Ebook Update on production of recombinant therapeutic protein – Transient gene expression: Part 2

3

Optimisation of Transient Gene
Expression for Therapeutic
Protein Production
Jianwei Zhu

Transient gene expression (TGE) is a well-established technology for
rapid generation of recombinant proteins with human embryonic
kidney (HEK) and Chinese hamster ovary (CHO) cell lines. Many
TGE protocols have been published over the last ten years. A number
of representative protocols are described in the last section of Chapter
2 for producing therapeutic proteins at different scales (from 10 mL
-100 L), facilitated by calcium phosphate or polyethylenimine (PEI),
using viral or non-viral vectors. The protocol using 25 kDa linear
PEI (Table 2.3) as a transfection component to form a plasmid/
PEI complex for transfecting CHO and HEK cells has been widely
accepted by many laboratories and the biopharmaceutical industry.
Scalability and expression levels of TGE have improved significantly
in the past few years to the extent that it is now feasible to produce
100 g of protein for preclinical studies or even early phase clinical
development of therapeutics. Successful reports from the past few
years are listed and summarised in Table 3.1.
Table 3.1 presents the representative productivity level with the TGE
technology platform. Since 2008, when the production titre first
reached the 1 g/L milestone [1], there have been repeatedly successful
results either through TGE [2] or stable transfection pools [3] at, or
over, 1 g/L expression levels, making it feasible to use the TGE system
to produce therapeutic proteins in sufficient quantity for preclinical
studies or even early stage clinical trials. In the successful examples
(Table 3.1) both CHO and HEK293 cells were predominantly used

as host cell lines. Great effort has been invested into optimisation of
many aspects of the TGE system in coexpression of antiapoptosis
genes [1], the use of new additives in the media [4, 5], and transfection

81


82

Expression

81 mg/L

1000 mg/L

50 mg/L

80 mg/L

Cell

HEK293E

HEK293E

HEK293E

CHOK1SV

Mab

r-protein

Mab

Mab

Product

4) Process scaled up to 20 L in Wave bioreactor

3) Product showed similar glycosylation patterns
from batch to batch

2) Fed-batch process maintained for 14 days
with 3 post-transfections at 2-4 day intervals

1) DMSO and lithium acetate increased
expression

Ten-fold increase in human SEAP expression
was obtained in HEK293E cells compared with
pcDNA3.1 vector.

3) Process scaled-up to 2L

10% hp21, 5% FGF resulting in 1 g/L expression
level

2) Cotransfection with 37.5% HC, 10% LC,

10% hp18

1) Human CMV + Intron + WPRE (40 mg/L)

Enhancement by cotransfection of WPRE by
five-fold

Expression and process description

[5]

[75]

[1]

[53]

Reference

Table 3.1 Recombinant therapeutics produced by transient gene expression technology platform

Update on Production of Recombinant Therapeutic Protein


90 mg/L

3.1 mg/L

4-9.1 mg/L

50-60 mg/L

40-50 mg/L

30-60 mg/L

69-165 mg/L

CHO-DG44

CHO-S

HEK293

CHO-S

HEK293F

CHO-DG44

HEK293

Human IgG
Antibody

Mab

Mab

Fc-Fusion protein

Mab

2) Transfection at a high density

250 mg/L (at
500 mL scale)

Large scale in Wave bioreactor

Disposable shake bioreactor at 30 L working
volume.

Flask at 1 L scale using FreeStyle TM MAX.
Human erythropoietin and human blood
coagulation factor IX were expressed in both
HEK293 and CHO cells

Comparison of two hosts: CHO versus HEK293

VPA increased mRNA and protein levels

3) Eliminating the culture dilution step after
transfection

1) Reduced pDNA by 50%

300 mg/L
Mab
(at 5 mL scale)

Process scaled up to 110 L bioreactor with 80 L
working volume

CHO

Mab

22 mg/L

CHO-DG44

[13]

[127]

[123]

[124]

[4]

[6]

[20]

Optimisation of Transient Gene Expression

83

84
r-protein

60 pg/cell/day

100-400 mg/L

40 mg/L

235 mg/L

160 mg/L

206 mg/L

PER.C6

CHOK1SV

HKB-11

CHO

1.0 g/L

1.4 g/L

Chimeric antibody

300-500 mg/L

PER.C6

CHO-GS

Fc-Fusion protein

63 mg/L

CHO

Mab

Various

Mab

Mab

Mab

r-protein

Human IgG

1000 mg/L

HEK293F

Mab

140 mg/L

Epi-CHO

Stable transfection pool at 2 × 25 L scale

Stable transfection pool at 5 L scale

CHO and lentiviral vector expression system

Produced in 10 L wave bioreactor

Seven different antibodies were expressed in the
system. Cells were transfected by eletroporation
and stable pools were selected. Productions were
at 100 L scale

Roller bottle scale

Roller bottle scale

Adenovirus-based expression at small scale

High density cell culture using Expi293 Medium

EBVNA1 and oriP in expression plasmid

[3]

[130]

[131, 132]

[25]

[129]

[129]

[128]

[2]

[78]

