CA1316853C - Growth of cells in a magnetically stabilized fluidized bed - Google Patents

Abstract

ABSTRACT

Growth of Cells in a Magnetically Stabilized Fluidized Bed The invention discloses methods of growing cells in a magnetically stabilized fluidized bed. Cells and magnetic particles are entrapped in a material, preferably a polysaccharide gel, and then cultured in a magnetically stabilized fluidized bed where nutrients and oxygen are supplied to the cells.

Description

13~853 Growth of Cells in a Magnetically Stabilized Fluidized Bed Field of the Invention This invention is directed to methods of growing cells.
More particularly this invention is directed to methods of growing plant and animal cells.
Background o~ the Invention Plants are the source of over 80% of all known natural products. Plants provide not only food, but also drugs, flavors, ~ragrances and coloring agents. Production of these commercially important cellular metabolites is normally by cultivation of the plants followed by extraction of the desired compound. However, this process is labor-intensive and is associated with many limitations including disease, drought and sporadic availability of plant sources. Since many of thes~ compounds are complex molecules which cannot be chemically synthesized, alternative means for production such as cell culture have been eagerly sought.
Application of techniques for the culture of microorganisms which have been known for many years to the culture of plant or animal cells has had limited success due to unique features of the plant cells themselves. For example, plant cells have much slower growth rates than microorganisms (with a ~' ~;
, . . ~ ~ , . .
. . ~

UPN-69 13 ~ 3 doubling time of days vs. hours for the latter) which makes them vulnerable to overgrowth by con~aminating microorganisms when left for long periods of time in reactore. Typical microorganism experiments performed in batch reactors run for one day as opposed to 1 to 3 months for plant cell experiments. One day is usually too short a time for a contaminant organism to have a signi~icant detrimental effect. Therefore, a plant cell reactor must be designed with as few potential sites for contamination as possible.
Plant cells are somewhat sensitive to shear stress because of their large size (compared to microorganisms) and their rigid walls. Reaction vessels which are agitated mechanically to aerate the cells produce currents within the growth medium and the resultant flow can break open the plant cells. Thus any possible reactor design must minimize the amount of shear stress to which the cells are subjected.
Animal cells, particularly hybridomas and genetically engineered mammalian cells, are becoming important sources of antibodies, hormones and other important products. These cells are even more sensitive to ehear than are plant cells~
Competing with the need for minimal shear stress is the need for adequate oxygenation of the cell cultures. Since plant cells in particular are typically grown in thick suspensions (up to 60g/l biomass), a low shear environment makes it difficult to maintain adequate oxygen transfer rates which are necessary for :: :

:;

. ~ :

~ 3 ~ 3 good cell growth. These two opposing constraints, high oxygen transfer rates and low shear must be optimized in any plant or animal cell reactor design~
Other differences between eukaxyotic (plant and animal) cells and microorganisms make growing eukaryotic cells in reactors developed for microorganisms difficult. Unlike the growth of microorganisms, growth of these cells is regulated by hormones.
These growth regulators can be excreted into the environment (of the reactor) by some cells and generally affect the growth of all cells. This process "conditions~ the medium. Since plant cells require a certain critical minimum concentration of growth regulators in order to proliferate in continuous flow fermentors, there exists a maximum dilution rate beyond which the cells cease to grow. Likewise in batch culture, there is a minimal cell inoculum density below which no cell growth occurs. This minimum cell density ensures that adequate quantities of growth factors are excreted into the medium.
Additionally, plant cells tend to grow in aggregates, whereas microorganisms frequently do not. These aggregates are known to aid in product formation. However, the exact nature of this process is not clearly understood. Typically, the rate of product formation increases as the cell aggregate increases in size~ Growth of animal cells is somewhat different because growth regulators normally are supplied as part of the nutrient medium.
Despite the drawbacks, the most commonly used .

