Production of tilapia insulin-like growth factor-2 in high cell
density cultures of recombinant Escherichia coli
Shao-Yang Hu
a, Jen-Leih Wu
b, Jan-Hsiung Huang
a,∗aLaboratory of Fermentation, Graduate Institute of Agricultural Chemistry, National Taiwan University, Taipei 106, Taiwan, ROC bLaboratory of Marine Biology and Biotechnology, Institute of Zoology, Academia Sinica, Nankang, Taipei 115, Taiwan, ROC
Received 16 July 2003; received in revised form 30 August 2003; accepted 12 September 2003
Abstract
An improved expression plasmid pET-insulin-like growth factor-2 (IGF2) was constructed and transferred into Escherichia coli BL21(DE3) for the expression of tilapia insulin-like growth factor-2. The recombinant insulin-like growth factor-2 was produced as inclusion bodies, and the recombinant insulin-like growth factor-2 content was as high as 10.3% of the total protein content. For production of recombinant insulin-like growth factor-2 in E. coli, pH-stat fed-batch cultures were used to achieve a high cell density culture. A cell concentration 183 g l−1dry cell weight (DCW) was obtained after 30 h cultivation and plasmid stability was maintained at high levels. Expression of insulin-like growth factor-2 was induced at three different cell concentrations, 50, 78.5, and 114.5 g l−1dry cell weight. When cells were induced at a cell concentration of 114.5 g l−1dry cell weight, the amount of insulin-like growth factor-2 produced was 9.69 g l−1(11.3% of the total protein). Using a simple purification process including inclusion body isolation, denaturation, refolding and Ni-NTA affinity chromatography, 19.51 mg of insulin-like growth factor-2 was obtained from a 22.5 ml of culture, and the recovery yield was 20.5%. The biological activity of the purified IGF-2 was demonstrated as promoting the growth of four different cell lines by the colorimetric bioassay and the best growth stimulation ratio was obtained for the Balb/3T3 clone 31A cell line.
© 2003 Elsevier B.V. All rights reserved.
Keywords: Insulin-like growth factor-2; High-cell density; Fed-batch culture; Escherichia coli; Growth stimulation
1. Introduction
Insulin-like growth factor (IGF) is a single-chain polypeptide hormone with three disulfide bonds, which exists in variety of mammals and vertebrates including humans, cattle, chickens, and fish. During the past 20 years, IGF proteins have been
character-∗Corresponding author. Tel.:+886-2-23630231x2481;
fax:+886-2-23630573.
E-mail address: [email protected] (J.-H. Huang).
ized from species as diverse as humans, chickens, rats and sharks, with the majority of the information being obtained during the last decade through cDNA cloning (Jansen et al., 1983; Kajimoto and Rotwein, 1991; Shimatsu and Rotwein, 1987). Depending on the structure and function of the IGF, two types of
IGFs, IGF-1 and IGF-2 has been classified (Baker
et al., 1993; Cohick and Clemmons, 1993). In fish, in addition to a growth promotion function, IGF-1 can modulate calcium concentrations to balance changes in osmotic pressure (McCormick et al., 1991; 0168-1656/$ – see front matter © 2003 Elsevier B.V. All rights reserved.
Moriyama et al., 1993, 2000). However, while much is known about the function of insulin-like growth factor-1, less is known about IGF-2 as to whether it exhibits any roles distinct from those in IGF-1. Recently, tilapia IGF-2 cDNA was isolated from the liver and the protein sequence was expressed by GST fusion in Escherichia coli (Chen et al., 1997). Mature tilapia insulin-like growth factor-2 which consists of the B, C, A, and D domains, is a small single-chain peptide of 70 amino acids. It can stimulate somatic growth and cell differentiation in vertebrates. Be-cause of these biological activities, tilapia IGF-2 has attracted attention in the field of aquaculture and it is currently being developed as a fodder additive for growth promoter. Furthermore, for a detailed under-standing of the function and role of IGF-2 in modu-lating physiology, a source of IGF-2 is necessary.
Although IGFs have been expressed in a variety of host systems, including yeast (Bayne et al., 1988; Gellerfors et al., 1989), bacteria, and mammalian cells (Ballotti et al., 1987), E. coli is still the most widely employed host for the production of IGFs because of ease of handling and high yield. In a
number of publications (Kim and Lee, 1996; Yamada
et al., 1996), IGF has been expressed as a fusion protein for availability of antibodies and ease of de-tection in E. coli. However, it is not an economical strategy from the production viewpoint because the separation steps required to release IGF protein have a high cost and this results in a low efficiency of recovery. For this reason, in this study, a new ex-pression plasmid for tilapia IGF-2 was constructed by improving the plasmid pGEX2T-IGF2, which was designed to express IGF-2 as a fusion protein with glutathione-S-transferase (GST) as previously reported (Chen et al., 1997). In this new expression plasmid, IGF-2 has six histidines at the C-terminal for easy purification and consequently the yields are sig-nificantly improved. In our present investigation, we were concerned with high-level production of tilapia IGF-2 in fed-batch cultures of E. coli BL21(DE3). An effective fed-batch culture procedure was set up that achieved a high cell concentration of 183 g l−1 dry cell weight (DCW). In addition, the process of purification was also developed to obtain biologically active IGF-2. We expected that the results will be useful as a production model for other recombinant proteins expressed as inclusion bodies.
