Cytotechnology 5: 223-231, 1991. 9 1991 KluwerAcademic Publishers.Printedin the Netherlands.

Growth and protein production kinetics of a murine myeloma cell line transfected with the human growth hormone gene C.A. Mitchell 1, J.A. Beall 2, J.R.E. Wells 2 and P.P. Gray 1 1Department of Biotechnology, University of New South Wales, PO Box 1, Kensington NSW 2033, Australia; 2Department of Biochemistry, University of Adelaide, GPO Box 498, Adelaide SA 5001, Australia Received 21 August 1990; accepted in revisedform 14 November1990

Key words: cell culture, kinetics, Ig promoter/enhancer, plasmacytoma, recombinant protein production Abstract A model mammalian cell system for the production of recombinant proteins was investigated. Murine myeloma cells which had lost the ability to produce both heavy and light chain immunoglobulin molecules were transfected with a vector containing the immunoglobulin heavy chain promoter and enhancer elements linked to the human growth hormone gene. The growth kinetics of G32, a clonal isolate, were found to be similar to both the parent myeloma and hybridomas. However, production of hGH by G32 was growth associated, rather than as a secondary metabolite as is the case for hybridomas. In addition, G32 produced hGH at molar levels greater than most hybridomas.

Abbreviations: ELISA - Enzyme Linked Immunosorbent Assay, Ig - Immunoglobulin, MAb Monoclonal Antibody, X63 - Murine Myeloma Cell Line P3X63-Ag8.653

Introduction There are now many examples in which knowledge of genetic regulatory signals has been used to obtain high level expression of proteins from specific genes transfected into a variety of cells. The requirement of post-translational modification necessary for either the stability or biological activity of some proteins constrains the choice of host cell to eukaryotic lines. For these reasons we have chosen to investigate a model animal cell production system, specifically the murine myeloma P3X63-Ag8.653 (Keamey et al., 1979) transfected with a novel construct in which im-

munoglobulin heavy chain (IgH) promoter and enhancer elements have been linked to the human growth hormone (hGH) gene. Since X63 has lost the ability to produce both heavy and light chain Ig molecules, it is one of the most popular fusion partners currently in use for the production of hybridomas (Goding, 1986). X63 should therefore have resources available for the production of a protein from an introduced gene. As a plasmacytoma, X63 is expected to have effective secretion pathways in place. Ig promoter/enhancer sequences are recognised as being powerful regulatory systems, albeit cell type specific (Doyen et al., 1986). Since hGH is

224 relatively small and non-glycosylated, it is an ideal marker protein for initial studies.

Materials and methods Cell line The murine myeloma cell line, P3X63Ag8.653 (X63), was obtained from the American Type Culture Collection (CRL 1580). This cell line, derived from the MOPC-21 plasmacytoma, is a variant which no longer expresses either immunoglobulin heavy or light chains (Keamey et al., 1979)

Cells were routinely passaged every 2-3 days in 75 cm 2 T-flasks (Coming, NY, USA). The flasks were gassed with 5% carbon dioxide in air and incubated at 37~ Detailed studies on growth, metabolism and hGH productivity were performed in duplicate in 500 ml glass spinner vessels (Techne) with an approximate culture volume of 250-300 ml. Cells were cultivated in batch mode in the medium described above, and sampled every 8 hours for growth, metabolic and productivity studies. Separate cultures were sampled daily for extracellular and intracellular hGH levels. The headspace of each culture was gassed with 5% carbon dioxide in air every time a sample was taken. Cell numbers, hGH, glucose, and lactic acid levels were determined.

Transfection of X63 Analytical methods The transfection of the cell line X63 was carried out using electroporation (Chu et al., 1987). The cells were suspended in phosphate buffered saline (PBS) at a concentration of 10 6 cells/ml. 0.5 ml of cells were mixed with 10 gg of linearized pNIgGhGH (see Fig. 1) in a shock chamber (Biorad) and incubated at 4~ for 10 minutes. The cold chamber was exposed to 180 volts from a 90 gfd capacitor (Gene-Pulser, Biorad). The cells were left for 10 minutes at room temperature, then diluted with DMEM containing 10% Foetal Calf Serum (FCS) before plating into microtitre plates at a final concentration of 10 3 cells]well. The cells were incubated at 37~ with 10% CO2 mixed with air. After 24 hours, 450 gg/ml of G418 was added to each well. The medium containing G418 was changed every 3 days for a period of 2 weeks.