;~ :

alternative process for the cultivation of plant or animal cells is the batch reactor. Batch cultivation is characterized by a constantly changing environment where the cells are inoculated into a medium in a relatively dormant state (lag phase) and are allowed to grow until they reach stationary phase, where the cells have reached their maximum possible cell population with the quantity of nutrients present. Between these two phases there is a rapid period of growth called the growth or logarithmic phase.
The length of this phase is controlled by varying the concentration of a limiting nutrient in the medium.
Suspension culture of plant cells in a batch reactor is the method most commonly used today to study plant growth kinetics. This simple system consists of growing the cells in an Erlenmeyer flask agitated on a platform shaker. Animal cells are grown in roller bottles and other specialized devices.
The first large scale batch culture system for plant cells was developed by Tulecke and Nickell, N.Y. Acad. Sci. 22:
196-206, (1960). Their system was a very simplistic desiqn consisting of a 20 liter carboy with inlets for medium, the inoculum, and air and an outlet for air. More sophisticated batch culture systems have been developed using forced aeration with magnetic stirring. All of these systems are limited by the shear stress which can be sustained by the plant cell walls, yet adequate mixing must be maintained, particularly at high cell densities. The maximum permissible plant cell mechanical : ::

:
' . ` '' agitation rates vary from 50 to 200 rpm depending on the cell line in question.
Kato et al., J. Ferment. Technol. 53: 744, (1975). used bubble-type reactors with sparged air entering the bottom of the agitator for agitation. Although this system is advantageous in that it provides a lower shear environment than a mechanically agitated reactor, adequate mixing cannot be provided in many cases. Wagner and Vogelman, in Plant Tissue Culture and Its Biotechnical Appl., Barz, Reinhard and Zenk, eds., Springer-Verlag, Berlin (1977~, studied a variety of original reactor typesincluding a draft tube with turbine and an air lift fermentor.
The draft tube reactor uses a turbine to force liquid up through the annular space between the draft tube and reactor wall which then flows into the center of the reactor. This design has disadvantages in that shear is very high near the turbine. The air lift fermentor is probably the best overall reactor for plant cell cultivation to date. In this system, agitation is accomplished by air rising in the draft tube causing the liquid to lift upwards. Good mixing can be achieved at low shear.
Although batch systems are convenient to operate, they have many limitations. Since the environment is continually varying with time and cells from past phases have affected the environment of present phases, it is difficult to examine cellular physiology. Also, it is not possible to maintain the system at the maximum product production rate for an indefinite period o~

.. . . . . .

., : .
: . :
- , . :
..
:: . : , : . . . '. ~ - -time, thus maximizing the potential productivity of the cells.
Finally, batch processes are not convenient for industrial scale operations since starting and harvesting a batch culture is very labor intensive. This leads to less efficient downstream processing and higher costs.
Continuous culture systems, an alternative to batch reactors, are designed so that the cells grow at a constant rate under environmental conditions which do not change with time.
There are several advantages associated with continuous culture systems. Steady states in growth rates, metabolism, and concentration of nutrients in the medium enable one to study growth and metabolism under controlled conditions. Also, the effects of a single growth-limiting nutrient on cell growth can be determined. Since the growth rate can be controlled, the system theoretically can be run indefinitely at a maximum productivity, unlike the situation which exists in a batch system. However, the continuous addition and withdrawal of media results in dilution of the product which makes product recovery difficult. Also, dilution rates are limited by complete washout of the cells or los3 of growth regulators. Washout occurs when the dilution rate exceeds the maximum cell growth rate and all cells are lost from the reactor. This is a particularly important problem for plant cells which grow at very low rates (with a doubling time of several days).
Application of techniques for enzyme and microbial , ', , - ' .
.
- :
.
, 8 ~ 3 immobilization to the growth of plant cells has had more promising results. The method of entrapment in beadæ or blocks of gel using alginate, agarose, agar or carrageenan has proven to be capable of growing cells e~Eiciently while maintaining good cell viability.
Immobilization of cells in a gel apparently mimics the natural condition of the cell within an aggregate of cells. Alginate is the most popular support material due to the ease with which beads may be prepared and the mild conditions required for gel formation.
Few investigations have been carried out using immobilized cell bioreactors. Brodelius et al., FEBS Letters l03:
93-97, (1979), studied the potential of batch and packed bed reactors. Continuous flow packed bed and fluidized bed reactors have also received limited analysis. Packed bed reaators have several disadvantageæ. In upflow, the flow rates are limited by the pressure drop. This low flow velocity often leads to problems with oxygen trans~er, control of temperature and pH, and removal of gaseous products. In downflow, the pressure drop adds to the weight of th~ bed. This compresses the beads and can choke fluid flow. Also, packed bed reactors can filter free cells and degraded support particles from the liquid stream causing blockage of flow.
In addition to immobilization in gel supportsj plant cells have also been immobilized within membranes and in vessels bounded by membranes. The most csmmon systems of this type are , . .
.
., ~ . - . : .
.
., . -:. . . . . . . , - .

