Comparison of influenza virus yields and apoptosis-induction in an adherent and a suspension MDCK cell line

Cell culture-based manufacturing of influenza vaccines is ideally based on easily scalable platforms using suspension cells that grow in chemically defined media. Consequently, different adherent cell lines selected for high virus yields were adapted to grow in suspension culture. This includes the MDCK suspension cell line MDCK.SUS2, which was shown to be a suitable substrate for influenza virus propagation in previous studies. In this study, we investigated options for further improvement of influenza A/PR/8 (H1N1) virus titres and cell-specific virus yields. Best results were achieved by performing a 1:2 dilution with fresh medium at time of infection. In shake flask cultivations, even for multiplicities of infection as low as 10⁻⁵, all cells were infected at 36 h post infection as determined by flow cytometry. In addition, these cells showed a better viability than cells infected without previous washing steps, which was reflected by a reduced level of apoptotic cells, and virus yields exceeding 3 log₁₀ HAU/100 μL. Comparison of bioreactor infections of MDCK.SUS2 cells to the parental adherent MDCK cells showed similar HA titres of 2.94 and 3.15 log₁₀ HAU/100 μL and TCID₅₀ of 1 × 10⁹ and 2.37 × 10⁹ infectious virions/mL. Surprisingly, virus-induced apoptosis differed between the two cell lines, with the MDCK.SUS2 cells showing a much stronger apoptosis induction than the adherent MDCK cells. Obviously, despite their resistance to anoikis, the suspension cells were more susceptible to virus-induced apoptosis. Whether this is related to the adaptation process itself and/or to changes in cell survival pathways influenced by adhesion molecules or influenza virus proteins needs to be clarified in additional studies.     

Use of MDCK cells for production of live attenuated influenza vaccine

To develop a cell-based live attenuated influenza vaccine (LAIV) manufacturing process, several different cell lines were evaluated by comparing the titer of viruses after infection with LAIV strains. While several cell lines have been reported to support influenza virus replication, the degree of replication and the ability to support replication of LAIV strains have not been systematically examined. MDCK cells, which have been considered as potential substrates for influenza vaccine production were evaluated in addition to Vero, MRC-5, WI-38 and FRhL cells. MRC-5, WI-38 and FRhL cells produced low to moderate titers of virus with titers equal or below 5.0 log₁₀ TCID₅₀/mL. Both Vero and MDCK cells could support a higher level of virus replication for certain strains, however, Vero cells only produced high titers when grown in the presence of serum. MDCK cells supported high levels of vaccine virus production for multiple different LAIV subtypes in both serum containing and serum-free media. These results suggest that MDCK cell-based production can be used as an alternative production platform to the currently used egg-based LAIV production system.     

Production of high-titer human influenza A virus with adherent and suspension MDCK cells cultured in a single-use hollow fiber bioreactor

Hollow fiber bioreactors (HFBRs) have been widely described as capable of supporting the production of highly concentrated monoclonal antibodies and recombinant proteins. Only recently HFBRs have been proposed as new single-use platforms for production of high-titer influenza A virus. These bioreactors contain multiple hollow fiber capillary tubes that separate the bioreactor in an intra- and an extra-capillary space. Cells are usually cultured in the extra-capillary space and can grow to a very high cell concentration. This work describes the evaluation of the single-use hollow fiber bioreactor PRIMER HF^® (Biovest International Inc., USA) for production of influenza A virus. The process was setup, characterized and optimized by running a total of 15 cultivations. The HFBRs were seeded with either adherent or suspension MDCK cells, and infected with influenza virus A/PR/8/34 (H1N1), and the pandemic strain A/Mexico/4108/2009 (H1N1). High HA titers and TCID₅₀ of up to 3.87 log₁₀ (HA units/100 μL) and 1.8 × 10¹⁰ virions/mL, respectively, were obtained for A/PR/8/34 influenza strain. Influenza virus was collected by performing multiple harvests of the extra-capillary space during a virus production time of up to 12 days. Cell-specific virus yields between 2,000 and 8,000 virions/cell were estimated for adherent MDCK cells, and between 11,000 and 19,000 virions/cell for suspension MDCK.SUS2 cells. These results do not only coincide with the cell-specific virus yields obtained with cultivations in stirred tank bioreactors and other high cell density systems, but also demonstrate that HFBRs are promising and competitive single-use platforms that can be considered for commercial production of influenza virus.     

