- Email: [email protected]
Bacterial expression and purification of biologically active human TFF3
Peptides 25 (2004) 785–792
Bacterial expression and purification of biologically active human TFF3 Min Fang, Wei Wang, Yanru Wang, Binggen Ru∗ Department of Biochemistry and Molecular Biology, National Laboratory of Protein Engineering, College of Life Science, Peking University, Beijing, PR China Received 9 September 2003; accepted 16 January 2004
Abstract A glutathione S-transferase (GST) fusion protein expression system for the production and purification of recombinant human trefoil factor family-domain peptide 3 (hTFF3) was established. The hTFF3 gene, prepared by PCR, was cloned into a pBluescript KS(+) plasmid, and inserted into a pGEX-4T-1 GST fusion vector. The GST-hTFF3 fusion protein was expressed in Escherichia coli, and hTFF3 was purified with Glutathione Sepharose 4B affinity chromatography, yielding about 3–4 mg of pure hTFF3 in one liter of culture broth. The biological activity of purified hTFF3 was tested in two previously reported rat gastric ulcer models. Oral administration of recombinant hTFF3 has a dose dependent protective effect against ethanol-induced or pylorus ligation-induced gastric mucosa injury in rat, which indicates that our recombinant hTFF3 is biologically active. © 2004 Elsevier Inc. All rights reserved. Keywords: Human trefoil factor family-domain peptide 3 (hTFF3); Glutathione S-transferase (GST); Bacterial expression system; Gastric ulcer; Rat
1. Introduction Since the first discovery of trefoil factor family peptides nearly 20 years ago [4], three members of mammalian TFF peptides, i.e. trefoil factor family peptide 1 (TFF1 or pS2), trefoil factor family peptide 2 (TFF2 or SP) and trefoil factor family peptide 3 (TFF3 or ITF) have been cloned [7,10]. TFF peptides are usually synthesized in mucin-secreting epithelial cells and are predominantly secreted in gastrointestinal tract [15]. The trefoil domain in TFF peptides is formed by six highly conserved cysteine residues, creating a characteristic three-leaved structure, which gives the peptide family its name [12]. It has been reported that TFF peptides play an important role in maintaining the epithelial mucosal barrier in the gastrointestinal tract [1,12]. TFF3 is mainly expressed in the goblet cells of the small and large intestine [3]. Human TFF3 (hTFF3), which has one characteristic trefoil domain within its total 59 amino acids, can also form homodimers through inter-molecular disulfide bridge at position 57. Human TFF3 has been found to have functions of promoting wound healing, stimulating epithelial cell migration, and maintaining intestinal mucosal barrier [2]. Orally administered TFF3 can protect rat gastric
∗
Corresponding author. Fax: +86-10-62751842. E-mail address: [email protected] (B. Ru).
0196-9781/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.peptides.2004.01.025
mucosa against indomethacin- and ethanol-induced injury [1]. Only very limited amount of hTFF3 can be prepared from human tissue extract. Many efforts have been made to express large quantity of TFF peptides in either yeast or Escherichia coli expression system [5,13,14]. Here a method using GST-fusion strategy to produce sufficient amount of hTFF3 for physiological and biochemical studies is described.
2. Materials and methods 2.1. Cloning of hTFF3 The hTFF3 cDNA was amplified by polymerase chain reaction (PCR) using human fetal colon cDNA as the template. The forward primer sequence was 5 -GGGGTACCAGATCTGAGGAGTACGTGGGC-3 , and the reverse primer sequence was 5 -CGGAATTCCTATCAGAAGGTG-3 . The PCR reaction was carried out in the following conditions: denature at 94 ◦ C for 1 min, annealing at 60 ◦ C for 45 s, and enlongation at 72 ◦ C for 45 s, followed by extension at 72 ◦ C for 10 min. A total of 30 cycles were used and the final product was then digested with KpnI and EcoRI before being ligated into the pBluescript KS(+) plasmid. The resulting plasmid was named as pBTF.
786
M. Fang et al. / Peptides 25 (2004) 785–792
2.2. Expression of hTFF3 in E. coli BL21 cells The hTFF3 cDNA was digested from pBTF with Bgl II and EcoRI, and inserted into a pGEX-4T-1 glutathione S-transferase (GST) fusion vector which had been digested with BamHI and EcoRI. This resulting plasmid, named pGTF, was then used to transform E. coli BL21 competent cells. Sterile 2-YT broth containing 100 g/ml of ampicillin was inoculated with an overnight culture of transformed E. coli at a dilution of 1:10, and incubated at 37 ◦ C with shaking. When the cells were grown to an OD550 of 0.5, GST-hTFF3 expression was induced by adding 0.4 mM isopropyl- -d-thiogalactoside (IPTG). After induction for four hours, the cells were pelleted by centrifugation at 4000 × g for 15 min, and re-suspended in 50 ml phosphate-buffered saline (PBS, 140 mmol/l NaCl, 2.7 mmol/l KCl, 10 mmol/l Na2 HPO4 , 1.8 mmol/l KH2 PO4 , pH 7.3). The re-suspended cells were disrupted by sonication at 50 W for 20–15 s cycles with a 1/8 in. diameter microtip (Xinzhi Ultrasonics Corporation, Zhejiang, China). The mixture was then centrifuged at 15,000 × g for 15 min. The supernatant was retained for hTFF3 purification. 2.3. Purification of hTFF3 Glutathione Sepharose 4B (GS4B) affinity column (1.0 cm × 5 cm) (Amersham Pharmacia Biotech) was equilibrated with 10 bed-volumes of PBS and 10 bed-volumes of PBS containing 1% Triton-X-100. The cell extract was loaded onto the column at a flow rate of 0.5 ml/min .The column was then washed thoroughly with 20 bed-volumes of PBS to remove unbound material and equilibrated in 1 bed-volume of PBS. The GST-hTFF3 fusion protein was cleaved on column with thrombin (10 U/mg fusion protein, Amersham Pharmacia Biotech) overnight (12 h) at room temperature, and then eluted by PBS. The solution was loaded onto a Sephacryl-S-100 column (2.6 cm × 80 cm), which has been pre-equilibrated by 10 mM NH4 HCO3 . UV-absorption was monitored at 280 nm. The fractions corresponding to hTFF3 were pooled and lyophilized. Protein samples were analyzed by sodium dodecyl sulfate–polycrylamide gel electrophoresis (SDS–PAGE) using 15% acrylamide. Gels were stained with 0.02% (w/v) Coomassie brilliant blue R-250 and de-stained with a 40% (v/v) methanol, 7% (v/v) acetic acid solution. The protein product was confirmed by Western blot analysis using specific rabbit anti-hTFF3 poly-clonal antibody. 2.4. Characterization of recombinant hTFF3 2.4.1. Mass spectrometry analysis Mass spectrometry analysis of purified hTFF3 was performed on a Voyager MALDI/TOF mass spectrometer (PerSeptive Biosystems, Framingham, MA) in the