:. ' ,. " ' . ~ . -:

hollow fiber and flat plate reactors. The choice of membrane is vsry critical since iks pore size controls the reactor performance. Also, there are mechanical constraints on the thickness of the membrane since it must be durable enough to with stand weeks of operation. However, it is also necessary that the membrane be thin enough to allow adequate mass trans~er of nutrients to the cell-. At present the major problem associated with membrane reactors is scale-up, since large membranes are costly.
In the case of animal cells, the most success~ul immobilization methods appear to those in which a membrane is used to isolate the cells, suspended in liquid, from a flowing liquid nutrient stream. Hollvw fiber bundles and polymeric bubbles have been used successfully for this purpose.
The use of magnetic fields to stabilize fluidized particles was first reported by Herschler in U.S Patents 3,219,318 and 3,439,89g. Herschler studied the effect of magnetic fields on liquid metals and magnetically susceptible solids that had been fluidized. Using an alternating current, he was able to decrease bubbling and prevent solids entrainment.
The first systematic studies of magnetically stabilized fluidized beds (MSFB's) were conducted in the 1970's at the Exxon Corporation by Rosensweig and his colleagues (Ind. Eng. Chem.
Fund. 18: 260-269, (1979), Science 204: 57-60, (1979), AIChE
25 Symposium Serie~ 77: 8-16, (1981)). They were able to stabillze ..... .
' ' .
., ~ .

., - ~. , :
- -UPN-69 ~- 3 ~ 3 fluidized magnetically susceptible solids in a uniform magnetic field. This stabilization phenomenon due to the magnetic field effect occurs when a magnetic dipole is created in the particles, causing them to align in a direction parallel to the field lines and attract one another. Stabilization of a fluidized bed has been defined as the elimination of mixing, bubbling and slugging.
Since the chainlike alignment of particles does eliminate these problems, it was possible with the MSFB to maintain high fluid flow rates and good mass transfer while eliminating particle collisions and washout from the bed. The MSFB is unique in that it retains some properties of both a fixed bed and a fluidized bed. It behaves li~e a fixed bed in that the dispersion, or range of contact times for fluid elements, is much more limited than in a fluidized bed. At the same time, the particulate phase undergoes almost no back-mixing. As in the fluidized bed, the pressure drop across the bed is low and the particles flow in a fluid-like fashion over a wide range of conditions. In the MSFB, the particles can be continually removed from the bed by flowing through an orifice in the fluid distributor at the bottom of the bed. As the lowest layer of particles flows out of the bed, the upper layers of particles move down in plug flow. With the continual addition of particles at the top and entry of fluid from the bottom, a continuous counter-current contacting scheme between the solid and fluid phases can be established. This cannot be accomplished either with a typical fixed or a fluidized bed. The _g_ .

:, - . - - . . :, . ~, , ... ~- ~ . . , -:
-. .

' :, ~

~ 3 ~

Exxon researchers were concerned mainly with gas-solids contacting.
Graves and Burns, U.S. Patent 4,675,113 used a MSFB
having a solid support of alginate beads with a magnetic core to conduct affinity chromatography separations of biomaterials. A
biomaterial is separated from a solution containing the biomaterial by passing the solution through the MSFB and contacting it with the beads which have been modified to have a ligand on the surface which will bind the biomaterial. The biomaterial binds to the bead and remains behind. ~ater the beads are r moved from the MSFB and the biomaterial is recovered from the beads.
Hu and Wu, Chem. Eng. Res. Des. 65: 238-242, (1987), used a MSFB having a solid support of cubic particles formed of polyacrylamide gel, containing magnetite and an unknown type of cell to remove phenol from a circulating liquid. This MSFB system bubbled air through the reactor to aerake the cells and was operated as a batch reactor. Hu and Wu did not attempt to culture the cells, but merely used the MSFB as a sophisticated filter to remove phenol from a circulating fluid.
Summary o~ the Invention The present invention discloses a process for the production of cellular metabolites comprising the steps of providing a plurality of cell beads comprising cells which produce the desired cellular metabolite and magnetic particles in .

