Home

2 - GE Healthcare Life Sciences

Contents

1. Flux LMH 8 8 O75 40 1425 1450 175 20 2 250 275 075 100 125 150 f 175 20 235 250 275 TMP bar TMP bar Fig 6 5 TMP excursion results The curves were created by the KTAcrossflow Evaluation Wizard from the raw data generated during the TMP scouting runs The optimal TMP ensures operation in the transition region between the pressure dependent and independent region for a robust process and so a TMP of 1 8 bar was chosen 114 KTAcrossflow Method Handbook 11 0012 36 Edition AB Applications Ce 6 3 5 Diafiltration time optimization To minimize the diafiltration time the optimal concentration for diafiltration was established by analysing the relationship between flux and concentration factor A five times concentration process was run at the optimized crossflow and TMP settings and a plot of the Flux concentration factor versus concentration factor was generated using the Evaluation Wizard The concentration factor that gives the highest value of the Flux concentration factor will give the shortest diafiltration time Fig 6 6 shows that the diafiltration will take the same amount of time whether it is performed at a four or five times concentration This is because the decrease in retentate volume at five times concentration is offset by the decrease in flux However since the buffer consumption is lower at five times concentration as the
2. Microfiltration cartridges are shipped dry and can be used immediately after wetting out with buffer Ultrafiltration cartridges contain an alcohol glycerin storage solution and must be thoroughly rinsed before use You can i feed and retentate poi on any hollow fiber cartridge Feed port inlet Retentate port outlet Permeate port Permeate port Feed port inlet Permeate port Permeate port Fig 1 6 Start AXM and Start AXH cartridges 22 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Introduction 1 10 CFF filter life cycle The long term stability of a used filter may vary depending on many factors including the following e Components in the process solution e Processing conditions e Aggressiveness of the cleaning protocols e Handling and storage conditions Typically the performance of the filter is checked before and after use by measuring the rate of water flow through the membrane under controlled conditions When the performance of the filter drops to unacceptable levels the filter should be replaced Fig 1 7 At the laboratory scale some users dispose of membrane filters after each use This avoids the use of cleaning chemicals and their disposal cleaning time and the possibility for cross contamination from the filter Air diffusion and bubble point tests are normally only completed when using pilot and production scale equipment Fail test Dispose of filter Fail test Fail test Dispose
3. Fill Sample gt Fig 2 18 Result file Detailed descriptions of creating and running a method and of evaluation the results can be found in AKTAcrossflow User Manual and the UNICORN 5 0 Evaluation for Cross Flow Filtration User Reference Manual 54 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cross flow filtration process considerations 3 Cross flow filtration process considerations 3 1 Factors influencing product yield The amount of product that can be recovered from a process step represents the product yield Increasing yield in laboratory processes ensures maximum product for testing efficient use of lab resources and accurate projections for scaling to pilot equipment 3 1 1 General considerations The system tubing design can affect product yield if it impedes recovery of the liquid For example every CFF system has a hold up volume tubing volume but well designed systems minimize this volume enabling maximum product recovery Poorly designed systems include long tubing runs poor tank drainage and other non recoverable volume such as poorly positioned drain valves Two methods can be used to recover most of the hold up volume from AKTAcrossflow system One method maximizes product yield at the expense of concentration and the other enables the highest concentration to be achieved at the expense of some yield 3 1 2 Measuring yield When yield becomes important the appropriate
4. e Method Editor to create and edit methods for pre programmed system control e System Control to monitor processes on line e Evaluation to evaluate and present stored results It also includes the Evaluation Wizard specifically designed for AKTAcrossflow result files AKTAcrossflow Method Handbook 11 0012 36 Edition AB 39 AKTAcrossflow system components and software 2 4 3 Common interface UNICORN provides a common control platform and one common user interface for all scales of operation In addition there is the same familiar interface for both chromatography and membrane systems which can be controlled and monitored from your office desk with easy to use software wizards 2 System Control 1 Asterix Method x Annika lgG PostProdigGSpec m08 Result T ao va Block Volume Block Time TifFlow FeedFlow RetFlow PermFlow FeedPress RetenPress PermPress Ret ol Perm ol Trans ol uv pH gt gt Cond gt gt DeltaP TMP Flux ConcFactor DF_X Fact Cond FeedFlow RetFlow PermFlow TriFlow RetenPress PermPress pH Temp Curves DeltaP TMP ConeFactor DF_X_Fact RetVol PermVol ConvRatio n 1l Flow Scheme BOJ Transfer Purgevalve Cartridge permeate For Help press F1 OEnd No watch Not Connected Fig 2 8 Common interface for chromatography an
5. Pa UV cell Retentate valve block t S7 _ Conductivity cell a je Valve R PCV ni Permeate y B pressure Transfer purge valve sensor Pp AO Transfer Feed pressure __ valve block 1 sensor Pr Air sensor Feed pump P 984 module A and B Reservoir CFF cassette Retentate Transfer cartridge pressure valve block 2 sensor Pp Fig 2 1 AKTAcrossflow instrument showing key components AKTAcrossflow Method Handbook 11 0012 36 Edition AB 27 2 AKTAcrossflow system components and software Transfer Flow Pressure Restrictor Sensor Transfer Pump Module A Transfer Valve Block1 ni Transfer Purge Valve Pkt Waste 1 Transfer line Air Sensor n2 n3 n4 Retentate Pressure Control Level amp Temperature Sensor Transfer Valve Block 2 Stirrer Permeate Retentate Valve Block Valve Block ns n6 n7 n8 Reservoir Retentate Permeate line Pressure Sensor Recirculation line Feed Permeate Permeate Pressure Pressure Pressure Control Sensor Sensor 7 CHE Permeate Pump Module B Feed Pump Cartridge Out 1 Fig 2 2 Diagram of the general flow scheme in KTAcrossflow Recycle Ou
6. The air sensor is located in the flow path for the sample inlet It is designed to continuously monitor for air bubbles and ensure that the maximum volume of external feed can be transferred into the system without any risk of introducing air into the transfer line When air is detected the system is paused or an action is performed that has been set in the method Avoiding air in the transfer line is important to ensure the high volume accuracy of the transfer pump and thereby the accuracy of the retentate volume content 2 4 UNICORN AKTAcrossflow system is controlled and monitored by UNICORN software UNICORN is a complete package for control and supervision of biotechnical systems It consists of control software and where applicable a controller card or interface unit for interfacing the controlling PC to the liquid handling module 2 4 1 Liquid chromatography system version UNICORN can be used with a number of systems including AKTAdesign liquid chromatography systems For practical reasons the user documentation for AKTAcrossflow also includes the user reference manuals for the UNICORN general liquid chromatography version the examples in the UNICORN User Reference Manual are based on an AKTAexplorer 100 system operating with the 100F400 strategy 2 4 2 Software modules The software consists of four integrated modules e UNICORN Manager for file handling and administration e g definition of systems and user profile etc
7. These tests are often completed by labs in vivo or in vitro For example United States Pharmacopoeia Class VI Plastics Test involves both the implantation and extraction of drug product contact surfaces to demonstrate that these materials are not toxic to various mammalian cells Blinded When a filter is blinded particles have filled the pores and the flow through the filter from the feed side to the permeate side is reduced or stopped AKTAcrossflow Method Handbook 11 0012 36 Edition AB 125 Bubble point The minimum pressure required to overcome the capillary forces and surface tension of a liquid in a fully wetted membrane filter The bubble point value is determined by observing when bubbles first begin to emerge on the permeate downstream side of a fully wetted membrane filter when pressurized with a gas on the feed upstream side of the membrane filter By knowing the surface tension of the liquid the pore size can be determined Bubble point test The test procedure for determining the bubble point of the largest pores in a microfiltration membrane Buffer exchange Filtration process used for the removal of smaller ionic solutes whereby the feed solution is washed usually repeatedly and one buffer is removed and replaced with an alternative buffer Cartridge or A filtration or separation device having a membrane cartridge filter encapsulated within a housing In the case of hollow f
8. amp South East Europe Tel 43 1 982 3826 Fax 43 1985 8327 Denmark Tel 45 16 2400 Fax 45 16 2424 Finland amp Baltics Tel 358 0 9 512 39 40 Fax 358 0 9 512 39 439 France Tel 01 69 35 67 00 Fax 01 69 41 96 77 Germany Tel 0761 4903 490 Fax 0761 4903 405 e Italy Tel 02 27322 1 Fax 02 27302 212 e Japan Tel 81 3 5331 9336 Fax 81 3 5331 9370 e Latin America Tel 55 11 3933 7300 Fax 55 11 3933 7304 Middle East amp Africa Tel 30 210 9600 687 Fax 30 210 9600 693 Netherlands Tel 0165 580 410 Fax 0165 580 401 Norway Tel 815 65 555 Fax 815 65 666 e Portugal Tel 21 417 7035 Fax 21 417 3184 e Russia amp other C I S amp N I S Tel 7 095 232 0250 956 1137 Fax 7 095 230 6377 South East Asia Tel 60 3 8024 2080 Fax 60 3 8024 2090 Spain Tel 93 594 49 50 Fax 93 594 49 55 Sweden Tel 018 612 1900 Fax 018 612 1910 Switzerland Tel 0848 8028 12 Fax 0848 8028 13 UK Tel 0800 616928 Fax 0800 616927 e USA Tel 800 526 3593 Fax 877 295 8102 imagination at work Method Handbook 11 0012 36 AB 01 2006 E Elanders Tofters 2006
9. it is not uncommon to reach 70 wet cell weight while maintaining steady state conditions However lysates tend to need a lower solids level to promote passage of the target material Start with solids in the 5 to 10 range and monitor transmission as well as TMP during the concentration phase Size of the target For separations with large target material high selectivity it may be best to avoid material any concentration but rather perform a constant volume wash from the start Using a more open membrane may require the use of short 30 cm cartridges and permeate flow control Use open UF membranes to clarify small proteins from either whole cell broths or lysate streams Shear sensitivity If the feed stream is particularly shear sensitive and the recirculation flow rate is reduced it may be necessary to lower the permeate flow rate when using permeate flow control to optimize throughput Volume When scaling a process cartridge housing diameter is increased in order to maintain constant volume to area ratio When using a fixed 50 cm filter estimate the flux rate so that the starting volume is suitable for the target process time Temperature As temperature decreases the filtration time often increases due to viscosity effects and larger cartridges might be appropriate For example cold room processing at 4 C can take twice as long as room temperature processing Other variables Selection considerations T
10. to the surface of the membrane Driven by pressure some of the feed solution passes through the membrane filter Most of the solution is circulated back to the feed tank The movement of the feed solution across the membrane surface helps to remove the buildup of materials on the surface AKTAcrossflow Method Handbook 11 0012 36 Edition AB 127 128 Cross flow rate Also called retentate flow rate The flow rate of feed solution that flows across the surface of the filter and exits the filter as retentate Higher cross flow rates help sweep away the debris that forms on the surface of the filter Cross flow rate is most often measured at the retentate outlet Cutoff See Molecular weight cutoff MWCO and Nominal molecular weight cutoff NMWC Dead ended filtration See normal flow filtration AP AP or pressure differential between the feed and retentate lines The AP equals the feed pressure minus the retentate pressure Depth filter A thick filter that captures contaminants within its pore structure using entrapment and adsorption A membrane filter primarily captures contaminants on its surface Depyrogenate The removal or decomposition of pyrogens lipopolysaccharides endotoxins from a process solution Diafiltration A unit operation that incorporates ultrafiltration membranes to remove salts or other microsolutes from a solution Small molecules are separated from a
11. A 1 Hollow Fiber Cartridges Configuration Catalogue Pore size Fiber Number of Nominal number diameter fibers surface mm area cm Start AXH UFP 3 C H24U 3 kD 0 5 4 40 60 ci pathlength UFP 10 C H24U 10kD 0 5 4 40 ultrafiltration UFP 30 C H24U 30 kD 0 5 4 40 UFP 100 C H24U 100 kD 0 5 4 40 UFP 300 C H24U 300 kD 0 5 4 40 UFP 500 C H24U 500 kD 0 5 4 40 Start AXM UFP 3 C 2U 3 kD 0 5 12 50 30 cm path length UFP 10 C 2U 10 kD 05 12 50 ultrafiltration UFP 30 C 2U 30 kD 0 5 12 50 UFP 100 C 2U 100 kD 0 5 12 50 UFP 300 C 2U 300 kD 0 5 12 50 UFP 500 C 2U 500 kD 0 5 12 50 UFP 500 E 2U 500 kD 1 0 6 50 UFP 750 E 2U 750 kD 1 0 6 50 Start AXM CFP 1 E 2U 0 1 um 1 0 6 50 30 cm path length CFP 2 E 2U 0 2 um 1 0 6 50 microfiltration CFP 4 E 2U 0 45 um 1 0 6 50 CFP 6 D 2U 0 65 um 0 75 8 50 Cartridge kit CFP CELL KIT 2U Contains 1 each of UFP 750 E 2U CFP 1 E 2U CFP 2 E 2U CFP 4 E 2U and CFP 6 D 2U a AKTAcrossflow Method Handbook 11 0012 36 Edition AB 123 A 2 Membrane Cassettes for AKTAcrossflow system Code number Model number Pore size kD Membrane area cm 11 0006 02 UFESTOOOSOSOST 5 50 11 0006 04 UFESTOO1O050SE 10 Select 50 11 0006 03 UFESTOO10050ST 10 Select 50 11 0006 05 UFESTOO30050ST 30 50 11 0006 06 UFESTOOSOOSOST 50 50 11 0006 08 UFESTO100050ST 100 50 11 00
12. Handbook 11 0012 36 Edition AB 79 Cell Processing es Table 4 4 illustrates typical starting conditions for the harvesting of different cell types Bacterial cell harvest Mammalian cell Virus particle clarification concentration 10X concentration 10X concentration 5X concentration followed by 3X followed by 3X followed by 3X diafiltration diafiltration diafiltration Average flux 25 Imh with Permeate flow control at Low TMP and 6 000 sec high cell density starting 30 Imh no retentate shear 20 50 Imh material and back pressure average productivity unrestricted permeate calculated below at 30 Imh This process description This process description This process description is for optimal recovery is for removal of cells is for purification of virus and washing of cells and optimal recovery of particles with gentle only expressed target protein process conditions only only Table 4 4 Recommended starting point for developing process conditions for cell harvesting 4 5 2 Recommendations for Start AXM and Start AXH cartridges The general guidelines for using Start AXM and Start AXH cartridges in cell harvesting are as follows typical operating flow rate e 8 000 to 16 000 sec shear rate for bacterial feed streams e 2 000 to 4 000 sec for shear sensitive and high viscosity feed streams including mammalian cells 4 5 3 Process sequence Flux and protein passage is dependent on the concent
13. Hold_Until Trans ol Greater_Than 5 ml INFINITE base Hold_Until DeltaP Stable_Baseline 1 00 Minutes 5 00 base Hold_Until Trans_Flow Less_Than 0 1 ml min 5 00 base Fig 2 11 A Block pane in the Method Editor AKTAcrossflow Method Handbook 11 0012 36 Edition AB 45 AKTAcrossflow system components and software 2 5 6 Run Set up Run Set up in the Method Editor is a dialog box with a number of tabs that define the method properties Fig Fig 2 12 Start Protocol Questions Result Name Variables Scouting Notes Gradient Columns Reference Curves Evaluation Procedures Method Information Variable Value Range Ret_Vol ml 13 00 0 00 5000 00 a E EE a 2 T a EE Sansen rs Sewe oo oa Sample voume Sample VoLHigh ind i oo ooon e SSSSCSCSCSC SN__ 00 BOONLOD Reon Ste i Prowse i o ooo Hadum feet SSCSCS S dOH S C d OOH mese S eee O po sa Sterio f Core Factor StepiCont oo oo a E E BOOOLOD efion odes Ret Flow ProdRec mi min J o ooe Parow S w oo ooa C Show details C Show unused variables Display tooltip for extended variable cells Edit Variable Fig 2 12 Run Set up dialog box in the Method Editor ee 46 AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software 2 6 Work flow The work flow for AKTAcrossflow can be divided into three distinct stages e Create a method e Run the method e Evaluate the
14. Pt Upstream the reservoir and is mainly used to measure the pressure in the reservoir for safety reasons Table 2 1 Pressure sensor location To protect the system pressure limits can be set in UNICORN for the sensors Pf Pr and Pp The pressure sensors have a pressure range of 0 10 bar 1 MPa 145 psi The pressure sensor housing is made of PEEK Other wetted parts are made of titanium and stainless steel 2 3 5 Reservoir level sensor The reservoir level sensor is located in the reservoir bottom end plate It is a highly sensitive pressure sensor that continuously reports the hydrostatic pressure in the reservoir and thus the weight of the retentate to the control software These data are then transformed to information on the retentate liquid volume The level sensor has also the function of a low volume alarm for the reservoir The level sensor is used to calibrate the volume of AKTAcrossflow system during start up and in addition it ensures efficient product removal at the end of the filtration process by protecting the filter against the introduction of air The level sensor has a pressure range of 0 100 mbar 10 kPa 1 45 psi A temperature sensor is integrated with the reservoir level sensor and allows for continuous temperature measurement of the liquid feed to the CFF cassette cartridge 38 AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software 2 3 6 Air sensor
15. Pump heads Each pump head has an inlet check valve and an outlet check valve for the liquid flow In addition each pump head has an outlet check valve for the rinsing flow system The individual pump heads are actuated in opposite phase to each other by microprocessor controlled individual stepper motors The synchronization of the pump heads generates a constant flow with low pulsation For the feed pump this synchronization is optimized to yield a low pulsation flow at the inlet and outlet However the pump heads of the permeate pump are synchronized such that the flow at the pump inlet has low pulsation and for the transfer pump the pump outlet has low pulsation Pressure and flow at the permeate side of the filter cartridge can thus be controlled with a high degree of accuracy The pump heads are made from titanium alloy 30 AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software 2 2 3 Piston rinsing system Leakage between the pump chamber and the drive mechanism is prevented by a seal The seal is continuously lubricated by the presence of buffer In order to prevent any deposition of salts from aqueous buffers and other organic compounds on the pistons and to prolong the life of the seals the pump has a piston rinsing system The low pressure chamber situated behind the piston can be flushed continuously with 10 mM sodium hydroxide in 20 ethanol A check valve in the system ensures that t
16. cells In a crossflow filtration system the protein solution or cell suspension is recirculated over the filter membrane and thus passes through the pump many times It is therefore essential that the pump causes as little protein aggregation or cell disruption as possible In order to measure the effects of shear induced by different pump designs a series of studies was made comparing AKTAcrossflow P984 pumps and traditional pumps and the influence of a piston rinsing system on protein aggregation and cell disruption C 1 Shear studies on protein solutions A P984 pump fitted with AKTA sanitary tubing i d 3mm max flow rate 800ml min was compared to a peristaltic pump fitted with Marprene II 25 hose i d 4 8mm max flow rate 690 ml min In these studies a rotary piston pump with no piston sealing or rinsing system was also tested C 1 1 Piston rinsing system Fig C 1 shows the results of a pump rinsing system on protein aggregation as measured by UV absorbance at 620nm using AKTAcrossflow pump P984 The rinsing system reduces the friction between the piston and the cylinder wall and reduces protein aggregation due to shear The choice of rinsing fluid is important Fig C 2 Sodium hydroxide dissolves the aggregates and ethanol reduces the viscosity making the rinsing more effective The recommended rinsing solution for AKTAcrossflow system is 10 mM sodium hydroxide in 20 ethanol xflow sys P1 02 without rinsing syste
17. chemicals that they are exposed to at a given concentration temperature and total exposure time Clean in place CIP The process of cleaning a filtration device without removing it from its filtration system CIP processes can remove proteins lipids cell debris micro organisms and other contaminants Composite A membrane that is made up of two or more layers membrane that are usually chemically or structurally different Concentrate Also called retentate The part of the process solution that does not pass through a cross flow membrane filter Concentration Cross flow filtration process in which the components that do not pass through the membrane remain in the feed loop and therefore increase in concentration as permeate leaves the system Concentration The concentration factor equals the ratio of the initial factor feed volume to retentate volume after separation For example if the initial feed volume is 100 and the final retentate volume is 20 the concentration factor is 5x Concentration The buildup of molecules of dissolved substances polarization solutes on the surface of the membrane filter during filtration The concentration polarization layer increases resistance to permeate flow and reduces the permeate flux thus decreasing filtration efficiency Cross flow filtration CFF Also called tangential flow filtration In cross flow filtration the feed solution flows parallel
18. contributes to higher recovery but the amount recovered after 3 washes is small The amount of additional protein recovered should be balanced against the time needed for the extra wash steps Membrane Permeate Wash 1 Wash 2 Wash 3 Wash 4 pore size ratings 0 1 um 53 14 8 5 3 0 2 um 60 16 9 5 3 Table 6 5 Results from GFP His wash volume recovery study KTAcrossflow Method Handbook 11 0012 36 Edition AB Applications Ce A full run was completed with the optimal process parameters and the result is shown in Fig 6 4 The run was repeated three times in order to confirm the robustness of the selected process parameters T 5 a J 3 a a Fig 6 4 Results of a full process run to confirm recovery at maximum flux During concentration the retentate volume is decreasing and then it is kept constant during the diafiltration step while the UV is decreasing as the target protein is transferred to the permeate KTAcrossflow Method Handbook 11 0012 36 Edition AB 111 E Applications 112 6 2 6 Conclusion The clarification of the E coli lysate used in this study requires membranes with relatively large pore sizes To achieve a fast effective and robust process for the clarification of this cell lysate containing GFP His a membrane pore size of 0 1 um was used at a shear rate of 16000 sec and a flux rate of 45 LMH In some cases when a higher recovery is desired f