Cells and cell culture The medium was a 1:1 mixture of DMEM and Coon's F12 (Gibco) supplemented with 10% FCS (CSL, Australia), 5 x 10.5 M 2-Mercaptoethanol, and 1.8 g/l Sodium Hydrogen Carbonate. Antibiotics were not used.

Cell numbers and viability were determined using a haemocytometer and the trypan blue dye exclusion method. Glucose concentration in the supernatant was determined with a glucose analyser (YSI, Ohio, USA). An enzymatic assay kit (Boehringer Mannheim) was used to determine lactic acid concentrations. A sandwich type ELISA (Belanger et al., 1973) was developed to assay hGH (manuscript in preparation). Samples for the determination of intracellular hGH levels were prepared as follows: 400 gl of culture medium was spun down in a microfuge. The supematant was retained as the extracellular sample. The pellet was resuspended with an equal volume of PBS, then spun down again. Other experiments have shown that a single wash is sufficient to remove more than 80% of hGH from the cell pellet (results not shown). This supernatant was assayed as the wash sample. The pellet was resuspended in 50 gl of lysis buffer (0.14 M NaC1, 1.5 mM MgC12, 10 mM TrisHCl, 0.5% NP40) and made up to the original volume with PBS, mixed, centrifuged, and the supernatant set aside for intracellular hGH analysis. All samples were routinely stored at -20~

225

plgH

Enhancer

xb~,l EooR~ Xbal Avail V ~ Promoter/~"~

~ Cu '-

BamHI EcoRI 1 200bp I IgHPromoter

H

4 ~ ori

7

m

p

~

S

IgHEnhancer

Xhol

PyT

BamH~-~

Bam.ITII RSV-LTR A

Sail [

700bp

a

l

l

A m P ~ I ~ - ~

SaIl

' ~

pBR322/

k ) "~'~-.

Neo r

pER329

Neor

IgH promoter & enhancer Genomal human ~f~-------..../Sall Growth Hormone/// N DHFR

Neor Fig. 1. Construction of pNIgGhGH. The RSV-LTR promoter was removed from the plasmid pAV009/A+. The mouse [.t IgH promoter and enhancer elements were removed from the plasmid Pu and inserted into the modified pAV009/A+ plasmid to form pNIg. The unique BamHI site was used to insert the genomal hGH gene to create pNIgGhGH.

Culture parameters T h e f o l l o w i n g e q u a t i o n s w e r e u s e d to c a l c u l a t e a

variety of culture parameters. T h e g r o w t h rate w a s c a l c u l a t e d u s i n g standard formula:

the

226 g= log(x2/xl) / (t2-tl)

(1)

where g = growth rate (/hr) x = viable cell density (cells/ml) t = time (hr) A log mean average of the cell density (Equation 2) was required to calculate specific glucose utilisation rates and hGH production rates (Equations 3 and 4 respectively). Xm = (X2--X1) /

In (xffxl)

(2)

where x m = log mean cell density (cells/ml) Qs = (s2-sl) / ((1000)(t2--tl))(Xm))

(3)

where Qs = specific glucose utilisation rate (mg/ 106 cells.hr) s = glucose concentration (mg/l) Qp = ( p 2 - p l ) / ((t2-tl) (Xm))

(4)

where Qp = specific hGH productivity (gg/106 cells-hr) P = hGH supematant concentration (gg/ ml) Specific intracellular hGH levels were calculated as follows: spi = (1000 * Pi) / x

(5)

where spi = specific intracellular hGH (ng/106 cells) Pi = intracellular hGH (p.g/ml)

