Viruses and infectivity titration
Simian adeno virus vectors ChAd63-METRAP and ChAdOx2-RabGP were prepared, purified and tested for quality by the Jenner Institute Viral Vector Core Facility, as previously described30,43. Viruses were dialysed against either a previously used storage buffer (10 mM Tris, 7.5% w/v sucrose, pH 7.8) or unbuffered 0.5 M (50:50) trehalose and sucrose and stored at − 80 °C as stock. ChAdOx1 expressing Photinus pyralis luciferase44 was prepared as previously described, with tangential flow filtration into unbuffered 0.5 M (50:50) trehalose and sucrose45. Typical preparations were supplied and stored at a titre of c. 1 × 1012 virus particles (VP) per mL, corresponding to c. 1 × 1010 infectious units (IU) per mL and hence a particle: infectivity ratio of c. 100.
For infectivity titration of ChAd63-METRAP and ChAdOx2-RabG, duplicate fivefold serial dilutions were prepared in complete DMEM (10% FCS, 100 U penicillin, 0.1 mg streptomycin/mL, 4 mM l-glutamine) and used to infect 80–100% confluent HEK293-TRex cells (ThermoFisher) grown in 96-well plates (BD Purecoat Amine, BD Biosciences, Europe). Infected cells were immunostained and imaged as previously described46. Wells containing 20–200 spots were used to back-calculate recovered infectious units.
Infectivity titration of ChAdOx1 luciferase was performed as above with the exception that, 24 h after infection, cells were lysed by addition of 1% Triton X-100 (Sigma) and luciferase activity measured using a Bright-Glo assay kit (Promega). Infectivity was calculated by interpolation on a standard curve prepared by serial dilution of a sample of known infectivity.
Drying, thermochallenge and reconstitution
Stock vaccines were thawed and diluted into unbuffered 0.5 M trehalose–sucrose solution. Dilution factors, final viral titres, and the trehalose:sucrose ratio varied between experiments, as stated in figure legends.
For data other than Fig. 7E, fibrous matrix was cut into 100 mm2 pieces, loaded with vaccine or sugar solutions (as described in individual figure legends) and transferred to a glove box (Coy Laboratory Products). An activated silica bed within the chamber and circulation of air through desiccation capsules containing anhydrous calcium sulphate (Drierite™, W.A. Hammond Drierite Co.), regulated by a humidity controller (Series 5000, Electro-tech Systems), was used to maintain relative humidity beneath 5%. A portable datalogger (AET-175, ATP instruments) was used for recording changes in relative humidity and temperature during the desiccation process. The temperature inside the enclosed glove box remained between 22 and 25 °C for all experiments.
For data shown in Fig. 7E, drying was performed similarly except for use of a lyophilizer rather than glovebox to provide a room temperature, atmospheric pressure low humidity environment: humidity was lowered by setting the condenser to − 70 °C, while the sample temperature was controlled by setting the shelf temperature to 20 °C.
Samples were transferred into 2 mL glass vials, stoppered and crimp sealed under dry conditions within the glove box prior to further use. Samples undergoing thermochallenge i.e. storage at elevated temperature (typically 45 °C for 1 or 4 weeks) were stored within secondary packaging (moisture barrier bags).
For experiments involving reconstitution of dried samples, this was performed by addition of phosphate buffered saline (Sigma), followed by brief vortexing of the vial (1 ± 0.5 s, three times). Virus infectivity after reconstitution was assayed as described above. Recovery of infectious virus was quantified by comparison to a control sample of the starting material, included on the same assay plate. Between the set-up of an experiment and the assay of recovered infectivity, control samples were stored at − 80 °C in aqueous buffer (under which conditions loss of infectivity is known to be negligible).
Recovery was calculated in terms of log10-fold loss in the total infectious virus content of the matrix i.e. log10-fold loss = log10(infectious units dried on matrix based on − 80 stored sample) − log10(infectious units recovered from matrix). 0.3 log10-fold loss thus implies c. 50% recovery, 0.5 log10-fold loss implies c. 30% recovery, and 1 log10-fold loss implies 10% recovery.
