METHOD FOR PURIFYING VIRUS

Abstract

The invention relates to a method for purifying an adenovirus comprising (a) providing a liquid sample comprising adenovirus, (b) clarifying said sample by depth filtration, (c) performing anion exchange chromatography comprising the steps of (i) directly applying the clarified sample of (b) to an anion exchange column, (ii) eluting adenovirus from the anion exchange column to provide an eluate. The invention also relates to adenovirus produced from said methods, and to compositions comprising same.

Claims

1. A method for purifying an adenovirus comprising (a) providing a liquid sample comprising adenovirus (b) clarifying said sample by depth filtration (c) performing anion exchange chromatography comprising the steps of (i) directly applying the clarified sample of (b) to anion exchange (ii) eluting adenovirus from the anion exchange to provide an eluate.

2. A method according to claim 1 wherein the anion exchange chromatography of (c) comprises membrane anion exchange chromatography or column anion exchange chromatography.

3. A method according to claim 2 wherein the membrane comprises quaternary amines.

4. A method according to claim 2 or claim 3 wherein the membrane comprises Sartobind Q membrane.

5. A method according to any preceding claim wherein the anion exchange chromatography is carried out with load and wash salt concentrations in the range 24 to 31 mS/cm, with 20 mM Tris pH 8.0.

6. A method according to any preceding claim wherein the anion exchange chromatography is carried out with load and wash salt concentrations in the range 280 mM NaCl to 361 mM NaCl, with 20 mM Tris pH 8.0.

7. A method according to any preceding claim wherein the anion exchange chromatography is carried out with wash conditions of 15 mS/cm to 30 mS/cm conductivity, with 20 mM Tris pH 8.0.

8. A method according to any preceding claim wherein the sample is adjusted to a conductivity of 28 mS/cm before applying to the anion exchange column.

9. A method according to any preceding claim wherein said depth filtration of step (b) comprises primary clarification followed by secondary clarification.

10. A method according to any preceding claim wherein said depth filtration of step (b) comprises combined primary and secondary clarification.

11. A method according to any preceding claim wherein said depth filtration of step (b) comprises use of a CoSP depth filter.

12. A method according to any preceding claim wherein the anion exchange chromatography of (c) is performed in-line with the clarifying depth filtration of (b) as a single unit operation.

13. A method according to any preceding claim wherein said liquid sample comprises cell lysate produced from cultured host cells comprising adenovirus.

14. A method according to claim 13 wherein said cell lysate is produced by treating a sample of cultured host cells comprising adenovirus with a detergent.

15. A method according to claim 13 or claim 14 wherein said cell lysate is produced by treating a sample of cultured host cells comprising adenovirus with a nuclease.

16. A method according to claim 15 wherein said nuclease comprises DNAse and/or RNAse.

17. A method according to claim 15 or claim 16 wherein said nuclease comprises, or consists of, endonuclease from Serratia marcescens.

18. A method according to any of claims 15 to 17 wherein said nuclease comprises, or consists of, Benzonase.

19. A method according to claim 18 wherein said Benzonase is added to said sample at a final concentration of 15 units/millilitre.

20. A method according to any preceding claim further comprising: (d) performing buffer exchange on the eluate of (c).

21. A method according to claim 20 wherein step (d) comprises tangential flow filtration.

22. A method according to any preceding claim wherein said adenovirus is, or is derived from, a simian adenovirus, preferably wherein said adenovirus is, or is derived from, a species E simian adenovirus, preferably wherein said adenovirus is ChAdOx1.

23. A method according to claim 22 wherein the adenovirus is ChAdOx1 nCoV-19.

24. A method according to any of claims 1 to 21 wherein said adenovirus is an adenovirus having a capsid with charge characteristics similar to those of ChAdOx1.

25. A method according to any of claims 1 to 21 wherein said adenovirus is an adenovirus having capsid charge characteristics such that its elution conductivity, in 20 mM Tris pH 8 on a Sartobind Q or Pall Mustang Q membrane chromatography unit, exceeds 25 mS/cm.

26. An adenovirus prepared by a method according to any preceding claim.

27. A composition comprising an adenovirus according to claim 26.

28. A composition according to claim 27 which is a pharmaceutical composition.

29. A composition according to claim 27 which is a vaccine composition.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0281] Embodiments of the present invention will now be described further, with reference to the accompanying drawings, in which:

[0282] FIG. 1 shows graphs, bar charts and a table.

