The use of field emission scanning electron microscopy to assess recombinant adenovirus stability

Scalable production of adenovirus vectors

Recombinant adenoviruses (AdV) are highly efficient at gene transfer for a broad spectrum of cell types and species. They became one of the vectors of choice for gene delivery and expression of foreign proteins in gene therapy and vaccination purposes. To meet the need of significant amounts of adenoviral vectors for preclinical and possibly clinical uses, scalable and reproducible production processes are required.In this chapter, we review processes used for scalable production of two types of first generation (E1-deleted) adenoviral vectors (Human and Canine) using stirred tank bioreactors. The production of adenovirus vectors using either suspension (HEK 293) or anchorage-dependent cells (MDCK-E1) are described to exemplify scalable production processes with different cell-culture types. The downstream processes will be covered in the next chapter.

Investigation on formulation and preparation of adenovirus encoding human endostatin lyophilized powders

A recombinant adenovirus encoding human endostatin gene, E10A, has finished phase II trials for head and neck cancer. However, the rigid storage temperature (-80°C) and the toxicity of glycerol in the E10A liquid preparation limited its clinical application. In this study, lyophilization was applied to develop a stable E10A lyophilized powder without glycerol that is able to maintain biological activity at 4°C and suitable for intravenous administration. The E10A lyophilized formulations composed of nontoxic and already clinically used excipients were characterized in terms of the pH change during freezing, the eutectic melting temperature (T(eu)) and the collapse temperature (T(c)). Freeze thawing tests were carried out to examine the protective effect of various excipients during freezing. Mannitol and its combinations with sucrose or inulin showed effective protection of E10A. The E10A lyophilized powders were analyzed by particle size measurement, residual humidity quantification, infectivity assay and gene expression level. An optimized formulation (formulation I1) yielded a good recovery of 76% of the starting infectivity after lyophilization and 89% of the original infectivity after storage at 4°C for 180 days. Also the gene expression capability of E10A in formulation I1 was maintained after lyophilization. In addition, it was found that the matrix of amorphous excipients, mannitol combinations with sucrose or inulin, was indispensible in protecting E10A against the stress of freezing and dehydration. Hereby, the E10A lyophilized powder with eliminated glycerol toxicity and improved stability could enhance the applicability of E10A for cancer gene therapy through intravenous administration.

Self-assembly of viral capsid protein and RNA molecules of different sizes: requirement for a specific high protein/RNA mass ratio

Virus-like particles can be formed by self-assembly of capsid protein (CP) with RNA molecules of increasing length. If the protein "insisted" on a single radius of curvature, the capsids would be identical in size, independent of RNA length. However, there would be a limit to length of the RNA, and one would not expect RNA much shorter than native viral RNA to be packaged unless multiple copies were packaged. On the other hand, if the protein did not favor predetermined capsid size, one would expect the capsid diameter to increase with increase in RNA length. Here we examine the self-assembly of CP from cowpea chlorotic mottle virus with RNA molecules ranging in length from 140 to 12,000 nucleotides (nt). Each of these RNAs is completely packaged if and only if the protein/RNA mass ratio is sufficiently high; this critical value is the same for all of the RNAs and corresponds to equal RNA and N-terminal-protein charges in the assembly mix. For RNAs much shorter in length than the 3,000 nt of the viral RNA, two or more molecules are assembled into 24- and 26-nm-diameter capsids, whereas for much longer RNAs (>4,500 nt), a single RNA molecule is shared/packaged by two or more capsids with diameters as large as 30 nm. For intermediate lengths, a single RNA is assembled into 26-nm-diameter capsids, the size associated with T=3 wild-type virus. The significance of these assembly results is discussed in relation to likely factors that maintain T=3 symmetry in vivo.