Karl Fischer moisture analysis
Residual moisture content in single matrix post-desiccation was determined with a Karl Fischer moisture analyser equipped with coulometer (Metrohm), and Hydranal-Coulomat titration solution (Honeywell, Fluka) in accordance with the manufacturers’ recommendations. A standard was used to calibrate the instrument performance (lactose standard 5%, MerckMillipore). Residual moisture content was expressed as a percentage (calculated from the measured mass of water as a proportion of the calculated mass of solutes in the loaded sample).
Measurement of subvisible particles
To produce the data shown in Fig. 2D, 100 mm2 pieces of glass fibre (GE Whatman Standard 14 [henceforth S14]) were cut, autoclaved at (121 °C, 15 min) and loaded with 50 µL of sugar solution prior to drying in the glove box, vialling and reconstitution as described above. Reconstituted solution from 10 vials was aspirated using a syringe with a conventional 20G needle (BD Biosciences), to produce a single pooled unfiltered sample. Reconstituted solution from a further 10 vials was aspirated using a 5-μm filter needle (BD Biosciences), to produce a single pooled filtered sample. Both pools were diluted to a final volume of 25 mL and tested for sub-visible particles by light microscopy according to Ph.Eur (2.9.19)/USP (leftlangle {788} rightrangle) by a commercial testing laboratory (Reading Scientific Services Limited).
The data shown in Fig. 7D was collected using a similar but non-pharmacopoieal in-house approach, in which unfiltered solution from reconstitution of samples dried on S14 and PVA matrices was subjected to light microscopy (Leica M205C).
Scanning electron microscopy
To produce the images shown in Figs. 2, 3, 4, 5, 6, untreated matrices (i.e. without sputter coating) were loaded onto aluminium mounts using carbon conductive tabs and imaged using a Zeiss-Evo LS15 variable pressure scanning electron microscope (SEM) equipped with variable pressure secondary electron detector (Carl Zeiss Ltd). Imaging was performed at a chamber pressure of 50 Pa air and accelerating voltage of 15 kV.
To produce the images shown in Fig. 7A,B, samples were sputter coated with gold with the coating thickness of 30 nm using a Quorum Q150RS coating unit. They were then imaged using a Jeol JSM-6610LV scanning electron microscope with 5 kV accelerating voltage and beam spot size at setting 40.
Further analyses of images such as measurement of fibre diameter were made using Image J software (freely available, National Institutes of Health). At least 40 measurements per matrix were performed, using images taken at varying magnification.
Confocal laser scanning microscopy
10 mg/mL Fluorescein-labelled lysozyme (Nanocs Inc, USA) was spiked into the 0.5 M trehalose:sucrose (50:50) solution and dried as above before examination by fluorescence confocal microscopy (DMI6000 B, Leica Microsystems, Germany).
Differential scanning calorimetry
The glass transition temperature (Tg) of sugar glass in matrices was measured immediately after drying had completed using Differential Scanning Calorimetry (DSC). The melting point of the binder in untreated glass fibre (S14) was measured using DSC (Q2000, TA instruments). The instrument was purged with dry nitrogen (50 mg/mL) continuously during sample measurement. Calibration was performed prior to measurements using a certified reference material (Indium) for temperature and heat flow accuracy.
Multiple 6-mm diameter discs were cut out of a matrix dried with 0.5 M trehalose:sucrose (50:50). Discs were weighed and, for each matrix type in turn, a total mass of 5–15 mg was loaded into Aluminium DSC pans (TA Instruments) and hermetically sealed. Samples were subjected to a temperature ramp from − 20 to 180 °C at a heating rate of 10 °C per minute. Measurements on all samples were performed in duplicates. Thermograms relating heat flow (W/g) to temperature (°C) were analysed using Trios software (TA Instruments) for identification of the glass transition (Tg) onset temperature.