'' ~ 3 ~ 3 combination with a material capable o entrapping the cells;
providing a magnetically stabilized fluidized bed to zontain the cell beads; establishing a nutrient stream around the cell beads whereby nutrients in the nutrient stream diffuse into the cells entrapped in the cell beads and are used for cellular metabolism;
and retrieving the cellular metabolite from the nutrient stream or the cells.
In the case of cells which secrete the desired cellular metabolite, the metabolite can be removed from the nutrient stream by methods known in the art. In the case of cells which do not secrete the desired cellular metabolite, the cells are removed from the bed and the desired cellular metabolite is then recovered from the cells by methods known in the art.
Additionally, cell beads can be arranged in the bed in layers, each layer having been in contact with the fluid for di~ferent periods of time. The layers can then be removed from the bed sequentially, in the order in which the layers of cell beads were placed in the bed. This removal sequence allows older cells to be removed from the bed to make room for younger, fresher cells, so that the process for production of cellular metabolites ca~ be run most efficiently.
Cells can also be added and removed at the same rate, so that a continuou~ cell culture is established. This variation of the invention also allows the culture of the plant cells to progre~s efficiently and also eliminates the problems of washout ~ ' .- ' ' ' ' ~ ' .
,: ~ , - .
', ' ' ' ' ~
' ' , .

~ 3 ~ 3 and dilution present with other types of continuous cell culture reactors containing freely suspended cells.
Further, in the case of cells which secrete the desired cellular metabolite, the cells can be removed from the bed and be reused after a period of rest and growth in another container.
This variation of the invention is particularly useful in the case of cells which secrete the desired cellular metabolite only in the presence of stressful conditions (low amounts of nutrient, etc.).
The cells can be removed from the bed before they die and be reused after a period of rest and growth in another container under conditions which promote normal growth.
The process of the present invention makes it possible to maintain high oxygen transfer rates yet maintain the cells in a low shear environment due to their immobilization in the cell beads. The environment created by the MSFB reduces cell loss due to low oxygen concentrations and mechanical breakage caused by high shear rates in the medium. Since the cell beads are relatively motionless, bead collisions which cause bead damage are also eliminatad. Slow cell growth does not limit the reactor flow rates which can be used since the cells are immobilized in Ca-alginate beads as opposed to being freely suspended. The beads are themselves retained by gravity plus the magnetic ~orces, thus washout is not a problem. The process takes place in a closed environment thus minimizing the risk of contamination and 5 overgrowth by fa~ter growing microorganisms.

- .
' .
:~ : . . ' , :

~3~ 3~

Brief Descriptio~ of the Drawi~g The Figure 1 is a schematic diagram of an embodiment of a magnetically skabilized fluidized bed reactor system for use in this invention.
S Detailed Description of the IDvention An embodiment of a magnatically stabilized fluidized bed reactor system useful in thi~ invention is diagra~med in the figure. Cell beads flow into the column 5 of the magnetically stabilized fluidized bed from the solids input reservoir 1 by gravity and exit the column through the solids output 2. The diameter and length of the column 5 can be any dimensions. The preferred dimensions are 5 cm in diameter by 43 cm long. The dimensions of the preferred embodiment were chosen for ease in use of the invention. The column 5 can be constructed from any material that is non-magnetic and autoclavable. The preferred reactor material is glass for ease in viewing the cell beads and sterilization. The column 5 is cooled by a water jacket which can be made of any material which is non-magnetizable and does not interfere with the magnetic field induced by the coils 3. A
uniform magnetic field is generated through the coils 3. The magnetic field holds the cell beads in position within the column 5 so that the cell beads are relatively motionless. The connecting passageway 15 connects the parts of the magnetically stabilized fluidized bed reactor system. The nutrient stream is 5 conducted from the nutrient stream input vessel 8 by the , .

~ .

~3:L~3 connecting passageway 15 through pump 19 and enters the column 5 through a porous distributor 13 at the bottom of the column 5.
The porous distributor 13 can be made of any material that is autoclavable. Polypropylene is the preferred material. The preferred pore size is 250 microns. The reactor is designed with an O-ring joint 4 at the bottom of the column 5 so that the distributor 13 can be replaced and cleaned with ease. The nutrient stream flows around the cell beads in the column 5 and exits at khe top of the column 5. The nutrient stream can then be lo conducted through pump 21 and collected in a nutrient stream output vessel 9 and the cellular metabolite retrieved from the nutrient stream collected. Alternatively, the nutrient stream can be recirculated through the recycle vessel 7, conducted through the oxygenation means 6 and pump 17 and returned to the column 5.
The nutrient stream can also be split as it exits the column 5 and part of the nutrient stream collected in the nutrient stream output vessel 9 and part of the nutrient stream recirculated. In this case, fresh nutrient stream is added to the used nutrient ~tream via the recycle ve~sel 7 before the nutrient stream reenters the column 5. Air is conducted to the oxygenation means 6 through air input 10 and is removed from the oxygenation means 6 through air output 11.
As used herein, the term cellular metabolite encompasses molecules used or produced by the cell for primary metabolism and molecules used or producad by the cell in secondary metabolism.