Results

ent in the plasmid for positive selection in mammalian cells. The construction of pNIgGhGH is shown in Fig. 1. The gIgH enhancer and promoter elements were isolated as XbaI/EcoRI and BamHI/ AvaII fragments respectively from the plasmid Pu (Grosschedl and Baltimore, 1985). A SalI restriction site was added to the XbaI end of the 700bp enhancer and an EcoRI site was added to the AvaII end of the 200bp promoter fragment by ligation of the appropriate linkers. The EcoRI ends of the promoter and enhancer were ligated together before being inserted into a modified version of expression vector, pAV009/ A § (Choo et al., 1986) in which the RSV long terminal repeat in front of the unique BamHI cloning site was removed as a BamHI/SalI fragment and replaced by the IgH enhancer/promoter BamHI/SalI fragment to create pNIg. The genomal hGH gene was inserted into the unique BamHI site to complete pNIgGhGH. The unique SalI site was used to linearize the plasmid prior to transfection into the X63 cell line.

Expression of hGH in X63 Linearized pNIgGhGH was introduced into the non-Ig secreting mouse myeloma P3X63Ag8.653 by electroporation. After 2 weeks in the presence of G418, the Neo R colonies were assayed for the expression of hGH by an ELISA. Of the Neo g colonies, 70% produced hGH. The positive colonies were expanded and the hGH production was measured using an RIA kit (Pharmacia). The levels varied between 10 n g - 2 gg/106 cells/24 hrs. The distribution of expression levels is shown in Fig. 2. The cell line G32 on initial test produced 2 gg/106 cells/24 hrs and was selected for further studies.

Construction of pNIgGhGH The plasmid pNIgGhGH contains the genomal human growth hormone (hGH) gene under the control of the mouse immunoglobulin (Ig) ~t heavy chain promoter and enhancer. In addition, the neomycin resistance (Neo g) gene is also pres-

Cell growth When grown in tissue culture flasks, G32 cells routinely reach a maximum viable cell density of 1.5-2 x 106 cells/ml, regardless of the inocula-

227

10

! 8

! !9

i.

I9

0.1

:;

i9

9149

r ct) 9

o

1.

CD

O

t9

CB tO

tion density (results not shown). In comparison, when grown in spinner vessels, the maximum cell density is 3-4 x 106 cells/ml, almost double that obtained in stationary culture (Fig. 3(a)). The cells grew from an inoculation density of 1 x 105 cells/ml to a maximum of 3.5 x 106 viable cells/ml after 6 days in spinner culture (Fig.3(a)). The doubling time in the exponential phase was 15-16 hours. The viable cell density continued to increase for about 6 days (Fig. 3(a)). This high cell density was maintained for approximately 24 hours, and then the cell viability dropped off steadily. Interestingly, the total cell density had not decreased significantly after 16 days in batch culture, even though the viability had dropped to less than 5% of the total cell count (Fig. 3(a)). The growth rate peaked after just 30 hours in culture, then dropped uniformly for a further 40 hours, stabilised briefly as the glutamine supply was exhausted (results not shown) and the cultures approached maximum cell density, then decreased to zero as the viable cell density started to decline (Fig. 3(c)).

om

,"

Glucose utilisation and lactate production

0.01

rO O O

-1r

pNIgGhGH

(50)

Fig. 2. Distribution of hGH production levels obtained from 50 different colonies. Each expanded colony was assayed for the amount of hGH produced by 106 cells over a 24 hr period. The levels varied between 10 ng-2 1.tg/106[24 hr.