19. density is summarized in Table C 1 The results show that cells are treated very gently in KTAcrossflow system resulting in high cell density and viability and low levels of LDH Cell count in Nigrosin BB1744 Sample Viability Cell density 10 ml Starting CHO cells 98 5 13 Control 99 7 0 99 CHO cells after 400 cycles in 93 4 0 73 P984 CHO cells after 400 cycles in 98 8 0 83 AKTAcrossflow system CHO cells after 200 minutes in a 97 2 0 93 tank fitted with magnetic stirrer Table C 1 140 ml of CHO cell suspension was recirculated for 400 cycles at 280 ml min In addition a CHO cell suspension sample was kept in a recirculation tank with a magnetic stirrer to monitor the effect of stirring on cell integrity A KTAcrossflow Method Handbook 11 0012 36 Edition AB 139 140 C 3 Conclusion A comparison of shear forces induced by AKTAcrossflow P984 and other pumps clearly shows that AKTAcrossflow P984 is a much gentler pump It is the pump of choice for CFF applications with protein solutions and cell suspensions causing less protein aggregation and cell disruption AKTAcrossflow Method Handbook 11 0012 36 Edition AB www gehealthcare com GE Healthcare Bio Sciences AB Bj rkgatan 30 SE 751 84 Uppsala Sweden Drop Design AKTA AKTAcrossflow AKTAdesign UNICORN Luer Lock Hi Trap and Kvick Start are trademarks of GE Healthcare Ltd a General Electric company GE imagination at work and GE monogram are trademarks of General E
20. e Use low pressure and low pump speed to prevent heat generation in the flow path e Use low volume to filter surface area ratios to shorten process time e Place the system in a cold room Enzymatic action Enzymes released during cell culture tend to concentrate in the concentration and gel layer During concentration lower molecular weight proteins pass through the membrane and the concentration of the target protein increases The enzymes may digest the proteins of interest which remain in high concentration in the feed retentate circulation loop Catalysts for enzymatic activity include heat metal ions and pH Enzymatic activity can be minimized by e Minimizing process time e Minimizing process volume to surface area ratio e Reducing temperature e Adjusting temperature and ionic strength 3 2 4 Concentration gradient layer During filtration the solvent nearest the surface of the membrane flows through the membrane into the permeate As the solvent flows away from the surface of the membrane the solutes or particles near the surface become more concentrated Fig 3 1 This region of increased concentration is called the concentration gradient layer The concentration gradient layer reduces the flux compared to clean water flux The concentration gradient layer cannot be eliminated under production conditions but it can be controlled to some degree Decreasing TMP can lower the concentration gradient layer and its effects
21. float to prevent vortex formation and foaming so that operation at lowest circulation volume with high low rate is facilitated There are two sizes of reservoir e 350 ml 375 ml without float e 1100 ml 1200 ml without float 2 2 5 Liquid connections Each reservoir has connections for the liquid flow positioned at the reservoir bottom end plate There is one outlet for delivering liquid to the feed pump via a manifold and the outlet is placed off centre at the bottom of the reservoir to prevent vortex formation The retentate return is positioned such that the liquid is injected tangentially to the bottom surface 2 2 6 Magnetic stir bar The reservoir is mounted on a reservoir holder which contains a motor unit for a magnetic stir bar The stir bar can be used with both reservoirs to improve mixing characteristics Recommended stir bar dimensions are e 375 ml reservoir length of stir bar 30 mm diameter 6 mm max diameter at pivot ring 7 mm e 1200 ml reservoir length of stir bar 35 mm diameter 6 mm max diameter at pivot ring 8mm The appropriate mixing rate is a function of application and retentate volume and can be adjusted by the control software As default the UNICORN control software adjusts the mixing rate automatically depending on the actual retentate volume At low retentate volume the stir bar and the float will be in contact such that the stir bar will rotate the float Under these conditions a low mix
22. mode the TMP value is a function of the permeate flux Flux control mode Control Element Flux Control Feed Pump Permeate R PCV Pump Flux control with constant Or gt 0 Flux gt 0 Offset Feed flowrate Unrestricted for Pp gt Offset Flux control with constant Qr gt O Flux gt 0 Offset Retentate flowrate Unrestricted for Pp gt Offset Flux control with constant shear rate gt 0 Flux gt 0 Offset Shear rate Unrestricted for Pp gt Offset Flux control with constant Pe Pp gt O Flux gt 0 Offset AP Unrestricted for Pp gt Offset Table 2 3 Flux control mode During flux control it is common that the feed pressure is so low that the permeate pressure drops below zero If the permeate pressure is below 0 2 bar the software will adjust the R PCV to increase the retentate pressure When the permeate pressure is above 0 2 bar the permeate pump can start A constant ramping during 60 seconds from flux zero to the set flux value is then performed oe AKTAcrossflow Method Handbook 11 0012 36 Edition AB 43 AKTAcrossflow system components and software 2 5 Programming a UNICORN method 2 5 1 Blocks The text pane in the Method Editor of UNICORN displays the method as a list of text instructions Fig 2 10 The instructions are usually organized in blocks which define a specific function e g load a sample or concentrate a sample Blocks are indicated by blue square symbols A block may contain other blocks or individual instru
23. of Dispose if end of filter or clean service life and test again Clean and test again if failure Fig 1 7 Life cycle of membrane filters AKTAcrossflow Method Handbook 11 0012 36 Edition AB 23 Introduction 24 1 11 Membrane filter specifications The following sections describe technical aspects of membrane filters The appendix provides detailed specifications 1 11 1 Materials of construction Three desirable characteristics of membrane filters are mechanical strength chemical and physical compatibility and low extractables and toxicity ratings The membrane materials polyethersulfone and polysulfone used in AKTAcrossflow membrane filters offer broad pH and thermal stability and provide good chemical compatibility with many bioprocess fluids and cleaning solutions 1 11 2 Kvick Start cassettes The construction materials for Kvick Start cassettes are as follows e Fluid path e Inner plates Polyester copolymer e Membrane screen Polypropylene e Membrane Polyethersulfone e Port sealer Solvent free urethane meth acrylate blend e Luer Lock adapters Polypropylene e Luer Lock adapter gasket EPDM Ethylene propylene diene monomer e Housing Epoxy e Wetting fluid 0 1 0 2N sodium hydroxide and 20 22 w v glycerine 1 11 3 Start AXM and Start AXH cartridges The construction materials for Start AXM and Start AXH cartridges are as follows e Housing Polysulfone e Membrane Polysu
24. of clarification Suspended Path length Lumen ID Rating Rating solids and cm mm um NMWC viscosity Monoclonal antibody Low to 30 or 60 1 0 0 2 from hybridoma cell moderate 0 45 culture 0 65 Clarification of Low 30 0 75 0 65 adenovirus from 293 cell culture Clarification of protein High 30 or 60 1 0 0 1 lt 60 kD from E coli 500 000 whole cell broth 750 000 Clarification of 20 kD Moderate 30 or 60 1 0 750 000 protein from E coli lysate Clarification of 40 kD High 30 1 0 0 1 protein from Pichia pastoris Table 4 5 Recommended cartridges for cell culture and lysate clarification es AKTAcrossflow Method Handbook 11 0012 36 Edition AB 85 Cell Processing E Table 4 6 illustrates the relationship between solution variables and cartridge selection Solution variables Selection considerations Cell concentration Determine wet cell percent to anticipate the degree of concentration that may be used The benefit of a highly concentrated cell mass may should be balanced against the possibility of high inlet pressure requirements or the necessity of using less efficient short 30 cm path length cartridges Solids loading For whole cells it is not uncommon to reach 70 wet cell weight while maintaining steady state conditions However lysates tend to need a lower solids level to promote passage of the target material Start with solids in the 5 to 10 range and monitor transmission
25. off MWCO Nominal filter rating A rating that indicates the percentage of particles of a specific size or molecules of a specific molecular weight that will be removed by a filter No industry standard exists hence the ratings from manufacturer to manufacturer are not always comparable Nominal molecular The size designation in Daltons D for ultrafiltration weight cut off membranes No industry standard exists hence the NMWC NMWC ratings of different manufacturers are not always comparable Normal flow Also called dead ended filtration In normal flow filtration filtration liquid flows perpendicular to the filter media and all of the feed passes through ormalized water permeability NWP The water flux at 20 C divided by pressure Common units LMH psi LMH bar Particle size The distribution of particle sizes number or weight distribution fraction in a fluid Permeate Also called filtrate Any components of the feed solution that passes through the membrane pH Negative logarithm of the hydronium ion H30 concentration in an aqueous solution Indicates the acidity or alkalinity of a substance AKTAcrossflow Method Handbook 11 0012 36 Edition AB Programmable logic controller PLC A device for industrial control Types of operations common to PLCs are polling or checking sensors and activating deactivating valves and switches compared against programme
26. on flux Increasing the cross flow rate produces a sweeping effect that helps to redistribute concentrated solutes back into the bulk feed stream and maintain flux AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cross flow filtration process considerations A Feed flow 2 ee ee Bulk stream AF a s Oe Fee ay 7 lt ee s 9 w a ws A of Oae o epo a y e 8 E Eeen Solutes Concentration gradient layer Permeate flow B Feed flow A e e e e e e Bulk stream 4 J2 TAE lt Be s e ee eee ee e e os o 2 g 9 e e te gt ote 32 soot A So d Solutes Concentration 5 gradient layer Gel layer J Membrane LI Permeate flow Solvent flow is greatly reduced once a gel layer forms on the membrane surface Fig 3 1 A Concentration gradient layer forms on the membrane surface during processing and B gel layer and concentration gradient layer formed on a membrane surface AKTAcrossflow Method Handbook 11 0012 36 Edition AB 61 Cross flow filtration process considerations 3 2 5 Gellayer A gel layer is a concentration gradient layer that has reached its highest value In a gel layer hydro colloids formed from concentrated proteins become packed so tightly against the membrane surface that they form a viscous or gelatinous layer The gel layer has a considerable effect on the filtration process influencing
27. process streams should be sampled before and after filtration By sampling the permeate stream information can be obtained on the types of non target proteins lipids or unwanted components that are being passed though the membrane AKTAcrossflow includes a UV sensor in the permeate line to measure the passage of protein through the membrane AKTAcrossflow Method Handbook 11 0012 36 Edition AB 55 Cross flow filtration process considerations 56 3 2 Specific actions that increase yield The key factors that influence yield are the following e Membrane selection e Non recoverable hold up volume e Denaturation shear temperature and enzymatic action e Concentration and gel layer formation 3 2 1 Membrane selection For ultrafiltration with cassettes membranes influence ultrafiltration yields in two ways Selectivity and protein binding If a membrane is selected with pores too large to retain the protein being concentrated some of the target protein passes through the membrane and is not recovered decreasing yield Protein binding usually becomes an issue when attempting to separate extremely small amounts of protein In this case binding of the protein to the membrane can show up as unexpected yield losses KvickStart polyethersulfone membrane exhibits low protein binding and minimizes this effect For cell harvesting with microfiltration cartridges membrane selection plays a less important role in yield results Cells
28. quite high closely monitor the inlet pressure to avoid over concentration Operate at low shear conditions using 0 65 micron membranes and permeate flow control set at 20 to 30 LMH AKTAcrossflow Method Handbook 11 0012 36 Edition AB Table 4 7 Recommended starting point for developing process conditions for clarification 87 Cell Processing 88 4 12 Three examples of clarification strategies The following examples illustrate the development of process conditions with mammalian cells bacterial cells and yeast 4 12 1 Mammalian cells Recently developed therapeutic proteins are derived from cell culture sources These are most often grown with mammalian cell lines in highly purified media Although there are a variety of cells that are suitable and cell densities range from 10 to 10 cells per ml the clarification process is similar for each type Membrane and cartridge selection If the protein of interest is an antibody 0 2 0 45 or 0 65 micron membranes can be chosen Permeate turbidity will be slightly lower if 0 2 micron filters are used The 0 65 micron rating usually provides the best throughput These cartridges can be tested with feed solution to determine which rating provides the best overall performance The above microfiltration membranes are available in only one fiber diameter and the only remaining variable is the cartridge path length 30 or 60 cm Initial testing should begin with the
29. results 2 6 1 Creating a new method There are two ways to create a new method e Using the Method Wizard where customized methods for most methods are pre programmed and the user sets appropriate values for the method variables e Using the Text Instructions editor in the Method Editor module where the user can choose more advanced editing facilities 2 6 2 Method Wizard Method wizards support all typical ultrafiltration and microfiltration product processing operations With the method wizard filtration methods can be easily and rapidly programmed It uses pre optimized and verified methods and no programming skills are needed The method wizard covers system functional tests and all the steps in a typical filtration process It is possible to rinse new filters CIP used filters and test water flux to check filter quality and status before and after each run Data for a given filter can be gathered over multiple cycles in order to check its membrane flux recovery A system sanitization method is also provided Wizard methods for flat sheets Wizard methods for hollow fibers Ultrafiltration Cell processing Proteins Cell harvest Cell clarification Lysate clarification 1 Concentration reduce volume 1 Concentration reduce volume 1 Concentration reduce volume 1 Concentration reduce volume 2 Diafiltration exchange buffer 2 Washing promote contaminant passage 2 Wa
30. size for application not being used Reevalaute pore size and membrane area selection Insufficient cross flow rate excessive gel layer formation Increase cross flow rate TMP too high excessive gel layer formation Lower TMP reduce permeate flow rate using permeate flow control Chemical incompatibility between cleaning agents and membrane membrane damage Check chemical compatibility between membrane and process fluid Water flux after cleaning less than 60 80 of initial water flux Insufficient cleaning ncrease cleaning temperature ncrease concentration of cleaning solution Increase cleaning circulation time or rate Use cleaning solution better able to solubilize contaminants Chemical incompatibility between cleaning agents and membrane membrane damaged Replace cartridge or cassette Water flux less than 60 of water flux when the cartridge was new data shows a gradual decrease over many runs Normal decline in operational efficiency Replace cartridge or cassette Water flux less than 60 of water flux when the cartridge was new data shows decrease was sudden compared to historical data Chemical incompatibility between cleaning agents and or process fluids and membrane membrane damaged Replace cartridge or cassette Insufficient cleaning Increase cleaning temperature Increase concentration of cleaning solution Increase cleaning circulation
31. solution CIP optional Recirculate cleaning solution Water flush Flush chemical cleaning solution from system Water flux test Determines the performance of filter after use and cleaning Storage solution Flushes the filter with storage solution to prevent bacterial growth in storage Table 4 1 Typical steps in a cell harvesting method AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing 4 2 3 Membrane and cartridge selection In cell harvesting microfiltration membranes will easily retain all cells The key to membrane selection is not based on retention but on process optimization For example smaller pore size membranes often provide the highest permeate flux once the system in a steady state Fig 4 3 The 500 000 NMWC ultrafiltration membrane is often the cartridge of choice for harvesting E coli even though it has a relative small pore size compared to the size of the cells Permeate Flux Fig 4 3 Flux of three membranes with all parameters held constant except pore size Membrane A has a larger pore size than membrane B which has a larger pore size than membrane C The 30 cm cartridge length allows low pressure drop for difficult separations using low TMP The 60 cm hoop cartridges have a similar membrane area but will require less circulation flow per unit area Therefore the 60 cm length is preferred for applications in which higher TMP does not adversely affect the separation Table
32. well as for sterilization and contaminant and particle removal in numerous biopharmaceutical processes AKTAcrossflow Method Handbook 11 0012 36 Edition AB Introduction 1 1 5 Ultrafiltration filters Membranes with 1 000 to 1 000 000 Daltons nominal molecular weight cutoff NMWC are called ultrafilters These membrane are used for concentrating and fractionating protein streams virus concentration desalting and buffer exchange The objective of most ultrafiltration processes is to retain and fractionate soluble macromolecules such as proteins while allowing liquid and unwanted smaller molecules to pass such as salts amino acids and mono or di saccharides Fig 1 2 Microfiltration Ultrafiltration EAE E S A S A D 4 441i 0 1nm inm 10nm 100nm 1pm 10um 100um 1KD 100KD 750KD 50KD 500KD 1 nm 7 nm Sucrose Haemoglobin Fig 1 2 Relative size of CFF feed components and operational scales for filtration KTAcrossflow Method Handbook 11 0012 36 Edition AB 11 Introduction 1 2 CFF terminology Feed The starting solution or suspension that is pumped to the filter for separation Permeate Any components of the feed that pass through the membrane Retentate Any component that does not pass through the membrane but instead circulates through the retentate line back to the feed tank Cross flow rate The rate of flow across the membrane surface Higher cross fl
33. 0 2000 i500 4 i000 4 500 _ 3 29 o La ees ee ob 1b 2b 3b 4b 5b 6b 7b eb ab 10 0 mi PnAMRetentate001 10 UV2 280nm PnAMRetentate001 10 Conc PnAMRetentateO01 10 pH PnAMRetentate001 10_UV2_280nm 02 BASEM1 ImAU 1380 160 140 120 100 80 60 40 20 RO pa a ob 1b 2b 2b 4b 5b 6b 75 3D ob 10 0 mi Fig 6 11 Hi Trap column chromatography results of IgG4 The column was equilibrated with 20 mM sodium phosphate buffer pH 7 0 100 ul of the samples were applied to the column at a fluid velocity of 1 ml min The IgG was eluted with 0 1 M sodium citrate buffer pH 3 5 A start material B concentrated material after diafiltration diluted forty times O KTAcrossflow Method Handbook 11 0012 36 Edition AB 121 6 Applications 6 4 6 Conclusion In this application AKTAcrossflow system was used with Kvick Start cassettes to concentrate IgG4 from a clarified cell culture supernatant The optimal process parameters were defined by running TMP excursions as well as a concentration run to find the most suitable concentration factor for performing diafiltration The optimized process can be summarized as follows e gG4 starting concentration 0 3 mg ml e Process Crossflow rate 125ml min e Process TMP 1 3 bar e Optimal time for diafiltration is at forty times concentration e IgG4 final concentration 12 mg ml 122 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Appendix A Membrane Filters for AKTAcrossflow system
34. 06 61 UFESTCPAKO4SST 5 10 Select 10 30 50 50 100 Sas 124 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Appendix B Glossary of terms B 1 Glossary of terms Adsorption The binding of molecules to a surface as a result of a chemical or physical interaction between the membrane surface and the molecule Air diffusion rate The rate at which air diffuses through the wetted pores of amembrane at a given differential pressure Measuring the air diffusion rate is a method used to check the integrity of a membrane filter Autoclave autoclavability An autoclave is a device that uses saturated steam at a specified pressure over time to kill microorganisms and thus achieve sanitization or sterilization Because many materials change properties when exposed to moisture heat and pressure products destined for this process must be specially engineered for autoclavability Back flushing backwash Reversing the permeate flow to mechanically clean the membrane Binding The process by which some components in a feed solution adhere to the membrane Binding can be desirable in some instances but often as in the case of protein binding during sterile filtration can result in a loss of valuable product Biosafety tests A class of tests that determine whether a filter s materials of construction can induce systemic toxicity skin irritation sensitization reaction or other biological responses
35. 1 Product Flux _ConcDiaBSAQ01 Product ConcFactor Cone DiaBSA001 Product Logbook bar 7 0 2 5 6 0 N j 20 5 0 mt 40 x 1 6 3 0 20 Fill Sample 0 0 Fig 6 7 Complete BSA process with optimized parameters During the diafiltration step the conductivity of the permeate and consequently the retentate was lowered from 16 mS cm to 4 mS cm and the pH was decreased from 7 4 to 5 9 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Applications Ce 6 4 Concentration of cell culture supernatant containing gG4 6 4 1 Objective The objective of this application was a forty times concentration of the IgG4 solution from 0 3 mg ml to 12 mg ml The sample was 2000 ml of cell culture supernatant containing 0 3 mg ml IgG4 To keep the sample load within the recommended range 30 200 I m three Kvick Start cassettes 50 cm 30 kD were connected in parallel giving a total membrane filter area of 150 cm The sample solution was applied in fed batch mode starting with an initial volume of 50 ml solution in the reservoir and continuously feeding sample to reservoir at the same rate as the permeate was removed 6 4 2 Process optimization The strategy for the optimization of process parameters was similar to that of the earlier example of BSA The IgG4 concentration in the various samples was determined using a Hi Trap Protein A column AKTAcrossflow Method Han
36. 30 cm path length to help maintain low TMP readings When the appropriate membrane and operating conditions have been chosen additional testing using the 60 cm path length may be used to determine if this design is suitable for scale up Process conditions and monitoring The circulation flow rate should be set up to 4000 sec If tests show damage to the cells the circulation flow rate should be reduced The permeate flow rate should be set to 30 to 50 LMH Using these conditions the initial TMP will begin at approximately 70 mbar Since the cells are completely retained and the protein will initially pass through the membrane quantitatively the objective of testing should be to determine the filtration capacity As a general rule once the TMP has increased by a factor of 4 to 5 from the initial reading the membrane is exhausted In this example working with permeate flow control set at 30 LMH if the TMP begins at 70 mbar when the TMP reaches 250 to 350 mbar the cartridge capacity has been reached Using the change in TMP as an indicator it is possible to compare a variety of membrane ratings and process controls Once a set of standard conditions has been adopted the filtration efficiency can also be studied as a function of the cell culture process For example with low cell viability membrane throughputs working with a 0 45 um membrane might be as low as 50 L m With healthy cells operating under the same conditions the throughput m
