Engineered CHO cells as a novel AAV production platform for gene therapy delivery.

The Herpes simplex virus (HSV)-based platform for production of recombinant adeno-associated viral vectors (rAAVs) yields higher titers and increased percentage of full capsids when compared to the triple transient transfection (TTT) method. However, this platform currently faces two major challenges. The first challenge is the reliance on commercial media, sometimes supplemented with serum, leading to costly manufacturing and a high risk for introduction of adventitious agents. The second challenge is that the production of HSV-1 relies on adherent complementing Vero cells (V27), making it difficult to scale up. We engineered serum-free-adapted CHO cells expressing key HSV-1 entry receptors, HVEM and/or Nectin-1 to address the first challenge. Using high-throughput cloning methods, we successfully selected a HVEM receptor-expressing clone (CHO-HV-C1) that yields 1.62 × 109, 2.51 × 109, and 4.07 × 109 viral genome copies/mL with rAAV6.2-GFP, rAAV8-GFP, and rAAV9-GFP vectors respectively, within 24 h post rHSV-1 co-infection. Moreover, CHO-HV-C1-derived rAAVs had comparable in vitro transduction, infectivity, and biodistribution titers to those produced by TTT. The second challenge was addressed via engineering CHO-HV-C1 cells to express HSV-1 CP27. These cells successfully produced rHSV-1 vectors, but with significantly lower titers than V27 cells. Taken together, the CHO/HSV system provides a novel, scalable, reduced cost, serum-free AAV manufacturing platform.

Process economics evaluation and optimization of adeno-associated virus downstream processing.

Adeno-associated virus (AAV) manufacturing has traditionally focused upon lab-scale techniques to culture and purify vector products, leading to limitations in production capacity. The tool presented in this paper assesses the feasibility of using non-scalable technologies at high AAV demands and identifies optimal flowsheets at large-scale that meet both cost and purity targets. The decisional tool comprises (a) a detailed process economics model with the relevant mass balance, sizing, and costing equations for AAV upstream and downstream technologies, (b) a built-in Monte Carlo simulation to assess uncertainties, and (c) a brute-force optimization algorithm for rapid investigation into the optimal purification combinations. The results overall highlighted that switching to more scalable upstream and downstream processing alternatives is economically advantageous. The base case analysis showed the cost and robustness advantages of utilizing suspension cell culture over adherent, as well as a fully chromatographic purification platform over batch ultracentrifugation. Expanding the set of purification options available gave insights into the optimal combination to satisfy both cost and purity targets. As the purity target increased, the optimal polishing solution moved from the non-capsid purifying multimodal chromatography to anion-exchange chromatography or continuous ultracentrifugation.

Accelerated cell culture process development and characterization for cilgavimab/tixagevimab (AZD7442) for the prevention and treatment of COVID-19.

The global COVID-19 pandemic ignited an unprecedented race to develop vaccines and antibody therapeutics. AstraZeneca's pursuit to provide AZD7442 (EVUSHELD), two long-acting, SARS-CoV-2 spike receptor binding domain-specific neutralizing monoclonal antibodies, to individuals at risk on highly accelerated timelines challenged our traditional ways of process development and spurred the rapid adoption of novel approaches. Conventional upstream development processes were replaced by agile strategies that combined technological advances and highly accelerated workflows. With calculated business risks and close cross-functional collaborations, this process paved the way for hyper accelerated antibody development from discovery through manufacturing, process validation, emergency use authorization filing, and global regulatory approvals. The result was initiation of commercial manufacturing at a contract manufacturing organization less than 6 months from the selection of cilgavimab and tixagevimab-a process that historically has taken close to 10 years.

Chemically Defined, High-Density Insect Cell-Based Expression System for Scalable AAV Vector Production.

The recombinant adeno-associated virus (AAV) vector is one of the most utilized viral vectors in gene therapy due to its robust, long-term in vivo transgene expression and low toxicity. One major hurdle for clinical AAV applications is large-scale manufacturing. In this regard, the baculovirus-based AAV production system is highly attractive due to its scalability and predictable biosafety. Here, we describe a simple method to improve the baculovirus-based AAV production using the ExpiSf Baculovirus Expression System with a chemically defined medium for suspension culture of high-density ExpiSf9 cells. Baculovirus-infected ExpiSf9 cells produced up to 5 × 1011 genome copies of highly purified AAV vectors per 1 mL of suspension culture, which is up to a 19-fold higher yield than the titers we obtained from the conventional Sf9 cell-based system. When mice were administered the same dose of AAV vectors, we saw comparable transduction efficiency and biodistributions between the vectors made in ExpiSf9 and Sf9 cells. Thus, the ExpiSf Baculovirus Expression System would support facile and scalable AAV manufacturing amenable for preclinical and clinical applications.

