Metabolism of hybridoma cells and antibody secretion at high cell densities in dialysis tubing

Metabolism of hybridoma cells and antibody secretion at high cell densities in dialysis tubing Claudia Ktisehagen,* Fritjof Linz,t Gerlinde Kretzmer, Thomas Scheper and Karl Sehiiged Institut far Technische Chemie, Universitiit Hannooer, Hannooer, FRG * Technische Oberwachungs-Verein Hannover e.V., Hannover, FRG t Department of Chemical Engineering, Rensselaer Polytechnic Institute, Troy, New York, USA

The experimental setup, consisting of a bundle of dialysis tubing 2.5 mm in diameter [10-15 kD cutoff, mean pore size 25 A, 20 txm (dry) and 40 tzm (wet) wall thickness] inserted into a l-I glass bioreactor supplied with oxygen and pH electrodes, a porous gas distributor, a sampling tube, and a holder for the eight pieces of dialysis tubing, was developed to investigate the properties and the microenvironment of hybridoma cells enclosed in the tubing during their batch cultivation. The concentrations of lowmolecular-weight medium components were the same inside and outside the tubing, and it was possible to control the microenvironment of the cells in the tubing easily. The cell damage caused by mechanical stress was less in the dialysis tubing than in stirred spinner flasks. The influence of the initial cell density in the range from 4 x l0 s to 1 × 108 cells ml -l and the cultivation time were evaluated according to the total and viable cell concentrations and the cell~cellfragment size distributions. Furthermore, the cell membrane properties, glucose consumption rate, lactate, ammonia and lipid storage material, and the monoclonal antibody production rates as well as intracellular enzyme activities in the culture medium were measured and compared to those in reference cultures in spinner flasks with the same inoculum at low initial cell densities. In dialysis tubing in a concentration range of 5 x 106 to 108 cells ml-i, the total and viable concentrations of cells remained the same during cultivation. At high cell concentrations, cell size decreased, membrane folding increased, the fraction of dead cells and cell fragments increased, and the metabolic activity of the cells and the monoclonal antibody productivity were higher than with low cell concentrations. Since the dialysis membrane is nonpermeable for proteins, high monoclonal antibody concentrations were attained.

Keywords: Hybridoma cells; monoclonal antibody; membrane reactor; high cell density; flow cytometry; electrorotation measurements

Introduction Different methods have been used to increase mammalian cell density in bioreactors. Cell immobilization by different carriers or in hollow fiber membrane modules and cell retention in membrane (perfusion) reactors are the most popular methods. The highest local densities are attained in hollow fiber membrane modules. The cells are retained in the extracapillary space of the module, and they are supplied with nutrients and oxy-

Address reprint requests to Dr. Schiiged at the Institut for Technische Chemie, Universit~it Hannover, Callinstrasse 3, D-3000 Hannover 1, Germany Received 15 February 1991; accepted 20 May 1991

© 1991 Butterworth-Heinemann

gen from the intracapillary to the extracapillary space by transport across the membrane. 1-4 The culture temperature, pH, pO 2, and composition are controlled in a separate reactor connected to the capillaries of the module through which medium is pumped continuously to maintain a constant environment for the cells. The membrane reactor can be operated for several weeks 2,3 (30 to 40 days 4 and 41 daysS). Depending on the choice of the membrane cutoff, particular medium components, metabolites, or products can be replaced repeatedly or retained in the extracapillary space. When using the low membrane cutoff, serum components and product are retained and enriched in the extracapillary space, which results in lower operation and recovery costs, z By this, the yield of monoclonal antibodies increased by 80%.6

Enzyme Microb. Technol., 1991, vol. 13, November

873

Papers One of the main problems of cell cultivation in membrane modules is that the cells cannot be examined during cultivation. The aim of the present investigations was to develop a reactor system which allows the study of the cells during cultivation. In order to retain the monoclonal antibody, dialysis tubing was used.

