Intensification of Influenza Purification From Clarified Harvest to Formulated Product in a Single Shift

M. Tajnik Sbaizero, M. Wolschek, M. Reiter, T. Muster, P. Gagnon, A. Štrancar

nfluenza is a global respiratory disease with an estimated mortality of up to a half million I people per year (1). The majority of traditional influenza vaccines are still produced in eggs. Downstream processing typically consists of Vero cells infected with purified influenza A: Infectious virus particles were stained clarification by centrifugation, with rabbit antiinfluenza A NP- and a goat antirabbit IgG antibody concentration by ultrafiltration, and conjugated with AlexaFluor 488 (green). Cell nuclei are stained blue with DAPI (4‘,6-diamidino-2-phenylindole). purification by ultracentrifugation (2). Recombinant vaccines are most often purified by . across the field of biopharmaceuticals. help companies focus resources on Chromatographic purification of One of those trends is process essential process functions. already has achieved major intensification, referring to As with many viruses, purification of improvements in recovery and development of processes that influenza virus from clarified cell scalability (3), but it also is important harmonize integration of fewer and culture harvests often begins with because it enables virus purification more capable steps to achieve higher tangential-flow filtration (TFF), which to keep pace with important productivity and reproducibility as simultaneously achieves multiple tasks. regulatory and manufacturing trends well as reduce manufacturing costs. The first is concentration of the virus, Intensification takes many forms. which is retained, while contaminant Continuous processing represents one levels are reduced by passing through Product Focus: Vaccines aspect, and it already has been applied membrane pores. Both capabilities often Process Focus: Downstream successfully to anion-exchange are referred to as ultrafiltration (UF). processing purification of influenza A (4). An The third capability is buffer exchange arguably more important aspect by a process called diafiltration (DF). Who Should Read: Process involves application of tools that The original fluid is replaced gradually development and manufacturing overcome limitations of traditional by buffer with a formulation suitable for options and thereby improve results an initial purification step. In this Keywords: Chromatography, and/or reduce the number of process discussion, reference to UF–DF is influenza, process intensification, steps. Application of single-use simplified to TFF. single-use, tangential-flow filtration, processing materials also contributes to Despite the valuable role of TFF, Vero cells intensification. By suspending the its benefits come at a price. Shear Level: Intermediate validation and manufacturing burdens stress produced during TFF has been of multiple use components, disposables documented to strip outer envelopes from lipid-enveloped viruses and to One key distinction of monolithic influenza B, both lacking TFF steps, fracture brittle capsids of non-lipid− chromatography media compared with and both using a single enveloped species (5, 6). Such liabilities columns packed with particles is that chromatography step with a cation- might be prevented by direct virus monoliths do not have void spaces. exchange monolith on a single-use capture on traditional porous particle Flow through the interconnected basis. The choice of process buffers columns, but this is impractical for channels is laminar, so eddy formation enables final formulation by simple two reasons. An obvious reason is the and shear stress are nil. Another key dilution of the product pool. DNA extended time that would be required distinction is that mass transport in digestion requires two hours. Capture, to load unconcentrated virus. The monoliths is exclusively convective, purification, and formulation are other is that chromatography columns rather than the situation in porous achieved within four hours. Host-cell packed with particles represent an particle columns where flow between DNA and host-cell (HCP) are additional source of shear stress. Flow particles is convective, but solutes reduced more than 99%, and final through the irregular spaces between move into and out of the pores virus recovery is 80%. particles causes formation of eddies, exclusively by diffusion. zones of recirculation. When bulk Diffusion is the reason columns Materials and Methods flow through the interparticle space packed with particles are slow. Influenza virus was produced in Vero