The glucose concentration in fresh medium is 3.8 g/1. As expected, the absolute glucose level dropped sharply in the exponential growth phase, and continued to decrease more slowly during stationary and decline phases to about 0.5 g/1 after 16 days (Fig. 3(b)). This corresponds with results of Hu et al. (1987) for both hybridomas and swine cells. The glucose consumption rate (Equation (3)) was initially high at around 125 gg/106 cells.h, then decreased sharply during the first three days of the culture while the cells were growing exponentially, and stabilised at approximately 5 gg/ 106 cells-hr after five days as the culture reached maximum viable cell density (Fig. 3(c)). Lactate concentration in the supematant rose steadily during the initial stage of the culture when the specific glucose consumption rate was greater than 10 gg/106 cells.hr (Fig. 3(b)). Initially, as a result of the Crabtree effect and high startup

228 -

increase in accumulated hGI-I levels occurs during late growth and stationary phases of the culture, when the specific productivity is high. As the viability of the culture drops during the final 100 hours or so, the accumulated hGH level increases slowly and the specific productivity rate declines. A strong positive correlation (p = 0.01) was found between the viability of the culture and the specific productivity rates. Long et al. (1988) found a similar strict correlation between antibody synthesis and cell viability in hybridomas. This suggests that cells which devote more energy to maintenance are better able to produce other proteins.

i(1

~!

',a)

! 4.e-

~

(b)

30-

..g_~

60

0o~0-11~

.~,

.>- ~.

20-

~.,

940

~J

0 040 "

"o

o +~

13_ o ::L v

x-

~

k..% ~

Growth Rale

- 5o

o _~ "~

I10(9

'20

As

.~

Glucoso Uplake OnO0---,

~'-'~"~ lOO

~ ' ~ r~r* ' ~ ' ~ - ~ ' ~ 200 300

'

o

0

400 1 00

T i m e (h)

200

40O

300

Time (h)

Fig. 3. Growthkinetics of G32 in batch culture. (a) Viable and total cell densities as determined by trypan blue dye exclusion (b) Cumulative glucose consumption and lactic acid (lactate) production (c) Growth rates (see Equation (1)) and specific glucose utilisationrates (Equation(3)). All bars represent average values from duplicate cultures.

Fig. 4. hGH productionby G32 in batch culture. Cumulative hGH levels in the culture supematant as measured by ELISA and specific hGH productivitybased viable cell density (Equation (4)). All bars represent average values from duplicate cultures. 25

glucose concentrations, most of the glucose taken up by G32 cells is shunted through to lactate, instead of utilising it for energy production in the TCA cycle. Production of lactate effectively ceases when the glucose consumption rate levels off after 100 hours as the cells move into the latter stage of exponential growth. The glucose consumption rate does not appear to be directly related to the growth rate.

,

,

20

,

Specific

,

.

60

Intraceilular

-

~_

45

30

~

~o

~

~o

xz

"m ~ 5 --

0

-

-

,,~

"

=

=

=

1 O0

~

Intracellula,

~'

;

200 Time

;-

_,z, ~

0 300

(h)

Production of hGH

Fig. 5. Secretion of hGH by G32 in batch culture. Cumulative extracellular and daily intracellularhGH levels, as determined

hGH is secreted by G32 cells throughout the culture period. Figure 4 shows that the major

by ELISA, and specific intracellular hGH calculated using Equation (5). All bars represent average values from duplicate cultures.

229 Extracellular and intracellular levels of hGH are shown in Fig. 5. Human growth hormone which is stored intracellularly is almost negligible as a proportion of the total hGH in the supernatant. Specific intracellular levels calculated using Equation (5) are around 30-50 ng/106 cells for the duration of the culture. This may be significant in terms of secretion holdup, but these results do not confirm this.

Discussion

A murine myeloma cell line, X63, which has lost the ability to produce both heavy and light chain Ig molecules was transfected with a DNA construct containing the murine g Ig promoter/enhancer sequences upstream of hGH and a clonal line expressing hGH was chosen for further study (G32). The parent plasmacytoma, X63, does however, produce a small amount of messenger RNA for both heavy and light chain immunoglobulin genes, but both these mRNA species are shorter than the native versions (Wallach et al., 1982). This suggests that the lack of Ig production is probably associated with missing segments from both genes, as opposed to a regulatory problem. In the process of developing and characterising a new cell line, it is useful to draw comparisons between the new line and existing data for similar cell types such as the parent cell line, X63, and hybridoma lines derived from the parent. In agitated culture, G32 cells have a doubling time of 15 hours which is comparable to that reported for the parent line, X63 (Sato et al., 1987; Jager et al., 1988). However, the maximum cell density of G32 in spinners is 2 - 4 times that for X63 in 24 well plates, and G32 in static culture (Sato et al., 1988; Jager et al., 1988; C. Mitchell, unpublished results). If diffusion is limiting the maximum viable cell density, then it is probable that the growth rate will also be limited, particularly at higher cell densities. It is possible that the conditions which allow continued 'normal' metabolism may not coincide precisely with conditions required for cellular division. Thus,