2. Material and methods
2.1. Bacterial strains and plasmids
Tilapia IGF-2 cDNA was cloned and the plasmid pGEX2T-IGF2 was constructed for expression in
E. coli as described previously (Chen et al., 1997). For expression of the mature igf-2 gene, pET-IGF2 was constructed by cloning the igf-2 gene into
pET28b as follows. The forward primer 5
-CCGA-CGCATCCATGGCGGAA ATGG was designed
to contain an NcoI restriction site (underlined
nu-cleotides). The reverse primer 5
-GACGACGCC-CTCGAGTTCGGACTT was designed to contain an XhoI restriction site (underlined nucleotides). The complete coding region for the mature tilapia IGF-2 was amplified by PCR (GeneAmp PCR Sys-tem 2400, The Perkin-Elmer Corporation, CA, USA) using the plasmid pGEX2T-IGF2 as the template. The PCR product was digested with NcoI and XhoI before it was cloned into the same enzyme site in the pET28b vector. The igf-2 gene was expressed under the strong T7 promoter by induction with
isopropyl--d-thiogalactopyranoside (IPTG) (Sigma,
St. Louis, MO, USA). E. coli DH5␣ was used as a
host strain for cloning and maintenance of the plas-mids. E. coli BL21(DE3) was used as a host for expression of IGF-2.
2.2. Growth medium and cultivation condition 2.2.1. Seed cultivation and shake flask cultivation
E. coli BL21(DE3) cells which contained plasmid
pET-IGF2 were stored in glycerol at−30◦C. Frozen
glycerol cell stock (500l) was inoculated into a
250 ml flask containing 50 ml Luria–Bertani (LB) medium (tryptone 10 g l−1, yeast extract 5 g l−1, NaCl 10 g l−1) supplemented with 50g ml−1of kanamycin
in a shaking incubator at 37◦C at 125 rpm. After
12 h cultivation, 1 ml of culture was transferred to a 500 ml Hinton flask containing 100 ml LB medium
supplemented with 50g ml−1of kanamycin and
in-cubated for 12 h under the same conditions as the seed culture. Shake flask experiments were carried out in SSP medium (yeast extract 10 g l−1, glucose 5 g l−1, N-Z-Amine AS 15 g l−1, K2HPO4 8 g l−1, KH2PO4
2 g l−1; pH 7.5). A 500 ml Hinton flask containing 100 ml SSP medium was inoculated with 2 ml of seed culture and incubated at 37◦C at 125 rpm. Following
a 3 h cultivation (OD6001), IPTG was added to a final
concentration of 0.06 mM for expression of IGF-2. Af-ter 12 h cultivation, the cells were harvested for 16.5% Tricine SDS-PAGE and estimation of expression level.
2.2.2. Fed-batch cultivation
Fed-batch cultivation was carried out at 37◦C in a 5 l bioreactor (Biostat B., B. Braun Biotech In-ternational GmbH, Melsungen, Germany) equipped with a built-in digital controller for pH, temper-ature, agitation, dissolved oxygen and peristaltic pumps for acid, base, antifoam and nutrient. The pH (InPro 3030/325 combination pH electrode, Mettler-Toledo GmbH, Switzerland) and dissolved oxygen concentration (InPro 6000 series O2 sensors,
Mettler-Toledo GmbH, Switzerland) were controlled at the set points by on-line monitoring. The con-troller was interfaced with a computer through the Multi-Fermentor Control System software (Biostat B., B. Braun Biotech International GmbH, Melsun-gen, Germany). The initial work volume was 1.30 l
of modified R-medium (Lee and Chang, 1993).
The medium contained citric acid 3 g l−1, KH2PO4
6.75 g l−1, (NH4)2SO4 5 g l−1, Na2HPO4·12H2O
3 g l−1, MgCl2 1.5 g l−1, NH4Cl 0.1 g l−1, glucose
20 g l−1, yeast extract 20 g l−1, N-Z-Amine AS
30 g l−1 and 6 ml l−1 of a trace metal solution. The composition of trace metal solution that contains (per
ml of 5 M HCl) FeSO4·7H2O 10 mg, ZnSO4·7H2O
2.25 mg, CaCl2·2H2O 1.35 mg, MnSO4·5H2O
0.5 mg, CuSO4·5H2O 1 mg, AlCl·6H2O 0.3 mg,
(NH4)6Mo7O24·4H2O 0.1 mg, H3BO3 0.2 mg, and
thiamine–HCl 2 mg. Glucose was sterilized separately and trace metal solution was sterilized by filtration. The entire content of seed culture (7.69% v/v) was in-oculated. The agitation speed and flow rate of aeration was set at 1000 rpm and 3 l min−1. The pH was kept at 7.0 by adding 28% (v/v) ammonium water. The dis-solved oxygen concentration was maintained above 20% of air saturation by increasing the percentage of pure oxygen using a gas mixer (Gas Mix Unit, B. Braun Biotech International GmbH, Melsungen, Ger-many). A nutrient feeding solution was added by us-ing the pH-stat feedus-ing strategy. The nutrient solution contained glucose 750 g l−1, yeast extract 50 g l−1, N-Z-Amine AS 75 g l−1, MgSO4·7H2O 20 g l−1, citric
acid 3 g l−1, KH2PO4 6.75 g l−1, (NH4)2SO4 5 g l−1,
Na2HPO4·12H2O 3 g l−1, MgCl2 5 g l−1, NH4Cl
0.1 g l−1. When the pH rose to a value greater than the set point (pH 7.0) by 0.1 U due to the depletion of glucose, the nutrient feeding solution was automati-cally added to increase the glucose concentration in the culture broth. Expression of the igf-2 gene was induced by adding IPTG to a final concentration of 0.06 mM. Foam was suppressed as necessary by the addition of sterilized antifoam (Sigma A-5758).