Update on Production of Recombinant Therapeutic Protein


121-405 mg/L

Mab

Stable transfection pools at 20-200 L WAVE or
300 L stirred-tank

CMV: Cytomegalovirus
DMSO: Dimethylsulfoxide
DNA: Deoxyribonucleic acid

EBVNA1: Epstein-Barr virus nuclear antigen 1
Epi-CHO: Transient expression system in Chinese hamster ovary
Fc: Fragment crystallisable
FGF: Fibroblast growth factor
HC: Heavy chain
HKB-11: Hybrid of human embryonic kidney 293 and a human B cell line
hp: Human cell cycle regulatory protein
Ig: Immunoglobulin
IgG: Immunoglobulin G
LC: Light chain
Mab: Monoclonal antibodies
mRNA: Messenger ribonucleic acid
oriP: Plasmid origin of viral replication
pDNA: Plasmid deoxyribonucleic acid
RNA: Ribonucleic acid
r-protein: Recombinant protein
SEAP: Secreted alkaline phosphatase
VPA: Valproic acid
WPRE: Woodchuck hepatitis virus post-transcriptional regulation element

CHO-GS

[25]

Optimisation of Transient Gene Expression

85


Update on Production of Recombinant Therapeutic Protein

at high cell density [6, 2]. However, compared with an optimised
stable transfection expression, TGE showed that the volumetric titres
using TGE are lower and this is still a limiting factor. The overall
expression yield is at least five- to ten-fold lower than that obtained
with stable expression cell lines for the production of therapeutic
recombinant proteins and Mab. The majority of published titres
in a TGE system are in the range of 10-1000 mg/L with a specific
productivity range of 1-10 pg/cell/day, whilst an optimised stable gene
expression (SGE) process could reach 100 mg-10 g/L with a specific
productivity of 10-80 pg/cell/day (Table 1.1, Table 2.3 and Table 3.1).
As time- and resource-consuming cell line development is eliminated,
every batch production run has to start with plasmid preparation
and transfection. Low productivity causes the manipulation of
increasing quantities of culture media and vector DNA. Optimisation
of the productivity through higher transfection rate, better culture
conditions for cell growth leading to better expression, simplification
of the procedure, and extension of production duration have been
highly desirable. Ultimately, the technology platform is aimed at
large scale production to supply clinical grade material for early
phase human clinical trials. In this chapter we will present recent
developments in the optimisation of TGE methodology in order to
maximise the production capacity for therapeutic applications.

3.1 Optimisation of the Transient Gene Expression
Conditions
Several strategies have worked well to improve overall yields using
CHO cells for TGE including transfection media optimisation,
high density transfections, mild hypothermia, and vector design
improvement [1, 7-12, 123]. A widely acceptable protocol for
transfection using linear 25 kDa PEI with CHO or HEK cells to

generate milligram to gram quantities of therapeutic proteins and
monoclonal antibodies (Mab) are available. However, to maximise
the capacity of the technology platform, studies are still needed to
optimise overall cell growth to increase the transfection efficiency,
to simplify the protocols, and ultimately to improve the expression
86


Optimisation of Transient Gene Expression
and productivity. In the following sections the optimisation of TGE
cell culture conditions including medium components, additives for
the medium, optimisation of construction and TGE procedure will
be updated and discussed.

3.1.1 Medium Optimisation
Over the last several years, extensive research into media formulations
for the manufacture of recombinant therapeutics has led to serumfree cultures becoming the standard procedure for mammalian cell
growth in suspension. Hence a number of chemically defined media
which support high cell densities and facilitate the expression and
purification of recombinant proteins have been developed [13,
14]. The following are commonly used: FreeStyle 293, OptiMEM,
OptiPRO SFM, and chemically-defined CHO (CD-CHO) from
Invitrogen; D/H from Biochrom; UltraCHO and ProCHO5 from
Lanza, as shown in Table 3.2, which summarises a list of media
which have been used successfully for TGE since 2007. In most cases,
these media have been developed specifically for the cultivation of
HEK293 or CHO cells, the major hosts for TGE.
The choice of culture medium has a significant impact on
transfection efficiency and productivity. Many serum-free media
formulations may not support good transfection efficiency [18,

21]. For example, transfection with calcium phosphate requires the
presence of serum, therefore transfections with this compound was
performed in media such as Dulbecco's modified Eagle’s medium
(DMEM)/F-12 (Ham’s nutrient mixture formulation) plus 1% fetal
bovine serum (FBS) [16, 17]. Whilst newly developed media greatly
improved cell growth and DNA transfection, it is also possible to
perform PEI-mediated serum-free transfection in a minimal medium
such as Roswell Park Memorial Institute 1640 (RPMI 1640) [15].
As stated previously, linear PEI with a molecular weight of 25,000
has been a standard key reagent to facilitate plasmid delivery
[18-20]. Transfection competent polycation-DNA complexes
have net positive charges and are thought to bind cells through
87


Update on Production of Recombinant Therapeutic Protein
ionic interaction with negatively-charged membrane-associated
proteoglycans [22]. As heparin or dextran sulfate is usually present
in serum-free formulations to reduce cell aggregation, the low
transfection efficiency observed when using such formulations is
most likely a result of polyplex neutralisation by these polyanionic
molecules. Indeed, it has been reported that heparin strongly inhibits
polycation-mediated TGE [22]. To overcome this inhibition, a
complete medium exchange for a transfection-competent medium
is often performed prior to transfection [23, 24]. Alternatively,
neutralisation of the polyanions in the medium could probably be
achieved by using higher PEI:DNA ratios, or by adding free PEI to
the culture prior to transfection.
Table 3.2 shows a summary of culture media used in TGE and their
suppliers.