[0283] FIG. 2 shows graphs, diagrams and tables.

[0284] FIG. 3 shows graphs, tables and a diagram.

EXAMPLES

[0285] 50 L and 200 L runs were performed at Pall Biotech, Portsmouth, UK.

[0286] All obtained with BalanCD293 medium/feed combination & HEK293 T-rex cells.

Example 1: Production/Upstream Process

[0287] We initially investigated medium/feed combinations for a fed-batch USP. BalanCD293 medium and feed (Fujifilm) was found to support growth of the producer cells (HEK293 T-rex, Thermo) to 1.210.sup.7 cells/mL with high viability (FIG. 1A). Using this combination in small-scale production of adenovirus vectors of two serotypes and carrying three transgenes, we attained productivity exceeding 510.sup.11 virus particles (VP) per mL, around 5-fold greater than typically obtained in our previous USP (FIG. 1B). To our knowledge such productivity has not previously been reported from a non-perfusion USP: a fall in cell-specific productivity to <110.sup.5VP/cell is commonly observed at cell densities exceeding 110.sup.6 cells/mL (Kamen, A. and O. Henry, Development and optimization of an adenovirus production process. J Gene Med, 2004. 6 Suppl 1: p. S184-92).

[0288] Using ChAdOx1 nCoV-19 starting material, we initially investigated the optimal multiplicity of infection (MOI), cell density and time of harvest (FIG. 1C-D). An MOI of 10, cell density of 2-310.sup.6/mL and time of harvest 42-48 hours post infection were selected for further work. This USP was subsequently scaled up, with USP productivities in 10 L, 50 L and 200 L stirred tank reactors (STR) in the range 2-410.sup.11 VP/mL, and acceptable cell growth and metabolite profiles (FIGS. 1E-F).

[0289] Early STR batches made use of a DSP very similar to that we had previously described, which again achieved recovery of 50-60% and quality characteristics compliant with a regulator-accepted specification for product for clinical use (FIG. 1G) (Fedosyuk et al 2019 Vaccine vol 37 pages 6951-6961).

[0290] We refer to FIG. 1 which shows development of fed batch process.

[0291] Panel A shows cell counts (solid lines, filled symbols) and viability (dashed lines, open symbols) attained during growth of HEK293 T-rex cells in BalanCD medium with (triangles) or without (squares) feed, as compared to the CD293 medium (circles, Thermo) used in our previous process. In the case of the BalanCD medium+feed condition, feed (5% v/v) was added at 36 and 108 hours.

[0292] Panel B shows small-scale USP productivity of ChAdOx1-luciferase, ChAdOx1-LassaGP, and ChAdOx2-GFP with BalanCD medium/feed (infected at 410.sup.6 cells/mL, MOI=10) as compared to our previously established conditions in CD293 medium (infected at 110.sup.6 cells/mL, MOI=3) (Fedosyuk et al 2019 Vaccine vol 37 pages 6951-6961).

[0293] ChAdOx1-luciferase infections were performed in a 3 L bioreactor; the other two viruses were produced in 30 mL volume in shake flasks.

[0294] Panels C-D show small-scale USP productivity of ChAdOx1 nCoV-19 in shake flasks at 30 mL working volume at MOI=3 (C) and MOI=10 (D) respectively. Legend indicates cell density represented by each line, in million cells/mL at point of infection.

[0295] Infectious unit (IU) titers broadly paralleled VP titers; results are representative of two replicate experiments.

[0296] For panels A-D, points indicate median and error bars show range of results for 2-3 replicate flasks.

[0297] Panels E-F: Examples of 50 L and 200 L USP runs (carried out by Pall Biotech). (E) shows cell growth (solid lines) and viability (dashed lines). (F) shows glucose (solid lines) and lactate (dashed lines) concentrations.

[0298] Panel G: Examples of quality of drug substance from 50 L and 200 L runs (carried out by Cobra).