Quantitative sizing of nano/microparticles with a tunable elastomeric pore sensor

The use of a "size-tunable" polyurethane resistive pulse sensor for quantitative sizing of nano- and microparticles is presented. A linear relationship, as first suggested by Maxwell, between particle volume and change in electric resistance across the pore was observed. Particle sizes were quantified for a given size-tunable membrane, by first creating a linear calibration curve to a series of monodisperse carboxylated polystyrene particles of various diameters and then applying this curve to calculate the size of "unknown" nanoparticles. The diameters of a selection of synthetic and biological particles, being PMMA and nonfunctionalized polystyrene particles, along with biological nanoparticles (adenovirus) were calculated using this methodology. Calculated particle diameters and coefficients of variation were shown to be in good agreement with both transmission electron microscopy and dynamic light scattering results.

Controlled inactivation of recombinant viruses with vitamin B2

Inactivated viruses are important tools for vaccine development and gene transfer. 8-Methoxypsoralen (8-MOP) and long-wavelength ultraviolet irradiation (LWUVI) inactivates many viruses. Toxicity limits its use in animals and humans. Toxicological and photosensitizing properties of riboflavin make it suitable for virus inactivation in preparations for biological use. Viruses expressing beta-galactosidase were mixed with either 8-MOP (1.5mM) or riboflavin (50 microM) and exposed to LWUVI (365 nm) for 2 h. Virus activity was determined by limiting dilution. The half-life of the adenovirus preparation treated with 8-MOP was 8.28 ns(-1) and 36.5 ns(-1) after treatment with riboflavin. Despite the difference in half-life, both preparations were completely inactivated within 45 min. In contrast, the half-lives for adeno-associated virus (AAV) preparations were similar (63 ns(-1) 8-MOP vs. 67 ns(-1) riboflavin). Each AAV preparation was fully inactivated within 90 min. The half-life of lentivirus was 193.4 ns(-1) after treatment with 8-MOP and 208 ns(-1) after exposure to riboflavin. Virus treated with riboflavin was inactivated within 20 min. Virus exposed to 8-MOP was inactivated in 90 min. DNA and RNA viruses can be inactivated by riboflavin and LWUVI and used in physiological systems sensitive to other photochemicals.

A simple macroscopic model for the diffusion and adsorption kinetics of r-adenovirus

The diffusion of viruses toward cells is a limiting step of the infection process. To be modeled correctly, this step must be evaluated in combination with the adsorption of the virus to the cell surface, which is a rapid but reversible step. In this paper, the recombinant adenovirus (rAd) diffusion and its adsorption to 293S cells in suspension were both measured and modeled. First, equilibrium experiments permitted to determine the number of receptors on the surface of 293S (R(T) = 3,500 cell(-1)) and the association constant (K(A) = 1.9 x 10(11) M(-1)) for rAd on these cells based on a simple monovalent adsorption model. Non-specific binding of the virus to the cell surface was not found to be significant. Second, total virus particle degradation rates between 5.2 x 10(-3) and 4.0 x 10(-2) min(-1) were measured at 37 degrees C in culture medium, but no significant virus degradation was observed at 4 degrees C. Third, free viral particle disappearance rates from a mixed suspension of virus and cells were measured at different virus concentrations. Experimental data were compared to a phenomenological dynamic model comprising both the diffusion and the adsorption steps. The diffusion to adsorption ratio, a fitted parameter, confirmed that the contact process of a virus with a cell is indeed diffusion controlled. However, the characteristic diffusion time constants obtained, based on a reversible adsorption model, were eightfolds smaller than those reported in the literature, based on diffusion models that assume irreversible adsorption.

Biophysical characterization of influenza virus subpopulations using field flow fractionation and multiangle light scattering: Correlation of particle counts, size distribution and infectivity

Adequate biophysical characterization of influenza virions is important for vaccine development. The influenza virus vaccines are produced from the allantoic fluid of developing chicken embryos. The process of viral replication produces a heterogeneous mixture of infectious and non-infectious viral particles with varying states of aggregation. The study of the relative distribution and behavior of different subpopulations and their inter-correlation can assist in the development of a robust process for a live virus vaccine. This report describes a field flow fractionation and multiangle light scattering (FFF-MALS) method optimized for the analysis of size distribution and total particle counts. The FFF-MALS method was compared with several other methods such as transmission electron microscopy (TEM), atomic force microscopy (AFM), size exclusion chromatography followed by MALS (SEC-MALS), quantitative reverse transcription polymerase chain reaction (RT Q-PCR), median tissue culture dose (TCID(50)), and the fluorescent focus assay (FFA). The correlation between the various methods for determining total particle counts, infectivity and size distribution is reported. The pros and cons of each of the analytical methods are discussed.

