Betacellulin-4, a Novel Differentiation Factor for Pancreatic -Cells, Ameliorates Glucose Intolerance in Streptozotocin-Treated Rats
http://www.100md.com
《内分泌学杂志》
Institute for Molecular and Cellular Regulation (T.O., Y.Y., I.K.), Gunma University, Maebashi 371-8512, Japan
Third Department of Medicine (T.O., Y.Y., Y.T.), National Defense Medical College, Tokorozawa 359-8513, Japan
GroPep Ltd. (A.J.D.), Adelaide SA 5000, Australia
Department of Bioscience and Biotechnology (M.S.), Graduate School of Natural Science and Technology, Okayama University, Okayama 700-8530, Japan
Abstract
We previously described a novel alternatively spliced mRNA transcript of the betacellulin (BTC) gene. This splice isoform, termed BTC-4, lacks the C-loop of the epidermal growth factor motif and the transmembrane domain as a result of exon 4 ‘skipping’. In this study, we expressed BTC-4 recombinantly to explore its biological function. When BTC-4 was expressed in COS-7 cells, it was secreted largely into the culture medium, in contrast to BTC. Unlike BTC, highly purified recombinant BTC-4 produced in Escherichia coli failed to bind or induce tyrosine phosphorylation of either ErbB1 or ErbB4, nor did it antagonize the binding of BTC to these receptors. Consistent with this, BTC-4 failed to stimulate DNA synthesis in Balb/c 3T3 and INS-1 cells. However, BTC-4 induced differentiation of pancreatic -cells; BTC-4 converted AR42J cells to insulin-producing cells. When recombinant BTC-4 was administered to streptozotocin-treated neonatal rats, it reduced the plasma glucose concentration and improved glucose tolerance. Importantly, BTC-4 significantly increased the insulin content, the -cell mass, and the numbers of islet-like cell clusters and PDX-1-positive ductal cells. Thus, BTC-4 is a secreted protein that stimulates differentiation of -cells in vitro and in vivo in an apparent ErbB1- and ErbB4-independent manner. The mechanism by which BTC-4 exerts this action on -cells remains to be defined but presumably involves an, as yet, unidentified unique receptor.
Introduction
MEMBERS OF THE epidermal growth factor (EGF) family are characterized by a high degree of sequence similarity, particularly with respect to a common six-cysteine 36- to 40-amino-acid residue EGF-motif with a spacing of CX7CX4–5CX10–13CX1CX8C that forms three intramolecular disulfide bonds (C1-C3, C2-C4, and C5-C6) and a characteristic three-loop structure (1). The EGF-motif of growth factor peptides belonging to this family all appear to be encoded by two exons with a precisely located intervening intron, corresponding to the border separating the first two disulfide loops (A loop, C1-C3; B loop, C2-C4) from the third loop (C loop, C5-C6) (2). A common feature of these molecules is that they are synthesized as larger transmembrane precursors that are proteolytically cleaved to release the soluble biologically active form of the growth factor. The consensus EGF-motif is crucial for binding to and activating members of the ErbB receptor tyrosine kinase family (1, 2). Four receptors belonging to this family have been identified (ErbB1/EGFR, ErbB2/HER2/Neu, ErbB3, and ErbB4). Ligand binding to ErbB receptors induces receptor homo- or heterodimerization, autophosphorylation, and subsequent activation of downstream signaling pathways, resulting in diverse physiological processes, including cell proliferation, differentiation, migration, and survival (3). Ligands belonging to this family include EGF; TGF-; heparin-binding EGF; epiregulin; amphiregulin; the neuregulin subfamily, which includes the products of four genes (NRG1–4); and betacellulin (BTC). BTC binds and activates ErbB1 and ErbB4 homodimers and is further characterized by its ability to activate, albeit weakly, the highly oncogenic ErbB2/ErbB3 complex (4, 5). BTC was originally isolated as a growth factor synthesized in an insulinoma cell line (6). BTC is expressed abundantly in the intestine and pancreas (7). In the pancreas, BTC is expressed in non--cells of the islet in adults and is expressed in endocrine precursor cells in the fetus (8). This suggests the possibility that BTC regulates the growth and/or differentiation of pancreatic -cells and their precursors. In accordance with this notion, BTC is able to induce differentiation of pancreatic AR42J cells and convert them into insulin-producing cells (9) and to convert glucagon-producing -cells into -cells (10). Furthermore, in vivo studies have shown that BTC promotes regeneration of pancreatic -cells in various animal models of experimental diabetes (11, 12, 13, 14). Collectively, BTC is a growth or differentiation factor for pancreatic -cells.