2.3. Purification of IGF-2 from the inclusion body
Cells were harvested from 22.5 ml of culture by cen-trifugation at 6000× g for 15 min at 4◦C and washed twice in PBS buffer (140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4·12H2O, 1.8 mM KH2PO4; pH 7.3),
then centrifuged again. The washed cell pellets were resuspended thoroughly in PBS buffer and disrupted by sonication using a sonicator (Model XL 2020, 419A microtipprobe, Misonix Inc., NY, USA) at a setting of 3. The sonicator was programmed to provide 10 s pulses with 5 s pause for a total period of 30 min. The disrupting pellets were washed with 50 ml PBS buffer containing 3 M urea to remove cell debris contami-nants and were further centrifuged at 17,500× g for 30 min to isolate the inclusion bodies. The inclusion bodies were completely soluble in 50 ml PBS buffer (pH 12) containing 5 M urea at room temperature and the solubilized solution was clarified by centrifugation at 17,500× g for 30 min at 4◦C. Refolding was initi-ated by diluting the solubilized protein in PBS buffer and the solution was then incubated at 4◦C for 36 h.
The refolded protein was centrifuged at 17,500× g
for 30 min to remove any insoluble material. The su-pernatant fraction was collated and concentration by centrifugation using a type YM3 ultrafiltration mem-brane (molecular weight cutoff, 3000; Millipore, MA, USA) and then loaded to bind with the Ni-NTA gel (Qiagen). The gel was washed with 30 mM imidazole and eluted with 250 mM imidazole. The purified IGF-2 protein solution was dialyzed against PBS buffer and stored at 4◦C prepared for bioactivity assay.
2.4. IGF-2 bioactivity assay
The cell lines, growth medium and culture con-dition in this assay are shown in Table 1. Four dif-ferent cell lines, zebrafish liver cell (ZFL), tilapia
Table 1
Cell lines, growth medium and culture condition used in IGF-2 bioactivity assay
Cell line Species and type Growth mediuma,b Culture condition Reference and source ZFL Zebrafish liver cell L15+ fetal bovine serum 28◦C Chen et al. (2001) TO-2 Tilapia ovary cell L15+ fetal bovine serum 28◦C Chen et al. (1997) Balb/3T3 Clone 31A Mouse embryol cell DMEM+ 10% calf serum 37◦C, 5% CO2 CCRC 60009c
MRC-5 Human lung cell EM+ 10% fetal bovine serum 37◦C, 5% CO2 CCRC 60023c aGrowth medium used contained penicillin G sodium 100 U l−1, streptomycin sulfate 100 U l−1 and amphotercin B 25g l−1. b The medium and reagent were purchased from Gibco BRL.
cCCRC: Culture Collection and Research Center, Hsinchu, Taiwan.
ovary cell (TO-2), mouse embryo cell (Balb/3T3 clone 31A) and human lung cell (MRC-5), were used as the models for the bioactivity assay. The biological activity of IGF-2 was assayed by evalu-ating the ability to promote cell proliferation using
the colorimetric bioassay method (Okajima et al.,
1992). Cells were seeded at a density of 3× 104
cells per well in 96-well microplates and cultured in growth medium under their optimal individual culture conditions. After 24 h incubation, the growth medium
was discarded, and replaced with 100l of growth
medium without serum and contained various concen-trations of purified IGF-2. After 48 h cultivation, the
cells were stained by adding 11l of
[3-(4,5-dime-thyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromi-de] (MTT) (Sigma M-5655) reagent to each well (final concentration was 0.5 mg ml−1) and incubated
for 4 h. Then 100l of dimethyl sulfoxide (DMSO)
(Sigma D-2650) was added to lyse the cells and the plates were then incubated for 1 h to extract the MTT-formazan product. Absorbances were read
at 570 and 650 nm (A570−650 nm) by a microplate
reader (Anthos 2020, Wals, Austria). Growth stim-ulation ratios were calculated using the following equation:
growth stimulation ratio(GSR) (% control) =Asample
Acontrol × 100
where Acontrolwas the avarage absorbance of the wells
which contained cells without added IGF-2 (control),
Asamplewas the average absorbance of the wells which
contained the cells stimulated with purified IGF-2 (sample).