As mentioned previously, a cell culture medium is often not able
to support high transfection efficiency, so changing the medium
is a common procedure in order to reach optimum transfection
efficiency. In Table 3.2, two sets of media are listed for cell culture and
transfection. Most of these media are used in published experiments
and are commercially available; however exact components may not
be disclosed except for the popularly used media DMEM and F12.
From the production scale-up point of view, if a medium change
could be avoided by using a single formulation to support growth
and transfection, the target of realising 1g/L expression level at
production scale would be quite promising.

3.1.1.1 Peptones
Protein hydrolysates (peptones) are used in serum-free cell culture
media for biosafety reasons and to facilitate downstream processing.
Peptones provide substantial nutrients including trace elements for
a serum-free cell culture medium. Supplementation of standard
protein-free media with peptones yielded a significant increase in
TGE productivity in HEK293E cells [9, 10]. The effects of many
protein hydrolysates were evaluated on cell proliferation, transfection
88


Culture medium

Ex-cell 293

FreeStyle + SFX4HEK

Ex-cell 293

FreeStyle 293

HyQSFM4TransFx293

M11V3

FreeStyle

Ex-cell V-Pro

DMEM/F12 or EpiSerf

Ex-cell 293

Protein Expression
Medium + supplement

DMEM/F12 (D/H) or
EpiSerf

Cell

HEK293E

HEK293E

HEK293E

HEK293F

HEK293
SF-3F6

HEK293T

HEK293E

HEK293E

HEK293E

HEK293

CAP-T

VERO

Biochrom or
Invitrogen

Invitrogen

Sigma-Aldrich

Biochrom or
Invitrogen

SAFC Biosciences

Invitrogen

Novartis Proprietary

Hyclone

Invitrogen

Invitrogen

Invitrogen/HyClone

SAFC Biosciences

Supplier

Invitrogen

Invitrogen

Invitrogen

Invitrogen

Supplier

DMEM/F12 (D/H) or
EpiSerf

OptiMEM

RPMI 1640 +
supplement

DMEM/F12 or
EpiSerf

DHI/SAFC medium

293 SFM II

M11V3

Biochrom or
Invitrogen

Invitrogen

Lonza

Biochrom or
Invitrogen

SAFC
Biosciences

Invitrogen

Novartis
proprietary

HyQSFM4TransFx293 Hyclone

OptiPRO SFM

DMEM + FBS

FreeStyle 293

FreeStyle 293

Transfection medium

Table 3.2 Culture media used for cell growth and DNA transfection

[26]

[133]

[15]

[26]

[19]

[134]

[13]

[135]

[123]

[34]

[24]

[1]

Reference

Optimisation of Transient Gene Expression

89


90

CD-CHO/DMEM/F12

CD-CHO +
supplement

CHO-S-SFMII + ACA* Invitrogen

CD-CHO

CD-CHO

DMEM

CHO-S

CHO-S

CHO-T

CHOK1SV

CHOK1SV

CHO
DUKX-B11
Lonza

CHO-DG44 ProCHO5

Invitrogen

ProCHO5

CHO-S-SFMII

DMEM

DMEM/FBS/GS
supplements

UltraCHO

CHO-S-SFMII

CD-CHO +
supplement

DMEM

Lonza

Invitrogen

Biochrom

JRH
Biosciences

Lonza

Invitrogen

Invitrogen

Invitrogen

DMEM + supplements Introgen

OptiPRO SFM

[4, 6, 11, 12,
20, 30]

[27]

[136]

[5, 38]

[5]

[78]

[32]

[16]

[34]

[123]

ACA: Anticlumping agent
CHOM: A CHO cell growth medium developed by Chiang and co-worker (Chiang GG, Sisk WP. 2005. Bcl-xL
which mediates increased production of humanised Mab in Chinese hamster ovary cells [85].
GS: Glutamine synthetase
VERO: Cell line derived from kidney of an African green monkey

Chiang and Sisk

CHO-DG44 CHOM

Biochrom

Invitrogen

Invitrogen

Invitrogen

Invitrogen

Lonza

ProCHO5

CHO-S

Invitrogen

FreeStyle CHO +
supplements

CHO-S

Update on Production of Recombinant Therapeutic Protein


Optimisation of Transient Gene Expression
efficiency, and volumetric productivity using a model protein such as
SEAP as a reporter gene [9]. The addition of the gelatin peptone N3
and removal of bovine serum albumin slightly enhanced transfection
efficiency and significantly increased volumetric productivity by

four-fold. A 293 cell line that stably expressed EBVNA1 (HEK293E)
and which was capable of growing in a low cost gelatin peptone
N3-fortified serum-free medium was developed [9]. Furthermore,
the same study group optimised the feeding process by a single
pulse of peptones (protein hydrolysates) to the cultures in a serumfree medium which resulted in a significant increase in volumetric
protein productivity. Sixteen peptones from different sources were
tested and almost all of them showed a positive effect on r-protein
production. The relative abundance of the mRNA of SEAP suggests
that the improvement in protein yield results from an increase of both
the translational activity and transcription efficiency [10]. Currently
optimised TGE protocols often include the addition of supplements
(which may contain peptones) and peptone in both the basal and
feeding media [25-27].

