90
To test whether weakly stained or negative rCHO cells represent a non-vital
subpopulation of the original culture, a dual labeling method was developed to
simultaneously detect intracellular mAb with Cy2-antibody probe, and prefixation
viability using EMA, a fluorescent photoaffinity label that, after photolysis, binds
covalently to nucleic acids in cells with compromised membranes.2 EMA is excluded
from viable cells. Cells weakly stained for mAb (Fig. 4A) were impermeable to EMA
(Fig. 4B), indicating lack of mAb-specific fluorescence was not related to viability. In
some cases cells were found containing dual label, demonstrating nonspecific adsorption
of anti-mAb probe to dead cells (data not shown).
Since recombinant protein is synthesized at very high levels in transfected CHO it was
necessary to determine whether low levels of staining were due to poor access of
fluorescently labeled antibody probes to 'saturated' secretory compartments. In addition,
since detection of cells which stained weakly for mAb required long exposure times, a
reagent was needed to allow visualization of all cells in a field. BiP (IgG heavy chain
binding protein, Grp78) is an ER lumenal protein which functions in post-translational
protein folding3,4. BiP was selected as a convenient marker to delineate cells and to test
reagent accessibility to ER lumenal targets to determine whether low-/ non-staining cells
were an assay artifact. rCHO cells were dual labeled to simultaneously probe mAb and
BiP. Cells with low anti-mAb (Fig. 5C) signal had correspondingly high anti-BiP signal
(Fig. 5D), demonstating reagent access to secretory compartments.
91
5. Discussion
A highly specific and sensitive immunofluorescence method for intracellular
characterization of individual rCHO cells has been developed utilizing Cy2- and Cy3-
labeled antibody probes and the viability indicator EMA. Targets detected include
expressed mAb, constitutive BiP and viability status. Preliminary application of this
method proved valuable in detecting unexpected heterogeneity of intracellular mAb
across a population of rCHO cells generated by clonal selection. Immunofluorescence
offers the ability to rapidly profile cell lines at the single cell level during early stages of
process development. Early detection of unstable and/or non clonal lines may allow
channeling of costly development resources into cell lines of highest potential. In
addition, work is underway to combine this technique with quantitative imaging analysis
to allow prediction of cell line productivity very early in the development process.
92
6. References
1Urlaub, G., Kaas, E., Carothers, A.M., Chasin, L., (l983),Cell, 33:405-412.
2Haughland, R.P., ed. Handbook of Fluorescent Probes and Research Chemicals,
Molecular Probes, Eugene, OR., 6:150.
3Warren, G., (1987), Nature, 327:17-18.
4Bole, D.G. et al (1986) Posttranslational Association of Immunoglobulin Heavy Chain Binding Protein with
Nascent Heavy Chains in Nonsecreting and Secreting Hybridomas, Journal of Cell Biology. 102:1558-
1566.
ELECTROLYZED REDUCED WATER WHICH CAN SCAVENGE ACTIVE
OXYGEN SPECIES SUPRESSES CELL GROWTH AND REGULATES
GENE EXPRESSION OF ANIMAL CELLS.
S. Shirahata, S. Kabayama, K. Kusumoto, M. Gotoh, K. Teruya, K. Otsubo*, J.S.
Morisawa*, H. Hayashi** and K. Katakura
Graduate School of Genetic Resources Technology, Kyushu University, Hakozaki,
Higashi-ku, Fukuoka 812-81, Japan; *Nihon Trim Co. Ltd., Meiji Seimei Jusou
Building 6F, 1-2-13 Shinkitano, Yodogawa-ku, Osaka 352, Japan; **Water Institute,
Nisshin Building 9F, 2-5-10 Shinjuku, Tokyo 160, Japan.
Abstract:
Active oxygen species are considered to cause extensive oxidative damage to biological
macromolecules, which bring about a variety of diseases as well as aging. Reduced water
produced near cathode during electrolysis of water exhibits high pH, low dissolved
oxygen, extremely high dissolved molecular hydrogen, and extremely negative redox
potential values. Recently we found that strongly electrolyzed reduced water scavenges
active oxygen species and protects DNA from oxidative damage (Shirahata, S. et al.,
Biochem. Biophys. Res. Commun., 234, 269-274 (1997)). Electrolyzed reduced water
suppressed the growth of human normal fibroblast TIG-1, human lung adenocarcinoma
A549, and human uterine cervix cancer HeLa, indicating that reduced water affects the
signaling pathway of cell cycle. The expression of the interleukin-6 gene was enhanced
by reduced water as well as ascorbic acid, (+)-catechin and tannic acid when added to the
culture of human osteosarcoma MG-63 cells, suggesting that reduced water acts as a
reductant to cells.
Introduction
Active oxygen species such as singlet oxygen superoxide anion radical
hydrogen peroxide and hydroxyl radical (OH) are considered to cause extensive
oxidative damage to biological macromolecules (DNA, membranes, enzymes and so on),
which bring about a variety of diseases as well as aging (1, 2). Antioxidative enzymes
such as superoxide dismutase, catalase, and glutathione peroxidase can scavenge active
oxygen species. Daily intake of antioxidants such as vitamin C, vitamin E,
(+)-catechin is also important to protect cells from oxidative damage. In spite of these
protective mechanism, a chronic state of oxidative stress exists in cells because of an
imbalance between prooxidants and antioxidants. Recent heavy environmental pollution
seems to strengthen oxidative stress to our bodies.
Water is a most abundant compound on the earth and indispensable for existence of life.
More than a half century ago, domestic devices to reform water have been developed in
Japan. The principal is to separate reduced water near cathode from oxidized water near
anode by semipeamiable membrane during electrolysis of water. Electrolyzed reduced
water exhibits alkaline pH and negative redox potential. The devices to reform water are
admitted as therapeutic devices by the Japanese ministry of public welfare because
reduced water is effective to suppress abnormal fermentation in intestine and hyperacidity,
although the action mechanism is unclear. Based upon the interesting clinical
improvement of a variety of diseases by intake of reduced water, Hayashi proposed the
hypothesis “water regulating theory’ since 1985 (3). Recently we found that electrolyzed
reduced water contains high concentration of dissolved hydrogen (DH), scavenges active
93
O.-W. Merten et al. (eds.), New Developments and New Applications in Animal Cell Technology, 93-96.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
94
oxygen species and protects DNA from oxidative damage (4). We proposed a new
hypothesis that active hydrogen in reduce water may scavenge active oxygen. Here we
report that electrolyzed reduced water suppresses the cell growth and regulates the gene
expression in animal cells.
Materials and methods
Electrolysis of water. Ultrapure water produced by an ultrapure system (ULTRAPUR
LV-10T, TORAY, Tokyo) was added 0.1 g/l NaCl to elevate electrical conductance (EC)
to about 20 ms/m. The water was then electrolyzed with various voltages (0 - 40 V) by an
electrolyzing device (Type TI-7000S and TI-7000SL, Nihon Trim Co., Osaka) equipped
with platina-coated titanium electrode to produce reduced water which exhibited various
RP. Ultrapure water containing NaCl was sent to the electrolyzing device using a water
current pump at a rate of 0.5 1/min to 1.0 l/min. Electrolyzing voltage was changed to
several to several tens V and currents several to 10 A. RP, EC, dissolved oxygen (DO)
and DH were measured using a RP meter (type, HM-14P), a EC meter (CM-14P), a DO
meter (DO-14P) and a DH meter (DHDI-1) of Toa Electronics Ltd. (Tokyo) at 25°C. pH
was measured using a pH meter (Beckman, Type pHI32) at 25°C.
Assay of scavenging effect of reduced water against superoxide onion radical. The
scavenging effect of reduced water against was examined by hypoxanthine-xanthine
oxidase system using luciferine analog and a chemiluminometer as described previously
(4). All reactions were performed in neutralized condition using 40 mM sodium
phosphate buffer (pH 7.0).
Measurement of growth curve of human normal and cancer cell lines. Normal human
fibroblast cell line TIG-1, human lung adenocarcinoma cell line A549 and human uterine
cancer cell line HeLa were inoculated into 24-well microplates and cultivated in 10% fetal
calf serum-MEM medium containing electrolyzed reduced water at 37°C under an
atmosphere of 5% Since strongly electrolyzed reduced water often contains HOCl
which was produced near anode, the effects of reduced water on the cell growth were
compared with those of HOCl of the same concentration. Tap water must contains more
than 0.1 ppm of residual chlorine. The concentration of HOCl was determined by o-
tolidine method.
Analysis of the interleukin-6 gene expression in MG-63 cells. Human osteosarcoma
MG-63 cells were cutured in 5% FBS-MEM medium at 37°C under an atmosphere of 5%
The cells were treated in PBS (+) containing 2.7% reduced water, 0.1 µM ascorbic
acid, 0.33 µg/ml (+)-catechin or 40 µM tannnic acid in the presence of
for 2 hour. RT-PCR was performed after extraction of mRNA.
Results and discussion
Characteristic of electrolyzed reduced water
Strongly electrolyzed reduced water exhibited high pH, low DO, high DH and extremely
negativeRP values. Marked changes in these values occur in water after electrolysis. For
example, the value of RP +350 mV in water before electrolysis changes to -659 mV after
electrolysis; 4.08 of pH changes to 10.47; 4.46 mg/1 of DO changes to 3.38 mg/l and
0.0046 mg/l of DH changes to 0.467 mg/l, respectively. It should be noticed that DH
value is higher in reduced water than in the original water by two orders of magnitude.
Strongly reduced water completely scavenged produced by hypoxanthine-xanthine
oxidase system (4). Reduced water was shown to scavenge It prevented single-
strand breakage of plasmid DNA caused by the Cu(II)-catalyzed oxidation of ascorbic
acid, suggesting that reduced water can also scavenge OH and The scavenging
activity of reduced water against was stable in closed bottle at 4 °C for about a month
95
but lost by opened autoclave or closed autoclave in the presence of tungsten trioxide
which is knwon to be a specific adsorbent of active hydrogen.
Suppression of the growth of several human cell lines by electrolyzed reduced water.
Electrolyzed water exhibits strong scavenging activity against active oxygen species. In
order to examine the effect of reduced water on animal cells, the growth of human normal
fibroblast TIG-1, lung adenocarcinoma A549, and human uterine cervix cancer HeLa
were examined in the medium containing reduced water. Since reduced water contained
HOCl which was produced near anode, effect of the same concentration of HOCl was
examined as control. As shown in Fig. 1, the growth of TIG-1 was suppressed by
reduced water and the cells reached confluence at lower cell density in medium containing
reduced water than in medium containing no reduced water. The morphology of cells in
medium containing reduced water seemed to be normal. One and 0.5 ppm of HOCl
slightly stimulated the growth of TIG-1 cells, suggesting that oxidative stress affects the
growth of the cells. The growth of lung cancer A549 cells was remarkably suppressed by
reduced water. HOCl did not affect the cell growth of A549 cells. The growth of uterine
cancer HeLa cells were significantly suppressed by reduced water, but HOCl did not
affect the growth of HeLa cells. These results suggested that reductive stress by reduced
water may affect the signaling pathway of cell cycle to slower the rate of cell division.
Sensitivity of animal cells to reductive stress may be different depending upon cell types.
96
Induction of the interleukin-6 gene expression in human osteosarcoma MG-63 cells by
electrolyzed reduced water and reducing substances.
In order to examine if reduced water regulates some gene expression in animal cells,
we examined the expression of the interleukin-6 gene in human osteosarcoma MG-63
cells. Super-induced MG-63 cells produce interferon The expression of interleukin-
6 gene was examined by RT-PCR method. Treatment of MG-63 cells with reduced water
induced the expression of the interleukin-6 gene (Fig. 2). Since reducing substances such
as ascorbic acid, (+)-catechin, and tannic acid also induced the expression of the
interleukin-6 gene, the effect of reduced water may be due to its reducibiltiy.