For the measurement of enthalpic recovery, which manifests as an endothermic peak at the glass transition, modulated DSC (temperature modulation ± 0.50 °C every 60 s and ramp rate 3 °C/min from − 20 to 100 °C) was employed. The enthalpic recovery was estimated by linear peak integration in the thermograms plotted between nonreverse heat flow (W/g) and temperature (°C) using Universal Analysis software (TA Instruments).
Protein recovery
A 10 mg/mL solution of lysozyme (Sigma-Aldrich) was made in 0.5 M trehalose sucrose. 25 µL was loaded into each matrix in triplicates. Protein was reconstituted from the matrices after desiccation overnight and recovery was quantified using EnzChek Lysozyme Assay Kit (ThermoFisher Scientific). Fluorescence measurements were performed in triplicates for each sample and protein recovery calculated by interpolation on a standard curve, using GraphPad Prism.
Protein loss was then calculated similarly to infectivity loss (above), i.e. log10-fold loss = log10(micrograms of protein dried on matrix) − log10(micrograms of protein recovered from matrix).
Thermogravimetric analysis
Degradation temperatures of matrix constituents were measured by thermogravimetric analysis using a TGA Q500 (TA instruments). Samples loaded into a tared platinum pan just prior to measurement were subjected to a temperature ramp at 5 °C per minute from ambient to 550 °C in a flowing nitrogen atmosphere (100 mL/min). The gas was switched to air at 550 °C (100 mL/min) and heat was continued at the rate of 5 °C/min to 730 °C. Data was analysed using Universal Analysis software (TA Instruments).
Polyvinyl alcohol (PVA) extraction and Fourier transform infrared spectroscopy
A 500 mm × 27 mm piece of glass fibre (S14) was dissolved in 100 mL of ultrapure water by stirring and heating at 90 °C for 48 h. The suspension was filtered through a 0.2 µm filter and the filtrate was freeze dried (Virtis AdVantage 2.0, SP Scientific) to a white amorphous material. This residue was subjected to single reflection attenuated total reflection (MIRAacle™, Pike Technologies) Fourier transform infrared spectroscopy ATR-FTIR (Tensor 37, Bruker) equipped with nitrogen-cooled mercury cadmium telluride detector.
To obtain a spectrum of the sample, an average of 64 interferograms was collected at a resolution of 4 cm−1 in the wavelength range from 4000 to 750 cm−1 and blank subtracted.
Spectra were analysed using OPUS 6.5 software (Bruker) and compared with RMIT University’s spectral library of organic compounds generated using Spectrum 10 software (Perkin Elmer). Fingerprint spectra shown in Fig. 5D were prepared using Prism 7.0 (GraphPad Software LLC).
Dynamic light scattering (DLS)
The molecular weight of the PVA binder present on the glass fibre matrix (S14) was estimated by DLS. PVA was extracted and freeze-dried as described above. Solutions of this extracted sample and standards of known polymer size were prepared in water to the concentration of approximately 1.3 mg/mL and passed through 0.22 µm and 300 kDa molecular weight cut-off filters. The sample was then concentrated approximately five-fold from 4 mL to 0.75 mL using a 3 kDa MWCO filter (Vivaspin, GE). Zetasizer Nano ZS and DTS software (Malvern Instruments) was used for measurement of hydrodynamic diameter based on size distribution by volume. Independent duplicate preparations of all standards and samples were tested. Prism 7.0 (GraphPad Software LLC) was used to generate a standard curve plotted between measured hydrodynamic diameter and known average molecular weight (kDa) to interpolate size of the PVA in the extract.