Chemistry Department of Peking University. Protein samples were diluted into freshly prepared saturated sinapinic acid dissolved in 50% acetonitrile, 0.3% TFA. Aliquots of 0.5 or 1.0 ml samples were spotted onto stainless steel sample plates and spectra were collected by averaging 10–20 laser shots. Samples were irradiated with a 377-nm nitrogen lazer (Laser Science, Franklin, MA). 2.4.2. N-terminal amino acid sequence analysis of purified hTFF3 Amino terminal sequence analysis of the recombinant hTFF3 was carried out with a PE Applied Biosystems Model 491A Protein Sequencer. The hTFF3 was first transferred to a polyvinylidene fluoride (PVDF) membrane (BioRad) from an SDS–PAGE gel, followed by analysis of the N-terminal amino acid sequence. 2.4.3. The amino acid composition analysis After hydrolysis in 6 M HCl at 110 ◦ C in vacuum-sealed tubes for 24, 48 and 96 h, protein samples (100 g) were analyzed on a Beckman (Model 121MB) automatic amino acid analyzer. After reduction of the disulfide bonds by tributylphosphine, cysteine was determined as the S--(4-pyridylethyl) derivative, followed by coupling with 4-vinylpyridine. Hydrolyses of 4-vinylpyridine-treated samples were performed in 4 M methanesulphonic acid or 3 M mercapto-ethanesulphonic acid at 110 ◦ C for 24 h as described above. 2.5. Biological assay of recombinant hTFF3 in rat gastric injury models Male Wistar rats weighing 180–220 g (Weiduolihua Lit., Beijing, China) were maintained at a controlled temperature (21 ± 2 ◦ C) and a 12 h/12 h light/dark cycle. Animals were housed in cages with wide mesh wire bottoms to prevent coprophagy. Food was withheld for 24 h before each experiment, but drinking water was allowed ad libitum. 2.5.1. Protection of gastric mucosa against ethanol-induced gastric injury Animals were divided into six groups with 5–8 rats in each group. Under light anesthesia, hTFF3 at the following dosage: 0.1, 0.25, 0.5, 1 and 2 mg/kg, was administered orally to rats of each group by gastric lavage using a curved spinal needle (Becton Dickinson, Franklin Lakes, NJ). One group was given normal saline as a control. One hour later, gastric mucosal injury was induced by intragastric administration of absolute ethanol (1.0 ml/rat). One hour after ethanol administration, the rats were immediately killed by cervical dislocation. The stomachs were kept moist with saline and photographed, and the lesions were scored independently by two investigators in a blinded manner. Gastric lesions were scored by a previously described scoring system with modification as following: grade 0, no lesions; grade 1, <2 mm of lesions;
M. Fang et al. / Peptides 25 (2004) 785–792
787
grade 2, 2–6 mm of lesions; grade 3, >6 mm of lesions. The tissues then were preserved in 10% buffered formalin for histological examination. A 5–7 m thickness sections were stained with haematoxylin and eosin and photographed under 100× magnification. 2.5.2. Protection of pylorus ligation induced ulceration in rats For further analysis of the biological activity of hTFF3, a pylorus ligation-induced gastric ulcer model was also tested in rats. Animals were divided into three groups with 6–7 rats in each group. One group received oral hTFF3 at a dose of 0.25 mg/kg, and the other group received 0.5 mg/kg. An additional group received normal saline as control. Gastric mucosal lesions were produced by pyloric ligation as described by Shay et al. [11]. Briefly, the abdomen was opened by a small midline incision below the xiphoid process under ether anesthesia; the pyloric portion of the stomach was slightly lifted and ligated, avoiding any damage to the adjacent blood vessels [9]. Food and water were not provided to the animals during the post-operation period. Rats were killed under ether anesthesia 4 h later. Rat stomach was excised and the gastric juice was collected and measured for the total volume and acidity. Then the stomach was cut open along the greater curvature, and the gross mucosal injury was assessed by a lesion index based on the method described above. After scoring, mucosal specimens were scraped off using a slide glass and immediately snap frozen in liquid nitrogen and stored at −80 ◦ C until analysis. 2.6. Statistical analysis Data were presented as mean ± S.E.M. Analysis of variance (ANOVA) and Student’s t-test were used for statistical analysis for the difference among groups. The difference was considered to be statistical significant when P value is less than 0.05.
Fig. 1. PCR amplification of TFF3. Lane 1: DNA ladder; lane 2: PCR product.
apparent molecular size of 31 kDa, which corresponds well with the predicted total molecular weight of the fusion protein (Fig. 2). The recombinant fusion protein constitutes approximately 15% of the total cellular proteins. 3.3. Purification of hTFF3 The purification process yielded hTFF3 polypeptide which is shown in Fig. 3A and B. The Western blot analysis of the purified recombinant hTFF3 was shown in Fig. 4. The final yield of hTFF3 was determined to be about 3–4 mg/l culture broth, which is higher than the typical yield of 2.5 mg/l for GST fusion proteins (Amersham Pharmacia Biotech). However, this yield may vary among different cultures, depending on specific growing conditions and incubation times, which can be further optimized.
3. Results
3.4. Characterization of recombinant hTFF3
3.1. Cloning of hTFF3
3.4.1. Mass spectrometry analysis Mass spectrometry analysis of hTFF3 shows two peaks corresponding to two molecular weights: 13.441 and 6.832 kDa (Fig. 5). The calculated molecular weight of hTFF3 homodimer is 13433 (including two N-terminal amino acid residues Gly and Ser from the vector). This is in a good agreement with the molecular weight determined. The calculated molecular weight of hTFF3 in a monomer form in which Cys-57 contains a free –SH group is 6717.5, which is 114.5 lower than the determined value of 6832. This is probably due to an additional Cys residue bonded to the free Cys-57 or the free Cys-57 has been oxidized.