We recently identified an alternately spliced mRNA transcript encoding a novel isoform of human BTC (15). This isoform, originally termed BTC- but now referred to as BTC-4, lacks 147 bp encoding exon 4 of the BTC gene, leading to the generation of an mRNA encoding an unusual BTC precursor in which the C-loop of the EGF domain and the transmembrane domain are deleted while the remainder of the mature molecule, including loops A and B and the ‘hinge’ valine, is fused in frame to the truncated C-terminal cytoplasmic tail (Fig. 1). Retention of the hydrophobic signal sequence and the absence of the transmembrane domain suggests that BTC-4 may be a secreted protein. Interestingly, a similar isoform lacking the third disulfide loop of the EGF domain has been reported for heparin-binding EGF (16). In this case, a 94-bp insertion between exons III and IV causes a frameshift and premature termination generating a protein that retains the signal peptide, proregion, heparin-binding domain and the first two conserved disuphide loops of the EGF motif, while a short nine-amino-acid tail replaces the third disulfide loop, transmembrane, and cytoplasmic domains. In this report, we have produced BTC-4 recombinantly to characterize it’s biological activity and, in particular, to determine whether it may act as a naturally occurring BTC antagonist. Our results indicate that whereas BTC-4 does not bind to either ErbB1, ErbB4, or the ErbB2–3 heterodimeric complex and shows no evidence of acting as an BTC antagonist, BTC-4 is biologically active and acts as an inducer of differentiation of pancreatic -cells in vitro and in vivo, possibly through a novel cell surface receptor.
Materials and Methods
Reagents
Recombinant human activin A was a generous gift from Dr. Y. Eto of Central Research Laboratory, Ajinomoto Inc. (Kawasaki, Japan). Streptozotocin (STZ) was purchased from Wako Pure Chemicals (Osaka, Japan). Mouse fibroblast Balb/c 3T3 cells were from the American Type Culture Collection (Manassas, VA). Human lung fibroblast AG2804 cells and CHO cells transfected with either ErbB 4 or ErbB2/3 were kindly provided by Dr. J. Gunn (Texas A&M University, College Station, TX) and Prof. Y. Yarden (Weizmann Institute, Rehovet, Israel), respectively. INS-1 cells were generously provided by Prof. C. Wollheim (University of Geneva, Geneva, Switzerland). Anti-FLAG M2 antibody was from Sigma (St. Louis, MO). Anti-ErbB1 (1005), anti-ErbB4 (C-18), and antiphosphotyrosine (PY20) antibodies were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) and horseradish peroxidase (HRP)-conjugated rabbit antisheep antibody from Zymed Laboratories (San Francisco, CA). HRP-conjugated sheep antimouse antibody was from Silenus Laboratories (Hawthorn, Australia). Lipofectamine 2000, Enterokinase, Escherichia coli TOP10 cells, Optimem-1, 10–20% Tris Tricine gels, and pcDNA3.1 were from Invitrogen (Carlsbad, CA). West Pico SuperSignal substrate was from Pierce Biotechnology, Inc. (Rockford, IL) pET32, BL21trxB (DE3) cells, and BugBuster Protein Extraction Reagent were from Novagen; Ni-NTA agarose was from QIAGEN (Valencia, CA). Recombinant human BTC and goat antihuman BTC ectodomain (Asp32-Tyr111) antibody were from R&D Systems, Inc. (Minneapolis, MN). Complete protease inhibitors and protein G-Sepharose were from Roche Molecular Biochemicals (Basel, Switzerland).
