CELLine - Cosmo Bio Co Ltdsearch.cosmobio.co.jp/cosmo_search_p/search_gate2/docs/... · 2006. 8. 18. · Obtain 25 x 106 viable cells from a pre-culture in log growth phase and suspend

Disposable Bioreactor for Efficient Protein Expression

CELLine

Minori Ito

Two-Compartment Technology

Efficient cell cultivation relies on maintaining an optimal supply of oxygen and nutrients, as well ason removing inhibiting metabolic waste products efficiently. These factors limit the maximum celldensity obtainable from standard, homogeneous cell culture techniques which are consequentlypoor at achieving high expression levels of proteins.

The Two-Compartment Bioreactor CELLine is designed to overcome such limitations. The innovativeCELLine technology separates the bioreactor into a medium and cell compartment by means of a10 kDa semi-permeable membrane. This membrane allows a continuous diffusion of nutrients intothe cell compartment with a concurrent removal of any inhibitory waste product. The individualaccessibility of the compartments allows to supply cells with fresh medium without mechanicallyinterfering with the culture.Efficient gas transfer is ensured by a silicone membrane which forms the cell compartment base.This membrane provides an optimal oxygen supply and control of carbon dioxide levels by provid-ing a short diffusion pathway to the cell compartment. All together, the technology built into CELLineallows simulating quasi in vivo cultivation conditions leading to high cell densities.

In vivo-type cultivation

Successful cultivation of different cell types demands not only skillful handling, but also dependsto a large extent on the chosen cultivation system. CELLine technology closely mimics the phys-iological conditions within the body enabling production of drastically increased cell densitiesand more organotypic cell morphologies, sometimes even to the extent of 3-dimensional cellgrowth.

1 : 4 1 : 4

0

10

20

30

0 2 10 12 14 16

Proteinmg/ml

0.0

1.0

2.0

3.0

10 6

cells/ml

20 2218

…

1 : 4

time (days)4 6 8

HarvestSplit-BackFresh Medium

HarvestSplit-BackFresh Medium

7 - 10 d 5 - 7 d

ApplicationEfficient protein expression

Cells growing under the optimal conditions created in CELLine reach densities of 107 to 108

cells per ml, a cell concentration that is about two magnitudes higher than the one obtainedwith conventional culture techniques. Consequently, the concentration of expressed protein istypically 50 to 100 times above what is found in standard cell culture disposables. In addition,CELLine has been designed to maintain cells for several months in culture allowing to periodi-cally harvest expressed proteins. By combining high product concentration with recurring prod-uct collection, large amounts of highly concentrated proteins are routinely obtained in CELLine.With classic culture techniques, contaminating proteins originating from serum or cells are asignificant part of the total protein fraction. In contrast, due to the high product concentrationsobtained with CELLine, the relative level of contaminating proteins to the expressed protein ismuch lower. In many cases the quality of monoclonal antibodies produced with CELLine is suf-ficient to perform standard laboratory tests, like Western Blot, without the need of any furtherpurification steps.

Further Reading:Efficient laboratory-scale production of monoclonal antibodies using membrane-based high-density cell culture technology. Trebak et al. (1999) J. Immunol. Methods, 230: 59-70

Long-Term High Level Protein Expression in Adherent, Protein-free Growing BHK Cells Using INTEGRA CELLine adhere 1000 Bioreactor Flasks. J. Mitter-maier and M. O. Zang-Gandor (2004) GeneticEngineering News, 24(12): 42

For more references go to www.integra-biosciences.com or contact us at [email protected].

10 µm 10 µm

Representative EM micrographs showing the flat morphology of HEp-2 cells growing as mono-layer in a standard T-Flask (left) compared to the rounded shape of the same cells when culti-vated in CELLine adhere (right). (With courtesy of W. Pfaller, Institute of Physiology, University of Innsbruck)

The choice between CELLine classic and CELLine adhere makes it possible to grow both sus-pension or anchorage-dependent cells, hence allowing production of different biomoleculesusing different expression systems such as monoclonal antibodies in hybridomas, recombinantproteins in transfected cell lines and virus particles in packaging cells.

Easy and economical protein expressionEasy operation

Due to its uncomplicated design working with the CELLine bioreactor is as simple as withany standard tissue culture flask. For a straightforward control of the growth process,microscopic observation of the cells is made possible by the transparent design of theCELLine classic bioreactor. The system operates independently of any complicated controltechnology and works without any pump systems or agitation devices. Easily and secure-ly stackable, CELLine flasks occupy a minimum of space in any standard CO2 incubator.A specific adaptation of the cell culture techniques or media composition is generally notnecessary when starting to work with CELLine and both, serum-supplemented or serumfree media are suitable.

Cost efficiency

CELLine has been designed to bring substantial cost savings to cell cultivation. Labourcosts are considerably reduced, because fewer disposable flasks need to be handled com-pared to other culture techniques in order to produce milligram amounts of protein. Alsothe expenses for media supplements are significantly reduced, since the addition of serumor other synthetic additives can be limited to the cell compartment. Furthermore, the highproduction yield and quality obtained in the relatively small culture volume of CELLine con-tribute to reduce costs and labour time in the subsequent downstream processing steps ofthe product.

Economic analysis for mAb Production

Units CELLine 1000 T-Flask 225cm2

Productivity

Hybridoma productivity (literature value) pg/h x cell 0.3 0.3

Cell density cell/ml 3x107 1x106

Culture volume ml 20 50

Yield per harvest (7 days) mg 30.24 2.52

mAb concentration mg/ml 1.51 0.05

Production Costs (250 mg mAb) Amount Cost (in $) Amount Cost (in $)

Harvests per disposable 8 1

Number of disposables 1 150 100 300

Medium ($ 20 per liter) litres 8 160 5 100

Serum consumption* ($ 300 per liter) ml 16 4.8 500 150

Labour** ($ 25 per hour) min 120 50 500 208

Total costs 364.8 758

Costs per mg mAb 1.46 3.03

* The medium is supplemented with 10% of serum. In the case of CELLine, only the medium in the cell compartment is supplemented.

** Labour is calculated as the time used for inoculating and harvesting one CELLine flask (15 min)or one T-Flask (5 min) multiplied by the number of harvests or flasks, respectively.

Animal welfare

CELLine is a disposable bioreactor that is competitive in costs and performance to the pro-duction of monoclonal antibodies using mice ascites. As an added benefit, when express-ing monoclonal antibodies in hybridomas using CELLine, the antibody preparation is freeof any contamination from mouse immunoglobulins. Over recent years, CELLine technolo-gy has been successfully adopted worldwide for the production of monoclonal antibodiesand thereby has contributed diminishing the use of laboratory mice.

CELLine Two-Compartment Disposable Bioreactors are manufactured from optically clear virgin polysterene with a gas transfer bottom made of a molded silicone membrane providinga 0.2 µm vent barrier. The compartments are separated by a 10 kDa semi-permeable cellu-lose acetate membrane and individually pressure tested for integrity. The bioreactors are eas-ily stackable owing to specific stabilisation interlocks, packed individually in easy to openmedical-grade blister packaging, sterilised by gamma irradiation and non-pyrogenic.

CELLine CL 350 CELLine CL 1000 CELLine AD 1000

Size: L x W x H (mm) 190 x 95 x 62 275 x 120 x 80 275 x 120 x 80

Weight (g) 185 334 336

Medium compartment cap 28 mm vented (0.2 µm), 38 mm vented (0.2 µm), 38 mm vented (0.2 µm), green polypropylene cap white polypropylene cap black polypropylene cap with polypropylene liner with polypropylene liner with polypropylene liner

Cell compartment cap 24 mm polypropylene cap 28 mm polypropylene cap 28 mm polypropylene cap with polyethylene liner with polyethylene liner with polyethylene liner

Cell compartment inlay none none PET matrix inlay

Microscopic viewing center window requires center window requires limited visibility due to (inverted microscope) objectice working objectice working inlay matrix

distances of 2.5 mm distances of 2.5 mm

Vertical and horizontal 50 - 350 ml 100 - 1000 ml 100 - 1000 mlvolume markings

Product Name Description Quantity/Case Item No.

CELLine CL 350 Disposable Two-Compartment Bioreactor 5 90010for suspension cells,350 ml media volume, 5 ml culture volume

CELLine CL 1000 Disposable Two-Compartment Bioreactor 3 90005for suspension cells,1000 ml media volume, 15 ml culture volume

CELLine AD 1000 Disposable Two-Compartment Bioreactor 3 90025with matrix inlay for anchorage-dependent cells,1000 ml media volume, 15 ml culture volume

Ordering Information

Technical Specifications

CELLine classic (CL) is ideal for laboratory scale applicationsusing suspension cells or adherent cells in combination withmicrocarriers. The unit is optimised for cultivation of hybrido-mas and many other cell types (e.g. CHO, NSO, SF cells).

CELLine adhere (AD) is specifically adapted to allow growth ofanchorage-dependent cells (e.g. HEK, BHK, CHO cells). Thebioreactor contains a woven, polyethylene terephtalate (PET)matrix in the cell compartment providing an ideal surface forcell attachement. PET matrix inlay of

CELLine adhere

CELLineclassic

CELLineadhere

Bioreactor Types

INTEGRA Biosciences AGCH-7000 Chur, SwitzerlandPhone: +41 81 286 95 30Fax: +41 81 286 95 33E-mail : [email protected]

9009

9_V

02_F

lyer

_CEL

Line

ww

w.a

aa-d

esig

n.ch

Having troubles expressing enough recombinant protein?

Efficient protein expression

50-100 times higher product concentrationscompared to classic cell culture disposables

Easy operation

as simple as using a tissue culture flask

Cost efficient

90% less media supplements and reducedhandling time

Applications

Monoclonal antibody production in hybridomas

Recombinant protein expression in transfected cells

Virus production

Continuous culture maintenance for long-term studies

High-density cell culture

Boost your production of monoclonal antibodies or recombinant proteins by cultivating cellsat highest densities with CELLine, the disposable bioreactor based on Two-CompartmentTechnology.

INTEGRA Biosciences 5

Stan

dard

Ope

ratin

g In

stru

ctio

ns2.2 CELLine CL 1000

2.2.1 Required Material and Preparation

● CELLine CL 1000 Bioreactor● Standard 25 ml serological pipettes

with Pipetting aid ● Preculture of totally 25 x 106 viable cells● 350 ml of fresh nutrient medium suitable for your individual cell type and

equilibrated to the desired culture temperature (see 3.1).● 5 ml of fresh complete medium

For more information on media composition please also refer to general note 3.2.

2.2.2 Equilibration of CELLine

Day 1 In order to obtain optimal performance of CELLine put 50ml of nutrient medium intothe medium compartment and let the semi-permeable membrane equilibrated for atleast 5 minutes (see 3.3).

2.2.3 Preparation of Inoculum

Obtain 25 x 106 viable cells from a pre-culture in log growth phase and suspendthe cells in 15 ml fresh medium resulting in a minimal concentration of about 1.5 x 106 viable cells / ml (see 3.4).

2.2.4 Inoculation of CELLine

Loosen the big, front cap of medium compartment in order to prevent air lock. Aspirate the 15 ml cell suspension into a serological pipette, open the cell com-partment and inoculate the cell compartment by inserting the pipette into the blacksilicone cone.

It is important to minimize the introduction of air bubbles into the cell compartmentduring seeding. In case air gets trapped within the cell compartment try to carefully remove the big bubbles by carefully drawing them back into the pipettetogether with fluid. Close the cell compartment by completely tighten the cap.

After seeding add 975 ml of equilibrated medium into the medium compartment and then completely tighten both caps. Place the CELLine into a standard CO2 incu-bator under culture conditions appropriate for your individual cell type.

Minori Ito

6 User Manual CELLine 2003

Stan

dard

Ope

ratin

g In

stru

ctio

ns 2.2.5 Culture monitoring (optional)

Day 3 After 72 hours, take a sample from the cell compartment for assessment of celldensity and viability, expression levels of recombinant protein or determination ofother individual critical culture parameters. This is especially important when cul-turing a new cell type in order to establish a working protocol.

2.2.6 Cell compartment harvest and Spilt back

Day 7 In general, the first harvest is recommended 7 days after inoculation (see also note 3.5).

In order to harvest the cells, simply pour off and discard all medium from the medium compartment.