3.1.1.2 Valproic Acid
Medium additives may play important roles in cell growth and DNA
transfection. VPA is a small molecule that inhibits histone deacetylase
activity. It has been reported recently that VPA enhances recombinant
mRNA and protein levels in transiently transfected CHO-DG44 cells
[4]. The steady-state levels of the messenger ribonucleic acids of IgG
light and heavy chains were nearly ten times higher than those in the
untreated control transfection, even though the level of transfected
plasmid DNA was the same in the presence or absence of VPA [4].
Addition of VPA to HEK293E cells transfected with a plasmid for
IgG expression led to an 11-fold increase in productivity. However, in
the same report, VPA increased the yield for Mab but not Fc-fusion
protein [28]. VPA, as a chemical additive, exerts wide ranging effects
on various aspects of protein expression, some direct (transcription,
translation, secretion), and some indirect (effects on cell cycle and
cell division), similar to those shown for other chemical additives

91


Update on Production of Recombinant Therapeutic Protein
such as sodium butyrate. It was reported that VPA has a negative
impact on cell growth [29]. Cell growth and viability were decreased
while productivity was increased five- to ten-fold by adding the
VPA [4]. In order to fully elucidate the role of VPA, further analysis
would be required using high throughput methods such as DNA
microarrays. Several reports showed that VPA significantly affects
both cell viability and specific productivity [1, 4]. VPA is added to
TGE media in current optimised protocols [15].

3.1.1.3 Other Additives
To achieve a scalable CHO cell TGE process with high level protein
expression, Ye and co-workers screened different media for optimum
transfection and product expression [5]. UltraCHO (Lonza), with
added DMSO and lithium acetate, was found to improve CHO
transfection expression level significantly. DMSO and lithium acetate
have been proposed to increase DNA transportation into the cells
by changing cell membrane or wall permeability, but the exact
mechanisms remain unknown [5]. Addition of growth factor(s) into
a culture system may improve cell growth, which may lead to better
productivity [7, 29].
While the TGE volumetric productivity has improved significantly over
the past decade, the amount of plasmid DNA needed for transfection
remains very high. The use of nonspecific (filler) DNA to partially
replace the transgene-bearing plasmid DNA in transfections of CHO
and HEK293E cells was reported [30]. When the optimal amount of
coding plasmid DNA (pDNA) for either host was reduced by 67% and

replaced with filler DNA, the r-protein yield decreased only by 25%
relative to the yield in the control transfections. Filler DNA did not
affect the cellular uptake or intracellular stability of coding pDNA,
but its presence led to increases of the percentage of transfected cells
and the steady-state level of transgene mRNA compared to control
transfections. The results suggest that filler DNA allows the coding
pDNA to be distributed over a greater number of DNA-PEI complexes,
leading to a higher percentage of transfected cells [30].

92


Optimisation of Transient Gene Expression

3.1.2 Optimisation of Transient Gene Expression
Conditions and Procedures
3.1.2.1 Process Design
Design-of-experiments (DoE) methodology has been widely used in the
development of biotechnological processes [31]. Methods such as factorial
design, response surface methodology, and DoE provide powerful and
efficient ways to optimise cultivations and other unit operations and
provide procedures using a reduced number of experiments. The
multitude of interdependent parameters involved within a unit operation
or between units in a bioprocess sequence may be substantially refined
and improved by the use of such methods [31]. The productivity of
TGE depends on multiple factors including the proportion of the cell
population transfected, the capacity of cells to grow over the time in a
given medium, the efficiency of transcription and translation in the host
cells, and post-translational apparatus. More specifically, DoE with
the Box-Wilson design [32] or Box-Behnken design [33, 34] is used to

design the optimum conditions for the effects of DNA concentration, PEI
concentration, and incubation time on transient transfection efficiency
[33] or cell density for transfection [32, 34]. The optimised conditions
were used in 5 L stirred-tank bioreactor runs [33] to validate the results
from small scale experiments. The main difficulty in the response surface
methodology lies in the choice of initial design space for the subsequent
response surface optimisation. Generally, using factorial designs, one can
gradually move towards an optimum response. However, a rapid method
of using empirical cell specific cytotoxicity to free PEI in the culture
medium was used in the report by Thompson and co-workers [32].