37. 4 2 The inside diameters of the hollow fibers in Start AXM and Start AXH cartridges range from 0 5 to 1 0 mm Larger diameter fibers should be used for solutions with high suspended solids high cell densities and high viscosity AKTAcrossflow Method Handbook 11 0012 36 Edition AB 75 Cell Processing A Type of cell Suspended Path length Lumen ID Rating Rating solids and cm mm u NMWC viscosity E coli Moderate 30 or 60 1 0 0 1 500 000 750 000 Yeast High 30 1 0 0 1 750 000 0 2 Mammalian Low 30 or 60 0 75 or 1 0 0 2 0 45 0 65 Blood cells Moderate 30 or 60 0 75 or 1 0 0 2 0 45 0 65 Table 4 2 Recommended cartridges for cell harvesting ee 76 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing able 4 3 illustrates the relationship between process variables and cartridge selection Process variables Selection considerations Cell harvesting Use microfiltration cartridges for cell harvesting Select membrane pore size based on the specific application to achieve a stable flux rate Solution variables Selection considerations Cell concentration Determine wet cell percent to anticipate the degree of concentration that may be used A highly concentrated cell mass may seem efficient but may also result in high inlet pressures or lead to using the less efficient short 30 cm path length Solids loading For whole cells
38. 8 C 2 Shear studies ON Cell SUSPENSIONS ssssssssssessessssssscsssssssssesessesscessseeees 138 C 2 1 COMPGTISON OF PUMP TUDES ssssnsescsssicerstesssaceasecvasstencdesbectesrseoinan ctincadattons 139 CB CONCIUSi Oirean a asa 140 E KTAcrossflow Method Handbook 11 0012 36 Edition AB vii Contents viii KTAcrossflow Method Handbook 11 0012 36 Edition AB Introduction 1 Introduction 1 1 Cross flow filtration and membrane filters 1 1 1 What is Cross Flow Filtration Cross Flow Filtration CFF is a filtration process in which the feed solution tangentially passes along the surface of the filter A pressure difference across the filter is used to drive those components through the filter that are smaller than the pores Components larger than the filter s pores are retained and pass across the membrane surface flowing back to the feed reservoir Fig 1 1 The key feature of CFF is the cross flow The cross flow of fluid along the membrane surface sweeps away the build up of material deposits on the filter surface and prevents the filter from fouling quickly CFF is simple in concept but its proper execution requires detailed knowledge and good filtration technique Higher pressure on the feed retentate side of the membrane drives the fluid and small components through the membrane Retentate Permeate Cross flow sweeps material buildup from the membrane surface Filter housing Membrane Fig 1 1 The fund
39. Edition AB 101 Concentration and Diafiltration 102 One shot Discontinuous Continuous fn 60 0 50 0 S Ss 400 S S 30 0 la O W 5 200 amp S 100 E 0 0 0 1 2 3 4 5 6 7 8 9 10 Diafiltration factor x Fig 5 8 Effect of diafiltration methods on the buffer concentration KTAcrossflow Method Handbook 11 0012 36 Edition AB Applications ea 6 Applications 6 1 Purification of B glucosidase from a Pichia pastoris cell culture broth using microfiltration 6 1 1 Objective The objective of this application was the clarification of a yeast cell suspension and the recovery of B glucosidase in the permeate P pastoris expresses B glucosidase extracellularly so the protein is in the culture broth B glucosidase is a complex consisting of five subunits with an approximate molecular weight of 400 kD P pastoris cells can be grown to avery high cell density in this application the dry cell weight content was 17 equivalent to 45 wet cell weight This high solids content can create problems at higher crossflow rates When using hollow fibers the cross flow rate is converted into a shear rate which represents the crossflow per fiber and fiber diameter Since the fibers are relatively open they rely on a high crossflow rate to create enough turbulent flow to prevent the cells from accumulating at the membrane surface The crossflow rate or shear rate and the amount of rest
40. GE Healthcare AKTAcrossflow Method Handbook AKTA Handbooks from GE Healthcare f 1 ot cer MT he E abi FSS be o i ee a Sn a Pe wwe 7 a de ory IA R i Antibody Purification Handbook 18 1037 46 The Recombinant Protein Handbook Protein Amplification and Simple Purification 18 1142 75 Protein Purification Handbook 18 1132 29 lon Exchange Chromatography amp Chromatofocusing Principles and Methods 11 0004 21 Affinity Chromatography Principles and Methods 18 1022 29 Hydrophobic Interaction Chromatography Principles and Methods 18 1020 90 Gel Filtration Principles and Methods 18 1022 18 Expanded Bed Adsorption Principles and Methods 8 1124 26 Microcarrier cell culture Principles and Methods 8 1140 62 Percoll ethodology and Applications 8 1115 69 Ficoll Paque Plus For in vitro isolation of lymphocytes 8 1152 69 GST Gene Fusion System Handbook 8 1157 58 2 D Electrophoresis using immobilized pH gradients Principles and Methods 80 6429 60 Contents Introduction 1 1 Cross flow filtration and membrane filters wo eceeeccssssesssseeesseseen 9 1 1 1 Whatt iS Cross FIOW Filtration ceccccsssssssssssessssssssessssssseesssssssseesssssssesessssseeees 9 L2 How does CFF differ from conventional filtration 1 1 3 Microfiltration and ultrafiltration 1 1 4 Microfiltration filters oes 10 1 1 5 Ultrafiltration NOMS costes icessscte ents nan
41. Lab cassette Kvick Pilot cassette Kvick Process cassette Start AXM Start AXH Pilot scale cartridges Process scale cartridges Process scale cartridges Table 3 3 Scale up pathway for cartridges and cassettes ooo AKTAcrossflow Method Handbook 11 0012 36 Edition AB 67 Cross flow filtration process considerations es 3 8 Membrane fouling and cleaning procedures Start AXM and Start AXH cartridges are designed for single use only The recommended cleaning procedure for Kvick Start cassettes is summarized in Table 3 4 Cleaning agent Cleaning conditions Membrane type 1 5 Alconox Contact time 60 minutes Polyethersulfone detergent Temperature 40 C 104 F 0 1 to 0 5N NaOH Contact time 60 minutes Polyethersulfone Temperature 40 C 104 F 200 to 300 ppm sodium Contact time 60 minutes Polyethersulfone hypochlorite in 0 1 to Temperature 20 C 68 F 0 5N NaOH Table 3 4 Recommended cleaning procedures for membrane cassettes Sx 68 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cross flow filtration process considerations 3 9 Troubleshooting The following chart can be used to troubleshoot fouling problems in CFF Symptom of fouling Possible cause of fouling Corrective action During operation slow gradual decrease of flux to about 90 of starting flux Normal operation None During operation moderate decrease of flux to about 75 or less of starting flux Best pore
42. R PCV is used to accurately control the retentate pressure over the pressure range 0 1 5 2 bar In this way the transmembrane pressure TMP can be adjusted In addition the R PCV can operate as an open close valve in product recovery and system cleaning procedures Permeate control valve P PCV The main function of the permeate control valve P PCV is to control the pressure downstream of the permeate pump in order to ensure accuracy in the permeate flow rate To ensure proper operation of the check valves the pressure downstream of the pump must be greater than the pressure upstream of the pump The P PCV valve is controlled by the software such that it will always maintain a higher pressure downstream of the pump 2 way transfer purge valve The transfer purge valve directs the liquid flow either from the transfer line or the permeate recycle line to the reservoir default or to waste 2 way transfer purge valve Pressure control valves R PCV and P PCV Fig 2 7 Two way transfer purge valve and pressure control valves R PCV and P PCV The lever actuated valve units have an EPDM encapsulated lever which is actuated by a solenoid to open or close a flow path The solenoid adjusts the force of the lever against the flow through the inlet port This novel and robust design results in the pressure upstream of the valve being maintained irrespective of changes in flow rate in contrast to conventional control valves 36 AKTAcrossflow Me
43. a very high final protein concentration The sample was 30 mg ml BSA 66 kD molecular weight in PBS buffer and the sample volume was 250 ml CFF experiments were carried out on AKTAcrossflow system using a 50 cm Kvick Start cassette 10 kD NMWC The diafiltration buffer used was 50 mM acetate buffer pH 5 5 6 3 2 Process Optimization The strategy for process optimization was as follows e Membrane selection e Optimization of the critical process parameters e Verification of the optimized parameters in a complete run 6 3 3 Membrane selection Membranes were screened to identify the highest membrane NMWC which retained the target molecule The BSA recovery was calculated by measuring the absorbance at 280 nm for both the retentate and permeate As a general rule select a membrane with a NMWC which is at least five times lower than the molecular weight of the target protein 6 3 4 Optimization of critical process parameters TMP excursions A series of TMP excursions were carried out Five TMP setpoints were tested at three crossflow rates in total recycle mode at the initial BSA concentration 30 mg ml and at the target BSA concentration 150 mg ml The TMP excursion plan is summarized in Table 6 6 The crossflow range tested was based on typical operating crossflow rates from 300 to 500 LMH which corresponds to 25 to 42 ml min for a 50 cm membrane a AKTAcrossflow Method Handbook 11 0012 36 Edition AB 113 6 Applications BSA concentra
44. amental concept and terminology of cross flow filtration AKTAcrossflow Method Handbook 11 0012 36 Edition AB 9 Introduction 10 1 1 2 How does CFF differ from conventional filtration CFF differs from conventional filtration in two ways First CFF filters use membranes exclusively while conventional filtration may use membranes paper or nonwovens to separate components in a feed stream Secondly the feed in a CFF filter circulates across the membrane surface multiple times In conventional filtration the feed is directed at the surface of the filtration media and does not circulate Hence as the filter cake builds up the filtration characteristics change the fluid flow decreases markedly and eventually the filtering ends Typically the filtration membrane is single use at the laboratory scale 1 1 3 Microfiltration and ultrafiltration Although cross flow filtration encompasses a wide range of membrane technologies for the purpose of this handbook CFF can be divided into two classes microfiltration and ultrafiltration Microfiltration filters have larger pores than their ultrafiltration counterparts 1 1 4 Microfiltration filters Membranes with 0 1 um to 10 um pore size ratings are classified as microfilters however in CFF the practical pore size ranges from 0 1 um to 1 um Membranes with 0 65 um 0 45 um 0 2 um and 0 1 um pore size ratings are used for separation of cultured cells from the growth medium broth as
45. ana 71 AZ Cell harvesting Process sescsssssesasessonsecssnvsesasssaseoossderdsessovdsvesus sae Sbessestorssoesys 71 4 2 1 Washing SEO P sessissc csssscodsssonnssnsdnsnvansestaasatepnccnzpntsaclacacastonssecsbedjcbsasiacususivinidbcrenda 71 4 2 2 Typical steps in a cell harvesting method eeeeecccssssssssssssssscsscsscsesseeceeeees 74 4 2 3 Membrane and cartridge selection i 4 3 Other product and processing factors wecceeccccsssssssseesssesecssssesssessseees BA Preparation before Usesenniinisieniinnnanireiiiimiraiii 4 4 1 Microfiltration COPEIAGE sesssssscssssssssssssssssssssssssssssssssesesseseseeeseeeeseesssssssnsssesse 4 4 2 Conditioning the system with buffer ccccccccccscssssssssssssssssssssssssssssseeses 45 Operating parameters ou 45 1 Permeate flow COMO cceeccssssssssssssssssssssssssssssssssssussssssssscsesssseseeeeeeeeees 4 5 2 Recommendations for Start AXM and Start AXH cartridges 4 5 5 PrOCeSS SEQUENCE eeeessssssesssssssseesseessssesseessneseeessnnecsesssnesecesssneeseessnneesesennetseessaneess 4 5 4 Process TEMPOTOTULES asssscssssecseasesssosssssscesasyasnecong2sceatsteesennsasedshonnseencnaiendeseesness 4 6 Cell harvesting Conditions ooecceeececcssssssssssssessessecssssssssssssuesssssssscsessessssunseesees AT Celkelarincati Ohessa 4 8 Lysate Clarification E 4 9 Membrane and cartridge selection wceecccccccsssssssssssseesssssscsssssssssuseeees 4 9 1 Membrane SOE CE ON se ccscscscscocosscasssneissssovscestetndvcadanaassaiasasasc
46. arameter EES AKTAcrossflow Method Handbook 11 0012 36 Edition AB 51 AKTAcrossflow system components and software 52 Flux vs TMP Ce a O CF 42ml min 100g L BSA CF 33ml min 100g L BSA CF 25mi min 100g L BSA X CF 42mlimin 20g L BSA CF 33mlmin 20g L BSA CF 25mi min 20g L BSA 150 F E 125 2 E me H 100 x C 3 B ma TS Be E _ _ a E T Bg 50 Le 06 o7 ws os 10 11 12 43 44 15 16 47 18 TMP bar Fig 2 17 Optimization curves generated in the evaluation module from the raw data of Fig 2 16 Normalized water flux is used to monitor membrane quality over several cleaning cycles This ensures that the cleaning process is still effective and also helps to determine the lifetime of the filter For agiven ultrafiltration process diafiltration time optimization allows the user to identify the factor of volume concentration where the least time is required to complete the diafiltration The analysis of experimental results in cell processing often includes plotting process parameters versus the membrane capacity Capacity plots allow the user to plot any process parameter including a system external result such as activity assay results versus the accumulating permeate volume normalized to the surface area capacity The any vs any evaluation operation is used to analyze resu
47. are ES 2 3 1 Valves The liquid flow in AKTAcrossflow system is controlled by valves of different functionality e Four membrane valve blocks of stepper motor actuated valves with open close functionality e Two pressure control valves R PCV and P PCV lever actuated e One 2 way switch valve transfer purge valve lever actuated All valves are sanitary designed with EPDM membranes for high chemical resistance Membrane valve block The valves are located in valve blocks to minimize hold up volumes A valve block consists of a connection block containing the ports and membranes and a mechanical housing containing the stepper motor cams and actuating pistons The valve blocks have different numbers of inlet and outlet ports depending on their location in the flow path see flow diagram There are four different types of membrane valve block e Inlet valves T VB In 1 4 e Inlet valves T VB In 5 8 e Outlet valves R VB Out 1 safety valve 2 3 e Outlet valves P VB Out recycle 1 2 3 safety valve Two of the outlet valves R VB Out 1 and P VB Out 3 have built in safety valve functionality with an opening pressure 7 bar 102 psi To From transfer pump transfer valve block 2 From buffer sample containers and air sensor Fig 2 6 Valve block ooo AKTAcrossflow Method Handbook 11 0012 36 Edition AB 35 AKTAcrossflow system components and software Retentate control valve R PCV The retentate control valve
48. are relatively large compared to the membrane pores So even selecting a microfilter with very large pores will still retain all of the cells and particle components For lysate clarification where the goal is to recover a protein while holding back cell debris membrane selection is critical in allowing the target protein to pass Membrane selectivity Membrane selectivity is defined as a membrane s ability to retain 100 percent of a single species Ultrafiltration filters have a broad pore size distribution and are therefore not highly selective To achieve the best possible retention with a typical ultrafiltration filter a NMWC that is 3 to 5 times less than the target molecule weight should be evaluated for performance In microfiltration membrane selectivity is not as critical For example when separating an antibody from a cell culture pore size distribution is not a key factor A membrane with a distribution of larger pores will provide good yield but the permeate may be slightly turbid and require a polishing filtration step If an excessively small pore size is chosen not all of the antibody will pass through the membrane decreasing the yield AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cross flow filtration process considerations The following guidelines represent a good starting point for microfiltration membrane selection e Yeast and bacteria cell harvest 0 1 um pore size microfiltration e Lysate clarificati
49. as well as TMP during the concentration phase Size of the target For separations with large target material high selectivity it may be best to avoid material any concentration but rather perform a constant volume wash from the start Remember that using a more open membrane may require the use of short 30 cm cartridges and permeate flow control Use open UF membranes to clarify small proteins from either whole cell broths or lysate streams Shear sensitivity If the feed stream is particularly shear sensitive and the recirculation flow rate is reduced it may be necessary to lower the permeate flow rate when using permeate flow control to optimize throughput Other variables Selection considerations Time constraints Increased membrane area and larger housing size shorten production time Heat sterilization Choose autoclavable or steam in place models for scale up Table 4 6 The influence of process variables in selecting a cross flow cartridge 4 10 Filter and system preparation The steps for filter and system cleaning are as outlined in the previous section on cell harvesting as 86 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing 4 11 Operating parameters 4 11 1 Permeate flow control In clarification processes as in cell harvesting the flux is often high even at low TMP values and steps should be taken to decrease the flux to prevent premature fouling of the membrane Please refer to the section on cel
50. ated proteins enzymes and cell debris The washing process is commonly a constant volume diafiltration process in which buffer is added to the cell suspension at the same rate as the permeate flow Unlike centrifugal techniques where cells are packed in a dense cake or pellet washing the cells in a buoyant state enables effective removal of contaminants In AKTAcrossflow UNICORN software user interface the diafiltration volume is used to set the washing parameters AKTAcrossflow Method Handbook 11 0012 36 Edition AB al Cell Processing 72 Cells and broth Fermentor Clarified product of interest Downstream processing steps Finished product Mechanical disruption of cells Cell Cells harvesting Cell debris and lysate Lysate clarification Cell debris and unwanted components Fig 4 1 Cell harvesting step in a typical biopharmaceutical manufacturing process AKTAcrossflow Method Handbook 11 0012 36 Edition AB 1 Start of Cell Harvesting Cartridge Feed contains cells filter and unwanted broth components Permeate collection vessel 3 Start of Washing step Cartridge Feed reservoir filter contains harvested cells and some unwanted broth components Cell Processing Ce 2 Cells Harvested Cartridge filter Feed reservoir contains harvested cells fewer unwanted broth components Permeate contains broth and unwanted components 4 End of Wash
51. benisiadenneeind guanaaacnsed 11 12 CPE TERMI OOO Yinsenin rin EAE 12 13 Applications OVERVIOW sisississcicssssivicsscantessissieeinsvsssscuterasssitennevsctsannunien 14 1 3 1 Cell processing and concentration diafiltration ccscssssssssssssssee 15 1 4 Technical parameters os 1 5 Approximate times for completing CFF runs wececccccccssssssssesessesecssen 17 1 6 Water quality require Me nts ceecccscccssssssssssssnusssssesscssssssssssssuessessesceses 17 1 7 Membrone filtration devices occeeccccccscsssssssssssssssssesssssseeseseceseeeceeceececeeneceeee 18 1 7 1 Fiber length and lumen diameter cartridges 1 2111121191 1121 22 18 17 2 Flow path length and channel height cassettes 1 7 3 Membrane surface area 1 7 4 Pore SIZE iannis 19 1 8 Membrane Structure oaccecccsccsssssssssssssssssssssssssssssssssssssssssnsssesssssseneneeeneneeenneneees 20 1 8 1 Ultrafiltration Membrane sssssssssssssssssscssssesecceeccsssssssssssssssssusnussssssssesseeeeeees 20 1 8 2 Microfiltration Membra Nessin 20 19 M mbran filter AESIGMA scccsisscisssssescecscsusvessisciesbsasosnssssbinessbscisebstoscsssstuacesion 21 1 9 1 Kvick Start cassettes Pe 1 9 2 Hollow fiber COPtIAGCS ssssssssssssssssssssssssssssssstussssssssssesessecseseeseeeseseessssssssssesss 21 110 CFE TIMER life Cycle nesie na 23 1 11 Membrane filter Specifications w cceccccccssssssssssssessessssssssssssssseueesssessssen 24 1 11 1 Materidls Of COMStFUCTION
52. both filter efficiency and selectivity To control the filtration process steps must be taken to minimize the formation of a gel layer Fig 3 1 The following operating conditions contribute to gel layer formation e Excessive TMP e Low cross flow rate High feed concentration e Incorrect ionic condition of the feed During the optimization of a CFF process with various TMPs the point just before the formation of a gel layer is identified At the optimum TMP and cross flow rate the highest flux rate is achieved without forming a gel layer that will diminish process control and flux rate In the case of protein concentration where the product of interest is retained in the retentate a gel layer can prevent the washing out of contaminants The result is a reduction in purity and product quality 3 2 6 Summary of concentration gradient and gel layer formation In summary three components resist the transfer of solvent through the membrane during concentration the membrane concentration gradient layer and gel layer With pure water only the membrane resists the transfer and there is no concentration gradient or gel layer When using process fluid the membrane and the concentration gradient layer resist the transfer of solvent through the membrane When a gel layer forms due to the incorrect operating conditions listed above all three components membrane concentration gradient layer and gel layer resist the transfer of solvent wit
53. cant increase in flux This enables a crossflow rate to be chosen which gives a high flux value for a reasonable process time AKTAcrossflow Method Handbook 11 0012 36 Edition AB 99 Concentration and Diafiltration TMP excursions TMP excursions BSA30g L BSA 150g L 250 CF Hamn CF mn CF Elma CF ae CF Wia CF TA 225 200 Ens Za 105 100 754 0 75 1 00 1 25 1 50 1 75 2 00 2 25 2 50 275 3 00 075 1 00 1 25 1 50 1 75 2 00 2 25 2 50 275 TMP bar TMP bar Fig 5 5 TMP excursion results at two concentrations of BSA 5 8 2 Diafiltration time optimization In a concentration process the optimized crossflow and TMP conditions established above can be used to identify the diafiltration point the point which provides the fastest buffer exchange and optimal buffer consumption A typical result file for diafiltration time optimization is shown in Fig 5 6 Evaluation DFTimeOptimization 10 x UNICORN Local Filnnika Result Wizard09T ID40001 res i File Edit view Integrate Operations Procedures Window Help ax Dg 6a Otawa 66 Sr Bii bo i ps po Pi hs j rd ho i a bs a AA bo 7 z p5 rd Cd Ca 2 0 Pi ai ee s pe R bs 1 ba ee a a ee oo ob 6b 40 0 46 0 20 0 26 0 300 350 a00 450 500 66 0 60 0 66 0 min Ready NUM Fig 5 6 Result file for diafiltration ti