2.5. Analytical methods
2.5.1. Cell and protein concentration
Approximately 7 ml of culture sample was ex-amined periodically. The cell concentration was measured using the optical density (OD) on a spec-trophotometer (Hitachi U-2000, Tokyo, Japan) at 600 nm 0.3–0.6, either undiluted or after dilution. Dry cell weight was determined by centrifugation of 2 ml culture in a pre-weighted microcentrifuge tube, followed by washing twice with distilled water and drying to a constant weight at 80◦C in an oven. The protein concentration was determined by the Bradford protein assay method using bovine serum albumin as the standard (Bradford, 1976).
2.5.2. Plasmid stability estimation
A serial dilution of the culture broth sample was spread onto LB agar plates without antibiotic and
incubated at 37◦C for 12 h. One hundred colonies
were picked up and transferred to LB agar plate with kanamycin. Plasmid stability was determined by calculating the percentage of colonies resistant to kanamycin.
2.5.3. Glucose and acetate assay
The glucose concentration was determined at 570 nm using a microplate reader by the DNS method (Miller, 1959). Culture samples were prepared for acetic acid analysis by microcentrifuge at 12,000 rpm
for 5 min. A 500l sample of the supernatant
cul-ture broth was acidified by adding 10l of 12 N HCl
and centrifuged again. A 200l sample of the
su-pernatant was mixed with an equal volume of 0.2% butyric acid solution that was used as the internal standard before acetic acid analysis. The quantity of
the acetic acid was analyzed using gas chromatog-raphy (Hitachi G-3000, San Jose, CA, USA) with
cp-wax 52 CB column (0.53 mm × 30 m, film
thick-ness 1m, Corompak, Varian, Palo Alto, CA, USA).
The conditions were as follows: carrier gas and flow rate, H2 and 3 ml min−1; inject temperature, 240◦C;
detector and detector temperature, FID and 260◦C;
oven program, 120◦C held for 5 min and than raised
the temperature 10◦C min−1 until the temperature
reached 200◦C and then held for another 5 min.
2.5.4. IGF-2 purity analysis
The IGF-2 content, separated by electrophore-sis, was quantified using a densitometer (ID Im-age Analysis Software, Kodak Digital Science, NY, USA). The purity of purified IGF-2 was analyzed by HPLC equipped with a betabasic-8 column (4.6 mm × 150 mm, 5 m, Keystone Scientific Inc., Bellefonte, PA, USA). Elution was performed using 25% acetonitrile containing 0.1% trifluoroacetic acid at a flow rate of 0.5 ml min−1.
2.5.5. Determination of N-terminal amino acid sequence and mass spectrum
Protein samples were analyzed using 16.5%
Tricine SDS-PAGE gel (Schagger and Jagow, 1987)
and stained with Coomassie brilliant blue R-250 (Bio-Rad). Electroblotting of purified IGF-2 to polyvinylidene difluoride (PVDF) immobilon mem-brane (Millpore, Bedford, MA, USA) was performed
according to the method of Matsudaira (1987). The
N-terminal sequence of IGF-2 was determined by Edman degradation using a model 476A sequencer (Applied Biosystems). The molecular mass of pu-rified IGF-2 was determined by mass spectrometry (Quattro Ultima, Micromass, UT, USA).
3. Results
3.1. Expression and characterization of the tilapia IGF-2 gene in E. coli
The expression plasmid pET-IGF2 encoding the tilapia IGF-2 are derived from pET28b and used to transform E. coli BL21(DE3) for expression (Fig. 1). The amino acid, Alanine, which generated from re-striction site, exists between the start codon ATG (Met)
Fig. 1. Construction of the recombinant expression plasmid pET-IGF2. (A) The PCR fragment (length, 224 bp) which encodes mature tilapia IGF-2 was digested with NcoI and XhoI and then cloned into pET28b at the same restriction sites. Kanr, kanamycin
resistance; lacI, lac repressor; ori, the origin gene of replication. (B) Diagram illustrating the main features of pET-IGF2. Expres-sion of the tilapia igf-2 gene is controlled by T7 promoter (PT7). RBS, ribosome binding site.
and the mature IGF-2. The C-terminal of mature IGF-2 was tagged with six histidines for purification using Ni-NTA affinity chromatography. Cells were grown in SSP medium and induced with 0.06 mM IPTG at
an OD600 of 1. More than 90% of the recombinant
IGF-2 was expressed as inclusion bodies even when
the culture temperature was down shifted to 27◦C.
The IGF-2 content level was about 10.3% of the total protein (Fig. 2).