Avoid to shake the CELLine during this process (see note 3.6)

Loosen the white medium compartment cap

Gently harvest all liquid from the cell compartment by aspirating content with a 25 ml serological pipette. Slowly pipette the liquid up and down several times to thoroughly mix the cell suspension The cell compartment will comprise about 15 mlcell suspension with the individual secreted product. Due to osmotic flux of liquidfrom the medium - to the cell compartment, the total volume might be slightly increased (see note 3.7).

Take 3 ml of mixed cell suspension and add to 12 ml fresh complete medium(1:4 Split Back) and gently return the 15ml of cell suspension back into the cellcompartment (see note 5.5)

Remove any air bubbles as described above. Tighten the green medium compart-ment cap.

Add 1000 ml of fresh, preheated nutrient medium to the medium compartment. Place CELLine back into the incubator until next harvest.

2.2.7 Harvesting Cycles

from Consecutive harvests can approximately be made every 5 to 7 days (dependingDay 14 on the individual application and cell type used, also see not 3.5). All harvests are

performed as outlined above and should include a change of the culturemedia.

Periodically, cells can be monitored for growth and production by removing a smallsample from the cell compartment . If the CELLine Bioreactor is handled with careand the sterility barrier is not broken individual cultures can be maintained over several months.

CELLine - Literature Thursday, August 17, 2006

General Publications

High density Cell Culture in a new passive membrane based bioreactor

M. Wolf, T. DeSutter; Biotechnology International 1999

Celline Overview, Application Introduction This report is intended to be an overview of the CELLine regarding itsoperation and the principle of cell compartment viable cell capacity. Thisreport describes the impact of nutritional medium exchange and thediffusion of solute accross the semi permeable membrane. The importanceof innoculation density and its impact on culture initiation is described. Thevariable volume of the cell compartment, expected osmotic water flux, andequivalency between the different CELLine models is also reported. >more

M. Wolf; Wilson Wolf Corporation, Minneapolis USA

Small-scale Biomanufacturing Benefits from Disposable Bioreactors

In this article different disposable cell culture systems are compared and a cost analysis for production of MAbs using CELLine or conventional methods is presented.

Fabrizio Baumann; Biopharm International 2005, 18 (12): 22-30

Antibody Expression in Hybridoma

Dialysis-based bioreactor systems for the production of mono-clonalantibodies - alternatives to ascites production in mice

M.P. Bruce, V. Boyd, C. Duch, J.R. White; Journal of Immunological Methods 2002, 264: 59-68

Manufacture of Pure Monoclonal Antibodies by HeterogeneousCulture without Downstream Purification

L.E. Scott, H. Aggett, D.K. Glencross; Biotechniques 2001, 31-3: 666-68

Efficient laboratory-scale production of monoclonal antibodies usingmembrane-based highdensity cell culture technology >more

M. Trebak, J.M. Chong, D. Herlyn, D.W. Speicher; Journal of Immunological Methods 1999, 230: 59-70

Comparison of batch vs. Celline culture for production of mono-clonal antibody in vitro as alternatives to ascites. Application: MurineHybridoma This report is a comparison of antibody production in the CELLine comparedto a traditional batch culture method. The seven hybridoma clones whichwere cultured in the CELLine had been previously cultured using a traditionalbatch method. Comparison of results obtained in the CELLine and in batchculture is provided. The benefits of a concentrated product, reducedhandling, and reduced serum use are demonstrated. The consumption ofnutrient medium by the two different methods is reported and provided foreach clone. >more


Antibody Manufacture in the Celline CL1000, Application: Murinehybridoma

This report includes results from the culture of over 30 different hybridomaclones in the CELLine. The results from the large number of different clonesprovides a range of performance which the user can use to compare theirindividual results. Despite significant differences which arise from clone toclone, mean target values for cell numbers, antibody yield and harvest timesand volume are provided derived from cultures of different hybridomaclones. Importantly, the clones were randomnly selected representingnumerous mAb isotypes and fusion partners. Additionally, all the cultureswere carried out in a small manufacturing environment aimed towardsreducing overall costs and labor. >more


Attachment dependent cell cultivation on Micro-Carriers: Secre-tedprotein production >more


Minori Ito

Membrane-based Cell cutlure systems - an alternative to in Vivoproduction of monoclonal antibodies

A new generation of membrane based cell culture devices especiallydesigned for small scale production of monoclonal antibodies (mab´s)entered the market in last few years. In contrast to conventional perfusionhollow fibre bioreactors these devices contain two functionally differentmembranes - one ultrafiltration membrane for nutrient supply and one gaspermeable membrane for direct oxygenation of cells. The latest systems ofthis generation are static culture systems are of moderate costs and eitherbetter than, or equal to, the ascites mice in terms of quality and quantity ofproduced monoclonal antibodies. We have investigated the advantages ofthe perfused Tecnomouse bioreactor and the static CELLine culture flasks incomparison to ascites production and conventional roller bottle cultures.>more

A. Nagel, S. Koch, U. Valley, F. Emmrich, U. Marx; Dev Biol Stand 1999, 101: 57-64

Rapid evaluation of hybridoma behaviour in high cell density pro-duction process

S. Koch, A. Kriszo, C. Kloth, U. Marx, A. Nagel; ESACT 1999, Lugano 1999

Membrane-based cell culture technologies: a scientifically andconomically satisfactory alternative to malignant ascites produc-tionfor monoclonal antibodies

U. Marx; Res Immunol 1998, 149: 557-9

Protein expression in CHO cells

A soluble TGF-b type-1 Receptor mimics TGFb responses F. Docagne, N. Colloc'h, V. Bougueret, M. Page, J. Paput, M. Tripier, P. Dutartre, E.T. MacKenzie, A. Buisson, S. Komesli, D. Vivien; Journal of Biological Chemistry 2001, 276-49: 46243-50

High density suspension culture for Recombinant Protein produc-tion from CHO cells >more


Protein expression in BHK cells

Long-Term High Level Protein Expression in Adherent, Protein-freeGrowing BHK Cells Using INTEGRA CELLine adhere 1000 Bio-reactorFlasks

J. Mittermaier, M. O. Zang-Gandor; EUGENEX Biotechnologies GmbH, Tägerwilen, Switzerland

Production in Baculovirus Infected SF9 Cells

Continuous Recombinant Protein Production in Baculovirus InfectedSF9 Cells using CELLine classic 1000 Two-Compartment Bioreactors

Izumi Matsumoto (GSI Creos GmbH, Tokio, Japan) and Alex Studer (INTEGRA Biosciences, Chur, Switzerland)

Fermentation

The Application of in vitro Models for Production of Metabolites:Isolation and Characterisation of Hydroxysecobarbital

D. Marshall, M. Robinson, P. Hincks, M. Dumasia, P. Teale, E. Houghton; hromatographia Supplement 2000, 52: 35-38

Protein production in plant cells

Production of Human alpha-1-Antitrypsin from Transgenic Rice CellCulture in a Membrane Bioreactor

McDonald KA et al. 2005, 21: 728-34

CELLine Technical Report ICELLine Overview • Application: IntroductionMartin Wolf Ph.D. • Wilson Wolf Corp. • Minneapolis • USA

Page 1

The CELLine culture devices are based on a compart-mentalization approach. Cells are cultured in a cell com-p a rtment separated from nutrient medium by a semi-per-meable membrane. Cells in the cell compartment are main-tained in a small volume and supported by a larger volumeof nutrient medium contained in the nutrient medium re s e r-v o i r. Successful culture in CELLine devices re q u i re manipu-lation of both the cell compartment contents and the nutri-ent medium. Attributes such as high cell density, re d u c e ds e rum use, ease of handling are available to the usert h rough this compartmentalization approach. The followingo v e rview is provided to highlight certain features and func-tions of the CELLine devices to assist the user in develop-ment of individual and varied applications.

High cell density: Cells can be cultured at high densitywithin the cell compartment of the CELLine. In comparisonto traditional static culture, the number of cells per milliliter,which can be maintained within the cell compartment of theCELLine is much higher. For example, hybridoma cells, lym-phocytes and leukemic cells can reach concentations whicha re 20 to 30 fold greater when compared to growth in sta-tic vessels. This provides benefits to the user by pro v i d i n gi n c reased concentrations of secreted effector molecules,high antibody concentrations, and unique cell culture envi-ronments not available in traditional low density culture .

The separation of the cells from the nutrient medium by as e m i - p e rmeable membrane allows for trapping or re t a i n i n gcell secreted factors and the ability to conserve the use ofexogenous culture factors. The 10,000 MWCO semi-per-meable membrane allows the retention of many cytokineswhich are approximately 10,000 MW or larg e r. Forinstance, IL-2 provided only to the cell compartment exert sfull biological activity on the cells in culture. Conversely,s e c reted cytokines produced within the cell compart m e n tcan reach levels not normally associated with traditionalstatic culture. Conditioned medium can be produced withsignificantly diff e rent levels of effector molecules in theCELLine than in traditional static culture .

For production of monoclonal antibodies the benefits of thec o m p a rtmentalized strategy provide cost savings and laborsavings. Hybridoma cells can be cultivated at high densitywithin the cell compartment of the CELLine producing high-ly concentrated titers of antibody within the cell compart-ment volume. Antibody titers in excess of 1 mg/ml are ro u-tinely achieved. Concentrated antibody in small super-natant volumes can be diluted and used directly or con-centrated supernatant can be applied directly to aff i n i t ypurification columns.

I m p o rt a n t l y, the consumption of serum can be significantlyreduced by providing serum only to the cell culture com-p a rtment and eliminating serum from the nutrient medium.A 1 liter culture in the CELLine consumes only milliliters ofs e rum in contrast to hundreds of milliliters of serum whichwould normally be consumed in static culture. These sav-ings in serum cost quickly accumulate as the duration of cul-t u re increases. An added benefit, is removal of interf e re n c ef rom serum proteins during antibody purification, as thep roduct is obtained at mg/ml concentrations in a 10%s e rum supernatant. This is in marked contrast to concen-trating serum components in conjunction with the desire dantibody during processing leading to difficulties in purifi-cation and contamination of antibody preparations withnon specific immunoglobulin molecules and other seru mp roteins. Use of serum to generate highly concentrateds u p e rnatant of antibody molecules is not as problematic asdown stream concentration steps are eliminated.

Cell compar t m e n t : Understanding the concept of aviable cell capacity for the cell compartment is important inthe operation of the CELLine for maximum perf o rmance. Thecell compartment in the CELLine has an upper limit to thenumber of viable cells which can be maintained within it.This is termed cell compartment viable cell capacity. As thenumber of viable cells within the cell compartment incre a s-es, the consumption of metabolic substrates and accumula-tion of metabolic byproducts also increases. Diff u s i o na c ross the semi permeable membrane begins to become

Minori Ito

CELLine Technical Report I • Application: IntroductionPage 2

limiting when the viable cell numbers reach the viable cellcapacity in the cell compartment even when a maximum dif-fusion gradient is provided across the semipermeable mem-brane. Import a n t l y, cell proliferation does not cease, asshown in Fig. 1. when viable cell capacity has beenreached. Cells continue to proliferate within the cell com-p a rtment after viable cell capacity is reached. There f o re, atviable cell capacity, the number of viable cells no longeri n c reases however total cell numbers do continue toi n c rease. This can lead to an accumulation of very highnumbers of total cells. Splitting back the cell compart m e n tinfluences the ratio of total and viable cells. Splitting backthe culture when capacity of the cell compartment isreached will maintain high cell viability. Splitting back sev-eral days after capacity is reached does not lead to loss ofviable cells but can lead to increased product concentrationand increased total cell numbers.

The handling strategy for the CELLine should be based onthe needs of the operator. For high percent viability it isi m p o rtant to split the cell compartment back when cellsreach viable cell compartment capacity. This prevents accu-mulation of non viable cells. For higher total cell concen-trations and increased product concentrations less fre q u e n tsplitting of the culture is acceptable. The ratio at which thecell compartment is split during harvest will determine thetime re q u i red for cells to re t run to maximum viable cellc a p a c i t y. Cultures split back two fold ,will re t u rn to maxi-mum viable cell numbers sooner than cultures split back 4fold. To provide extended durations between handling, thecell compartment can be be split further than 50%. Resultsobtained with the CELLine will be dependent upon the strat-egy used in manipulating the cell compartment contentsand in the exchange of nutrient medium as discussedb e l o w.