54. ct yield sssansisiasrorsmnniisusiise 55 3 14 General CONSIGEFALIONS eesessssssssssssssssssssssssssssusussssssssesseseeseseeseeseessesssssssssssees 55 3 1 2 Measuring HCl erick cacczzscsccattadhavochescashdcivendcaonosritersnnstoartatiatandvliasavvanszisiaviniencinse 55 3 2 Specific actions that increase yield ooecccccccsssssssseseessssessseeesssnunsees 56 3 2 1 Membrane selection c cccecsssssssssssesesssssssssssssssssssssussssssssssssseeseeeeseeesesssesssnsssssssed 56 3 2 2 RECOVER Unasa 58 3 2 3 Denaturation shear temperature and enzymatic action 59 3 2 4 Concentration Gradient AYEM sssssssssssssessssssscsscsecseseeseesessssssssssssnesseeess 60 3 2 5 GOL GY ikresen Gs crteie cls e SANN 62 3 2 6 Summary of concentration gradient and gel layer formation 62 3 3 Flushing product out With buffer oc cccccccccsssssssssssseesssssssecssssesssees 63 3 4 Recovering product from the membrane surface s s s 63 3 5 Product recovery ANd ASSAY specificity ecceeescccccssssssssesseesssesscssssssssees 64 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Contents 3 6 Operating PORGIMETOLS isinru 65 3 6 1 Flux Versus TMP hysissiuinonme ninun a A ERNEA 65 3 6 2 TMP ONA ErOSSIOWinaneran nea a S 66 Sf Scaling UP PATAMELETS eeeecesscessssscccssssssssssseeeseeeeee 67 3 8 Membrane fouling and cleaning procedures 68 39 TrOUDICSHOOTING pissin 69 4 Cell Processing 4L COM TGS SEI Gl sinsin nn
55. ctions The blocks can be expanded to show the instructions within the block 2 5 2 Base Every method must start with a base instruction defining the base for calculating breakpoints Different blocks can use different bases In AKTAcrossflow the default method base refers to column volume and thus needs to be changed to one of the following e Volume the unit depends on the scale defined in the system strategy e Time minutes e SameAsMain all blocks will inherit the base defined in the main block E Main E 0 00 Block Info_Inlets_and_Outlets E 0 00 Block INITIAL_SETUP E 0 00 Block FILL_SAMPLE E 0 00 Block CONCENTRATION_1 CONCENTRATION_1 0 00 Base SameAsMain 0 00 Set_Mark Concentration 1 S 0 00 Block Flowpath_PermConc Flowpath_PermConc 0 00 Base Time 0 00 Permeate_ alve_Block P B Out2 0 00 Block DelayOp05min 0 00 End_Block W 0 00 Block RetFlow_Step1 E 0 00 Block TMP_Step1 E 0 00 Block Chase_Inlet_to_Reservoir E 0 00 Block Stepi_Conc E 0 00 Block Restore_CONCENTRATION 0 00 End_Block H W 0 00 Block RECOVERY 1 A Fig 2 10 UNICORN text pane in the Method Editor 44 AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software 2 5 3 Calls To execute the instructions contained within a block in a method the block must be called by the program When a block is called the instructions in the block are executed in the order that they are wri
56. d membrane operations 40 KTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software 2 4 4 Special features The scouting feature gives automatic support to process development and optimization Method wizards and pre programmed cleaning methods provide a high degree of efficiency in scale up and production processes and simplify purification tasks The ability to generate customized reports saves time and the software is supplied with comprehensive documentation that helps fulfil regulatory requirements UNICORN fully conforms to all applicable regulations including 21 CFR Part 11 Complete documentation and protection Logbook Method System Control Result Method Editor Evaluation Controlled user access fa Validation support Main Menu Fig 2 9 Complete documentation and protection 2 4 5 Control modes AKTAcrossflow system with UNICORN software supports the process control modes commonly used in ultrafiltration diafiltration and microfiltration applications such as TMP control and flux control These control modes can be combined with selectable feed pump instructions such as feed flow rate feed pressure AP retentate flow rate or shear rate UNICORN also reports real time process parameters such as retentate volume concentration factor diafiltration exchange factor total buffer used retentate volume and accumulated permeate volume AKTAcrossflow Meth
57. d not contribute much to the overall recovery of the target molecule so three washes is enough for this process Ea 106 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Applications Ce 6 1 6 Conclusion For recovering B glucosidase from this high cell density P pastoris culture using microfiltration the following optimal process parameters were identified e Use a membrane pore size of 0 45 um e Re circulate feed at a shear rate of 4000 1 s e Perform the process with a flux setpoint of 35 LMH e Concentrate to a factor of 1 4 x to 21 dry cell weight e Wash the cells with three retentate volumes of buffer Fig 6 2 shows the results from a complete run using the parameters above Fig 6 2 Clarification of P pastoris culture using 0 45 um Start AXM cartridge This graph view gives the possibility to monitor for example the decrease in UV i e absorbance at 280 nm as the target protein is transferred to the permeate AKTAcrossflow Method Handbook 11 0012 36 Edition AB 107 E Applications 108 6 2 Purification of Green Fluorescent Protein His GFP His from an Escherichia coli cell homogenate 6 2 1 Objective The objective of this application was clarification of the E coli strain BL21 DE3 homogenate removal of cells and cell debris and the recovery of the target protein in the permeate GFP His is an intracellular protein which is over expressed in the cells and occurs as monomers with a mo
58. d presets or default levels Pleating Folding flat sheet filter media to increase the surface area that can be fitted into a given separation device Generally used in dead end filtration Pore size The range of pore sizes in a membrane The tighter distribution the pore size distribution the better control one has over the filtration process Porosity A measurement of the open space in a membrane The higher the membrane porosity the more pores there are and hence a higher flow rate is anticipated Pressure drop The difference in pressure between two points Protein passage The passing of protein from the feed stream into the permeate stream Pyrogen A substance that produces a fever within a warm blooded animal when injected into the bloodstream Filtration materials of construction that come in contact with injectable liquids must meet pyrogenicity standards Recovery Percentage of the target substance that can be collected in the retentate or permeate solution after processing Retentate The portion of the feed solution that does not pass through a cross flow membrane filter Any component that does not pass through the membrane flows out of the filter and back to the feed container Any components in the return line is called retentate Retention The ability of a separation device to retain an entity of a given size Reverse osmosis Type of crossflow filtrati
59. dbook 11 0012 36 Edition AB 117 6 Applications 118 6 4 3 Optimization of critical process parameters Generation of TMP excursion curves A series of TMP excursions were carried out Five TMP setpoints were tested at three crossflow rates at the initial lgG4 concentration 0 3 mg ml and at the target IgG4 concentration 12 mg ml See Fig 6 8 At the starting concentration the process is pressure dependent over the whole range of TMPs tested However at the target concentration most of the TMP values are in the pressure independent region When analyzing the curves an optimal window of operation for TMP is found between 1 2 and 1 5 bar The evaluation of the correlation between crossflow and flux showed that the highest flux was achieved at a crossflow rate of 125 ml min Flux vs IMP 03 mamil Lo c Ca C CF 125mLmin CF 1d0mLmin cF7S5mLmin B Flex LMH BS 8 8 70 Flex LMH Fig 6 8 TMP excursion results for the IgG4 runs Determination of the optimal concentration factor for diafiltration As with the previous ultrafiltration application using BSA the optimal concentration for diafiltration was established by analyzing the relationship between flux and concentration factor A forty times sample concentration was KTAcrossflow Method Handbook 11 0012 36 Edition AB Applications Ce obtained with the optimal crossflow and TMP settings and a concentration factor Flux versus concentration factor plot was created
60. duct of interest Excessive shear damages protein Membrane selection is key to success Open membranes can lead to turbidity and process control issues Biopharmaceuticals Concentration Concentrating and buffering target molecules Products can be shear sensitive Membrane selection is key Diafiltration downstream Final formulation of bulk drug Includes concentrating protein and exchanging processing substance buffer to final product formulation Table 1 2 Typical uses of cross flow filtration ee 14 AKTAcrossflow Method Handbook 11 0012 36 Edition AB 1 3 1 Introduction Cell processing and concentration diafiltration The terms cell processing and concentration diafiltration are used in this handbook to differentiate between the main types of biopharmaceutical applications Table 1 3 Main types of CFF applications in the biopharmaceutical industry 1 Cell processing Cell harvesting Cell clarification Lysate clarification Recovers cells from fermentation broth Recovers cells all fragments and other particles from the target protein in the cell broth mixture Remove the cell fragments and macrosolutes from the arget protein Cells in the retentate are the product of interest Protein in the permeate is the product of interest Protein in the permeate is the product of interest 2 Concentration Diafiltration Concentrate and buffer prote
61. e stream 2 Toensure removal of trapped air increase the retentate flow rate and run for several minutes until no bubbles appear in the retentate stream 3 Circulate the buffer through the retentate and permeate at a feed pressure of 1 6 to 2 8 bar 25 to 40 psi for four minutes to condition the system for pH and ionic stability 4 Remove the buffer from the feed reservoir Keep buffer in other parts of the system to prevent air from entering the system AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing es 4 5 Operating parameters 4 5 1 Permeate flow control In ultrafiltration applications pressure TMP is applied to the filtration process to drive flux and efficiency In most microfiltration applications such as cell harvesting the flux is often so high even at low TMPs that the flux must decreased to prevent premature fouling of the membrane In AKTAcrossflow system the permeate flow is controlled using the permeate pump AKTAcrossflow software wizard enables the permeate flow rate to be set to control flux Restricting permeate flow generates back pressure on the permeate side of the filter The back pressure effectively lowers the TMP on the feed side of the filter reducing the flux and pore fouling Fig 4 4 gt S Unrestricted permeate flow amp OT Permeate flow control i wW Time Fig 4 4 Using permeate flow control results in more stable flux E KTAcrossflow Method
62. e selection is influenced by process objectives and operating variables This is summarized in Table 3 2 Process objective Cartridge selection Cell concentration Cell protein separation Virus removal Protein concentration Desalting Use microfiltration or open ultrafiltration cartridges for bacterial removal and cell concentration Select membrane pore size based on the specific application Use ultrafiltration cartridges for molecular scale applications such as desalting and protein concentration Solution variables Cartridge selection Solids loading Viscosity Shear sensitivity High solids loading and high viscosity fluids work best with larger hollow fibers and shorter lengths With fluids that are not shear sensitive small diameter fibers can be used Other variables Cartridge selection Time constraints Increased membrane area and larger housing size shorten production time Pump constraints Larger diameter large surface area cartridges with many large fibers require pumps with high flow rate capacities Heat sterilization Choose autoclavable or steam in place models 58 Table 3 2 The influence of process objectives and operating variables in selecting a hollow iber cartridge for microfiltration 3 2 2 Yield decreases as the quantity of process fluid that cannot be recovered from a system increases AKTAcrossflow software Method Wizard supports two meth
63. e solution Often reported in NTU nephelometric turbidity unit Ultrafiltration The separation of macrosolutes based on their molecular weight or size Upstream The feed side of a separation process Upstream Cellular separations including cell lysates cell processing harvesting clarification and cell culture perfusion Viral clearance The removal of viral contamination using methods such as filtration chromatography heat and low pH Viscosity A measurement of a fluid s resistance to shear A slow flowing liquid such as gear oil has a higher viscosity than a free flowing liquid such as water In a given separation process higher viscosity Newtonian fluids are operated at a lower flow rate through a cartridge than do lower viscosity fluids Void volume Quantity of fluid required to completely fill a section of piping Also the amount of open space within membrane filter media Water flux Measurement of the amount of water that flows through a given membrane surface area in a set time See also flux The water flux test is commonly used to assess cleaning efficacy Yield The amount of particulates or molecules of interest product that can be recovered from the cross flow filtration process Also called recovery AKTAcrossflow Method Handbook 11 0012 36 Edition AB 135 136 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Appendix C Shear effects on proteins and
64. ebe teases cdosiscectbaincs 4 9 2 Carnage selectii csriscstes 65 8 sscselsscecnsdoaensctonsshiscrestestanscnsvlastnansanGurtaitadgeaten 4 10 Filter and system preparation cceessecsssssssscssssessssusseessscecsesssssssassesesseee 4 11 Operating parameters 4 11 1 Permeate flow CONMTFOD ceeeceesssssssssssssssssssssssssseee 4 12 Three examples of clarification strategies sty 4121 MOMMallign Ce1Seeecsccssssssssssssssssccssccsssssssssssssnssssnssnusssssseseeseeseseeseeeeeeessessnssssssass 412 2 BOCLCNOLCONS srs E tiettanaiaunadaentnas ART AS 0 E E A A E EEEE 5 Concentration and Diafiltration 51 AIMEPOAUCHON sinuni 5 2 Product and process considerations a BS IDIGHROUON aitescintiisciicetiindnnansinnneineamnAdhnnscnaatan 5 3 1 EFNCICNC Unkon naan in Ninis 5 3 2 Discontinuous diafiltrati Nisisisseessiiiiiinininisieninii 5 3 3 S guentialdiafiltrat oN iiisiiseoeiianiinnt 5 4 Membrane and cassette selection ae 5 4 1 M mbrane selectio Nasenne 5 4 2 Cassette SCLCCtION eeeeessessssssssscssssssssssesecseeesesssessssssssssssssssitususssssscssssecseeeeeeees 5 5 Device and system preparation ANd cleaning sssssesccsscccccsseeeesees 96 Ee AKTAcrossflow Method Handbook 11 0012 36 Edition AB Vv Contents 5 6 Operating conditions for Kvick Start cassette o ecceseeseeseececeee 97 Sf COMCEFERATIOM ACTON s seccstscessesesrracssssarmcercerrtnasmnctenroannnaten 97 59r Optimizat OF IMP sisssss ssrsiciessovsssisssssscasvancennnas
65. em should be programmed to perform the following steps 1 Atthe end of the process of harvesting cells or concentrating a protein close the permeate valve or reduce the feed pressure to 0 3 bar 5 psi 2 Reduce the cross flow rate to 1 10 of the recommended processing cross flow rate 3 Circulate the remaining product for 15 minutes This procedure will help recover product that has accumulated on the surface of the membrane 4 Recover the product by pumping it from the system to a collection vessel AKTAcrossflow Method Handbook 11 0012 36 Edition AB 63 Cross flow filtration process considerations es 3 5 Product recovery and assay specificity Measuring recovery requires a reliable assay for the product The assay must have a specificity for only the product of interest and not any degradation products that may be present Mass balance estimates for recovery require feed samples before and after filtration and permeate samples after filtration An analysis of the permeate samples provides insight into the rate of product passage over the processing time The formula mass balance determination is as follows VCs V C VpCp VhCh Where e V volume e C concentration e s starting e r retentate e p permeate e h hold up eee 64 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cross flow filtration process considerations 3 6 Operating parameters 3 6 1 Flux versus TMP In CFF the key optimization parameter is the flu
66. ength and thus the device s ability to maintain integrity under pressure Media exchange A filtration step used during aseptic cell culturing to remove growth media or fermentation broth so that fresh growth media can be added to the bioreactor Membrane A thin layer of a highly engineered material with controlled pore size and used to separate particles biological matter and molecules from a solution Membrane The degree to which the original performance of a recovery membrane can be restored by cleaning Microfiltration The process of removing particles from a liquid by passing it through a porous membrane under pressure Microfiltration usually refers to removing submicron size particles Micron micrometer um One one millionth of one meter AKTAcrossflow Method Handbook 11 0012 36 Edition AB 131 Microporous membrane A thin porous film or hollow fiber having pores ranging from 0 1 to 10 um Cross flow microfilters typically have pores ranging from 0 1 to 1 0 um Minimum process volume Also called minimum operating volume The least amount of fluid able to be handled effectively by a filtration system Molecular weight Mass of one molecule of a nonionic substance in atomic mass units Molecular weight cut off MWCO See NMWC Nanofiltration Separation processes targeted for solutes having molecular weights from 500 to 1 000D See Molecular weight cut
67. ent diafiltration due to membrane polarization effects and to protein precipitation AKTAcrossflow Method Handbook 11 0012 36 Edition AB 97 Concentration and Diafiltration 5 8 Optimization of TMP TMP excursions are an important part of process optimization Increasing TMP when ultrafiltering pure water results in a proportional increase in flux With a process fluid that contains solutes there is a similar pressure dependent region of operation where an increase in TMP corresponds to an increase in flux but there is also a pressure independent region where an increase in pressure does not lead to an increase in flux Fig 5 3 The pressure independent region is a result of the build up of solutes at the surface of the membrane called a gel layer which creates resistance to flow This gel layer can be reduced by increasing the crossflow Pressure dependent region Pressure independent region Process Fluid Water wb CF gt CF gt CF CF gt CF Filtrate Flux Rate Cross Flow Transmembrane Pressure Fig 5 3 TMP excursions are an important part of process optimization 5 8 1 Concentration The standard procedure is to perform a TMP excursion experiment in which a series of TMP setpoints is measured at different crossflow rates From these experiments the effect on flux is evaluated and optimal crossflow and TMP may be identified AKTAcrossflow software contains a method Wizard where the user can input the de
68. ermine the relationship between process conditions and yield losses due to shear As a quick feasibility study a protein solution can be circulated across the feed retentate path and the bioactivity of the protein analyzed to relate protein activity to process time on a number of pump passes Where feasible low pressures and low pump speeds should be used to minimize shear in the flow path When using hollow fiber filters cross flow rates are often expressed in terms of shear rate This convention makes it possible to scale up or down between cartridges By using a shear reference chart it is possible to approximate the flow rate that will yield the same shear at the new scale The formula below can be used to calculate flow rates and shear rates for hollow fiber units y 4q mR Where e y shear rate sec e q flow rate through the fiber lumen cm3 sec fiber e R fiber radius cm Calculation of shear for cassettes is more complicated because of the influence of screens and is beyond the scope of this handbook KTAcrossflow Method Handbook 11 0012 36 Edition AB 59 Cross flow filtration process considerations 60 Temperature For heat sensitive proteins the process solution temperature can be modified in a number of ways during processing e Precondition the AKTAcrossflow system with cooled buffer before starting e Lower the protein temperature before beginning the filtration process e Use chilled buffer during diafiltration
69. ers C 2 Shear studies on cell suspensions A P984 pump fitted with KTA sanitary tubing i d 3mm max flow rate 800ml min was compared to a peristaltic pump fitted with Marprene II 25 hose i d 4 8mm max flow rate 690 ml min using a sample of wild type CHO cells Cell integrity was measured as a function of LDH activity and cell viability by counting free nuclei In addition temperature and cell densities were measured A control sample was kept in an end over end mixer at low speed to monitor the natural increase in LDH activity A sample of the cell suspension was homogenized three times at 800 bar to establish the LDH maximum value AKTAcrossflow Method Handbook 11 0012 36 Edition AB C 2 1 Comparison of pump types Compared with the peristaltic pump the P984 pump showed a higher cell viability a higher cell count a lower increase in temperature and a lower increase in LDH activity Fig C 4 Viability second run 100 97 4 94 si 90 90 80 70 60 3 50 40 30 20 10 0 T T T Sart P984 Watson Marlow_505U Control Fig C 4 400 ml of CHO cell suspension was recirculated for 370 cycles at 400 ml min for each pump Measurements were made of cell viability cell density LDH and temperature The results of a comparison of a stand alone P984 pump and KTAcrossflow system using the criteria of cell viability and cell
70. f the cartridge due to the high permeability of these membranes in the presence of permeate pressure This phenomenon is most often associated with the operation of microfiltration membranes using permeate flow control Sterilization A process that removes destroys all microorganisms from a solution or a filtration system See Autoclave EtO Ethylene oxide sterilization Gamma sterilization Surface filter A filter in which particles larger than the pores are retained on the surface of the filter Tangential flow filtration See cross flow filtration Thermal stability The ability of a membrane and filtering device to maintain its performance during and after exposure to excursions of temperature such as the elevated temperatures experienced during autoclaving or steam sterilization AKTAcrossflow Method Handbook 11 0012 36 Edition AB Throughput The volume of solution that will pass through a separation device before the permeate output drops to an unacceptable level Transmembrane The pressure differential between the upstream pressure TMP feed and downstream permeate sides of a membrane It is calculated by TMP Preed P retentatel 2 Ppermeate Turbidity Measure of relative clarity of a liquid Measurements are based on the amount of light transmitted through a sample The more light that is scattered by fine solids or colloids the less clear and more turbid th
71. flow applications the membrane pore size selection is based on the size of the target molecule The general guideline for selecting a membrane for product concentration is to start with a NMWC that is 3 to 5 times smaller than the target molecule For example a 50 kD or 30 kD membrane would be a suitable choice to retain IgG 160 kD and a 30 kD or 10 kD membrane would be a suitable choice for albumin 66 kD 5 4 2 Cassette selection Kvick Start cassettes are used for downstream processes such as concentration of proteins and diafiltration before chromatography and for final concentration and purification of post chromatography product Kvick Start cassettes incorporate polyethersulfone membrane and when operated correctly a high level of protein recovery can be expected As with any polymeric membrane a low level of non specific protein binding is possible 5 5 Device and system preparation and cleaning Safety and operating instructions are included with all cassettes and these should be followed when preparing the cassette for use A new cassette contains an aqueous solution of 0 1 to 0 2 N NaOH and 20 to 22 glycerin Before using a new or previously stored Kvick Start cassette the storage solution must be AKTAcrossflow Method Handbook 11 0012 36 Edition AB Concentration and Diafiltration removed In addition some applications require the completion of other preparatory steps before using the cassette The main preparator