Developing an Anion Exchange Chromatography Assay for Determining Empty and Full Capsid Contents in AAV6.2.

Adeno-associated virus (AAV) vectors are clinically proven gene delivery vehicles that are attracting an increasing amount of attention. Non-genome-containing empty AAV capsids are by-products during AAV production that have been reported to potentially impact AAV product safety and efficacy. Therefore, the presence and amount of empty AAV capsids need to be characterized during process development. Multiple methods have been reported to characterize empty AAV capsid levels, including transmission electron microscopy (TEM), analytical ultracentrifugation (AUC), charge detection mass spectrometry (CDMS), UV spectrophotometry, and measuring capsid and genome copies by ELISA and qPCR. However, these methods may lack adequate accuracy and precision or be challenging to transfer to a quality control (QC) lab due to the difficulty of implementation. In this study, we used AAV serotype 6.2 (AAV6.2) as an example to show the development of a QC-friendly anion exchange chromatography (AEX) assay for the determination of empty and full capsid percentages. The reported assay requires several microliters of material with a minimum titer of 5 × 1011 vg/mL, and it can detect the presence of as low as 2.9% empty capsids in AAV6.2 samples. Additionally, the method is easy to deploy, can be automated, and has been successfully implemented to support testing of various in-process and release samples.

Production and characterization of active recombinant human factor II with consistent sialylation.

Coagulation factor II (prothrombin; FII) is the pre-proteolyzed precursor to thrombin in the coagulation cascade. It has 10 sites of gamma-carboxylation, which are required for its bioactivity, and is N-glycosylated at three of four putative sites. Production of recombinant human FII (rhFII) using a platform fed-batch process designed for monoclonal antibody production resulted in low levels of gamma-carboxylation and sialylation. There have not been any prior reports of successful process development and clinical manufacture of rhFII with optimal, consistent gamma-carboxylation and sialylation. In order to develop such a fed-batch process, various process parameters were evaluated to determine their impact on product quality. Process temperature and temperature shift timing were important for both sialic acid level and gamma-carboxyglutamate (Gla) level. In addition, vitamin K concentration and the type of surfactant used for preparation of vitamin K stock solution were also important for gamma carboxylation. A fed-batch study performed with various medium additives known to be involved in the N-glycosylation pathway, such as N-acetyl-d-mannosamine (ManNAc), galactose (Gal), dexamethasone, and manganese sulfate, increased the level of sialylation and enabled the elucidation of some potential bottlenecks in the sialylation pathway. The optimized process based on these studies yielded a reduction in the level of missing Gla by 0.4 moles per mole of rhFII in cell culture and a nearly threefold increase in sialic acid level. The process was successfully implemented at the 2000 L scale where a high Gla level and sialylation levels were achieved in all GMP lots. Biotechnol. Bioeng. 2017;114: 1991-2000. © 2017 Wiley Periodicals, Inc.

A bicistronic vector with destabilized mRNA secondary structure yields scalable higher titer expression of human neurturin in E. coli.

Human neurturin (NTN) is a cystine knot growth factor with potential therapeutic use in diseases such as Parkinson's and diabetes. Scalable high titer production of native NTN is particularly challenging because of the cystine knot structure which consists of an embedded ring comprised of at least three disulfide bonds. We sought to pursue enhanced scalable production of NTN in Escherichia coli. Our initial efforts focused on codon optimization of the first two codons following AUG, but these studies resulted in only a marginal increase in NTN expression. Therefore, we pursued an alternative strategy of using a bicistronic vector for NTN expression designed to reduce mRNA secondary structure to achieve increased ribosome binding and re-initiation. The first cistron was designed to prevent sequestration of the translation initiation region in a secondary conformation. The second cistron, which contained the NTN coding sequence itself, was engineered to disrupt double bonded base pairs and destabilize the secondary structure for ribosome re-initiation. The ensemble approach of reducing NTN's mRNA secondary structure and using the bicistronic vector had an additive effect resulting in significantly increased NTN expression. Our strain selection studies were conducted in a miniaturized bioreactor. An optimized strain was selected and scaled up to a 100 L fermentor, which yielded an inclusion body titer of 2 g/L. The inclusion bodies were refolded to yield active NTN. We believe that our strategy is applicable to other candidate proteins that are difficult-to-express due to stable mRNA secondary structures. Biotechnol. Bioeng. 2017;114: 1753-1761. © 2017 Wiley Periodicals, Inc.