Materials and methods

Cell lines, precultures, and media Two cell lines were used: (a) hybridoma 3C2, which is a mouse/mouse hybridoma constructed by the fusion of a myeloma cell P3 X63-Ag8 with the activated Blymphocyte of the mouse, donated by Nabet, Nancy, and (b) the mouse/rat hybridoma 187.1 (ATCC No. HB 58). The mouse/mouse hybridoma 3C2 produces a monoc|onal IgG1 antibody against the HCG hormone. The mouse/rat hybridoma 187.1 produces a rat antibody of the class IgG against the kappa-light chain of mouse immunoglobulin. The hybridoma stock cultures were incubated at 37°C, 95% relative humidity, and 5% CO2 in air in Roux flasks (Falcon) in incubators (Heraeus Type B 5060 EK COz) and were diluted two times per week with fresh medium to 105 cells m1-1. These cells were used as inoculum for the precultures in 2-1 Bellco-spinner flasks at 60-70 rev min -1. The hybridoma cell line 3C2 was cultivated in serumfree IMDM medium (Flow Laboratories) and in serumfree DIF medium 4'7 as well as in DIF + 10% FCS medium. The hybridoma cell line 178.1 was cultivated in serum-free DIF medium, 5'8 which is a 1 : 1 mixture oflMDM (Gibco) and HAM F 12 (Gibco) supplemented with transferrin, insulin, sodium pyruvate, 1-glutamine, NaHCO 3, and BSA.

Analytical methods The determination of cell concentration and viability was performed in a hemocytometer by using trypan blue exclusion. The following enzymatic reactions were used for the determination of glucose and lactate, ammonia, and urea concentrations: Glucose was converted by glucose dehydrogenase (Gluc-DH) and NAD ÷ into gluconolactone and NADH. Lactate was converted by lactate dehydrogenase (LDH) and NAD ÷ to pyruvate and NADH. To shift the equilibrium to pyruvate, the latter was converted by glutamate-pyruvate transaminase (GPT) to alanine and a-ketoglutarate. In both cases, the N A D H concentration was measured at 450 nm. The NH~- ion was converted to glutamate, NAD + , and H20 in the presence of a-ketoglutarate and NADH by glutamate dehydrogenase (GIDH). The NADH concentration was measured at 450 nm. Urea was converted by urease into ammonia and CO2. The ammonia was determined by G1DH through the latter method. Amino acid concentrations were measured by HPLC and OPA-precolumn derivatization with a C-18 column (SP 8700 HPLC and the SP 8780 XR autosampier, Spectra Physics) at 30°C, detected by Fluoromoni-

874

tor Ili (Milton Roy), and evaluated by an integrator (SP 4200 Spectra physics). The monoclonal antibody concentration was determined by an enzyme-linked immunosorbent assay (ELISA) with goat anti-rat IgG and horseradish peroxidase (POD) conjugated antibody on microtiter plate (Nunc) with an immunoreader (NJ 2000, Nunc) (for more details, see ref. 7). The extracellular activities of the intracellularly released enzymes LDH, glutamate-oxalacetate transaminase (GOT), and GIDH were determined as well. L D H was used to convert pyruvate and N A D H into lactate and NAD +, GOT was applied to convert 2-oxoglutarate and aspartate into glutamate and oxalacetate, and the latter by malate dehydrogenase and N A D H into malate and NAD ÷ . G1DH was used to convert 2-oxoglutarate, NADH, and NH~- into glutamate, NAD +, and HzO. The N A D H concentration was obtained for all of them. The overall protein content inside and outside of the tubing was detected by Coomassie blue. The cell/cell fragment size distributions were determined with a flow cytometer (Partograph FMP. Kratel) by measuring the He-Ne laser light extinction at 90°; the lipid content was determined with nile red9; DNA and RNA contents were determined by propidium iodide staining, by argon ion laser excitation, and measuring the backward fluorescence intensity. 9 The plasma membrane properties of the cells were studied by means of the electrorotation technique. 10Membrane capacity and membrane resistance were evaluated by this technique.

