Evaluation of an alkali stable Protein A matrix versus Protein A Sepharose Fast Flow and considerations on process scale-up to 20,000L.

Transcription

1 Evaluation of an alkali stable Protein A matrix versus Protein A Sepharose Fast Flow and considerations on process scale-up to 20,000L. Simon C.B. Jones & Martin P. Smith. LONZA, 224 Bath Road, Slough, SL1 4DX. Presented at, 3 rd International Symposium on Downstream Processing of Genetically Engineered Antibodies and Related Molecules, Nice, France, October 2004.

2 Introduction 2004 LONZA brings on-line 3x20,000L stirred tank mammalian cell culture reactors. Complimenting: 2x5k, 2x2k Air-lift in US 2x2k, 2x200L Air-lift in UK Multi-product facilities. Different operational scales within same site. unique challenges to both operations and process development. 28/09/2004 / 2

3 Historical application of Protein A LONZA s core expertise with production of recombinant monoclonal antibodies. Contract manufacture set-up. Varying mix of purification process designs. >95% employ Protein A as affinity capture. Experienced with compressible and incompressible matrices Wide variety of Protein A Sepharose matrices in use Rmp, recombinant, native Sepharose 4 FF. 28/09/2004 / 3

4 Protein A Sepharose 4 Fast Flow In widespread use high binding capacities, but flow rate limited Highly compressible Minimum Residence time 6min (150cm/h, 15cm Ho) Historically PnA Sepharose cleaned with 6M Guanidine- HCl Pressure Drop (psi) Linear velocity (cm/h) Pressure Drop versus Flowrate Column Diameter 1.6cm 2.6cm 5.0cm 10cm 20cm 28cm 40cm 140cm Stickel JJ, Fotopoulos A. Biotechnol Prog Jul-Aug; 17(4): Packed bed Height = 15cm Matrix = Sepharose 4 Fast Flow Buffer 22 C 28/09/2004 / 4

5 Issues associated with 6M Guanadine-HCl regeneration Materials Handling (Large Scale) 6M Gu-Hcl FW=573 g/l. 2.0m column, 15cm Bed height CV = 462L. 2CV of 6M Gu-HCl (942L) = 539kg. Logistical considerations, disposal, health and safety. Buffer preparation and solution properties Buffer preparation is ENDOTHERMIC. Jacketed prep and hold tanks are required. Solution has very high viscosity & density. Corrosive-tanks, pipe work, columns, terminators. 28/09/2004 / 5

6 Importance of Materials of Construction LSBO Purification suite P6B Purification suite P6-B 1.4m diameter stainless steel column to rear 2m diameter stainless steel chromatography column 28/09/2004 / 6

7 Regeneration Process Economics Considering Large Scale Operations 20,000L at 1g/L, 1.4m column, 15cm bed height 150cm/h, DBC=21g/L, 5 cycles/batch 1 CIP cycles per 5 cycles Guandine-HCl cleaning costs during campaign ( ) 1,400,000 1,200,000 1,000, , , , ,000 As much spent on cleaning after 140 cycles as on procurement of matrix Purification cycles 28/09/2004 / 7

8 Current challenges Fermentation titres are increasing, increased productivity changes ratio of fermentation to purification batches. LONZA currently operates capacity limited by fermentation throughput. Increase in number of fermentation systems per site in line with customer requirements and business growth. Downstream processing optimisation required. Eliminate capacity bottlenecks before they occur. operational improvements. new purification technologies. 28/09/2004 / 8

9 LONZA-Amersham Collaboration Goal Evaluate latest developments in chromatography matrices. LONZA experienced in antibody purification Assess to high producing cell lines Industrially relevant feed steams Strong analytical support for technology development projects Amersham are a leading developer of chromatography matrices Supply resource for experimentation 28/09/2004 / 9

10 Tracking advances in Matrix Development Agarose from sea weed to separation media Gracilaria and Gelidium red algae Amersham Matrix launches Protein A Seph 4FF 1996 rprotein A Seph 4FF 2000 rmp Protein A Seph 4FF 2001 MabSelect 2005.MabSelect SuRe, Xtra 28/09/2004 / 10

11 MabSelect SuRe Developed from MabSelect technology. Designed to withstand harsh CIP agents (NaOH) without losing its functional properties. Claims towards. Increased Cost effectiveness. Less carry over, purer product. Extended life length. 28/09/2004 / 11

12 Staphylococcal protein A Immunoglobulin binding domains Ss E D A B C XM Gly29Ala mutation Z domain Nilsson et al. (1987) Protein Engineering, vol 1, pp /09/2004 / 12

13 MabSelect SuRe design The ligand is based on a genetically engineered variant of the Z domain where a number of Aspargines have been replaced with other amino acids. The ligand is a tetramer construct with four identical domains. The ligand is single point attached to MabSelect base matrix. 28/09/2004 / 13

