Gene Therapy Analytics and Manufacturing, a recent conference organized by the Cambridge Healthtech Institute and held in San Diego, described itself as a way to take “an in-depth look at the challenges facing the formulation, characterization, analysis, and scale-up of gene therapies.” The conference lived up to its promise, and it was, as this article maintains, particularly strong in its coverage of viral vectors. In the following text, five presenters from the conference offer their views on viral vector–based gene therapies. (Readers may notice that this conference was, along with the Cell Therapy Analytics and Manufacturing conference, part of theCell and Gene Therapy “pipeline” at PepTalk: The Protein Science Week.)
“Viral vectors are complex structures of several megadaltons, consisting of nucleic acid and a protein shell,” said Klaus Richter, PhD, group leader and analytical ultracentrifugation (AUC) expert at Coriolis Pharma. “In some viral species, a membrane is additionally present with proteins embedded in it.”
To make a viral-based therapeutic, all the parts of the virus must be functional. “This is very different from small molecules or classical biopharmaceuticals, where the administered drug substance already contains the actual active agent,” noted Richter. “In addition, the whole structure needs to be stable because the viral vectors need to maintain their ability to infect cells.” These requirements make quality control a very complex activity.
To assess the quality of a virus, such as an adeno-associated virus (AAV), explained Richter, scientists can use sedimentation velocity (SV)-AUC. For example, SV-AUC can be used to analyze the DNA in the protein shell, the percentage of empty or partially assembled viruses in an AAV preparation, and the percentage of aggregated AAVs. Richter added, “These are critical parameters to confirm the quality of the final product and to assess its stability and shelf lifetime.”
Manufacturers can add SV-AUC to their existing process, Richter observed, “provided that enough time is allowed for performing the experiment and data analysis, which together can be a few hours using optimized procedures for AUC analysis and data handling.” According to Richter, this method can be used in development to detect viral particles at particular steps in manufacturing, for example, to evaluate enrichment or on a final product to determine if the “critical parameters for quality control are met.”
Swapping out silver staining
The detection of AAV capsid proteins can be used to assess the quality of the particles. Often, scientists perform this detection with silver staining of sodium dodecyl sulfate-poly-acrylamide gel electrophoresis (SDS-PAGE), but scientists may try alternative techniques. For example, as a presentation from PerkinElmer demonstrated, scientists may use capillary electrophoresis-SDS (CE-SDS).
“The AAV capsid is composed of three proteins, which are designated as VP1, VP2, and VP3,” said James Atwood, PhD, general manager of applied genomics at PerkinElmer. “The proteins differ in size with reported values of 87, 73, and 62 kilodaltons for VP1, VP2, and VP3, respectively.” In producing a recombinant AAV, a manufacturer must analyze the distribution, size, and ratio of the viral capsid proteins. Atwood noted that this quality- control step is used “to validate capsid protein expression, to screen for protein contaminants, and to verify assembly of the capsid in the appropriate stoichiometry.”
At the conference, PerkinElmer discussed a microfluidic CE-SDS method for characterizing capsid proteins. “AAVs are diluted in a nonreducing sample buffer and heated to dissociate the viral capsid into individual protein constituents,” Atwood explained afterward. “Standard protocols of the LabChip ProteinEXact Assay are then followed for preparing the LabChip assay with a gel and dye solution.” The samples are analyzed on the LabChip GXII Touch instrument, which automatically analyzes the AAV protein capsid size for each sample.
“A typical silver staining SDS-PAGE experiment takes approximately 60 to 120 minutes, but it only takes approximately 60 seconds for each sample run on the LabChip system,” Atwood asserted. “The LabChip ProteinEXact Assay can detect protein concentrations as low as 0.2 nanograms per microliter, which is about an order of magnitude lower than silver staining SDS-PAGE.”
Overall, Atwood described this approach as “a fully automated and validated solution delivering quantitative and reproducible digitized results to monitor the quality of AAV particles.”
Making more use of MVM
Analyzing a biopharmaceutical process for its ability to clear virus, said David Cetlin, founder and CEO of MockV Solutions, “requires spiking with a live infectious agent— virus.” He added that contract research organizations (CROs) “are typically the only avenue for conducting viral clearance work,” which is often so costly and complex that it is limited to late-stage validation. So, Cetlin and his colleagues created the first viral clearance prediction kit.