Development of a CHO-Based Cell-Free Platform for Synthesis of Active Monoclonal Antibodies.

Chinese Hamster Ovary (CHO) cells are routinely optimized to stably express monoclonal antibodies (mAbs) at high titers. At the early stages of lead isolation and optimization, hundreds of sequences for the target protein of interest are screened. Typically, cell-based transient expression technology platforms are used for expression screening, but these can be time- and resource-intensive. Here, we have developed a cell-free protein synthesis (CFPS) platform utilizing a commercially available CHO extract for the rapid in vitro synthesis of active, aglycosylated mAbs. Specifically, we optimized reaction conditions to maximize protein yields, established an oxidizing environment to enable disulfide bond formation, and demonstrated the importance of temporal addition of heavy chain and light chain plasmids for intact mAb production. Using our optimized platform, we demonstrate for the first time to our knowledge the cell-free synthesis of biologically active, intact mAb at >100 mg/L using a eukaryotic-based extract. We then explored the utility of our system as a tool for ranking yields of candidate antibodies. Unlike stable or transient transfection-based screening, which requires a minimum of 7 days for setup and execution, results using our CHO-based CFPS platform are attained within 2 days and it is well-suited for automation. Further development would provide a tool for rapid, high-throughput prediction of mAb expression ranking to accelerate design-build-test cycles required for antibody expression and engineering. Looking forward, the CHO-based CFPS platform could facilitate the synthesis of toxic proteins as well.

Liquid-liquid phase separation causes high turbidity and pressure during low pH elution process in Protein A chromatography.

Turbid elution pools and high column back pressure are common during elution of monoclonal antibodies (mAbs) by acidic pH in Protein A chromatography. This phenomenon has been historically attributed to acid-induced precipitation of incorrectly folded or pH-sensitive mAbs and host cell proteins (HCPs). In this work, we propose a new mechanism that may account for some observations of elution turbidity in Protein A chromatography. We report several examples of turbidity and high column back pressure occurring transiently under a short course of neutral conditions during Protein A elution. A systematic study of three mAbs displaying this behavior revealed phase separation characterized by liquid drops under certain conditions including neutral pH, low ionic strength, and high protein concentration. These liquid droplets caused solution turbidity and exhibited extremely high viscosity, resulting in high column back pressure. We found out that the droplets were formed through liquid-liquid phase separation (LLPS) as a result of protein self-association. We also found multiple factors, including pH, temperature, ionic strength, and protein concentration can affect LLPS behaviors. Careful selection of process parameters during protein A elution, including temperature, flow rate, buffer, and salt can inhibit formation of a dense liquid phase, reducing both turbidity (by 90%) and column back pressure (below 20 pounds per square inch). These findings provide both mechanistic insight and practical mitigation strategies for Protein A chromatography induced LLPS.

A novel cholesterol/lipid delivery system for murine myeloma cell lines.

Murine myeloma NS0 cells are cholesterol-dependent auxotrophs and require externally provided cholesterol for sustained growth. Traditionally, cholesterol is provided to these cells by supplementing cell culture media with a concentrated solution of cholesterol and other water insoluble components dissolved in 200-proof ethanol. However, the solubility of cholesterol in ethanol is limited, and for processes requiring large amounts of cholesterol, the consequential increase in added ethanol may negatively impact cell growth. Additionally, the flammability of 200-proof ethanol may restrict the preparation scale and storage volumes at a large-scale facility, thus resulting in a more complex preparation procedure due to safety guidelines. This study proposes 1-propanol as an alternative solvent, which can dissolve up to 40 g L-1 of cholesterol along with other water insoluble components, as compared to ethanol, which can dissolve up to 10 g L-1 of the same. A concentrated formulation simplifies the preparation method and ameliorates the procedural and operational challenges, as well as reduces the total amount of alcohol added to a cell culture by ∼80% when compared to the ethanolic solution, to deliver the same amount of cholesterol, thereby significantly minimizing alcohol exposure to the cells and mitigating the fire hazards at a large-scale facility. © 2017 American Institute of Chemical Engineers Biotechnol. Prog., 33:795-803, 2017.