72. h the gel layer providing most resistance 62 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cross flow filtration process considerations 3 3 Flushing product out with buffer Flushing product out of the filtration system with buffer enables the highest yield to be obtained In this technique the product should be slightly over concentrated collected from the system and a small volume of buffer or permeate added into the system The added volume flushes out the residual product from the feed retentate loop AKTAcrossflow system should be programmed to perform the following steps 1 Asthe CFF process nears completion decrease the pump speed to minimize flow rate vortexing in the feed tank and the possibility for product foaming 2 When the slightly over concentrated volume is reached pump the concentrated product to the collection vessel 3 Add an appropriate volume of buffer or permeate to the reservoir via the transfer pump The buffer should be circulated for two to three minutes with the permeate valve closed to help bring the residual product into suspension 4 Pump the buffer solution from the system into the collection vessel 3 4 Recovering product from the membrane surface Recovering product from the membrane surface enables the most highly concentrated product to be obtained In this technique product is recovered from the membrane surface without adding buffer or permeate to the system AKTAcrossflow syst
73. he level to which cells can be concentrated during cell clarification varies Some typical values are as follows e E coli cells 5x concentration e Yeast cells 2x concentration e Mammalian cells 10x to 20x concentration 4 8 Lysate Clarification The components in lysates have a propensity to foul membrane pores To minimize pore fouling and to enable the filter to operate under equilibrium conditions initial lysate clarification trials often include a constant volume wash with little or no concentration Because the starting volume is typically exchanged 5x during washing the wash is large Small initial volumes 100 to 200 ml are typical When working with lysates ultrafiltration filters are normally used The small pores in ultrafiltration filters help minimize pore plugging by the submicron particles found in lysates AKTAcrossflow Method Handbook 11 0012 36 Edition AB 81 Cell Processing oS ou The key process variables during clarification are permeate flow rate flux and TMP Cells and broth ae Downstream BiOreaCtOr u Clarification processing steps Capture and Finished product disposal of cells Cells Fermentor and broth Cell Mechanical harvesting disruption of cells Clarified product of Downstream e interest processing Lysate steps clarification Finished product Cell debris and unwanted components Fig 4 5 Two upstream clarification processes A Separating a ta
74. here is a continuous flow of rinsing fluid Feed pump P 984 Optional path RS1 t Waste Rinsing solution Transfer amp Permeate pump P 982 Optional path N RS1 j E Waste Rinsing solution Fig 2 4 Piston rinsing system Feed pump P 984 and Transfer and Permeate pump P 982 KTAcrossflow Method Handbook 11 0012 36 Edition AB 31 AKTAcrossflow system components and software 2 2 4 Reservoir The reservoir contains the liquid sample to be processed It provides a gentle but efficient mixing of the process liquid with returning retentate and additional liquid added via the transfer line Permeate may be recycled into the reservoir to achieve steady state conditions during process development studies Lid Top flange Float Stir bar Bottom end plate Flow outlet Reservoir level sensor Fig 2 5 Reservoir 350 ml A magnetic stir bar in the bottom of the reservoir ensures uniform mixing between the bulk fluid the retentate returned from the filter and liquid added via the transfer line An integrated level and temperature sensor continuously monitors and reports the retentate liquid volume and the temperature of the liquid fed into the filter device The level sensor can also be used to protect the filter against the introduction of air 32 AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software The reservoir is equipped with a
75. iber cartridges the housing normally has feed and permeate ports and in the case of cross flow filters a retentate port All of these ports may be used to control the flow parameters of fluid into and out of the housing and through the membrane Cassette A device used for cross flow filtration typically in a rectangular form comprised of stacked flat sheets of membrane integrally bonded together Most cassettes are typically designed to fit into a standard cassette holder where the feed permeate and retentate ports mate with appropriate fittings on the cassette holders Cell harvesting The process of concentrating dewatering the cell mass after fermentation Cell slurries in excess of 70 wet cell weight are achievable The cells may also be washed to prepare them for further processing such as freezing or lysing Unlike clarification processing with cell harvesting the cells are the target material Channel height The height of the path that the feed retentate solution must pass through for a flat membrane cassette Channel length See flow path length 126 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Chemical compatibility The ability of the components of a filter to resist chemicals that can influence the filter s performance For example some chemicals could cause the filter to swell or dissolve filter components Repeatable performance requires that filters are resistant to all the
76. ight be as high as 120 L m Operating at a higher flux rate usually decreases throughput capacity For high yields working with monoclonal antibodies derived from CHO cells it is normally possible to concentrate the cells 10X and follow with a 3X to 5X wash AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing Using the 30 cm long Start AXM hollow fiber cartridge with 50 cm and with a starting volume of 400 to 500 ml the process time will be approximately two hours 4 12 2 Bacterial cells E coli and related bacterial cells have been used for many years for the expression of a wide range of recombinant proteins vaccines and enzymes The fermentation times can range from under a day to a week Due to the short doubling time of these cells prolonged fermentation can result in a significant cell mass Moreover the nutrient media is usually much more complex and less purified than the material used with mammalian cells As a result separating these cells from the target protein can be a more complex process Membrane and cartridge selection For relatively small proteins lt 40 kD the open ultrafiltration membranes rated at 500 kD and 750 kD should be tested first These membranes will provide a stable flux rate and resist rapid fouling For large proteins microfiltration membranes rated at 0 1 or 0 2 um should be used but only in conjunction with permeate flow control In order to make a comparison of any of the
77. ime constraints Increased membrane area and larger housing size shorten production time Heat sterilization Choose autoclavable or steam in place models for scale up Table 4 3 The influence of process variables and feed solution on the CFF process AKTAcrossflow Method Handbook 11 0012 36 Edition AB 77 Cell Processing 78 4 3 Other product and processing factors There are some other factors that should be considered when defining a product and developing a process A faster processing time may offer the benefit of less exposure of the product to shear forces temperature increases and enzymatic action 4 4 Preparation before use 4 4 1 Microfiltration cartridge Using a new microfiltration cartridge requires no rinsing When using Start AXM and Start AXH hollow fiber cartridges proceed to the water flux step AKTAcrossflow software automatically measures water flux normalizes the results and presents and stores the data for analysis 4 4 2 Conditioning the system with buffer Before processing samples it is recommended to precondition the system with a buffer similar in pH and ionic strength to that of the sample Conditioning the system removes trapped air and minimizes unwanted chemical reactions between the sample and the wetted parts of the system 1 Circulate 1 liter of buffer through the system with approximately 0 3 to 1 bar 5 to 15 psi retentate pressure Run until no bubbles appear in the permeat
78. in Protein in the retentate is the product of interest Table 1 3 Application terminology AKTAcrossflow Method Handbook 11 0012 36 Edition AB 15 Introduction ee 1 4 Technical parameters An optimized CFF process starts with a characterization of the feed material as follows e Temperature sensitivity of the feed material e pH stability range of the target molecule e Sensitivity of the target molecule or cell to shear forces e Target molecule solubility e Availability of a suitable assay for monitoring yield and finished product activity e Is it possible to concentrate the feed to the target concentration given the starting volume and the system s working volume e Will increases in viscosity due to cell mass concentration exceed the capability of the CFF system eS 16 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Introduction 1 5 Approximate times for completing CFF runs Typically the time for completing a cross flow filtration run using AKTAcrossflow system can range from 3 to 8 hours A run can be divided into three parts preparing the filter and system for processing conducting the separation concentration diafiltration process and cleaning and flushing the system and filter for storage Table 1 4 AKTAcrossflow completes many of these tasks automatically In add ition it fully monitors and records process parameters throughout the run freeing operators to perform other tasks Fu
79. in the Evaluation Wizard The results are shown in Fig 6 9 As described earlier the highest value on the Y axis at the highest concentration factor indicates the fastest diafiltration with the lowest buffer consumption The shortest process time and the lowest buffer consumption occurs when the diafiltration is run at the target concentration factor of 40 since the diafiltration time optimization parameter Flux concentration factor is continually increasing ConcFactor Flux vs ConcFactor 2 ConcFackr Fig 6 9 Diafiltration time optimization for the IgG4 sample AKTAcrossflow Method Handbook 11 0012 36 Edition AB 119 6 Applications 6 4 4 Concentration and diafiltration process The complete process for concentration and diafiltration of an IgG4 solution was run using the optimized parameters TMP 1 3 bar and feed flow 42 ml min The total sample processed was 2000 ml using three Kvick Start cassettes connected by a manifold 150 cm total area at constant TMP control mode The sample was concentrated 40 times from an initial concentration of 0 3 mg ml to 12 mg ml and then a five times buffer exchange was performed The collected data of various parameters are shown in Fig 6 10 From the curves in Fig 6 10 it can be seen how the flux decreases as the concentration factor increases during concentration In the diafiltration step UV the absorbance at 280 nm decreases from 2100 MAU to 40 mAU as some of the contaminants in the pr
80. ing Step Cartridge filter Feed reservoir contains mostly cells and buffer Permeate contains buffer and unwanted components Fig 4 2 Cell harvesting often includes a washing step to help flush unwanted components from the fermentation broth AKTAcrossflow Method Handbook 11 0012 36 Edition AB 13 Cell Processing 74 Successful cell harvesting relies on knowledge of the product such as e Robustness of the cultured cells e Starting volume and concentration of cells e Desired finished concentration and volume e Desired yield and quality viability of the cells 4 2 2 Typical steps in a cell harvesting method The typical steps in a cell harvesting method are outlined in Table 4 1 Step Comments Rinsing Rinse storage solution from filter CIP Circulate cleaning solution to clean system and filter Water flush Flush cleaning solution from system Water flux test Determines performance of filter before processing Buffer conditioning Conditions filter and system components before adding product to minimize adverse chemical reactions Cell harvest Harvest cells concentration Cell washing Helps drive unwanted components through the diafiltration membrane Product recovery Recovers cells from AKTAcrossflow system Buffer flush Flushes residual product from system without risking precipitation of components on the membrane or flow path CIP Recirculate cleaning
81. ing rate is selected as default by the control software At higher retentate volume where the float is not in contact with the stir bar the user can select a higher mixing rate The following mixing rates are recommended as maximum mixing rates that will ensure sufficient mixing for all conditions e 375 ml reservoir 200 rom e 1200 ml reservoir 300 rpm AKTAcrossflow Method Handbook 11 0012 36 Edition AB 33 AKTAcrossflow system components and software 2 2 7 Materials The reservoir consists of the following material e Glass tube Borosilicate e Bottom end plate top flange and lid Polyetherimide e Sealing lid Thermoplastic elastomer e Float Polypropylene e Stir bar Polytetrafluoroethylene 2 3 Sanitary design AKTAcrossflow has been designed to allow effective sanitization using 1M sodium hydroxide NaOH as a sanitizing agent Sanitization is the use of a chemical agent to reduce a microbial population to an acceptable predetermined level Microbial challenge tests are used to evaluate the efficiency of the sanitizing agent A study including two challenging organisms has been carried out The system was subjected to a high level of microbial challenge 1x10 Colony Forming Units CFU ml The results show that the method used efficiently reduced the numbers of viable organisms and was sufficient for sanitization 34 AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and softw
82. is 40 LMH i Ine Thee vi foo foo Unstable resulting TMP joo Stable resulting TMP a a 20 ao echo a Fig 6 3 Results of a flux excursion to determine the maximum flux set point for a given shear rate 109 E Applications 110 When using the 0 1 um membrane pore size with a shear rate of 16000 sec a high flux 45 LMH could be achieved and the recovery was relatively high 83 see Table 6 4 The highest product recovery 93 was reached using the 0 2 um pore size at this high shear rate but the flux was low 20 LMH The high protein recovery is offset by the low flux and resulting longer process time This however could be compensated by a larger membrane area but would result in higher costs in pump capacity and membranes In summary for the fastest process time and acceptable recovery the 0 1 um membrane pore size was chosen at a shear of 16000 sec Membrane pore Shear rate Flux Protein transmission size ratings 1 sec Imh of total content 0 1 um 10 000 35 70 0 1 um 14 000 40 85 0 1 um 16 000 45 83 Table 6 4 Results of flux optimization 6 2 5 Optimization of retentate wash for protein recovery The effect of washing buffer on protein recovery is shown in Table 6 5 During the three times concentration 50 60 of the total GFP His content passed into the permeate depending on the membrane pore size used Increasing the number of washes
83. kly creates the desired process methods Table 6 3 summarizes the results of the membrane screening experiments Protein recovery was 75 and 83 respectively for membranes with pore sizes 0 1 um and 0 2 um and these were selected for further investigations AKTAcrossflow Method Handbook 11 0012 36 Edition AB Applications Ce ae Pore size Concentration Wash volume Protein factor recovery 750 kD 5 3X 61 0 1 um 3 3X 75 0 2 um 3 3X 83 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Table 6 3 Results of pore size screening with GFP His 6 2 4 Optimization of shear flux settings The objective of these experiments was to determine the maximum flux rate for a given shear rate accessed by TMP stability A constant shear rate was selected in the example shown in Fig 6 3 the shear rate was 8000 sec and the flux increased in a stepwise manner while monitoring the TMP These experiments were done at the target concentration of the sample and the permeate was recycled back into the retentate A rapid increase in TMP indicates reduced membrane permeability due to gel layer formation The process was repeated for shear rates of 8000 and 16000 sect for both the 0 1 and 0 2 um membrane pore sizes and the data analyzed Fig 6 3 shows the results of an experiment using a shear rate of 8000 sec and a pore size of 0 1 um In this test it can be seen that the maximum flux set point which results in a stable TMP
84. l experiments and was not included in further tests The pore sizes were tested with two different methods e Noconcentration and 4 times washing of the retentate e 1 4x concentration and 4 times washing of the retentate The washing steps are essential for improved protein recovery particularly when clarifying a high solid content sample with little or no possibility of an initial concentration step Since as can be seen in Table 6 1 approximately 100 protein recovery was obtained using the 0 45 um filter this membrane was selected for further process optimization Membrane Shear rate Recovery pore size 1 s No concentration 1 4x concentration 0 1 um 5500 58 60 0 1 um 8000 56 0 2 um 4000 54 1 0 2 um 4500 67 72 0 45 um 4000 62 100 0 45 um 6000 70 Table 6 1 Membrane selection AKTAcrossflow Method Handbook 11 0012 36 Edition AB Applications Ce 6 1 4 Optimization of shear flux settings For the membrane pore size of 0 45 um the effect of shear rate on maximal achievable flux was studied using methods with no concentration and 1 4 x concentration A higher shear rate results in increased flux which shortens process times However at a shear rate of 6000 1 s the concentration of 1 4 x could not be achieved because the feed pressure exceeded its operating limit With no concentration at this shear rate a flux rate of 55 LMH could be achieved Increased flux gives shorter p
85. l harvesting for more details Table 4 7 describes typical starting conditions for the clarification of different cell types Bacterial Mammalian Yeast Adeno Clarification of fermentation cell culture fermentation associated VLP from a yeast clarification of Mab clarification of virus lysate target protein clarification target protein clarification expressed expressed from 293 or HeLa extracellularly extracellularly cells 5X concentration followed by a 3 to 5x diafiltration 10X concentration followed by 3X diafiltration Partial concentration 1 5 to 2X at best followed by 3 to 5X diafiltration 5X concentration followed by 5X wash First try a constant volume diafiltration process with no initial concentration with high shear and permeate flow control set at 20 LMH With good transmission a 5X wash may suffice For large target proteins use microfiltration membranes with permeate flow control set at 20 to 30 LMH For smaller molecules use 750 or 500 kD UF membranes with unrestricted permeate flow and TMP readings at 1 to 1 5 bar Use 0 2 or 0 45 microfiltration membranes with permeate flow control set at 30 LMH No retentate back pressure This process description is for the removal of cells and optimal recovery of the target protein Membranes rated at 750 kD UF and 0 1 micron MF have worked well with unrestricted permeate flow If the cell density is
86. lectric companies Alconox is registered trademark of Alconox Inc Microsoft and Excel are either registered trademarks or trademarks of Microsoft Corporation in the United States and or other countries All goods and services are sold subject to the terms and conditions of sale of the company within GE Healthcare which supplies them GE Healthcare reserves the right subject to any regulatory and contractual approval if required to make changes in specifications and features shown herein or discontinue the product described at any time without notice or obligation Contact your local GE Healthcare representative for the most current information 2006 General Electric Company All rights reserved GE Healthcare Bio Sciences AB a General Electric company GE Healthcare Bjorkgatan 30 751 84 Uppsala Sweden GE Healthcare Europe Gmbh Munzinger Strasse 5 D 79111 Freiburg Germany GE Healthcare UK Ltd Amersham Place Little Chalfont Buckinghamshire HP7 9NA UK GE Healthcare Bio Sciences Corp 800 Centennial Avenue P O Box 1327 Piscataway NJ 08855 1327 USA GE Healthcare Bio Sciences KK Sanken Bldg 3 25 1 Hyakunincho Shinjuku ku Tokyo 169 0073 Japan Asia Pacific Tel 852 2811 8693 Fax 852 2811 5251 e Australia Tel 61 2 9899 0999 Fax 61 2 9899 7511 Austria Tel 01 57606 1619 Fax 01 57606 1627 Belgium Tel 0800 73 888 Fax 03 272 1637 Canada Tel 800 463 5800 Fax 800 567 1008 Central East
87. lecular weight of approximately 30 kD It has been observed that GFP His adsorbs to cell debris at pH values lower than 8 2 so the pH of the homogenate prepared from 25 g l cell suspension was adjusted to 8 3 Hollow fiber cartridges are often chosen for E coli homogenate clarification because of their open feed channel design Optimization of the clarification process involves selecting a filter pore size and operating conditions that retain solids while yielding the highest recovery of the target protein in a specified process time CFF experiments were carried out on AKTAcrossflow system using 50 cm Start AXM hollow fiber cartridges 6 2 2 Process Optimization The strategy for process optimization was as follows e Membrane selection e Optimization of shear flux settings e Find the number of retentate washes needed for improved protein recovery 6 2 3 Membrane selection The objective was to choose the smallest membrane pore size which provides an excellent recovery of target protein Smaller pore sizes reduce the tendency for particles to become embedded in the membrane Three membrane pore sizes were screened 750 kD 0 1 um and 0 2 um at a cell concentration of 25 g l The sample was concentrated three times and then washed three times with buffer As mentioned previously in microfiltration applications full process runs have to be completed under different experimental conditions but the Method Wizard in AKTAcrossflow software quic
88. lecules are separated from a solution while larger molecules remain in the retentate In general microsolutes are easily washed through the membrane so that for a fully permeated species approximately three volumes of diafiltration solution will eliminate 95 99 of the microsolute Table 1 1 CFF key terminology The appendix includes a glossary that defines additiona terms and concepts AKTAcrossflow Method Handbook 11 0012 36 Edition AB 13 Introduction eS 1 3 Applications overview CFF is used in research product development and production in the biopharmaceutical industrial and medical industries Table 1 2 Industry Application Comments Biopharmaceuticals cell processing upstream processing Harvesting cells from fermentation broths Cells harvested include bacteria insect Cell is the product of interest Most cells are insensitive to shear Wide range of membranes will work Cells are large compared to pores A broad range of microfiltration membrane may be used Separating proteins from fermentation broths Separating proteins from intact cells and potential cell debris Protein in broth is product of interest Usually mammalian cells Shear damages cells and proteins Yeast cells are difficult to process Membrane selection is the key to success Open membranes can lead to turbidity and process control issues Separating proteins from cell lysates Protein in lysate is pro
89. leted and an evaluation audit trail shows all access and operations performed All batch details are included in a single result file for easy back up and all method details are embedded in a result file In addition the design of the evaluation module allows the automation of repetitive tasks ooo AKTAcrossflow Method Handbook 11 0012 36 Edition AB 53 AKTAcrossflow system components and software Evaluation ConclG001 Product _ x UNICORN Local Fil Annika Result lgG ConclgG001 res 2i File Edit View Integrate Operations Procedures Window Help 89 ETLE LEFT Se ConclgG001 Product RetFlow Recent Runs Files Find Conclg 001 Product Logbook B Recent Rests ra JE 91 81 min 125 0 ml min U EmptyReservs A EmptyReserv at a iar ahi alae aie Taare eT TST ce oe a EmptyReservd f120 EmptyReserve EmptyReserv EmptyReserve PostProdIgGSy EmptyReservg EmptyReserv EmptyReserv EmptyReservg 20 EmptyReserv EmptyReservs EmptyReserv EmptyReserve EmptyReserv 70 EmptyReservi EmptyReservi EmptyReservd s0 4 EmptyReservi EmptyReservi EmptyReserv EmptyReserv EmptyReservd 45 PostProdigGSy TonsigG00 1 Product PermFlow ConclaG001Product TMP i ee EEEEEEEEEEEEEEEEEEEEEEEEE E ee c a gt Remove Remove all o Refresh Pref ele WO _ __ __ _ 5 3 3 Ready NUM