In summary, the cell compartment can be treated as a stan-d a rd tissue culture flask except that cells are present at highconcentrations. Splitting back cells in the cell compart m e n tp rovides control over cell numbers and percentages ofviable cells. When cells have reached the maximum capac-ity of the cell compartment, proliferation continues andaccumulation of total cells will take place.

Nutrient Medium: The rate of nutrient mediumexchange also impacts culture results. Diffusion across thes e m i - p e rmeable membrane is driven by concentration gra-

dients established between the nutrient medium and the cellc o m p a rtment medium. Glucose flux across the semi-perm e-able membrane is shown in Fig. 2. As nutrient medium isdepleted, the driving force for solute diffusion across thes e m i - p e rmeable membrane is also reduced. A decrease inviable cell mass is associated with depleted nutrient medi-um and accumulation of metabolic byproducts.

The nutrient medium does not change color as significantlyas it does in static culture flasks. The direct gas exchangea c ross the bottom of the cell compartment reduces the accu-mulation of acid within the nutrient medium. Color changeof nutrient medium can be used as an indicator of meta-bolic activity but is not as accurate for assessing when nutri-ent medium should be exchanged. Tracking cell numbers isthe most accurate.

As shown in Fig. 3, increased nutrient medium exchangedoes not significantly increase cell compartment capacity.Note the maintenance of the viable cell mass and the fur-ther accumulation of total cells after viable cell capacity hasbeen reached.

In summary, nutrient medium can be exchanged to main-tain the highest possible gradient for diffusion across thes e m i p e rmeable membrane, or it may be exchanged lessf requently to maximize the efficiency of medium use. Theuser can determine which strategy is most suitable for ap a rticular application.

Cell inoculation: For certain cells a relationship betweencell density at inoculation and the initial rate of cell gro w t hin the cell compartment has been observed. The cell densi-ty dependent outgrowth of a murine hybridoma cell isshown in Fig. 4. When cells were inoculated at lower con-centrations they took longer to reach cell compart m e n tc a p a c i t y, in comparison to cells inoculated at higher densi-t y. The responses of certain cells to low density inoculationmay be due to dilution of growth promoting or condition-ing effects currently not well characterized. Certain celllines may benefit from higher initial inoculation densities.For production purposes maximum perf o rmance is attainedwhen the CELLine is operated at or near viable cell capac-ity and inoculating a higher number of cells at culture initi-ation can leaad to more rapid attainment of viable cellc a p a c i t y.


Osmotic Flux: Water flux across the semiperm e a b l emembrane is driven by diff e rences in protein concentra-tions across the membrane.. When serum is not pro v i d e din the nutrient medium an oncotic gradient between the cellc o m p a rtment and the nutrient medium compartment isestablished. This can lead to increases in the cell compart-ment volume during culture. The change in volume can bec o n t rolled by manipulating the oncotic gradient across thes e m i p e rmeable membrane through the addition of an inex-pensive protein hydrolyzate (lactalbumin) to the basalmedium. In most applications the slight flux of water into thecell compartment is insignificant and can be compensatedby providing slightly higher concentrations of serum orsupplements to the cell compartment. As shown in F i g . 5.changes in the cell compartment volume can be influencedby serum concentration diff e rences across the semi perm e-able membrane, the amount of cells in the cell compart m e n tand the duration between handling of the cell compart-m e n t .

Cell compartment volume capacity: The cell compart m e n tvolume is variable. As shown in Fig. 6, the cell compart-ment can be varied significantly due to the complianceassociated with the semi permeable membrane. For the CL1000 units, a slight change in volume of the cell compart-ment can be induced with no significant increase in pre s-s u re. This allows the user to explore protocols with diff e re n tcell compartment volumes. It is recommended that for ro u-tine use, the specified cell compartment volumes bea d h e red to as increased stress on the semi permeable mem-brane is associated with significant depart u re from re c o m-mended cell compartment volumes. Additionally, theimpact of stretch on the semi-permeable membrane andsolute diffusion is not fully characterized.

S c a l e : Results obtained in any of the INTEGRA BiosciencesCELLine models can be useful in predicting results expectedin the other devices. As shown in F i g . 7, the cell concentra-tions and growth rates obtained in the small scale CL 6WELL were also obtained in the CL 350 and CL 1000 inindependent cultures of the same cell under identical con-ditions.

This allows experimental work to be conducted in the smallscale CL 6WELL and carried directly into the larger CELLineunits for scale up purposes. The scale of CELLine devices isrelated to the surface area of the semipermeable mem-

brane. The larger the surface area the greater the viablecell capacity. The viable cell capacity of the CL 6WELL andCL 350 are approximately 1/30 and 1/3 of the CL 1000re s p e c t i v e l y.

h t t p : / / w w w. c e l l @ i n t e g r a - b i o s c i e n c e s . c o m

Tel.: + 1 - 8 0 0 - 8 8 6 - 8 6 7 5

Te l . : +41-1-877 46 46

Te l . : +49-6404 809-0

Te l . : +33-1-34 46 71 71

Te l . : +44-1462 48 65 48

CH

F

D

GB

U S A


10864200

20

40

60

80Viable Cells

Total Cells

Non-Viable Cells

Day of Culture

Cel

l Num

ber

x 1

0^

6

Cell Compartment Capacity

Figure 1: Cell compartment capacity Figure 2: Glucose flux (all)

Figure 1: Growth curve example: Theoretical growth of cellswithin the cell compartment of the CELLine is depicted. Theopen circles represent viable cell numbers and plateau whenthe cell compartment capacity is reached. Total cells repre-sented by open squares continue to increase throughout theculture period. Non-viable cells represented by filled circlesbegin to accumulate after cell compartment capacity isreached.

Figure 2: Glucose diffusion across semipermeable membrane.Tests conducted at room temperature. RPMI-1640 was placedin nutrient reservoir of individual CL 6WELL compartments.The cell compartment filled with distilled water. Cell compart-ment harvested at indicated times and glucose concentrationdetermined. The driving force for glucose flux decreased withtime as the cell compartment glucose concentrationsincreased. Flux of nutrients across the semi permeable mem-brane during culture will be influenced by depletion of nutri-ents during culture.

1008060402000

100

200

Glucose Flux-CL6WELL

Time (minutes)

Glu

cose

mg/1

00

ml


Figure 3: Growth curve of murine hybridoma cells in CL 1000.Cell compartment: 10% FBS, Nutrient compartment 0.8% FBS,.1% vitacyte. Concentrations of viable and total cells are plot-ted during culture. The viable cell concentration remained rel-atively constant after cell capacity had been reached. Totalcell concentration continued to increase with culture. Notethat exchange of basal medium did not significantly increasethe viable cell numbers.

Figure 4: Cell outgrowth and intial cell concentration. AC.04murine hbyridoma cells were inoculated at 0.5 x 106 and 2.0x 106 cells/ml into individual wells of a CL 6WELL. Cell com-partment 15% FBS, nutrient compartment 0% FBS, RPMI-1640.At days indicated cells were counted and viable cell concen-trations determined by trypan blue exclusion.

0 1 2 4 5 80

20

40

60

80

Viable cells/ml

Total Cells/ml

Day of Culture

Cel

l C

onc

entr

atio

n (x

10

^6

/ml)

Basal medium exchanged

Figure 3: 10.1 Cell conc bar

432100

10

20

300.5

2.0

0.5

2.0

0.5

2.0

Day of Culture

Via

ble

cells

x 1

0^

6/m

l

Figure 4: Incoculation density


20019018017016015014013012011010090807060504030201000

2

4

6

8

10

12

14

16

18

20

Membrane Burst Pressure

Vol CC

Gau

ge

Pres

sure

(m

m H

g)

Figure 6: Volume pressure curve

10864200

10

20

30

Day

Cell

Com

par

tment

Volu

me

10864200

10

20

30

15% 10%

Figure 5: 10 and 15 % water flux

Figure 5: Osmotic water flux into cell compartment of CL 1000and CL 350 units during culture. Cell compartment volumesare plotted at beginning and end of harvest intervals takenduring culture. Cultures were conducted with cell compart-ment serum concentrations of 10% (0.8% nutrient compart-ment) and 15% (0% nutrient compartment).

Figure 6: Cell compartment volume and pressure relationship.The volume of infused air into a wetted cell compartment ofindividual CL 1000 units is plotted versus gauge pressure.Units were tested with 250 ml of water placed atop of cellcompartment. Pressure was incrementally increased in 20 secintervals until membrane burst or developed leak.

Figure 7: Murine hybridoma cells cultured under identical con-ditions in different CELLine devices: Cell compartment 15%FBS, RPMI-1640, Nutrient compartment-0% FBS, RPMI-1640.Cell sample removed from cell compartments on day indicat-ed and cells counts determined. Counts were obtained from asingle CL 1000, CL 350 and 2 CL -6WELL units. The resultsfrom the CL 6WELL were obtained from a single well. Cultureswere split back on indicated days.

201816141210864200

10

20

30

CL1000

CL350

CL6

CL6

Device Comparison: CL6WELL, CL350, CL1000

Day

Via

ble

Cells

x 1

0^

6/m

l

Culture split 1:1 cell: 7G7

Figure 7: Device comparison

As an alternative to ascites, in vitro production of mono-clonal antibody can be accomplished with a variety ofmethods (1-5). Comparison of in vitro methods indicatesnot all methods are equally cost effective or user friendly.The ascites method has a number of perceived advan-tages over most in vitro methods, primarily a low cost ofp roduction based on mg of antibody produced and aconcentrated product which eases down stream pro c e s s-ing re q u i rements.

Many systems and instruments (Bioreactors) designedspecifically to produce mAb in vitro are burdened by cap-ital instrument expense and additional culture - w a re costs.These systems also re q u i re a learning curve for successfuluse. For small scale production, many of these systems arenot cost effective. These systems become more cost eff e c-tive as production increases, due to amortization of costsover larger amounts of produced antibody. In contrast, theascites method does not become less expensive as theamount of antibody to be produced increases. Ascites pro-duction costs are generally linearly related to the amountof antibody re q u i red (number of mice).

For small production runs, ascites has been considere dm o re cost effective than most other methods, as it can bereadily scaled to small production needs. A single mousecan produce milligram amounts of antibody and pro v i d eit in a concentrated form. To effectively compete withascites for production of modest amounts of monoclonala n t i b o d y, in vitro methods must have no capital costs andp rovide additional cost saving benefits. For example, pro-duction of concentrated antibody, reduced serum use, andreduced labor are benefits which make in vitro methodscompetitive to ascites for production of modest amounts ofa n t i b o d y. These benefits are unfortunately available onlywith more costly and complicated systems and are not coste ffective for producing smaller amounts of antibody. The INTEGRA Biosciences CELLine culture devices have nocapital equipment costs, a low purchase cost, provide con-

centrated product, reduce serum use, and reduce han-dling. Importantly all of these attributes are available in adevice as simple to use as a standard tissue culture flask.The CELLine CL 1000 was used to produce modestamounts of monoclonal antibody in a manufacturing lab-o r a t o ry. The CL 1000 is a member of a product family ofhigh density cell culture devices available from INTEGRABiosciences (Ijamsville, MD). The CL 1000 has a re s e rv o i rvolume of 1 liter with a cell compartment capacity of 20-30 ml. The results obtained with the INTEGRA BiosciencesCELLine CL 1000 are presented to provide a demonstra-tion of the cost savings and additional benefits obtainedby producing antibody in these devices.

The results from 7 diff e rent murine hybridoma cultures car-ried out in the CL 1000 are presented. The cultures werec a rried out by a commercial manufacturing laboratorywhich produces and sells antibody reagents to there s e a rch community. A minimal handling protocol (contin-uous batch process) was employed by the manufacturingl a b o r a t o ry to reduce the handling of the cultures. Thehybridoma cells were maintained in the cell compart m e n tof the CL 1000 and harvested at approximately 7 dayi n t e rvals. During harvest, cells and supernatant were col-lected from the cell compartment, a fraction of the harv e s twas reinoculated into the cell compartment with fre s hmedium. Nutrient medium was removed and re p l a c e dwith fresh medium on day of harvest. Results obtainedf rom cultures in the CL 1000 are compared to pre v i o u sresults obtained with the same cell lines cultured in a batchmethod in traditional tissue culture flasks at the facility.

M e t h o d s : Murine hybridoma cell lines (Table 1) werethawed from frozen stocks and expanded in static culture(RPMI-1640, 10-15% FBS, 2X L-Glutamine, Pen-Stre p ) .After demonstration of consistent cell doubling in static cul-t u re, cells were inoculated into the CL 1000 devices.