The amplified PCR fragment from human fetal colon tissue was examined by 2% agarose gel electrophoresis. A band of approximately 200 nucleotides was observed (Fig. 1). The cDNA fragment was successfully sub-cloned into the GST expression vector. The orientation and DNA sequence was verified (results not shown). 3.2. Expression of hTFF3 in E. coli BL21 cells The recombinant hTFF3 was expressed in a soluble form, as a fusion protein with GST. The protein has an
788
M. Fang et al. / Peptides 25 (2004) 785–792
Fig. 2. SDS–PAGE analysis for expressed protein: (1) crude lysate of BL21transformed with pGTF with IPTG induction; (2) crude lysate of BL21 with IPTG induction; (3) crude lysate of BL21 transformed with GEX-4T-1 with IPTG induction; (4) molecular weight marker.
Fig. 3. (A) Gel filtration pattern on Sephacryl-S-100 of eluent from Glutathione Sepharose 4B after digested with thrombin. (A) Thrombin (B) hTFF3. (B) SDS–PAGE analysis for purified hTFF3 (1) molecular weight marker (2) purified TFF3.
According to Fig. 5, the purified recombinant hTFF3 was both in dimer form and monomer form . The presence of only the monomer form on the SDS–PAGE gel (Fig. 3B) was most likely due to the typical reducing process used in the experiment. To confirm this, we run another SDS–PAGE gel which shows less reduced hTFF3.(Fig. 6). According to lane 3, it seemed that the purified recombinant TFF3 mainly existed as dimmers.
Fig. 4. Western-blot of recombinant hTFF3. Lane 1: recombinant hTFF3; lane 2: standard hTFF3.
3.4.2. N-terminal amino acid sequence analysis of purified hTFF3 The N-terminal amino acid sequence analysis revealed that the final product was contaminated by a peptide with an N-terminal sequence of Gly-Ser-Glu-Glu-Tyr-Val-Gly-LeuSer-Ala. Two amino acid residues (Gly and Ser) before the
M. Fang et al. / Peptides 25 (2004) 785–792
789
Fig. 5. Mass spectrometry of recombinant hTFF3. The two peaks corresponding to two molecular weights: 13.441 and 6.832 kDa.
Glu residue were from the cleavage site of pGEX-4T-1 vector. The residues after the Glutamine residue are the same with the natural hTFF3. 3.4.3. The amino acid composition analysis Table 1 shows the amino acid composition of the purified hTFF3, which was in a fairly good agreement with the expected values (including two amino acid residues Gly and Ser from the pGEX-4T-1 vector). 3.5. Biological assay of recombinant hTFF3 in rat gastric injury models 3.5.1. Protection of gastric mucosa against ethanol-induced gastric injury in rats According to Table 2 and Fig. 7, hTFF3 shows a dose dependent protective function in rat gastric injury models. Compared with rats pretreated with saline, protection was observed at a dose of 100 g/kg although the P value is more than 0.05 which means not significant. But the degree of protection increased with increasing doses of peptide from 100 g/kg to 2 mg/kg. The P values were all less than 0.05 when the doses were more than 250 g/kg. The dose dependent protection degree indicates that the purified recombinant hTFF3 has its biological activity.
3.5.2. Protection of pylorus ligation-induced ulceration in rats Similar protective effect can also be observed for hTFF3 in pylorus ligation-induced gastric injury model (Table 3). However, hTFF3 has no effect in regulating gastric acid secretion in this model. In 4 h period, the control rats secreted 36.47 ± 14.5 meq. of HCl, while the 0.25 mg/kg hTFF3 treated group secreted 33.9 ± 7.24 meq. and the 0.5 mg/kg hTFF3 treated group secreted 34.49±9.35 meq. (P > 0.05). 4. Discussion In this study, we described the usage of a GST fusion protein expression system for the production of recombinant hTFF3 in milligram quantity. This is the first description of using GST system for the expression of trefoil peptides. The expression level is fairly high and the purification is relatively simple and efficient as compared to the reported yeast expression system [5,13,14]. The purified recombinant hTFF3 peptide was found both in monomer and in dimer form. We have previously reported that the dimmer form is the biologically active form in vivo [3,8]. The purified recombinant hTFF3 in this study contains two additional amino acid residues in the N-terminal (Gly and Ser), however it has the same biological activity as shown in the gastric mucosa injury models.
790
M. Fang et al. / Peptides 25 (2004) 785–792 Table 1 Amino add composition of recombinant hTFF3 Amino acid
Experimental
Theoretical
Asp Asn Thr Ser Glu Gin Pro Gly Ala Cys Val Met Ile Leu Tyr Phe Lys His Trp Arg
2.86 2.81 1.68 2.71 5.36 1.79 5.94 4.93 3.72 7.56 5.18 0.21 1.35 1.89 1.77 3.13 2.89 1.14 1.34 3.29
3 3 2 3 5 2 6 5 4 7(8) 5 0 1 2 2 3 3 1 1 3
Total
61.55
61(62)
Epithelium is more intact, compared with that of the damaged rat corpus seen in B and C. This indicates that hTFF3 shows a dose dependent protective function in rat gastric injury model. Fig. 6. SDS–PAGE analysis for reduced and non-reduced hTFF3: (1) purified recombinant hTFF3 with reducing process/the monomer form of TFF3 was observed; (2) molecular weight marker; (3) purified recombinant hTFF3 with less reducing agent (2-mercaptoethanol) used in the SDS–PAGE process. Both the dimer form and monomer form of TFF3 were observed. But it seemed that the purified recombinant TFF3 mainly existed as dimmers.
Trefoil factors are known to have protective effect in gastric intestinal tract against indomethacin- and ethanol-induced mucosal injury. Here the protective effect of hTFF3 against ethanol-induced gastric mucosal injury was used to confirm that our purified hTFF3 had the known biological property of gastric mucosal protection. Furthermore, a rat model of pyloric ligation-induced gastric ulcer was also used in our study to further examine the function of hTFF3. Prophylactic oral administration of hTFF3 has a dose-dependent protective effect against pyloric ligation-induced gastric ulcer in rat. Since the gastric acid secretion was not affected by administration of hTFF3, its gastric mucosal protection is unlikely mediated by changing other factors such as acid secretion. The mechanisms of TFF3 for gastrointestinal mucosal protection are still not clear. Some studies have indicated that TFF3 formed a protective layer by combining with mucins in the gastric intestinal tract that can prevent the damage of injurious factors such as ethanol, acids, etc. [6]. TFF3 can also promote the restitution, i.e. the migration of epithelial cell from the normal mucosa to the adjacent damaged mucosa [16]. The distinct trefoil structure formed by three
Table 2 Effect of recombinant hTFF3 pretreatment on the ethanol-induced gastric mucosal injury SI. no.