14 Evaluation Project Stage 1 Review of dynamic Binding Capacity (DBC) pure IgG 4 as function of residence times. Stage 2 Determine DBC at maximum recommended flow rate and bed height using cell culture supernatant (protein free GS-NS0) Stage 3 Assess contaminant and impurity clearances at operational capacities. Review process performance and product quality. Stage 4 Resin re-use exploring alkali stability over 100 cycles 28/09/2004 / 14

15 Stage 2 Breakthrough capacity determination on IgG 4 ex Cell Culture Supernatant rmppna Seph 4 FF Mabselect Prototype SuRe Rmp Protein A 4 FF XK16, 15cm Ho, 150cm/h, 6min RT 5% C/Co = 26g/L Op.DBC = 20.8g/L Breakthrough (C/Co) % Antibody Loaded (mg/ml) MabSelect TC5, 20cm Ho, 500cm/h, 2.4min RT 5% C/Co = 19g/L Op.DBC = 15.2/L MabSelect SuRe TC5, 20cm Ho, 500cm/h, 2.4min RT 5% C/Co = 26g/L Op.DBC = 20.8g/L 28/09/2004 / 15

16 Protein A Step throughput analysis Single batch SuperPro v5.5 process model 2000L fermenter, 0.5g/L titre Elution Wash 1 Eqm Strip Batch S-101 1x CCS Load / V-101 PnA / C-101 Storage PBA Chromatography PnA Waste Column diameter fixed at 30cm Bed Height 15cm rmp.protein A 20cm MabSelect, SuRe Flowrates (all steps) 150cm/h rmp.protein A 500cm/h MabSelect, SuRe Buffers 5CV Eqm, Wash 4CV Elution, 3CV Strip, 2CV CIP 15min contact NaOH, 60min contact Gu-HCL. 28/09/2004 / 16

17 Rmp protein A Sepharose 4 Fast Flow throughput modelling single batch day h Operations Gantt Chart (Single Batch) - rmppna.spf day h Overview Complete Recipe Load in V-101 CHARGE-1 ANSFER-OUT-1 PnA in C-101 PnA (Cycle #1 ) LOAD-1 CHARGE-1 (0.00 h) LOAD-1 (3.79 h) TRANSFER-OUT-1 (25.37 h) 5x 5.5h cycles 27.4h Step time WASH-1 WASH-1 (0.50 h) ELUTE-1 ELUTE-1 (0.40 h) REGENERATE-1 EQUILIBRATE-1 PnA (Cycle #2 ) REGENERATE-1 (0.20 h) EQUILIBRATE-1 (0.50 h) 901L Buffers consumed LOAD-1 LOAD-1 (3.79 h) WASH-1 WASH-1 (0.50 h) ELUTE-1 REGENERATE-1 EQUILIBRATE-1 ELUTE-1 (0.40 h) REGENERATE-1 (0.20 h) EQUILIBRATE-1 (0.50 h) Resin utilisation = 91% PnA (Cycle #3 ) LOAD-1 LOAD-1 (3.79 h) WASH-1 WASH-1 (0.50 h) ELUTE-1 ELUTE-1 (0.40 h) REGENERATE-1 REGENERATE-1 (0.20 h) EQUILIBRATE-1 EQUILIBRATE-1 (0.50 h) PnA (Cycle #4 ) LOAD-1 LOAD-1 (3.79 h) WASH-1 WASH-1 (0.50 h) ELUTE-1 ELUTE-1 (0.40 h) REGENERATE-1 REGENERATE-1 (0.20 h) EQUILIBRATE-1 EQUILIBRATE-1 (0.50 h) PnA (Cycle #5 ) LOAD-1 LOAD-1 (3.79 h) WASH-1 WASH-1 (0.50 h) ELUTE-1 ELUTE-1 (0.40 h) REGENERATE-1 REGENERATE-1 (0.20 h) EQUILIBRATE-1 EQUILIBRATE-1 (0.50 h) P-1 in C-101 CIP-1 CIP-1 (1.00 h) 28/09/2004 / 17