Here we first demonstarted that electrolyzed reduced water can regulate the growth of
animal cells and gene expression. Water can permeate everywhere in the body and
penetrates every membrane in eluding the blood-brain barrier. To neutralize the toxic
action of active oxygen species, electrolyzed-reduced water may be an ideal and very
powerful antioxidant and may be applied for prevention and therapy of various diseases.
Further intensive investigation on the effect of reduced water on cell biology, immunology
and oncology should be promoted.
REFERENCES
1. Sohal, R. S. and Weindruch, R. (1996) Oxidative stress, caloric restriction, and
aging.Science 273, 59-63.
2. Feig, D. I., Reid, T. M. and Loeb, L. A. (1994) Reactive oxygen species in
tumorigenesis. Cancer Res. 54, 1980s-1984s.
3. Hayashi, H. (1995) Water, the chemistry of life, part IV. Explore 6, 28-31.
4. Shirahata, S., Kabayama, S., Nakano, M., Miura,T., Kusumoto, K., Gotoh, M.,
Hayashi, H., Otsubo, K., Morisawa, S., and Katakura, Y. (1997) Electrolyzed-
reduced water scavenges oxygen species and protects DNA from oxidative damage.
Biochem. Biophys. Res. Commun. 234, 269-274 (1997).
5. Kabayama, S., Osada, K., Tachibana, H., Katakura, Y. and Shirahata, S. (1997)
Enhancing effects of food components on the production of interferon from animal
cells supressed by stress hormones.Cytotechnology 23, 119-125.
DIFFERENTIAL GENE EXPRESSION OF CYTOCHROME P450 IN
IMMORTALISED HEPATOCYTE CELL LINES
H.T.KELLY1, K.ANDERSON2, E.HILL1, H.GRANT2, C.MacDONALD1
Biological Sciences1, University of Paisley and Bioengineering Unit2, University of
Strathclyde, Glasgow, UK.
INTRODUCTION
Cytochrome P450 isoenzymes and the associated mixed function oxidase system are
important factors in the metabolism and detoxification of xenobiotics in the liver.
However these functions are amongst the most labile activities in primary cultures of
hepatocytes (Steward et al., 1985; Wortelboer et al., 1990). As an alternative approach
to the use of primary hepatocyte cultures for the study of xenobiotic metabolism and
toxicity several immortalised rat hepatocyte cell lines have been generated by either
calcium phosphate precipitation (MacDonald et al, 1994) or electroporation of SV40
DNA (Yin et al., 1996,). In order to examine the retention of drug metabolising enzyme
activity in our immortalised lines we employed the reverse transcriptase polymerase
chain reaction (RTPCR) to determine the expression of P450 isoform mRNA, RNA
arbitrarily primed polymerase chain reaction (RAP-PCR) to examine differential gene
expression of cytochrome P450 and immunoblotting to determine the presence of the
P450 protein.
METHODS
1.Cell Culture: Hepatocytes were isolated from male Sprague Dawley rats by
collagenase perfusion and transfected with SV40 early region DNA by either calcium
phosphate precipitation ('C' lines) or electroporation ('LQC' and 'KC' lines) and cultured
as described previously (MacDonald et al., 1994; MacDonald and Willett, 1997; Yin et
al., 1996).
2.RNA Preparation: Total RNA from freshly perfused hepatocytes and immortalised
cell lines was prepared using the total RNA isolation reagent TRIzol (Life
Technologies). The RNA obtained was subsequently used for RTPCR and RAP-PCR.
3. First Strand Synthesis: Total RNA (5µg/sample) was converted to cDNA using the
Superscript First Strand Synthesis kit and method (Life Technologies). Oligo(dT)12-18
was used to prime the first strand synthesis of all cDNAs prepared in this study.
4. PCR: A 10mM primer mix specific for either 2B1, 2B2 or 3A1 isoforms of CYP450
was used here (Omiecinski et al., 1990). To a thermowell tube (Costar) 5µl of each
primer mix was added. In addition 5µl of 10xReaction buffer [500mM KCl, 15mM
100mM Tris HCl (pH9 at room temp.)] (Pharmacia Biotech), 4µl 10mM dNTP
mix (10mM dATP, dTTP, dCTP, dGTP), 30.75µl and 0.25µl Taq DNA
polymerase (Pharmacia Biotech) were added to each reaction and pulse spun. cDNA
solution (5µl) was subsequently added to each tube and the total reaction overlayed
97
O.-W. Merten et al. (eds.), New Developments and New Applications in Animal Cell Technology, 97-99.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
98
with mineral oil. The reaction was pulse spun and placed in the thermocycler
(Techne) to undergo a series of denaturation, annealing and elongation cycles as
follows: 1 cycle @ 93°C for 4min, 30 cycles @ 93°C for 1min, 54°C for l.5min and
72°C for 1min and 1 cycle @ 72°C for 5min. The resulting PCR products were resolved
on a 1.5% agarose DNA gel and visualised by ethidium bromide staining. A 100bp
DNA ladder was used as a sizing marker.
5.RAP-PCR: For the purpose of RNA fingerprinting a RAP-PCR kit (Stratagene ) was
used. cDNA synthesis was performed as described above using an 18mer arbitrary
primer Al (Stratagene) to replace oligo(dT). To a thin wall tube of 10x reaction
buffer, _ mix (25mM each dATP, dTTP,
dCTP, dGTP), Taq DNA polymerase RAP-PCR primer Al,
of 1:10 cDNA dilution and were added. The reaction
was gently mixed, pulse spun and overlayed with mineral oil. RAP-PCR was
performed using the following cycling conditions: 1 cycle of 1min @ 94°C, 5 min @
36°C and 5min @ 72°C; 40 cycles of 1min @ 94°C, 2min @ 50°C and 2min @ 72°C
and finally 1 cycle @ 72°C for 10min. To analyse the product of the above reaction
was added to of a stop solution containing 95% formamide, 20mM EDTA, 0.05%
Bromophenol Blue and 0.05% xylene cyanol FF and the mixture heated to 80°C for
2min. of the reaction was run in a 6% acrylamide sequencing gel in 1xTBE
buffer.
6. Western Blotting: All microsomes used in the immunoblotting were prepared from
cell homogenates by ultracentrifugation. The samples were separated by
electrophoresis, transferred onto nitrocellulose and stabilised in a 3% (w/v) gelatin
solution. To immunostain 20ml of relevant primary antibody (a gift from
Glaxo/Wellcome) were added for 2h. The protein samples on the nitrocellulose were
then washed twice in TTBS (Tween 20 Tris Buffered Saline) and the secondary
antibody, goat conjugated with alkaline phosphatase (Biorad), added for 2h.
The samples were subsequently washed once in TTBS and once in TBS (Tris Buffered
Saline). An alkaline phosphatase detection kit (Biorad) was used for colour
development after which the nitrocellulose paper was washed with 3 rinses of water and
dried between sheets of blotting paper before being photographed. The immunoblots
were analysed using densitometry on a Biorad GS 690 scanning densitometer using
Biorad Molecular Analyst software.
RESULTS
The presence of PCR products specific for the 2B1, 2B2 and 3A1 isoforms of
cytochrome P450 was detected in primary hepatocytes using RTPCR. In contrast
however larger PCR products, specific for the 2B1, 2B2 and 3A1 primers used here,
were identified in both the calcium phosphate transfected and electroporated
immortalised hepatocyte cells.
Immunoblotting with antibodies specific for 1A1, 2B(1/2), 3A(l-9), 2C6 and 2C11
isoforms of P450 demonstrated that freshly isolated hepatocytes had a higher level of
staining for all isoenzyme forms compared with that observed for 24h cultured primary
hepatocytes. Immortalised cells, in addition to staining positively for the CYP450
99
isoenzyme proteins examined, also demonstrated multiple stained products smaller than
that expected for the proteins studied. Using densitometry analysis the majority of
immortalised lines were found to stain positively for CYP1A1, 2B(l/2), 2C6, 2C11 and
3A(1-9) albeit at levels much lower than that observed in the primary cultures.
In order to compare the RNA present in both the primary hepatocytes and immortalised
cells, arbitrarily primed PCR (McClelland et al., 1995) was used to generate RNA
fingerprints of these two hepatocyte populations. Using this technique RAP-PCR
products were identified in the immortalised hepatocytes that were absent in the
primary hepatocytes.
CONCLUSIONS
RTPCR analysis demonstrated that mRNAs encoding the cytochromes 2B1, 2B2 and
3A1 are present in our immortalised lines although the product amplified appeared
larger than that observed with primary hepatocytes. Cytochrome P450 isoenzyme
proteins were detected in both freshly isolated and 24h cultured primary hepatocytes
with the levels of isoenzyme lower in the latter cells. In addition to giving the expected
protein band, multiple bands were identified for the immortalised cells suggesting
cross-reactions had occurred. This was unexpected as monoclonal antibodies were used
throughout and no cross reaction was observed with other proteins in freshly isolated or
primary rat hepatocytes. These proteins were also smaller in the immortalised cells and
some degradation of the cytochrome P450 may have occurred.
RNA fingerprinting, using RAP-PCR, identified differences in the RNA transcripts
from immortalised cell lines compared with those obtained from primary hepatocytes.
This suggests the occurrence of differential gene expression in our immortalised
hepatocytes and highlights RAP-PCR as a useful screening technique for identifying
those immortalised lines which retain the most 'normal' phenotype in relation to primary
hepatocytes.
REFERENCES
1. MacDonald C., Vass M., Willett B., Scott A., Grant M.H. (1994). Expression of liver functions in
immortalised rat hepatocyte cell lines. Human Experimental Toxicology 13, 439-444.
2. MacDonald C and Willett B. (1997). The immortalisation of rat hepatocytes by transfection with SV40
sequences. Cytotechnology 23, 161-170.
3. McClelland M., Mathieu-Daudi F., Welsh J., (1995). RNA fingerprinting and differential display using
arbitrarily primed PCR. TIG 11, 242-246.
4.Omiecinski C.L., Hassett C., Costa P., (1990). Developmental expression and in situ localisation of the
phenobarbital-inducible rat hepatic mRNAs for cytochromes CYP2B1, CYP2B2, CYP2C6 and CYP3A1.
Molecular Pharmacology 38, 462-470.
5. Steward A.R., Dannan G.A., Guzelian P.S., Guengerich F.P., (1985). Changes in the concentration of
seven forms of cytochrome P450 in primary cultures of adult rat hepatocytes. Molecular Pharmacology 27,
125-132.
6. Wortelboer H.M., de Kruif C.A., van Lersel A.A.J., Falke H.E., Noordhoek and Blaauboer B.J. (1990).
The isoenzyme pattern of cytochrome P450 in rat hepatocytes in primary culture, comparing different
enzyme activities in microsomal incubations and in intact monolayers. Biochemical Pharmacology 40, 2525-
2534.
7. Yin L., MacDonald C., Grant H. (1996). Electroporation conditions for retention of rat hepatocyte
viability. The Genetic Engineer and Biotechnologist 16. 27-34.
POST-TRANSCRIPTIONAL CONTROL OF RECOMBINANT GENE
EXPRESSION: mRNA RETARGETING AND REGULATION OF
EXPRESSION
K.A. PARTRIDGE1, A. JOHANNESSEN2, A. TAULER3, I.F. PRYME2
and J.E. HESKETH1,*
1Intracellular Targeting Group, Rowett Research Instititue, Bucksburn,
Aberdeen, AB21 9SB, UK; 2Department of Biochemistry and Molecular
Biology, University of Bergen, Arstadveien 19, 5009 Bergen, Norway;
3Unit of Biochemistry, Faculty of Pharmacy, University of Barcelona,
Diagonal 643, 08028 Barcelona, Spain; *author for correspondence.