the cells may grow at the same rate, but not continue to divide to the same level. In addition, the viable cell density of G32 continues to increase for 6 days in culture, in comparison to 3 - 4 days observed for hybridomas, derived from the parent, growing in shake flasks and internal-loop airlift reactors in medium with 10% FCS (Moo-Young and Chisti, 1988) and the parent itself in serum-free medium (Jager et al., 1988). Shacter (1989) suggests most hybridomas generate 1-100 gg monoclonal antibody (MAb)/ml of culture fluid in stationary flasks. Evans and Miller (1988) report that less than 10% of mouseX-mouse hybridomas examined produce more than 100 gg MAb/ml (0.6 nmol Mab/ml) in static culture. In these experiments, G32 produced around 20 ~tg hGH/ml (0.9 nmol hGH/ml). If one is comparing the effectiveness of a given promoter in two separate situations, involving two different proteins, then the comparison must be on a molar basis, since the figure of interest is the number of protein molecules produced. Thus, the immunoglobulin promoter/enhancer system in G32 is superior to a significant proportion of hybridomas. However, it should be noted that there are several intermediate steps where degradation could occur. In addition, it is possible that the levels of trans-activating factors required for maximum expression of Ig promoter/enhancer cis regions may be limiting in the X63 cell line. The addition of the recently found 3' IgH enhancer (Petterson et al., 1990) may increase production levels. Fully differentiated plasma cells in vivo are able to produce several thousand antibody molecules/cell second (Alberts et al., 1989). As an example, ten thousand antibodies per cell per second corresponds to around 9 gg/106 cells-hr. The specific productivity of G32, at 30-50 ng/ 106 cells.hr, is two orders of magnitude less. However, fully differentiated plasma cells devote so much energy towards antibody production that these cells are unable to divide and usually die within a few days. The bulk of monoclonal antibody production in hybridomas occurs during stationary phase, so

230 that the MAb can be thought of as a secondary metabolite (Gunnersen et al., 1989). In contrast, the specific hGH productivity of G32 seems to have two phases. The high specific productivity region extends across exponential growth and stationary phases. Lower specific productivity levels occur towards the end of the culture as the viability decreases. This study indicates that it is possible to produce significant levels of a heterologous protein in plasmacytoma cells, utilising transcriptional control elements from immunoglobulin heavy chain genes. It forms a basis for further optimisation of the system.

Acknowledgements This work was supported in part by the Commonwealth Special Research Centre grant to J.R.E. Wells. We thank R. Grosschedl for the plasmid Pu, K. Choo for the plasmid pRV009/A +, and R. Sturm for contribution to the construction of the vector pNIgGhGH. We also thank M. Stuart for the kind donation of antibodies for the hGH ELISA.

References 1. Alberts B, Bray A, Lewis J, Raff M, Roberts K and Watson JD (1989) Molecular Biology of the Cell, 2nd Ed. Garland Publishing, Inc. New York. 2. Belanger L, Sylvestre C and Dufour D (1973) Enzymelinked immunoassay for alpha-fet0protein by competitive and sandwich procedures. Clin. Chim. Acta 48: 15-18. 3. Choo KH, Filby RG, Jennings IG, Peterson G and Fowler K (1986) Vectors for expression and amplification of cDNA in mammalian cells: Expression of rat phenylalanine hydroxylase. DNA 5: 529-537. 4. Chu G, Hayakawa H and Berg P (1987) Electroporation for the efficient transfection of mammalian cells with DNA. Nucl. Acid Res. 15(3): 1311-1326. 5. Doyen N, Leblond-Francillard M, Holm I, Dreyfus M and Rougeon F (1986) Analysis of promoter and enhancer cell-type specificities and the regulation of immunoglobulin gene expression. Gene 50: 321-331. 6. Evans TL and Miller RA (1988) Large-scale production of murine monoclonal antibodies using hollow-fibre bioreactors. BioTechniques 6(8): 762-767.