Example 2: Purification/Downstream Process

[0299] The inventors proceeded to develop the purification process in such a manner as to be scalable to 1000-2000 L. Handling of the large volumes of lysate, diafiltration buffer and waste during the initial TFF step was identified as a process bottleneck. We therefore developed a simplified DSP by loading the clarified lysate directly on an AEX membrane, followed by a single TFF diafiltration polish/formulation step. As well as removing the initial TFF step, this process change provides the option of execution of clarification and AEX as a single unit operation (FIG. 2A). Small-scale experiments were used to demonstrate feasibility and optimise conditions of the AEX step (FIG. 3 A-D), prior to execution of the complete revised DSP at 10 L scale (FIG. 2B). Binding capacity of the AEX membrane was reduced (to c. 310.sup.13 VP per mL of membrane, as compared to >510.sup.13 VP with loading of a diafiltered feed), but it was typically possible to recover c. 90% of loaded virus in the eluate. After a final TFF step, overall recovery was typically c. 70% and host cell protein and DNA were reduced to acceptable levels (FIG. 2C). The low levels of host cell DNA achieved, and the maintenance of acceptable binding capacity, were particularly surprising in view of the lack of a DNA precipitation step prior to the depth filter clarification. Efficient recovery of adenovirus from such a direct load AEX has not, to the inventors' knowledge, previously been disclosed. The ability to achieve this enabled the simplification of the DSP from four steps to two or three. This efficient recovery with a direct load AEX method is even more surprising given than no DNA precipitation step was performed before the depth filtration step. Without such a DNA precipitation step, it would be expected that the host cell DNA could interfere with the virus binding in the AEX, such that the host cell DNA bound competitively to the AEX membrane, and/or formed undesirable aggregates with the virus.

[0300] The simplicity and high productivity of the method of the invention are beneficial from an economic perspective. Process economic modelling of drug substance manufacturing suggests a cost of goods of <USD 2 per dose (FIG. 2D). Final prices may be somewhat higher due to additional costs, for example fill and finish.

Example 3: Distributed Manufacturing

[0301] COVID-19 poses a unique challenge for vaccine manufacturing, both due to the volume and speed required, and due to concerns about equitable vaccine distribution and so-called vaccine nationalism. Distributed manufacturing (manufacturing the same vaccine across multiple facilities in different countries) is a potential solution to these issues, but requires a readily transferable process (Gomez, P. L. and J. M. Robinson, Vaccine Manufacturing. Plotkin's Vaccines, 2018: p. 51-60.e1). Our process is well-suited to this strategy as it uses unit operations which are standard across the bioprocess industry, single-use product-contact materials throughout, and a viral vector of good biosafety (BSL1-2, dependent upon jurisdiction). Many contract manufacturing organisations (CMOs) are already equipped to execute such processes with little or no capital expenditure. We therefore pursued a distributed manufacturing strategy from an early stage of the pandemic. ChAdOx1 nCoV-19 (sometimes referred to as AZD1222 drug substance) is now being manufactured at 1000 L scale in facilities in multiple countries (FIG. 2E).

[0302] We refer to FIG. 2 which shows Improved process, enabling cost-effective large-scale distributed manufacture

[0303] Panel A shows schematics of previous and revised DSPs. The dashed box indicates the feasibility of execution of depth filter clarification and AEX as a single unit operation. Previous DSP means known DSP (i.e. not part of the invention). Revised DSP means part of the invention.

[0304] Panel B shows chromatogram obtained when running in-line depth filter clarification and AEX at 10 L scale. Absorbance at 280 nm is shown (line between 4-5 arbitrary units from 50-100 minutes), alongside conductivity (other line). For schematic of process skid, see FIG. 3E. A 150 mL Sartobind Q capsule was loaded at 3.810.sup.13 VP per mL of membrane. Numerals indicate stages: 1=loading, 2=wash, 3=elution, 4=1M sodium hydroxide sanitisation.

[0305] Panel C shows product recovery and quality from the 10 L process shown in Panel B and after final formulation by TFF.

[0306] Panel D tabulates modelled costs of drug substance production at 200 L scale. This includes capital for purchase of all equipment needed for the process; many facilities do not require this.

[0307] Panel E shows, in red, countries in which ChAdOx1 nCoV-19/AZD1222 is currently being manufactured.

[0308] We refer to FIG. 3 which shows small-scale optimisation of anion exchange with direct loading of clarified lysate.

[0309] Panel A shows gradient elution of ChAdOx1-luciferase from Sartobind Q anion exchange membrane. Filtered lysate containing 910.sup.13 VP of ChAdOx1-luciferase was loaded onto a 3 mL Sartobind nano Q capsule, followed by elution with a gradient of increasing salt concentration. Two peaks were observed (chromatogram) and analysed. Coomassie-stained SDS-PAGE, with comparison to virus purified by caesium chloride gradient ultracentrifugation (+C), shows that peak 1 contains impurities (notably free hexon protein) while Peak 2 contains virus. This was corroborated by infectivity assay. n.d. indicates not detected.