3.2. High cell density culture of recombinant E. coli
Fed-batch cultivations using a pH-stat strategy were carried out in order to achieve high cell densi-ties of the recombinant E. coli strain. When the initial glucose (20 g l−1) in the bioreactor was exhausted at a cultivation time of 6 h, the addition of feeding solution was started due to the pH rising over the set value of 7.1. With glucose being the limiting param-eter during the batch phase, a specific growth rate () of 0.65 h−1and a yield coefficient for glucose of 0.453 g g−1 was attained. The glucose concentration was kept under 1.5 g l−1during the fed-batch culture
Fig. 2. Expression of tilapia IGF-2 mature peptide in E. coli BL21(DE3) by a flask culture. (A) The proteins were analyzed by 16.5% Tricine SDS-PAGE. Lane: 1, molecular mass standard; 2, the total protein without IPTG induction at 37◦C; 3, the total protein with IPTG induction at 37◦C; 4, soluble protein with IPTG induction at 37◦C; 5, insoluble protein with IPTG induction at 37◦C; 6, soluble protein with IPTG induction at 27◦C; 7, insoluble protein with IPTG induction at 27◦C.
to avoid accumulation of acetic acid. Hence, during the process of cultivation, acetate concentration was maintained below 1 g l−1. In the fed-batch phase, the specific growth rate () descends to 0.057 h−1 due to nutrient limitation by glucose (<1.5 g l−1) and the yield coefficient for glucose was maintained about
0.427 g g−1. Pure oxygen was supplied to
accommo-date the increasing rate of oxygen consumption ac-companying cell growth. By this mean, the dissolved
Fig. 3. High cell density fed-batch cultivation of recombinant Escherichia coli BL21(DE3). Log OD600 (䊉), dry cell weight (䉱), plasmid
stability (䊊), glucose ( ), acetic acid (䊏). Fed-batch culture started at 6 h after unlimited growth in the batch phase.
oxygen concentration could be maintained above 20% of air saturation during the culture. By keeping excel-lent growth conditions for cell growth in the bioreac-tor, a high density of biomass of 183 g l−1DCW was achieved after 30 h cultivation and 91% of the cells still contained the expression plasmid (Fig. 3).
3.3. Expression of tilapia IGF-2 at varying cell densities
In order to evaluate the efficiency of expression at the time of induction, cells were induced with 0.06 mM IPTG at three different cell concentrations
(50, 78.5 and 114.5 g l−1 DCW). Fig. 4 shows the
time profile for cell density (OD600), dry cell weight,
and IGF-2 content expressed as a percentage of the total protein content. When cells were induced at the low cell concentration (50 g l−1 DCW), the fraction of IGF-2 in the total protein increased for 15 h af-ter induction and then decreased slightly (Fig. 4A). The dry cell weight and the maximal IGF-2 content were 94.5 g l−1 and 15.27% of total protein, respec-tively. When cells were induced at the intermediate cell density (78.5 g l−1DCW), the fraction of IGF-2 increased for 18 h after induction and than decreased. The cell dry weight and the maximal IGF-2 content
were 125.78 g l−1 and 12.11% of total protein,
re-spectively (Fig. 4B). When cells were induced at the high cell concentration (114.5 g l−1DCW), the IGF-2
Fig. 4. Time profiles for cell density (OD600) (䊉), cell dry weight
(䊊), and IGF-2 content (䉱) during fed-batch cultivation with induction at the low cell concentration (A), the intermediate cell concentration (B), and the high cell concentration (C). The arrow indicated the time of induction.
Table 2
Protein recovery in a representative experiment: extraction from inclusion bodiesa Purification step Volume (ml) Concentration
of protein (mg ml−1)b Amount of total protein (mg) Amount of IGF-2 (mg) IGF-2 recovery (%) Purity (%)
Solubilized inclusion body 50 5.88 294 94.962 100 32.3c
Dilution and refolding 1000 0.183 183.2 65.05 68.50 35.51c
Ni-NAT affinity chromatography 40 0.542 21.68 19.512 20.54 90d
a A 22.5 ml fed-batch culture was induced at the high cell concentration of 114.5 g l−1DCW. b Concentration of total protein was determined with the Bradford method.
c Determined by densitometric scanning of Tricine SDS-PAGE gels. d Determined by HPLC.
content increased continually for 18 h. The maximal dry cell weight and maximal IGF-2 content were 156 g l−1and 11.3% of the total protein, respectively (Fig. 4C). In the main, no large variation happened in the IGF-2 content when IPTG was induced at different cell concentration. Therefore, the largest amount of IGF-2 was produced when cells were induced at the high cell concentration (114.5 g l−1 DCW), and the concentration of IGF-2 was as high as 9.69 g l−1.
3.4. Purification of recombinant IGF-2
Recombinant IGF-2 was purified from 22.5 ml of culture from a fed-batch culture induced at the high cell concentration as described above. The results are summarized inTable 2. The final amount of the IGF-2 and recovery yield was 19.512 mg and 20.54%, respec-tively. The purity of IGF-2 as determined by 16.5% Tricine SDS-PAGE to be greater than 91% and the purity of the IGF-2 as determined by HPLC was 90% (Fig. 5). This results of purification indicated that
869.67 mg l−1 of purified IGF-2 was obtained from
the fed-batch culture. A mass spectrum analysis of the purify IGF-2 was 9072.8 Da, which is consistent with the molecular mass of the tilapia IGF-2 plus the amino acid generated by the restriction site and plus the six of histidines.