Page 1CELLine Technical Report IIComparison of batch vs. CELLine culture for production ofmonoclonal antibody in vitro as alternatives to ascites.Application: Murine Hybridoma

Martin Wolf Ph.D. • Wilson Wolf Corp. • Minneapolis • USA

Minori Ito

CELLine Technical Report II • Comparison of batch vs.CELLine culture for production of monoclonal antibody in vitro asa l t e rnatives to ascites • Application: Murine Hybridoma

Page 2

Table 1

H y b r i d o m a I s o t y p e Fusion Inoculation P a rt n e r Concentration

Viable Cell/ml

A C . 1 I g G 1 N s - 1 7.5 x 106

A C . 2 I g G 2 a 6 5 3 6.8 x 106

A C . 3 I g G 1 N s - 1 3.7 x 106

A C . 4 I g G 2 a S p 2 0 5.6 x 106

A C . 5 I g G 1 S p 2 0 7.2 x 106

A C . 6 I g G 1 n . k . 7.2 x 106

A C . 7 I g G 1 n . k . 5.5 x 106

n.k. = not known

Cell compartment medium: RPMI-1640; 2X L-glutamine (5mM ), penicillin G (66 mg/L), streptomycin sulfate (144mg/l). Basal medium was supplemented with 10% FBS(Hyclone, Logan Utah). Additional supplementation ofmedium with a hybridoma growth supplement (0.1% Vi t a-cyte, J. Brooks Irvine, CA)) was done to remain consistentwith prior batch production runs of the same cell lines intraditional flasks.

Nutrient medium: RPMI-1640; 2X L-glutamine (5 mM ) ,penicillin G (66 mg/L) streptomycin sulfate (144 mg/l)with 0.8% FBS, 0.1% Vi t a c y t e .

I n o c u l a t i o n : Cells were inoculated from static culture atDay 0 in a 20 ml volume into the cell compartment of theCL 1000 devices. Inoculation density was maintainedabove 3.0 x 106 cells/ml. Cells were removed fro mf rozen stock initiated cultures and re suspended in fre s hcell compartment medium prior to inoculation. Nutrientmedium (1000ml) was supplied to the nutrient mediumc o m p a rtment and the devices placed into a 5% CO2, 37°C. humidified tissue culture incubator.

H a rvest: At harvest, the total cell compartment volume wasremoved from the CL 1000 units by pipette. Cell numbersw e re determined by diluting and counting samples usinga standard hemacytometer. Viable cells were discriminat-ed from non-viable cells by trypan blue staining and phasecontrast micro s c o p y. Cell compartment contents were splitback between 3-5 fold determined by cell numbers in thecell compartment at time of harvest. Fresh cell compart-ment medium (17-15 ml) was added to the cell fraction (3-5 ml) to achieve a 20 ml volume and the cell suspensionre t u rned to the cell compartment. The harvested cell con-

taining supernatant fraction was kept sterile and stored at4 °C until purification by affinity chro m a t o g r a p h y. Nutrientmedium (1000ml) was removed and replaced with fre s hmedium at the time the cell compartment was harv e s t e d .Devices were re t u rned to incubator until next harv e s t .Devices were stacked atop of each other in the incubator.

Antibody purification: C u l t u re supernatant was pro c e s s e dby eluting antibody from protein A affinity chro m a t o g r a-phy columns following manufacturers protocol. Eluted anti-body fractions were collected, pooled and antibody quan-tified by spectrophotometer and ELISA. Sandwich ELISAwas perf o rmed with polyclonal goat anti-mouse IgG orIgM capture antibody and polyclonal anti-mouse IgG orIgM antibody labeled with peroxidase. Color was devel-oped with ABTS. Antibody purity was assessed by SDS-PAGE and Coomassie blue staining. Purified antibody wassubjected to further labeling reactions to generate Fluore s-cein and Phycoerythrin conjugates. Purified antibody wassubjected to internal quality testing and re l e a s e d .

R e s u l t s : A re p resentative cell growth curve is shown inF i g . 1. The total cells and total viable cells are plottedagainst day of culture. The cell compartments were splitback at harvest and growth of the reinoculated cellsresumed. The semi continuous batch process generatedover 1x109 total cells within the cell compartment at har-vest. The total viable cell numbers obtained prior to or atday of harvest indicate the maximum capacity for viablecells in the cell compartment. The maximum cell numberscounted from a single harvest during the entire culture peri-od for the individual cultures are shown in Table 2.

Table 2

H y b r i - Maximum Maximum d o m a Viable Cells Total Cells

(cell compart m e n t ) (cell compart m e n t )

A C . 1 561 x 106 1701 x 106

A C . 2 660 x 106 3304 x 106

A C . 3 520 x 106 2140 x 106

A C . 4 731 x 106 2080 x 106

A C . 5 682 x 106 1957 x 106

A C . 6 570 x 106 1476 x 106

A C . 7 504 x 106 1120 x 106


Page 3

The production results for the individual cultures are shownin Table 3. The mean number of harvests for the 7 cultures was 5. A mean harvest volume of 114 ml wasp rocessed for antibody purification. Clone AC.5 had anunusually low yield of antibody and analysis indicated thatthe majority of the antibody present was not intact (SDS-PAGE; size analysis). Supernatant from batch culture ofthis clone in traditional flasks also contained partial anti-body fragments. Results from this clone were not includedin the following mean values as it was considered to be anon productive clone. It should be noted that a modestamount of recoverable antibody was obtained from theCELLine culture while supernatant from batch cultures wereunable to be used.

A mean of 118 mg of purified antibody per culture wasobtained. The mean duration for the 6 cultures was 39days. The mean antibody concentration re c o v e red fro mh a rvested supernatant was 1.02 mg/ml. This concentra-tion was calculated by dividing the volume of pro c e s s e ds u p e rnatant (116 ml) by the amount of antibody whichwas re c o v e red following purification (118 mg). The mg ofantibody produced per liter of nutrient medium consumedwas 23.7 mg/l. The mean antibody concentrationobtained from the standard batch culture method used pre-viously in production for the 6 diff e rent clones was

3 1 mg/l. The batch cultures were maintained until viablecell numbers were nearly exhausted.

The CELLine 1000 devices consumed significantly lesss e rum to produce antibody when compared to results fro mthe traditional batch cultures. Nutrient medium in theCELLine CL 1000 was supplemented with only 0.8% FBSand the cell compartment was supplemented with 10%FBS. About 50-60 ml of FBS was consumed for an individ-ual culture. In contrast a batch culture consumed appro x i-mately 380 ml of FBS to produce equivalent amounts ofa n t i b o d y. Supplementation of the nutrient medium in theCELLine CL 1000 with FBS has been demonstrated not tobe necessary for many hybridoma clones.

Purification of the culture supernatant from the CELLineC L 1000 resulted in both time and labor savings whenc o m p a red to traditional batch culture. Instead of concen-trating more than 3 liters of medium to recover compara-ble amounts of antibody, the CELLine supern a t a n t(mean:116 ml) was centrifuged and applied directly to thea ffinity purification columns. Consequently, there was nosimultaneous concentration of serum protein in the super-natant which can lead to purification difficulties associat-ed with concentrating traditional culture supernatant.

Table 3

H y b r i - Number of Total harvest Mg Ab C u l t u re Duration mAb concentra- mAb mg/literd o m a H a rv e s t s volume ml Total mg d a y s tion mg/ml nutrient medium

A C . 1 5 1 2 8 1 2 1 4 2 1 . 0 6 2 4 . 2A C . 2 6 1 3 0 1 5 8 4 2 1 . 2 1 2 6 . 3A C . 3 4 1 0 0 8 0 3 0 0 . 8 1 9 . 9A C . 4 6 1 2 0 1 4 1 5 0 1 . 1 7 2 3 . 5A C . 5 5 1 0 4 2 3 3 8 0 . 2 2 4 . 6A C . 6 4 1 1 0 1 2 1 3 7 1 . 1 3 0 . 4A C . 7 5 1 1 0 8 7 3 6 0 . 8 1 7 . 6M e a nn = 6 5 . 0 1 1 6 . 3 1 1 8 . 1 3 9 . 5 1 . 0 2 2 3 . 7


Page 4

Summary: The results obtained from hybridoma culturesin INTEGRA Biosciences CELLine CL 1000 culture flasksindicate the following: 1. Cell growth was obtained in all of the 7 cultures (Oneclone AC.5 was not considered a productive clone andexcluded from the mean values derived for the cultures). 2.The mean production of the remaining 6 clones was 118mg of purified antibody over a mean culture duration of39 days. 3. Significant reductions in serum consumption wasobtained compared to traditional batch cultures of thesame cell lines. 4. Nutrient medium consumed per mg of antibody pro-duced in the CELLine CL 1000 was slightly greater thanthat obtained in batch culture, 23.7 mg/l compared to 31mg/l. The nutrient medium in the CL 1000 was less expen-sive when compared to that used in batch culture as it con-tained only 0.8% FBS. The lower efficiency in nutrientmedium consumption was more than offset by re d u c e dnutrient medium costs by significant reduction in consump-tion of FBS. 5. A concentrated supernatant (1 mg/ml; mean) wasobtained from the CL 1000 cultures when compared toprior batch cultures, reducing downstream pro c e s s i n gre q u i rements and eliminating the need for a concentratingstep prior to antibody purification.

The results indicate that a single CL 1000 can pro d u c eg reater than 100 mg of purified antibody in appro x i-mately 39 days while requiring only 5 harvests. The semi-continuous batch method allowed for sequential harv e s t sand reduced handling. The CL 1000 units were no mored i fficult to handle than traditional tissue culture flasks anddid not re q u i re syringes to access cells or medium. The CL1000 units were stacked in the incubator and did not takeup extra incubator space or re q u i re any support systemssuch as a roller apparatus or perfusion pump. A concen-trated product was obtained from all of the cultures whichsimplified volume handling and processing downstream ofthe culture, eliminating any concentration steps norm a l l yassociated with processing traditional culture supern a t a n t .

In conclusion; the CELLine CL 1000 units are considere dcost competitive to ascites as an in vitro method for pro-ducing antibody. The operating costs (medium and seru m )w e re significantly less than those associated with tradition-al in vitro batch methods using standard tissue culture flasksor gas permeable bags. Additional benefits such as con-

centrated product, reduced handling, and simplified down-s t ream processing were also obtained with the CL 1000devices. Most import a n t l y, the ease of use and re d u c e dhandling needed to achieve the above mentioned benefitsmake the CELLine culture flasks a viable alternative toascites for producing monoclonal antibody.

R e f e re n c e s :

1. Modified CelliGen packed bed bioreactors forhybridoma cell cultures. Wang G, Zhang W, Jacklin C,F reedman D, Epstein L, Kadouri A, Cytotechnology, 9(1-3):41-49, 1992

2. In vitro production of monoclonal antibodies in highconcentration in a new and easy to handle modularm i n i f e rm e n t o r. Falkenberg FW, We i c h e rt H, Krane M, Bar-tels I, Palme M, Nagels HO, Fiebig H, J. Immunol. Meth-ods, 179(1):13-29, 1995

3. Evaluation of hollow fiber bioreactors as an altern a t i v eto murine ascites production for small scale monoclonalantibody production. Jackson LR, Trudel LJ, Fox JG, Lip-mann NS. J. Immunol. Methods, 189 (2):217-231, 1996

4. Increase of hybridoma productivity using an originaldialysis culture system. Mathiot B, Perani A, Dumas D,Maugras M, Didelon J, Stoltz JF, Cytotechnology 11(1):41-48, 1993

5. Factors affecting monoclonal antibody production inc u l t u re. Reuveny S, Velez D, Macmillan JD, Miller L., Dev.Biol Stand. 66:169-175, 1987


Tel.: + 1 - 8 0 0 - 8 8 6 - 8 6 7 5

Te l . : +41-1-877 46 46

Te l . : +49-6404 809-0

Te l . : +33-1-34 46 71 71

Te l . : +44-1462 48 65 48

CH

F

D

GB


Page 5

F i g u re 1: G rowth Curve (AC.5). The number of total cells andviable cells which were re c o v e red from the cell compart m e n tof the CL 1000 during culture are shown. The cells were count-ed using a standard hemacytometer and non viable cellsd e t e rmined by trypan blue staining. The culture was harv e s t-ed at times indicated by the arrows. The cell numbers weresplit back as indicated and cell growth between harvests isshown. The semi continuous batch process; harvests followedby resumed cell expansion of the inoculum is readily seen. Theviable cell capacity of the CL 1000 was demonstrated during

the initial 5 days of culture. Note, that the number of viablecells did not continue to increase, but declined as nutrientmedium was consumed. Addition of fresh nutrient mediumdoes not lead to increased viable cell numbers within the cellc o m p a rtment beyond its capacity. The viable cells do contin-ue to proliferate at maximum capacity and lead to the veryhigh total cell numbers seen at harvest. Splitting back of thecell compartment allowed for continued cell proliferation andremoval of excess cells. The growth of all the hybridomaclones was similar to the growth shown for AC.5.