1. Introduction
Mammalian cell lines are used to produce commercially significant amounts of
recombinant proteins in cell factories. There are numerous examples where normally
secreted proteins are being produced in this way (e.g. Lin et al, 1986), as well as a
number of commercial vectors for the secretion of heterologous proteins. However such
systems have been used only for the production of secreted proteins rather than
proteins which are normally intracellular. The ability to secrete intracellular proteins
would greatly increase the range of potential products which could be made by cell
factories and therefore the secretion of intracellular proteins in mammalian expression
systems is an important aim in biotechnological research (James and Simpson, 1996).
Secreted proteins are synthesised on the endoplasmic reticulum (ER) and the secretion
of an intracellular protein from a cell factory would require as a first step the
redirection of the mRNA for synthesis of the protein to the ER.
2. Experimental Approach
Targeting of mRNAs coding for secreted and membrane proteins is achieved by a
signal peptide in the nascent polypeptide chain whilst there is growing evidence for
targeting of mRNAs coding for intracellular proteins to different cytoplasmic domains
by signals in the 3'untranslated region (3'UTR) (Hesketh, 1996). The feasibility of
retargeting a mRNA coding for a recombinant intracellular protein to the ER was
investigated using stable transfected cell lines expressing gene constructs containing
the reporter sequences and luciferase linked to different combinations of 5'
and 3' signals. In particular we have studied the effects of adding the 5'albumin signal
sequence in combination with either the native globin 3'untranslated region (3'UTR) or
the 3'UTR from the c-myc mRNA since the latter is known to contain a signal for
perinuclear localisation and association with the cytoskeleton (Veyrune et al, 1996).
The subcellular localisation and stability of the transcripts were determined.
101
O.-W. Merten et al. (eds.), New Developments and New Applications in Animal Cell Technology, 101-103.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
102
3. Results and Discussion
Expression of the exogenous gene in the stable transfected fibroblast cell lines was
assessed by Northern hybridisation to detect globin transcripts. The highest levels of
globin expression from all four genes was greatest in cells at 70% confluence and all
subsequent experiments were carried out at that stage of growth. Detergent/salt
fractionation (Vedeler et al, 1991; Veyrune et al, 1996) was carried out to investigate
the compartmentation of the transcripts between free polysomes, cytoskeletal-bound
polysomes and membrane-bound polysomes. Each cell line was fractionated,
polysomes pelleted, and then RNA extracted from the pellets and analysed by Northern
hybridisation for abundance of globin transcripts.
Globin transcripts with either the native 3'UTR or with the 3'UTR of c-myc were found
to be present predominantly in free/cytoskeletal-bound polysomes. In contrast, globin
transcripts with the native globin 3'UTR and the albumin signal sequence were found
at highest abundance in the membrane-bound polysomes derived from the ER. Globin
transcripts with both albumin signal sequence and c-myc 3'UTR showed a markedly
lower enrichment of the mRNA in the ER fraction. Reprobing of filters with the 18S
rRNA probe allowed correction for RNA loading and quantification of the
hybridisation data per unit RNA. This data confirm that the addition of the albumin
signal sequence redirects the globin mRNA to the ER (p=0.01). The signal sequence
redirected globin with the c-myc attached (p<0.05) but presence of both c-myc 3'UTR
and the signal sequence reduced the extent to which redirection occurred (p<0.05).
The data show that the addition of the signal sequence from rat albumin leads to
efficient redirection of the globin mRNA with the native 3'UTR to membrane-bound
polysomes. The presence of the c-myc 3'UTR reduced the extent of redirection,
suggesting that there is competition between 5' and 3' localising signals in the
chimaeric mRNA. Measurement of mRNA stability showed that the addition of the
signal sequence does not destabilize the mRNA. Translational efficiency was found to
be unaffected by addition of the signal sequence or replacement of the native 3'UTR by
that of c-myc.
Overall the results show that it is possible to retarget a mRNA from free/cytoskeletal-
bound polysomes to the ER while maintaining stability and translational efficiency but
that removal of 3'UTR sequences may be necessary in order to achieve efficient
retargeting. This is a prerequisite for secretion of an intracellular protein and is the
first step in developing appropriate technology to achieve secretion of intracellular
proteins from a cell factory; as such it has important implications for the development
of vectors to promote secretion of intracellular proteins from cell factories.
Acknowledgements: This work was supported by a grant from the European Union
Biotechnology Programme (Bio2-CT94-3069) and the Scottish Office Agriculture,
Environment and Fisheries Department. We thank Eli Johannesen and Gill Campbell
for excellent technical assistance.
103
References:
Hesketh, J.E. (1996) Sorting of mRNAs in the cytoplasm: mRNA localisation and the cytoskeleton. Expt. Cell
Res. 225, 219-236.
James, J. and Simpson, B.K. (1996) Application of enzymes in food processing. Crit. Rev. Food Sci. Nutr. 36,
437-463.
Lin, F-K., Lin, C-H., Lai, P-H., Browne, J.K., Egrie, J.C., Smalling, R., Fox, J.M., Chen, K.K., Castro, M. and
Suggs, S. (1986) Monkey erythropoietin gene: cloning, expression and comparison with the human
erythropoietin gene. Gene 44, 201–209.
Vedeler, A., Pryme, I.F. and Hesketh, J.E. (1991) The characterization of free, cytoskeletal and membrane-bound
polysomes in KrebII ascited and 3T3 cells. Mol. Cell Bioch. 100, 183-193.
Veyrune, J-L., Campbell, G.P., Wiseman, J., Blanchard, J-M. and Hesketh, J.E. (1996) A localisation signal in
the 3'untranslated region of c-myc mRNA targets c-myc mRNA and beta-globin reporter sequences to the
perinuclear cytoplams and cytoskeletal-bound polysomes. J. Cell Science 109, 1185-1194.
TRANSIENT TRANSFECTION IN MAMMALIAN CELLS
A basic study for an efficient and cost-effective scale up process
E.-J. Schlaeger , J.Y. Legendre*, A..Trzeciak, E.A. Kitas, K. Christensen, U. Deuschle
and A. Supersaxo
F. Hoffmann La Roche Ltd. PRPN-G, PRPV CH-4070 Basel, Switzerland
*UPSA Laboratoires, 128 rue Danton, 92506 Rueil-Malmaison
Abstract
The aim of the presented work was to develop a cost-effective and easily scaleable
transient transfection system with mammalian cells grown in serum-free suspension
culture. For this purpose the cationic polyethylenimine (PEI) and the novel hybrid
molecule dioleoyl-melittin (DOM) were used. Both substances are highly efficient
transfection reagents for mammalian cells and have been described recently.
The transfection complexes were made directly within the cell culture by consecutively
adding plasmid and PEI (direct method). Alternatively, the DNA-PEI transfection
complexes were performed in fresh medium (1/10 culture volume) and then added to the
cells (indirect method).
Here we have optimized the PEI-mediated transfection conditions for HEK293 and
293EBNA cells grown in serum-free suspension culture for large scale experiments.
In a second part of the investigation the efficiency of gene expression was studied in
labscale as well as in fermentor scale by combining DOM with PEI to form
transfection complexes.
Our data show the PEI-mediated transfection into EBNA cells can be used as a cost-
effective transfection system in spinner culture scale up to 2-5L, to produce milligram
amounts of recombinant protein. Fermentor scale experiments, however, are less
efficient because the PEI- mediated transient gene expression is inhibited by the
conditioned medium. Transfection experiments performed in the present of both DOM
and PEI exhibited synergistic enhancement of gene expression in EBNA cells.
The synergistic enhancement between DOM and PEI at low dose DNA (0, l-0,2µg/ml)
was seen in spinner scale as well as in 10L and 23L fermentor
Keywords : Transient transfection, HEK293 cells, serum-free, gene
expression
Introduction
Efficient gene delivery in gene therapy using non-viral vectors where the DNA is
generally complexed with cationic lipids or polycationic polymers has recently become
of great interest (1, 2). More recently some of these transfection agents were
conjugated with proteins or peptide in order to target then to specific cell types (3, 4).
In addition to being important for gene therapy protocols improved non-viral gene
delivery techniques are also of significant interest in gene expression in cell culture.
105
O.-W. Merten et al. (eds.), New Developments and New Applications in Animal Cell Technology, 105-112.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
106
Obtaining stable cell clones is generally a time consuming processs lasting from 4-6
weeks untill 6 months. In order to speed up the whole process of expressing
milligram amounts of recombinant proteins we were interested in scaling up the
transient gene expression system in suspension culture.
Here we have optimized the PEI-mediated transfection conditions of HEK293(EBNA)
cells grown in serum-free suspension culture for large scale experiments. In a second
part of the investigation the efficiency of gene expression was studied by combining
dioleoyl-melittin (DOM), which represents a new class of peptide-based reagent for
efficient transfection of mammalian cells, with PEI to form transfection complexes
(5,6).
Materials and Methods
Cell culture
Cell line: HEK 293EBNA (Invitrogen) adaptated to growth in suspension
Medium: HL serum-free, Ca-free with transferrin, insulin, Primatone RL, Pluronic F68,
fructose, maltose) (7 ).
Feed solution.: Mixture of gln, glucose, essential and non-essential. amino acids,
vitamins, Primatone RL
Culture vessel: Spinner flasks (Bellco), 12 L bioreactor, stirred (MBR)
Transfection
Transfection reagents: Polyethelenimine (Fluka), dioleoyl-melittin (Roche)
Reporter gene: Extracellular domain of human TNF receptor p55 (TNFRp55).
Plasmid: pREP 7-TNFRp55 (EBV-based expression vectors containing the EBV ori
P and the EBNA-1 gene which replicate extrachromosomally in dividing cells,
Invitrogen).
Transfection assay
HEK293(EBNA) cells grown to a cell titer of 0,6-1.0x10e6cells/ml were centrifuged,
washed once and resuspended in fresh medium. The cell titer was adjusted to 5-6xlOe5
cells/ml.
Preparation of the transfection complexes
Direct method: The DNA-PEI transfection complexes were preformed
directly in the cell culture by adding DNA and PEIconsecutively
Indirect method: The DNA-PEI transfection complexes were preformed
in fresh medium (1/10 volume) and then added to the cells.
The DNA-DOM-PEI transfection complexes were prepared using the indirect method
Incubation
The transfected culture was grown for 2-5 days.After the second day the culture was fed
once daily with a nutrient mixture of of sugars, amino acids and vitamins.
107
Product analysis
The recombinant TNFRp55 were measured by ELISA.
Results and Discussions
First we adaptated the adherently growing HEK293 cells and the subclone 293EBNA
to serum-free growth in suspension culture using HL medium.
In order to quickly generate a large set of data on the transient gene expression,
experiments were performed in 12 well dishes by transferring 1 ml suspended cells into
each well. The transfection complexes were then formed by adding first the plasmids
followed by the cationic polymer PEI directly into the culture (direct method). First the
PEI nitrogen : DNA phosphate ratio was examined, which is very important for high
transfection efficiency (6). The relatively high ratio of 26:1 which was found to be
optimal could partially explained by the DNA-PEI complex formation in the culture
in the presence of cells.
This ratio was subsequently used for gene delivery experiments in spinner culture. Fig.
1 shows the expression levels from an experiment in which 1µg/ml DNA was
complexed either by using the direct method or by the indirect method in 1/10 volume
of fresh medium. Two clear results were obtained. First the expression levels in EBNA
cells were significantly higher compared to 293 cells if the direct method was used. This
could be due to the higher growth rate observed during the post-transfection time, (see
the final cell density). In contrast, the 293 cell titer is lower after 3 days incubation.