Treatment and dose (mg/kg)
Animal number
Ulcer index (mean ± S.E.)
1 2 3 4 5 6
Control (saline) hTFF3 (0.1) hTTF3 (0.25) hTTPS (0.5) hTFF3 (1.0) hTFF3 (2.0)
8 8 7 8 8 5
42.36 37.10 30.85 28.63 27.25 18.80
± ± ± ± ± ±
10.38 10.71 9.13 7.94 6.41 5.82
P
>0.05 <0.05 <0.05 <0.05 <0.01
Table 3 Effect of recombinant hTFF3 pretreatment on the induction of gastric ulcers by pylorus ligation SI. no.
Treatment and dose (mg/kg)
Animal number
Ulcer index (mean ± S.E.)
1 2 3
Control (saline) hTFF3 (0.25) hTFF3 (0.50)
7 6 7
8.4 ± 2.2 4.5 ± 2.3 3.1 ± 0.9
P
<0.01 <0.01
intra-chain disulfide bonds between six highly conserved cysteine residues is thought to contribute its remarkable resistant to acid and protease digestion of the gastrointestinal tract and therefore remains biologically active when it is administered orally.
M. Fang et al. / Peptides 25 (2004) 785–792
791
Fig. 7. Gastroprotection by recombinant hTFF3 microscopic pathology. Fasted rats were treated intragastrically with saline or 0.25 mg/kg TFF3 or 0.5 mg/kg TFF3. One hour later, gastric mucosal injury was induced by intragastric administration of absolute ethanol (1.0 ml/rat). Rats were killed one hour later. (A) Normal rat gastric epithelium (100×). (B) Rat gastric epithelium seriously injured by ethanol (100×). Note that the gastric corpus fovcolar epithelium is seriously damaged, compared with that of the undamaged rat corpus seen in A. (C) Rat gastric epithelium treated with 0.25 mg/kg recombinant hTFF3 before ethanol administration (100×). Note that the gastric corpus foveolar epithelium is intact, compared with that of the damaged rat corpus seen in B. (D) Rat gastric epithelium treated with 0.50 mg/kg recombinant hTFF3 before ethanol administration (100×). Note that the gastric corpus foveolar.
Acknowledgments We thank Prof. Fangwen Dishen for providing human fetal colon cDNA, Yunpeng Su for providing rabbit anti-hTFF3 poly-clonal antibody, Xiuyun Dong (The Third Clinical Hospital of Peking University) for assistance with the animal models, Yiyou Chen, Tiegang Han and Jing Lin for the helpful comments on the manuscript. References [1] Babyatsky MW, DeBeaumont M, Thim L, et al. Oral trefoil peptides protect against ethanol-induced and indomethacin-induced gastric injury in rats. Gastroenterology 1996;110:489–97. [2] Chinery R, Bates PA, Amitabha DE, et al. Characterization of the single copy trefoil peptides intestinal trefoil factor and pS2 and their ability to form covalent dimmers. FEBS Lett 1995;357(1):50–4. [3] Hauser F, Poulsom R, Chinery R, et al. hP1.B, a human P-domain peptide homologous with rat intestinal trefoil factor is expressed also in the ulcer-associated cell lineage and the uterus. Proc Natl Acad Sci USA 1993;90(15):6961–5. [4] Jorgensen KH, Thim L, Jacobsen HE. Preparation and characterization of PSP, a new polypeptide from porcine pancreas. Diabetalogia 1981;21:288–94.
[5] Kannan R, Tomasetto C, Staub A, Bossenmeyer-Pourié C, Thim L, Nielsen PF, et al. Human pS2/Trefoil Factor 1: production and characterization in Pichia pastoris. Protein Exp Purif 2001;21(1):92– 8. [6] Kindon H, Pothoulakis C, Thim L, et al. Trefoil peptide protection of intestinal epithelial barrier function: cooperative interaction with mucin glycoprotein. Gastroenterology 1995;109(2):516–23. [7] Kou RQ, Wang W, Li LY, Ru BG. Trefoil factor family. Prog Biochem Biophys 2000;27(4):367–72. [8] Kou RQ, Wang W, Li LY, Ru BG. Construction and bioactivity of Cys57 mutant of human intestinal trefoil factor. Chin J Biochem Mol Biol 2000;16(5):606–11. [9] Rastogi L, Patnaik GK, Dikshit M. Free radicals and antioxidant status following pylorus ligation induced gastric mucosal injury in rats. Pharmacol Res 1998;38(2):125–32. [10] Sands BE, Podolsky DK. The trefoil peptide family. Annu Rev Physiol 1996;58:253–73. [11] Shay M, Kamarov SA, Fels D, Meranze D, Gruen-stein H, Siplet H. A simple method for the uniform production of gastric ulceration in the rat. Gastroenterology 1945;5:43–61. [12] Thim L. A new family of growth factor-like peptides: trefoil disulphide loop structures as a common feature in breast cancer associated peptide(pS2), pancreatic spasmolytic polypeptide(pSP) and frog skin peptides (spasmolysins). FEBS Lett 1989;250:85– 90.
792
M. Fang et al. / Peptides 25 (2004) 785–792
[13] Thim L, Wödike HF, Nielsen PF, et al. Characterization of human and rat intestinal trefoil factor produced in yeast. Biochemistry 1995;34(14):4757–64. [14] Thim L, Kjeld N, Fanny N, et al. Purification and characterization of the trefoil peptide human spasmolytic polypeptide (hSP) produced in yeast. FEBS Lett 1993;318:345–52.
[15] Williams GR, Wright NA. Trefoil factor family domain peptides. Virchows Archiv Int J Pathol 1997;431(5):299– 304. [16] Wright NA, Hoffmann W, Otto WR, et al. Rolling in the clover: trefoil factor family (TFF)-domain peptides, cell migration and cancer. FEBS Lett 1997;408(2):121–3.