7. Goding JW (1986) Monoclonal Antibodies: Principles and Practice, 2nd Ed. Academic Press, London. 8. Grosschedl R and Baltimore D (1985) Cell-type specificity of immunoglobulin gene expression is regulated by at least three DNA sequence elements. Cell 41: 885-897. 9. Gunnersen JM, Fowles LF, Snelgar JM, Mitchell CA, Patava JV, Osborne KJ and Mahler SM (1989) Downstream processing of monoclonal antibodies. Aust. J. Biotech. 3: 69-71. 10. I-Iu WS, Dodge TC, Frame KK and Himes VB (1987) Effect of glucose on the cultivation of mammalian cells. Develop. Biol. Standard 66: 279-290. 11. Jager V, Lehmann J and Friedl P (1988) Serum-free growth medium for the cultivation of a wide spectrum of mammalian cells in stirred bioreactors. Cytotechnology 1: 319329. 12. Keamey JF, Radbruch A, Liesegang B and Rajewsky K (1979) A new mouse myeloma cell line that has lost immunoglobulin expression but permits the construction of antibody-secreting hybrid cell lines. J. Immunol. 123: 1548-1550. 13. Long WJ, Palombo A, Schofield TL and Emini EA (1988) Effects of culture media on murine hybridomas: Definition of optimal conditions for hybridoma viability, cellular proliferation, and antibody production. Hybridoma 7(1): 6977. 14. Moo-Young M and Chisti Y (1988) In: Durand G, Bobichon L and Florent J (eds) Proceedings: 8th International Biotechnology Symposium, Societ6 Frangaise de Microbiologie, Paris, pp. 454-466. 15. Muller MM, Gerster T and Schaffner W (1988) Enhancer sequences and the regulation of gene transcription. Eur. J. Biochem. 176: 485--495. 16. Petersson S, Cook GP, Bruggeman M, Williams GT and Neuberger MS (1990) A second B cell-specific enhancer 3' of the immunoglobulin heavy-chain locus. Nature 344: 165-168. 17. Sato JD, Kawamoto T and Okamoto T (1987) Cholesterol requirement of P3-X63-Ag8 and X63-AgS.653 mouse myeloma cells for growth in vitro. J. Exp. Med. 165: 17611766. 18. Sato JD, Cao H-T, Kayada Y, Cabot MC, Sato GH, Okamoto T and Welsh CJ (1988) Effects of proximate cholesterol precursors and steroid hormones on mouse myeloma growth in serum-free medium. In Vitro Cellular and Developmental Biology 24(12): 1223-1227. 19. Scheidereit C, Heguy A and Roeder RG (1987) Identification and purification of a human lymphoid-specific octamer-binding protein (OTF-2) that activates transcription of an immunoglobulin promoter in vitro. Cell 51: 783-793. 20. Shacter E (1989) Serum-free media for bulk culture of hybridoma cells and the preparation of monoclonal antibodies. Tibtech 7: 248-253. 21. Wallach M, Ishay-Micheli R, Givol D and Laskov R (1982) Analysis of immunoglobulin mRNA in murine myeloma

231 cell variants defective in the synthesis of the light or heavy polypeptide chains. J. Immunol. 128: 684-690. 22. Weidle UH, Koch S and Buckel P (1987) Expression of antibody cDNA in murine myeloma cells: possible involvement of additional regulatory elements in transcription of immunoglobulin genes. Gene 60: 205-216.

Address for offprints: C.A. Mitchell, Department of Biotechnology, University of New South Wales, PO Box 1, Kensington NSW 2033, Australia