3.1.2.2 Optimisation of Culture Conditions
There are many published methods which discuss the simplification
of the TGE procedure [6, 34], by lowering the culture temperature
during expression, and using hyperosmolarity.
A simply optimised method for TGE in suspension-adapted CHO
cells, using PEI for DNA delivery, was presented by Wurm’s group.

93


Update on Production of Recombinant Therapeutic Protein
The procedure was further simplified via elimination of a dilution step
after transfection [6]. Both the transfection and production phases of
the bioprocess were performed at a density of 4 × 106 cells/mL at a
low temperature of 31 °C as previously reported by the same group.
In addition, the amounts of both PEI and plasmid DNA were reduced
by up to 50% on a per cell basis compared to previously published
protocols from this laboratory, resulting in higher cell viability after
transfection and higher volumetric recombinant protein yields. In batch

cultures of up to 14 days, reproducible recombinant antibody yields up
to 300 mg/L were achieved at the small scale of 5 mL working volume,
and up to 250 mg/L at the large scale of 500 mL. The simplicity and
improved yields are expected to increase the utility of CHO cells for the
rapid production of recombinant proteins by TGE at larger scales [6].
To ensure maximum productivity of recombinant proteins during
production culture, it is usual to encourage an initial phase of rapid
cell proliferation to achieve high biomass, followed by a stationary
phase where cellular energies are directed towards the production
of r-protein. During many such biphasic cultures, the initial phase
of rapid cell growth at 37 °C is followed by a growth arrest phase
induced by reduction of the culture temperature. Low temperatureinduced growth arrest is associated with many positive phenotypes
including increased productivity, sustained viability and an extended
production phase, although the mechanisms regulating these
phenotypes during mild hypothermia are poorly understood [35, 36].
Temperature during biopharmaceutical product expression has been
one of the factors that has a significant impact on the productivity.
Besides the improvements achieved through medium components
and additives, expression levels were also increased more than threefold at the low temperature of 31 oC compared with 37 oC [12].
The increase in TGE correlated with the accumulation of cells in the
G1 phase of the cell cycle, increased cell size, higher cell viability,
higher steady-state levels of transgene mRNA, reduced consumption
of nutrients and decreased accumulation of waste products. The
enhancement of TGE was not vector-dependent, but the presence of
WPRE in the 3′ untranslated region of the transgene mRNA increased
the transient recombinant Mab expression more than three-fold
at 31 °C as compared to expression at 37 °C. The yields achieved
94

Optimisation of Transient Gene Expression
by the low temperature enhancement of TGE in CHO cells makes
this technology feasible for a rapid production of gram amounts of
secreted recombinant proteins at large scale (up to 100 L).
The effect of hyperosmolarity on transient r-protein production in
CHO cells was reported. Addition of 90 mM sodium chloride to the
production medium ProCHO5 (Lonza) increased the volumetric yield
of recombinant antibody up to four-fold relative to the transfection
in ProCHO5 medium alone. Volumetric yields up to 50 mg/L were
achieved in a six day batch culture of 3 L. In addition, hyperosmolarity
reduced cell growth and increased cell size [37].
Optimisation of culture conditions can be a wide ranging task. In
many cases final improvements in production yield resulted from
many factors, including genetic construction, medium additives,
culture environment (temperature and osmolality), and a combination
of all of the above [1, 12, 38]. One typical example was due to a
combination of the addition of growth factor(s) into the culture
system and reduction of the culture temperature [38]. In a different
study, the addition of recombinant insulin-like growth factor (LR3IGF) and a reduction in the culture temperature to 32 °C were found
to increase product titre two- and three-fold, respectively. However,
mild hypothermia and LR3-IGF acted synergistically to increase
product titre by 11-fold. Although increased product titre in the
presence of LR3-IGF alone was solely a consequence of increased
culture duration, a reduction in culture temperature post-transfection
increased both the integral of viable cell concentration and cellspecific Mab production rate. Galbraith and co-workers called
this a significant improvement when using a combination of mild
hypothermia and growth factor(s) to yield an extended ‘activated
hypothermic synthesis’ [38]. Addition of WPRE in expression
construction and reducing culture temperature from 37 oC to 31 oC as
mentioned earlier in this section increased productivity by three-fold

[12]. The most impressive example is still the report by Backliwal
and co-workers of reaching 1 g/L expression through optimisation
of construct by including intron and WPRE, cotransfection of the
antiapoptosis genes hp18 and hp21, and introducing a growth factor
into the expression system [1].
95


Update on Production of Recombinant Therapeutic Protein

3.1.3 Construction Optimisation
Gene expression levels in mammalian TGE system largely depend,
among other things, on the strength of the transcriptional
regulation elements including the promoter, enhancer, intron, and
polyadenylation signal [39, 40]. Modifying the vector in both the
backbone and the insertion containing the gene of interest (GOI) can
greatly influence the total amount of protein produced from each cell.
Optimisation of a construct which focuses on the promoter, signal
sequence, and other genetic elements in TGE will most likely lead to
a much better expression result.

3.1.3.1 Promoter
Most attention has so far been directed to the promoter and its
associated elements in order to increase transcription levels. The
promoters used are generally derived from the genes of viruses or
highly expressed mammalian genes, with the viral CMV and Simian
vacuolating virus 40 (SV40) promoters being the most commonly
used. As well as these two, human gene promoters including
glucose-regulated protein (Grp78), a combination of the CMV early
enhancer element and chicken β-actin promoter (CAG), elongation

factor-1α (EF-1a), ubiquitin C (UbC), ferritin heavy chain (FerH),
and phosphoglycerate kinase-1 (PGK) promoters, and others are
frequently used to drive high levels of gene expression in mammalian
expression vectors. Here we illustrate one experiment within
which a number of promoters were evaluated as described in the
protocol of Hopkins and co-workers [24] to screen the best choice
of a promoter before deciding on the final production construct.
An r-protein as a model product was expressed in HEK293E cells
that were transiently transfected with a plasmid similar to the
procedure shown in Figure 1.1 for product expression. Supernatants
were analysed by sodium dodecyl sulfate polyacrylamide gel
electrophoresis (SDS-PAGE) and Western blotting that was scanned
and quantified. The intensity of the product signals was plotted as
shown in Figure 3.1 to indicate the effectiveness of each promoter
in the construct.
96


Optimisation of Transient Gene Expression

Figure 3.1 Comparisons of promoters for r-protein expression
with TGE. Relative expression levels versus seven promoters
(CMV, Grp78, EF-1α, UbC, FerH, PGK and CAG) are compared
Seven promoters and one expression enhancer were cloned into a
typical plasmid vector as shown in Figure 1.1. Small quantities of
the recombinant plasmids were amplified in Escherichia coli and
purified through a process described in Chapter 2. The plasmids
were transfected into HEK293E cells which were cultured for an
additional three days before harvest. Expression levels of the product
in harvested conditioned media were measured through SDS-PAGE

and Western blotting. Intensities of the Western blot images were
analysed and calculated for comparison.
As shown in Figure 3.1, of the three better expressed samples
associated with EF-1a, CMV, and CAG promoters, the expression level
by the EF-1a promoter was the highest for the specific product and
construct. In addition to the seven promoters tested, the effectiveness
of WPRE in combining with CMV was also explored in the same
experiment. However, some outstanding expression results were
reported elsewhere. Among the constructs tested, significant differences
in product expression was observed, which justified the initial
97


Update on Production of Recombinant Therapeutic Protein
screening of the promoters prior to finalising a production construct.
Incorporating promoter screening results with other construction
optimisation, such as leader sequence and codon optimisation, prior
to finalising a production construct would be a wise approach.
The promoter CMV has been the most popularly used in mammalian
cell expression systems due to its strong promoter activity. The
human, mouse and rat CMV along with another strong promoter, the
myeloproliferative sarcoma virus (MPSV) promoter, were investigated
in the presence or absence of intron A for protein expression [40]. The
protein expression levels of four GOI driven by these promoters were
evaluated in HEK293EBVNA and CHO-K1 cells that were either stably
or transiently transfected. In general, the full length human CMV,
when in the presence of intron A, gave the highest levels of protein
expression in transient transfections in both cell lines. However, the
MPSV promoter provided the highest levels of stable protein expression
in CHO-K1 cells. Using the CMV driven constitutive promoters in the

presence of intron A, a yield of greater than 10 mg/ml of recombinant
protein was obtained using transient transfections[40].
Viral replicon-based expression systems have been established for
transient protein expression. Influenza A replicon for the expression
of recombinant proteins was compared with CMV promoter-driven
green fluorescent protein (GFP) expression in HEK293 cells [41]. The
overall expression level was much lower than the CMV promoter
but 100-fold higher than the SV40 promoter. When a single secreted
protein, for example, an antibody light chain, was expressed by the
influenza replicon, it resulted in five-fold higher expression levels
compared to the commonly used CMV promoter-based expression.
Thus the influenza A replicon system may be considered for high level
expression of complex proteins in mammalian cells [41].

3.1.3.2 Codon Optimisation
Codon optimisation of DNA vectors can enhance protein expression
both by enhancing translational efficiency, and by altering RNA
stability [42]. Several codon optimisation programs companies
98


Optimisation of Transient Gene Expression
are available commercially: GeneGPS (DNA2.0), GeneArt (Life
Technologies), OptimumGene (GenScript). guanine-cytosine content
(GC content) can be one of the optimisation targets. The GC-rich
genes can be several-fold to over a 100-fold more efficient than their
GC-poor counterparts [43, 44].

3.1.3.3 Leader Sequence
A signal peptide at the N-terminal of a protein directs the molecule

into the endoplasmic reticulum and ultimately outside the cell
membrane. Signal peptidase is cleaved off once it has served its purpose
of translocating the protein [45]. Selection of the signal peptide is
vitally important when aiming to produce maximum amounts of
r-protein in a mammalian expression system. Similar to the promoter
screening for an optimum expression, screening signal leaders may
have a huge impact on final expression levels [44]. There are many
choices which can be considered in the screening: native leader, human
immunoglobulin K, immunoglobulin E (IgE), preprotrypsin, human
albumin, chymotrypsinogen, tissue plasminogen activator (tPA),
human interleukin-2, granulocyte macrophage colony-stimulating
factor (GMCSF), human trysinogen-2 and others [44, 45, 46]. A native
signal peptide is not necessarily always the most effective [44, 45, 47].
Replacing the native secretory signal peptide of the cytokine product with
three different signal peptides (tPA, GMSCF and IgE) resulted in faster
secretion in the medium and increased accumulation of the cytokine in
the extracellular compartment [44]., The ability of a leader sequence to
secrete protein may be improved via DNA mutagenesis. By increasing
the basicity of the n-region and the hydrophobicity of the h-region in the
interleukin (IL)-2 signal peptide, Zhang and Robinson [48] observed up
to a 3.5-fold increase in secretion levels of a reporter protein.
Figure 3.2 shows a case analysis, in which several signal leader
sequences were compared in HEK293E cell transient expression
of a recombinant protein. Supernatant samples were analysed by
SDS-PAGE, Western blotting, and size exclusion chromatography
(SEC)-high performance liquid chromatography (HPLC). The
samples labeled M01 to M15 represent constructs with combination
99

Update on Production of Recombinant Therapeutic Protein
of different leader and promoter sequences. Reference material was
used as ‘reference’ and GFP was used as the ‘control’.

Figure 3.2 Analysis of r-protein expressed by transient gene
expression. a) SDS-PAGE analysis of r-protein expression by
TGE; b) Western blotting analysis of r-protein expression by
100


Optimisation of Transient Gene Expression
TGE; and c) SEC-HPLC analysis of r-protein expression by TGE.
A molecular weight standard (MW STD) such as Mark 12 from
Invitrogen was used as reference of molecular size. A commercial
source of the r-protein was used as the positive control
Plasmids were transfected into HEK293E cells for transient
expression. Three days post-transfection, supernatants were taken
from cell culture containers for SDS-PAGE analysis (Figure 3.2a);
Western blotting analysis (Figure 3.2b); and SEC-HPLC analysis
(Figure 3.2c). Western blotting was conducted using a specific
monoclonal antibody.
Figure 3.2 shows a typical set of analysis data for the cell culture
supernatant samples. The purpose of the assay was to screen the
best signal leader sequence and optimum combination of leaders
and promoters for expression of an r-protein. As it can be clearly
seen on Figure 3.2a and 3.2b, M10-M12 expressed products with
only trace amounts which were barely detectable by either SDSPAGE or Western blotting. Interestingly, these three samples were
associated with one specific leader in the three constructs. The poor
expression could be due to low activities at either transcriptional
or translational levels, however, it was most likely caused by low

efficiency of the leader sequence in secretion. For the rest of the sample
groups, expression of the r-protein was reasonably satisfactory. To
validate the results, a SEC-HPLC method was developed to have
sufficient resolution for product quantitation. As shown in Figure
3.2c, the product peak was well separated from other impurities,
which demonstrated the usefulness of this analytical tool to quantify
samples from the cell culture.

3.1.3.4 Other Genetic Elements
Low yield from TGE in mammalian cells limits its application to
areas where large amounts of proteins are needed. One effective
approach to enhance TGE levels is to introduce genetic elements
including post-transcriptional regulatory elements (PTRE), WPRE,
and matrix attachment regions (MAR).
101


Update on Production of Recombinant Therapeutic Protein
3.1.3.4.1 Post-transcriptional Regulatory Elements
The effects of five PTRE including the DNA sequence in the 5ʹ
untranslated region, intron A, WPRE, and leader sequence on the TGE
of a number of proteins in different cell lines were evaluated [49-51].
Most of the elements increased expression but exhibited cell-specific and
gene-specific effects. The tripartite leader sequence of human adenovirus
mRNA linked with a major late promoter enhancer provided the
most universal and highest enhancement of gene expression levels. It
increased the expression of the proteins that expressed in CHO and
HEK cells by 3.6- to 7.6-fold. Combinations of multiple PTRE increased
protein expression by as much as 10.5-fold [51].
3.1.3.4.2 Woodchuck Hepatitis Virus Post-transcriptional

Regulatory Element
WPRE has been known to enhance gene expression by its effect
on nuclear mRNA processing, mRNA export, and translation
[52]. Wulhfard and co-workers used WPRE to improve antibody
production in transiently transfected CHO cells in combination with
the effect of low culture temperature [12]. WPRE was employed
to increase the antibody production by HEK293E cells [53]. As a
result, the antibody production was increased by 5.5-fold through
the enhancement of total mRNA levels of HC and LC, and the
efficient export of nuclear mRNA into the cytoplasm. Using WPRE,
1.9 mg of cumulative recombinant antibody was obtained in
transiently transfected adherent HEK293E cells from one 100 mm
dish transfection with 10 mL medium exchange every three days
for 24 days of cultivation [53]. In addition, the highest recombinant
antibody concentration of 81 mg/L was obtained. The highest titre of
production so far reported using TGE technology was 1 g/L, where
WPRE was used in the expression vector [1].
3.1.3.4.3 Matrix Attachment Regions
MAR are DNA sequences that help to generate and maintain an
open chromatin domain that is favourable for transcription and
102