Bacterial expression and purification of biologically active human TFF3 Min Fang, Wei Wang, Yanru Wang, Binggen Ru∗ Department of Biochemistry and Molecular Biology, National Laboratory of Protein Engineering, College of Life Science, Peking University, Beijing, PR China Received 9 September 2003; accepted 16 January 2004
Abstract A glutathione S-transferase (GST) fusion protein expression system for the production and purification of recombinant human trefoil factor family-domain peptide 3 (hTFF3) was established. The hTFF3 gene, prepared by PCR, was cloned into a pBluescript KS(+) plasmid, and inserted into a pGEX-4T-1 GST fusion vector. The GST-hTFF3 fusion protein was expressed in Escherichia coli, and hTFF3 was purified with Glutathione Sepharose 4B affinity chromatography, yielding about 3–4 mg of pure hTFF3 in one liter of culture broth. The biological activity of purified hTFF3 was tested in two previously reported rat gastric ulcer models. Oral administration of recombinant hTFF3 has a dose dependent protective effect against ethanol-induced or pylorus ligation-induced gastric mucosa injury in rat, which indicates that our recombinant hTFF3 is biologically active. © 2004 Elsevier Inc. All rights reserved. Keywords: Human trefoil factor family-domain peptide 3 (hTFF3); Glutathione S-transferase (GST); Bacterial expression system; Gastric ulcer; Rat
1. Introduction Since the first discovery of trefoil factor family peptides nearly 20 years ago [4], three members of mammalian TFF peptides, i.e. trefoil factor family peptide 1 (TFF1 or pS2), trefoil factor family peptide 2 (TFF2 or SP) and trefoil factor family peptide 3 (TFF3 or ITF) have been cloned [7,10]. TFF peptides are usually synthesized in mucin-secreting epithelial cells and are predominantly secreted in gastrointestinal tract [15]. The trefoil domain in TFF peptides is formed by six highly conserved cysteine residues, creating a characteristic three-leaved structure, which gives the peptide family its name [12]. It has been reported that TFF peptides play an important role in maintaining the epithelial mucosal barrier in the gastrointestinal tract [1,12]. TFF3 is mainly expressed in the goblet cells of the small and large intestine [3]. Human TFF3 (hTFF3), which has one characteristic trefoil domain within its total 59 amino acids, can also form homodimers through inter-molecular disulfide bridge at position 57. Human TFF3 has been found to have functions of promoting wound healing, stimulating epithelial cell migration, and maintaining intestinal mucosal barrier [2]. Orally administered TFF3 can protect rat gastric
∗
Corresponding author. Fax: +86-10-62751842. E-mail address: [email protected] (B. Ru).
0196-9781/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.peptides.2004.01.025
mucosa against indomethacin- and ethanol-induced injury [1]. Only very limited amount of hTFF3 can be prepared from human tissue extract. Many efforts have been made to express large quantity of TFF peptides in either yeast or Escherichia coli expression system [5,13,14]. Here a method using GST-fusion strategy to produce sufficient amount of hTFF3 for physiological and biochemical studies is described.
2. Materials and methods 2.1. Cloning of hTFF3 The hTFF3 cDNA was amplified by polymerase chain reaction (PCR) using human fetal colon cDNA as the template. The forward primer sequence was 5 -GGGGTACCAGATCTGAGGAGTACGTGGGC-3 , and the reverse primer sequence was 5 -CGGAATTCCTATCAGAAGGTG-3 . The PCR reaction was carried out in the following conditions: denature at 94 ◦ C for 1 min, annealing at 60 ◦ C for 45 s, and enlongation at 72 ◦ C for 45 s, followed by extension at 72 ◦ C for 10 min. A total of 30 cycles were used and the final product was then digested with KpnI and EcoRI before being ligated into the pBluescript KS(+) plasmid. The resulting plasmid was named as pBTF.
786
M. Fang et al. / Peptides 25 (2004) 785–792
2.2. Expression of hTFF3 in E. coli BL21 cells The hTFF3 cDNA was digested from pBTF with Bgl II and EcoRI, and inserted into a pGEX-4T-1 glutathione S-transferase (GST) fusion vector which had been digested with BamHI and EcoRI. This resulting plasmid, named pGTF, was then used to transform E. coli BL21 competent cells. Sterile 2-YT broth containing 100 g/ml of ampicillin was inoculated with an overnight culture of transformed E. coli at a dilution of 1:10, and incubated at 37 ◦ C with shaking. When the cells were grown to an OD550 of 0.5, GST-hTFF3 expression was induced by adding 0.4 mM isopropyl- -d-thiogalactoside (IPTG). After induction for four hours, the cells were pelleted by centrifugation at 4000 × g for 15 min, and re-suspended in 50 ml phosphate-buffered saline (PBS, 140 mmol/l NaCl, 2.7 mmol/l KCl, 10 mmol/l Na2 HPO4 , 1.8 mmol/l KH2 PO4 , pH 7.3). The re-suspended cells were disrupted by sonication at 50 W for 20–15 s cycles with a 1/8 in. diameter microtip (Xinzhi Ultrasonics Corporation, Zhejiang, China). The mixture was then centrifuged at 15,000 × g for 15 min. The supernatant was retained for hTFF3 purification. 2.3. Purification of hTFF3 Glutathione Sepharose 4B (GS4B) affinity column (1.0 cm × 5 cm) (Amersham Pharmacia Biotech) was equilibrated with 10 bed-volumes of PBS and 10 bed-volumes of PBS containing 1% Triton-X-100. The cell extract was loaded onto the column at a flow rate of 0.5 ml/min .The column was then washed thoroughly with 20 bed-volumes of PBS to remove unbound material and equilibrated in 1 bed-volume of PBS. The GST-hTFF3 fusion protein was cleaved on column with thrombin (10 U/mg fusion protein, Amersham Pharmacia Biotech) overnight (12 h) at room temperature, and then eluted by PBS. The solution was loaded onto a Sephacryl-S-100 column (2.6 cm × 80 cm), which has been pre-equilibrated by 10 mM NH4 HCO3 . UV-absorption was monitored at 280 nm. The fractions corresponding to hTFF3 were pooled and lyophilized. Protein samples were analyzed by sodium dodecyl sulfate–polycrylamide gel electrophoresis (SDS–PAGE) using 15% acrylamide. Gels were stained with 0.02% (w/v) Coomassie brilliant blue R-250 and de-stained with a 40% (v/v) methanol, 7% (v/v) acetic acid solution. The protein product was confirmed by Western blot analysis using specific rabbit anti-hTFF3 poly-clonal antibody. 2.4. Characterization of recombinant hTFF3 2.4.1. Mass spectrometry analysis Mass spectrometry analysis of purified hTFF3 was performed on a Voyager MALDI/TOF mass spectrometer (PerSeptive Biosystems, Framingham, MA) in the