Semi-perfusion cultures of suspension MDCK cells enable high cell concentrations and efficient influenza A virus production

Control and prevention of rapid influenza spread among humans depend on the availability of efficient and safe seasonal and pandemic vaccines, made primarily from inactivated influenza virus particles. Current influenza virus production processes rely heavily on embryonated chicken eggs or on cell culture as substrate for virus propagation. Today’s efforts towards process intensification in animal cell culture could innovate viral vaccine manufacturing using high-yield suspension cells in high cell density perfusion processes. In this work, we present a MDCK cell line adapted to grow as single cell suspension with a doubling time of less than 20 h, achieving cell concentrations over 1 × 10⁷ cells/mL in batch mode. Influenza A virus titer obtained in batch infections were 3.6 log₁₀(HAU/100 µL) for total- and 10⁹ virions/mL for infectious virus particles (TCID₅₀), respectively. In semi-perfusion mode concentrations up to 6 × 10⁷ cells/mL, accumulated virus titer of 4.5 log₁₀(HAU/100 µL) and infectious titer of almost 10¹⁰ virions/mL (TCID₅₀) were possible. This exceeds results reported previously for cell culture-based influenza virus propagation by far and suggests perfusion cultures as the preferred method in viral vaccine manufacturing.     

Influenza A virus production in a single-use orbital shaken bioreactor with ATF or TFF perfusion systems

Driven by the concept of plug-and-play cell culture-based viral vaccine production using disposable bioreactors, we evaluated an orbital shaken bioreactor (OSB) for human influenza A virus production at high cell concentration. Therefore, the OSB model SB10-X was coupled to two hollow fiber-based perfusion systems, namely, tangential flow filtration (TFF) and alternating tangential flow filtration (ATF). The AGE1.CR.pIX avian suspension cells grew to 50 × 10⁶ cells/mL in chemically defined medium, maintaining high cell viabilities with an average specific growth rate of 0.020 h⁻¹ (doubling time = 32 h). Maximum virus titers in the range of 3.28–3.73 log₁₀(HA units/100 µL) were achieved, corresponding to cell-specific virus yields of 1000–3500 virions/cell and productivities of 0.5–2.2 × 10¹² virions/L/d. This clearly demonstrates the potential of OSB operation in perfusion mode, as results achieved in a reference OSB batch cultivation were 2.64 log₁₀(HA units/100 µL), 1286 virions/cell and 1.4 × 10¹² virions/L/d, respectively. In summary, the SB10-X bioreactor can be operated with ATF and TFF systems, which is to our knowledge the first report regarding OSB operation in perfusion mode. Moreover, the results showed that the system is a promising cultivation system for influenza A virus vaccine production. The OSB disposable bioreactor has the potential for simplifying the scale-up from shake flasks to the large-scale bioreactor, facilitating rapid responses in the event of epidemics or pandemics.     

Mutations affecting the stability of the haemagglutinin molecule impair the immunogenicity of live attenuated H3N2 intranasal influenza vaccine candidates lacking NS1

The isolation and cultivation of human influenza viruses in embryonated hen eggs or cell lines often leads to amino acid substitutions in the haemagglutinin (HA) molecule. We found that the propagation of influenza A H3N2 viruses on Vero cells may trigger the appearance of HA destabilising mutations, affecting viral resistance to low pH or high temperature treatment. Two ΔNS1 reassortants, containing the HA sequences identical to the original human H3N2 influenza virus isolates were constructed. Passages of these viruses on Vero cells led to the appearance of single mutations in the HA₁ L194P or HA₂ G75R subunits that impaired virus stability. The original HA sequences and the stable phenotypes of the primary isolates were preserved if reassortants were passaged by infection at pH 5.6 and cultivation in medium at pH 6.5. Corresponding ΔNS1 reassortants were compared for their immunogenicity in ferrets upon intranasal immunisation. Vaccine candidates containing HA mutations demonstrated significantly lower immunogenicity compared to those without mutations. Thus, the retaining of the original HA sequences of human viruses during vaccine production might be crucial for the efficacy of live attenuated influenza vaccines.     