[0310] Panel B shows initial estimation of binding capacity and product recovery by step elution. Clarified lysate containing 210.sup.14 VP of ChAdOx1 nCoV-19 was loaded onto a 3 mL Sartobind nano Q capsule. Binding capacity and quantity of product bound was determined by collection of serial fractions of flowthrough during loading (step 1). After washing with equilibration buffer (step 2), product was eluted with a step to 39-40 mS/cm (step 3); steps 4 and 5 indicate regeneration with 1M NaCl and 1M NaOH respectively. Flowthrough and elution fractions were analysed by qPCR to calculate binding capacity and recovery: 10% breakthrough occurred at a load of 3.410.sup.13 VP per mL of bed volume; 90% of bound product was recovered.

[0311] Panel C shows optimisation of salt concentration/conductivity during loading and washing. Three runs were performed, in each of which filtered lysate containing 210.sup.14 VP of ChAdOx1 nCoV-19 was loaded onto a 3 mL Sartobind nano Q capsule, after adjustment of conductivity to the indicated values by addition of salt. Each run used wash buffer with conductivity matching the load. The chromatogram overlays results from the three runs. Eluates were analysed by Coomassie-stained SDS-PAGE, qPCR and HCP ELISA. Recovery and HCP are tabulated: the basal condition (24-25 mS/cm) achieved nearly 2log.sub.10 reduction in HCP; increasing the conductivity of the load and wash achieved modest improvement in HCP clearance, with substantially reduced recovery with loading at 30-31 mS/cm. As virus in the flowthrough was not quantified, recovery is shown as % of loaded product (rather than bound product, as in panel B); the low recovery for the basal condition here thus reflects the overloading of the column.

[0312] Panel D shows initial results obtained using a 150 mL Sartobind Q capsule with relatively low load challenge. Clarified lysate containing 7.410.sup.14 VP of ChAdOx1 nCoV-19 was loaded, using conditions as indicated, based upon the results of the experiment shown in Panel C. The eluate was analysed by qPCR, HCP ELISA and infectivity assay. Recovery as % of loaded product, HCP and P:I ratio are tabulated.

[0313] Panel E illustrates system used for in-line clarification and anion exchange (P: pump; DF: depth filter; F: 0.2 m filter), results of which are shown in FIG. 2B-C.

[0314] Viruses

[0315] The ChAdOx1 nCoV-19, ChAdOx1 Lassa-GP, ChAdOx1 luciferase and ChAdOx2 GFP vectors used here have previously been described.sup.1-4. Virus used as seed to infect shake flask cultures and as standards in quality control assays was produced by caesium chloride density-gradient ultracentrifugation by the Jenner Institute Viral Vector Core Facility. Virus used as seed to infect 50 L and 200 L cultures was prepared using our previously described process in 3 L shake flasks or bioreactors, up to the point of the first tangential flow filtration (TFF) step 5. After this the concentrated and diafiltered lysate was aliquoted and frozen at 80 C.

[0316] Cells and Upstream Process

[0317] HEK293 T-rex cells (ThermoFisher) were banked and adapted to low-serum suspension culture in CD293 medium (ThermoFisher) as previously described 5. Cells were then adapted to increasing proportions of BalanCD293 medium (Fujifilm-Irvine Scientific), supplemented with 4 mM GlutaMAX (ThermoFisher), over one week.

[0318] For upstream process experiments, seed culture at 2 the specified final density was diluted by addition of 1 volume of fresh medium to reach the cell density specified for each experiment at the point of infection. A multiplicity of infection of 10 was used unless otherwise stated.

[0319] Each feed was with 0.05 volumes of BalanCD293 feed (Fujifilm-Irvine Scientific). Pre-infection, cultures were fed on the day cell density exceeded 110.sup.6 cells/mL. Cultures for which the intended cell density at the point of infection was 310.sup.6 cells/mL received a second pre-infection feed when cell density exceeded 410.sup.6 cells/mL. Post-infection, all cultures were fed at 0.5 h and 22 h after infection.

[0320] Shake flask experiments were performed in Erlenmeyer flasks (Corning), with a working volume of 25-35 mL in a 125 mL flask unless otherwise stated. BioBlu 3c and 14c (Eppendorf) single-use bioreactor vessels were used in accordance with the manufacturers' instructions. A GX bioreactor controller unit and C-BIO software (both from Global Process Control) were used to control both vessel types. Dissolved oxygen (DO) was regulated at a setpoint of 55% air saturation by addition of medical air via macrosparger. pH was regulated in the range 7.2-7.3 as previously described 5.