Chemistry Department of Peking University. Protein samples were diluted into freshly prepared saturated sinapinic acid dissolved in 50% acetonitrile, 0.3% TFA. Aliquots of 0.5 or 1.0 ml samples were spotted onto stainless steel sample plates and spectra were collected by averaging 10–20 laser shots. Samples were irradiated with a 377-nm nitrogen lazer (Laser Science, Franklin, MA). 2.4.2. N-terminal amino acid sequence analysis of purified hTFF3 Amino terminal sequence analysis of the recombinant hTFF3 was carried out with a PE Applied Biosystems Model 491A Protein Sequencer. The hTFF3 was first transferred to a polyvinylidene fluoride (PVDF) membrane (BioRad) from an SDS–PAGE gel, followed by analysis of the N-terminal amino acid sequence. 2.4.3. The amino acid composition analysis After hydrolysis in 6 M HCl at 110 ◦ C in vacuum-sealed tubes for 24, 48 and 96 h, protein samples (100 g) were analyzed on a Beckman (Model 121MB) automatic amino acid analyzer. After reduction of the disulfide bonds by tributylphosphine, cysteine was determined as the S--(4-pyridylethyl) derivative, followed by coupling with 4-vinylpyridine. Hydrolyses of 4-vinylpyridine-treated samples were performed in 4 M methanesulphonic acid or 3 M mercapto-ethanesulphonic acid at 110 ◦ C for 24 h as described above. 2.5. Biological assay of recombinant hTFF3 in rat gastric injury models Male Wistar rats weighing 180–220 g (Weiduolihua Lit., Beijing, China) were maintained at a controlled temperature (21 ± 2 ◦ C) and a 12 h/12 h light/dark cycle. Animals were housed in cages with wide mesh wire bottoms to prevent coprophagy. Food was withheld for 24 h before each experiment, but drinking water was allowed ad libitum. 2.5.1. Protection of gastric mucosa against ethanol-induced gastric injury Animals were divided into six groups with 5–8 rats in each group. Under light anesthesia, hTFF3 at the following dosage: 0.1, 0.25, 0.5, 1 and 2 mg/kg, was administered orally to rats of each group by gastric lavage using a curved spinal needle (Becton Dickinson, Franklin Lakes, NJ). One group was given normal saline as a control. One hour later, gastric mucosal injury was induced by intragastric administration of absolute ethanol (1.0 ml/rat). One hour after ethanol administration, the rats were immediately killed by cervical dislocation. The stomachs were kept moist with saline and photographed, and the lesions were scored independently by two investigators in a blinded manner. Gastric lesions were scored by a previously described scoring system with modification as following: grade 0, no lesions; grade 1, <2 mm of lesions;
M. Fang et al. / Peptides 25 (2004) 785–792
787
grade 2, 2–6 mm of lesions; grade 3, >6 mm of lesions. The tissues then were preserved in 10% buffered formalin for histological examination. A 5–7 m thickness sections were stained with haematoxylin and eosin and photographed under 100× magnification. 2.5.2. Protection of pylorus ligation induced ulceration in rats For further analysis of the biological activity of hTFF3, a pylorus ligation-induced gastric ulcer model was also tested in rats. Animals were divided into three groups with 6–7 rats in each group. One group received oral hTFF3 at a dose of 0.25 mg/kg, and the other group received 0.5 mg/kg. An additional group received normal saline as control. Gastric mucosal lesions were produced by pyloric ligation as described by Shay et al. [11]. Briefly, the abdomen was opened by a small midline incision below the xiphoid process under ether anesthesia; the pyloric portion of the stomach was slightly lifted and ligated, avoiding any damage to the adjacent blood vessels [9]. Food and water were not provided to the animals during the post-operation period. Rats were killed under ether anesthesia 4 h later. Rat stomach was excised and the gastric juice was collected and measured for the total volume and acidity. Then the stomach was cut open along the greater curvature, and the gross mucosal injury was assessed by a lesion index based on the method described above. After scoring, mucosal specimens were scraped off using a slide glass and immediately snap frozen in liquid nitrogen and stored at −80 ◦ C until analysis. 2.6. Statistical analysis Data were presented as mean ± S.E.M. Analysis of variance (ANOVA) and Student’s t-test were used for statistical analysis for the difference among groups. The difference was considered to be statistical significant when P value is less than 0.05.
Fig. 1. PCR amplification of TFF3. Lane 1: DNA ladder; lane 2: PCR product.
apparent molecular size of 31 kDa, which corresponds well with the predicted total molecular weight of the fusion protein (Fig. 2). The recombinant fusion protein constitutes approximately 15% of the total cellular proteins. 3.3. Purification of hTFF3 The purification process yielded hTFF3 polypeptide which is shown in Fig. 3A and B. The Western blot analysis of the purified recombinant hTFF3 was shown in Fig. 4. The final yield of hTFF3 was determined to be about 3–4 mg/l culture broth, which is higher than the typical yield of 2.5 mg/l for GST fusion proteins (Amersham Pharmacia Biotech). However, this yield may vary among different cultures, depending on specific growing conditions and incubation times, which can be further optimized.
3. Results
3.4. Characterization of recombinant hTFF3
3.1. Cloning of hTFF3
3.4.1. Mass spectrometry analysis Mass spectrometry analysis of hTFF3 shows two peaks corresponding to two molecular weights: 13.441 and 6.832 kDa (Fig. 5). The calculated molecular weight of hTFF3 homodimer is 13433 (including two N-terminal amino acid residues Gly and Ser from the vector). This is in a good agreement with the molecular weight determined. The calculated molecular weight of hTFF3 in a monomer form in which Cys-57 contains a free –SH group is 6717.5, which is 114.5 lower than the determined value of 6832. This is probably due to an additional Cys residue bonded to the free Cys-57 or the free Cys-57 has been oxidized.