403530252015105010

100

1000

10000

total viable

total cells

Day

Cel

ls x

10

^ 6

Figure 1

The successful production of antibody in vitro is dependentupon numerous factors. A variety of methods and devicesfor producing monoclonal antibody are available. One ofthe most important variables which impacts pro d u c t i o nrates and costs is the clone itself from which antibody is tobe produced (1-3). Not only do the same clones perf o rmd i ff e rently under varied culture conditions, but diff e re n tclones produce markedly diff e rent amounts of antibodywhen cultured under the same conditions (4,5).

For many manufacturers of antibody, a large number ofd i ff e rent clones are routinely cultured to produce mono-clonal antibody. The production amounts of antibodyre q u i red may be only 100-200 mg. For these applica-tions, a simple to use culture method which provides ben-efits associated with more expensive and complicated sys-tems for larger production amounts is of benefit. The abil-ity of the INTEGRA Biosciences CELLine culture flasks tomeet the needs of the small scale manufacturing laborato-ry is provided below.

To assess the perf o rmance of the INTEGRA B i o s c i e n c e sCELLine culture devices in the production of antibody inv i t ro under manufacturing conditions; a large number ofhybridoma clones (32 individual clones) were cultured inCL 1000 CELLine units by an independent manufacturingc o n c e rn. The clones were selected at random and culture dby the independent commercial laboratory pro d u c i n gmonoclonal antibody for use in re s e a rch. The antibodyp roduced was purified from culture supernatant by aff i n i t yc h ro m a t o g r a p h y. All produced antibody passed qualitys t a n d a rds (purity, activity, specificity) except for one clone(see below) and was released as product.

The results indicate a range of production achieved usingthe CELLine CL 1000 in routine in vitro production of mon-oclonal antibody under manufacturing conditions. Themanufacturing laboratory chose a reduced handling pro-tocol to minimize labor and handling costs. Changes in

FBS supplementation and medium were evaluated duringthe process by the manufacturing laboratory. The pro t o c o lwas not optimized for the individual clones.

M e t h o d s : Murine hybridoma cell lines were thawed fro mf rozen stocks and expanded in static culture (RPMI-1640,10-15% FBS, 2X L-Glutamine, Pen-Strep). After demonstra-tion of consistent cell doubling in static culture, cells wereinoculated into the CL 1000 devices. The hybridomaclones were obtained from sources around the world andincluded both clones obtained under license and clonesgenerated by the manufacture r. The clones (fusion part-ners, isotypes) were randomly selected based on pro d u c-tion needs. The cell lines cultured were derived from fusionp a rtners which included 653, NS-1, SP20. Clones arecoded by the manufacturer for confidential reasons.

Cell compartment medium: RPMI-1640 or DMEM; 2X L-glutamine (5 mM), penicillin G (66 mg/L), stre p t o m y c i nsulfate (144 mg/l). Basal medium was supplemented with10%, 15% or 20% FBS (Hyclone, Logan Utah). Addition-al supplementation of medium with an additional hybrido-ma growth supplement (0.1% Vitacyte, J. Brooks Irv i n e ,CA)) was done for some cultures and not others.

Nutrient medium: RPMI-1640 or DMEM; 2X L-glutamine (5mM ) , penicillin G (66 mg/L), streptomycin sulfate (144mg/l) with 0%, 0.8%, or 5% FBS. Additional supplemen-tation of medium with an additional hybridoma gro w t hsupplement (0.1% Vitacyte) was used for some culture s .

I n o c u l a t i o n : Cells were inoculated from static culture atDay 0 in a 20 ml volume into the cell compartment of theCL 1000 devices. Inoculation density was maintainedabove 2.0 x 106 cells/ml. Cells were removed fro mf rozen stock initiated cultures and re suspended in fre s h

CELLine Technical Report IIIAntibody Manufacture in the CELLine CL1000Application: Murine hybridomaMartin Wolf Ph.D. • Wilson Wolf Corp. • Minneapolis • USA

Page 1

Minori Ito

CELLine Technical Report III • Antibody Manufactur ein the CELLine CL1000 • Application: Murine hybridoma

Page 2

cell compartment medium prior to inoculation. Nutrient medium (1000ml) was supplied to the nutrient mediumc o m p a rtment and the devices placed into a 5% CO2,3 7 ° C. humidified tissue culture incubator.

H a rv e s t : At harvest, the total cell compartment volume wasremoved from the CL 1000 units by pipette. Cell numbersw e re determined by diluting and counting samples usinga standard hemacytometer. Viable cells were discriminat-ed from non-viable cells by trypan blue staining and phasecontrast micro s c o p y. Cell compartment contents were splitback between 3-4 fold determined by cell numbers. Fre s hcell compartment medium was added to the cell fraction(4-5 ml) to achieve a 20 ml volume re t u rned to the cellc o m p a rtment. The harvested cell containing supern a t a n tfraction was kept sterile and stored at 4°C until purifica-tion by affinity chro m a t o g r a p h y. Nutrient medium(1000ml)was removed and replaced by pouring duringcell compartment harvest. Devices were re t u rned to incu-bator until next harvest. Devices were stacked atop ofeach other in incubator

Antibody purification: C u l t u re supernatant was pro c e s s e dby eluting antibody from protein A affinity chro m a t o g r a-phy columns following affinity resin manufacturers pro t o-col. Eluted antibody fractions were collected, pooled andantibody quantified by spectrophotometer and ELISA.Sandwich ELISA was perf o rmed with polyclonal goat anti-mouse IgG or IgM capture antibody and polyclonal anti-mouse IgG or IgM antibody labeled with pero x i d a s e .Color was developed with ABTS. Antibody purity wasassessed by SDS Electro p h o resis and Coomassie bluestaining. Purified antibody was subjected to further label-ing reactions to generate Fluorescein and Phycoery t h r i nconjugates. Labeled antibody was subjected to intern a lquality testing and re l e a s e d .

R e s u l t s : As shown in Table 1, a total of 33 individualmurine hybridoma cultures were completed in the CELLinedevices. Antibody isotypes included IgG1, IgG2a, IgG2band Ig-M. A myeloma (AC.19 MOPC) producing IgG1was also cultured. Cultures are presented in the sequencethat they were run. Changes in serum supplementation,and medium were implemented by the laboratory as indi-cated. If cell growth was not established following inocu-

lation, increased serum was provided to either cell com-p a rtment or the nutrient medium as indicated. The hybridoma clones were obtained from sources aro u n dthe world and included clones obtained under license(majority) and clones generated by the manufacture r. Theclones were randomly selected based on manufacture r sp roduction needs and re p resented a random sampling ofisotypes and fusion partners. Many of the clones werenewly received by the manufacturing laboratory and didnot have extensive production re c o rds indicating expectedp roduction levels.

Clones AC 11,12,13 and 16 were difficult to establish ini-tially but were successfully propagated after incre a s i n ginoculation densities and or providing increased seru mconcentrations in the nutrient medium. One clone AC.9g rew well in the devices but did not produce suff i c i e n tamounts of antibody. Sub-cloning (AC17) of clone AC.9was not satisfactory in restoring antibody production toacceptable levels. The clone was unable to produce suit-able product amounts when cultured in static culture flasks.Several clones (AC 18,31) were treated with an antimycoplasma agent (enrofloxacin, Baytril) to treat suspect-ed mycoplasma contamination. These clones were diff i c u l tto establish initially but grew and produced antibody ats a t i s f a c t o ry levels following treatment. Cells were tre a t e doutside of CL 1000 flasks.

The total cell compartment supernatant volumes collectedf rom each culture are shown in Table 1 (Harvest Vo l ) .AC15 was ran simultaneously in two units to increase pro-duction. The number of harvests are also shown indicatingthe total number of harvests/handling operations for eachc u l t u re. Clone AC 2 was cultured in 3 CL 350 flasks. Theh a rvest volume was determined from pooling the thre eflasks. Cell numbers are those counted from a single flask.The CL 350 has 1/3 the surface area as the CL 1000 andhas 1/3 the capacity.

Maximum viable and maximum total cell numbers whichw e re counted in supernatant from the cell compart m e n tduring harvest are also shown for cultures which weretracked by counting cells. The maximum cell counts re f l e c tthe highest numbers obtained during a harvest during thec u l t u re period and reflect the maximum cell capacitiesattained in the cell compartments during culture for thep a rticular hybridoma. Representative growth curves fro mindividual cultures are shown in Fig. 1.


Page 3

Cells proliferated throughout the interval between harv e s t sas indicated by the accumulation of total cells. Maximumviable cell numbers ranged between 318 and 800 x 106

cells per cell compartment for the diff e rent clones. Themaximum number of total cells counted in a harvest fro mthe cell compartment ranged between 648 and 3304 x1 06 cells for the diff e rent clones. The continued pro l i f e r a-tion of cells after viable cell capacity was re a c h e daccounted for the large numbers of total cells re c o v e red ath a rvest and contributed to overall antibody production.

Nutrient medium was exchanged during harvest and wasusually depleted as evidenced by a decline in the totalviable cell numbers in the cell compartment at harv e s t .Nutrient medium contained 0.8% FBS in the initial experi-ments, supplemented with 0.1% Vitacyte. Supplementationwith 0.1%Vitacyte was done to duplicate culture conditionsused in prior production runs in batch culture. Cell com-p a rtment medium was initially supplemented at 10% FBS.Subsequent experiments were ran with 0% FBS in the nutri-ent medium and 20% FBS in the cell compartment and noadditional supplementation. The protocols were selected bythe manufacture r. No significant diff e rence in antibody pro-duction was re p o rted by the manufacturing concern. Thefinal protocol conditions were left at 20% FBS in the cellc o m p a rtment and 0% FBS in the nutrient medium. The 20%FBS was chosen by the manufacturer to allow for pro-longed intervals between handling to compensate for waterflux into the cell compartment. Water flux into the cell com-p a rtment was observed in all cultures resulting in incre a s e dcell compartment volumes compared to inoculation vol-umes at harvest. The volume increase was influenced byduration between harvests, and cell numbers present ath a rv e s t .

A mean of 136 mg of purified antibody was re c o v e red fol-lowing affinity column purification of supernatant from the30 individual cultures (Table 2). The average culture duration was 36 days. The mean har-vest volume was 144 ml of supernatant. The mean (puri-fied) antibody concentration in the harvest supernatant wasnearly 1 mg/ml. This was determined by dividing the har-vest supernatant volume by the amount of antibody re c o v-e red following purification of the supernatant. The range of(purified) antibody concentration was between 2.21mg/ml and 0.36 mg/ml for the individual cultures. Thelongest culture was 54 days (AC 18) which included a lackof initial growth which was subsequently established. Cell

lines AC29, AC9 and 17 were excluded from the determ i-nation of the mean values for the following reasons: AC 9and AC 17 produced a partial antibody molecule whichdid not pass manufacturing criteria, AC17 was sub-clonedf rom AC9 and was also incapable of producing satisfacto-ry antibody. AC29 was excluded due to poor growth. Them a n u f a c t u rer has re p o rted that this cell does not appear tobe producing in static culture and can not be considered ap roductive clone.

S u m m a ry : The CL 1000 proved to be a suitable pro d u c-tion vessel for manufacturing limited amounts of mono-clonal antibody. It provided cost savings through re d u c t i o nof serum use, handling and processing. A concentratedp roduct was obtained in a small volume of culture super-natant when compared to traditional batch cultures. Theapplicability of the CL 1000 was established in over 30d i ff e rent hybridoma cell lines which comprised variousimmunoglobulin isotypes, and which were derived fro mvarious fusion partners.