Comparison of egg and high yielding MDCK cell-derived live attenuated influenza virus for commercial production of trivalent influenza vaccine: In vitro cell susceptibility and influenza virus replication kinetics in permissive and semi-permissive cells

Currently MedImmune manufactures cold-adapted (ca) live, attenuated influenza vaccine (LAIV) from specific-pathogen free (SPF) chicken eggs. Difficulties in production scale-up and potential exposure of chicken flocks to avian influenza viruses especially in the event of a pandemic influenza outbreak have prompted evaluation and development of alternative non-egg based influenza vaccine manufacturing technologies. As part of MedImmune's effort to develop the live attenuated influenza vaccine (LAIV) using cell culture production technologies we have investigated the use of high yielding, cloned MDCK cells as a substrate for vaccine production by assessing host range and virus replication of influenza virus produced from both SPF egg and MDCK cell production technologies. In addition to cloned MDCK cells the indicator cell lines used to evaluate the impact of producing LAIV in cells on host range and replication included two human cell lines: human lung carcinoma (A549) cells and human muco-epidermoid bronchiolar carcinoma (NCI H292) cells. The influenza viruses used to infect the indicators cell lines represented both the egg and cell culture manufacturing processes and included virus strains that composed the 2006–2007 influenza seasonal trivalent vaccine (A/New Caledonia/20/99 (H1N1), A/Wisconsin/67/05 (H3N2) and B/Malaysia/2506/04). Results from this study demonstrate remarkable similarity between influenza viruses representing the current commercial egg produced and developmental MDCK cell produced vaccine production platforms. MedImmune's high yielding cloned MDCK cells used for the cell culture based vaccine production were highly permissive to both egg and cell produced ca attenuated influenza viruses. Both the A549 and NCI H292 cells regardless of production system were less permissive to influenza A and B viruses than the MDCK cells. Irrespective of the indicator cell line used the replication properties were similar between egg and the cell produced influenza viruses. Based on these study results we conclude that the MDCK cell produced and egg produced vaccine strains are highly comparable.     

MDCK cells that express proteases TMPRSS2 and HAT provide a cell system to propagate influenza viruses in the absence of trypsin and to study cleavage of HA and its inhibition

Cleavage of the influenza virus hemagglutinin (HA) by host cell proteases is essential for virus infectivity and, therefore, relevant proteases may present promising new drug targets. We recently demonstrated that serine proteases TMPRSS2 and HAT from human airways activate influenza virus HA with monobasic cleavage site in vitro. In the present study we generated MDCK cells with inducible expression of either TMPRSS2 or HAT. MDCK-TMPRSS2 and MDCK-HAT cells supported growth of human and avian influenza viruses of different subtypes in the absence of exogenous trypsin. Further, we used these cell lines to investigate the efficacy of protease inhibitors to prevent proteolytic activation of HA by TMPRSS2 and HAT. Multicycle viral replication in both cell lines was markedly suppressed in the presence of serine protease inhibitors and we found that particularly in MDCK-HAT cells proteolytic activation of progeny viruses was very susceptible to inhibitor treatment. Taken together, our data demonstrate that MDCK-HAT and MDCK-TMPRSS2 cells are useful experimental systems to study cleavage of HA by host cell protease and its inhibition and in addition represent applicable cell lines to propagate influenza viruses in the absence of trypsin.     

The stability and immunogenicity of inactivated MDCK cell-derived influenza H7N9 viruses