Apparatus and sampling The aim of~this research was the investigation of the biological state of the cells and their microenvironment in the membrane module as a function of the cultivation time. The sophisticated equipment made this possible. It consisted of eight pieces of regenerated cellulose dialysis tubing, 2.5 mm in diameter (10-15 kD cutoff, mean pore size 25 A, 20 ~m (dry) and 40 ~m (wet) wall thickness) (Thomapor "Rapid," Reichelt Chemietechnik GmbH & Co.) inserted into a 1-1 glass bioreactor supplied with oxygen and pH electrodes, a porous gas distributor, a sampling tube, and a holder for the eight pieces of dialysis tubing (Figure 1). This equipment was filled with physiological NaC1 solution and sterilized for 40 rain at 120°C in an autoclave. Before inoculation, the NaCI solution was drained, the cell suspension was concentrated by centrifuge to 1.3 x 10 6 tO 1 × 10 8 cell density, and the tubing was filled by pipets under aseptic conditions. Finally, the reactor was filled with protein-free culture medium. This equipment was installed in an incubator (Heraeus Type B 5060 E K CO2), operated at 37°C under 95% relative humidity with a magnetic stirrer, pH and pO2 control, and controlled pressurized air/CO2 mixture. For sampling the cell-containing medium, every day one dialysis tubing was closed at both ends with hot nippers and the tubing was then removed from the

Enzyme Microb. Technol., 1991, vol. 13, November

Hybridoma cells in dialysis tubing: C. K~sehagen et al. |

I

. . . .

' I I,pO2IpH ,

I I

pH measurement

I l l

I

--

I

,

and control pO 2 measurement and control

~

SV

B

Figure 1 Experimental setup. B, Incubator; F, sterile filter; G, porous gas distributor; MV, magnetic valve; M, pressure meter; P, pressurized air; R, magnetic stirrer; SV, control valve

holder. Each sample consisted of 1-1.5 ml cell suspension. After separation of the cells by centrifugation at 200g for 5 min, 0.7-0.8 ml cell-free sample was recovered and used for medium analysis. Parallel to these investigations, cells were cultivated in Bellco-spinner flasks. They were inoculated with the same preculture but with cell densities of 4 x 105 cells ml-1 under the same time conditions, and sampled just as the dialysis tube cultures were. These cultures were used as references. The following abbreviations are used in the subsequent representations: IDT, medium inside the dialysis tubing; ODT, medium outside the dialysis tubing; REF, cells and medium of the reference culture in the spinner flasks. Results

Influence of the initial cell concentration on cell growth in stationary flasks The culture media consisting of RPMI 1640 and 4% HS in monolayer flasks were inoculated with different densities (0.1,0.5, and 1 x 106 cells ml- l, respectively) of hybridoma 3C2 cells. The total and viable cell concentrations were measured as a function of the cultivation time. For all three cultures, the same maximum cell densities (4 x 106 cells ml- 1) were achieved. However, they were attained faster with high initial cell concentrations than with lower ones. With the initial concentration of 0.1 x 106 cells ml -~, the maximum was attained after 150 h (671% increase), with 0.5 x 106 cells

ml -~ after 130 h (654% increase), with 1 x 106 cells m1-1 after 80 h (118% increase), and with 2 x 106 cells ml-1 after 50 h (41% increase). The higher the initial cell concentration, the lower the growth rate in static cultures.

Influence of medium formation on cell growth in microtiter plates Three different media were used: serum-free DIF medium, DIF with 10% FCS, and serum-free IMDM/F 8 medium (Flow Laboratories). Two milliliters of medium was inoculated with 6 x 104 hybridoma 187.1 cells in each of the holes of the microtiter plates (Costar). On the third, fourth, and fifth day, the cell concentrations and the monoclonal antibody (MAB) concentrations were determined (Figures 2a and b). These figures show the well-known positive effect of the serum on growth and MAB production. Therefore, all of the following investigations were carded out--with few exc e p t i o n s - w i t h the same (DIF/FCS) culture medium.