[0321] 50 L and 200 L upstream processes were performed using Pall Allegro stirred tank reactors (STRs).

[0322] Unless otherwise stated, bioreactors were seeded at 0.4-0.610.sup.6 cells/mL in c. 35% of the maximum working volume. Antifoam C emulsion (SigmaAldrich) was used in 50 L and 200 L STRs. 0.05 culture volumes of BalanCD feed was added when the density reached 1.010.sup.6 culture cells/mL. At a cell density of 4.010.sup.6/mL (range 3.0-6.010.sup.6), cells were diluted with 1 volume of medium and infected, using an MOI of 10 unless otherwise stated. 0.05 volumes of BalanCD feed were added 30 minutes after infection, and again after 22 hours.

[0323] Lysis, Nucleic Acid Digestion and Clarification

[0324] Lysis was performed as previously described.sup.5, in the culture vessel, with the exception that the concentration of Benzonase (MerckMillipore) was reduced to 15 units/mL. No DNA precipitation step was performed before the clarification by depth filtration. Lysis was initiated at 42-48 h after infection, with the exception of the productivity kinetic experiments shown in FIGS. 1C-D. Two hours after addition of lysis buffer, clarification was initiated, using Millistak+ HC Pro CoSP depth filters as in our previous work 5. During 200 L runs, an Advanced MVP skid (Pall Biotech) was used for filtration steps.

[0325] Tangential Flow and Bioburden Reduction Filtration

[0326] Tangential flow filtration was performed essentially as we have previously described 5, scaled appropriately and with the following modifications. Where TFF was performed before anion exchange (AEX), i.e. for the 200 L run producing product as reported in FIG. 1G, only 2-fold concentration was performed, prior to 6 diavolumes of diafiltration. For TFF after AEX, Omega T-series 300 kDa cut-off flat sheet filters (Pall Biotech) were used. For TFF during 200 L runs, an Allegro CS 4500 single-use TFF skid was used (Sartorius). A Supor EKV 0.2 m filter was used for bioburden reduction filtration after the final TFF.

[0327] Anion Exchange Chromatography

[0328] Where preceded by TFF (run reported in FIG. 1G), AEX was performed as previously reported 5, with scaling of the chromatography capsule and buffer volumes based upon anticipated binding capacity of 710.sup.13 VP per mL of membrane volume.

[0329] For direct-load AEX (loading clarified lysate), small-scale studies (FIG. 3) were performed using an Akta Pure (GE) and 3 mL bed volume/8 mm bed height SartobindQ Nano capsules (Sartorius). Equilibration buffer comprised 20 mM Tris-HCL pH8.0, 1 mM MgCl.sub.2, 0.1% v/v polysorbate 20, 5% w/v sucrose; this was also used as a base for wash buffers. Elution buffer comprised 20 mM Tris-HCL pH8.0, 1 mM MgCl.sub.2, 0.1% v/v polysorbate 20, 5% w/v sucrose, 600 mM NaCl, except where salt concentration was varied, as stated. Adjustment of the conductivity of the sample and wash buffers, to target values as stated in the descriptions of individual experiments, was performed using 5M NaCl (Sigma).

[0330] Column equilibration was in accordance with the manufacturer's instructions. After loading, capsule was washed with 10 MV of equilibration buffer before elution step (both at 5 MV/min).

[0331] For the direct-load AEX purification from a 10 L bioreactor (FIG. 2B) a peristaltic pump-driven rig was constructed, as shown in FIG. 3E, incorporating a CoSP depth filter (as above), Millipak-20 0.2 m filter, and 150 mL/8 mm bed height Sartobind Q capsule (Sartorius), plus single-use UV absorbance, conductivity and pressure sensors (Pendotech). Buffers, column equilibration, sample loading, washing and elution were as described above, with the exception that a flow rate of 0.7 membrane volumes/minute was used for sample loading, washing and elution.

[0332] Product Quantification and Assessment of Product Quality

[0333] Product quantification was as previously reported, using qPCR and UV spectrophotometry assays for viral particles in impure and pure samples respectively, and an immunostaining-based infectivity assay 5.