The amplified PCR fragment from human fetal colon tissue was examined by 2% agarose gel electrophoresis. A band of approximately 200 nucleotides was observed (Fig. 1). The cDNA fragment was successfully sub-cloned into the GST expression vector. The orientation and DNA sequence was verified (results not shown). 3.2. Expression of hTFF3 in E. coli BL21 cells The recombinant hTFF3 was expressed in a soluble form, as a fusion protein with GST. The protein has an
788
M. Fang et al. / Peptides 25 (2004) 785–792
Fig. 2. SDS–PAGE analysis for expressed protein: (1) crude lysate of BL21transformed with pGTF with IPTG induction; (2) crude lysate of BL21 with IPTG induction; (3) crude lysate of BL21 transformed with GEX-4T-1 with IPTG induction; (4) molecular weight marker.
Fig. 3. (A) Gel filtration pattern on Sephacryl-S-100 of eluent from Glutathione Sepharose 4B after digested with thrombin. (A) Thrombin (B) hTFF3. (B) SDS–PAGE analysis for purified hTFF3 (1) molecular weight marker (2) purified TFF3.
According to Fig. 5, the purified recombinant hTFF3 was both in dimer form and monomer form . The presence of only the monomer form on the SDS–PAGE gel (Fig. 3B) was most likely due to the typical reducing process used in the experiment. To confirm this, we run another SDS–PAGE gel which shows less reduced hTFF3.(Fig. 6). According to lane 3, it seemed that the purified recombinant TFF3 mainly existed as dimmers.
Fig. 4. Western-blot of recombinant hTFF3. Lane 1: recombinant hTFF3; lane 2: standard hTFF3.
3.4.2. N-terminal amino acid sequence analysis of purified hTFF3 The N-terminal amino acid sequence analysis revealed that the final product was contaminated by a peptide with an N-terminal sequence of Gly-Ser-Glu-Glu-Tyr-Val-Gly-LeuSer-Ala. Two amino acid residues (Gly and Ser) before the
M. Fang et al. / Peptides 25 (2004) 785–792
789
Fig. 5. Mass spectrometry of recombinant hTFF3. The two peaks corresponding to two molecular weights: 13.441 and 6.832 kDa.
Glu residue were from the cleavage site of pGEX-4T-1 vector. The residues after the Glutamine residue are the same with the natural hTFF3. 3.4.3. The amino acid composition analysis Table 1 shows the amino acid composition of the purified hTFF3, which was in a fairly good agreement with the expected values (including two amino acid residues Gly and Ser from the pGEX-4T-1 vector). 3.5. Biological assay of recombinant hTFF3 in rat gastric injury models 3.5.1. Protection of gastric mucosa against ethanol-induced gastric injury in rats According to Table 2 and Fig. 7, hTFF3 shows a dose dependent protective function in rat gastric injury models. Compared with rats pretreated with saline, protection was observed at a dose of 100 g/kg although the P value is more than 0.05 which means not significant. But the degree of protection increased with increasing doses of peptide from 100 g/kg to 2 mg/kg. The P values were all less than 0.05 when the doses were more than 250 g/kg. The dose dependent protection degree indicates that the purified recombinant hTFF3 has its biological activity.
3.5.2. Protection of pylorus ligation-induced ulceration in rats Similar protective effect can also be observed for hTFF3 in pylorus ligation-induced gastric injury model (Table 3). However, hTFF3 has no effect in regulating gastric acid secretion in this model. In 4 h period, the control rats secreted 36.47 ± 14.5 meq. of HCl, while the 0.25 mg/kg hTFF3 treated group secreted 33.9 ± 7.24 meq. and the 0.5 mg/kg hTFF3 treated group secreted 34.49±9.35 meq. (P > 0.05). 4. Discussion In this study, we described the usage of a GST fusion protein expression system for the production of recombinant hTFF3 in milligram quantity. This is the first description of using GST system for the expression of trefoil peptides. The expression level is fairly high and the purification is relatively simple and efficient as compared to the reported yeast expression system [5,13,14]. The purified recombinant hTFF3 peptide was found both in monomer and in dimer form. We have previously reported that the dimmer form is the biologically active form in vivo [3,8]. The purified recombinant hTFF3 in this study contains two additional amino acid residues in the N-terminal (Gly and Ser), however it has the same biological activity as shown in the gastric mucosa injury models.
790
M. Fang et al. / Peptides 25 (2004) 785–792 Table 1 Amino add composition of recombinant hTFF3 Amino acid
Experimental
Theoretical
Asp Asn Thr Ser Glu Gin Pro Gly Ala Cys Val Met Ile Leu Tyr Phe Lys His Trp Arg
2.86 2.81 1.68 2.71 5.36 1.79 5.94 4.93 3.72 7.56 5.18 0.21 1.35 1.89 1.77 3.13 2.89 1.14 1.34 3.29
3 3 2 3 5 2 6 5 4 7(8) 5 0 1 2 2 3 3 1 1 3
Total
61.55
61(62)
Epithelium is more intact, compared with that of the damaged rat corpus seen in B and C. This indicates that hTFF3 shows a dose dependent protective function in rat gastric injury model. Fig. 6. SDS–PAGE analysis for reduced and non-reduced hTFF3: (1) purified recombinant hTFF3 with reducing process/the monomer form of TFF3 was observed; (2) molecular weight marker; (3) purified recombinant hTFF3 with less reducing agent (2-mercaptoethanol) used in the SDS–PAGE process. Both the dimer form and monomer form of TFF3 were observed. But it seemed that the purified recombinant TFF3 mainly existed as dimmers.
Trefoil factors are known to have protective effect in gastric intestinal tract against indomethacin- and ethanol-induced mucosal injury. Here the protective effect of hTFF3 against ethanol-induced gastric mucosal injury was used to confirm that our purified hTFF3 had the known biological property of gastric mucosal protection. Furthermore, a rat model of pyloric ligation-induced gastric ulcer was also used in our study to further examine the function of hTFF3. Prophylactic oral administration of hTFF3 has a dose-dependent protective effect against pyloric ligation-induced gastric ulcer in rat. Since the gastric acid secretion was not affected by administration of hTFF3, its gastric mucosal protection is unlikely mediated by changing other factors such as acid secretion. The mechanisms of TFF3 for gastrointestinal mucosal protection are still not clear. Some studies have indicated that TFF3 formed a protective layer by combining with mucins in the gastric intestinal tract that can prevent the damage of injurious factors such as ethanol, acids, etc. [6]. TFF3 can also promote the restitution, i.e. the migration of epithelial cell from the normal mucosa to the adjacent damaged mucosa [16]. The distinct trefoil structure formed by three
Table 2 Effect of recombinant hTFF3 pretreatment on the ethanol-induced gastric mucosal injury SI. no.