: :

1 3 ~
UPN~69 The cell beads used in this invention comprise cells which produ~e the desired cellular metabolite and magnetic particles in combination with a material capable of entrapping the cells. ~enerally, the cells and magnetic particles will be dispersed within the material capable of entrapping cells~
although it may be possible to ~orm a ~ead having a core of magnetic particles coated with a mixture of cells and the material capable of entrapping the cells or a bead having a core of magnetic particles coated with an inner layer of the material capable of entrapping the cells and an outer layer of cells immobilized on the outer surface of the inner layer. The preferred bead diameter is 1 to 5 millimeters.
TAe cell beads can have various arrangements in the column 5. For example, the cell beads can be placed in the column 5 at the same time. In this case, the cell beads can be removed from the column 5 at the same time or in separate batches at different times. The cell beads can also be arranged in layers in the column 5. The layers can be placed in the column 5 at the same time or at different times. With this arrangement, the layers can be removed from the column 5 at the same time or at different times. Additionally, the cell beads can be added to the column 5 substantially continuously and simultaneously removed from the column 5 at a similar rate to establish a continuous flow of cell beads through the column 5.
Any type of cell which can be cultured while entrapped ~3~5~

in the material capable of entrapping the cells can be used in the invention. It is envisioned that plant cells, mammalian cells, hybridomas and cells which have been genetically manipulated by means such as gene splicing or insertion of plasmids or viruses can be used in the invention. The preferred cells are plant cells. It may also be possible to grow microorganisms such as bacteria or fungi in the invention.
The magnetic particles used in the invention are preferably ferromaynetic or superparamagnetic particles. The preferred magnetic particles are superparamagnetic magnetite.
The material capable of entrapping the cells can be any of the substances known in the art to be suitable for immobilized culture of cells. Polysaccharide gels such as alginate, agar, agarose and carrageenan have been used to immobilize cells. The preferred material for use in the invention is alginate.
The cell beads may be made by methods known in the art;
~ee U.S. Patent 4,675,113, The prefe~red method of preparing the cell beads comprises dropwise addition of a suspension of cells, alginate and magnetite in culture medium into a vessel containing culture medium supplemented by CaC12.
The nutrient stream is made up of medium used to culturs the cells. The medium will vary according to the kind o~ cells used in the reactor. If the cells will produce the cellular metabolite only under stre8~, the culture medium can be varied so ~ -16-.

.

~ 3 ~ 3 that it contains fewer nutrients than need d for optimum growth.
The medium used in the preferred embodiment of the invention with plant cells is Murashige-Skoog medium (with glucose as the carbon source) supplemented with 10 mg/ml L-cysteine, 1.0 mg/l thiamine-HCl, 1.0 mg/l 2,4-D and 0.2 mg/l kinetin. For animal cells, suitable media are known in the art. The nutrient stream should be sterile to minimize the risk of contamination of the reactor by microorganisms.
The nutrient stream may be recycled through a vessel where fresh medium is added. An amount of used medium equal in volume to that o~ the fresh medium is removed prior to entering the recycle vessel. The beads flow into the column from a reservoir by gravity. The entry of the cell beads may optionally be enhanced by a flowing stream of nutrient or retarded by a valve or strong magnetic field applied locally to the inlet tube.
Flow of the nutrient stream through the column must be fast enough to fluidize the bed yet slow enough to avoid damage to the cell beads. The preferred minimum flow rate is 20 ml per square centimeter of column cross sectional area and the maximum preferred flow rate is 40 ml per square centimeter of column cross sectional area in the preferred embodiment of the invention.
The flow rate of the nutrient stream when the column dimensions vary from the preferred embodiment may need to be adjusted accordingly.
Stabilization of the bed is accomplished by a series of magnetic coils powered by an adjustable output power supply. The magnetic coil~ can be of any number that can be arranged to provide a unifor~ magnetic field around the length of the column.
The preferred number of magnetic coils is four for a column size 5 cm in diameter and 43 cm long. The preferred coils are wound with 16 gauge copper magnet wire coated with a polyester-amide-imide insulating material and have 750 turns of copper wire.
Magnetic field strengths can range from 25 to l,000 Oersteds depending on cell bead and column size, the amount of magnetic material per cell bead and the nutrient stream flow rate.
A magnetic field strength of 200 Oersteds generated at 2 amps and 35 volts is the preferred magnetic field for the preferred flow rate and a bead size of 3-4 mm. The strength of the magnetic field may need to be adjusted to stabilize the bed when using column or cell bead sizes other than the preferred column or cell bead size.
The nutrient stream may be oxygenated by an oxygenation means. The oxygenation meane can be any system of oxygenation which maintains the sterile environment of the MSFB and provides adequate oxygenation. A capillary tube hemo dialyzer, for example a renal dialysis cartridge, is the preferred oxygenation means because it is readily available and economical to use. A
commercial oxygenation cartridge with micorporous wall capillary tubes may be used if a higher amount of oxygen i5 needed than can be provided by a capillary tube hemo dialyzer.
A more detailed description of the preferred embodiment .~ ',.