Secondly, the transient expression of TNFRp55 displays a significantly lower titer if
the indirect method was used, indicating that the conditions of the separately prepared
DNA-PEI complexes is suboptimal.
108
In an additional experiment the scaleability of the direct method was investigated using
1 µg/ml plasmid. In Table 1 we demonstrate scaleability of the spinner culture system.
No difference in expression levels were observed if increased culture volume up to 2L
was used in corresponding spinner flasks. From this finding it is evident that milligram
amounts of recombinant proteins can be produced in a cost-effective way using this
simple indirect method of DNA-PEI complex formation.
A lot of effort was invested to reduce the DNA concentration in transient transfection
experiments further. From previous experiments we know that the transient gene
expression was always slightly if PEI (Mw 25kDa) was used instead of PEI (800kDa)
higher compared to data obtained with PEI Mw 800 kDa.
Therefore, studies were performed using both 25 and 800 kDa PEI to determine the
optimal ratios of PEI: DNA with respect to plasmid concentration and the method used
for complex formation. The results clearly shows that in the presence of 1µg/ml DNA
the ratios of PEI:DNA is different depending on the method and DNA concentration, see
Table 2. While the direct method again shows the ratio of 26:1 the indirect complex
formation shows a N : P ratio of 13: 1. The latter finding confirms results described in
the literature (6). Highly remarkable is the fact that at low DNA concentration
(0,2µg/ml) high expression levels were measured if the amount of PEI was increased .
The optimal ratio of PEI: DNA was found to be 33 : 1. Table 2 shows the TNFRp55
titer of transient transfection experiments with different DNA concentrations and
different molecular weights of PEI ( 25 and 800 kDa), using the direct and indirect
method.
109
Before each transfection experiment 293EBNA cells were transferred
into fresh medium. This step is very important for the DNA uptake process as is
shown in the next experiment. EBNA cells were washed and diluted into fresh medium,
followed for 48 hrs. incubation. During culturing the medium become conditioned.
After 24 and 48 hrs. incubation cells were adjusted to a cell titer of 5×10e5 c/m1 with
the corresponding spent medium and transfected using the indirect method with DNA-
PEI complex formation in 1/10 culture volume with fresh medium. In the control
sample, cells from the 48 hrs. time point were carefully transferred into fresh medium
and then transfected as described. The results in Fig.2 clearly demonstrate that transient
gene expression is strongly inhibited by condioned medium depending on the age of the
culture. However, the cells are still competent for DNA uptake as illustrated by the
control result.The inhibition ot transient transfection by conditioned medium presents a
serious problem in large scale experiments.
To overcome this difficulty we have used the combination of two transfection agents
dioleyl-melittin (DOM) and PEI. DOM is highly toxic in serum-free suspension culture
if the described DNA-DOM ratio of 1 : 10 was used. By lowering the DNA
concentration from 1µg/ml to 0.1-0.2 µg/ml the toxic effect is drastically reduced but
no TNFRp55 could be measured. However, if PEI was also added to the DNA-DOM
complexes (here we used the 25 and the 800 kDa polymer) a significantly increase of
the expression levels was observed as shown in Fig.3. This clearly demonstrates a
synergistic enhancement of transient transfection efficiency in the presence of DOM and
PEI.
In a set of experiments we optimized the ratios of DNA-DOM-PEI.
Conditions could be found which show high transfection efficiency at low DNA
concentration. Furthermore, the inhibitory effect of conditioned medium could be
significantly reduced by using DOM and PEI together. From this it is evident that if
no medium exchange is performed in a large scale transient transfection experiment
DOM and PEI must both be used and the culture must be diluted with fresh medium
to give the correct cell titer.
110
In a last experiment we show a transient gene expression experiment using a 12L
bioreactor. EBNA cells were transferred into fresh medium and used to inoculate the
fermentor. If the cell titer of 5-6x10e5 c/ml was achieved then 1L of fermenter volume
was replaced by 1L freshly prepared transfection complexes (DNA-DOM-PEI). The
released 1L portion was transfected under identical conditions and incubated in a spinner
flasks. The results of this fermentor scale experiments is illustrated in Fig. 3. About 16
milligram TNFRp55 was produced within 5 days post-transfection.
Conclusions
PEI-mediated gene transfer into HEK293EBNA cells can be used as a simple
transfection system to produce milligram amounts of protein in a cost effective way
during 3-5 days in serum-free medium.
Transient transfection is inhibited in conditioned culture.
The combination of DOM and PEI significantly reduces the inhibitory effect and yields
higher expression values >Furthermore, the action of both agents together are extremely
useful in fermentor scale transfection at low DNA concentration.
111
References
1. X Gao and L Huang (1995) Cationic liposome-mediated gene transfer, Gene Therapy
2, 710-722.
2. Barthel F., Remy J.-S., Loeffler J.-P. and Behr J.P. (1993) Gene Transfer
Optimization with Lipospermine-Coated DNA, DNA and Cell Biology 12, 553-560.
3. Gottschalk S., Sparrow JT., Hauer J, Mims MP., Leland FE, Woo SLC and
Smith LC. (1996) A novel DNA-peptide complex for efficient gene transfer and
expression in mammalian cells, Gene Therapy 3, 448-457.
4. Kircheis R., Kichler A., Wallner G., Kursa M., Ogris M., Felzmann T.
Buchberger M. and E. Wagner (1997) Coupling of cell-binding ligands to
polyethelenimine for targeted gene delivery, Gene Therapy
5. Legendre JY., Trzeciak A., Bohrmann B., Deuschle U., Kitas E.A. and Supersaxo A.
(1996) Dioleoylmelittin as a Novel Serum-Intensitive Reagent for Efficient
Transfection of Mammalian Cells, Bioconjugate Chem. 8, 57-63
6. Boussif O., Lezoualc'H F., Zanta M., Mergny M., Scherman D., Demeneix B. and
Behr J.-P. (1995) A versatile vector for gene and oligonucleotide transfer into cells
in culture and in vivo: Polyethelenimine, Proc. Natl. Acad. Sci.. USA 92,7297-
7301.
7. Schlaeger E.-J. (1996) The protein hydrolysate, Primatone RL, is a cost-effective
multiple growth promotor of mammalian cell culture in serum-free media and
displays anti-apoptosis properties, J. Immunol. Methods 194, 191-199.
112 What is the transfection efficiency in the reactor?
Discussion I do not know as I only looked for the accumulation of TNF
Wurm: receptor. From earlier experiments 30-60% of the cells were
Schlaeger: transfected.
Mueller:
Some time ago NC2 precipitation of DNA was used to transfect
Schlaeger: cells and those reactions were very strongly pH dependent. I
would expect the same in your system. Have you looked for pH
Merten: dependence on your precipitation efficiency?
Schlaeger. No. The whole system works very well on the plate, and we have
used different types of medium, but the pH was constant at about
7.3. I do not believe that there is any difference in the complex
formation with pH. The system is good in that it is simple, but we
have a problem with the conditioning medium in the 60 1 fermenter
because of the number of passages.
I know that DOM is still produced in your laboratory. Will it
become available on the market as it would be useful in large-scale
applications?
You can buy the 2 molecules separately but you have to do the
conjugation yourself. It has been given to various companies for
testing, but I do not know what the decision was.
LARGE SCALE TRANSIENT GENE EXPRESSION IN
MAMMALIAN CELLS
E.-J. Schlaeger, K. Christensen, G. Schmid, N. Schaub, B. Wipf, and
A. Weiss
F. Hoffmann-La Roche Ltd., Pharmaceutical Division - PRP
Department PRPN-G, CH-4070 Basel, Switzerland
Keywords: transient transfection, plasmid production, plasmid purification
•
The aim of this work was to develop a fast, robust and cost-
effective integrated process for the production of mg
amounts of protein for research purposes. The process
consists of 3 distinct parts: Plasmid production, plasmid
purification and large-scale transient transfection of
mammalian cells grown in serum-free suspension culture.
•
We investigated plasmid amplification using various E. coli
strains cultivated in different complex media followed by
the establishment of plasmid production conditions in
10 L fermenters
•
The efforts in the area of plasmid purification were geared
towards the development of a fast and cost-effective
method that generates material suitable for our PEI-
mediated transfection protocol
•
Different plasmid preparations were tested in PEI-mediated
transient transfection experiments with HEK293 (EBNA)
cells grown in serum-free suspension culture at lab and
bioreactor scale
113
O.-W. Merten et al. (eds.), New Developments and New Applications in Animal Cell Technology, 113-116.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
114
Process flow diagram for the purification
of plasmid DNA for transient transfection
Alkaline Lysis and 2-Propanol Precipitation
115
Cell culture HEK293(S), 293EBNA (Invitrogen)
Cell line: HL-0%FCS w heparin
Medium: Spinner flaks (Bellco), 12L bioractor (MBR)
Culture vessel: pREP7, pC1,(soluble human TNFRp55)
Plasmid: pGL3-CMV(luciferase)
Reagent: Polyethylenimine (PEI), Fluka, Aldrich
Dioleolyl-melittin (DOM), Roche
Transfection assay:
Log cells(5xl0e5c/ml) in HL w/o heparin
Formation of DNA-PEI transfection complexes:
indirect method: The complexes were performed in fresh medium
(1/10 volume) an then added to the cells.
Incubation: 3-5 days, spinner flask, 100rpm, (feed)
Test: sol. human TNFRp55 ELISA
116
Summary
•
We established an E. coli fermentation process for plasmid production in
10 L bioreactors yielding between 50-150 mg of unpurified plasmid
depending on the plasmid and the host strain.
•
The E. coli strain was selected to give a high yield of monomeric
supercoiled pREP7-TNFRp55 DNA.
•
A standard alkaline extraction method was scaled up and a 1-step
chromatographic procedure yielded plasmid DNA suitable for our
transfection protocol.
•
Transient gene expression with different plasmid preparations exhibited high
transfection efficiencies.
•
We developed a cost-effective large scale transient transfection
process to produce mg amounts of recombinant protein using
HEK293 (EBNA) cells grown in serum-free suspension
culture.
Synergistic enhancement of transient expression by dioleoyl-
melittin (DOM) and polyethylenimine (PEI) in mammalian
cells in suspension culture
E.-J. Schlaeger , J.Y. Legendre*, A. Trzeciak, E.A Kitas, K. Christensen, U. Deuschle
and A. Supersaxo
F. Hoffmann La Roche Ltd. PRPN-G, PRPF, CH-4070 Basel, Switzerland
*UPSA Laboratoires, 128 rue Danton, 92506 Rueil-Malmaison, France
Abstract
Dioleoyl-melittin, which is a conjugate of dioleoyl-phosphatidyl-ethanolamine-N-{3-(2-
pyridyldithio)propionate} with the amphipatic peptide melittin { Gly-Cys} 1 represents a
new class of peptide-based reagent for efficient transfection of mammalian cells.
In this work we investigated the transfection efficiency of dioleoyl-melittin (DOM)
combined with polyethylenimine (PEI) using HEK293(EBNA) cells grown in serum-
free suspension cultures. Gene expression was monitored using the luciferase reporter
gene and the human soluble TNF receptor p55 gene (TNFRp55) inserted into different
vectors. Our data clearly show that DOM together with PEI exhibited synergistic
enhancement for gene expression in EBNA cells. At the optimal DNA-DOM-PEI
weight ratio the efficiency of transfection increases significantly compared to
corresponding PEI and DOM transfection at low DNA concentration.
In summary, our data show that dioleoyl-melittin and polyethylenimine act
synergistically in transfecting 293(EBNA) cells grown in serum-free suspension culture.