In this kit, Cetlin explained, a noninfectious “mock virus particle” (MVP) replaces the usual live, infectious virus. He added that the MVP “mimics the physicochemical characteristics of MVM (minute virus of mice), a parvovirus commonly used for spiking studies.” This so-called MVM-MVP system offers a range of benefits, starting with the ability to use it in biosafety level 1 (BSL-1) conditions.
“Whereas live MVM requires a BSL-2 laboratory, MVM-MVP can be used on any common benchtop, thereby enabling studies to be performed on site, as opposed to on location at a CRO,” Cetlin asserted. “The MVM-MVP kit, which contains a stock solution of MVM- MVP, is also much cheaper than contracting a CRO-led study.” At $4000 per kit, which can run about 10 small-scale studies, this method only costs about $400 per experiment. Cetlin added, “Sample and data analysis can also be conducted in a matter of hours, as opposed to the weeks that would be required for a plaque assay readout at a CRO.”
In his talk, Cetlin described the physicochemical similarities between MVM and MVM-MVP, explained how to conduct an experiment utilizing the MVM-MVP, and showed data sets from three collaborations in which MVM-MVP was used in process development/characterization efforts. “The data not only demonstrated that MVM- MVP can predict MVM’s removal within an AAV process, but also demonstrated that Thermo Fisher’s AAVX resin can distinguish between AAV and MVM, which are both parvoviruses,” he stated. An MVM-MVPstudy, he concluded, can serve “as a good viral clearance reduction step in an AAV process.”
Learning more from light scattering
Developing gene- and cell-based therapies requires methods for characterizing and quantifying AAV and other viral vectors. Michelle Chen, PhD, vice president of analytical services at Wyatt Technology, discussed how light scattering can be used in those processes with three tools: dynamic light scattering (DLS) for fast screening of viral vector size distribution and particle concentration; size-exclusion chromatography (SEC) combined with multiangle light scattering (MALS) for some crucial AAV quality attribute measurements; and field flow fractionation (FFF) combined with MALS for characterizing AAV aggregates, large viral vectors, and other bionanoparticles.
“Our SEC method, coupled with UV, MALS, and differential refractive index (dRI) detectors, is used to measure three critical quality attributes of an AAV product—total capsid concentration, capsid content, and degree of aggregation—within one single SEC run,” Chen said. She noted that this method requires no calibration or label, runs in 30 minutes or less, and is fully automated. Chen added that it is “robust and has the potential to be implemented in AAV manufacturing, quality control, and quality assurance.”
Using the combination of different detectors in Wyatt’s protein conjugate-analysis method provides the “molecular weight and eluted mass for the protein capsid and DNA,” Chen detailed. “These quantities can then be readily converted to capsid-particle concentration and empty-to-full ratio.” Plus, she noted that “the data from the same run can be used to quantify aggregation, detect impurities, and measure particle size.”
Results indicate the benefits of this method. “Compared to the other techniques,” Chen asserted, “the SEC-MALS tool provides an orthogonal and complementary approach with easy implementation and validation throughout the AAV production process.”
Looking at lentiviral vectors
Severe combined immunodeficiency syndrome (SCID) arises from a genetic defect that prevents the development of the adaptive immune system and leads to “a wide range
of life-threatening infections like pneumonia, meningitis, and sepsis,” said Alfred Luitjens, director cell technology, Batavia Biosciences. “Babies with SCID die within their first year.”
SCID is associated with 20 or more genes. RAG1-SCID, a common form of SCID associated with a RAG1 variant, is the focus of a consortium led by the Leiden University Medical Center (LUMC). As a partner in the consortium, Batavia is developing good manufacturing practices for the production of a lentiviral vector–based therapy.
“One of the next steps in this project will involve bringing the process to commercial manufacturing scale with the aim to treat all babies with RAG1-SCID worldwide,” Luitjens said. For this process, Batavia developed a model that uses Univercells’ scale-X bioreactors and NevoLine biomanufacturing system. The consortium hopes to develop “an autologous transplant system,” Luitjens noted. “The patients’ own blood-forming stem cells will be collected and sent to the transduction site, the LUMC in the Netherlands. Then the modified stem cells will be returned to the clinical centers and transplanted into the patients.”
When asked to summarize the process, Luitjens said it involves “getting the blood- forming stem cells in good and consistent condition at the LUMC, performing the transduction, and subsequently transporting the genetically modified stem cells back to the treatment center.” If the results of the Phase I study are encouraging, the process will be scaled up with the scale-X bioreactor.