3.5. Characterization and bioactivity of IGF-2
For characterization of the purified IGF-2 pro-tein, N-terminal amino acid sequencing was ana-lyzed. The amino acid sequence following Ala was Glu–Met–Ala–Ser–Ala–Glu, which was identical to that of the native IGF-2. The biological activity of the
Fig. 5. Samples from each purification step analyzed by 16.5% Tricine SDS-PAGE gel. (A) Lane 1, molecular mass standards; 2, inclusion body fraction after washing with 3M urea; 3, supernatant after 5 M urea and pH 12 denaturation and centrifugation; 4, supernatant after refolding and centrifugation; 5 and 6, sample after Ni-NTA affinity chromatography and dialysis. (B) Reverse phase HPLC analysis of purified IGF-2. Arrow indicated the peak of purified IGF-2.
purified IGF-2 was further confirmed. MTT method
as described in Section 2was used to determine the
effect of purified IGF-2 with different concentration on promoting proliferation of the cell lines. Compared with the growth of the cells without the addition of IGF-2, an effective growth-promoting action in the presence of different concentrations of IGF-2 from 5 to 120 nM was observed. Furthermore, the IGF-2 gave different growth stimulation responses on different cell lines (Fig. 6). The growth stimulation ratios on the four different cell lines at IGF-2 concentration from 30 to 120 nM were significantly different (P < 0.05). The greatest increase in growth-promotion was with the Balb/3T3 clone 31A cell line and a growth
stimu-Fig. 6. The bioassay of recombinant tilapia IGF-2. The bioactiv-ity of IGF-2 was measured with MTT method as described in Section 2. The growth stimulation ratios (GSR) were evaluated after 48 h cultivation and the results were compared with cells in the absence of IGF-2 treatment. Growth stimulation ratios are shown as means± S.D. (n = 6). The growth stimulation ratios on the four different cell lines at IGF-2 concentration from 30 to 120 nM were significantly different (P < 0.05).
lation ratio of 187% was obtained in the presence of 120 nM of IGF-2.
4. Discussion
IGF-2 has previously been purified from E. coli using the GST fusion system; however, low recovery yields were obtained during the separation and pu-rification process for this fusion protein. In addition, the separation of the fusion into its two components by thrombin increased the cost of the production sig-nificantly. Because of these problems, a new expres-sion plasmid for IGF-2 production was constructed. The igf-2 gene on pET-IGF2 encodes a recombinant tilapia IGF-2 tagged with an extension of six his-tidyl residues (6× His) at the carboxyl terminus. This tag can be used for purification by Ni-NTA affinity chromatography. In order to improve productivity and obtain a large amount of IGF-2, a high cell den-sity culture technique with pH-stat strategy was used in this study. The nutrient feeding strategy is based on the finding that pH increases when the carbon source is depleted. Therefore, glucose was controlled at low concentration (below 1.5 g l−1) to avoid the
accumulation of acetic acid during the fed-batch stage. It is well-known that acetate is overproduced when
E. coli is grown under anaerobic or oxygen-limiting
conditions (Luli and Strohl, 1990). Thus, pure oxygen was supplied to maintain the dissolved oxygen above 20% of air saturation and the acetic acid concentration could be maintained effectively below 1 g l−1during the process of cultivation. This acetate concentration was lower than the 5 g l−1that causes a lower growth rate, a lower biomass yield, and a lower maximum attainable cell density in high cell density culture (Lee, 1996). Therefore, a high density of biomass 183 g l−1DCW was achieved after 30 h cultivation by maintaining these optimal growth conditions.
In this study, the final biomass concentration
(183 g l−1 DCW) approaches to the maximum
theo-retical cell mass concentration (160–200 g l−1DCW) as described in Märkl et al. (1993). Only two other publications describe high cell densities close to
this. These are 190 g l−1 DCW for E. coli K-12
in dialysis bioreactor (Nakano et al., 1997), and
204.3 g l−1 DCW for a recombinant E. coli
produc-ing poly(3-hydroxybutyrate) (PHB) (Wang and Lee,
1997). When these two cases are compared with this
study, the dialysis system was an expensive approach and not easy to scale up to a factory process. The product of the other case, PHB, is a high-molecular weight polymer that cause no osmotic stress on the cell and does not adversely affect overall bacterial growth. Because most of the research in case of PHB was aimed at maximizing the PHB content, the real cell mass (real cell mass= total biomass − PHB) was generally below 30–40 g l−1 DCW (Srinivasan et al., 2002). In our study, a 183 g l−1DCW of cell concen-tration was almost the maximum cell concenconcen-tration that could be obtained without any changes to the bioreactor system.
The efficiency of IGF-2 expression was estimated by IPTG induction at different cell concentrations; however, this resulted in little change to the IGF-2 content. Furthermore, cell growth was manifested as a deceased cell growth rate regardless of when the induction was performed. This phenomenon resulted from the expression of recombinant IGF-2, which caused a metabolic burden on growth as mentioned byGlick (1995). Therefore, when IPTG was induced at early stage of growth, cell growth was restricted strictly by metabolic stress and a lower biomass was
obtained. For this reason, the largest amount of IGF-2 was obtained at higher cell concentration through induction during the late growth phase and this im-proved the productivity of IGF-2. When cells were
induced at a high cell density of 114.5 g l−1 DCW,
the maximum cell mass and amount of IGF-2
pro-duced was 156 g l−1 DCW and 9.69 g l−1 (11.3% of
total protein), respectively, and the specific IGF-2 productivity was highest (2.07 mg g−1DCW h−1).