90. lfone e Luer Lok fittings Polycarbonate e Potting Epoxy e Wetting fluid Glycerine in case of UF membranes MF membranes are delivered dry AKTAcrossflow Method Handbook 11 0012 36 Edition AB Introduction 1 12 Testing procedures 1 12 1 Water flux test The water flux test measures the flow rate of water through the membrane under controlled conditions The flow rate provides an indication of the performance capability of the membrane By tracking the water flux measurements over time itis possible to determine the effectiveness of cleaning cycles and determine when a cassette reaches the end of its service life A filter will normally lose up to 20 percent of its performance as measured by the water flux test after its first use and cleaning The performance level should remain stable from that time forward Water flux testing is usually carried out when the filter is new and after each use or cleaning cycle AKTAcrossflow system software contains a method to automatically measure the water flux of the filter and calculate and plot the results in LMH bar Details of the water flux test procedure can be found in AKTAcrossflow User Reference Manual 1 13 Quality assurance and documentation Quality assurance documentation and evidence of consistent performance process validation are key process requirements when using CFF systems and filters in biopharmaceutical applications 1 13 1 Hollow fiber cartridges GE Healthca
91. lts from routine concentration diafiltration and cell processing runs It allows any process parameter captured as a curve in a given result file to be plotted on either the X axis or the Y axis AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software ee 2 7 Comprehensive report generation UNICORN software provides a flexible environment for presenting data and information It includes configurable report formats for easy report generation and has a simple click and print report interface All information about a run can be included and reports can be printed automatically after a run It is also possible to export data to Microsoft Excel for further analysis if required 2 8 Security UNICORN software has an extensive data analysis capability while maintaining data security Feature Function Access security Only authorized users can access UNICORN Each user is assigned an access level which defines the functions that the user is permitted to use Data security Result files from an ongoing run can be saved automatically at preset intervals to minimize data loss if the system fails The results are saved locally if the network communication fails Electronic signatures Method and result files can be signed electronically for enhanced security and accountability Table 2 6 Main security functions in UNICORN for AKTAcrossflow Original data cannot be modified or de
92. m Xflow sys P1 02 with rinsing system 500 450 400 350 300 250 200 150 100 50 0 0 50 100 150 200 250 Cycle no Absorbance 620nm mAu Fig C 1 Influence of using a piston rinsing system on protein aggregation using P984 pump KTAcrossflow Method Handbook 11 0012 36 Edition AB 137 138 180 4 Z 160 a 100 e 10 mM NaOH 120 20 EtOH 100 gt lt 100mM NaOH A 80 PBS buffer S 60 PBS buffer and 10 glycerin 40 50 mM NaOH 3 20 0 0 50 100 150 200 250 Cycle no Fig C 2 Influence of rinsing fluids on protein aggregation C 1 2 Comparison of different pump types Human IgG 20 mg ml in 100 mM PBS buffer pH 7 0 containing 20 mg ml glycine was used as the sample and the absorbance at 620 nm measured as a function of the number of cycles Fig C 3 Protein aggregation was highest when using the rotary piston pump P984 pump showed the lowest protein aggregation of all pumps 450 400 Peristaltic pump 350 300 250 200 150 P984 synchronized and with rinsing system second run IVEK P984 synchronized and with rinsing system Peristaltic pump second run 100 50 Absorbance 620nm 0 50 100 150 200 250 Cycle no Fig C 3 Comparison of different pump designs and the formation of protein aggregates as a function of cycle numb
93. m processing Most downstream concentration and diafiltration processes use membrane cassettes The main steps in the downstream purification process are illustrated in Fig 5 1 and Fig 5 2 Membrane Growth Media or Buffers Pyrogen die Membrane Membrane Membrane Membrane E Media Cell Harvest Concentration Exchanger Fermentation Diafiltration Polishing Membrane Purification Membrane Capture m a Sample Concentration Sample Concentration Buffer Exchange Buffer Exchange Membrane Membrane Membrane Sample Concentration Buffer Exchange Virus Removal Sterilization Final Product Fig 5 1 Downstream steps in the purification of IgG SEE AKTAcrossflow Method Handbook 11 0012 36 Edition AB 93 Concentration and Diafiltration 1 Start of concentration 2 End of concentration Feed reservoir Product Small molecules and ions 3 Start of diafiltration 4 End of diafiltration Feed reservoir with Feed reservoir Permeate Permeate collection collection vessel Result vessel Product concentrated but still in original buffer Feed reservoir d Filter with product in concentrated new buffer product in original buffer eee Permeate Result Permeate collection Product concentrated collection vessel small molecules and ions vessel removed and product in buffer of choice Fig 5 2 Comparison of concentration and diafiltration processes 5 2 Product and proce
94. me optimization Flux is the grey curve concentration actor is the blue curve es 100 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Concentration and Diafiltration See A curve of Flux concentration factor versus concentration factor can be created in the Evaluation module which enables the optimization of diafiltration time Fig 5 7 The highest value on the y axis at the highest concentration identifies the fastest diafiltration with the lowest buffer consumption This example also shows that diafiltration takes the same time if performed at four times or five times concentration because the decrease in retentate volume at five times concentration is offset by the decrease in flux ConcFactor Flux vs ConcFactor T z 8 1 0 1 5 2 0 25 3 0 3 5 4 0 45 5 0 5 5 6 0 ConcFactor Fig 5 7 Diafiltration time optimization 5 8 3 Diafiltration factor The Diafiltration factor DF is the percentage of original buffer remaining in the feed Diafiltration factor Sample Volume Buffer Volume The volume of buffer required to achieve a desired diafiltration factor can be calculated using the following formulae Continuous diafiltration C7 C 1 B F One shot diafiltration C C 1 DF 1 Discontinuous diafiltration C C 1 2 F Where e C final concentration e Cy Original concentration e 8 sample turn over ratio per unit time a AKTAcrossflow Method Handbook 11 0012 36
95. n the initial processing volume L Method Laion UN TILED DOSE Sy 20 poet Unused Number of Steps 3 b t Sip Concertestion Diafivation sp2 OConcertation Disf haton Smp Concermaton O Diatination DC losd of Sampie Semple Voke 100 si 000000 mi Hotel Recommended Min Working Voha 23 m Brest port imiu 00 S mn Racie Pamese O Tiama AlamctMo OWetch Cema Hep Reas Fig 2 14 Product step selection in the method wizard It is also possible to choose the pre product and post product steps to be included in the method Note the methods are slightly different for hollow fibers cartridges and flat sheet cassettes Pre product steps Post product steps Rinsing Flush Filter CIP Filter CIP Water flush Water flush Water Flux test Water Flux test Buffer conditioning Filter storage solution Table 2 5 Pre product and Post product steps contained in the Method Wizard AKTAcrossflow Method Handbook 11 0012 36 Edition AB 49 AKTAcrossflow system components and software 2 6 5 Process optimization A special option in the Method Wizard is the process optimization method where specific recirculation conditions are selected for scouting a series of TMP values to find the optimal TMP for the application being run This method is automatically linked to the evaluation module where the results can be analyzed and presented 2 6 6 Evaluation module The evaluation m
96. nction Steps Time required Preparing the filter and system for processing Sanitization optional Rinsing Water flush Water flux test Buffer conditioning Up to 120 minutes Product processing Cell processing and washing or Protein concentration and diafiltration Time dependent on surface area applied per feed volume target concentration factor and diafiltration exchange volume Cleaning and flushing the system Buffer flush Cleaning sanitization Water flush Water flux test Storage Up to 120 minutes Table 1 4 Typical times for completing CFF procedures using AKTAcrossflow system 1 6 Water quality requirements To prevent plugging the pores of the membrane filter always use deionized water ultrafiltered water 10 000 NMWC or water for injection when rinsing or flushing when making up cleaning solutions or when adding water for diafiltration of process fluids AKTAcrossflow Method Handbook 11 0012 36 Edition AB 17 Introduction 18 1 7 Membrane filtration devices The membrane inside a membrane filter performs the separation by size sieving the components in the feed stream Hence membrane characteristics represent a key variable in selecting a CFF filter Both flat sheet cassettes and hollow fiber cartridges are available from GE Healthcare Filter Type Selection Open Channel Hollow Fiber Cartridges J 3 Permeate Screen Channel EGN Fla
97. nnel height can affect the fluid dynamics of the system and will directly affect pump requirements and differential pressure of the filtration step Flux Flux represents the volume of solution flowing through a given membrane area during a given time Expressed as LMH liters per square meter per hour Fouling A build up of material on the membrane surface that reduces the filtration rate Unlike the gel layer this material is not redispersed in the bulk stream by higher shear rates Gel layer During the filtration process the thin layer of particles or molecules that may build up at the membrane surface is called the gel layer It is also referred to as the concentration polarization layer High TMP can be the result of an increase in the thickness of the gel layer Gel layer formation can negatively impact the filtration process by reducing flux and inhibiting passage though the membrane Operating at a higher shear rate may reduce the thickness of the gel layer Hold up volume Quantity of fluid remaining within the system after the filtration step is complete Hollow fiber A tube made of membrane When sealed inside a cross flow cartridge the feed stream flows into the inner diameter of one end of the hollow fiber and the retentate the material that does not permeate through the walls of the hollow fiber flows out the other end The material that passes through the membrane walls of the hollow fiber is called
98. o the high viscosity of the feed stream it is critical that all characterization testing be done while operating in total recycle Even a slight increase in the solids concentration will result in a significant increase in the pressure drop Process conditions and monitoring When working with the 750 kD ultrafiltration membrane the circulation flow rate should be kept constant while gradually increasing the TMP to see if there is a proportionate and stable increase in flux If the increased TMP only provides a AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing partial increase in flux the TMP should not be increased as this may result in a drop in protein transmission which will necessitate extensive washing With either the 0 1 or 0 2 um microfiltration membranes the flux rate should be gradually increased while monitoring the TMP As in the earlier example if the TMP increases over time the flux rate should be adjusted to a lower setting until it remains stable The TMP readings with these microfiltration membranes will probably remain below 0 34 bar throughout the process Unless the starting cell density is quite low lt 35 it is unlikely that the process will allow any initial concentration of the feed stock Instead it is best to operate under a constant volume wash mode from the beginning Since as much as 50 of the feed is actually cells the wash volumes double their effectiveness High yields are possible u
99. od Handbook 11 0012 36 Edition AB 41 AKTAcrossflow system components and software 42 2 4 6 TMP control TMP control is usually used in ultrafiltration where the system forces the retentate through the relatively small pores of the membrane The TMP control mode is used at a constant feed flow a constant retentate flow or a constant AP TMP control mode Control element AP Feed pump Permeate R PCV pump TMP control with constant Or gt 0 Offset TMP Feed flowrate TMP control with constant Qr gt O Offset TMP Retentate flowrate TMP control with constant Pe PR gt 0 Offset TMP Table 2 2 TMP Control mode The TMP is mainly controlled by the retentate control valve R PVC In the TMP control mode the software adjusts the retentate valve and permeate pump to maintain a constant TMP In TMP control mode the offset is 0 2 bar as default and is used to avoid low or negative pressure on the permeate side which would affect the permeate pump s function as a flow meter AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software Ss 2 4 7 Flux control mode Flux control is usually used in microfiltration where the system limits the permeate flow through the relatively large pores of the membrane The control mode is used at a constant feed flow a constant retentate flow a constant shear rate or a constant AP In this
100. ods for recovery of the product Recovery e No recovery An option to select if the retentate volume is to be drained manually e Recovery The retentate is first emptied until the reservoir volume is zero Then a maximum of two flushes are performed to flush the retentate side of product AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cross flow filtration process considerations Minimum working volume The minimum working volume represents the amount of feed retentate fluid required to operate the system at the desired cross flow rate without drawing air into the feed pump The minimum working volume is determined by the design of the system retentate tubing volume reservoir bottom design the device retentate hold up volume and the crossflow rate It is important to consider the minimum working volume of a system in the design of a CFF process in particular to confirm that the final target retentate volume is not less than the system s minimum working volume For further details please refer to AKTAcrossflow Instrument Handbook 3 2 3 Denaturation shear temperature and enzymatic action Excessive shear temperature and enzymatic action can denature the product and lower yield Shear The shear sensitivity of a biomolecule generally increases with molecular size Most proteins are relatively resistant to shear denaturation If the shear sensitivity of the protein is not known trials should be carried out to det
101. odule eliminates manual transfer of data to spread sheets The module allows flexible and direct presentation of optimization results with five different evaluation operations elect opersnan Process opimizaton ONomaizes Wate Fh O Disttrston ime optimization Cen e Fig 2 15 Operations available in the evaluation module 50 KTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software aT The operations include process optimization normalized water flux diafiltration time optimization capacity plots and any vs any plots Process optimization is used to analyze process characterization experiments where a series of set points are tested lt Evaluation O Dse a Saou Wether aes iva j Mather poston 0 15 0 15 ber 600o Lee Time 104 06 nin Car Besim T ae DJA BSA CF mimes 20 BSA CF Medina on Fig 2 16 Raw data from process optimization for BSA concentration diafiltration using TMP scouting with 100g I and 20g I BSA solutions The most common experiments are TMP excursions at different retentate flow rates and protein concentrations Process optimization makes a new plot from user identified points along original data curves for example flux vs TMP Process optimization also allows multiple plots to be overlaid at different retentate flow rates or protein concentrations This capability can be used for any process p
102. on factor Alternatively flux can be increased during diafiltration to shorten processing time In both cases crossing the optimum process point results in decreasing returns For example if the protein solution becomes too concentrated prior to diafiltration the flux decreases offsetting the benefits of reduced buffer consumption AKTAcrossflow Method Handbook 11 0012 36 Edition AB 95 Concentration and Diafiltration 96 5 3 2 Discontinuous diafiltration Discontinuous diafiltration is a process where the protein solution is repeatedly concentrated and re diluted It is less efficient than continuous diafiltration because of the extra volume of buffer required 5 3 3 Sequential diafiltration Occasionally it may not be possible to exchange the existing buffer with a new buffer directly without damaging effects to the target protein In sequential diafiltration several buffer formulations moving from weaker to stronger chemical solutions are introduced sequentially to the product to achieve the final buffer exchange 5 4 Membrane and cassette selection 5 4 1 Membrane selection A membrane with too large an NMWC will allow the molecule of interest to pass through and significantly reduce yields Conversely a membrane with too small an NMWC will reduce flux and lead to slower processing times oversized systems increased capital cost and plant space requirements working volume and hold up volume For typical membrane based cross
103. on microfiltration pore size about 10x larger than the target protein e Mammalian cell clarification 0 2 to 0 65 um pore size microfiltration Membrane protein binding The level of protein binding depends upon the membrane material and the protein characteristics and increases with increasing hydrophobicity Table 3 1 Normally in terms of yield protein binding remains insignificant at the laboratory scale but for tight ultrafiltration membranes it can be an indicator of a propensity towards membrane fouling Membrane type BSA yg cm Lysozyme g cm 10 kD select 1 8 4 5 10 kD 1 6 4 2 30 kD 2 4 5 2 50 kD 2 4 Sul 100 kD 9 6 J2 Table 3 1 Typical dynamic protein binding capacities for membranet 1 Data from Validation Guide for Amersham Biosciences Membrane Cassettes document number 18 1171 70 AA published by GE Healthcare The dynamic protein binding test involved installing membrane into a stirred cell pre wetting the installed membrane with buffer and then passing the protein solution through the membrane Following exposure to the protein solution membrane discs were washed three times to remove unbounded proteins Proteins that remained on the membrane were analyzed using a BCA kit Membrane dynamic protein binding capacity is reported in pg cm AKTAcrossflow Method Handbook 11 0012 36 Edition AB 57 Cross flow filtration process considerations Cartridge selection Cartridg
104. on that separates proteins from the cell lysate 12 KTAcrossflow Method Handbook 11 0012 36 Edition AB Introduction Concentration The process of concentrating and buffering proteins in diafiltration preparation for chromatographic processing or for final formulation of an end product Hollow fiber Hollow fiber cartridges consists of bundles of cylindrical cartridge fibers with lumen diameters ranging from 0 25 to 3 mm Feed flows through the lumen under pressure and the permeate passes from the inside to the outside of the hollow fibers Membrane cassette Membrane cassettes consist of layers of flat sheets of membrane sandwiched together often with a spacer between the layers Shear rate The ratio of velocity and flow section expressed in units of sec The shear rate for a hollow fiber cartridge is based on the flow rate through the fiber lumen While excessive shear excessive feed stream flow rate can potentially damage cells and proteins higher shear rates generally result in flux improvements Concentration The concentration factor is the ratio of the initial feed factor volume to retentate volume after separation For example if the initial feed volume is 100 and the final retentate volume is 20 the concentration factor is 5x Diafiltration Diafiltration is a unit operation that incorporates ultrafiltration membranes to remove salts or other microsolutes from a solution Small mo
105. on used for removal of very small solutes lt 1 000 Daltons and salts It uses a semi permeable membrane under high pressure to separate water from ionic materials High pressure is necessary to overcome the natural osmotic pressure created by the concentration gradient across the membrane AKTAcrossflow Method Handbook 11 0012 36 Edition AB 133 Sanitization A cleaning process that destroys most living microorganisms reducing the microbial population to an acceptable predetermined level Separation Dividing a liquid or gas feed stream into separate components Shear rate A ratio of velocity and distance expressed in units of sec The shear rate for a hollow fiber cartridge is based on the flow rate through the fiber lumen Sieving Removal of particles from a feed stream as a result of entrapment within the depth of the membrane pore structure Steam in place SIP The process of sterilizing a device such as a hollow fiber cartridge with steam without removing the device from the separation system Size exclusion Mechanism for removing particles from a feed stream based strictly on the size of the particles Retained particles are held back because they are larger than the pore opening Solute An ionic or organic compound dissolved in a solvent Starling flow A portion of permeate that is driven back through the membrane in the reverse direction near the outlet o
106. or example with proteins of high value a membrane pore size of 0 2 um may be the most appropriate choice Determination of the maximum feed volume for a given process time Another critical parameter for running a successful microfiltration process is finding the right feed load to membrane area ratio for improved process economy The calculation of maximum feed volume is based on the selected maximum flux the desired process time and in the case of the example above the 3x concentration and three wash steps Max Feed Volume Qmax X tx m x Vf Vp Where Qmax Maximum flux rate Imh e t Target process time for scale up h user defined e m Surface area of cartridge used for optimization e V Feed volume L e Vp Full process accumulated permeate volume L Conclusions e Perform process verification at calculated load capacity to evaluate recovery and TMP stability e Decreasing the feed volume from what was used for optimization experiments will shorten the process time with no impact on recovery e Increasing the feed volume may impact recovery and TMP stability and thus must be evaluated AKTAcrossflow Method Handbook 11 0012 36 Edition AB Applications Ce See 6 3 Optimization of a concentration diafiltration process for a BSA solution 6 3 1 Objective The objective of this application was to optimize the conditions for a process which includes 5x concentration followed by five diafiltrations while targeting