Cultivation in the dialysis tubing The following conditions were kept constant during the course of the cultivation: Dialysis tubing length: 22-24 cm (volume 1.75-2 ml); Dissolved oxygen concentration: 60% and 90% of the saturation value; pH value: 6.5-7 (monitored outside the tubing and controlled at the end of the measurements in the culture medium from inside the tubing);

E n z y m e Microb. Technol., 1991, vol. 13, N o v e m b e r

875

Papers total cell concentration [ml-l]

viable cell conc. [m1-1]

iOS 1.10 B ml-1

lOG! 5,4.10 6 mi-]

C] IM.DM/F El

/;7

DIF

E] DIFIFCS

5 cultivation time (day)

1061 O

1,3.10 6 mini/

4,10 5 ml-1

IO S

cultivation time (day) MAB-conc. [#g ml"1] Figure 3 Cell concentration courses with different initial cell concentrations. (El) REF culture

:~10o IMDM/F D DIF g DIF/FCS

0

:3

4

cultivation time

5

(day)

Figure 2 (a) Viable cell concentrations with serum-free IMDM/F, seru m-free DI F, and DI F + 10% FCS media at different cultivation times. (b) M o n o c l o n a l a n t i b o d y (MAB) concentrations with serum-free IMDM/F, serum-free DIF, and DIF + 10% FCS media at different cultivation times

Stirrer speed: 60-70 rev min-t; Filling volume of the bioreactor: 1 1 protein-free medium; Filling volume of the tubing: 1-1.5 ml. The same preculture was used for the inoculation of the reference Bellco-spinner flasks. The dialysis cultures and the spinner flasks were sampled simultaneously. The cell concentrations and extracellular enzyme activities were determined immediately after the sampling, as were the measurements with the laser flow cytometers. Cell concentration and viability. At an initial concentration of 1.3 x 10 6 cells m1-1, the growth proceeded as usual: the lag phase was followed by the exponential growth phase and a stationary phase. The ratio of viable to total cells was constant. In this case the cell growth in the spinner flasks and in the tubing proceeded in the same way. During 7 days, the cell densities increased by a factor of 500. At an initial cell concentration of 5 × 10 6 cells ml-1 (IDT) and above, up to 1 x 108 cells ml-i (IDT), the 876

total and viable cell concentrations remained constant during the 7 days (Figure 3). At densities of 5 × 10 6 cells ml- ~ and higher in the dialysis tubing, the fraction of nonviable cells was very high; in fact, it varied between 45% (at 5 × 10 6 cells m1-1) and 80% (1 x 108 cells ml-=). Since the dissolved oxygen concentration was sufficiently high in the tubing as well (above 60% of the saturation outside the tubing), concentration depletion of the substrates in the dialysis tubing was assumed. To check the latter, the concentrations of low-molecularweight components were determined in IDT and ODT. The difference between the concentrations of glucose, lactate, and amino acids in the IDT and ODT were slight. One, therefore, must assume that there is no substrate depletion in IDT cultures in well-controlled medium composition in the ODT.

Variation in the concentrations of low-molecular-weight medium components during cell cultivation. The variations of the glucose and lactate concentrations in the IDT/ODT and in the spinner culture were similar at low cell densities: glucose concentration decreased and lactate concentration increased with the cultivation time. There were close relationships between the growth rate and glucose utilization and lactate production rates. At high cell densities in the IDT, no net cell growth was observed. The glucose concentration, however, decreased and the lactate concentration increased. The particular amino acid concentrations behaved in three different ways: (1) Serine and glycine concentrations remained constant; (2) Alanine and lysine concentrations increased; (3) The concentrations of all other amino acids decreased. The concentration variations of the amino acids in the IDT/ODT and in the R E F were similar. The specific

Enzyme Microb. Technol., 1991, vol. 13, November

Hybridoma cells in dialysis tubing: C. K~sehagen et Table I Specific ammonia concentrations in the dialysis tubing (IDT) and reference (REF) cultures

MAB-conc. [/~gml-1] I~O0

Cultivation time (days)

mg NH~ (106 cells) -1 in IDT

mg NH~ (106 cells) -1 in REF

al.

a

IOOCIJ

\ \ \

1 2 3 4 5 6 7

-0.174 0.207 0.309 0.427 0.492 0.738

0.164 0.334 0.310 0.161 0.068 0.037 0.034 I

fl

.