In recent years, cell-based influenza vaccines have gained a great interest over the egg-based vaccines. Several inactivated H7N9 vaccines have been evaluated in clinical trials, including whole-virion vaccines, split vaccines and subunit vaccines. Recently, we developed a new suspension MDCK (sMDCK) cell line for influenza viruses production. However, the properties of purified antigen from sMDCK cells remain unclear. In this study, the stability of influenza H7N9 vaccine bulk derived from sMDCK cells was investigated, and the data were compared with the vaccine antigen derived from our characterized adhesion MDCK (aMDCK) cells in serum-free medium. The influenza H7N9 bulks derived from sMDCK and aMDCK cells were stored at 2–8 °C for different periods of time, and a number of parameters selected to monitor the H7N9 vaccine antigen stability were evaluated at each interval (1, 3 and 12 months). The monitored parameters included virus morphology, hemagglutinin (HA) activity, HA concentration, antigenicity, and immunogenicity. The sMDCK-derived H7N9 bulk showed similar morphology to that of the aMDCK-derived H7N9 bulk, and there were no obvious changes after the extended storage periods. Furthermore, the HA titer, HA concentration, and antigenicity of sMDCK-derived H7N9 bulk were stable after 28 months of storage. Finally, the results of hemagglutination inhibition and neutralization tests showed that sMDCK- and aMDCK-derived H7N9 vaccines had comparable immunogenicity. These results indicated that sMDCK-derived H7N9 bulk has good stability compared to that of aMDCK-derived H7N9 bulk. Thus, the newly developed suspension MDCK cell line shows a great alternative for manufacturing cell-based influenza vaccines.     

A fast and efficient purification platform for cell-based influenza viruses by flow-through chromatography

Since newly emerging influenza viruses with pandemic potentials occurred in recent years, the demand for producing pandemic influenza vaccines for human use is high. For the development of a quick and efficient vaccine production, we proposed an efficient purification platform from the harvest to the purified bulk for the cell-based influenza vaccine production. This platform based on flow-through chromatography and filtration steps and the process only involves a few purification steps, including depth filtration, inactivation by formaldehyde, microfiltration, ultrafiltration, anion-exchange and ligand-core chromatography and sterile filtration. In addition, in the proposed chromatography steps, no virus capture steps were employed, and the purification results were not affected by the virus strain variation, host cells and culturing systems. The results from different virus strains which produced by Vero or MDCK cells in different culturing systems also obtained 33–46% HA recovery yields by this platform. The overall removal rates of the protein and DNA concentration in the purified bulk were over 96.1% and 99.7%, respectively. The low residual cellular DNA concentrations were obtained ranged from 30 to 130 pg per human dose (15 µg/dose). All influenza H5N1 purified bulks met the regulatory requirements for human vaccine use.     

Cloned cDNA of A/swine/Iowa/15/1930 internal genes as a candidate backbone for reverse genetics vaccine against influenza A viruses

Reverse genetics viruses for influenza vaccine production usually utilize the internal genes of the egg-adapted A/Puerto Rico/8/34 (PR8) strain. This egg-adapted strain provides high production yield in embryonated eggs but does not necessarily give the best yield in mammalian cell culture. In order to generate a reverse genetics viral backbone that is well-adapted to high growth in mammalian cell culture, a swine influenza isolate A/swine/Iowa/15/30 (H1N1) (rg1930) that was shown to give high yield in Madin-Darby canine kidney (MDCK) cells was used as the internal gene donor for reverse genetics plasmids. In this report, the internal genes from rg1930 were used for construction of reverse genetics viruses carrying a cleavage site-modified hemagglutinin (HA) gene and neuraminidase (NA) gene from a highly pathogenic H5N1 virus. The resulting virus (rg1930H5N1) was low pathogenic in vivo. Inactivated rg1930H5N1 vaccine completely protected chickens from morbidity and mortality after challenge with highly pathogenic H5N1. Protective immunity was obtained when chickens were immunized with an inactivated vaccine consisting of at least 2(9) HA units of the rg1930H5N1 virus. In comparison to the PR8-based reverse genetics viruses carrying the same HA and NA genes from an H5N1 virus, rg1930 based viruses yielded higher viral titers in MDCK and Vero cells. In addition, the reverse genetics derived H3N2 and H5N2 viruses with the rg1930 backbone replicated in MDCK cells better than the cognate viruses with the rgPR8 backbone. It is concluded that this newly established reverse genetics backbone system could serve as a candidate for a master donor strain for development of inactivated influenza vaccines in cell-based systems.     

Immortalized chick embryo cell line adapted to serum-free growth conditions and capable of replicating human and reassortant H5N1 influenza strains for vaccine production

The current process of influenza vaccine production can take 6-9 months and is dependent on the availability of embryonated eggs. Additionally, this process selects for receptor-binding variants with reduced antigenicity and requires significant downstream production for purification. We have established an immortalized chick embryo cell line, termed PBS-12SF, which is adapted to growth in serum free conditions, and is capable of replicating human and reassortant H5N1 influenza strains to high titers. In many cases, PBS-12SF cells produced higher growth titers of influenza virus than those of primary chick embryo kidney (CEK) cells, Madin-Darby Canine Kidney (MDCK) cells and African green monkey kidney cells (Vero). Additionally, in PBS-12SF cell cultures, influenza virus is released into the culture fluid without need for exogenous proteases, which can simplify downstream processing for vaccine production.     