Treatment and dose (mg/kg)
Animal number
Ulcer index (mean ± S.E.)
1 2 3 4 5 6
Control (saline) hTFF3 (0.1) hTTF3 (0.25) hTTPS (0.5) hTFF3 (1.0) hTFF3 (2.0)
8 8 7 8 8 5
42.36 37.10 30.85 28.63 27.25 18.80
± ± ± ± ± ±
10.38 10.71 9.13 7.94 6.41 5.82
P
>0.05 <0.05 <0.05 <0.05 <0.01
Table 3 Effect of recombinant hTFF3 pretreatment on the induction of gastric ulcers by pylorus ligation SI. no.
Treatment and dose (mg/kg)
Animal number
Ulcer index (mean ± S.E.)
1 2 3
Control (saline) hTFF3 (0.25) hTFF3 (0.50)
7 6 7
8.4 ± 2.2 4.5 ± 2.3 3.1 ± 0.9
P
<0.01 <0.01
intra-chain disulfide bonds between six highly conserved cysteine residues is thought to contribute its remarkable resistant to acid and protease digestion of the gastrointestinal tract and therefore remains biologically active when it is administered orally.
M. Fang et al. / Peptides 25 (2004) 785–792
791
Fig. 7. Gastroprotection by recombinant hTFF3 microscopic pathology. Fasted rats were treated intragastrically with saline or 0.25 mg/kg TFF3 or 0.5 mg/kg TFF3. One hour later, gastric mucosal injury was induced by intragastric administration of absolute ethanol (1.0 ml/rat). Rats were killed one hour later. (A) Normal rat gastric epithelium (100×). (B) Rat gastric epithelium seriously injured by ethanol (100×). Note that the gastric corpus fovcolar epithelium is seriously damaged, compared with that of the undamaged rat corpus seen in A. (C) Rat gastric epithelium treated with 0.25 mg/kg recombinant hTFF3 before ethanol administration (100×). Note that the gastric corpus foveolar epithelium is intact, compared with that of the damaged rat corpus seen in B. (D) Rat gastric epithelium treated with 0.50 mg/kg recombinant hTFF3 before ethanol administration (100×). Note that the gastric corpus foveolar.
Acknowledgments We thank Prof. Fangwen Dishen for providing human fetal colon cDNA, Yunpeng Su for providing rabbit anti-hTFF3 poly-clonal antibody, Xiuyun Dong (The Third Clinical Hospital of Peking University) for assistance with the animal models, Yiyou Chen, Tiegang Han and Jing Lin for the helpful comments on the manuscript. References [1] Babyatsky MW, DeBeaumont M, Thim L, et al. Oral trefoil peptides protect against ethanol-induced and indomethacin-induced gastric injury in rats. Gastroenterology 1996;110:489–97. [2] Chinery R, Bates PA, Amitabha DE, et al. Characterization of the single copy trefoil peptides intestinal trefoil factor and pS2 and their ability to form covalent dimmers. FEBS Lett 1995;357(1):50–4. [3] Hauser F, Poulsom R, Chinery R, et al. hP1.B, a human P-domain peptide homologous with rat intestinal trefoil factor is expressed also in the ulcer-associated cell lineage and the uterus. Proc Natl Acad Sci USA 1993;90(15):6961–5. [4] Jorgensen KH, Thim L, Jacobsen HE. Preparation and characterization of PSP, a new polypeptide from porcine pancreas. Diabetalogia 1981;21:288–94.
[5] Kannan R, Tomasetto C, Staub A, Bossenmeyer-Pourié C, Thim L, Nielsen PF, et al. Human pS2/Trefoil Factor 1: production and characterization in Pichia pastoris. Protein Exp Purif 2001;21(1):92– 8. [6] Kindon H, Pothoulakis C, Thim L, et al. Trefoil peptide protection of intestinal epithelial barrier function: cooperative interaction with mucin glycoprotein. Gastroenterology 1995;109(2):516–23. [7] Kou RQ, Wang W, Li LY, Ru BG. Trefoil factor family. Prog Biochem Biophys 2000;27(4):367–72. [8] Kou RQ, Wang W, Li LY, Ru BG. Construction and bioactivity of Cys57 mutant of human intestinal trefoil factor. Chin J Biochem Mol Biol 2000;16(5):606–11. [9] Rastogi L, Patnaik GK, Dikshit M. Free radicals and antioxidant status following pylorus ligation induced gastric mucosal injury in rats. Pharmacol Res 1998;38(2):125–32. [10] Sands BE, Podolsky DK. The trefoil peptide family. Annu Rev Physiol 1996;58:253–73. [11] Shay M, Kamarov SA, Fels D, Meranze D, Gruen-stein H, Siplet H. A simple method for the uniform production of gastric ulceration in the rat. Gastroenterology 1945;5:43–61. [12] Thim L. A new family of growth factor-like peptides: trefoil disulphide loop structures as a common feature in breast cancer associated peptide(pS2), pancreatic spasmolytic polypeptide(pSP) and frog skin peptides (spasmolysins). FEBS Lett 1989;250:85– 90.
792
M. Fang et al. / Peptides 25 (2004) 785–792
[13] Thim L, Wödike HF, Nielsen PF, et al. Characterization of human and rat intestinal trefoil factor produced in yeast. Biochemistry 1995;34(14):4757–64. [14] Thim L, Kjeld N, Fanny N, et al. Purification and characterization of the trefoil peptide human spasmolytic polypeptide (hSP) produced in yeast. FEBS Lett 1993;318:345–52.
[15] Williams GR, Wright NA. Trefoil factor family domain peptides. Virchows Archiv Int J Pathol 1997;431(5):299– 304. [16] Wright NA, Hoffmann W, Otto WR, et al. Rolling in the clover: trefoil factor family (TFF)-domain peptides, cell migration and cancer. FEBS Lett 1997;408(2):121–3.