Selected amino acids protect hybridoma and CHO cells from elevated carbon dioxide and osmolality.

Elevated pCO(2) inhibits cell growth. This growth inhibition is accompanied by a decrease in intracellular pH (pHi), as well as a decrease in glycolysis. Elevated concentrations (mM) of some amino acids have been shown by others to protect cells exposed to two very different environmental stresses: nutrient starvation and hyperosmolality. The fact that many of the amino acids shown to have protective effects against other stresses are transported into the cell through a pHi-sensitive transporter led us to study the possibility of using these amino acids as protective agents under elevated pCO(2). Screening experiments using 5, 15, and 25 mM of each amino acid showed that not all amino acids that protect cells from hyperosmolality protect them from elevated pCO(2). Glycine betaine and glycine were chosen for further characterization in both hybridoma and CHO cells. Asparagine and threonine were also tested in hybridoma and CHO cells, respectively. All amino acids tested under 195 mm Hg pCO(2)/435 mOsm/kg (50% growth inhibition) restored the specific growth rate (mu) in hybridoma cells to that observed under control conditions (40 mm Hg/320 mOsm/kg). Addition of each amino acid resulted in an increase in the consumption rate and intracellular accumulation of that amino acid. In CHO cells, glycine betaine also restored mu to control values, while glycine and threonine partially restored mu. In hybridoma cells, the higher specific antibody productivity obtained at elevated pCO(2) was maintained with the lowest amino acid concentration (5 mM). Productivity decreased toward control values with increasing amino acid concentrations. Elevated pCO(2) decreased the specific tPA productivity in the CHO cell line studied. Only glycine betaine resulted in a 20% increase in productivity at 195 mm Hg/435 mOsm/kg. With the exception of glycine betaine in hybridoma cells, amino acids did not mitigate the associated pHi decrease of at least 0.2 pH units at 195 mm Hg/435 mOsm/kg. pHi in hybridoma cells under elevated pCO(2) in the presence of glycine betaine was about 0.1 pH units below that of control. Amino acids had no effect on the cell size response of hybridoma cells, while they partially offset the increase in CHO cell size at elevated pCO(2). Glycine betaine, asparagine, and glycine increased the specific glucose consumption rate observed at 195 mm Hg/435 mOsm/kg (50% of control) to values greater than 70% of control in hybridoma cells. In CHO cells, only glycine betaine increased q(glc) (by 20%) under elevated pCO(2). All amino acids tested improved the cell yield from glutamine at 195 mm Hg/435 mOsm/kg in both cell lines.

Hyperosmotic stress and elevated pCO2 alter monoclonal antibody charge distribution and monosaccharide content.

Medium osmolality increases with pCO2 at constant pH. Elevated pCO2 and osmolality inhibit hybridoma growth to similar extents in both serum-containing and serum-free media. The combination of osmolality and elevated pCO2 synergizes to negatively impact cell growth. IgG2a glycosylation by hybridoma cells was evaluated under elevated pCO2 (to 250 mmHg pCO2) and/or osmolality (to 476 mOsm/kg). IgG2a site occupancy did not change significantly under any of the conditions studied, which is consistent with the robust glycosylation of other antibodies produced under various environmental stresses. However, changes were observed in the IgG2a charge distribution. Changes in the isoelectric point (pI) were greater under hyperosmotic stress, increasing by 0.32 and 0.41 pH units at 435 mOsm/kg in serum-containing and serum-free medium, respectively. Hyperosmotic stress also resulted in a concomitant increase in the heterogeneity of the charge distribution. The mean pI in serum-containing medium decreased by 0.16 pH units at 250 mmHg pCO2 when osmolality was controlled at 320 mOsm/kg but increased by 0.20 pH units when the osmolality increased with pCO2 (195 mmHg pCO2-435 mOsm/kg). In serum-free medium, elevated pCO2 did not alter pI, regardless of medium osmolality. In contrast to elevated osmolality at control pCO2, elevated pCO2 did not significantly alter the IgG2a charge heterogeneity under any of the conditions studied. The IgG2a was not sialylated, so sialylation changes were not responsible for changes in the charge distribution. IgG2a galactose content decreased with elevated osmolality, as a result of either elevated NaHCO3 or NaCl. However, when osmolality was controlled at elevated pCO2, the galactose content tended to increase. The mannose content decreased with increasing stress, while the fucose content remained relatively unchanged. It is likely that the observed increases in the pI of murine IgG2a were due to increased organellar pH, which is reflected by increased specific beta-galactosidase activity in the supernatant.