Keywords : HEK293, serum-free culture, transient transfection, bioreactor
Introduction
Efficient gene delivery in gene therapy by using non-viral vectors has recently become
of great interest (1 ). Substantial work has been done to improve DNA uptake where
the DNA is generally complexed with cationic lipids or polycationic polymers. More
recently some of these transfection agents were conjugated with proteins or peptide in
order to target it to specific cell types (2,3). In this context we were interested in dioleyl-
melittin (DOM) which is a novel peptide-based gene transfer agent as described recently
(4 ). The authors described that the highest expression levels for gene delivery
experiments were found by using the ratio DNA:DOM of 1:10. If we transferred the
weight ratio of DNA: DOM to our transfection system it became clear that if lµ g/ml
DNA was used in a transfection experiment the corresponding amount of DOM is
highly toxic especially in serum-free suspension culture. Therefore, it was very
important to minimize the cell toxicity by reducing the plasmid concentration.
In this work we investigated the transient transfection efficiency of dioleoyl-melittin
(DOM) combined with polyethylenimine (PEI) at low DNA concentration using
HEK293(EBNA) cells grown in serum-free suspension cultures.
117
O. - W. Merten et al. (eds.), Mew Developments and New Applications in Animal Cell Technology, 117–120.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
118
Materials and Methods
Cell culture
Cell line: HEK 293EBNA (Invitrogen) serum-free HL medium (5).
Culture vessel: Spinner flaks (Bellco), 12 L bioreactor, stirred (MBR)
Transfection
Transfection agents: Polyethelenimine (PEI), dioleoyl-melittin (DOM)
Reporter gene: Extracellular domain of human TNF receptor p55 .
Plasmid: pREP7-TNFRp55 (EBV-based expression vectors (Invitrogen)
Transfection assay
HEK293(EBNA) cells were centrifuged, washed once and resuspended into fresh
medium. The cell titer was adjusted to 5-6xl0e5 cells/ml.
Preparation of the transfection complexes
The DNA-PEI transfection complexes were prepared in fresh medium (1/10 volume) and
than added to the cells (6).
The DNA-DOM-PEI transfection complexes were prepared in fresh medium (1/10
volume) and then added to the cells.
Incubation: The transfected culture was grown for 2-5 days.After the second day the
culture was fed once daily .
Product analysis: The secreted TNFRp55 were measured by ELISA.
Results and Discussions
Before establishing data for a scaleable transient gene expression system we first
adaptated the adherently growing HEK293 cells and the subclone 293EBNA to serum-
free growth in suspension culture using HL medium .
In a set of experiments we tested the expression levels of DOM-mediated transient
transfection by using different DNA and DOM concentrations.
The results clearly indicate significantly toxic effects on cells if more than
DNA and DOM is used. However, lowering the DNA concentration further
caused a significant drop in efficiency ( data not shown).
From our experience with PEI-mediated transient gene expression we tried to combine
both transfection agents in order to improve the gene delivery system in suspension
cell culture. Fig. 1 shows the time course of a transfection experiment in which 0.25
DNA was incubated with DOM in the absence and in the presence of
PEI. Using the DNA-DOM complexes high cell viability was observed (> 85%).
However, 3 days post-infection only background expression levels were measured.
Similar results were obtained if DNA-PEI transfection complexes were used. Quite
different results were obtained in the sample where PEI was added with the
DNA-DOM complex. A high yield of TNFRp55 was accumulated in the culture
medium. The conclusion of this simple experiment clearly indicates that DOM and PEI
act synergistically in transfecting 293EBNA at low plasmid concentration. This result
was subsequently confirmed on several occasions. The concentrations of DNA, DOM
and PEI were optimized in order to improve transfection efficiency.
119
After establishing lab-scale data on transient gene expression we scaled up this
technique. First we applied the transfection conditions in spinner culture up to 2L
volume without lost of expression levels, demonstrating the possibility of production
of several mg recombnant protein. Finally transient gene expression was performed in a
12 L bioreactor. Figure 2 shows the results. About 2,5L cells were grown up to a cell
titer of 1,2×l0e6 cells/ml, transferred into fresh medium by a centrifugation step and
used to inoculate the fermentor giving a final volume of 11L.
120
Conclusion
The results of this study clearly show that DOM together with PEI exhibited
synergistic enhancement for gene expression in EBNA cells at low DNA concentration.
At the optimal DNA-DOM-PEI weight ratio (0.2-0.15-0.5) the efficiency of transfection
increases significantly (up to 10-fold) compared to corresponding DNA-PEI and DNA-
DOM transfection. No toxic effect was observed.
Therefore, the combination of DOM and PEI exhibits a cost effective transient gene
expression system with high transfection values and is extremely useful for large scale
production of milligram amounts of recombinant proteins at low plasmid
concentration.
References
1. X Gao and L Huang (1995) Cationic liposome-mediated gene transfer, Gene Therapy
2, 710-722.
2. Gottschalk S., Sparrow JT., Hauer J, Mims MP., Leland FE, Woo SLC and
Smith LC. (1996) A novel DNA-peptide complex for efficient gene transfer and
expression in mammalian cells, Gene Therapy 3, 448-457.
3. Kircheis R., Kichler A., Wallner G., Kursa M., Ogris M., Felzmann T.
Buchberger M. and E. Wagner (1997) Coupling of cell-binding ligands to
polyethelenimine for targeted gene delivery, Gene Therapy
4. Boussif O., Lezoualc’h F., Zanta M., Mergny M., Scherman D., Demeneix B. and
Behr J.-P. (1995) A versatile vector for gene and oligonucleotide transfer into cells
in culture and in vivo: Polyethelenimine, Proc. Natl. Acad. Sci.. USA 92,7297-
7301.
5. Schlaeger E.-J. (1996) The protein hydrolysate, Primatone RL, is a cost-effective
multiple growth promotor of mammalian cell culture in serum-free media and
displays anti-apoptosis properties, J. Immunol. Methods 194, 191-199.
6. Legendre JY., Trzeciak A., Bohrmann B., Deuschle U., Kitas E.A. and Supersaxo A.
(1996) Dioleoylmelittin as a Novel Serum-Intensitive Reagent for Efficient
Transfection of Mammalian Cells, Bioconjugate Chem. 8, 57–63
TRANSIENT EXPRESSION OF A SOLUBLE AND SECRETED
FORM OF HETERODIMERIC T-CELL RECEPTOR IN HEK-293
M. JORDAN* , C. L. BLANCHARD°,1. BERNASCONI°,
I. LUESCHER° and F. M. WURM*
* Center of Biotechnology UNIL-EPFL, Swiss Federal Institute of
Technology, 1015 Lausanne, Switzerland
°Ludwig Institute for Cancer Research, University of Lausanne, 1066
Epalinges, Switzerland
1. Abstract
Breicoolomgbicinaallnyt parcotitveein hiunm2a9n3 Tc-eclelsll. rCeocerrpetcotr (TCRc)haiwnaspaierxinprgessweads afsacilaitatseecdretbeyd
complementary charged peptides (leucine zipper) at the C terminus. Two plasmids,
coding for the chains were constructed and prepared using commercially
available purification kits. After optimizing transfection conditions for secretion of
heterodimeric TCR in 12 well plates, transfections were performed in spinner flasks.
The transfection efficiency was determined with a co-transfected expression vector
and correlated with TCR levels as determined by ELISA. Product was isolated by
affinity chromatography. Biological activity, i.e. ligand binding was tested. Transiently
produced recombinant soluble TCR showed similar affinity to the ligand as receptors
purified from cell membranes.
2. Introduction
For determination of the structure of TCR, mg amounts of soluble protein are desirable.
Before a quantitative ELISA with a purified standard was available, relative amounts of
heterodimers were detected by an ELISA using antibodies against both chains.
Biological activity was measured using ligand. Stable expression in CHO
cell lines gave poor signals in both assays, indicating that the formation of
heterodimers is not very efficient and that TCR might be difficult to express at high
levels. This article describes the transient expression of TCR in the human kidney cell
line 293 T.
3. Materials and Methods
Plasmids and vectors: The chain were cloned seperately into the pCI-neo
expression vector (Promega). In order to have a soluble and secreted version of TCR the
transmembrane domain and cytoplasmic tail were deleted. In addition, a leucine zipper
121
O.-W. Merten et al. (eds.), New Developments and New Applications in Animal Cell Technology, 121-123.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
122
sequence was introduced into each chain, which should facilitate formation of
heterodimers. The Qiagen kit was used to purify mg amounts of plasmids.
Transfections in plates: Exponentially growing 293 T cells were seeded at
10-20 hours before transfection in DMEM/F12 medium supplemented with 2% FCS1.
Per 10 ml of medium, . of 250mM containing a total of DNA was
mixed with HEPES buffer (1.5mM 140mM NaCl, 50 mM HEPES,
pH = 7.05). The transfection cocktail was added to cells after 1 minute and the plates
were incubated for 4 hours at . Precipitate was removed at this point by exchanging
the medium.
Transfections in spinner flasks: For 1 L spinner flask, cells were seeded in 500 ml of
fresh medium at Immediately before the precipitate was formed 15 ml
of 250mM was added into the spinner flasks to increase the calcium
concentration. Subsequently, 20 ml of transfection cocktail
was added2. After 4 hours, the culture was diluted by
adding 500 ml of fresh medium. Protein was harvested after 2 - 3 days.
4. Results and Discussion
IS OPTIMAL FOR HETERODIMER EXPRESSION
Expression of TCR was first optimized in 12 well plates to identify an optimal ratio of
the two chains, encoded by two different expression vectors. Ratios of 1:9 through 9:1
were tested (Fig. 1). ELISA data as well as photo-affinity labeling tests indicated that
the is limiting and that at an excess
of expression vector significantly
improves transient expression.
4.2. T-FLASKS VERSUS SPINNER FLASKS
TCR was transiently expressed in T-flasks or in 1 L spinner flasks. The ratio of
chain expression vectors was 1:2 . For the spinner experiment, cells were
adapted for one month to grow in suspension containing single cells and small
aggregates (less than 10 cells per aggregate). Transfections of this adapted population in
plates showed that the cells still grew adherent and there was no change in transfection
efficiency. The expression of TCR in the spinner flasks was comparable to the
T flasks (Table 1). In other experiments no difference could be found between 12 well
plates and T-flasks.
123
4.3. DOES AN INTERNAL CONTROL CORRELATE WITH PRODUCT TITERS?
Although the transfection procedure is well defined, the problem of variation in
transfection efficiency between different experiments remains. For the transient
expression of proteins such variation can cause a problem, in particular when there is
no quantitative assay available for the protein to be expressed. To minimize variation
and to guarantee at least a reasonable transfection efficiency before efforts for product
isolation are started, we included in one series of transfections in 100 ml spinner flasks
as an internal standard 10 % of pCMV The cells were fixed 24 hours
after transfection and stained for 4 hours with X-gal. Positive cells were counted in the
microscope and compared with the relative ELISA-signal (Fig.2). Although none of the
assays were optimized for a high accuracy, the internal marker seems to correlate with
the levels of produced TCR. Such an
internal standard, co-transfected with any
protein has a clear advantage, because it
allows detection with one assay all the
variations in transfection efficiency, which
would directly affect the product yield.
5. Conclusions
Different chains seem to be expressed at different efficiency, even when both chains are
cloned into the same expression vector.
For two chain proteins the levels of transient expression can be improved by
optimizing the ratio of the two vectors used for co-transfection.
Transfections in spinner flasks resulted in the same level of expression as transfections
in T flasks or in 12 well plates.
Co-transfection of a marker (10% or less of total DNA) correlates with the product level
for a particular protein and generally can be used as a control for reproducible
transfection efficiencies.
200 µg of recombinant TCR was purified from 2.5 L of supernatant harvested 2 days
after transfection.