0

~

.

. E~

. ~

.

. IO0

. 120

I~U

1 ~180

180

cultivation time [h] (with reference to the cell number calculated) variations, however, were higher in IDT/ODT than in REF cultures. Ammonia is produced during the utilization of glutamine and the deamination of amino acids. Since ammonia is toxic for the cells, they detoxify it. The ammonia Concentration increases in IDT/ODT as well as in REF cultures. At high cell density, however [e.g., at 1 x 108 cells ml-1 (IDT)], the specific ammonia production rate was higher (with reference to the cell number calculated) (Table 1). In the dialysis tubing (IDT), the ammonia production remained nearly constant during 7 days and had an average value of 0.105 mg N H ; 10 - 6 cells per day. In the reference (REF) cultivation, nearly the same value (0.103 mg NH~- 10 - 6 cells per day) was attained, but only during the first 3 days; during the following 3 days, the ammonia was consumed at a specific rate of 0.103 mg NH~ 10 - 6 cells per day. Monoclonal antibody production in the I D T and R E F cultures. Since the dialysis tubing is not permeable to

high-molecular-weight components, no MAB was present in the medium outside the dialysis tubing (ODT). MAB concentration usually increased with the cultivation time during the first 7 days in IDT as well as in REF cultures. The MAB concentrations, however, were considerably higher in the IDT than in the REF culture. In Figure 4, the MAB concentrations are shown in IDT and in REF cultures with DIF + 10% FCS. The volumetric productivities were: at 10,500/~g ml- t, 1,575/zg MAB ml- ~day- ~in IDT, and at 295/~g m1-1, 36.8/~g MAB m1-1 day -1 in REF cultures. On the seventh day, however, the MAB concentration dropped in the IDT. In Figure 5, another example without FCS is depicted. The MAB concentration showed an increase during the first 6 days, and after that it dropped. The volumetric productivity amounted to 3.33 /zg MAB ml -I day (IDT) during that time. The volumetric productivity in the REF was considerably lower. With respect to specific productivity based on the total cell concentration, the productivity in IDT was lower than in REF cultures. However, the specific productivity based on the viable cell concentration re-

MAB-conc. [#g ml-1] 31313

b

Ioo

13

cultivation time [h] Figure 4

Monoclonal antibody concentrations as a function of

cultivation time. Cultivation with 10% FCS (a) in dialysis tubing, (b) in spinner flask

I~B-conc. [#g ml "1]

20 \ \ \ \

10

0 0

20

LO

~ 60

80

100

120

1/-,0 160

180

cultivation time [h] Figure 5 Monoclonal antibody concentration as a function of cultivation time. Cultivation with old hybridoma cells in serumfree medium (<3) in dialysis tubing, (0) in spinner flasks

E n z y m e M i c r o b . T e c h n o l . , 1991, vol, 13, N o v e m b e r

877

Papers LDH activity [U m1-1]

GOT-activity [U F1]

7

3[3-

6 5' 20

4" 3" tO"

2" 1" 0

0

"1

2'0

4'0

60

8'0

13

i

160 120 140 1t50 180

,-

0

,

40

~

,

t:21D

,.

f~D

2~13

cultivation time [h]

cultivation time [11] Figure 6 Lactate dehydrogenase (LDH) activity as a function of the cultivation time. Cultivation (×) in dialysis tubing, (+) in spinner culture

Figure 7 Glutamate-oxalacetate transaminase (GOT) activity as a function of cultivation time. Cultivation ( × ) in dialysis tubing, ( + ) in spinner flasks