Cloning of BTC and BTC-4 FLAG-tagged constructs
Full-length BTC (amino acids Met1-Ala178) and BTC-4 (amino acids Met1-Ala129) were cloned into pcDNA3.1 to generate expression vectors for the mammalian production of BTC and BTC-4 as follows. The open-reading frames (ORFs) of BTC and BTC-4 were excised from pBlue-BTC-4 and pBlue-BTC (15) with ApaI and BamHI and recloned into ApaI/BamHI digested pcDNA3.1. To generate ‘Flag-tagged’ pcDNA3.1-BTC-4 and pcDNA3.1-BTC constructs, in which the Flag epitope DYKDDDDK is inserted between amino acids S35–T36, pcDNA3.1-BTC-4 or pcDNA3.1-BTC was used as a template in PCR using the primers; 5'-CTCGGGAATTCCGACTACAAGGACGACGATGACAAGACCAGAAGTCCTGAA-3' (sense) (underlined nucleotides correspond to an EcoRI restriction site; double underlined nucleotides encode the FLAG epitope tag) and 5'-CTCCTGCAGTTAAGCAATATTTGTCTCTTC-3' (antisense) (underlined nucleotides correspond to a PstI restriction site). The resulting PCR products were purified and digested with EcoRI and PstI and cloned into EcoRI/Pst1 digested pBlue-BTC or pBlue-BTC-4. Subsequently, the resultant Flag-tagged pBlue-BTC and pBlue-BTC-4 constructs were digested with ApaI/BamH1 and cloned into ApaI/BamH1 digested pcDNA3.1 to generate pcDNA3.1-FLAG-BTC-4 and pcDNA3.1-FLAG-BTC.
Expression and analysis of BTC and BTC-4 FLAG-tagged constructs in COS-7 cells
For transient expression of pcDNA3.1-FLAG-BTC-4 and pcDNA3.1-FLAG-BTC, COS-7 cells were plated into 12-well plates, at 2 x 105 cells/well, in DMEM/10% fetal bovine serum. After overnight incubation, cells were transfected with 2 μg construct DNA using Lipofectamine 2000 and Optimem-1 media according to the manufacturer’s instructions. Twelve hours later, the cells were washed twice and replenished with fresh Optimem-1 media. Seventy-two hours post transfection, culture medium (CM) was collected and cell lysates prepared. CM was clarified by centrifugation, and the presence of BTC or BTC-4 in the media was analyzed by ELISA or Western blotting using an anti-FLAG-M2 antibody (Sigma). BTC or BTC-4 present in the cell lysate was analyzed by Western blotting with the anti-FLAG-M2 antibody. Cell lysates were prepared by washing the cells twice with PBS after the removal of CM and then lysing the cells directly in SDS-PAGE sample buffer and heating to 95 C for 5 min. For ELISA analysis of the CM, 90 μl of CM was mixed with 10 μl of 10x coating buffer (0.15 mol/liter Na2CO3, 0.35 mol/liter NaHCO3, pH 9.3) and loaded into 96-well immunosorbent plates. Plates were coated overnight at 4 C, then blocked in 2% BSA in PBS/0.1% Tween (PBS-T). Plates were washed four times with PBS-T and then incubated for 1 h at 37 C with anti-FLAG antibody diluted in PBS-T (2.5 μg/ml). After incubation, plates were washed as above and incubated with HRP-conjugated sheep antimouse antibody (1:2000) diluted in PBS-T. Plates were then washed four times with PBS-T and developed with o-phenylamine diamine substrate and stopped with 2 mol/liter H2SO4. Absorbance was read at 490 nm. Results are expressed as the mean ± SD of triplicate determinations. To analyze the presence of BTC or BTC-4 on the cell surface by ELISA, CM was removed and the cells washed three times with PBS. The cells were then fixed with 4% paraformaldehyde and stained with anti-FLAG M2 antibody as described above. For Western blot analysis with the anti-FLAG-M2 antibody, CM or cell lysates were resolved by SDS-PAGE (10–20% Tris-Tricine gels). Proteins were then transferred to nitrocellulose (Hybond-C extra; Amersham Pharmacia Biotech, Little Chalfont, UK) and membranes probed with mouse anti-FLAG-M2 antibody (2.5 μg/ml) and then HRP-conjugated sheep antimouse antibody (1:10,000). HRP-labeled proteins were visualized using SuperSignal West Dura Extended Duration Substrate.