18 MabSelect throughput modelling single batch day h Operations Gantt Chart (Single Batch) - Mab.spf day h Overview Complete Recipe Load in V-101 CHARGE-1 ANSFER-OUT-1 PnA in C-101 CHARGE-1 (0.00 h) TRANSFER-OUT-1 (8.61 h) 5x 1.9h cycles PnA (Cycle #1 ) LOAD-1 WASH-1 LOAD-1 (1.14 h) WASH-1 (0.15 h) 9.7h Step time ELUTE-1 ELUTE-1 (0.12 h) REGENERATE-1 REGENERATE-1 (0.06 h) CIP-1 EQUILIBRATE-1 PnA (Cycle #2 ) CIP-1 (0.25 h) EQUILIBRATE-1 (0.15 h) 1414L Buffers consumed LOAD-1 LOAD-1 (1.14 h) WASH-1 WASH-1 (0.15 h) ELUTE-1 REGENERATE-1 CIP-1 EQUILIBRATE-1 ELUTE-1 (0.12 h) REGENERATE-1 (0.06 h) CIP-1 (0.25 h) EQUILIBRATE-1 (0.15 h) Resin utilisation = 93% PnA (Cycle #3 ) LOAD-1 LOAD-1 (1.14 h) WASH-1 WASH-1 (0.15 h) ELUTE-1 ELUTE-1 (0.12 h) REGENERATE-1 REGENERATE-1 (0.06 h) CIP-1 CIP-1 (0.25 h) EQUILIBRATE-1 EQUILIBRATE-1 (0.15 h) PnA (Cycle #4 ) LOAD-1 LOAD-1 (1.14 h) WASH-1 WASH-1 (0.15 h) ELUTE-1 ELUTE-1 (0.12 h) REGENERATE-1 REGENERATE-1 (0.06 h) CIP-1 CIP-1 (0.25 h) EQUILIBRATE-1 EQUILIBRATE-1 (0.15 h) PnA (Cycle #5 ) LOAD-1 LOAD-1 (1.14 h) WASH-1 WASH-1 (0.15 h) ELUTE-1 ELUTE-1 (0.12 h) REGENERATE-1 REGENERATE-1 (0.06 h) CIP-1 CIP-1 (0.25 h) EQUILIBRATE-1 EQUILIBRATE-1 (0.15 h) 28/09/2004 / 18

19 MabSelect SuRe throughput modelling single batch day h Complete Recipe Load in V-101 CHARGE-1 ANSFER-OUT-1 PnA in C-101 PnA (Cycle #1 ) LOAD-1 WASH-1 ELUTE-1 REGENERATE-1 CIP-1 EQUILIBRATE-1 PnA (Cycle #2 ) LOAD-1 WASH-1 ELUTE-1 REGENERATE-1 CIP-1 EQUILIBRATE-1 PnA (Cycle #3 ) LOAD-1 WASH-1 ELUTE-1 REGENERATE-1 CIP-1 EQUILIBRATE-1 PnA (Cycle #4 ) LOAD-1 WASH-1 ELUTE-1 REGENERATE-1 CIP-1 EQUILIBRATE-1 Operations Gantt Chart (Single Batch) - SuRe.spf CHARGE-1 (0.00 h) TRANSFER-OUT-1 (7.88 h) LOAD-1 (1.42 h) WASH-1 (0.15 h) ELUTE-1 (0.12 h) REGENERATE-1 (0.06 h) CIP-1 (0.25 h) EQUILIBRATE-1 (0.15 h) LOAD-1 (1.42 h) WASH-1 (0.15 h) ELUTE-1 (0.12 h) REGENERATE-1 (0.06 h) CIP-1 (0.25 h) EQUILIBRATE-1 (0.15 h) LOAD-1 (1.42 h) WASH-1 (0.15 h) ELUTE-1 (0.12 h) REGENERATE-1 (0.06 h) CIP-1 (0.25 h) EQUILIBRATE-1 (0.15 h) LOAD-1 (1.42 h) WASH-1 (0.15 h) ELUTE-1 (0.12 h) REGENERATE-1 (0.06 h) CIP-1 (0.25 h) EQUILIBRATE-1 (0.15 h) Overview day h 4x 2.2h cycles 8.9h Step time 1131L Buffers consumed Resin utilisation = 85% Room for further reduction in column diameter to gain maximum process efficiency 28/09/2004 / 19

20 Stage 3. Purification evaluation at operational capacities. Operating Flow rate of 500cm/h, Tricorn 5x20 (4mL) Equil 10mM NaPO4, 150mM NaCl, ph 7.2, 5CV Load 1x Clarified Cell Culture Supernatant 0.5g/L titre, loaded to 1%C/C 0-20% Wash Elution Strip 10mM NaPO4, 150mM NaCl, ph 7.2, 5CV 100mM Glycine, ph 3.5, 4CV 100mM Citric Acid, ph 2.1, 3CV 28/09/2004 / 20

21 Stage 3. Purification Assessment Criteria Process Analysis: Chromatogram comparisons Yields, Elution volumes, concentrations Impurity and Contaminant Clearance Protein A ligand leakage Elute DNA, HCP levels Product Quality (SDS, IEF, gphplc) 28/09/2004 / 21