Effects of osmoprotectant compounds on NCAM polysialylation under hyperosmotic stress and elevated pCO(2).

Elevated osmolality and pCO(2) have been shown to alter sialylation in a protein-specific manner. In Chinese hamster ovary (CHO)MT2-l-8 cells, tPA sialylation changed only slightly from 40 to 250 mm Hg pCO(2), whereas neural cell adhesion molecule polysialic acid (NCAM PSA) content decreased by up to 70% at 250 mm Hg pCO(2), pH 7.2. NCAM PSA content also decreased with increasing NaCl or NH(4)Cl concentration. This suggests that PSA content is a sensitive indicator of conditions that may alter glycosylation. Amino acids and their derivatives have been used to protect hybridoma and CHO cell growth under hyperosmotic stress. We examined the impact of osmoprotectants on NCAM PSA content in CHO MT2-1-8 cells under hyperosmolality (up to 545 mOsm/kg) and at 195 and 250 mm Hg pCO(2). NCAM PSA content at 545 mOsm/kg was at least two-fold greater in the presence of glycine betaine or L-proline compared to that without osmoprotectant. Surprisingly, in the presence of 20 mM glycine betaine, PSA levels were 50-60% of the control level for osmolalities ranging from 320 to 545 mOsm/kg. Thus, glycine betaine inhibits NCAM polysialylation at osmolalities below 435 mOsm/kg and is beneficial at higher osmolalities. In contrast to glycine betaine, L-proline increased PSA content by 25-120% relative to the unprotected culture at < or =545 mOsm/kg. The decrease in NCAM PSA levels of CHO MT2-1-8 cells cultured at 195 mm Hg pCO(2)-435 mOsm/kg was not mitigated by the presence of 25 mM glycine betaine, glycine, or L-threonine, even though all of these compounds enhanced cell growth. At 250 mm Hg pCO(2), all osmoprotectants tested (20 mM L-threonine, L-proline, glycine, or glycine betaine) increased NCAM polysialylation, with 20 mM glycine betaine restoring NCAM PSA to near control levels. Thus, osmoprotectants may (partially) offset changes in glycosylation, as well as the inhibition of growth, in cells under environmental stress. Supernatant beta-galactosidase levels, which increase upon alkalization of acidic organelles, did not differ significantly under elevated pCO(2) and hyperosmolality from that at control conditions.

Characterization of hybridoma cell responses to elevated pCO(2) and osmolality: intracellular pH, cell size, apoptosis, and metabolism.

CO(2) partial pressure (pCO(2)) in industrial cell culture reactors may reach 150-200 mm Hg, which can significantly inhibit cell growth and recombinant protein production. The inhibitory effects of elevated pCO(2) at constant pH are due to a combination of the increases in pCO(2) and [HCO(-) (3)], per se, and the associated increase in osmolality. To decouple the effects of pCO(2) and osmolality, low-salt basal media have been used to compensate for this associated increase in osmolality. Under control conditions (40 mm Hg-320 mOsm/kg), hybridoma cell growth and metabolism was similar in DMEM:F12 with 2% fetal bovine serum and serum-free HB GRO. In both media, pCO(2) and osmolality made dose-dependent contributions to the inhibition of hybridoma cell growth and synergized to more extensively inhibit growth when combined. Elevated osmolality was associated with increased apoptosis. In contrast, elevated pCO(2) did not increase apoptotic cell death. Specific antibody production also increased with osmolality although not with pCO(2). In an effort to understand the mechanisms through which elevated pCO(2) and osmolality affect hybridoma cells, glucose metabolism, glutamine metabolism, intracellular pH (pHi), and cell size were monitored in batch cultures. Elevated pCO(2) (with or without osmolality compensation) inhibited glycolysis in a dose-dependent fashion in both media. Osmolality had little effect on glycolysis. On the other hand, elevated pCO(2) alone had no effect on glutamine metabolism, whereas elevated osmolality increased glutamine uptake. Hybridoma mean pHi was approximately 0.2 pH units lower than control at 140 mm Hg pCO(2) (with or without osmolality compensation) but further increases in pCO(2) did not further decrease pHi. Osmolality had little effect on pHi. Cell size was smaller than control at elevated pCO(2) at 320 mOsm/kg, and greater than control in hyperosmotic conditions at 40 mm Hg.