GIDH activity [U 11]

suited in higher productivity in IDT than in REF culture s. Activities of intracellular enzymes in the IDT and REF cultures. All three investigated enzymes play an important role in cell metabolism. LDH catalyses the reduction from pyruvate to lactate. Its extracellular activity indicates cell damage. In Figure 6, the course of LDH activity in IDT and REF cultures is shown. The activity in IDT cultures is high and passes a maximum, while only low activities are recorded for the REF. The specific activities with regard to the cell concentration, however, are lower in IDT than in REF cultures (Table 2). In general, the following observation is valid: the specific activities in the IDT cultures are considerably lower than in the REF cultures. GOT also is an important enzyme for metabolism. Figure 7 shows the variation of GOT activity in IDT and REF cultures. Both pass a maximum. The activity in 1DT, however, was lower. The specific GOT activities with regard to the cell concentrations were lower by several orders of magnitude in IDT than in REF cultures.

IS'

10"

S"

'

'

33

40

E~3

~3

.

IOI3

.

.

120

.

14(3

16El

I013

cultivation time [h] Figure 8 Glutamate dehydrogenase (GIDH) activity as a function of cultivation time. Cultivation in dialysis tubing

GIDH catalyses the conversion of glutamate to 2oxyglutarate and ammonia. Figure 8 shows GIDH activity variation in the IDT culture medium during cultivation. The GIDH activities in the REF culture were negligible. The high GIDH activity is in agreement with the high ammonia concentration in the IDT culture medium.

Table 2 Specific LDH activities in IDT and REF cultures Cultivation time (days)

LDH activity U (106 cells) -1 in IDT

LDH activity U (10e cells) -1 in REF

1 2 4 5 7

0.084 0.022 0.059 0.055 0.058

0.615 0.640 0.413 0.201 0.131

878

Total protein concentration in the IDT and REF cultures. In serum-free culture media, the protein concentration was nearly constant in the IDT and amounted to about 0.9 g 1-1. Laser flow cytometer measurements. The number of cell components as well as the cell/cell fragment size distributions can be determined by means of flow cytometers. The latter were measured during cultivation in Petri dishes and in IDT and REF cultures, a'9

Enzyme Microb. Technol., 1991, vol. 13, November

Hybridoma cells in dialysis tubing: C. K~sehagen et al. Frequency

1. day

b

1. day

1. day

I

r

T

I

r

I

I

T

5. day

5. day

14. day

O

>=

5

10

15

20

5

10

15

20

?

r

I

I

T

t

;

6. day

Diameter [pm] Figure 9 Size distribution of cells in Petri dishes on different days after inoculation

In Petri dish cultures and during the exponential growth phase, the cell size varied on the fourth day between 13.3 and 11.2 /~m. Simultaneously, the size distribution function became narrower. The size distributions remained constant up to the beginning of the death phase on the eighth day, when the mean cell size decreased to 8.1 /zm, and at the same time, the distribution became broader (Figure 9). During the initial 7 days, the shift of the size distribution was caused mainly by the reduction of cell size, and after the eighth day by cell fragmentation. The size distributions in IDT and REF cultures (Figure 10) differed from those in the petri dishes. Two maxima appeared already at the beginning of the cultivation. The first maximum at about 2/zm was due to the cell fragments; the second maximum at 8-15/zm was caused by the cells. The positions of the first maxima were similar in IDT and REF cultures. With increasing cultivation time, however, the height of the first maximum diminished in the REF cultures, contrary to that in the IDT cultures. The cell size invariably was smaller in IDT cultures than in REF cultures. The formation of lipid droplets in the cytoplasm consisting of neutral lipids, often of triglycerides or cholesterol ester, is a well-known phenomenon. These droplets can be stained by nile red and observed in a laser flow cytometer or fluorescence microscope. 9 In Figure 11, the relative fluorescence frequency diagrams are shown for cells cultivated on Petri dishes. As depicted, no lipid droplets are present in the cells on the first day. With progressing time, the lipid content increases and attains a maximum on the eighth day and then gradually diminishes.