Expression and purification of BTC and BTC-4
BTC-4 [constituting amino acids Asp32-Ala129 (BTC-432–129)] (Fig. 1) was expressed as a thiroredoxin fusion protein in pET32a. The ORF sequence of BTC-432–129 was amplified by PCR using the primer set, 5'-CGTCCATGGCTGATGGGAATTCCACCAGAAGT-3' (sense) and 5'-CGTCTCGAGTCATTAAGCAATATTTGTCTCTTC-3' and pBlue-BTC-4 (15) as template. NcoI and XhoI recognition sites were attached to the sense and antisense primers, respectively (underlined). BTC (constituting amino acids 32–111; BTC32–111) was also expressed using the pET system as a positive control, as described previously (7), or produced as a thioredoxin fusion protein with the plasmid pET32a. For this construct, the ORF sequence of BTC32–111 was amplified by PCR using the primer set, 5'-CGTCCATGGCTGATGGGAATTCCACCAGAAGT-3' (sense) and 5'-CGTCTCGAGTCAGTAAAACAAGTCAACTGT-3' (antisense) and pBlue-BTC (see Ref.18) as template. NcoI and XhoI recognition sites were also attached to the sense and antisense primers, respectively (underlined). The PCR products were digested with NcoI/XhoI and cloned into NcoI/XhoI digested pET32a vector as above. The resultant plasmids, pETBTC-432–129 and pETBTC32–111, were maintained in Escherichia coli JM109 and, for expression, transformed into Escherichia coli BL21trxB (DE3) cells. Large-scale expression of BTC or BTC-4 in BL21trxB (DE3) cells was carried out as described in the pET system manual (Novagen, Madison, WI). BTC or BTC-4 thioredoxin fusion proteins were purified from clarified cell lysates by Ni-NTA agarose affinity, followed by cleavage with Enterokinase to isolate authentic BTC or BTC-4. Authentic BTC or BTC-4 was separated from the thioredoxin fusion partner by additional Ni-NTA agarose affinity chromatography and reverse-phase HPLC (RP-HPLC). Briefly, frozen cell pellets were thawed on ice and resuspended in 18 ml BugBuster Protein Extraction Reagent containing lysozyme (100 μg/ml) and incubated at room temperature with gentle shaking for 10 min. After incubation, the cell lysate was sonicated (3 x 5-sec bursts) and clarified by centrifugation (20 min, 16,000 rpm, 10 min). The clarified cell lysate was adjusted to 10 mmol/liter imidazole and applied to a Ni-NTA agarose affinity column preequilibrated with 50 mmol/liter NaH2PO4, 0.3 mol/liter NaCl, 10 mmol/liter imidazole (pH 8.0). After column washing with 50 mmol/liter NaH2PO4, 0.3 mol/liter NaCl, 40 mmol/liter imidazole (pH 8.0), BTC or BTC-4 was eluted with 50 mmol/liter NaH2PO4, 0.3 mol/liter NaCl, 250 mmol/liter imidazole (pH 8.0) and subsequently dialyzed overnight (Spectra/Por, 3.5 kDa MWCO; Spectrum Laboratories, Houston, TX) against several changes of Enterokinase cleavage buffer (50 mmol/liter Tris-Cl, 1 mmol/liter CaCl2, 0.1% Tween 20, pH 7.4). To remove the thioredoxin fusion partner from BTC or BTC-4, Enterokinase was added to a final concentration of 0.1 U/20 μg protein and incubated overnight at 37 C with gentle mixing. BTC or BTC-4 was separated from the thioredoxin fusion partner by further Ni-NTA agarose affinity chromatography as described above. In this case, the cleaved thioredoxin fusion partner is captured on the resin and BTC or BTC-4 is collected in the flowthrough fraction flow-through fraction. BTC or BTC-4 present in the flow-through fraction was further purified by RP-HPLC. Briefly, the Ni-NTA agarose flow-through fraction was diluted 1:4 (vol/vol) with 0.1% trifluoroacetic acid (TFA) and applied to a C4 Prep-Pak RP-HPLC column (25 mm x 100 mm; 300, 15 μm; Millipore-Waters, Bedford, MA) at a flow rate of 10 ml/min. The column was washed with 0.1% TFA and eluted with a gradient of 8–80% (vol/vol) acetonitrile over 150 min in the presence of 0.08% TFA at a flow rate of 10 ml/min. Twenty-milliliter fractions were collected, and aliquots of each (50 μl) were analyzed by SDS-PAGE and Western blotting using a goat antihuman BTC ectodomain (Asp32-Tyr111) antibody to identify pure BTC or BTC-4 containing fractions. The purity of all BTC or BTC-4 preparations were analyzed by analytical RP-HPLC, electrospray ionization mass spectrometry, and SDS-PAGE. Analytical RP-HPLC was performed using a Brownlee aquopore C4 RP-HPLC column (2.1 x 100 mm) (Alltech, Deerfield, IL) with a linear gradient of 8–80% (vol/vol) acetonitrile and 0.08% TFA over 40 min at a flow rate of 0.5 ml/min. Alternatively, BTC-432–129 (containing an initiation methionine residue) was cloned into the expression vector pET3b to construct pBO604, and the recombinant protein was solubilized and refolded from inclusion bodies, purified through cation exchange column and gel filtration column chromatography as previously described (7, 17). The authenticity of BTC-432–129 was confirmed by amino-terminal sequence analysis and amino acid composition. All the recombinant proteins prepared in this study were lyophilized and stored at –80 C before use.