References
1 Jordan, M., Schallhorn, A. and Wurm, F.M. (1996) Transfecting mammalian cells: optimization of critical
parameters affecting calcium-phosphate precipitate formation. Nucleic Acids Res. 24, 596-601
2 Jordan, M., Koehne, C. and Wurm, F.M. (1997) Calcium phosphate mediated DNA transfer into HEK-
293 cells in suspension: control of physicochemical parameters allows transfection in stirred media.
Cytotechnology, in press
MEASURABLE PARAMETERS OF CELLS AND PRECIPITATE
PREDICT TRANSFECTABILITY WITH CALCIUM PHOSPHATE
M. Jordan and F. Wurm
Center of Biotechnology UNIL-EPFL, Swiss Federal Institute of
Technology, 1015 Lausanne, Switzerland
1. Abstract
Transient expression of proteins in mammalian cells at 1 - 100 L scale has the
potential to become a powerful technique for the rapid production of research proteins.
We have shown that the calcium phosphate technique is compatible with routine
handling in bioreactors and that the yield of proteins is comparable to that seen in
standard transient transfections1. However, although the reproducibility of the method is
excellent within triplicates, there can be dramatic variations in transfection efficiencies
with different batches of cells, solutions, plasmid preparations etc. In addition,
expression levels of novel proteins may differ from those seen with proteins like tPA
or hGH. There is a need therefore, to monitor the transfection procedure itself and to
predict protein expression through defined and quantitative assays. We suggest two
tests: the first assesses, through OD-measurements of the transfection cocktail, the
quality of the DNA-carrier complexes. The second employs a commercially available
vector encoding a strongly fluorescent green fluorescent protein (GFP). Transfection
with this vector provides, within a day, a quantitative signal. Together, these assays
allow a fast assessment of transfections with calcium phosphate for optimization
purposes. Moreover, co-transfecting GFP with the vectors of interest allows, to a
certain degree, prediction of transfectability and expression of the desired protein.
2. Introduction
Transfection efficiency depends on physical and biological parameters, some of them
not known or difficult to control. For the calcium phosphate technique in particular,
physical parameters of the precipitation buffers are most critical2. The effect of such
parameters on the precipitation of DNA can be studied by a turbidity assay described
earlier3. Here we suggest a modified protocol without the addition of DNA, to test and
to compare new batches of transfection solutions rapidly.
Transfectability of cells is more difficult to control. Even with the same cell line,
transfection results can vary several fold depending on the actual status of the cell
population. It is desirable therefore, to include an internal standard in order to monitor
125
O.-W. Merten et al. (eds.), New Developments and New Applications in Animal Cell Technology, 125-128.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
126
the actual transfection efficiency. Such an internal standard should be applicable for
various cell lines and different methods applied. In addition, it should allow comparison
of transfection efficiencies at various scales.
3. Materials and Methods
Turbidity assay: One volume of 250 mM (without any DNA) is quickly mixed
with one volume of HEPES buffer
HEPES, pH = 7.05). After 1 minute, the absorption at 320 nm is measured.
GFP measurements:The plasmid pEGFP-Nl (Clontech) is transfected in multiwell
plates using the calcium phosphate method. Fluorescent signals are directly measured
from 12 well plates containing cell culture medium (CytoFluor II, extinction at
485/20, emission at 530/25). Since plastic and cell culture medium significantly
contribute to the signal, a control measurement immediately taken after the transfection
when fresh medium had been added, is to be subtracted from the derived values.
Alternatively, cells were counted and resuspended in PBS. cells per well (96
well quartz plate) typically resulted in good signals much above the background from
negative cells.
4. Results and Discussion
4.1. CONTROL OF NEW BATCHES OF SOLUTIONS
Although many parameters have been described to optimize the solutions needed to
form the precipitate, only a transfection experiment with a known reference solution
proves equivalency. It is obvious that this can be tricky. A more direct method is the
quantification of the turbidity of the forming precipitate itself. There is no absorption at
320 nm in the absence of "precipitate" particles. When the phosphate concentration in
the precipitation buffer is increased, the physical nature of the precipitate changes and
the turbidity increases in a typical mode (Fig. 1.). For transfections, an absorption of
0.13 - 0.145 was found to be optimal for precipitates formed in the presence of 25µg of
DNA per ml of precipitate. Once this
value is known for a transfection
protocol, the turbidity can be used as a
quick assay assessing the capacity of the
used solutions to generate an efficient
DNA-carrier complex with calcium
phosphate.
127
4.2. GFP SIGNAL IS PROPORTIONAL TO DNA AMOUNT ADDED PER WELL
Preliminary results with showed that an internal standard correlates with product
titers, but too many variations were seen with this assay. We investigated the use of
GFP protein, eliminating lysis of cells, enzymatic reaction etc. A major drawback of
GFP is that signals are relatively weak, which makes detection not very sensitive.
Second generation GFP's overcame this limit. We tested the sensitivity by transfecting
different amounts of GFP plasmid into
293 cells and measured the signal directly
in the 12 well plates (Fig. 2). Even at the
lowest amount of DNA, 200 ng per well,
the signal was clearly above the
background level. Up to a dose of 1 µg
per well, signal intensity increased
linearly with the amount of DNA, yet
became saturated above 2.5 µg per well.
4.3. GFP SIGNAL INCREASES FOR 70 HOURS AFTER TRANSFECTION
3 – 5 hours after the addition of precipitate to the cells, first positive cells can be
identified in the microscope. The quantification of fluorescence however is difficult at
time point earlier than 12 hours. Intensities of signals increase for the first 3 days and
stay almost constant for a few more days
(Fig. 3). The plateau after 3 days was
seen at all the DNA concentrations tested.
It appears therefore that such a plateau
does not reflect a saturation of
intracellular GFP accumulation, but a
loss of expression after a certain period of
time. A similar behaviour can also be
observed for transient expression of
secreted proteins.
128
5. Conclusions
The turbidity at 320 nm quantitatively describes the precipitate. Solutions can be
quickly tested without the addition of DNA.
The EGFP-N1 vector in combination with an efficient transaction method results
within a day after transfection in fluorescent signals detectable directly from growing
cells. Since the signal derives directly from the expressed protein, any pretreatment of
samples or incubation steps can be avoided.
Expression of GFP appears not to interfere with the expression of other proteins, when
representing 10% or less of plasmid quantity of the transfection cocktail. Even at this
low molar ratio, the product concentration is high enough to produce easy detectable
fluorescent signals.
Signals increase for about 70 hours and reflect GFP expression with a high accuracy.
Intracellular accumulation of GFP shows a trend typical for the expression of secreted
proteins, such as tPA or rDNase (data not shown). The fluorescent signal might not
only correlate with transfection efficiency, but with protein expression in general as
well.
References
1 Jordan, M., Koehne, C. and Wurm, F.M. (1997) Calcium phosphate mediated DNA transfer into HEK-
293 cells in suspension: control of physicochemical parameters allows transfection in stirred media.
Cytotcchnology in press
2 Jordan, M. and Wurm, F.M. (1995) High level transient expression in mammalian cells: identification and
optimization of physico-chemical parameters of the calcium phosphate transfection method, in E.C.
Beuvery et al. (eds.) Animal Cell Technology: Developments towards the 21st Century, Kluver
Academic Publisher, Dordrecht. 49-55
3 Jordan, M., Schallhorn, A. and Wurm, F.M. (1996) Transfecting mammalian cells: optimization of
critical parameters affecting calcium-phosphate precipitate formation. Nucleic Acids Res. 24. 596-601
GLYCOSYLATION
INFLUENCE OF Na-BUTYRATE ON THE PRODUCTION
AND SIALYLATION OF HUMAN INTERFERON- BY
2,6-SIALYLTRANSFERASE ENGINEERED CHO-CELLS
D. LAMOTTE1, L. MONACO2, N. JENKINS3 & A. MARC1
1 LSGC-CNRS, BP 451, 54001 Nancy Cedex - France
2 DIBIT, via Olgettina 60. 20132 Milano - Italy
3 De Montfort University, L91 9BH Leicester - UK
1. Introduction
Mammalian cells are widely used for the production of recombinant drugs because of
their ability to perform extensive post-translational modifications. However, they
exhibit low production yields in comparison with procaryotic cells. Nevertheless, the
use of stimulating agents such as sodium butyrate in medium formulation can enhance
the synthesis of recombinant glycoproteins1,2.
Some conflicting results demonstrated that Na-butyrate addition affects the cell's
glycosylation machinery. For some, sodium butyrate may reduce the
sialyltransferase synthesis3,4. Conversely, an increase in the specific activity of
sialyltransferases in CHO cells was pointed out as a result of sodium butyrate addition5.
Although the CHO-derived glycan structures are similar to those present on natural
human proteins, they are not identical6. In CHO cells, sialic acids are exclusively added
to terminal galactose via Since it is known that the absence of
linkages can dramatically enhance clearance of a drug from bloodstream6, we have
recently expressed the cDNA coding for the rat into the CHO 320 producing
(IFN) line and allowed the presence of on IFN7.
Here, we report the effect of 1 mM sodium butyrate on cell growth, IFN production and
sialylation. Experiments were performed in controlled bioreactor cultures with parental
and engineered CHO cells, in order to determine the influence of Na-butyrate
on the IFN sialylation.
2. Materials & Methods
The producing IFN CHO cell line (CHO 320) was supplied by the Wellcome
Foundation Laboratories. The engineered producing IFN CHO
cell line (CHO C5) was constructed as described in ref. 7. Cultures were performed
with a serum-free medium based on RPMI-1640 supplemented with BSA, insulin and
salts. The culture conditions were similar to those described in ref. 8. Sodium butyrate
was added from a stock solution (500 mM in PBS - Sigma B2503) 24 or 36 hours
further the seeding of the bioreactor. IFN purification and HPLC resolution and
detection of sialic acids to discriminate between and are described
in ref. 7.
131
O.-W. Merten et al. (eds.), New Developments and New Applications in Animal Cell Technology, 131-133.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
132
3. Results
3.1. INFLUENCE OF Na-BUTYRATE ON GROWTH AND IFN PRODUCTION
Since preliminary shake-flask cultures have shown that the 1 mM Na-butyrate
concentration is optimal for enhancing the IFN production without compromising the
cell viability, bioreactor cultures were performed with 1 mM Na-Butyrate.
1 mM Na-butyrate concentration restricted the maximal viable cell density and
increased the final IFN concentration by 90-130 % (fig. 1). Butyrate exposure resulted
in a major increase in the maximal specific IFN production rate reaching 45 and 150
for CHO 320 and CHO C5 respectively (data not shown). For both cell
lines, the production of the IFN occurred exclusively during the growth phase.
3.2. INFLUENCE OF Na-BUTYRATE ON IFN SIALYLATION
IFN was purified from culture supernatants at the end of the growth phase of the batch
cultures (80-100 hours of culture) performed with and without butyrate addition. The
percentages of sialic acids and the molar ratio of total sialic acids to IFN
molecules were determined (table 1). In control conditions, transfection of the
cDNA into CHO 320 cells to produce C5 resulted in 68% of total sialic acids linked in
the and nearly double the overall IFN sialylation, in comparison
with the IFN from the parental cell line.
133
The effect of adding butyrate was determined for the two cell lines. As expected, 100 %
of sialic acids on IFN from CHO 320 were still in the in the culture
performed with Na-butyrate. Conversely, 82 % of sialic acids were in the
conformation on IFN secreted by CHO C5 line withNa-butyrate
compared to 68% without butyrate. Furthermore, the overall degree of IFN sialylation
was twofold higher for CHO C5 cells subjected to butyrate, whereas it was not
enhanced by butyrate exposure in the CHO 320 cell line. Therefore, the increase in the
overall degree of sialylation on IFN produced by CHO C5 cells with butyrate results
predominantly from an increase in the molar ratio of total sialic acids
(+130%) rather than sialic acids (+15%).