Nuclear magnetic resonance (NMR) spectroscopy
The degree of hydrolysis of the PVA binder present on the glass fibre matrix (S14) was estimated by 1H NMR measurement as the intensity of the peak attributable to the acetyl group present in non-hydrolysed PVA. A 5 mg/mL aqueous solution of PVA extracted from the glass fibre sample (S14) was prepared in deuterium oxide. Reference spectra for PVA with varying degrees of hydrolysis were obtained by mixing > 99% hydrolysed PVA and 80% hydrolysed PVA in appropriate proportions to produce standards containing c. 0%, 2%, 5%, 10% and 20% acetyl groups, again at 5 mg/mL in deuterium oxide. An AVIII 700 instrument (Bruker Biospin) was used to generate 1H 1D spectra (employing a quantitative 1D NOESY (Nuclear Overhauser Effect Spectroscopy) presaturation sequence with recovery delay d1 = 30 s).
Application of PVA to matrices
Aqueous solutions of each type of PVA to be investigated were prepared at 10 mg/mL. Matrices were cut into approximately 100 mm2 pieces and loaded with until the matrices were saturated by the solution and air-dried overnight. The polyamide matrix (33100L) required surface modification by washing in 100% ethanol for 20 min and air drying before PVA could be loaded. Following PVA application, vaccines were dried on the matrices and thermostability assessed as described above.
Study of physical characteristics of matrices
The physical characteristics of matrices were studied using NWSPs (Nonwoven Standard Procedures) prescribed by the EDANA, the international trade association for the nonwoven industry.
Areal density was measured as ratio of mass in grams and area in square metres according to EDANA NWSP 130.1R0 (15).
Matrix thickness was measured at an applied pressure of 0.5 kPa according to EDANA NWSP 120.1.R0 (15) using a thickness tester (ProGage, Thwing-Albert Instrument Company).
Fibre length was measured based on 60 individual fibres extracted from the matrix using a Leica MD G41 optical microscope in transmission mode and Image J software.
Minimum mean and maximum pore size were obtained using a POROLUX 100 Automated Capillary Flow Porometer using Galpore liquid of surface tension 15.6 mN∙m-1. A total of five measurements were taken per sample. Porosity (ɛ) was determined by the equation ɛ = (1 − ∅) × 100% where ∅ is the volume fraction measured as a ratio of bulk matrix density to bulk fibre density.
The surface tension of aqueous sugar solution on the matrix was measured on a Kruss K100 tensiometer using the Wilhelmy plate method in which the force exerted on a suspended plate when it touches the surface of a liquid is related to the surface tension and the contact angle according to the equation (sigma = frac{{text{F}}}{{{text{L}} cdot cos theta }}) where σ = surface tension of the liquid, F = measured force, L = wetted length, and Θ = contact angle.
Wettability was evaluated using the Washburn method, again on a Kruss K100 tensiometer. The rate of mass uptake when the porous substrate (matrix) comes in to contact with a liquid was used to determine the capillary constant of the substrate by applying the Washburn equation (frac{{{text{m}}^{2} }}{{text{t}}} = frac{{{text{c}} cdot rho^{2} cos Theta }}{eta }), where m = mass, t = flow time, c = capillary constant of the nonwoven, ρ = density of the liquid, σ = surface tension of the liquid, Θ = contact angle, η = viscosity of the liquid. The capillary constant of the matrix was determined using n-hexane which has a contact angle of 0°.
In-house production of glass-fibre-based matrix
A glass fibre matrix (Fig. 6C) was custom made by a wet-lay process using commercially available glass fibre and polyvinyl alcohol aiming for similar porosity, thickness, areal density and wetting behaviour as the glass fibre sample (S14).
Wet-lay processes are similar to paper manufacturing. Glass fibre with geometry similar to the fibres used in S14 was obtained (Johns Manville, type 475, diameter 4 µm). A slurry (suspension of fibres in water) was produced and agitated in a rig connected to water supply and drainage. Upon opening of the rig’s drainage outlet, the fibres were captured and retained on a fine metal mesh at the outlet, forming a nonwoven web. The fibrous mesh was blot dried, roller pressed and further dried at 110 °C for 15 min. 1 g/L PVA solution (> 99% hydrolysed, Mw 146,000–186,000 kDa, Sigma) was then applied to each side using a spray gun before drying at 110 °C for a further 15 min.