107. orbance and provides high performance detection for the wavelengths 214 254 and 280 nm The UV cell housing is made of PEEK and other wetted parts are made of glass and titanium The UV cell is used for measuring the UV absorbance of the permeate This information is used to ensure protein rejection during ultrafiltration diafiltration and also to monitor applications in cell processing Conductivity measurement The conductivity cell is positioned after the permeate pressure sensor in the permeate line The conductivity cell is useful for measuring for example buffer exchange during diafiltration The cell also contains a temperature sensor Temperature variations influence the conductivity and in some applications where precise conductivity values are required it is possible to program a temperature compensation factor that recalculates the conductivity relative to a set reference temperature The measurement range of the conductivity cell is 1 uS cm to 250 mS cm a AKTAcrossflow Method Handbook 11 0012 36 Edition AB 37 AKTAcrossflow system components and software E 2 3 4 Pressure sensors KTAcrossflow system is equipped with four pressure sensors Pressure sensor Location Pf Close to the CFF filter in the feed line to measure the feed pressure Pr Close to the CFF filter in the retentate line to measure the retentate pressure Pp Close to the CFF filter in the permeate line to measure the permeate pressure
108. otein solution are washed from the retentate to the permeate ConeDialgG004 Product UV ConcDialgGO04 Product Cond ConeDialgG004 Product ConcFactor ConcDialgG004 Product pH ConcDialgG004 Product DeltaP ConcDialaG004 Product DF X Fact ConeDialgG004 Product TMP ConcDialaG004 Product Logbook foo 4 ps0 4 poo po Fill Sample po 0 40 0 30 40 50 60 70 80 90 100 110 120 150 Fig 6 10 Complete concentration process of IgG4 with optimized parameters Va Pt min 6 4 5 Analysis of IgG samples using a Hi Trap Protein A column A Hi Trap Protein A column was used to determine the relative IgG4 concentration of the samples during the process and the results are shown in Fig 6 11 The chromatography runs are very reproducible for both the starting material and the concentrated IgG4 taken after diafiltration The retention times were identical and the peak heights of the start sample and the concentrated 120 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Applications Ce sample were the same after the concentrated sample had been diluted 40 times Gel electrophoresis confirmed there was no IgG4 in the permeate PnAMStart2001 10 UV2_ 280nm PnAMStart2001 10 Cone PnAMStart2001 10 pH PnAMStart2001 10_UV2_280nm 02 BASEM nAU 1 a ae 000 4 I 250
109. ow rates help sweep away the debris that forms on the surface of the filter Cross flow rate is most often measured at the retentate outlet Flux Flux represents the volume of solution flowing through a given membrane area during a given time and commonly expressed as LMH liters per square meter of membrane per hour Flux is a key process criterion directly affecting production rate and determining filter performance AP The pressure differential between the feed and retentate lines The differential pressure equals the feed pressure minus the retentate pressure Transmembrane The pressure that drives components of the feed pressure TMP solution through the membrane As a key process variable TMP can help drive the process or if not controlled properly blind the filter resulting in low uncontrolled flux rates TMP is calculated as feed pressure retentate pressure 2 permeate pressure Cell processing The broad term used to describe the processes of cell harvesting cell clarification and lysate clarification Also called upstream processing Cell harvesting A cell processing application that separates cells from fermentation broth with the goal of recovering the cells C oO ll clarification A cell processing application that separates cells from the fermentation broth with the goal of recovering the broth and a protein s in the broth Lysate clarification A cell processing applicati
110. ration of particles With the high particle load typical of cell harvesting low to moderate transmembrane pressures should be used lt 1 bar 15 psi 4 5 4 Process temperature Room temperature is recommended but only if process components are stable at this temperature otherwise operate at 4 to 12 C but with lower flux 4 6 Cell harvesting conditions When working with cells which may still be partially active rapid methods may be important This can be achieved by decreasing the process volume to oO 80 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing membrane area ratio Using more membrane not only allows higher permeate flow rate but it disperses any fouling agent over a broader membrane area This results in a thinner fouling layer and consequently a higher average flux rate 4 7 Cell clarification Cell clarification refers to the separation of a target molecule from a cell culture The cells are filtered and remain in the feed retentate loop The permeate contains the protein or molecule of interest Fig 4 5 Separating a protein from a cell culture is similar to cell harvesting except the product of interest is the protein inthe permeate An effective cell clarification process enables the passage of the greatest amount of target molecules To promote target molecule transmission awash step is often added to the cell clarification process to help flush the target molecules through the membrane Fig 4 6 T
111. re is a smaller retentate volume to exchange the diafiltration was performed at this concentration In some processes it saves time to run the diafiltration step at a lower concentration and then concentrate to the desired retentate volume but this was not the case here ConcFactor Flux vs ConcFactor BSA 5x concentration peer T DIAL DD CPO wO N a f ConcFactor Fl 1 0 1 5 2 0 2 5 3 0 3 5 4 0 4 5 5 0 ConcFactor Fig 6 6 Diafiltration time optimization results AKTAcrossflow Method Handbook 11 0012 36 Edition AB 115 6 Applications 116 6 3 6 Conclusion The optimal process in this ultrafiltration application was defined by running TMP excursions and a concentration run with the optimal crossflow and TMP settings for the diafiltration time optimization It was simple hands free and the results were created automatically from the raw data by the software The optimized process can be summarized as follows e BSA starting concentration 30 mg ml e Process Crossflow rate 42 ml min e Process TMP 1 8 bar e Start diafiltration at five times concentration e BSA final concentration 150 mg ml The complete BSA process was run to verify the optimized parameters and some of the resulting curves are shown in Fig 6 7 ConcDiaBSA001 Product_Cond ConcDiaBSA001 Product DeltaP ConcDiaBSA001 Product PermPress Conc DiaBSA001 Product_pH ConcDiaBSA00
112. re supplies each hollow fiber cartridge with a Certificate of Test stating the model number batch number and test results from quality assurance testing Each cartridge is individually tested at the factory for fiber and cartridge integrity All cartridges and cartridge components meet the specifications of the following tests e USP Class VI Pastics 70 C e Hemolysis Rabbit Blood Direct Contact ISO 10993 e L929 MEM Elution Test ISO 10993 1 13 2 Cassettes GE Healthcare supplies each cassette with a Certificate of compliance The certificate of compliance includes the lot and serial number and states the cassette has been manufactured and tested in accordance with standard operating procedures and is certified to meet the specifications established by GE Healthcare AKTAcrossflow Method Handbook 11 0012 36 Edition AB 25 Introduction __ee All cassettes and cassette components meet the specifications of the following tests e USP Class VI Plastics 70 C e Hemolysis Rabbit Blood Direct Contact ISO 10993 e L929 MEM Elution Test ISO 10993 See 26 AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software 2 AKTAcrossflow system components and software 2 1 System overview Transfer pressure Permeate Transfer pump P 982 sensor P7 pump P 982 Power Permeate Buffer bag module A Manifold module B indicator valve block holder pH electrode Valve P PCV
113. ree structure which provides high tensile strength high temperature resistance and stable performance throughout the service life Fig 1 4 The membrane structure includes a skin layer and a supporting substructure 3 um skin layer 100 um substructure Fig 1 4 Scanning electron micrograph showing the structure of a hollow fibre ultrafiltration membrane from GE Healthcare 1 8 2 Microfiltration membrane Microfiltration hollow fiber membranes from GE Healthcare have a uniform microporous sponge like structure AKTAcrossflow Method Handbook 11 0012 36 Edition AB Introduction 1 9 Membrane filter design 1 9 1 Kvick Start cassettes Kvick Start cassettes include a feed port a retentate port and two permeate ports The ports use 5 16 24 UNF female fittings for connections Fig 1 5 Adaptors enable the connection of the ports to male Luer Lok fittings on laboratory tubing if desired Tubing from AKTAcrossflow eS Permeate 2 Vent R etentate iy 5 1 2 fitting Permeate 1 Drain Feed Fig 1 5 Kvick Start cassette 1 9 2 Hollow fiber cartridges AKTAcrossflow hollow fiber cartridges include a feed port a retentate port and two permeate ports The ports on the Start AXM and Start AXH cartridges use 5 16 24 UNF female fittings for quick and easy connection to AKTAcrossflow system tubing Fig 1 6 AKTAcrossflow Method Handbook 11 0012 36 Edition AB 21 Introduction
114. rget protein from a cell culture and B separating a target protein from a lysate E 82 KTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing 1 Start of clarification Proteins of interest h and other components in broth or lysate Cartridge filter Permeate collection vessel Cell or cell debris Protein of interest 3 Start of washing step Feed reservoir contains mostly cells and cell debris and some protein of interest Cartridge filter 2 End of first stage of clarification Feed reservoir contains mostly cells or cell debris and some protein of interest Cartridge filter Permeate contains protein of interest and broth or clarified lysate ea ee ee on 4 End of washing step Feed reservoir contains mostly cells or cell debris and buffer Cartridge filter Permeate contains protein of interest and buffer Fig 4 6 Operating principles of the clarification process AKTAcrossflow Method Handbook 11 0012 36 Edition AB 83 Cell Processing 84 Successful clarification of feed streams to recover target proteins requires knowledge of the starting product and the finished product specifications such as e Target molecule molecular weight morphology and robustness e Starting volume and concentration of the target protein e Desired finished protein concentration and volume e Desired yield and quality acti
115. riction on the permeate flow are crucial parameters to processing such high density cell suspensions without immediate membrane fouling and clogging of the filter To find the optimal parameters CFF experiments were carried out on an AKTAcrossflow system using 50 cm Start AXM hollow fiber membrane cartridges 6 1 2 Process Optimization The strategy for process optimization was as follows e Membrane selection e Optimization of shear flux settings taking into consideration shear limiting factors such as protein sensitivity to shear damage and feed pressure e Find the number of retentate washes needed for improved protein recovery In microfiltration applications there is no equivalent to TMP excursions so full process runs have to be completed under different experimental conditions and evaluated by analyzing the protein recovery However with the Method Wizard a method can be generated in minutes AKTAcrossflow Method Handbook 11 0012 36 Edition AB 103 6 Applications 104 6 1 3 Membrane selection The filter pore size chosen should be the one that retains the solids while yielding the highest recovery of the target protein in the shortest process time In the following example four membrane pore sizes were screened 750 kDa 0 1 um 0 2 um and 0 45 um Different shear rates were also tested for each membrane pore size to find the combination that gave the highest recovery The 750 kDa membrane gave a very low recovery in initia
116. rocess time but that has to be weighed against the risk of shear damage to a protein of such high molecular weight 400 kD and that no concentration was possible A concentration would decrease the washing time and buffer consumption Therefore results showed that the optimal parameters were to concentrate 1 4 x and run at a shear rate of 4000 1 s and a flux of 35 LMH Shear rate 1 s Flux LMH No concentration 1 4x concentration 4000 45 35 6000 55 Table 6 2 Shear flux optimization results 6 1 5 Optimization of retentate wash for protein recovery The final step in the process optimization was to evaluate the effect of washing on product recovery Fig 6 1 As already described due to the high cell density of the sample it was not possible to achieve a high concentration of the starting material and the process would therefore largely consist of washing the retentate with buffer to promote protein passage AKTAcrossflow Method Handbook 11 0012 36 Edition AB 105 6 Applications 120 100 80 60 m Concentrated g gt m Not concentrated 40 20 0 gt 5 D X se amp o s amp amp amp K A x RX RY s S O Q Q ee Fig 6 1 Evaluation of ion effect of washing on product recovery Using the process parameters established earlier the protein recovery was measured after each of four washing steps for both concentrated and unconcentrated samples As seen in Fig 6 1 the fourth wash di
117. sasdsccsastesebeaconstadsasiaedatsedas 38 2 3 6 Air sensor 24 UNICORN ana eresse Errera Er rear E TERE EEE 2 4 1 Liquid chromatography System version ssssssssceseeeesssssssssssseceeeseeeeeseeeee 39 2 4 2 SoftWare Mod le Ssss annis 39 2 4 3 COMMON INCCLFOCE uniniisnneinnin nin RRR RE a 40 2 4 4 Special features 2 4 5 COMET OL MOI Sessione wantin tau EOI 2 4 6 TMP CONTO seseina nnee i a 2 4 7 Flux control Modesa E needa 43 2 5 Programming a UNICORN Method ou eesssssssesseesssssssscccsssssssseueessesecesee 44 2 5 1 Blocks 2 5 2 BB OSC EE E ESAE EINE EENE 2 5 3 CONS annan EEO ROE AE TEEKS 2 5 4 Watch and Hold Util cccccccceeescsseesssssssseessesssseesscsssssessssssseeessssssseeessssseeseessssees 45 2 5 5 BIO GK DOING een iepak e E EN EE UREAN 45 2 5 6 Run Set up 26 WOTKIOWN merran teaa Ea i 47 2 6 1 Cr ating a new MEtHO cecccscssssssssssssessssssssssssssssssessssssssesecseeseeseseeeseesssssssssssts 47 2 6 2 Method WIZOrO gaiiean ataa a ini 47 2 6 3 Cho singa filter tUpessssdiideniiiauai nann 48 2 6 4 Creating a method 2 6 5 Process OPUMUZOT OM scscscsasssssccsesosceissiiaavascceccnassacascessestcoasostesestesedbosisanboeisebivdies 50 2 6 6 EVICTION PROG Cio iees ts dasssscdstestassssasesshetstestdsbnaaagdasiabdebeh aan 50 2 7 Comprehensive report generation u eeesssssssessessessccssscsssssssseseessesecesse 53 28 SO CUPMUM aE aere EEE E EAE 53 Cross flow filtration process considerations 3 1 Factors influencing produ
118. se selections without a significant contribution from a secondary rejection layer uniform operating conditions with a relatively high shear rate in the circulation flow and low TMP should be used Initial testing should use the 1 mm inside diameter fiber design with a 30 cm path length When the membrane and operating conditions have been chosen additional testing using the 60 cm path length may be used to determine if this design is suitable for scale up Several full scale production processes have successfully utilized the 110 cm ultrafiltration cartridge design with E coli fermentation processes Process conditions and monitoring Unlike fragile mammalian cells bacterial cells can withstand significant shear forces without damage Test results have shown that high circulation flows with 12 000 to 16 000 sec shear rates provide better transmission of the target protein and more stable flux rates Insufficient shear force or excessive TMP will cause the formation of a secondary rejection layer on the membrane surface that will prevent the passage of the target protein Therefore initial testing will require a reliable assay to establish a stable process with good yields With feed streams containing a high cell mass and a large target molecule testing should begin with the permeate flow control set as low as 10 to 20 LMH When working with ultrafiltration membranes the TMP should be gradually increased to see if there is a proportionate and
119. shing promote product passage 2 Washing promote product passage 3 Recover product retentate 3 Recover product retentate 3 Recover product permeate 3 Recover product permeate Table 2 4 There are Method Wizards for all typical product processing operations AKTAcrossflow Method Handbook 11 0012 36 Edition AB 47 AKTAcrossflow system components and software 2 6 3 Choosing a filter type Before starting the method the filter type is defined in UNICORN manager which determines the choice of cell processing or ultrafiltration in the method wizard To get the Method Wizard for hollow fibers Hollow Fiber must be selected as the filter component in the System Setup Component dialog To get the Method Wizard for flat sheets Flat sheet is selected Component Select components B Filter Type Hollow Fibre v Flat Sheet Auxiliary Equipment Fig 2 13 During system preparation the filter type is chosen in UNICORN manager 48 AKTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software 2 6 4 Creating a method Creating a method is easy and straightforward by choosing the method from a number of pre defined basic settings For product steps up to three concentration or diafiltration wash steps can be selected It is also possible to run in fed batch or tank batch mode depending o
120. sing only 2 5 wash volumes Moreover because the membrane is operating under equilibrium conditions flow rates and pressure readings should remain constant Initial testing should be directed at selecting the membrane with the best passage of the target protein This is usually the membrane that also provides the highest flux rate Optimization of the operating conditions will involve minor adjustments to the pressure readings and or flux rates Capacity will be a function of the protein transmission as this will lead to the determination of the required wash volumes Effective processes will be between 2 5 and 5X wash volumes Flux rates can range from 15 to 60 LMH When working with high cell density feed streams it is possible to achieve throughputs of 80 L m based on the starting material Using a 50 cm hollow fiber cartridges and a target 50 LMH flux rate 400 ml of feed material with a 2 5X wash objective will require a process time of four hours for a complete test AKTAcrossflow Method Handbook 11 0012 36 Edition AB 91 Cell Processing 92 AKTAcrossflow Method Handbook 11 0012 36 Edition AB 5 Concentration and Diafiltration Concentration and Diafiltration 5 1 Introduction CFF is used extensively in conjunction with chromatography to concentrate and purify proteins for use in pharmaceuticals The multi step process of CFF concentration diafiltration and chromatographic purification of proteins is referred to as downstrea
121. sired parameters and the software creates the method to perform the experiment The standard method may be used without any changes or may be easily modified to meet most processing needs Fig 5 4 is an example of one of the result files of a TMP excursion displayed in the UNICORN Evaluation window for Result files 98 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Concentration and Diafiltration CE Define plot data File c UNICORN SLocal Fil D efault Result temp Result files ProcessO ptimization Processoptimization002 res Marker position 0 55 0 53 bar 50 4 48 4 LMH Time 2 57 min Delete point Replace point 0 52 48 c optimization002 res Add point from all chromatograms Rename point list Export to Excel Fig 5 4 Result file of a typical TMP excursion Fig 5 5 shows the results of CF TMP optimization for a solution of BSA at a concentration of 30g L original concentration and 150g L target concentration In this example 6 TMP points are measured at 3 CF rates in total recycle mode The data were analyzed by the Evaluation module in AKTAcrossflow software and plotted as TMP excursion curves In this example it can be seen that the excursions done at low protein concentration indicate that all TMP setpoints operate in the pressure dependent region At high protein concentration the pressure independent region of operation is beginning to be seen and increasing the crossflow results in a signifi
122. solution while retaining larger molecules in the retentate Microsolutes are generally so easily washed through the membrane that for a fully permeated species about three volumes of diafiltration solution will eliminate 95 99 of the microsolute Diafiltration exchange factor Diafiltration exchange factor Diafiltration buffer volume Sample volume Dialysis Removal of small molecules from a solution of macromolecules by allowing them to diffuse through a semi permeable membrane into water or a buffer solution This osmotic pressure separations method is controlled by the concentration gradient of salts across the membrane Differential pressure See AP AKTAcrossflow Method Handbook 11 0012 36 Edition AB Diffusion Movement of liquid or gas particles from a region of higher concentration to a region of lower concentration Direct flow See normal flow filtration filtration Downstream Starting with a feed stream free of cells and cell processing debris the purification sequences involving chromatography and membrane separations to achieve final product purity Effective filtration area Ina membrane separations device the active area of the membrane exposed to flow Endotoxin The outer cell wall of gram negative bacteria also known as LPS lipopolysaccharides and pyrogens Ethylene oxide A sterilization process still common for biomedical EtO sterilization prod
123. ss considerations Successful protein concentration and diafiltration using AKTAcrossflow system relies on specifying the pre and post concentration product as follows A 94 KTAcrossflow Method Handbook 11 0012 36 Edition AB Concentration and Diafiltration e The characteristics of the target protein size and shape solubility shear sensitivity temperature sensitivity ionic condition and the role of the chromatography step e The starting characteristics of the protein solution pH ionic condition solubility compatibility with the chromatographic step and the target characteristics for the next step e The starting volume and concentration of the target protein e The desired finished concentration and volume e Process time and cost e Diafiltration volumes e The desired yield and quality stability of the target protein 5 3 Diafiltration The goal of diafiltration is buffer exchange to optimize the next chromatographic purification step or for final formulation of the protein solution From an operational perspective the goal is to minimize consumption of the diafiltration buffer and to shorten diafiltration processing time The diafiltration process usually consists of concentration followed by continuous diafiltration 5 3 1 Efficiency To increase process efficiency the product concentration can be increased before diafiltration thereby reducing the amount of buffer required to achieve a specific diafiltrati
124. sssessseeeessessssssttnsnesnsnssnsseeseeeeeeeeeseeseettesteeeeteet 6 3 4 Optimization of critical process parameters eeeeesssssssssssssseeeeeeeeseeseetee 113 6 3 5 Diafiltration time OPtiMiZAtiON eesssssesssssessssssssssssscseesesseceeeeeeecssnsssssssssee 115 6 3 6 CONCUSSION inicas terete rte rater nee Street aren ete 116 6 4 Concentration of cell culture supernatant containing GGA sesiuni iiaii 117 6 4 1 Bloa l EEEE NEEE 117 6 4 2 Process Optimizat Meisiinennenniiiiin nin 117 6 4 3 Optimization of critical process parameters v ereeessssssssssseseceeeseeeeseseseee 118 6 4 4 Concentration and diafiltration process 3 6 4 5 Analysis of IgG samples using a Hi Trap Protein A column 120 6 4 6 CONCUSSION iine na ER EEN e ESNE IAS 122 AppendixA Membrane Filters for KTAcrossflow system AL Hollow Fiber COrtrid Qe vscsssccatssssisccssssevassisesosnsststessersevansbecsssvnvsssissansnnsians 123 A 2 Membrane Cassettes for AKTACrOSsflow SUStOM essssccsssesssseeeen 124 Appendix B Glossary of terms B 1 Glossary OF terms wecceccsccsssssssssssssssussuessessssssccssssssssussusssssesscesssesssssssssueeseeseees 125 KTAcrossflow Method Handbook 11 0012 36 Edition AB Contents E Appendix C Shear effects on proteins and cells C 1 Shear studies on protein SOLULIONS ccsseesssesssssccssssssssssseesesseseceseeeeens 137 Cid Piston rinsing system C 1 2 Comparison of different PUMP types vrerereessessssssssseeeseesesesseeenneenenseesee 13