4. References
1. Chevalot I., Dardenne M., Cherlet M., Engasser J.-M. and Marc A. (1995) Effect of sodium butyrate on
protein production in different culture systems, in Beuvery E.C., Spier R. and Griffiths B. (eds.), Animal Cell
Technology: Developments towards the 21st century, Kluwer Academic Publishers, Dordrecht, pp. 143-147
2. Oster T., Thioudellet C., Chevalot I., Masson C., Wellman M., Marc A. and Siest G (1993) Induction of
recombinant gamma-glutamyl transferase by sodium butyrate in transfected V79 and CHO Chinese hamster
cells, Biochem Biophys Res Comm 193, 406-412.
3. Shah S., Lance P., Smith T.J., Berenson C.S., Cohen S.A., Horvath P.J., Lau J.T.Y. and Baumann H.
(1995) n-butyrate reduces the expression of in Hep G2 cell. J Biol Chem
267, 10652-10658.
4. Li M., Andersen M.L. and Lance P. (1995) Expression and regulation of glycosyltransferases for N-
glycosyl oligosaccharides in fresh human surgical and cultured cell lines, Clin Sci 89, 397-404.
5. Chotigeat W., Watanapokasin Y., Malher S. and Gray P.P. (1994) Role of environmental conditions on the
expression levels, glycoform pattern and levels of sialyltransferase for hFSH produced by recombinant CHO
cells, Cytotechnology 15, 217-221.
6. Jenkins N., Parekh R.B. and James D.C. (1996) Getting the glycosylation right: implications lor the
biotechnology industry, Nature Biotechnology 14, 975-981.
7. Monaco L., Marc A., Eon-Duval A., Acerbis G., Distefano G., Lamotte D., Engasser J.-M., Soria M and
Jenkins N. (1996) Genetic engineering of in recombinant CHO cells and its effect on
the sialylation of recombinant interferon-gamma, Cytotechnology 22, 197-203.
8. Lamotte D., Eon-Duval A., Acerbis G., Distefano G., Monaco L., Soria M., Jenkins N., Engasser J.-M and
Marc A. (1997) Controlling the glycosylation of recombinant proteins expressed in animal cells by genetic
and physiological engineering, in Carrondo M.J.T., Griffiths B. and Moreira J.L.P. (eds.), Animal Cell
Technology: From Vaccines to Genetic Medicine. Kluwer Academic Publishers, Dordrecht, pp. 761-765.
ELEVATED INHIBITS THE POLYSIALYLATION OF THE
NEURAL CELL ADHESION MOLECULE IN CHO MT2-1-8 CELL
CULTURES
JAMES A. ZANGHI1,3, THOMAS P. MENDOZA1, RICHARD H.
KNOP2, WILLIAM M. MILLER1,#
1 Department of Chemical Engineering, Northwestern University, 2145
Sheridan Road, Evanston, IL 60208-312
2 Division of Medical Oncology, Evanston Hospital, 2650 N. Ridge
Avenue, Evanston, IL 60201-1794
3 Present location: Institute of Biotechnology, Swiss Federal Institute of
Technology, ETH Hönggerberg, HPT, CH-8093 Zürich, Switzerland.
Email: [email protected]
# Corresponding author. Phone: (847)491-4828. Email:
wmmiller@nwu. edu
Animal cells are important for the production of complex glycoproteins for use in
therapeutics and diagnostics. Conditions during culture, such as the depletion of
nutrients and accumulation of metabolites, may adversely affect the glycosylation
process or increase the deterioration of glycoproteins after secretion (1, 2).
Glycosylation has a critical role in protein quality by regulating its physiochemical and
biological properties such as specific activity, thermal stability, solubility, clearance rate
from blood stream, and immunogenicity (1). We and others have shown that ammonia,
a metabolic waste product of animal cells, can specifically inhibit protein glycosylation
(3, 4, 5, 6). Using the neural cell adhesion molecule (NCAM) as our model cell surface
glycoprotein (Fig. 1), we showed that ammonia inhibits the NCAM polysialylation in
Chinese hamster ovary (CHO) and small cell lung cancer (SCLC) cells in a dose-
dependent and reversible manner (6). The response was very strong, with a 90% decrease
in CHO cell PolySia surface content following an exposure to 10 mM for 4
days. A possible explanation for ammonia inhibition of sialylation or polysialylation
is decreased (poly)sialyltransferase activity due to perturbations in the trans golgi pH
where the enzymes reside. The weak base readily diffuses across biological
membranes, while its conjugate acid is virtually impenetrable so that the base
accumulates in acidic compartments in the cell, including the trans golgi, resulting in
collapse of the pH gradient (7, 8, 9, 10).
Carbon dioxide is another important waste product of animal cells,
though problematic only under specific conditions such as high density culture systems
where significant differences between the mass transfer coefficients of oxygen and carbon
dioxide leads to a buildup of dissolved
135
O.-W. Merten et al. (eds.), New Developments and New Applications in Animal Cell Technology, 135-140.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
136 in water may be summarized as
By neglecting
follows (13):
Analogous to ammonia, freely diffuses across cellular membranes while the
conjugate base is essentially membrane impermeable. For this reason, cells
exposed to elevated at constant pH experience an initial decrease in pH. While the
trapping effect observed for ammonia is less likely to occur with because of specific
transporters such as the antiporter and symporter (14), the decrease
in intracellular pH may or may not return to the original value (15, 16). We therefore
hypothesized that elevated may also disrupt intracellular pH and other ion
gradients, and thus inhibit polysialylation in a manner analogous to ammonia.
Elevated inhibits CHO cell growth and protein production (12, 17, 18).
Kimura et al. (18) showed that CHO cell growth inhibition was more severe when
osmolality was allowed to increase with while an increase in osmolality alone
had no effect. In addition, it was shown that elevated had no effect on recombinant
tissue plasminogen activator (tPA) glycosylation with the exception of an increase in N-
glycolylneuraminic acid (Neu5Gc; balance N-acetylneuraminic acid, Neu5Ac) (19). A
similar observation was reported by Grampp (20). Here we examined the effects of
elevated as well as pH and osmolality, on CHO MT2-1-8 cell NCAM
polysialylation. The experiments were conducted with the same CHO MT2-1-8 cell
line, culture medium and set-up as that previously published for the studies with tPA
(18, 19, 21, 22) (Fig. 2). This allows a direct comparison of the effects of on the
glycosylation of the more robust tPA molecule and the hyper-sensitive NCAM
137
polysialic acid. Cells were cultured for 1-5 days under an atmosphere of 38 (control, 5%
, 90, 140, or 195 torr and subsequently analyzed by flow cytometry using
anti-NCAM and -PolySia mAbs. We separately measured the effects of osmolality and
pH on NCAM polysialylation by adjusting the NaCl or NaOH concentration in the
medium, respectively.
The results of the experiments are summarized in Table 1 (23). At normal
levels, both increasing pH and osmolality inhibited polysialylation. The effect
was most pronounced at a pH above 7.4 and when the osmolality exceeded 400
mOsmol/kg. As increased the effect was much greater, with PolySia content
decreasing until no immunoreactivity remained (i.e. anti-PolySia mAb fluorescence was
identical to background fluorescence) after four days exposure to 140 at pH 7.5.
At moderately elevated levels (90 torr), PolySia was unaffected or only slightly
inhibited at pH 7.2 but was significantly reduced at pH 7.5. In all cases, NCAM was
not affected, suggesting that the increase in pH, osmolality and were specific to
polysialic acid.
138
Kimura et al. showed that increasing osmolality exacerbates the growth
inhibition caused by elevated (18). In addition, increases with increasing
pH which may disrupt exchangers in the cell membrane or other cell functions.
Cell counts prior to flow cytometry confirm that there was a synergism between
increased pH and elevated At pH 7.5, CHO cell growth was completely inhibited
at 195 torr, and there was considerable cell death at 250 torr This contrasts the
results obtained at pH 7.2, where 195 torr and 250 torr decreased cell growth by 24% and
44%, respectively (18). In addition, at values of 195 and 250 torr, there were
gross changes in morphology, including larger cell size and decreased adherence to the
culture flask only at the higher pH. An increase in increases the buffering
capacity of the medium, thus requiring more acid or base to change the pH. As
illustrated in Fig. 3, small changes in pH will greatly affect the amount of NaOH added
to the cell culture at elevated which will also affect osmolality.
For ammonia, an increase in extracellular pH results in greater accumulation of
ammonia in the cell and greater inhibition of both cell growth and sialylation (4, 6, 24,
25). It is clear the carbon dioxide has a similar effect. The increased buffering capacity
at elevated will likely reduce proton gradients within the cell, especially in acid
organelles where pH can be a low as 5.0, and increase the energy required to maintain the
ATPase activity needed to pump protons from the organelles. If exchangers are
present in membranes of these organelles, then elevated may directly disrupt
intracellular pH gradients by way of Either way, changes in pH may result in a
decrease in the activity of pH-sensitive enzymes within these organelles and may explain
the observed decrease in NCAM polysialylation.
In addition to the relevance to biotechnology, the regulation of PolySia by
factors in the extracellular environment can have broad implications. NCAM
polysialylation is developmentally regulated and mediates a variety of cell-cell adhesive
interactions including cell migration during morphogenesis (26) and malignant growth
(27). Conditions present in large scale, high density culture systems such as elevated
ammonia and carbon dioxide levels are also present within fast-growing solid tumors
139
such as SCLC. The data provide evidence of a link between the tumor
microenvironment and metastasis through the regulation of NCAM polysialylation.
Acknowledgments. We thank Merck, Sharp & Dohme for sponsoring the oral presentation of
this work at the 15th ESACT meeting. We also acknowledge the organizing committee for
making this possible through a bursary. We thank Albert Schmelzer for his laboratory assistance.
This research was supported by NSF grants BCS-9058416 and BES-9402030 (W.M.M.) and the
Jean Ruggles Romoser Research Endowment (R.H.K).
References
1. C. F. Goochee, T. Monica, Bio/technol. 8, 421-426 (1990).
2. D. C. Anderson, C. F. Goochee, Curr. Opin. in Biotechnol. 5, 546-549 (1994).
3. D. C. Anderson, C. F. Goochee, Biotechnol. Bioeng. 47, 96-105 (1995).
4. M. C. Borys, D. I. H. Linzer, E. T. Papoutsakis, Biotechnol. Bioeng. 43, 505-514 (1994).
5. M. Gawlitzek, U. Valley, M. Nimtz, R. Wagner, H. S. Conradt, in Animal Cell Technology:
Developments towards the 21st Century. E. C. Beuvery, Ed. (Kluwer Academic, Netherlands, 1995).
6. J. A. Zanghi, T. P. Mendoza, R. H. Knop, W. M. Miller, J. Cell Physiol. submitted (1997).
7. W. F. Boron, P. de Weer, J. Gen. Physiol. 67, 91-112 (1976).
8. R. T. Dean, W. Jessup, C. R. Roberts, Biochem J. 217, 27-40 (1984).
9. C. de Duve, et al., Biochem. Pharmacol. 23, 2495-2531 (1974).
10. R. G. Anderson, R. K. Pathak, Cell 40, 635-643 (1985).
11. S. S. Ozturk, Cytotechnol. 22, 3-16 (1996).
12. D . R. Gray, S. Chen, W. Howarth, D. Inlow, B. L. Maiorella, Cytotechnol. 22, 65-78 (1996).
13. W. W. Umbreit, in Manometric and biochemical techniques W. W. Umbreit, R. H. Burris, J. F.
Stauffer, Eds. (Burgess Publishing, Minneapolis, 1972) pp. 20-29.