I

f

I

;

t

t

I

r

5

I0

15

20

5

10

15

20

Diameter [#m]

Figure 10 Size distribution of cell/cell fragments (a) in dialysis tubing, (b) in spinner flasks on different days after inoculation

In Figure 12, the fluorescence frequency diagrams are shown for cells in IDT and R E F cultures. Once more, the lipid content is very low at the beginning of the cultivation, but it quickly increases and attains its maximum on the third day. Thereafter, the cytoplasmic lipid content decreases faster in IDT than in REF cultures. The DNA and RNA content of the cells can be determined by propidium iodide staining. The DNA content increased from the first day onward and attained its maximum, which is twice as high as on the first day, on the sixth day. The RNA content was high during the first few cultivation days and decreased thereafter. Further investigations, however, are necessary to gain more information on the variation of the RNA content. Electrorotation measurement. The properties of the

cell membrane can be evaluated ~°from the variation of the rotation spectrum (rotation speed of the cells in a rotational electrical field as a function of the frequency). The frequency at which the minimum rotation speed is attained is called the critical frequency. A linear

Enzyme Microb. Technol., 1991, vol. 13, November

879

Papers 1. day

2. day /'~

t~

2. day

c

== g a. day

)

lO. day

4.day

3. day

relative fluorescence

5,day

8. day

Figure 11 Relative fluorescence frequency distribution of nile red-stained cells cultivated in Petri dishes on different days after inoculation

relationship exists between the critical frequency and the electrical conductivity of the outer medium for cells oflDT as well as of REF cultures, as was to be expected in accord with the theory. Therefore, the evaluation of the membrane properties was possible. In Figure 13, the cytoplasmic membrane capacities of the cells of the IDT and R E F cultures are plotted as a function of the cultivation time. The membrane capacities were nearly constant during the cultivation, but those of the cells from IDT cultures were invariably larger than those of cells from the reference cultures. Increased membrane capacity was probably caused by the increased folding of microvilli, which is in good agreement with the reduced cell size in IDT cultures. In Figure 14, the membrane conductivities of the cells from the IDT and R E F cultures are shown as a function of the cultivation time. Increased conductivity means that the membrane permeability increased for charged particles, i.e., more ions can be transported across the membrane per unit of time. The higher membrane conductivity of cells from IDT cultures compared to that of cells from the R E F cultures could be caused by the higher metabolic activity of the first-mentioned cells.

rel~tWe ~ u o r ~ c e n c e i

Figure 12 Relative fluorescence frequency distribution of nile red-stained cells cultivated (a) in dialysis tubing, (b) in spinner flasks on different days after inoculation

membrane capacity [/zF em "2]

f

.-.7,¸ [] a

[] b

;'3";;.; "/7/ii~' !!ili

///.~

Discussion and conclusions The experimental setup developed for the present inVestigations permitted the characterization of the cells and their local environment enclosed in dialysis tubing (DT) as a function of the initial cell density and cultivation time. For comparison, cultivations were carried out with the same inoculum in Bellco spinner flasks.

880

o

.

.

¢ / 7 ,

day Figure 13

Plasma membrane capacities of cells cultivated (a) in dialysis tubing, (b) in spinner flasks on different days after inoculation

Enzyme Microb. Technol., 1991, vol. 13, November

Hybridoma cells in dialysis tubing: C. Kgsehagen et al. membraneconductivity[mS cm"2] r

,,

50;; ;; ]

///~/

"//~, J

<,~.>/.

,/~//'!

<,~:;:

3O

-

";f6: ;'/,,., 4/,:: ,,,,.