125. sssssisisessnisieiesesrirnisiniinsnisnsisisinirsessiss 24 1 11 2 Kvick Start cassettes Leese 24 1 11 3 Start AXM and Start AXH cartridges 24 1 12 Testing procedures ssus a25 1 12 1 Water flUX teSta ccccesesssssssssssssssssssssssscsssscscceececseesesssssssnssnnsntunusnssesseseeseceeeees 25 1 13 Quality assurance and documentation ou eeceeceeesssssssssssssssssseseeseeeee 25 1 13 1 Hollow fiber cartridgeS 1212122 25 Tlse Cassetes aininn na ee e enicegasstbe i aiBananianee 25 KTAcrossflow system components and software 2l System OVeNiEW ipsas eeri anar ate aeea Ers t Eassa EEi iia 27 2 2 KTAcrossflow system COMPONENTS eesssscssssssssssessssssssseeesesesteeteeeee 29 2 2 1 PUMP Sienno eA E E EEEN EET 29 2 2 2 Pump headS 30 2 2 3 Piston rinsing system 31 2 2 4 RESEIVOII cssssssssessssssseseees asl 225 Lig id COMO cN Sacau an a Aa EE 33 KTAcrossflow Method Handbook 11 0012 36 Edition AB iii Contents 2 2 6 M gn tie stir OG cis ctia isssctactecsinGticcesctchcosdtsisbdbinuadtndedishicsielitiieectbitee 33 2 2 7 MAGE IONS eaaa R ASE 34 2 3 2 5 1 2 3 2 Flow restrictor in the transfer line sssssssssssessssssssssssssscssssecccsceescessesssssseees 37 2 3 3 Dete t rsand MOnt S aiscin 37 2 3 4 Pressure Sens OFS ccssssctcescsesspcsnintcnsazedersetdoisiehsatesetebcnanstspsebes E EnEn 38 2 3 5 RESCFVOIF level Sens Or sssscsescsascsssziusescetasatessutasdisveschv
126. stable increase in flux With microfiltration membranes the flux rate should be gradually increased while monitoring the TMP If the TMP increases over time the flux rate should be adjusted to a lower setting until it remains stable Even with a relatively high starting cell mass it is often possible to perform a 5X concentration without sustaining a significant increase in the pressure drop along the cartridge If the inlet pressure begins to AKTAcrossflow Method Handbook 11 0012 36 Edition AB 89 Cell Processing 90 rise abruptly it is normally not due to cartridge plugging but to the increased viscosity of the feed and additional concentration is not advised With initial testing a 2X increase in the pressure drop should be used as an upper limit The constant volume wash should be initiated without interrupting the circulation flow If the flux rate has decreased by more than 4x it is advised to temporarily open the back pressure valve and shut off the permeate flow This technique may help to diminish the effects of the secondary rejection layer Due to the relative difficulty of working with bacterial cell suspensions testing should emphasize selecting the membrane that provides the best long term transmission of the target protein coupled to a process sequence that maintains the process in an equilibrium status With regard to capacity testing should start at 25 L m Using a 5X concentration followed by a 5X wash with a 50 cm
127. t 1 Out 2 Out 3 28 KTAcrossflow Method Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software 2 2 AKTAcrossflow system components 2 2 1 Pumps The pumps of AKTAcrossflow system are high precision metering pumps The pump heads have a sanitary design with a self contained rinsing system to prevent contamination and pump damage For more information regarding shear effects on cells and proteins see Appendix C The low shear design ensures that sensitive cellular material is not damaged during operation Furthermore this design guarantees negligible heat transfer from the pump heads to the process fluid Pressure sensor Pr Fig 2 3 Feed pump P 984 Pump P 982 and Pump P 984 Pump P 982 and P 984 are high performance laboratory pumps for use in applications where accurately controlled liquid flow is required Twin reciprocating pump heads work in unison to deliver a smooth and pulsation free flow P 982 is used as the transfer pump module A and as the permeate pump module B P 984 is used as the feed pump module A and B Pump P 982 features e Four pump heads arranged in two pairs of two e Pressure range 0 520 kPa 5 2 bar 75 4 psi e Flow rate range 0 1 200 ml min AKTAcrossflow Method Handbook 11 0012 36 Edition AB 29 AKTAcrossflow system components and software Pump P 984 features e 4pump heads e Pressure range 0 520 kPa 5 2 bar 75 4 psi e Flow rate range 1 600 ml min 2 2 2
128. t Sheet Cassettes AS SS SS wg Permeate Fig 1 3 Mechanism of action for hollow fiber cartridges and flat sheet cassettes Filter characteristics that influence performance include the following e Fiber length and lumen diameter hollow fiber cartridges e Flow path length and channel height cassettes e Membrane surface area e Pore size e Material in membrane 1 7 1 Fiber length and lumen diameter cartridges Fiber length and lumen diameter in Start AXM and Start AXH cartridges are controllable variables that influence a CFF process Available fiber lengths are 30 cm and 60 cm Fiber lumen diameter for these cartridges range from 0 5 mm to Imm AKTAcrossflow Method Handbook 11 0012 36 Edition AB Introduction 1 7 2 Flow path length and channel height cassettes Kvick Start filters available for AKTAcrossflow system use two layers of flat sheet membrane separated by a screen The flow path length is 17 cm The channel height equals the thickness of the screen 1 7 3 Membrane surface area AKTAcrossflow system works best with ultrafiltration cassettes having 50 cm to 150 cm of membrane surface area and hollow fiber microfiltration cartridges having 50 cm of membrane surface area When choosing the membrane surface area consideration must be given to the starting volume of the product the nature of the product the desired processing time and operating pressures Typical values for the selection of the membrane s
129. test cartridge and 150ml of starting material the process time is expected to be 2 to 3 hours 4 12 3 Yeast Pichia pastoris and other types of yeast have been extremely popular for expressing target proteins Often a successful fermentation will result in a highly viscous material with as much as 50 percent cell mass This represents a challenge for any of the candidate clarification technologies However the target proteins are usually small Moreover the hollow fiber technology is linearly scalable and does not require pre dilution in order to provide good yields Membrane and cartridge selection The most popular membranes to ensure good passage of target proteins as large as 70 kD molecular weight have been the 0 1 micron microfiltration and 750 kD ultrafiltration membranes Larger proteins have been successfully processed at full scale with the 0 2 micron microfiltration membranes With such a high initial viscosity there is even greater need to use the most dense membrane that will effectively pass the target protein More open membranes working with high inlet pressures will result in cells being trapped on the membrane surface near the cartridge inlet Inlet pressures will rise and the process will fail The 1 mm fibers with the 30 cm path length are used exclusively with viscous yeast feed streams Moreover the circulation flow rate rarely exceeds 4 000 to 6 000 sec in order to keep the inlet pressure readings less than 0 7 bar Due t
130. the permeate Housing The mechanical structure that surrounds and supports the membrane or filter element The housing normally has feed retentate and permeate ports that direct the flow of process fluids into and out of the filter assembly 130 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Hydrophilic Filters that wet easily with water and work well with aqueous solutions Hydrophobic Filters that do not wet easily with water but typically do wet easily with nonpolar solvents such as alcohol Once wetted a hydrophobic filter may be used to process many aqueous solutions In vitro An experiment performed in a test tube Petri dish or other lab apparatus with parts of a living organism such as testing a drug with tissue samples From Latin meaning in glass In vivo An experiment performed using a living organism From Latin meaning in live subjects Inlet pressure The pressure of a fluid at the feed port of a separation device Isoelectric point The pH at which a protein carries no net electric charge Lumen The inner open space of a hollow fiber Macrovoid A generally undesirable open space in the substructure of a membrane filter that is appreciably larger than the average pore size Macrovoids can lead to pinhole defects resulting in unwanted passage that directly affects final product yield Macrovoids can also affect the overall membrane str
131. thod Handbook 11 0012 36 Edition AB AKTAcrossflow system components and software 2 3 2 Flow restrictor in the transfer line A flow restrictor is positioned downstream of the transfer pump in order to ensure a proper operation of the check valves at the pump heads and thus accuracy in the transfer flow rate The restrictor generates a constant back pressure of 3 bar 2 3 3 Detectors and monitors AKTAcrossflow is equipped with detectors for continuous in line measurement of pressure temperature pH conductivity and UV absorbance for accurate and reliable monitoring The flow cells for UV conductivity and pH in the permeate line are situated close together to minimize volume and time delay between components The flow cells are easily accessible from the front panel to facilitate maintenance pH measurement The pH electrode is positioned downstream of the pressure control valve P PCV The pH electrode is optimized for continuous pH measurement in AKTAcrossflow path The electrode is of the sealed combination double junction type with a glass tip and the cell holder is made of titanium The pH monitor provides pH measurement in the range 1 14 2 12 within specification and can be used for example to monitor buffer exchange during diafiltration UV measurement The UV cell is normally positioned after the conductivity cell in the permeate line but it can be moved to the retentate side if required It is designed for continuous measurement of UV abs
132. time or rate Use cleaning solution better able to solubilize contaminants Table 3 5 CFF troubleshooting chart AKTAcrossflow Method Handbook 11 0012 36 Edition AB 69 Cross flow filtration process considerations 70 AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing 4 Cell Processing 4 1 Cell harvesting Cell harvesting is the process of separating cells from the fermentation broth in which the cells grow After harvesting the cells are mechanically disrupted and the protein of interest is separated from the cell debris by clarifying the lysate Fig 4 1 Hollow fiber microfiltration or higher NMWC ultrafiltration cartridges may be used effectively for cell harvesting 4 2 Cell harvesting process The concentration factor that can be achieved is based on the starting concentration which can in the case of yeast cells be as high as 70 to 80 percent by cell weight Typical concentration factors are e E coli cells 5x concentration e Yeast cells 2x concentration e Mammalian cells 10x to 20x concentration 4 2 1 Washing step Cell harvesting usually includes a cell washing step to promote the transmission of broth components through the filter Fig 4 2 After washing the ideal end product would consist of the concentrated cells suspended in the buffer used to wash the cells However in practice the harvested cells and buffer can contain varying levels of unwanted elements such as precipit
133. tion Cross flow rate TMP set points bar mg ml ml min in 0 5 bar increments 30 and 150 42 0 8 2 8 33 0 8 2 8 25 0 8 2 8 Table 6 6 TMP excursion plan for the BSA runs The Method Wizard in AKTAcrossflow software was used to create a method for running the TMP excursions The Flux vs TMP curves were generated from the collected data using the Evaluation Wizard and the results are shown in Fig 6 5 Runs done at 30 mg ml BSA indicate that the TMP setpoints operate in the pressure dependent region until the TMP becomes higher than 2 3 bar where the pressure independent region of the operation starts As anticipated higher crossflow rates result in higher flux because of the more efficient reduction of the gel layer at the membrane surface At 150 mg ml BSA the pressure independent region of operation is more apparent The optimal crossflow rate is that which gives the highest flux shortest process time However this must be balanced against the sensitivity of the protein to shear a higher crossflow rate results in the protein passing through the pump more times and the disadvantage of requiring high pump capacity if scaled up In this case a crossflow rate of 42 ml min was chosen Flux vs TMP Flux vs TMP BSA30 g L BSA 150 g L re ma m re m m CF 42 mL min CF 33 mL min CF 25 ml min CF 42 mUmin CF 33 mL min CF 25 mLimin 120 s x Pe ze fe H d 5 2 p ig pa
134. tten until the block is finished or the End_Block instruction is executed There are two types of calls e Unconditional calls which are made with a Block instruction e Conditional calls which are made with a Watch instruction This makes it possible to call a specified block or instruction when a particular monitor signal meets a given condition 2 5 4 Watch and Hold_Until The breakpoint when the Watch instruction is issued determines when the Watch begins A Watch remains active until the condition is met or a new Watch instruction is issued for the same monitor The Watch is cancelled automatically when the condition is met A Watch can also be turned off with the WATCH_Off instruction The Hold_Until instruction is a special kind of Watch instruction The method is put on hold until a specific condition is met signal test or value or the time out is reached Thereafter the remaining instructions in the method are executed 2 5 5 Block pane The organization of blocks in the method is shown graphically in the Block Pane of the Method Editor fig Fig 2 11 Each block is represented by a gray bar with the block name and the length of the block The line is shifted down to indicate calls to other blocks CONCENTRATI Flowpath Perm RetFlow Step DelayUpUSmin Hold Until Delt 0 05 min 0 00 min 0 0 05 min Watch_ConcFactor Greater_Tha Watch RetVol Less Than 13 m Hold_Until Trans ol Greater_Than 0 00 ml INFINITE base
135. ucts in which product is subjected to steam and highly toxic ethylene oxide gas Because many materials change properties when exposed to moisture and EtO by products products destined for this process must be specially engineered for EtO sterilization Extractables Substances that may dissolve or leach from a membrane device during filtration and contaminate the process solution For example the these might include wetting agents in the membrane membrane cleaning solutions or substances from the materials used to encase the membrane Feed Material or solution that is introduced into a membrane separation system Feed pressure The pressure measured at the inlet port of a cartridge or cassette Filter area The surface area of filter media inside a separation device Filter efficiency Filter efficiency represents the percentage of a given size particle removed from the fluid by the filter Filtrate Also called permeate The portion of the process fluid that passes through the membrane AKTAcrossflow Method Handbook 11 0012 36 Edition AB 129 Flow path length The total length that a feed solution travels from inlet nominal flow path to outlet Flow path length is an important parameter length to consider when doing any process development system design or scale up or scale down experiments The flow path length and other fluid channel geometries such as lumen diameter or cha
136. urface area are as follows e 30 to 100 liters of feed per square meter of membrane surface area for microfiltration e 100 to 200 liters of feed per square meter of membrane surface area for ultrafiltration 1 7 4 Pore size Membrane pore size determines the size of the particles or molecules that pass through the membrane The pores in a membrane vary in size so the size distribution of the pores determines the sharpness of the separation Ultrafiltration The pore size of ultrafiltration membranes is expressed as nominal molecular weight cutoff NMWC AKTAcrossflow ultrafiltration membrane filters are available in both hollow fiber cartridge and membrane cassette formats Microfiltration The pore size for microfiltration cartridges is expressed in microns AKTAcrossflow cartridges have average pore size ratings from 0 1 um to 0 65 um Guidelines for selecting membrane pore size are found in later sections of this handbook The automation and minimum operating volume of AKTAcrossflow system make it easy to screen different pore sizes to find the best performing membrane for a given application In Appendix A you will find information of all micro filtration cartridges and ultra filtration membrane filters supplied by GE Healthcare AKTAcrossflow Method Handbook 11 0012 36 Edition AB 19 Introduction 20 1 8 Membrane structure 1 8 1 Ultrafiltration membrane Ultrafiltration membrane from GE Healthcare has a macrovoid f
137. vity of the protein e Level of suspended solids As with cell harvesting rapid processing times may reduce the exposure of the target protein to shear forces enzymatic action and temperature increases 4 9 Membrane and cartridge selection In the case of cell culture and lysate clarification microfiltration membranes will easily retain all cells and cell debris The key to selection is not based on retention but passage of the target molecule 4 9 1 Membrane selection When selecting membranes for clarification smaller pore size filters resist fouling more than filters with larger pore sizes such as 0 45 or 0 65 um A general guideline is to select the smallest pore size ratings that is at least 10x larger than the size of the target protein in it largest state or longest dimension When working with lysates which can contain a wide range of particle sizes and many types of proteins and sticky cell components choosing a small pore size can help prevent fouling of the membrane pores Table 4 5 presents typical cartridge and membrane characteristics for common clarification applications 4 9 2 Cartridge selection The presence of particles in the feed stream requires the selection of short path length cassettes 30 to 60 cm with large lumen diameters 0 75 to 1 0 mm See Table 4 5 for cartridge specifications AKTAcrossflow Method Handbook 11 0012 36 Edition AB Cell Processing Lee Type
138. vssrsansasiasrvarsenrndsoveasnisieorvens 98 5 8 1 Concentrato Niarn a Ea i 98 5 8 2 Diafiltration time OPtiMIZAtiON sssssssssssessessssssssssssseseesseceeeeeeessessnsssnssssses 100 5 8 3 DICE CEI OND tacto innerer oni sna ae EREE 101 6 Applications 6 1 Purification of B glucosidase from a Pichia pastoris cell culture broth using microfiltration s ss 1s11ss1ss111s1s110112511s111 103 6 1 1 ODjOCTIVE OEE 103 6 1 2 Process Optimization 103 6 1 3 Membrane Selection 104 6 1 4 Optimization of shear flux settings eesssssssssseeesssesssssessceeeeesesesesstsetttnees 105 6 1 5 Optimization of retentate wash for protein recovery 105 6 1 6 CONGCIUSION ssscccrscecsstttscddSndresennctshsteaneban cusntadeonetncamnianananitentsass 107 6 2 Purification of Green Fluorescent Protein His GFP His from an Escherichia Coli cell homogenate weeecccscssscssssssssessesseesscsee 6 2 1 OBJCCHV Esta tee teste tices etree E EE 6 2 2 Process Op mizA UoN irssi RR 6 2 3 Membrane selectes sieniin eiA N AA 6 2 4 Optimization of shear flux settings 121 11211911115115111111101111112112 6 2 5 Optimization of retentate wash for protein recovery si 6 2 6 Conclusio Maseni na nn in ni 6 3 Optimization of a concentration diafiltration process for a BSA solution weeecsccssscssssssssssseesesssssescessssssssssueeeessssesesssess 113 6 3 1 DEJ E EA E E E T EA 6 3 2 Process Optimization 6 3 3 Membrane SCICCTION eeesesesssssss
139. w and TMP At the optimum combination of the highest cross flow and TMP just before gel layer accumulation the highest flux rate will be achieved Fig 3 3 Optimal operating conditions Permeate flux operating region Gel layer control TMP Fig 3 3 Relationship of TMP and flux at five cross flow rates 66 KTAcrossflow Method Handbook 11 0012 36 Edition AB Cross flow filtration process considerations es 3 7 Scaling up parameters The ability to scale a process from the laboratory to manufacturing is a key factor in process development Normally the scale up sequence is completed in multiple steps lab scale to pilot scale and pilot scale to production scale Reasonable scale up increments are typically 5 to 20 times When scaling up the following parameters should be kept constant e Ratio of filter area to feed volume e Membrane e Screen type if applicable e Fiber or cassette path length e Channel height cassettes or lumen size hollow fiber cartridges e Cross flow rate e TMP e Temperature e Feed concentration e Process steps and sequence The scale up pathway from AKTAcrossflow system to the UniFlux system and production systems is summarized in Table 3 3 GE Healthcare CFF systems and devices for scale up Lab scale Pilot scale Production scale AKTAcrossflow system UniFlux system Engineered system Kvick Start cassette Kvick Flow 5 cassette Kvick Flow 25 cassette Kvick
140. x rate as a function of TMP For a given cross flow rate TMP controls flux at the beginning of a run If a gel layer forms increases in TMP will not result in increases in flux Therefore increasing TMP will provide little performance gain The optimal TMP range for efficient and economic operation is just before the gel layer starts influencing the flux Fig 3 2 i Gel layer control region A CF1 Permeate flux Optimal TMP range TMP Fig 3 2 Optimal TMP range under a constant cross flow rate CF1 KTAcrossflow system software includes a method that can perform an ultrafiltration optimization based on default values or values input by the operator KTAcrossflow system runs the optimization experiment and can display the results in table or graphic format Data can be exported to Microsoft Excel for advanced analysis See KTAcrossflow System User Reference Manual for additional information KTAcrossflow Method Handbook 11 0012 36 Edition AB 65 Cross flow filtration process considerations 3 6 2 TMP and crossflow At a given TMP increasing the cross flow rate helps reduces the concentration gradient layer and increases flux Cross flow rates may be increased until process yield or product quality are adversely affected through for example shear effects Optimization of a CFF process such as protein concentration must include an examination of the interaction of the two most important variables cross flo
141. y steps are summarized below e Rinsing the Kvick Start cassette e Rinsing storage solution from a previously used Kvick Start cassette e Measuring the water flux before and after use Determines the cleaning effectiveness and monitors the Kvick Start cassette s performance from run to run e Sanitizing the Kvick Start Cassette Kvick Start cassettes and AKTAcrossflow system can be treated with sanitizing agents such as 0 5N NaOH e Conditioning the system with buffer This exposes the system s wetted parts to an appropriate buffer before introducing the product The buffer can also bring the system to the proper operating temperature before processing begins All these steps are included in the method wizard of AKTAcrossflow system software 5 6 Operating conditions for Kvick Start cassette The following operating conditions are recommended starting points for downstream concentration trials e Typical operating flow rate 25 to 50 ml min cassette e Typical transmembrane pressure 1 7 to 2 4 bar 25 to 35 psi e Maximum operating pressure 4 bar 60 psi at 23 C e Maximum operating temperature 50 C at 2 bar 30 psi e Recommended pH operating range 1 to 14 e Kvick Start non recoverable drained volume less than 3 ml 5 7 Concentration factor The maximum available concentration factor is limited by the ratio between starting volume and minimum working volume In addition over concentration of protein can lead to ineffici

Download Pdf Manuals

Related Search

Related Contents

Philips LED bulb 8718291753254  テレビジョン受信機 - 省エネルギーセンター  User Manual - data    RFID System V680 Series  StarTech.com Portable Wireless N WiFi Travel Router / Access Point / Repeater - USB Powered  MDO 1.5 Le Fort I    Activateur Aloès    

Copyright © All rights reserved. Failed to retrieve file