14. I. H. Madshus, Biochem. J. 250, 1-8 (1988).
15. R. Krapf, C. A. Berry, R. J. Alpern, F. C. J. Rector, J. Clin. Invest. 81, 381-389 (1988).
16. L. Simchowitz, A. Roos, J. Gen. Physiol. 85, 443-470 (1985).
17. D. Drapeau, Y.-T. Luan, J. C. Whiteford, D. P. Lavin, S. R. Adamson, Paper presented at the Annual
Meeting of the Society of Industrial Microbiology, Orlando, FL (1990).
18. R. Kimura, W. M. Miller, Biotechnol. Bioeng. 52, 152-160 (1996).
19. R. Kimura, W. M. Miller, Biotechnol. Prog. 13, 311-317 (1997).
20. G. E. Grampp, et al., Paper presented at the Cell Culture Engineering IV meeting, San Diego, CA
(1994).
21. R. Kimura, PhD Thesis, Northwestern University (1996).
22. J. A. Zanghi, PhD Thesis, Northwestern University (1997).
23. J. A. Zanghi, T. P. Mendoza, A. Schmelzer, R. H. Knop, W. M. Miller, Manuscript in preparation.
(1997).
24. C. Doyle, M. Butler, J. Biotechnol. 15, 91-100 (1990).
25. I. Lüdemann, R. Pörtner, H. Märkl, Cytotechnology 14, 11-20 (1994).
26. G. M. Edelman, Ann. Rev. Biochem. 54, 135-169 (1985).
27. E. P. Scheidegger, P. M. Lackie, J. Papay, J. Roth, Lab.Invest. 70, 95-106 (1994).
140 It is very difficult to separate the effects of pH and bicarbonate and
Discussion because they are all coupled together. So I was wondering if
Ozturk:
it is possible to decouple the bicarbonate and
Zanghi:
Constant pH and osmolality was maintained and we varied the
Ozturk: but the bicarbonate level went up. It may be possible to alter
Zeng: the pH and together and maintain the bicarbonate
Zanghi: concentration.
Bicarbonate maybe more important than the itself.
Have you measured, or calculated, the data?
The was measured using a blood-gas analyser in all cases
INFLUENCE OF CULTIVATION CONDITIONS ON GLYCOSYLATION
PATTERN-
A FED-BATCH AND CONTINUOUS CULTURE STUDY
NATASCHA A. SCHILL, MORRIS Z. ROSENBERG,
REBECCA L. DABORA
Biogen Inc., 14 Cambridge Center, Cambridge MA 02142, USA
1. Introduction
Glycosylation is a post-translational event taking place in the Golgi apparatus of
eukaryotic cells. The glycosylation pattern of a protein may have a significant effect on
its physical, chemical and biological properties. Therefore, process development for a
potential therapeutic glycoprotein is only successful if it can be demonstrated that the
product is consistently glycosylated from batch to batch. One of the most critical
process steps with respect to glycosylation pattern of a protein is the cell culture
process (e.g. Jenkins et al. 1996). From the literature it appears as though factors
affecting glycosylation are different from protein to protein; both environmental and
kinetic variables were reported to influence the glycosylation pattern of proteins.
This paper discusses the influence of different process variables on the glycosylation
pattern of a recombinant fusion protein which contains eight N-linked glycosylation
sites. The carbohydrate groups contribute approximately 30% to the total molecular
mass of the protein. The protein is produced by a recombinant CHO cell line in fed-
batch mode. It was observed that minor changes in cultivation methods, timing of feed
addition or time of product harvest, could lead to a significantly different glycosylation
pattern of the protein. To identify the process variables influencing glycosylation of
this fusion protein, the study included fed-batch and continuous cultivations. Results
suggest that only specific productivity affects the glycosylation pattern of this
recombinant fusion protein.
2. Materials and methods
A CHO derived cell line was used in the study. It grew in serum-free low protein
medium. Methotrexate (MTX) was always present in the medium to maintain selective
pressure. All fed-batch and continuous cultures were performed in bench-scale reactors
controlled at 36 C, pH 7.2 and 50% air saturation. Appropriate amounts of fusion
protein were purified in a one column step, and glycosylation pattern was analyzed by
IEF (isoelectric focusing), asialo-FACE (fluorophore assisted carbohydrate
141
O.-W. Merten et al. (eds.). New Developments and New Applications in Animal Cell Technology, 141-147.
© 1998 Kluwer Academic Publishers. Printed in the Netherlands.
142
electrophoresis), OGS assay (Oxford Glyco-systems), and SEC (size exclusion
chromatography).
3. Results and Discussion
3.1. FED-BATCH CULTIVATION
Changes in protein glycosylation pattern during a typical fed-batch cultivation were
followed by analyzing samples taken during exponential, stationary and death phase.
Results from IEF, SEC and OGS analysis show that the glycosylation pattern of the
fusion protein changed over the course of the fed-batch culture. Between days 6 and 8
the IEF banding pattern shifted towards more acidic pH (see Figure 1), suggesting an
increase in the sialic acid attached to the carbohydrate groups of the fusion protein.
Indeed, analysis of sialic acid content confirmed that the sialic acid content of the
fusion protein increased by 20% during the cultivation. In addition to changes in sialic
acid content it was noted that the antennary structure was also altered. Figure 2 shows
the antennary structure of the fusion protein at different cultivation times, as measured
by OGS analysis: biantennary structures decreased, whereas tetraantennary structures
increased with cultivation time.
The observation that the cells only started to produce a more completely glycosylated
protein after the exponential growth phase suggested that one or more process variables
could be responsible for this. Throughout the cultivation specific growth rate
specific productivity (qP), growth phase, nutrient composition, ammonium
concentration, and culture osmolality changed. Because the culture environment is
altered steadily with time during fed-batch or batch cultivation, these methods are not
amenable to identifying critical process parameters affecting glycosylation
microheterogeneity. However, continuous culture experiments allow the control of the
culture environment independent of time. This technique is thus suitable for a thorough
investigation of the relationship of each of these growth variables with the
glycosylation pattern of the fusion protein.
3.2. CONTINUOUS CULTURE EXPERIMENTS
During long-term continuous cultures selective pressure was maintained with
methotrexate (MTX) to avoid a loss in fusion protein gene copy number.
During the experiment the following groups of variables were tested for their influence
on product quality: 1) environmental variables, i.e. asparagine concentration, glutamine
concentration, and ammonium concentration; and 2) kinetic variables, i.e. specific
growth rate and specific productivity. The glycosylation pattern was determined by
IEF, SEC, asialo-FACE, and OGS analysis.
None of the environmental variables tested in pulse and/or shift-experiments affected
the protein glycosylation pattern, although concentrations were increased one to three
fold for the amino acids, and seven fold for the ammonium concentration.
143
Steady states at dilution rates of 0.35/d and 0.20/d were established to investigate the
influence of specific growth rate on glycosylation pattern. Analysis by IEF, OGS and
SEC showed that glycosylation pattern was similar for both steady states although the
difference in growth rate was nearly 2-fold. Thus, the glycosylation pattern of the
fusion protein was independent of specific growth rate. Interestingly, glycosylation
pattern resembled the pattern obtained at the end of fed-batch cultivations
The glycosylation pattern did change significantly during a transient increase in ferric
citrate concentration (Figure 3), which resulted in an increase in both specific growth
and specific productivity. Coincident with these changes were observed an increase in
the amount of aggregation from 5.6 to 13.9 percent, and a decrease in terminal
sialylation as evidenced by the disappearance of the more acidic bands on an IEF gel
(not shown). Figure 4 shows that the change in ferric citrate concentration also resulted
in a decrease in tetraantennary glycoforms, with a corresponding increase in the
fraction of biantennary structures.
The observation that the complexity in antennary structure and sialic acid content
decreases when the specific productivity increases, agrees with the results obtained
during the fed-batch cultivation discussed above. For example, Figure 5 illustrates the
relationship between sialic acid content and specific productivity as observed during
fed-batch and transient continuous cultures. This strongly suggests that the specific
productivity is affecting the glycosylation of the fusion protein.
3.3. CHANGES IN INTRACELLULAR PROTEIN GLYCOSYLATION DURING
FED-BATCH CULTIVATION
There are two possible explanations for the observed relationship between
glycosylation pattern and specific productivity: 1) at high specific productivities the
glycosylation enzymes are saturated and can not glycosylate all proteins flowing
through the Golgi apparatus; or 2) the residence time of the protein in the Golgi
apparatus is too short to ensure complete glycosylation.
The following experiment was performed to investigate which of the two explanation is
most likely for the fusion protein of interest. A typical fed-batch experiment was
performed in duplicate reactors, where samples were taken throughout the cultivation
to analyze intracellular fusion protein quantity and quality.
Figure 6 shows a SDS PAGE Western blot of the intracellular fusion protein at
different time points of cultivation. The intensity of the protein band on the SDS PAGE
Western blot increased with cultivation time indicating an increase in intracellular
fusion protein concentration. This suggests that the fusion protein is accumulating in
the Golgi apparatus over the course of the fed-batch cultivation. As a consequence, an
increasing fraction of the product would be exposed to the glycosylation enzymes for a
longer period of time, and therefore, become more completely glycosylated. That
intracellular fusion protein becomes more fully glycosylated with cultivation time is
supported by the following three analysis: 1) the molecular mass of the intracellular
fusion protein appears to increase with cultivation time as evidenced by a shift to a
144
higher molecular mass of the protein band on the SDS PAGE Western blot (Fig. 6); 2)
the mass of glycan per gram of protein increased with cultivation time as shown by the
carbohydrate structure analysis (not shown); and 3) a clear shift towards more acidic
pH with cultivation time was observed on the IEF Western blot (not shown), suggesting
an increase in sialic acid content.
In summary, these results show that for studied cellular system, glycosylation is
inversely related to the specific productivity of the cell. Results further indicate that the
residence time of the protein in the Golgi apparatus is crucial for the extent of
glycosylation. However, this hypothesis would need confirmation by a pulse/chase type
experiment, which allows flow rates through the Golgi apparatus to be measured.
4. Conclusions
This project focused on the relationship between cultivation conditions and
glycosylation pattern of a recombinant fusion protein produced by a CHO derived cell
line. During the investigation it was found that the glycosylation pattern changed
during fed-batch cultivation: sialic acid content and antennary complexity increased
with cultivation time although ammonium concentration and most likely glycosidase
activities - two factors which reportedly can decrease sialic acid concentration -
increased simultaneously. Continuous cultures were performed to identify which
environmental and kinetic variables impact the glycosylation process. A change in
glycosylation pattern was only found when the specific productivity during a
continuous culture was increased transiently by an increase in ferric citrate
concentration. This indicated that it is a kinetic variable which influences the
glycosylation of the studied fusion protein. Further experiments focused on the
glycosylation of intracellular fusion protein and support the hypothesis that specific
productivity, or rather the residence time of the protein in the Golgi apparatus, is
crucial to the extent of glycosylation. Notably, concentration, molecular weight, and
sialic acid content of the fusion protein increased with cultivation time, suggesting that
a prolonged residence time in the Golgi apparatus leads to more complete
glycosylation.
5. References
Jenkins N., Parekh R.B. and James D.C. (1996) Getting the glycosylation right: implications for the
biotechnology industry. Nature Biotechnol. 14: 975-982.
The words you are searching are inside this book. To get more targeted content, please make full-text search by clicking here.
12_ANIMAL CELL TECHNOLOGY_707
Discover the best professional documents and content resources in AnyFlip Document Base.
Search
12_ANIMAL CELL TECHNOLOGY_707
- 1 - 50
- 51 - 100
- 101 - 150
- 151 - 200
- 201 - 250
- 251 - 300
- 301 - 350
- 351 - 400
- 401 - 450
- 451 - 500
- 501 - 550
- 551 - 600
- 601 - 650
- 651 - 700
- 701 - 707
Pages: