Escherichia coli is a valuable organism for the
high level production of recombinant protein in high
cell density culture. The expressed heterologous proteins
are usually accumulated inside the cells, which
can be easily degraded by cellular protease[ 11. Moreover,
foreign proteins can form inactive inclusion
bodies [2] caused by misfolding, and the refolding is
quite inefficient and expensive. Secretion of the heterologous
proteins reduces not only the toxic effects
on the host cells, but also the difficulties and cost in
recovery and purification.
Another disadvantage of E. coli is accumulation
of acetate during cultivation, which inhibits cell
growth and the foreign gene expression [3,4], and it is
more deleterious to the recombinant cells[S]. Many
strategies have been adopted to reduce the formation
of acetate by process control[3,6] and metabolic engineering[
7-9]. An acetate-tolerant mutant of E. coli
DHSa, DA19, was isolated by Zhu et aZ.[10], which
forms less acetate on glucose than DHSa and grows
better in the presence of acetate. This feature is especially
useful in large-scale fermentors, where uniform
distribution of the added concentrated glucose feed is
more difficult than in small ones.
Human epidermal growth factor (hEGF) is a peptide
composed of 53 amino acids. It can promote proliferation
of epidermal cells and inhibit the secretion
of gastric acid, and has wide cosmetic and medical
applications. We have studied hEGF production in E.
coli YKS37 using complex media, and the secreted
hEGF concentration reached 686mg.L - [ 11,121. A
preliminary study has been carried out with DHSa and
DA19 in shake flask cultures, which indicates that
DA19 produces 7.54mg-L-' of secreted hEG7 in a
complex medium, compared with 3.98mg.L- producedby DHSa[13]. The presence of large amount of
proteins in complex medium increases the difficulty in
downstream separation and purification of target protein,
but study on foreign protein production under control
of the phoA promoter in defined media is rare. In
this study, hEGF was produced by the acetate-tolerant
strain DA19 in a chemically defined culture medium.
It was found that pH had a significant effect on phoA
promoter-controlled gene expression, and hEGF production
was improved remarkably at an alkaline pH
Initially, the cultivation was carried out in a batch
mode. When glucose initially added in the medium
was depleted, feeding of glucose was started to allow
the total cell mass in the reactor to increase exponentially.
To maintain a constant specific growth rate of
0.2h-', the feeding rate of limiting substrate, glucose,
is given by
F = XO"0 e x p w
yXISsF
where F is the volumetric feeding rate (L.h-'), SF is
the glucose concentration in the feed solution (g.L-'),
p is the specific growth rate (h-') to be controlled,
Y X / S is the cell yield coefficient based on consumed
glucose at the p to be controlled, XO and VO are the
biomass concentration (g.L-') and culture volume (L) at
the start of feeding. When the specific growth ratelwas
0.2h-', a specific glucose supply rate of 0.19g.g- .hwas
obtained according to the above equation. The
biomass concentration in the culture was measured
every hour, and the quantity of glucose to be added
until next sampling was calculated according to the
above formula but was added at a constant volumetric
rate to deliver the same amount of glucose.
The preliminary fed-batch culture of DA19
(pAET8) was carried out in the defined main culture
medium with pH controlled at 7.0 during the whole
fermentation process. The initial glucose of 10g.L-l
was exhausted at 8.25h indicated by a sudden increase
of DO, and glucose was added with an average specific
supply rate of 0.19g.g-'K1. At 14.85h, the p,hosphate
concentration decreased to 1.6mmol.L - , at
which point the specific growth rate of 0.2h-I could
not be maintained any more despite continued addition
of the glucose feed. Meanwhile, acetate started to
accumulate, which indicated that nutrients other than
glucose were limiting. In the expression phase, glucose
was added at the same specific supply rate, and
the residual glucose was gradually increased up to
6.1g.L-', while the NH4' concentration was very low.
The hEGF production level was 16.lmg.L-' (data not
shown) .
This preliminary experiment indicated that in the
expression phase the glucose feeding rate was too high.
Fig.1 shows the typical trends of biomass (DCM), glucose,
phosphate, ammonium, and extracellular hEGF
concentrations in fed-batch culture with reduced glucose
feeding rate in the expression phase. The process
could be divided into four phases. Phase I is the
batch culture period that lasted until exhaustion of the
initial glucose; Phase I1 is an exponential growth
period, during which glucose was added according to
the above-mentioned method until the phosphate concentration
decreased to about 1.6mmol~L-'; in Phase
I11 glucose was added at a constant rate of 6.6g-h-'
until phosphate in the medium was nearly consumed;
and Phase IV was the expression period in which
glucose was added at a constant specific rate of
0.16g.g-'.h-'. Because of the reduced feeding rate of
glucose, no residual glucose was detected during theexpression phase. However, the extracellular hEGF
concentration remained a very low level of 14.6mg.L-'.
3.2 The effect of pH on hEGF production
Because alkaline phosphatase functions better in
an alkaline environment and expression of hEGF was
controlled by the promoter of alkaline phosphatase,
hEGF production at higher pH was examined. The
fermentation operation was the same as the above experiment
except for pH in the expression phase being
controlled at 7.5, 7.8, and 8.0, respectively. The results
are summarized in Table 1. It can be seen that
with the increase in pH, the hEGF production was
improved and the highest level of 75.5mg.L-' was
reached at pH 7.8 (Fig.2), 5.2-fold of that achieved at
pH 7.0. The specific hEGF production (YEGF/xa)v, erage
specific hEGF production rate (QEGF)a, nd average
volumetric hEGF production rate (TEGF) also reachedtheir maximal values at pH 7.8. However, further increase
in the pH to 8.0 resulted in decreased hEGF,
YEGFIX, QEGF, and TEGF.
Figure 3 shows the profiles of extracellular hEGF
and acetate during the expression phase controlled at
different pH values. At pH 7.5, acetate accumulation
and hEGF production were similar to those at pH 7.0;
the higher acetate concentration at the end of expression
phase was caused by the longer expression phase.
At pH 7.8, the EGF concentration significantly increased
even though acetate accumulation was enhanced
at the same time. At llh post induction, acetate
reached 4g.L-', but hEGF still showed an increasing
trend. When pH was controlled at 8.0, acetate
accumulation further increased, and the hEGF concentration
decreased during the period between 4 and
8h post induction. During the period between 8 and
10h, the acetate concentration rose from 3.9g.L-' to
5.6g.L-' and hEGF stopped increasing, while the
hEGF concentration continued to increase within a
similar acetate level when pH was controlled at 7.8.
Therefore, the higher acetate concentration at pH 8.0
was not responsible for the observed lower hEGF
concentration.Table 1 shows that the ammonia feeding rate increased
with an increase in pH due to more acetate
production. Because the concentration of ammonium
was low in the expression phase (Fig.l), adding more
ammonia solution should be beneficial to the expression
of hEGF in the acetate-tolerant strain. When thepH was controlled at 8.0, the hEGF concentration was
decreased even though more ammonium was provided.
These facts suggested that pH 7.8 was most suitable
for the expression of hEGF.
Studies on phosphatases in Neurospora crassa
show that pH is an important factor for enzyme production
and secretion [18]. Northern analysis and enzyme
activity assay confirm that the enzyme synthesis
in Metarhizium anisopliae is considerably improved at
the pH they function effectively, irrespective of
whether the medium contains an inductive substrate[
191. pH regulation is found in bacteria including
Salmonella typhimurium, Vibrio cholera, E. coZi[20],
as well as in fungi and animal cells, indicating that pH
regulation of gene expression is a general phenomenon[
21]. The present study indicated that at pH 7.8
best expression of the gene controlled by the phoA
promoter was obtained. The hEGF production level in
the minimal medium was not high compared with
those obtained in complex media[ 113. However, when
a mixture of tryptone and yeast extract was supplied,
the overall hEGF level reached 765mq-Lp' but the
secreted hEGF was only 78.7mg.L- , suggesting
limitation of secretion [22].
80
70
60
50 2 40 rE;
30
20
10
c
- 0
3.3 Tolerance of DA19 to acetate
One of the major drawbacks of E. coZi is formation
of acetate that seriously inhibits the expression of
foreign genes. In this study, an acetate-tolerant strain,
DA19, was used. In order to examine the effect of
acetate accumulation on hEGF production, glucose
was added at a higher specific rate of 0.247gSg-'.h-'
in the expression phase to produce more acetate with
the pH controlled at 7.8. 13.7g.L-' of acetic acid were
formed, but the hEGF concentration profile was similar
to that when glucose was added at a lower specific
rate of 0.159g.g-'.h;', where the highest acetate production
was 5.6g.L- (Fig.4 and Table 1). This strongly
indicated the advantage of this acetate-tolerant strain in
expression of a recombinant protein.3.4 Factor limiting cell growth in phase III
In complex media containing yeast extract and
tryptone, cells of DA19 (pAET8) could grow at a high
specific rate until phosphate was almost depleted (datanot shown), but in the defined medium, the phosphate
consumption rate decreased when the phosphate concentration
reduced to 1.6mmol.L-'. Examination of
the fermentations in the defined medium suggested
that the reduced phosphate consumption could be attributed
to limitation of ammonium that was below
5mmol.L-' (Fig.1). To understand the effect of nitrogen
source, NH4+ was added to reach 40rnm0l.L-~'
when the concentration had decreased to 1Ommol.L- .
As shown in Fig.5, supplementation of ammonium
facilitated phosphate uptake. Phosphate was depleted
within two hours after addition of ammonium, compared
with six hours without ammonium supplementation.
However, the secreted hEGF was only slightly
improved (78.7mg.L-').The pH in the expression phase remarkably affected
hEGF production controlled by the alkaline
phosphatase (phoA) promoter. At pH 7.8, the extracellular
hEGF increased by 5.2-fold compared with
that at pH 7.0, even though more acetate was formed.
During the expression phase, when DA19 (pAET8)
was fed with glucose at a higher specific rate to result
in an acetate concentration of 13.7g.L-', the hEGF
production was similar to that obtained at an acetate
concentration of 5.6g.L-', suggesting the advantage to
use the acetate-tolerant strain, DA19.