>:'; i

>9;<:4 o

/"

b

,./,

;.//...'.i 10

[]

": 3

4

5

n

day Figure 14 Plasma membrane electrical conductivities of cells cultivated (a) in dialysis tubing, (b) in spinner flasks on different days after inoculation

The cells enclosed in DT were supplied with a sufficient amount of dissolved oxygen and substrate. No essential differences between low-molecular-weight medium component concentrations inside or outside the dialysis tubing were found. Thus, no concentration depletion within the DT occurred. On account of the sufficient substrate concentrations in the medium, no growth limitatlon prevailed in DT. At low initial cell concentration (1.3 × 106 cells ml-l), no significant differences were found between the cultivated reference cells enclosed in DT and in spinner flasks. In the concentration range of 5 × 10 6 and 1 × l0 s cells m1-1, no cell growth could be observed in DT. The total cell concentrations remained constant during the cultivation time, but the fraction of nonviable cells increased from 45% (at 5 × l06 cells m1-1) to 80% (at 1 x l08 cells ml- r). The specific glucose utilization rate and the specific lactate and ammonia production rates of the cells cultivated in DT were higher than those in the REF culture. This indicates a higher metabolic activity of the cells in DT. The lipids were enriched in the cytoplasm of the cells cultivated in Petri dishes during the entire cultivation time. In spinner flasks and in DT, they decreased on the third day of cultivation, whereby the reduction of the number of these cell storage components happen faster in DT than in REF cultures. This again supports the supposition of high metabolic activity of cells in DT cultures. The activities of the intracellular enzymes LDH, GOT, and GIDH, which were set free during cell damage, increased during cultivation. The specific activities of L D H and GOT, however, were lower in

DT cultures than in REF cultures. This indicates that cell damage by mechanical stress in DT cultures is less than in the stirred spinner cultures. The size distribution measurements by a flow cytometer indicate that cells in DT are smaller than those in the reference culture. The measurements of the membrane capacity with cell rotation technique support these results. The membrane folding of the cells was higher in DT than in the reference cultures. The concentrations of MAB increased during cultivation and were higher in DT than in the REF cultures. The specific MAB productivities with regard to the total cell concentrations were lower, and those with regard to the viable cell concentrations were higher, in DT than in the REF cultures. This could be explained by the higher metabolic activity of the cells in the DT cultures. Reuveny et al. 4 explained the increased MAB production with regard to the viable cells by the lysis of the plasma membrane, which sets MAB free from the cells. This, however, cannot explain the increase of MAB concentration during cultivation. Only if one assumes that there is an increase of cell concentration during their growth and a decrease at their death, provided these two rates are equal, i.e., the total and viable cell concentrations during the cultivation remained the same, the explanation of Reuveny could be accepted. Further investigations are necessary to evaluate the influence of the high cell density on the cell cycle.

Acknowledgement Part of these investigations were carried out in the frame of the BAP-project 0132 D. The authors gratefully acknowledge the financial support of the European Community.

References 1 2 3 4 5 6 7 8 9 10

Altshuler, G. L., Dziewulski, D. M., Sowek, J. A. and Belfort, G. Biotechnol. Bioeng. 1986, 28, 646-658 Calabresi, P., McCarthy, K. L., Dexter, D. L., Cummings, F. J. and Rotman, B. Proc. Am. Assoc. Cancer Res. Am. Soc. Clin. Oncol. 1981, 22, 302 Knazek, R, A., Kohler, P. O. and Gullino, P. M. Exp. Cell Res. 1974, 84, 251-254 Reuveny, S., Velez, D., Riske, F., Macmillan, J. D. and Miller, L. Deu. Biol. Standardization 1985, 60, 185-197 Kflsehagen, C., Dissertation, University of Hannover, 1990 Hopkinson, J. Bio/Technology 1985, 3, 225-230 K~isehagen,C., Schgged, K. Untersuchungdes Verhaltens yon Hybridom-Zellen bei hohen Zelldichten Chem. Ing. Techn. 1991, 63, 756 JAger,V. Dissertation, University of Hannover, 1988 Linz, F., Dissertation, University of Hannover, 1990 Freitag, R., Schiigerl, K., Arnold, W. M. and Zimmermann, U. J. Biotechnol. 1989, 11, 325-336

Enzyme Microb. Technol., 1991, vol. 13, November

881