The protocol was based on a continuous batch pro d u c t i o nmethod which reduced handling and was tailored by themanufacturing concern to its needs. The positive attributesof reduced overall costs, concentrated product and mostsignificantly reduced labor were confirmed in this study.The data provides a range of production results from thec u l t u re of a large number of distinct murine hybridomalines. It should be pointed out that multiple units can beemployed for the same cell lines if increased production isd e s i red. Scale up can be accomplished by operation ofmultiple flasks simultaneously. In conclusion, the CL 1000flasks proved well suited for the small manufacturing labo-r a t o ry and provided cost savings, reduced handling andease of use when compared to prior batch methods used.


Page 4

R e f e rences:

1. Analysis of nutritional factors and physical conditionsa ffecting growth and monoclonal antibody production ofthe Hybridoma KB-26.5 cell line. Sanfeliu A, Cairo JJ,Casas C, Sola C, Godia F. Biotechnol Prog., 12 (2) 209-216, 1996

2. Fractional factorial study of hybridoma behavior. 1.Kinetics of growth and antibody production. Gaertner JG,Dhurjati P, Biotechnol. Prog., 9 (3), 298-308, 1993

3. Role of metabolic waste products in the control of cellp roliferation and antibody production by mouse hybrido-ma cell. Duval D, Demangel C, Miossec S, Geahel I,Hybridoma 11(3); 311-322, 1992

4. Physiology of cultured animal cells. Doverskog M,L j u n g g ren J, Ohman L, Haggstrom L. J. Biotechnol, 59 (1-2); 103-115, 1997

5. Effect of initial cell density on hybridoma gro w t h ,metabolism, monoclonal antibody production. Ozturk SS,Palsson BO, J. Biotechnol. 16 (3-4), 1990.


Tel.: + 1 - 8 0 0 - 8 8 6 - 8 6 7 5

Te l . : +41-1-877 46 46

Te l . : +49-6404 809-0

Te l . : +33-1-34 46 71 71

Te l . : +44-1462 48 65 48

CH

F

D

GB

U S A


Page 5

Clone Ig Harvest mAB mAB Duration Harvests Nutrient Cell Com- 0.1% Medium Maximum MaximumVolume total conc. (days) Medium partment Vitacyte Viable cells Total cells(ml) (mg) (mg/ml) (%FBS) (x 106) (x 106)

AC.1 G1 120 89 0.74 47 5 0.8 10 Y RPMI 616 1664AC.2 G1 65 101 1.55 25 3 0.8 10 Y RPMI 220 576AC.3 G1 128 121 0.94 42 5 0.8 10 Y RPMI 561 1701AC.4 G1 45 33 0.73 22 3 0.8 10 Y RPMI 543 1940AC.5 G1 110 122 1.10 37 4 0.8 10 Y RPMI 570 1476AC.6 G1 100 80 0.80 30 4 0.8 10 Y RPMI n.a. n.a.AC.7 G2a 120 141 1.17 50 6 0.8 10 Y RPMI 731 2080AC.8 G2a 130 158 1.22 42 6 0.8 10 Y RPMI 660 3304AC.9 G1 104 23 0.22 38 5 0.8 10 Y RPMI 682 1957AC.10 G1 110 88 0.80 36 5 0.8 10 Y RPMI 504 1120AC.11 G1 108 95 0.88 35 4 10-5 15-20 Y RPMI 800 1540AC.12 G1 140 109 0.78 48 5 5-0.8 10-20 Y RPMI/ 557 1425

DMEMAC.13 G1 144 71 0.49 28 4 2-0.8 20 N RPMI 312 1172AC.14 G2a 123 44 0.36 39 5 3 20 Y RPMI/ 456 1707

DMEMAC.15 G1 190 165 0.87 39 6 0.8 20 N RPMI 615 1660AC.16 M 160 124 0.78 32 4 2-0.8 20 N RPMI 300 992AC.17 G1 n.d n.d. n.d. 38 4 5 15 N DMEM 682 3000AC.18 G1 130 138 1.10 54 5 5 10-20 Y RPMI 535 1435AC.19 G1 110 243 2.20 32 4 0.8 20 N RPMI 644 1092AC.20 M 135 95 0.70 30 4 0.8 20 N RPMI 394 1160AC.21 G1 180 120 0.70 29 5 0.8 20 N RPMI 402 1608AC.22 G2b 190 105 0.60 38 6 0.8 20 N RPMI 444 1110AC.23 G2b 180 123 0.70 29 5 0.8 20 N RPMI 504 2102AC.24 G1 150 179 1.20 31 5 0.8 20 N RPMI 318 648AC.25 G1 150 209 1.40 32 5 0 20 N RPMI n.a n.a.AC.26 G1 170 163 1.00 38 6 0 20 N RPMI n.a. n.aAC.27 G1 170 163 1.00 36 6 0 20 N RPMI n.a. n.aAC.28 G1 200 152 0.80 42 6 0 20 N RPMI n.a. n.aAC.29 G2a 168 24 0.14 31 5 0 20 N RPMI n.a. n.aAC.30 M 185 169 0.91 37 6 0 20 N RPMI n.a. n.aAC.31 G1 198 135 0.68 39 6 0 20 N RPMI n.a. n.aAC.32 G2b 144 197 1.40 27 5 0 20 N RPMI n.a. n.aAC.33 G1 220 176 0.80 38 7 0 20 N RPMI n.a. n.a

n.a. = not available

Table 1: Cumulative Production Record CL 1000

Table 2

Mean (n=30) Mean (n=30) Mean (n=30) Mean (n=30) Mean (n=30)H a rvest Volume (ml) mAb T o t a l mAb Concentration Duration (days) H a rvests Mean (n=30) ( m g ) ( m g / m l )H a rvest Volume (ml)144 ml 1 3 6 . 3 0 . 9 4 3 6 5s = 4 0 s = 4 7 . 5 s = . 2 6 s = 7 . 5 s = 1 . 0


Page 6

504540353025201510500

500

1000

1500

2000

250022A5

Day

Cel

ls x

10

^6

Figure 1: Cell line 22A5

40353025201510500

500

1000

1500

2000

Day

Cel

ls x

10

^6

24-31

Figure 1: Cell line 24-31

504030201000

1000

2000

3000

total viable

total cells

HA5-6

Day

Cel

ls x

10

^6

40353025201510500

500

1000

1500

2000

Day

Cel

ls

x 1

0^

6

P2W7R

Figure 1: Cell line HA5-6 Figure 1: Cell line P2W7R

F i g u re 1: G rowth curves of re p resentative cultures: Thenumber of viable cells and total cells present in the cellc o m p a rtment of the CL 1000 during culture are shown.Viable cells are plotted with open circles, total cells areplotted with filled circles. The cell compartment was sam-pled on days indicated (usually at harvest) and the cells

counted. Cultures were split back as indicated and cellg rowth can be seen by increased numbers at next harv e s t .C u l t u res were split back by mixing cells with fresh cellc o m p a rtment medium and reinoculating into the cell com-p a rt m e n t .

Abstract: Culture of CHO cells at high density as suspen-sion cells in a protein free medium is described. The pro-duction of a recombinant protein from the cells was com-pared during early adaption phase in the presence of sev-eral different commercially available serum and proteinfree medium. Following adaption and medium evalua-tion, long term continuous culture was maintained by aregular schedule of harvesting and feeding from multipleCELLine CL 1000 bioreactor flasks. Soluble recombinantprotein at high concentration (greater than 40X) wasobtained from multiple CELLine cultures when comparedto traditional monolayer cultures.

Introduction: We evaluated commercial culture mediumfor the ability to support CHO cell proliferation and pro-tein production in high density suspension culture. Screen-ing and medium selection was done in small scale CL 6-well units. Following selection of a suspension mediumthe cultures were scaled up into the larger CL 1000 units(1000 ml) for production.

CHO cells are routinely cultured as a monolayer but havealso been cultured as suspension cells, and specializedmedia are available for this purpose (1-6 ). Other mono-layer forming cell types have also been shown to beadaptable to suspension culture (7-10). A CHO cell linesecreting a recombinant fusion protein construct (rec pro-tein) was evaluated. The CHO cells were previously engi-neered and selected for protein production in traditionalmonolayer culture. The protein construct is proprietary (Bcell activation molecule which binds to cell surface lig-and) and is functional when tested in cell based assaysand flow cytometry . A sandwich ELISA was developed tomeasure this protein in culture supernatant and duringpurification. Affinity purified rec protein was used to gen-erate the standard curve in the ELISA.

The following medium were evaluated: CHO -S-SFM II(Life Technologies, Rockville, MD), CD CHO (Life Tech-nologies, Rockville, MD), SFX-CHO (Hyclone, Logan, UT)

and HyQ PF CHO (Hyclone, Logan, UT). Control mono-layer culture medium was RPMI-1640 supplemented withEXCYTE VLE (1:300) (Bayer, Kankakee, IL), ITS (1ml/liter) (Collaborative Biomedical Products, Bedford,MA) and 0.5% FBS (Hyclone, Logan, UT). All mediumwas further supplemented with antibiotics and L-Gluta-mine (4 mM).

Results: To adapt cells to the various suspension medium,established monolayer cultures (6.0 x 106 cells) in T-25flasks (10 ml) in control monolayer culture medium wereweaned into the various medium formulations by stepwisedilution into the suspension medium. Starting at a 50%suspension medium concentration, cells were switched to75% suspension medium after 3 days. Following an addi-tional 2 days of culture, cells were harvested by trypsinfrom the flasks and reseeded in 100% suspension medi-um into T-25 flasks and into the CL 6-well cell compart-ments. At harvest in 75% suspension medium the follow-ing cell numbers were harvested from the flasks and thecorresponding concentration of recombinant protein wasassayed by ELISA present in the culture supernatant.

Medium Cells/Flask Rec Protein (µg/ml)CHO-S-SFM II 10.3 0.920CD CHO 9.1 0.657SFX-CHO 6.5 0.574HyQ PF CHO 4.8 0.822Control 6.0 0.969

Following transfer to 100% suspension medium and newculture flasks (1.0 x 106/flask), the cells no longerdemonstrated monolayer growth on the surfaces of the T-25 culture flasks. Although a few cells could be seenattached to the flask surfaces, the majority of cells grewin loose aggregates of viable cells. The control mediumcells were in a traditional confluent monolayer. In the con-trol medium, cells still formed a confluent monolayer. Allsuspension medium tested resulted in suspension culturesand the absence of monolayer formation.

CELLine Technical Report VIHigh Density Suspension Culture for Recombinant Protein Production from CHO cells. Martin Wolf Ph.D. • Wilson/Wolf Corp. • Minneapolis • USA

Page 1

Minori Ito

Cells harvested from the various (75% suspension medi-um) flasks were inoculated (2.0 x 106/750 µl) into thecell compartment of CL 6-well units in duplicate. The nutri-ent medium was matched to the cell compartment medi-um for each of the various suspension medium. Following6 days of culture in 100 % suspension medium, the super-natant from the reseeded T-25 flasks and the cell com-partments of the CL 6-well were assayed for recombinantprotein concentration.

Rec Protein Concentration (µg/ml)

Medium T-25 CL 6-wellCHO-S-SFM II 0.43 9.32CD CHO 0.38 2.91SFX-CHO 0.30 3.11HyQ PF CHO 0.43 >10Control 1.67 3.19

On day 9 of culture, cell compartment medium wasremoved and replaced with fresh medium leaving thecells in the cell compartment. Cells were left in the cellcompartment to maximize cell numbers. The nutrientmedium was replaced. At day 13 the cell compartmentswere harvested by mixing and removal of 400 µl for cellcounting and ELISA assay of supernatant. Nutrient medi-um was replaced.

Medium Viable Cells Rec Protein (µg/ml)x 106/ml

CHO-S-SFM II 4.49 17.33CD CHO 0.14 1.01SFX-CHO 1.08 10.58HyQ PF CHO 10.44 22.65Control 1.58 7.10

On day 15 of culture the cell compartment contents wereremoved from the CL 6-well and again analyzed for cellnumbers and protein concentrations. At this time two ofthe suspension medium were clearly supporting more cellgrowth as determined by cell counts and visible appear-ance of the cell mass in the cell compartment and the sizeof cell pellet following centrifugation of the cell compart-ment suspension (CHO-S-SFM II and HyQ PF CHO).