Due to the complicated structure of IGFs (three disulfide binds), the recombinant IGFs is usually expressed as inclusion bodies in E. coli. However, pro-duction of a recombinant protein from inclusion bodies has several advantages, such as resistance to prote-olytic degradation and simple primary recovery from the total protein. Therefore, biological active IGFs from different species have been produced from inclu-sion bodies in a number of studies (Fine et al., 1997; Upton et al., 1995; Wangsa-Wirawan et al., 1997). In our study, 19.512 mg of biologically active IGF-2 that was more than 90% pure was obtained from a 22.5 ml culture after a series of simple recovery steps (recov-ery yield, 20.54%). In the case of fed-batch cultivation with induction at the high cell concentration, a total of 33.9 g of IGF-2 was produced from 3.5 l of culture broth (final volume of the culture), and 3.04 g of bio-logically active IGF-2 was obtained after purification. The recombinant IGF-2 showed efficient growth-promoting activity in a variety of cell lines, including ZFL, TO-2, Balb/3T3 clone31A and MRC-5 and this indicates that the process of purification is effective. According toChen et al. (1997), recombinant tilapia IGF-2 is biologically active on the TO-2 cell line at
concentrations between 15 and 120 nM using a [3H]
thymidine incorporation assay. In our study, the pro-duced IGF-2 was biologically active and functioned in this range of concentrations, and it gave different growth stimulation ratios on the different cell lines. Presumably the membranes of the four cell lines have different IGF receptors with different binding affini-ties for IGF-2. Because the amino acid sequence of the tilapia and mammalian IGF-2 mature peptide is quite conserved (above 80% similarity), the recombi-nant tilapia IGF-2 is also functional for the Balb/3T3 clone31A and MRC-5 cell lines. As mention by
DeChiara et al. (1990), IGF-2 plays an important role in the development of the embryo as a regulatory fac-tor in embryo growth; therefore it is also called a fetal
growth factor. This explains why the highest growth stimulation ratio was for the Balb/3T3 clone31A cell line. Given the above results, we suggest that it should be possible to develop the recombinant tilapia IGF-2 as growth promoters in livestock industry or as a medical reagent.
In this report, we have described an improved ex-pression plasmid for IGF-2, an efficient production system for large quantities of recombinant tilapia IGF-2 by fed-batch cultivation and a simple pro-cess for the purification and refolding of IGF-2 into a biologically active form. The strategies described may be useful for the efficient production of other recombinant proteins expressed as inclusion bodies in E. coli.
Acknowledgements
This work was supported by grants from the Na-tional Science Council NSC89-2317-B-002–006,
NSC90–2317-B-002-002, NSC91–2317-B-002-009
and Mithra Bioindustry Co. Ltd.
References
Baker, J., Liu, J.P., Robertson, E.J., Efstratiadis, A., 1993. Role of insulin-like growth factors in embryonic and postnatal growth. Cell 75, 73–82.
Ballotti, R., Nielsen, F.C., Pringle, N., Kowalski, A., Richardson, W.D., Van Obbereghhen, E., Gammeltoft, S., 1987. Insulin-like growth factor-1 in cultured rat astrocytes: expression of the gene, and receptor tyrosine kinase. EMBO J. 6, 3633–3639. Bayne, M.L., Applebaum, J., Chicci, G.C., Hayes, N.S., Green,
B.G., Cascieri, A., 1988. Expression, purification and char-acterization of recombinant human insulin-like growth factor-1 in yeast. Gene 66, 235–244.
Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. Chen, J.Y., Chang, C.Y., Chen, J.C., Shen, S.C., Wu, J.L., 1997. Production of biologically active recombinant tilapia insulin-like growth factor-2 polypeptides in Escherichia coli cells and characterization of the genomic structure of the coding region. DNA Cell Biol. 16, 883–892.
Chen, H.C., Lin, G.H., Gong, H.Y., Weng, C.F., Chang, C.Y., Wu, J.L., 2001. The characterization of prepro-insulin-like growth factor-1 Ea-2 expression and insulin-like growth factor-1 gene (devoid 81 bp) in the zebrafish (Danio rerio). Gene 268, 67–75. Cohick, W.S., Clemmons, P.R., 1993. The insulin-like growth
factors. Annu. Rev. Physiol. 55, 131–153.
DeChiara, T.M., Efstratiadis, A., Robertson, E.J., 1990. A growth deficiency phenotype in heterozygous mice carrying an insulin-like growth factor-2 gene disrupted by targeting. Nature 345, 78–80.
Fine, M., Amuly, R., Sandowski, Y., Marchant, T.A., Chan, S.J., Gertler, A., Funkenstein, B., 1997. Recombinant gilthead seabream (Sparus aurata) insulin-like growth factor-1: subcloning, expression in Escherichia coli, purification and characterization. J. Endocrinol. 153, 139–150.