~ 3 1 ~

of the invention is set forth below.
Cultivation of Cell~ Prior To Vse In Cell Beads Suspension cultures of Coffea_ara~ica c211s from callus cultures are initiated by adding 8.8 g of cells to 50 ml of Murashige-Skoog medium (with glucose as the carbon source) supplemented with10 mg/l L-cysteine, 1.0 mg/l thiamine-HCl, 1.0 mg/l 2,4-D, and 0.2 mg/l kinetin in a 125 ml Erlenmeyer flask. The cells are cultivated on a gyratory shaker at 150 rpm. The suspension cultures are sub-cultured every 10 days by doubling the volume of the suspension by adding fresh medium, then splitting this into two new flasks.
Preparation of Beads With Immobilized Cells The following procedure is carried out under sterile conditions. Coffea arabica cells are washed with medium on a glass filter. A 50/50 (w/w) mixture of cells to alginate solution is prepared by adding 20 g of cells to 20 g of alginate solution containing 3.0%(w/w) Protanol LF (sodium alginate) and 20%(w/w) magnetite (Fe304) in culture medium. Beads approximately 3-4mm in size are made by placing the cell/alginate suspension in a 50ml syringe and adding the suspension dropwise into 100 ml of medium supplemented with 50 mM CaC12. Mild stirring is used to keep the solution well-mixed. The beads are left in the medium for 1 hr to allow for complete reaction of alginate with the Ca++ ions.
R~moval of Cells Fr~ The Beads Approximately 1.0 g of beads is suspended in 10 ml of :

medium containing 0.1 M sodium citrate at pH 6Ø The beads are dissolved by shaking the suspension for 30 min to 1 hr on a nrock'n'roll~ shaker. After the beads are completely dissolved, the suspension is filter~d on a nylon filter (50um pore size) in a buchner funnel. To ~llow the cells to recover from the above treatment, the cells are incubated in fresh medium with mild aeration from an external air supply for 1 hr.

Cultivation of c. arabica cells Using A ~SFB To Produce the Alkaloids Caffeine and Theobromine A glass column 43 cm long and 5 cm in inside diameter is used for cell cultivation. The column is positioned in the center of four coils. Each coil has an inside diameter of 11.44 cm, an outside diameter of 17.78 cm and a height of 5.72 cm. The coils are placed 2.5 cm apart. The coils are wound with 16 gauge copper magnet wire coated with a polyester-amide-imide insulating material. Each coil has 750 turns of copper wire. A magnetic field strength of 200 Oersteds is generated at 2 amps and 35 volts. A Plexiglas*water jacket 39.5 cm in length and an inside diameter of 7.0 cm is sealed over the glass column. Using a water flow rate of 100 ml/min in the jacket, heat is removed from the coils, thus maintaining the column at 25C.
Medium enters the bottom of the column and flows through a porous polypropylene liquid distributor (250um pores) and exits from a port at the top. Most of the medium (90 - 100%) is recycled through a recirculation loop where first fresh medium is added to ~ l -20-*Trademark "f'"'~' ' . ' ~ ' . ~