CHO-S-SFM II 4.39 >50CD CHO 0.14 1.53SFX-CHO 2.50 8.05HyQ PF CHO 6.82 17.06Control 1.21 4.22

Control cultures (monolayer medium) in the CL 6-well con-tained viable cells however no increase in cell numberscould be seen. Cells did not form monolayers in the con-trol medium in the cell compartment or in any of the othermedium. The contents of the cell compartment from theCHO-S-SFM II and HyQ PF CHO cultures were reseeded(1.1 and 1.51 x 106/ml respectively ) into CL 6-well unitsin replicates of 6 cell compartments to further assess cellproduction with the two different medium types.

Following 4 days of culture a 50 µl sample was taken fromeach replicate well and cell counts and protein concentra-tion determined. At this time considerable cell mass wasevident in both medium types. Cells were present as aggre-gates within the cell compartment. The aggregates con-tained viable cells as determined by trypan blue staining.Although the majority of cell aggregates could be dissoci-ated by pipetting with a P200 pipette tip, significant num-bers of cells were still present as multicellular aggregatescontaining predominantly viable cells. This made accurateassessment of cell numbers difficult due to presence ofaggregates still remaining. Results for the mean and stan-dard error for the replicate wells are shown below.


CHO-S-SFM II 5.17 ± 1.09 32.17 ± 2.47HyQ PF CHO 3.46 ±1.16 56.29 ± 11.17

After an additional 3 days in culture the cell compartmentcontents were removed entirely and cell counts and pro-tein concentrations determined. The entire cell compart-ment volume was removed by pipetting to suspend cellsfollowed by an additional rinse of 500 µl medium torecover as many cells as possible from the cell compart-ment. Cell counts were determined for each cell compart-ment and the individual supernatants were pooled forELISA determination of protein levels.


CHO-S-SFM II 11.95 ± 1.08 44.38 HyQ PF CHO 12.47 ± 0.67 80.86

CELLine Technical Report VI • High Density SuspensionCulture for Recombinant Protein Production from CHO cells.

Page 2

At this time cell growth in both medium was marked in thecell compartments of the CL 6-well. Although greater con-centrations of rec protein were obtained from the HyQ PFCHO cultures both medium appeared to support cellgrowth and production.

Assessment of a nutrient medium substitute with an alter-native less expensive medium was done in cultures con-tinued in the presence of alternative nutrient mediumchoices as well as the control medium (cell compartmentmedium). The alternative nutrient medium choices werebased on the basal medium from which the suspensionmedium was derived. For the CHO-S-SFM II medium, a S-MEM (Life Technologies, Rockville, MD) was selected andfor the HyQ PF CHO medium, RPMI-1640 (Hyclone,Logan, UT). The nutrient medium was used without anysupplementation or in the presence of 2% FBS. The con-trol monolayer medium was also evaluated as a nutrientmedium as it maintained cell viability although in theabsence of cell proliferation..

Cells were reinoculated into the cell compartments of theCL 6-well at 2.4 x 106 cells/ml in duplicate with the var-ious nutrient medium choices and the control medium. Byday 4 visual inspection revealed differences between thevarious nutrient medium cultures. Cell mass was greatestin the control medium cultures (cell compartment = nutri-ent compartment). The next best condition appeared tobe the control monolayer medium. The cultures were har-vested after 7 days and the cell counts and protein con-centrations determined for the various culture conditionsin duplicate.

Cell Compartment-HyQ PF CHO

Medium Viable Cells Rec Protein x 106/ml (µg/ml)

RPMI-1640 4.08 19.31RPMI-1640 2% FBS 7.42 22.32Monolayer Medium 9.94 26.89S-MEM 1.50 20.92S-MEM 2% FBS 0.44 26.24HyQ PF CHO 19.14 46.21

Cell Compartment-CHO-S-SFM II

Medium Viable Cells Rec Protein x 106/ml (µg/ml)

RPMI-1640 1.96 15.41RPMI-1640 2% FBS 2.32 24.13Monolayer Medium 3.10 15.13S-MEM 0.76 11.42S-MEM 2% FBS 0.82 17.13CHO-S-SFM II 8.18 19.65

At this time the substitution of an alternative nutrient medi-um to replace the more expensive cell compartment medi-um was not shown to be as effective as using the samemedium in both the cell compartment and the nutrientmedium reservoir. Although continued evaluation of smallmolecular weight supplements to a basal mediumappears promising in that the presence of ITS and Excytepresent in the monolayer medium appeared to be benefi-cial when compared to RPMI-1640 with or without 2%FBS. The poor performance of the S-MEM as an alterna-tive nutrient medium is not understood at this time. HyQPF CHO medium was chosen as the medium for produc-tion due to its protein free composition and increased pro-duction levels when compared to CHO-S-SFM II.

In parallel to evaluation of alternative nutrient medium,large cultures were established in CL 1000 units for pro-duction purposes. The cultures were maintained on atwice weekly handling protocol in which the cell com-partment was harvested and the nutrient reservoir medi-um exchanged. At harvest the nutrient medium waspoured from the bioreactors and the cell compartmentcontents mixed by gentle sliding of the bioreactors in aswirling motion on the work surface. This allowed nearcomplete removal of the cell compartment contents bypipetting. If mixing was not done it was difficult to removethe cell contents entirely by pipetting with visible cell massremaining in the cell compartment. After removal of thecell compartment contents, cells were centrifuged and thesupernatant removed and saved for harvest. Roughly 1/2of the cell pellet (2.5 ml) was resuspended in fresh medi-um (20 ml) and returned to the bioreactors. The nutrientmedium was replaced at the same time. Cell pelletsapproaching 5 ml in volume were routinely harvestedfrom the CL 1000 bioreactors.


Page 3


Page 4

Representative harvests from a 4 day culture period fromthe cell compartment of 4 individual bioreactors areshown below. The cultures were maintained for greaterthan 30 days and production levels were consistentthroughout the culture duration.

CL 1000 Viable Cells Harvest Rec Protein x 106/ml Volume (µg/ml)

No. 1 5.15 15 39.31No. 2 7.85 15 44.65No. 3 10.75 18 67.93No. 4 13.10 20 57.05

Conclusion: CHO cells grew and produced when culturedas suspension cells. Although the original clone wasselected from adherent monolayer forming cultures, theclone still produced at acceptable levels when cultured insuspension medium and in the absence of monolayer for-mation. The clone was easily maintained in the CELLineflasks after a brief adaption period. This resulted in amuch easier production method than standard monolayerpropagation involving trypsinization and continued pas-saging on tissue culture plastic. Additionally, the produc-tion of a concentrated supernatant reduced downstreamprocessing issues and allowed for use of supernatantdirectly for certain applications.

Of the commercial medium tested, both a serum free anda protein free medium appeared to be suitable for CHOcultivation in suspension. The protein free medium waspursued due to its protein free characteristics and slightlybetter production values. There is no explanation at thistime as to why the other medium types tested were not assuccessful. This may be due to differences in adaptiontimes, cell line specific sensitivities or other unknown rea-sons.

After 6 days of culture, the CHO line at confluence pro-duced a concentration of 1.67 µg/ml in a 10 ml culturevolume in a standard T-25 flask. In contrast much higherconcentrations were achieved in CELLine cultures. Amean of 52.23 ± 6.42 µg/ml was obtained in 4CL 1000 units after 4 days of culture. This represents a32 fold increase in product concentration in a smallersupernatant volume. A single CL 1000 flask in a 4 dayperiod produced equivalent to a 6 day (544 ml) mono-layer culture. This was achieved with minimal handling,in a single device and resulted in a significant reductionin supernatant volume for down stream processing.

Alternative medium choices for the nutrient medium maybe possible with further evaluation of the role of smallmolecular weight components present in the nutrientmedium. Although the addition of FBS to the RPMI-1640nutrient medium appeared somewhat beneficial whencompared to the absence of any FBS supplement, thepresence of ITS and Excyte VLE appeared to have a morebeneficial effect in the presence of only 0.5% FBS. Furtherevaluation of small molecular weight components on therole of cell proliferation and survival in the CELLine arewarranted and may provide further significant cost sav-ings by reducing the consumption of the more expensivecomplete medium. The dilution of the complete mediumwith a basal medium may prove to be an acceptablealternative, as it appears likely that dilution of a compo-nent across the semi permeable membrane reduces theeffectiveness of the complete commercial medium formu-lations when used only in the cell compartment. If a smallmolecular weight component which is essential to the suc-cess of the culture is present in the commercial formulas,it may be present at excess concentrations and be capa-ble of being diluted with less expensive basal mediumtypes and used in the nutrient reservoir. Alternatively, sup-plementation of a basal medium with specific compo-nents may duplicate the effects achieved with the com-plete medium for use as a nutrient medium.

References1. Production of recombinant proteins in Chinese hamsterovary cells using a protein-free cell culture medium. ZangM, Trautmann H, Gandor C, Messi F, Asselbergs F, LeistC, Fiechter A, Reiser J Biotechnology (N Y) 1995Apr;13(4):389-392 Institute of Biotechnology, Swiss Fed-eral Institute of Technology, ETH-Honggerberg, Zurich,Switzerland.

2. High-level expression of secreted proteins from cellsadapted to serum-free suspension culture. Berg DT,McClure DB, Grinnell BW Biotechniques 1993Jun;14(6):972-978 Lilly Research Laboratories, Lilly Cor-porate Center, Indianapolis, IN 46285-0424.

3. Production of a membrane-bound protein, the humangamma-glutamyl transferase, by CHO cells cultivated onmicrocarriers, in aggregates and in suspension. ChevalotI, Visvikis A, Nabet P, Engasser JM, Marc A Cytotechnol-ogy 1994;16(2):121-129 Laboratoire des Sciences duGenie Chimique, UPR CNRS 6811, Nancy.


Page 5

4. Influence of ammonium on growth, metabolism, andproductivity of a continuous suspension Chinese hamsterovary cell culture.Hansen HA, Emborg C Biotechnol Prog1994 Jan;10(1):121-124 Department of Biotechnology,Technical University of Denmark, Lyngby.

5. Scaled-up production of recombinant human renin inCHO cells for enzymatic and X-ray structureanalysis.Asselbergs FA, Rahuel J, Cumin F, Leist C JBiotechnol 1994 Feb 14;32(2):191-202 BiotechnologyDepartment, CIBA-GEIGY Ltd., Basle, Switzerland.

6. CHO transfectants produce large amounts of recom-binant protein in suspension culture.Schutt C, Furll B, Stel-ter F, Jack RS, Witt S Immunol Methods 1997 May12;204(1):99-102Institute of Immunology and Transfusion Medicine, Ernst-Moritz-Arndt-University of Greifswald, Germany.

7. Structure and function in rhodopsin: high level expres-sion of a synthetic bovine opsin gene and its mutants instable mammalian cell lines.Reeves PJ, Thurmond RL, Kho-rana HG Proc Natl Acad Sci U S A 1996 Oct15;93(21):11487-11492 Department of Biology, Mass-achusetts Institute of Technology, Cambridge 02139,USA.

8. Rapid generation of stable cell lines expressing corti-cotropin-releasing hormone receptor for drug discov-ery.Horlick RA, Sperle K, Breth LA, Reid CC, Shen ES,Robbins AK, Cooke GM, Largent BL Protein Expr Purif1997 Apr;9(3):301-308 Applied Biotechnology Depart-ment, DuPont Merck Pharmaceutical Co., Wilmington,Delaware 19880-0400, USA.

9. Problems with BHK 21 cells.Brown F Dev Biol Stand1998;93:85-8 Plum Island Animal Disease Center,Greenport, NY 11944, USA.

10. An experimental rabies vaccine produced with a newBHK-21 suspension cell culture process: use of serum-freemedium and perfusion-reactor system.Perrin P, Madhusu-dana S, Gontier-Jallet C, Petres S, Tordo N, Merten OWVaccine 1995 Sep;13(13):1244-50 Laboratoire desLyssavirus, Institut Pasteur, Paris, France.

E-Mail: [email protected]

Tel.: +1-800-886-8675

Tel.: +41-1-877 46 46

Tel.: +49-6404 809-0

Tel.: +33-1-34 46 71 71

Tel.: +44-1462 48 65 48

D

CH

F

UK

USA


Trou

ble

shoo

ting

and

FAQ 4 Trouble shooting and FAQ

1. Why is it that removal or addition of cell compartment volume is slow? Why does cellcompartment volume come back out of port when I remove the pipette?

Be sure that the medium compartment cap is loosened during manipulation of the cell com-partment volume. Changes in cell compartment volume create pressure in the nutrient com-partment if the cap is not loosened. Tighten cap after cell compartment manipulation is com-plete.