Optimisation of Transient Gene Expression
may also facilitate the integration of several copies of the transgene.
By incorporating MAR into expression vectors, an increase in the
proportion of high producer cells as well as an increase in protein
production was reported [58]. MAR are DNA elements that bind
to the nuclear matrix, a protein structure in the cell nucleus. There
are many reports on the influence of different MAR on transgene

expression. For example, the chicken lysozyme matrix attachment
region and the human β-globin matrix attachment region have been
shown to have a positive effect on transgene expression in CHO cells,
whereas the chicken α-type matrix attachment region has a negative,
transcription-reducing effect. Not much is known about how MAR
work at the molecular level and why they can confer beneficial effects
on transgene expression [59].
To improve mammalian cell expression systems, a variety of
matrix/scaffold attachment region elements were screened for their
ability to insulate transgene expression from the position effects
in CHO cells [8]. The human β-globin matrix attachment region
element is particularly effective as the frequency of β-galactosidasepositive colonies was increased by up to 80%. Furthermore, the
expression levels of these colonies were enhanced seven-fold. These
improvements appear to be related to the increased copy numbers and
a higher efficiency of expression of the integrated genes. A uniform
growth property by a simple two-step amplification process involving
two concentrations of methotrexate was established to generate high
producers. This eliminates the need to isolate individual colonies
followed by multistep treatments of methotrexate and thereby greatly
simplifies this mammalian expression system [8].

3.1.4 Coexpression of Growth Factors
Growth factors are crucial for cell growth and product expression.
Addition of growth factors to chemically defined media, which
normally contain only insulin and transferrin, is a common approach
in mammalian cell culture. Here, TGE offers the opportunity to
coexpress those growth factors together with the GOI. It was reported
that coexpressing acidic fibroblast growth factor (aFGF) led to
103

Update on Production of Recombinant Therapeutic Protein
transient protein titres of approximately 300 mg/L. This represents
a 50% increase in both antibody titres and specific productivity in
transiently transfected HEK293E cells in the experiment [7, 121, 122].
Similar effects were observed with Fc-tagged fusion proteins [47].
It seems that this increase in expression starts to take effect four to
six days after transfection, possibly because of the requirement of
a minimum cut-off concentration of aFGF in the medium. Whereas
one could also add growth factors such as aFGF directly to the
medium, direct coexpression offers the advantage of not having to
source expensive growth factor protein and might avoid the need to
add growth factors.
FGF enhanced transgene expression in CHO cells but not in HEK293
cells [121, 122]. Increased productivity in CHO cells may be due to
a FGF-induced recombinant ribonucleic acid synthesis [121, 122].
The use of protein kinase B to potentiate VPA increases heterologous
gene expression in mammalian cells, especially CHO cells [121, 122].
Fusion partners can increase the expression of recombinant
interleukins via transient transfection in 2936E cells. The expression
levels of five secreted target interleukins (IL-11, IL-15, IL-17B, IL32, and IL-23 p19 subunit) were tested with three different fusion
partners in 2936E cells. When fused to the N-terminus, human serum
albumin was found to enhance the expression of both IL-17B and
IL-15 cytokines which did not express at measurable levels on their
own [47].

3.2 Extension of Protein Production after Transfection
One major advantage of the TGE technology platform is the
short development time: the production phase begins shortly after
transfection, which eliminates time- and resource-consuming cell line

selection procedures. Although TGE can provide milligram to gram
amounts of recombinant proteins within a few days, each production

104


Optimisation of Transient Gene Expression
will have to start from plasmid preparation and transfection to
prepare the host cell for expression. Normally, the production stage
with an expression plasmid maintained in the host cells lasts 3-14
days, depending upon the culture conditions. The limited production
period is a bottleneck for the production via TGE. The technical
limitations and high cost of pDNA hamper the application of this
technology at larger scales. Therefore, extension of the production
phase of TGE is desirable to improve overall productivity.
Extension of protein production may utilise stable transfection pools
which allow transfected cells to grow for a much longer time when
they are in the production phase. It is, in principle, consistent with
the overall goal of quickly producing therapeutic biologics whilst
avoiding the costs associated with making pDNA and selecting high
producers. Such stable transfection pools can be further improved
by incorporating several genetic elements for protein expression
including transposons, ubiquitous chromatin opening element
(UCOE), expression augmenting sequence elements (EASE), a gene
targeting element, and matrix attachment region sequences.

3.2.1 Stable Transfection Pool
Using stable transfection pools is a strategy that can extend the
post-transfection expression time by keeping those cells that contain
a GOI. Stable transfection pools are generated by transfecting a

vector with a genetic selection marker such as antibiotic resistance or
essential metabolic enzyme markers. This is an alternative approach
for producing recombinant proteins and Mab which combines the
advantages of both TGE and SGE, along with being time-, labour-,
and cost-effective, and a reasonably high productivity. However, the
potential weakness of this system is unstable transgene expression.
The pools are known to be heterogeneous, where cells are in different
status in terms of their transgene integration sites. Consequently, the
stability and product expression levels are not the same.

105