~1 31~8~

it at a flow rate of 5 ml/min in a recycle vessel having an inside diameter of 7.62 cm and a length of 10.16 cm. Then, the stream flows into a hollow fiber dialyzer (23.5 cm in length, 400 cm in diameter, and effective surface area of 1.2 sq. m) where the medium is saturated with air before reentering the column. The air flow rate in the dialyzer is 1 ml/min. The medium which is not recycled is pumped into a collection vessel at a rate of 5 ml/min.
Calcium alginate/magnetite beads with immobilized C.
arabica cells having a diameter of approximately 4 mm are added to the top of the column semi-continuously every six (6) days. The beads are gravity fed into the column from a 100 ml boiling flask into a column inlet port. A total bead volume of 65 ml is added each time. The same total bead volume is removed every 6 days from the bottom of the column by gravity flow through a 12 mm glass tube placed in the center of the liquid distributor. The beads exit through this tube into a 100 ml boiling flask.
The beads are fluidized with a medium flow rate of 400-600 ml/min. A bed height of 20 cm is maintained throughout the experiment. From each sample that is collected on six day intervals, a dry weight and cell viability measurement is made.
When each bead sample is collected., a 5 ml liquid sample is collected from the liquid stream entering the collection vessel.
This liquid sample is analyzed for glucose and alkaloid (caffeine and theobromine) concentration.

' , Cell Viability A 100 mg cell sample is added to a test-tube which contained 3ml of a TTC (triphenyl tetrazolium choride) solution.
The solution is prepared by adding 0.6~ (w/v) triphenyl tetrazolium chloride to a 0.05 M sodium phosphate buffer at pH
7.5. The suspension is incubated overnight at 25C in the dark.
The solution is removed and the cells are washed with water. This is then centrifuged at 1000 rpm for 5 min. and the water layer is removed. The cells are extracted with 7 ml of 95% ethanol for 5 min at 80C. After cooling, 3 ml of 95% ethanol is added. This suæpension is again centrifuged at ~000 rpm for 5 min. The supernatant is removed and analyzed using a spectrophotometer at 485 nm. The per cent cell viability is determined relative to callus culture (100% viability).
Alkaloid analysis Determination of the alkaloid concentrations (caffeine and theobromine) in the medium are carried out using reverse phase HPLC. The column is Spherisorb C-l~*with the dimensions 4.6 ~m x lOO mm. A guard column is placed before the chromatography column having the same packing material and with the dimensions 4.6 x 30 mm. The mobile phase is an isocratic solution of MeOH:H2O (34:66) with l~M M~S at pH 5.8. The flow rate of tbe mobile phase is 1 ml/min. The sample enters the system through a lOul injection loop~ Detection of the sample is monitored with an ISCO* W
detector at 280-300nm. Alkaloid production of the cells *Trademark ``

-, UPN-6s 1 3~ 3 immobilized in cells beads is comparable to alkaloid production of cells grown in free suspension culture or immobilized in alginate alone.
Dry Weight ~amples are filter~d on a nylon filter (50um pore size), then weighed (total fre~h weight). A portion of the sample, 400 mg, is removed and dried o~ernight at 50C in an oven, then weighed. The total dry weight of ~he sample is calculated from the known dry ~eight of the portion of the sample.
Cells entrapped in cell beads increase in dry weight at rates comparable to cells grown in free suspension culture and in alginate alone.
Glucose As~ay The glucose concentration of the medium is determined using the Sigma Chemical Company 510-A Glucose Assay Kit. All samples are diluted 10 fold before using the assay.
Cells in bead cells utili~e glucose at rates similar to cells grown in free suspension culture and entrapped in alginate alone.
The invention is also applicable in other ways to the production of cellular metabolites. Typically, secondary plant metabolites are produced during the stationary phase when the growth limiting substrate has been depleted and the cells are under stress. The cell~, however, can survive only for a fixed length of time after this occurs. The MSFB is ideal for operation in such a mode. Cells which are already in the stationary phase '~'''''' '' . . .
.
- :. ' .