2. When I pour medium from the CELLine, I sometimes have a drop of medium left on theoutside of the neck, what can I do to stop this from happening?

When pouring medium from the CELLine it is recommended that the flask be held with thebottom of the flask in the palm. This provides adequate neck pouring angle and preventsaccumulation of the medium on lip of neck after pouring. The angle of the neck on theCELLine is required to prevent air locks in the nutrient reservoir and thus pouring is bestdone with flask upside down in palm of hand.

3. After incubation, I notice that the outside of the CELLine is wet, why is this?

If medium is placed in flask that is not pre-warmed, there will be considerable condensa-tion accumulation on the outside of the CELLine. Due to the large volume of medium con-tained in the CELLine, this condensation can be significant. The condensate takes time toevaporate in the humidified incubator. Test color of liquid by blotting with white paper, ifthere is no coloration, the liquid is water and due to condensate.

4. Can I place more than the recommended volumes into the cell compartment?

The protocol recommends a working volume of 5 ml (CL 350), and 15 ml (CL or AD 1000)for the CELLine products. This assures that volume in the cell compartment never exceedsbursting threshold for the membrane even with osmotic flux of water into the cell compart-ment over an extended period. The membrane is fragile but compliant and will distend sig-nificant distance when wet. Increased cell compartment volumes up to 1.3 times above rec-ommendation are not problematic.

Minori Ito


Trou

ble

shoo

ting

and

FAQ5. Will I be able to recover all of my cells from the cell compartment?

When working with cells growing in suspension in CELLine classic, the recovery of cellsfrom the cell compartment should be nearly 100%. Suspension cell types have not beenobserved to form aggregates and are readily disbursed with gentle pipetting. Followingpipetting the cells are easily recovered. An additional rinse of clean medium may be usedto further assure complete cell recovery if required.Other cells types may form cellular aggregates or attach to the bottom of the cell compart-ment. In these cases, cell recovery may require the use of a dissociating agent to separatecells and to aid in their recovery.

When working with anchorage-dependent cells in CELLine adhere, cell are attached to thePET ilay matrix. Depending on the individual cell type, sometimes recovery of the cells canbe achived mechniaclly by pipetting up and down, but in most of the cases a completerecovery involves the addition of a dissociating agent to the cell compartment.

6. Can I change nutrient medium by tracking color change of the nutrient medium as I doin my other cultures?

It is not recommended. Cell growth is dependent on the diffusion gradients present betweenthe cell compartment and nutrient medium compartment. The greater the gradient, the larg-er the flux of soluble metabolic substrates that is available for cell metabolism. The mediumcolor change is not an accurate assessment of nutrient and waste status in the CELLine dueto the ability of the cell compartment to balance pH directly with the incubator atmosphere.Nutrient medium will become more yellowish during culture but will not take on the char-acteristic color associated with spent medium in traditional flasks. Careful tracking of cellnumbers within the cell compartment and medium exchange rate can be done to determineoptimum conditions for your cell type.

7. When I harvest from the cell compartment, I always have a greater volume than what Iinoculated why is this happening?

Osmotic gradients across the upper semi-permeable membrane will drive water through themembrane. If a protein gradient is present across the semi-permeable membrane such aswhen no serum is used in the nutrient compartment, water is driven into the cell compart-ment. Because small solutes will move across the membrane also, this change in volumeonly affects colloid protein concentrations.

This is the reason for the recommended use of 15% serum in the cell compartment when noserum is used in the nutrient medium, as it assures that serum concentrations with in the cellcompartment do not become excessively diluted with continued culture.


Trou

ble

shoo

ting

and

FAQ 8. The cell compartment volume in my CELLine is less than what I inoculated, where is the

volume going?

If the CELLine is used in a non humidified incubator or warm room, evaporative losses fromthe cell compartment can lead to reduced volumes. The CELLine is intended to be used ina standard humidified incubator. If required it may be worth evaluating restricting the bot-tom gas vents to reduce evaporation. This should be done experimentally to determine thebalance between evaporative loss and adequate gas exchange.

9. How much nutrient medium can I place in the nutrient reservoir?

The maximum capacity of the reservoir is marked on the sides of the devices. For the CL350 (350 ml) and for the CL or AD 1000 (1000 ml) are the maximum volumes. Do notexceed these volumes as the design requires that there be an air passage to the green cap.If excess medium is placed in the devices, an air lock can be created within the reservoircompartment.

10. I notice that the distribution of cells in the cell compartment sometimes is not even acrossthe bottom of the cell compartment, should I mix the cell compartment to provide a moreeven distribution?

This is not necessary, experiments which involved re-suspending cells in the cell compart-ment did not lead to increased cell numbers or antibody production. An excessive accu-mulation of cells in one area of the CELLine due to a non-level incubator should however bedisbursed and allowed to resettle.

11. I have followed the protocol and my cells are not growing, what is wrong?

Check to be sure that the protocol has been adhered to. The condition of the cells prior toinoculation is critical. Cells should be taken from logarithmic growth and inoculated at therequired cell numbers. Be sure the minimum cell numbers for inoculation are present.Absence of hybridoma cell growth has been seen under certain conditions. Poor growthhas been associated with mycoplasm contamination. Mycoplasm accumulates within thecell compartment and may exert effects not seen in static culture flasks. Treat cells to erad-icate the mycoplasm and robust cell growth should be established. There may be some cells that are not capable of growth in the absence of serum compo-nents which can diffuse across the upper semi-permeable membrane. Increasing serum con-centrations in the nutrient medium and or cell compartment can be evaluated to determineif this is required. Finally, if insufficient numbers of cells are inoculated, there may be a sig-nificant lag phase prior to cell proliferation. It is recommended that a new inoculum withhigher numbers of cells be placed in the cell compartment and results evaluated. If cells require a conditioning factor that diffuses across the membrane it may be necessaryto start with a small volume in the nutrient medium compartment until cell numbers increase.When cell growth is established adequate nutrient medium must be added to supply theincreasing cell mass.


Trou

ble

shoo

ting

and

FAQ

15. How strong is the semi-permeable membrane?

The upper semi-permeable membrane is only 8 µm thick (dry). The membrane is delicatebut easily withstands normal handling. Shaking or banging of flask against hand or sur-face can lead to membrane failure if the cell compartment has liquid in it. This places sig-nificant stress on the membrane due to the rapid displacement of the liquid present in thecell compartment.

12. Will the CELLine function in a 7.5% CO environment?

Yes, cultures in the cell line will be under the same CO2 tensions as in static flasks. Use ofa medium formulated for use in 7.5% CO2 is required.

13. I can not view cells in the CELLine using my inverted microscope, what can I do to beable to view them?

Due to to PET matrix inlay in CELLine adhere cells can not be viewed under a microscope.

For CELLine classic bioreactors, the microscope objective must travel above the stage of themicroscope to allow for viewing into the cell compartment. If you have a mechanical artic-ulated stage or other attachments on the microscope, it may be required to remove them.Alternatively, the objective can be un-threaded from the nose piece several turns to allowsufficient travel up past the plane of the stage. Additionally, the working distance of theobjective must be sufficient, most 10X and some 20X objectives are suitable. The distancefrom objective to CELLine is 2.5 mm both for the classic 350 and 1000 reactors.

14. I have viewed my cells in the CELLine classic but when I tried to view them again afterseveral days of culture I was not able to focus on them, what has changed?

When the CELLine is removed from the incubator and the temperature of air in the CELLineis cooled slightly, contraction of the air takes place and will draw the membranes of thecell compartment up into the device. This contraction lifts the bottom membrane and takesit out of focusing distance. The loosening of the medium compartment cap will equilibratepressure and return membrane to original position.

16. Why is it recommended to wet the membrane before placing cells into the cell com-partment?

It is important to wet the membrane prior to inoculation to assure that the membrane is com-pliant. The wet membrane is compliant and capable of distension. The dry membrane ismore susceptible to tensile stress due to volume changes. The air trapped in the cell com-partment can not be removed until the membrane is wet and liquid is added into the cellcompartment. The dry membrane is stressed significantly if volume is added directly into thecell compartment prior to wetting of membrane.


Trou

ble

shoo

ting

and

FAQ 17. Does the semi-permeable membrane become clogged with use?

Performance of the CELLine devices does not decrease with time of culture. This indicatesthat solute transfer across the membrane does not decrease significantly during culture andthere is thus no significant clogging or fouling of the membrane.

18. Are there different membranes available for the CELLine?

Currently, only a 10,000 MWCO membrane is available.

19. Do you have any tips on handling which will reduce the risk of contamination?

Every time the culture is handled the sterile field is broken to add and remove cells andmedium. The medium is easiest removed using a Vacuum system like the INTEGRAVACUSAFE. In case the medium is simply poured out, drops on the neck of the bottlesshould be remove using a sterile Pasteur pipette and should not be wiped off with alcohol.Its recommended to perform all working steps with CELLine in a Class II biosafty cabinet.Cleaning of working surfaces with alcohol can provide additional protection against con-tamination risk. Sterile alcohol wipes or spray alcohol is routinely used to reduce contami-nation risk in many laboratories. There has been no indication that this produces adverseeffects on the device itself or on cell growth in the devices.

20. Are there any special storage conditions required for unopened CELLine products?

The devices are packaged in a sterile barrier blister within a foil vapor barrier pouch. Greatcare has been taken to provide as robust a package as possible. The devices can be storedunder ambient conditions with no demonstrated deterioration in performance. Care shouldbe taken to prevent the devices from being exposed to high temperature to prevent dimen-sional changes in the membrane and excessive tensile stress. It is recommended thatdevices be stored at room temperature.

21. Will my hybridoma stop secreting after prolonged culture in the CELLine?

We have no data that indicates selection of non-secreting clones takes place during culturein the CELLine. Even for low antibody secreting cells, no evidence has been found indi-cating that selection for non-secreting cells takes place. Importantly, a low secretinghybridoma will not revert to a high secretor when cultured in the CELLine. However, inmany cases, a low secretor can be tolerated due to the increase in product concentrationachieved with the CELLine. In many instances, cells which have very low production canbe salvaged by culture in the CELLine allowing useable amounts of product to be recov-ered without excessive amounts of culture supernatant processing.


Trou

ble

shoo

ting

and

FAQ22. Is the antibody produced in the CELLine classic equivalent to antibody produced in

static culture?

Analysis by flow cytometry indicates that antibody produced in the CELLine yields equiva-lent binding per mg (fluorescent profiles) when compared to control antibody cultured instatic culture flasks recovered without excessive amounts of culture supernatant processing.

23. How many cells can be cultured in the CELLine?

For a typical murine hybridoma the viable cell concentrations reached in CELLine arearound 2 - 3 x 107 cells/ml A fundamental principle of the CELLine is the cell capacity ofthe devices. If adequate nutrient medium exchange is provided, cell proliferation will con-tinue within the cell compartment even when maximum viable cell capacity has beenreached. This can result in very large numbers of total cells within the device. For the pro-duction of antibody total cell accumulation is not problematic.

24. Can I culture leukemic cell lines in the CELLine?

Lymphoblastic cells grow very well in the CELLine classic. Cell concentrations of certain lym-phoblastic cells can reach nearly twice that achieved for hybridoma cells. Some cell linesmay be dependent upon the use of serum on both sides of the semi-permeable membraneand this can be readily examined.

25. Do you recommend the use of a high glucose containing medium in the CELLine?

Most protocols were developed using standard RPMI-1640 medium. Some customers andothers who use hollow fiber bioreactors do use richer mediums. A slight performanceincrease may be obtained with richer media, however, this is dependent upon cell line andshould be evaluated experimentally. In general excellent results are obtained in mediumwhich is currently used to culture the cell line in static flasks.

26. Will serum free medium work in the CELLine?

Yes many customers report excellent results using serum free medium. The serum free medi-um is placed on both sides of the semi-permeable membrane in most cases. Importantly,the use of serum free medium may no longer be necessary when the CELLine is used. Asthe secreted protein is recovered at high concentration from the CELLine, it is no longernecessary to concentrate culture supernatant to recover antibody. This eliminates much ofthe interference associated with serum protein during purification. At antibody concentra-tions in excess of mg/ml, cell compartment supernatant can be applied directly to an affin-ity column.

Documents

CELLine - Cosmo Bio Co Ltdsearch.cosmobio.co.jp/cosmo_search_p/search_gate2/docs/... · 2006. 8. 18. · Obtain 25 x 106 viable cells from a pre-culture in log growth phase and suspend