Statistical optimization of influenza H1N1 production from batch cultures of suspension Vero cells (sVero)

Efficient vaccine production requires the growth of large quantities of virus produced with high yield from a safe host system. Human influenza vaccines are produced in embryonated chicken eggs. However, over the last decade many efforts have allowed the establishment of cell culture-derived vaccines.We generated a Vero cell line adapted to grow in suspension (sVero) in a serum-free medium and evaluated it for its potential as host cell for influenza vaccine production. Initially we studied the capacity of sVero cells to grow in the presence of incremental concentrations of trypsin. In comparison with adherent Vero cells (aVero), we found that sVero cells maintain their growth kinetics even with a three-fold increase in trypsin concentration.The influence of the conditions of infection on the yield of H1N1 produced in serum-free suspension cultures of sVero cells was investigated by a 2² full factorial experiment with center point. Each experiment tested the influence of the multiplicity of infection (m.o.i.) and trypsin concentration, on production yields at two levels, in four possible combinations of levels and conditions, plus a further combination in which each condition was set in the middle of its extreme levels.On the basis of software analysis, a combination of m.o.i. of 0.0066 TCID_50%/cell and trypsin concentration of 5 μg/1.0 × 10⁶ cells with a desirability of 0.737 was selected as the optimized condition for H1N1 production in sVero cells.Our results show the importance of proper selection of infection conditions for H1N1 production on sVero cells in serum-free medium.     

Development of a live-attenuated influenza B ΔNS1 intranasal vaccine candidate

We discovered a unique, single amino acid mutation in the influenza B M1 protein promoting viral growth of NS1 truncation mutants in Vero cells. Due to this mutation, we were able to generate an influenza B virus lacking the complete NS1 open reading frame (ΔNS1-B virus) by reverse genetics, which was growing to titers of 8 log₁₀ TCID₅₀/ml in a Vero cell culture-based micro-carrier fermenter. The ΔNS1-B vaccine candidate was attenuated in IFN-competent hosts such as human alveolar epithelial cells (A549) similar to influenza A ΔNS1 viruses. In ferrets, the ΔNS1-B virus was replication-deficient and did not provoke any clinical symptoms. Importantly, a single intranasal immunization of ferrets at a dose as low as 6 log₁₀ TCID₅₀/animal induced a significant HAI response and provided protection against challenge with wild-type influenza B virus. So far, the lack of a ΔNS1-B virus component growing to high titers in cell culture has been limiting the possibility to formulate a trivalent vaccine based on deletion of the NS1 gene. Our study closes this gap and paves the way for the clinical evaluation of a seasonal, trivalent, live replication-deficient ΔNS1 intranasal influenza vaccine.     

Wave microcarrier cultivation of MDCK cells for influenza virus production in serum containing and serum-free media

A process for equine influenza virus vaccine production using a microcarrier system (Cytodex 1) in a 2 L Wave bioreactor is described. Growth of Madin Darby canine kidney (MDCK) cells in serum containing GMEM medium (SC) is compared to growth in serum-free Ex-Cell MDCK medium (SF) without washing steps and medium exchange before infection. Cultivations with microcarrier concentrations of 2 and 4 g/L for both media are shown. Metabolic data from carbon and amino acid metabolism are discussed. Additionally, in roller bottle experiments the influence of multiplicity of infection (moi) and trypsin concentration on the HA value was investigated. Analysis of HA and TCID(50) at 37 degrees C showed a stable HA of maximum 2.6 log HA/100 microL for 2 weeks. Peak TCID(50) titers of 10(7.7) viruses/mL were achieved 20h post infection, but infectivity was below detection limit after 150 h. Cell attachment onto microcarriers under serum-free conditions was improved by Ca(2+) addition and by cell harvesting without trypsin using only an EDTA/PBS solution. For the wave cultivation maximum virus titers of 2.3-2.6 log HA units/100 microL were reached from infection with a moi of 0.05. However, in SF medium pH dropped to less than pH 6.8 which resulted in lower HA titers of 1.7 log HA units/100 microL. For the higher microcarrier concentration (4 g/L) medium exchange steps (500 mL) were needed for both media. Omission of the washing step and medium exchange before infection in SF medium clearly simplified the influenza production process; however, for higher virus yields a better pH control of the wave bioreactor would be required. Higher cell densities (2.8 x 10(6) cells/mL for 2 g/L microcarrier) and better attachment compared to stirred tank bioreactors showed, that the wave bioreactor is a good alternative to stirred tank processes for expanding production capacities in case of a pandemic.     