Monitoring the homogeneity of adenovirus preparations (a gene therapy delivery system) using analytical ultracentrifugation

This study explores the capability of modern analytical ultracentrifugation (AUC) to characterize the homogeneity, under product formulation conditions, of preparations of adenovirus vectors used in gene therapy and to assess the lot-to-lot consistency of this unique drug product. We demonstrate that a single sedimentation velocity run on an adenovirus sample can detect and accurately quantify a number of different forms of virus particles and subvirus particles. These forms include (a) intact virus monomer particles, (b) virus aggregates, (c) empty capsids (ECs), and (d) smaller assembly intermediates or subparticles formed during normal or aberrant virus assembly (or as a result of damage to the intact adenovirus or EC material during all phases of virus production). This information, which is collected on adenovirus samples under the exact formulation conditions that exist in the adenovirus vial, is obtained by direct boundary modeling of the AUC data generated from refractometric and/or UV detection systems using the computer program SEDFIT developed by Peter Schuck. Although both detectors are useful, refractometric detection using the Rayleigh interferometer offers a key advantage for providing accurate concentration information due to the similar response factors for both protein and DNA and its insensitivity to light scattering effects. Additional AUC data obtained from analytical band sedimentation velocity and density gradient sedimentation equilibrium experiments in CsCl with UV detection were also generated. These results further support conclusions concerning the solution properties of adenovirus, the identity of the different virus species, and the overall capability of boundary sedimentation velocity analysis.

Analysis of baculovirus aggregates using flow cytometry

Aggregation of viral particles represents a significant problem for baculoviral stock processing and storage. Aggregation may also affect the results of viral particle counting. A method using flow cytometry was previously developed in our lab to measure the concentration of baculovirus particles produced in insect cell cultures. In the present study, the use of the flow cytometry method was extended to the detection of baculovirus aggregates. Flow cytometry analysis of freshly prepared baculovirus stocks, stained with SYBR Green, generally exhibited a single unimodal distribution; while, baculovirus stocks stored at 4 degrees C for a few months exhibited a bimodal distribution of the fluorescent intensity signal. The bimodal distribution was associated with a decrease in the size of the original viral population and an emergence of a new viral population with a high fluorescence intensity. Treatment of these samples with an endonuclease (Benzonase) confirmed that the new population observed in the flow cytometry analysis is not free cellular DNA. Filtration through 0.22 and 0.45 microm membranes of the stored samples prior to flow cytometry analysis confirmed that the high fluorescence intensity population involved particles larger than a single baculovirus. Exposing freshly amplified baculovirus stocks with a unimodal distribution to a pH of 5.3, a condition known to induce aggregation, showed the emergence of a second population with a bimodal distribution. These results suggest that flow cytometry analysis could be used to detect baculovirus aggregates. The aggregates were associated with high fluorescence intensity populations and the mean green fluorescence intensity of these populations could be used as an indicator of the mean aggregate size.

Use of an automated image processing program to quantify recombinant adenovirus particles

Electron microscopy has a pivotal role as an analytical tool in pharmaceutical research. However, digital image data have proven to be too large for efficient quantitative analysis. We describe here the development and application of an automated image processing (AIP) program that rapidly quantifies shape measurements of recombinant adenovirus (rAd) obtained from digitized field emission scanning electron microscope (FESEM) images. The program was written using the macro-recording features within Image-Pro Plus software. The macro program, which is linked to a Microsoft Excel spreadsheet, consists of a series of subroutines designed to automatically measure rAd vector objects from the FESEM images. The application and utility of this macro program has enabled us to rapidly and efficiently analyze very large data sets of rAd samples while minimizing operator time.