22 Chromatographic Performance Yields: Mabselect 100% MabSelect SuRe 100% Eluate Volume (CV) MabSelect 1.65 MabSelect SuRe 1.87 Eluate Conc. (mg/ml) MabSelect 9.75 MabSelect SuRe /09/2004 / 22

23 MabSelect SuRe purification by SDS-PAGE Reduced Matrices were identical by Reduced SDS-PAGE IgG standard MabSelect MabSelect Xtra MabSelect SuRe 28/09/2004 / 23

24 MabSelect SuRe purification by SDS-PAGE Non-Reduced Matrices were identical by non-reduced SDS IgG standard MabSelect MabSelect Xtra MabSelect SuRe 28/09/2004 / 24

25 MabSelect SuRe eluate product quality by Isoelectric Focusing 28/09/2004 / 25 MabSelect MabSelect Xtra MabSelect SuRe pi markers

26 Contaminant and Impurity Clearances Host Cell Protein The three protein A Eluate samples were comparable. Lane 2 MabSelect eluate Lane 4 MabSelect SuRe Lane 4 MabSelect Xtra Lane 8 Ab. reference Lane 10 - Markers Lane 11 HCP IAC Lane 12 BSA/IgG IAC 28/09/2004 / 26

27 Contaminant and Impurity Clearances CHO DNA and Protein A ligand DNA MabSelect and MabSelect SuRe both <1000 pg/mg Protein A ligand MabSelect MabSelect SuRe 20ng/mg 5.4ng/mg Conclusions from Stage 3 purification assessment MabSelect SuRe equalled or exceeded purification capabilities of MabSelect, rmp Protein A 4FF 28/09/2004 / 27

28 MabSelect SuRe Resin Reuse Evaluation Small scale re-use study Tricorn 5x10, 2mL column, Flow rate 200cm/h Loaded w/50ml 1xCCS, 0.55g/L to 14g/L capacity- material constrained Equil, 5CV, 50mM NaPO4, 150mM NaCl, ph 7.4 Wash 5CV, 50mM NaPO4, 150mM NaCl, ph 7.4 Elute 4CV, 100mM Glycine ph 3.5 Strip 2CV, 500mM Acetic Acid+0.1M NaSO4 Wash 1CV, 50mM NaPO4, 150mM NaCl, ph 7.4 CIP 5CV, 0.1M NaOH, 15min contact time 28/09/2004 / 28

29 MabSelect SuRe re-use; Analysis 100 cycles in total Process Analysis: Chromatogram profile Yield Full breakthrough capacity rechecked every 25 th cycle Impurity and Contaminant Clearance Protein A ligand leakage Elute DNA, HCP levels 28/09/2004 / 29

30 MabSelect SuRe. Re-use Performance Recovery vs. Reuse Capacity vs. Reuse Recovery (%) Decrease in 10% breakthough capacity (%) Dynamic Binding Capacity 10% C/Co (mg/ml) Purification and CIP cycles Purification and CIP cycles 28/09/2004 / 30

31 MabSelect SuRe Re-use, Protein A ligand leakage Ligand leakage from prototype SuRe during 100 IgG 4 purification cycles Conc. [ng/mg] CIP / Purification cycles 28/09/2004 / 31

32 MabSelect SuRe Re-use, HCP clearance Key Lane Description 1 MW marker 2 BSA/IgG IAC 4 HCP IAC 6 SuRe Re-Use cycle 1 7 SuRe Re-Use cycle 5 8 SuRe Re-Use cycle 10 9 SuRe Re-Use cycle SuRe Re-Use cycle SuRe Re-Use cycle SuRe Re-Use cycle SuRe Re-Use cycle SuRe Re-Use cycle SuRe Re-Use cycle SuRe Re-Use cycle SuRe Re-Use cycle Carry over 18 Carry over after Load material 28/09/2004 / 32

33 Summary of MabSelect SuRe Reuse study Apparent 20% decrease in capacity measured at 10% breakthrough after 100 cycles No decrease in yield Acceptable levels of ligand leakage Consistent HCP clearance across cycles No trend in DNA clearance Overall acceptable performance Could performance be further increased by reducing frequency of clean (every 2 nd -5 th cycles) 28/09/2004 / 33

34 Conclusions MabSelect SuRe found to maintain rmp Protein A capacities at 3x higher flow rates Significant improvement in Protein A capture Increased throughputs from 27h processing to 9h Increased Protein A stability to caustic cleaning, no significant impurity/contaminant clearance issues Acceptable levels of purification across 100 cycles of re-use with 0.1M NaOH with 15min contact time Vendor collaboration resulted in successful partnership. 28/09/2004 / 34

35 Acknowledgements Amersham Biosciences Hans J Johansson Aman Mottaqui-Tabar Anders Ljunglof Mats Ramberg Andy Masters Stuart Aitken 28/09/2004 / 35