, ~ :

UPN-69 1 3 1 68l~3 can be put in at the top of the column and irrigated with a non-growth, product producing liquid in order to produce the cellular metabolites of interest. When the cells begin to lose viability or productivity, they can be removed ~rom the bottom of the column and either discarded or recycled after recovering in a medium which allows normal cellular metabolism.
The present invention is also applicable to the production of cellular metabolites which are not secreted by the cell, but are retained inside the cell. Cell beads are removed continuously from the bottom of the reactor when they have accumulated the maximal amounts of product. Cells are then removed from the cell beads and disrupted to release the product.
The product is then purified by conventional methods known in the art.
Another application of the present invention to cells which retain the desired cellular metabolite intracellularly is to alter the nutrient stream so that it includes a solvent which can permeabilize or damage the cell wall and cause the desired cellular to be released into the nutrient stream. With the MSFB
it is possible to instantaneously alter the nutrient stream for such a purpose by adding the solvent through a separate portal in the recirculation route or through the vessel in which the nutrient stream is supplied.
The present invention i~ also applicable to the ~5 production o~ biomass.

.

: . :
~,.

~ 3 ~ 3 Because solids can be removed from the bed in a sequential fashion, it is possible to sample from all regions in the bed by rapidly removing all solids sequentially. Each layer will contain cells which have been in contact with the medium being recirculated for a defined period of time. Thus, it is possible to analyze the behavior of c~lls at discrete time intervals where each sample contains cells o~ a given ~age~ (where age is defined as the time the cells have spent in the bed). This makes it possible to study cells at different time intervals in one experiment. Finally, since the fluid composition can be instantly altered, this allows one to easily study the effect of medium composition and growth factors on cell growth and metabolism by a pulse or step response technique.
It is anticipated that this invention can be used to grow mammalian cells with appropriate changes in materials and methods.

~. .
' . , -~ .

Claims (22)

1. A process for the production of cellular metabolites comprising the steps:
(a). providing a plurality of cell beads comprising cells which produce said cellular metabolites and magnetic particles in combination with a material capable of entrapping said cells;
(b). providing a magnetically stabilized fluidized bed to contain said cell beads:
(c). establishing a nutrient stream around said cell beads whereby nutrients in said nutrient stream diffuse into said cells entrapped in said cell beads and are used for cell metabolism;
(d). retrieving said cellular metabolite from said nutrient stream or said cells.

2. The method of claim 1 wherein said material capable of entrapping said cells is a Polysaccharide gel.

3. The method of claim 2 wherein said polysaccharide gel is alginate.

4. The method of claim 1 wherein said magnetic particle is a superparamagnetic particle.

5. The method of claim 4 wherein said superparamagnetic particle is magnetite.

6. The method of claim 1 wherein said cell beads are arranged in layers in said bed.

7. The method of claim 6 wherein additional layers are periodically added to said bed and layers are periodically removed from said bed.

8. The method of claim 6 wherein said layers are removed from said bed in the order in which said layers were added to said bed, the first layers added being removed first and later added layers being removed later in time.

9. The method of claim 1 further comprising the step of adding said cell beads to said bed substantially continuously and simultaneously removing said cell beads from said bed at a similar rate to establish a continuous flow of cell beads through said bed.

10. The method of claim 1 further comprising the step of recirculating said nutrient stream around said cell beads.

11. The method of claim 10 further comprising the step of conducting said nutrient stream through oxygenation means prior to recirculating said nutrient stream around said cell beads.

12. The method of claim 11 wherein said oxygenation means comprises a capillary tube hemo dialyzer.

13. The method of claim 12 wherein said capillary tube hemo dialyzer is a renal dialysis cartridge.

14. The method of claim 9 further comprising the step of recirculating said nutrient stream around said cell beads.

15. The method of claim 14 further comprising the step of conducting said nutrient stream through oxygenation means prior to recirculating said nutrient stream around said cell beads.

16. The method of claim 15 wherein said oxygenation means comprises a capillary tube hemo dialyzer.

17. The method of claim 16 wherein said capillary tube hemo dialyzer is a renal dialysis cartridge.

18. The method of claim 1 wherein said cell are eukaryotic cells.

19. The method of claim 18 wherein said eukaryotic cells are plant cells.

20. The method of claim 1 further comprising the steps of (1) harvesting said cells from said bed;
(2). recovering said cells from said beads;
(3). culturing said cells in a location which is not said bed, thereby allowing cells to rest:
(4). returning said cells to said bed whereby said cells are reused to produce said cellular metabolites.

21. The method of claim 1 wherein said nutrient stream flow rate is between 20 and 40 milliliters per square centimeter per minute.

22. The method of claim 1 wherein said nutrient stream comprises cell nutrients and a solvent capable of permeabilizing or damaging the walls of said cells whereby said solvent permits said cellular metabolite to escape from said cell into said nutrient stream.