High titer growth of human and avian influenza viruses in an immortalized chick embryo cell line without the need for exogenous proteases

The current method of growing influenza virus for vaccine production is through the use of embryonated chicken eggs. This manufacturing system yields a low concentration of virus per egg, requires significant downstream production for purification, and demands a considerable amount of time for production. We have demonstrated an immortalized chick embryo cell line, termed PBS-1, is capable of growing unmodified recent isolates of human and avian influenza A and B viruses to extremely high titers. In many cases, PBS-1 cells out perform primary chick embryo kidney (CEK) cells, Madin-Darby Canine Kidney (MDCK) cells and African green monkey kidney cells (Vero) in growth of recent influenza isolates. PBS-1 cells are free of any exogenous agents, are non-tumorigenic, and are readily adaptable to a variety of culture conditions, including growth on microcarrier beads. Influenza viruses grown in PBS-1 cells are released into the culture fluid without the need for exogenous proteases, thus simplifying downstream processing. In addition to offering a significant improvement in vaccine production, PBS-1 cells should prove valuable in diagnostics and as a cell line of choice for influenza virus research.     

Suspension-Vero cell cultures as a platform for viral vaccine production

Since Vero cells are currently considered as an acceptable cell substrate to produce a wide range of viruses, we developed a virus production platform using Vero cells adapted to grow in suspension in serum-free media. After adapting anchorage-dependent Vero cells to grow as a free-cell suspension, vesicular stomatitis virus, herpes simplex virus 1 and polio virus 1 production rates were evaluated in batch cultures using spinner flasks and perfused cultures in a bioreactor. The achieved results constitute valuable information for the development of a low-cost high-productivity process using a suspension culture of Vero cells to produce viral vaccines.     

Optimization of influenza A vaccine virus by reverse genetic using chimeric HA and NA genes with an extended PR8 backbone

The yield of influenza antigen production may significantly vary between vaccine strains; for example the A/California/07/09 (H1N1)-X179A vaccine virus, prepared during 2009 influenza pandemic, presented a low antigen yield in eggs compared to other seasonal H1N1 reassortants. In this study a bi-chimeric virus expressing HA and NA genes with A/Puerto Rico/8/34 (H1N1) (PR8) and X179A domains was rescued by reverse genetics using a mixture of Vero/CHOK1 cell lines (Medina et al. [7]). The bi-chimeric virus obtained demonstrated to yield much larger amounts of HA than X179A in eggs as measured by single-radial-immunodiffusion (SRID), the reference method to quantify HA protein in influenza vaccine. Such kind of optimized virus using PR8 backbone derived chimeric glycoproteins could be used as improved seed viruses for vaccine production.     

An animal component free medium that promotes the growth of various animal cell lines for the production of viral vaccines

IPT-AFM is a proprietary animal component free medium that was developed for rabies virus (strain LP 2061) production in Vero cells. In the present work, we demonstrated the versatility of this medium and its ability to sustain the growth of other cell lines and different virus strains. Here, three models were presented: Vero cells/rabies virus (strain LP 2061), MRC-5 cells/measles virus (strain AIK-C) and BHK-21 cells/rabies virus (strain PV-BHK21). The cell lines were first adapted to grow in IPT-AFM, by progressive reduction of the amount of serum in the culture medium. After their adaptation, BHK-21 cells grew in suspension by forming clumps, whereas MRC-5 cells remained adherent. Then, kinetics of cell growth were studied in agitated cultures for both cell lines. In addition, kinetics of virus replication were investigated.