encounters eddy flow in the opposite Diffusion is even slower for large cells. Viruses were constructed as direction, it generates shear stress at products such as viruses because previously described (9) and propagated their interface (Figure 1). Shear stress diffusion constants become slower with in Nunc Cell Factory CF-10 systems in column void spaces is directly increasing particle size. By contrast, (Thermo Fisher Scientific) using proportional to flow rate (7). convective flow can be thought of as a serum-free medium Opti Pro SFM river. Objects flow at the same rate as (Invitrogen, Thermo Fisher Scientific) Figure 1: Shear stress in the void volume of the current no matter what size they supplemented with 4 mM GlutaMax I packed particle columns; solid areas represent are. The result is that binding (Invitrogen, Thermo Fisher Scientific). particle surfaces. Interparticle void space is indicated in white. Black arrows indicate flow, efficiency, dynamic binding capacity, Upon , cells were grown using and line thickness indicates flow velocity. and elution efficiency in monoliths are recombinant trypsin (Thermo Fisher Circular patterns indicate eddies. Areas marked all independent of product size and Scientific) and incubated at 5.0% CO2 in red highlight zones of countercurrent flow flow rate. Flow rates can be 20–50× and 95% relative humidity at 37 °C that generate shear stress. Axial shear stress at higher than in packed-particle columns (influenza A) or 33 °C (influenza B) particle surfaces is negligible with flow rate approaching zero there because of friction. without compromising performance (8). for three days. Harvests were clarified Altogether, the combination of by low-speed centrifugation. Clarified convective efficiency and laminar flow harvest was treated with 20 U/mL enables rapid virus concentration on Benzonase endonuclease (Merck, monoliths without the flow-rate Germany) in the presence of 2 mM restrictions of traditional columns and magnesium for two hours at room without the shear forces created by temperature. either TFF or particle-packed Virus purification was conducted columns. In short, they enable the with 1, 8, or 80 mL CIMmultus™ SO3 sequential functions provided by TFF cation-exchange monoliths. They were and traditional column capture to be sanitized in advance with 1 M NaOH, combined into a single faster process then washed with water and equilibrated step with less risk of damaging a to 50 mM HEPES, 200 mM sucrose, product. pH 7.0. was diluted In this report we describe processes 2:1 (sample:buffer) and applied to the for purification of influenza A and monoliths without further preparation. was diluted 1:1 with Figure 2: Elution of Influenza B from a 1-mL CIMmultus SO3 monolith — the inset represents a zoomed image of the elution step; blue = UV absorbance at 280 nm; red = UV absorbance at equilibration buffer, then loaded at the 260 nm; black = conductivity same flow rate. The monolith was washed with 50 mM HEPES, 200 mM ) 2.5 300 (mcm) Conductivity In uenza B 2.5 50 lution sucrose, pH 7.5 containing 100 mM claried harvest 2.0 2.0 40 250 NaCl (influenza A) or 50 mM NaCl CIMmultus SO3 1.5 30 200 1.5 (influenza B) and eluted with a linear 1.0 20 150 gradient. The gradient endpoint buffer 1.0 0.5 10 100 in both cases was 50 mM HEPES, 200 0.0 0 26 nm, 28 nm 0.5 oad 630 635 640 50 mM sucrose, 2.0 M NaCl, pH 7.5. CIP Specific gradient configurations for A ( 0.0 0 0 100 200 300 400 500 600 700 influenza A and B are discussed in the lution Volume (CV) next section. Columns were cleaned and Figure 3: Elution of Influenza A from a 1 mL CIMmultus SO3 monolith — right image represents a zoomed view of the elution step; blue = UV absorbance at 280 nm; red = UV absorbance at 260 nm; brief wash step with pH 7.5 buffer black = conductivity containing 50 mM NaCl, the virus was ) 400 300 (mcm) Conductivity eluted with a 50-CV linear gradient to 250 100% buffer B. The six-hour total 300 oad 100 chromatography time included 80 200 sanitization, equilibration, load, wash 200 60 150 In uenza A 40 elution, and cleaning in place (CIP). 20 lution CIP 100 100 claried harvest The infectious virus was recovered in a 26 nm, 28 nm 0 50 CIMmultus SO3 525 535 545 555 4.1-mL fraction (indicated in the

A ( 0 0 colored background area) representing 0 100 400 500 600 lution Volume (CV) an overall concentration factor of 73-fold from the undiluted cell culture. Virus concentration was 2.43 × 1010 Figure 4: Elution of Influenza B from an 80 mL CIMmultus SO3 monolith with 2 µm channels; FFU/ mL, and recovery was 74%. The blue = UV absorbance at 280 nm; red = UV absorbance at 260 nm; black = conductivity; virus fraction was formulated with CV = column volume about a fourfold dilution to lower buffer

