-Cy7, Invitrogen #25731182), IL-5 (0.7 μL/test, PE, eBioscience #12–7052–82), IL-10 (2 μL/test, BV711, BD Biosciences #564081), and IL-17A (2 μL/test, AF700, BD Biosciences #560820). After two washes in staining buffer, cells were resuspended in 100 μL of staining buffer and analyzed on an LSRFortessa flow cytometer (BD Biosciences). After initial gating for live CD4+ or CD8+ lymphocytes, cells were gated for cytokine positivity. Quality of the response was determined by gating on cells that were double or triple positive for these markers. Cells triple positive for TNFα, IL-2, and IFNγ were considered activated polyfunctional T cells.

T-cell ELISpot assay

ELISpot plates (MilliporeSigma) were coated with either IFNγ (BD Biosciences, #51–2525KZ), IL-17A (Invitrogen, #88–7371–88), or IL-5 (BD Biosciences, #51–1805KZ) capture antibodies at a 1:200 dilution in Dulbecco’s PBS (DPBS; Gibco). After overnight incubation at 4 oC, plates were washed and then blocked with complete RPMI (cRPMI) medium for at least 2 h. Splenocytes harvested as described above were plated at 2 × 105 cells per well. A subset of each sample was stimulated with PepMix SARS-CoV-2 (JPT Peptide Technologies, #PM-WCPV-S-1) at a final concentration of 1 µg/mL. Plates were then incubated at 37 °C and 5% CO2 for 48 h. After a wash with PBS with 0.1% Tween 20, 100 µL of detection antibody (IFNγ, BD Biosciences, #51–1818KA; IL-17A, Invitrogen, #88–7371–88; and IL-5, BD Biosciences, #51–1806KZ) was added at a 1:250 dilution in ELISpot diluent (eBioscience) overnight at 4 °C. Plates were washed and developed with Vector NovaRED Substrate Peroxidase (Vector Laboratories, Burlingame, CA; #SK-4800) for 15 min. The reaction was stopped by washing the plates with deionized water, and plates were left to dry in the dark. Spots were counted and data were analyzed using ImmunoSpot 7 software (Cellular Technology Limited, Cleveland, OH).

Bone marrow IgA and IgG antibody-secreting cell ELISpot assays

Presence of antibody-secreting bone marrow-resident cells was measured by ELISpot. MultiScreen IP Filter plates (0.45 µm, MilliporeSigma) were treated with 15 µL of 35% ethanol for 30 s. After a wash with PBS, plates were coated with 100 µL of Recombinant SARS-CoV-2 Spike His Protein, Carrier Free, (R&D Systems, #10549-CV-100) at a concentration of 2 µg/mL diluted in DPBS (Gibco). Plates were incubated overnight at 4 °C, washed with PBS with 0.1% Tween 20, and blocked with cRPMI medium for at least 2 h.

To prepare bone marrow, both femurs were removed from each mouse and inserted into a snipped-end 0.6 mL tube (Eppendorf, Hamburg, Germany) inserted into a 1.5 mL Eppendorf tube containing 1 mL of cRPMI medium. Femurs were centrifuged for 15 s at 840 × g, and supernatant was discarded. The cell pellets were briefly vortexed and resuspended in 200 µg of RBC lysis buffer (pH 7.1–7.4) (Invitrogen), and they were incubated on ice for 30 s. After addition of 800 μL of cRPMI medium, cells were centrifuged 5 min at 400 × g, and supernatant was decanted. Cells were resuspended in 1 mL of cRPMI medium, counted, and transferred to prepared filter plates described above at 1 million cells per well followed by a threefold dilution across five adjacent wells.

After a 3-h incubation, plates were washed with PBS with 0.1% Tween 20, and secondary antibody (Goat Anti-Mouse IgG-HRP or IgA-HRP [SouthernBiotech, Birmingham, AL; #1030–05 and #1040–05]) was added at a 1:1000 dilution in PBS with 0.1% Tween and 5% FBS overnight at 4 °C. Plates were then washed three times in PBS with 0.1% Tween 20 and two times in PBS. For development, 100 µL of Vector NovaRED Substrate Peroxidase (Vector Laboratories, #SK-4800) was applied for 7 min. The reaction was stopped by rinsing plates with distilled water for 2 min, and plates were dried in the dark. Spots were counted and data were analyzed using ImmunoSpot software (Cellular Technology Limited).

Statistical analyses

Log-normalized IgG levels for vaccine storage analysis were analyzed by one-way ANOVA at each timepoint followed by Dunnett’s multiple comparison test comparing each stored sample to the freshly complexed positive control at each timepoint. Log-normalized IgG titers to assess prime and boost at multiple doses or with different vaccine constructs at prime and boost were assessed by mixed-effects analysis, or two-way ANOVA if there were no gaps in the data, both with and without multiple comparisons as needed. Log-normalized pseudovirus variant neutralization was compared using mixed-effects analysis or two-way ANOVA with multiple comparison correction. Bone marrow ELISpots were analyzed with one-way ANOVA with multiple comparisons. All statistical analyses were conducted using Prism 9 (GraphPad Software, San Diego, CA).

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

留言 (0)

沒有登入
gif