[0334] Residual host-cell protein (HCP) was quantified using the HEK293 HCP ELISA kit (Cygnus Technologies), according to the manufacturer's instructions. Residual host cell DNA was quantified using a previously reported quantitative PCR method targeting a 94 base pair amplicon within the Alu repeats.sup.6. The lower limit of quantification was 100 pg/mL for intact HEK293 DNA.

REFERENCES TO EXAMPLES 1, 2 AND 3

[0335] 1 van Doremalen, N. et al. ChAdOx1 nCoV-19 vaccine prevents SARS-CoV-2 pneumonia in rhesus macaques. Nature, doi:10.1038/s41586-020-2608-y (2020). [0336] 2 Purushotham, J., Lambe, T. & Gilbert, S. C. Vaccine platforms for the prevention of Lassa fever. Immunol Lett 215, 1-11, doi:10.1016/j.imlet.2019.03.008 (2019). [0337] 3 Dicks, M. D. et al. Differential immunogenicity between HAdV-5 and chimpanzee adenovirus vector ChAdOx1 is independent of fiber and penton RGD loop sequences in mice. Sci Rep 5, 16756, doi:10.1038/srep16756 (2015). [0338] 4 Morris, S. J. S., Sarah; Spencer, Alexandra J.; Gilbert, Sarah C. Simian adenoviruses as vaccine vectors. Future Virology 11, 649-659, doi:10.2217/fvl-2016-0070 (2016). [0339] 5 Fedosyuk, S. et al. Simian adenovirus vector production for early-phase clinical trials: A simple method applicable to multiple serotypes and using entirely disposable product-contact components. Vaccine, doi:10.1016/j.vaccine.2019.04.056 (2019). [0340] 6 Zhang, W. et al. Development and qualification of a high sensitivity, high throughput Q-PCR assay for quantitation of residual host cell DNA in purification process intermediate and drug substance samples. J Pharm Biomed Anal 100, 145-149, doi:10.1016/j.jpba.2014.07.037 (2014).

Example 4: 50 L Stirred Tank Bioreactor Production & Purification of ChAdOx1 nCoV-19

[0341] A 40 L culture was infected and harvested in a 50 L STR, as previously described (Fedosyuk et al 2019 Vaccine vol 37 pages 6951-6961 (doi: 10.1016)), with the exception of the use of the BalanCD293 medium/feed strategy, with a starting cell density of 2e6 cells/mL (diluted from 4e6 cells/mL) and an MOI of 10.

[0342] The product was subsequently purified using a strategy as previously described (Fedosyuk et al 2019 Vaccine vol 37 pages 6951-6961 (doi: 10.1016)). The final purified product yield was 1.810.sup.11 VP per mL of the upstream process (FIG. 1 E-G).

Example 5: Attainment of High Purity and Recovery by Simplified Downstream Process

[0343] We refer to FIG. 2 and FIG. 3.

Example 6: Comparative Data

[0344] An important advantage of the method of the invention (for example with direct AEX loading of clarified lysate) is that it allows reduction of the complexity of the purification process (downstream process) to two or three unit operations (as shown in FIGS. 2A & 3E). This is compared to known methods which require four unit operations (as shown in FIG. 2A). We refer also to the disclosure of Fedosyuk et al 2019 (Vaccine vol 37 pages 6951-6961) FIG. 5Ait should be noted that this known Fedosyuk process also contains a fourth [clarification] step before the three steps shown in FIG. 5A of Fedosyuk et al.

[0345] A further advantage of the method of the invention is higher product recovery compared to known methods. This is supported by the AEX step recovery data shown in FIGS. 2C and 3A/D (80, 83, 93%). (It should also be noted that the lower figures in FIG. 3C are not representative of the achievable recovery as the columns were over-loaded in this example.) This may be compared to Fedosyuk et al 2019 (Vaccine vol 37 pages 6951-6961) Table 2 which shows that up to 20% loss may occur in the TFF1 step of the known method. By comparison, this known TFF1 step has been eliminated from the method of the invention.

[0346] A further advantage which may be ascribed to the method of the invention is that losses on the known AEX step were greater than they are on the AEX step of the method of the invention. For comparison, we refer to Fedosyuk et al 2019 (Vaccine vol 37 pages 6951-6961) Table 2 compared to the recovery on the new AEX shown in the Figures herein.

[0347] Although illustrative embodiments of the invention have been disclosed in detail herein, with reference to the accompanying drawings, it is understood that the invention is not limited to these precise embodiments and that various changes and modifications can be effected therein by one skilled in the art without departing from the scope of the invention as defined by the appended claims and their equivalents.