) capacity, adjust salt concentration, and 2.5 In uenza B 300 (mcm) Conductivity introduce a nonanimal−derived claried harvest 250 2.0 stabilizer before final sterile filtration. CIMmultus SO3 200 1.5 lution Figure 3 shows the elution profile of 150 influenza A from a 1-mL CIMmultus 1.0 2.5 h 100 SO3 monolith. Initial material was 26 nm, 28 nm 0.5 oad CIP 50 diluted 1:2 (buffer:sample) with

A ( 0.0 0 equilibration buffer at pH 7.0. Similar to 0 20 40 60 340 360 380 400 420 the influenza B process, loading volume lution Volume (CV) was >500 CV. After loading, the column was washed using 100 mM NaCl containing HEPES buffer at pH 7.5. sanitized in place with 1 M NaOH, 2 analyzed by enzyme-linked The product was eluted during 50-CV M NaCl after each run. All immunosorbent assay (ELISA). linear salt gradient to 2 M NaCl. chromatography steps were performed at Infectious virus was recovered in a 5.1- a flow rate of five bed volumes per Results mL fraction (indicated in the colored minute. Infectious virus–containing Omitting TFF in advance of background area) with concentration of fractions were diluted to the formulation chromatography required that samples 4.44 × 108 FFU/mL. Pooled infectious salt concentration, and a nonanimal- be equilibrated by dilution. Filtered virus was formulated as described above. derived stabilizer was added. harvest was diluted at a proportion of Overall recovery was 82%. Infectious particle concentration one part equilibration buffer to two The influenza B purification was was determined by focus-forming parts sample, then applied to the scaled up 80-fold with an 80-mL assay (FFA) using a Cytation 5 equilibrated cation exchanger. This CIMmultus SO3 monolith to supply multimode fluorescent reader using resulted in 0.4% of the virus particles material for clinical trials (Figure 4). influenza-specific . Briefly, in the flow through for influenza A About 360 CV of 1:1 diluted FFA analysis was performed by and 3.5% for influenza B. Increasing Benzonase-treated material was loaded dispensing fivefold sample dilutions the dilution to a proportion of 1:1 at a flow rate of 3 CV/min, followed on Vero cells growing in 96-well flat- reduced influenza B losses in the flow by a wash step with 50 mM HEPES, bottom plates (Thermo Fisher through to 0.1% or less. With this 50 mM NaCl, pH 7.5. Product was Scientific). After incubation for 24 approach, binding capacity of >3.16 × eluted during a 25-CV linear salt hours, cells were fixed and 1010 FFU/mL SO3 resin was reached, gradient to 2 M NaCl. Total immunostained using an antibody while achieving >99% reduction of chromatography time was four hours. specific for influenza virus NP. The host-cell DNA and HCP. Infectious virus recovery was 83.4%. number of infected cells was counted Figure 2 illustrates the elution Host-cell protein was reduced 99.7% in triplicate for each sample. FFA titer profile of influenza B from a 1-mL and host DNA by 99.99%. was calculated and expressed as FFU/ CIMmultus SO3 monolith with 1:1 mL. DNA concentration was dilution of the initial material and Discussion and Conclusions determined by using a Picogreen column equilibration at pH 7.0. The Direct virus capture from diluted feed dsDNA assay kit or with qPCR. Total dominant feature is the extended load streams with monolithic columns protein content was determined using interval at 600 column volumes (CV). makes worthy contributions to process a Coomassie (Bradford) protein assay Because of the 5 CV/min flow rate, intensification during product kit (Thermo Fisher Scientific). Host- however, it corresponds to a loading development as well as during scale-up cell protein (HCP) content was time of only two hours. Following a and manufacturing (Figure 5). Cell Figure 5: Instrument-operational steps in purification of influenza virus 3 Morenweiser R. Downstream Processing of Viral Vectors and Vaccines. Gene Clari ed benzonase-treated Ther. 12(1): S103–S110. cell culture harvest 4 Fischer L, Wolff M, Reichl U. Purification of Cell Culture−Derived Influenza Traditional Process Intensi ed Process A Virus Via Continuous Anion Exchange Chromatography on Monoliths. Vaccine 36(22) 2018: 3153–3160; doi:10.1016/j. Tangential-ow ltration vaccine.2017.06.086. (TFF, UF/DF) 5 Besnard L, et al. Clarification of Vaccines: An Overview of Filter Based Cation-exchange Cation-exchange Technology Trends and Best Practices. chromatography (SO3) chromatography (SO3) Biotechnol. Adv. 34(1) 2016: 1–13; doi:10.1016/j. biotechadv.2015.11.005. 