Gellerfors, P., Axelsson, K., Helander, A., Johansson, S., Kenne, L., Lindqvist, S., Pavlu, B., Skottner, A., Fryklund, L., 1989. Isolation and characterization of a glycosylated form of human insulin-like growth factor-1 produced in Saccharomyces cerevisiae. J. Biol. Chem. 264, 235–244.
Glick, B.R., 1995. Metabolic load and heterologous gene expression. Biotechnol. Adv. 13, 247–261.
Jansen, M., Van Schaik, F.M.A., Ricker, A.T., Bullock, B., Woods, D.E., Gabbay, K.H., Nussbaum, A.L., Sussenbach, J.S., Van Den Brande, J.L., 1983. Sequence of cDNA encoding human insulin-like growth factor-1 precursor. Nature 306, 609–611. Kajimoto, Y., Rotwein, P., 1991. Structure of the chicken
insulin-like growth factor-1 gene reveals conserved promoter. J. Biol. Chem. 266, 9724–9731.
Kim, S.O., Lee, Y.I., 1996. High-level expression and simple purification of recombinant human insulin-like growth factor-1. J. Biotechnol. 18, 97–105.
Lee, S.Y., 1996. High cell-density culture of Escherichia coli. Trends Biotechnol. 14, 98–105.
Lee, S.Y., Chang, N.H., 1993. High cell density cultivation of Escherichia coli W using sucrose as a carbon source. Biotechnol. Lett. 15, 971–974.
Luli, G.W., Strohl, W.R., 1990. Comparison of growth, acetate production, and acetate inhibition of Escherichia coli strains in batch and fed-batch fermentations. Appl. Environ. Microbiol. 56, 1004–1011.
Märkl, H., Zenneck, C., Dubach, A.C., Ogbonna, J.C., 1993. Cultivation of Escherichia coli to high cell densities in a dialysis reactor. Appl. Microbiol. Biotechnol. 39, 48–52. Matsudaira, P., 1987. Sequence from picomole quantities of
proteins electroblotted onto polyvinylidene difluoride membra-nes. J. Biol. Chem. 262, 10035–10038.
McCormick, S.D., Sakamoto, T., Hasegawa, S., Hirano, T., 1991. Osmoregulatory actions of insulin-like growth factor-1 in rainbow trout (Oncorhynchus mylciss). J. Endocrinol. 130, 87– 92.
Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal. Chem. 31, 426–428. Moriyama, S., Duguay, S.J., Conlon, J.M., Duan, C., Dickhoff,
W.W., Plisetskaya, E.M., 1993. Recombinant coho salmon insulin-like growth factor-1: expression in Escherichia coli, purification and characterization. Eur. J. Biochem. 218, 205– 211.
Moriyama, S., Ayson, F.G., Kawauchi, H., 2000. Review: growth regulation by insulin-like growth factor-1 in fish. Biosci. Biotechnol. Biochem. 64, 1553–1562.
Nakano, K., Rischke, M., Sato, S., Märkl, H., 1997. Influence of acetic acid on the growth of Escherichia coli K-12
during high-cell density in dialysis reactor. Appl. Microbiol. Biotechnol. 48, 597–601.
Okajima, T., Nakamura, K., Zhang, H., Ling, N., Tanabe, T., Yasuda, T., Rosenfeld, R.G., 1992. Sensitive colorimetric bioassays for insulin-like growth factor (IGF) stimulation of cell proliferation and glucose consumption: use in studies of IGF analogs. Endocrinology 130, 2201–2212.
Schagger, H., Jagow, G.V., 1987. Tricine-sodium dodecyl sulfate polyacrylamide gel electrophoresis for the separation of protein in the range from 1 to 100 kDa. Anal. Biochem. 166, 368–379. Shimatsu, A., Rotwein, P., 1987. Mosaic evolution of the insulin-like growth factors: organization, sequence, and expression of the rat insulin-like growth factor-1 gene. J. Biol. Chem. 262, 7894–7900.
Srinivasan, S., Barnard, G.C., Gerngross, T.U., 2002. A novel high-cell-density protein expression system based on Ralstonia eutropha. Appl. Environ. Microbiol. 68, 5925–5932.
Upton, Z., Francis, G.L., Kita, K., Wallace, J.C., Ballard, F.J., 1995. Production and characterization of recombinant chicken insulin-like growth factor-2 from Escherichia coli. J. Mol. Endocrinol. 14, 90–94.
Wang, F., Lee, S.Y., 1997. Production of poly(3-hydroxybutyrate) by fed-batch culture of filamentation-suppressed recombinant Escherichia coli. Appl. Environ. Microbiol. 34, 382– 385.
Wangsa-Wirawan, N.D., Lee, Y.S., Falconer, R.J., Mansell, C.J., O’Neill, B.K., Middelberg, A.P.J., 1997. Novel fed-batch strategy for the production of insulin-like growth factor-1 (IGF-1). Biotechnol. Lett. 19, 1079–1082.
Yamada, H., Saito, Y., Fujimoto, T., Noguchi, Y., Mori, T., Miura, T., Kobayashi, M., Shimomura, K., 1996. Large scale purification of recombinant insulin-like growth factor-1 (IGF-1, Mecasermin) from a fused-protein produced in Escherichia coli. J. Ferment. Bioeng. 82, 134–139.