6 Negrete A, Pai A, Shiloach J. Use of Hollow Fiber Tangential Flow Filtration for Formulation by TFF the Recovery and Concentration of HIV Virus- Like Particles Produced in Insect Cells. J. Virol. Met. 195, 2014: 240–246; doi:10.1016/j. jviromet.2013.10.017. Sterile ltration Sterile ltration 7 Gagnon P. The Emerging Generation of Chromatography Tools for Virus Purification. BioProcess Int. 6(6) 2008: S24–S30. 8 Kramberger P, Urbas L, Štrancar A. culture harvests for purification process and validation. It also reduces the Downstream Processing and Chromatography development often are very limited in number of pieces of major equipment Based Analytical Methods for Production of Vaccines, Vectors, and the early stages. Small-volume TFF required for a process, which lowers . Hum. Vaccin. Immunother. 11(4) systems are becoming more available capital costs, simplifies operator 2015: 1010–1021; doi: 10.1080/ but still not at a size to provide scalable training and certification, and reduces 21645515.2015.1009817. results from laboratory scale to the need to set up supply chains and 9 Jindra C, et al. Attenuated Recombinant manufacturing. Use of a larger-than- maintain inventories of disposable Influenza A Virus Expressing HPV16 E6 and necessary filter area increases risk of media. It also reduces the number and/ E7 As a Novel Therapeutic Vaccine Approach. PLoS One 10(9s) 2015; doi:10.1371/journal. product loss because many virus species or volume of process buffers, tankage, pone.0138722. are known to bind nonspecifically to and buffer storage space; it increases 10 Oliver D. Virus Interactions with membrane surfaces (10). Neither of facility capacity by decreasing process Membrane Filters. Biotechnol. Bioeng. 10(6) 1968: these liabilities is a factor with dilution- time; and it reduces waste. 877–889; https://doi.org/10.1002/ to-direct capture, and omission of Not all purifications can be bit.260100612. membrane concentration bypasses a accomplished with one or even two 11 Weigel T, et al. A Flow-Through documented source of virus particle chromatography steps, but the same Chromatography Process for Influenza A and B Virus Purification. J. Virol. Methods. 207 (2014): damage by shear stress. principles do apply, and the 45–53. Capture on a porous particle opportunities for intensification 12 Transfiguracion J, et al. Particle column still would require a TFF increase. The tactic of dilute-to-direct- Quantification of Influenza Viruses By High concentration step because of the low load on monoliths equally serves the Performance Liquid Chromatography. Vaccine flow rates, but the ability of the need to reequilibrate buffers between 33(1) 2015: 78–84; doi:10.1016/j. monolith to accommodate high flow purification steps, and it has been used vaccine.2014.11.027. rates without compromising effectively to do so in many influenza 13 Banjac M, et al. Purification of Vero Cell Derived Live Replication Deficient Influenza A performance suspends that limitation. purification processes (4, 11–13). and B Virus By Ion Exchange Monolith That ability to achieve concentration Whatever the number of process steps Chromatography. Vaccine 32(21) 2014: 2487– and purification simultaneously also a purification process might include, 2492; doi:10.1016/j.vaccine.2014.02.086. bypasses two sources of shear stress pursuing a strategy of single-use and makes use of monoliths even more compounds the overall degree of Mojca Tajnik Sbaizero is project manager attractive. This is important in itself intensification. at BIA Separations in Ajdovscina, Slovenia. because process intensification is Markus Wolschek is CSO, Manfred Reiter meaningless if it is achieved at the References is CTO, and Thomas Muster is CEO of expense of product recovery. 1 Influenza (Seasonal). World Health BlueSky Vaccines in Vienna, Austria. Combining multiple operations into Organization (WHO: Geneva, Switzerland, Corresponding author Pete Gagnon is CSO 2018); www.who.int/en/news-room/fact-sheets/ a single step further compounds and Aleš Štrancar is CEO of BIA detail/influenza-(seasonal). Separations in Ajdovscina, Slovenia, BIA process intensification in a number of 2 Mischler R, Metcalf IC. Inflexal V a Separations d.o.o., Mirce 21, 5270 ways, beginning with reduction of the Trivalent Virosome Subunit Influenza Vaccine: Ajdovščina, Slovenia; Pete.Gagnon@ number of steps requiring development Production. Vaccine 20(5s): B17–23; biaseparations.com. doi:10.1038/sj.gt.3302624.

Electronic permissions to BIA Separations from BioProcess International October © 2018 Informa.