ErbB-receptor binding and mitogenic activity of BTC-4
The binding affinity of BTC or BTC-4 to ErbB receptors was determined by measuring the ability of BTC or BTC-4 to competitively displace [125I]-labeled recombinant human BTC from ErbB1 or ErbB4 receptors present on AG2804 fibroblasts or CHO cells stably transfected with ErbB4 (18) as described previously (19). The mitogenic activity of BTC and BTC-4 toward Balb/c 3T3 fibroblasts and INS-1 cells was performed as previously described (19).
Analysis of ErbB-1 and ErbB-4 receptor tyrosine phosphorylation
AG2804 or CHO-ErbB4 cells were grown to confluence in 10-cm dishes and subsequently incubated for 12–14 h in serum-free medium. Cells were then stimulated with 10 nmol/liter BTC, BTC-4, or a combination of both for 10 min at room temperature. After stimulation, cells were washed twice in PBS and then lysed in lysis buffer (0.5 ml) (50 mmol/liter Tris-Cl, pH 7.4; 150 mmol/liter NaCl; 1% deoxycholate; 1% Triton X-100; 0.1% sodium dodecyl sulfate; 5 mmol/liter sodium orthovanadate; 10 mmol/liter sodium fluoride; 1 mmol/liter EGTA; and complete protease inhibitors. Cell lysates were cleared by centrifugation (20 min, 15,000 x g at 4 C) and ErbB1 or ErbB4 immunoprecipitated by incubating the lysate with 1 μg rabbit polyclonal anti-ErbB1 antibody (1005, Santa Cruz) or rabbit polyclonal anti-ErbB4 antibody, respectively, for 2 h at 4 C with gentle shaking. After incubation, protein-G Sepharose was added and the mixture incubated for a further 1 h at 4 C. Immune complexes were collected by centrifugation, washed three times in lysis buffer, and heated (3 min, 95 C) in SDS-PAGE sample buffer. Proteins were separated by SDS-PAGE (10–20% Tricine gels) and transferred to nitrocellulose filters (Hybond C; Amersham). Membranes were probed with antiphosphotyrosine monoclonal antibody (PY20) and then with HRP-conjugated sheep antimouse antibody. HRP-labeled proteins were visualized using enzyme-linked chemiluminescence (Amersham). To confirm equal loading, blots were stripped and reprobed with rabbit polyclonal anti-ErbB1 or rabbit polyclonal anti-ErbB4 antibodies and HRP-conjugated rabbit antisheep antibody.
Measurement of differentiation and apoptosis of AR42J cells
AR42J-B20 cells were cultured in DMEM medium containing 10% fetal calf serum as described previously (9). To assess differentiation into insulin-producing cells, cells were incubated for 48 h with 2 nmol/liter activin A and either 1 nmol/liter BTC or BTC-4. Cells were then fixed and stained with antiinsulin or anti-C peptide antibody as described previously (9, 20), and the number of insulin-positive cells was counted. Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI). Apoptosis was assessed by using a terminal deoxynucleotidyl transferase-mediated deoxyuridine 5-triphosphate nick end labeling (TUNEL) technique (Wako Pure Chemicals.). Changes in the number of viable cells were assessed by using 3-[4,5-dimethylthiazole-2-yl]-2,5,-dipheltetrazodium bromide (MTT) (21). Anti-C-peptide antibody was obtained from Linco Research, Inc. (St. Charles, MS).
Treatment of neonatal rats with streptozotocin
One-day-old Sprague Dawley rats were treated with streptozotocin and tested for blood glucose levels as described previously (14). The animals were included in the study only if the blood glucose concentration was between 200 and 350 mg/dl on the next day (designated as d 0). Animals whose blood glucose was more than 350 mg/dl on d 0 developed severe diabetes, and the blood glucose concentrations in these animals were more than 600 mg/dl on d 1 or 2. Neonatal STZ-treated rats were injected with 3 pmol/g BTC-4, BTC, or saline every day for 5 d starting from d 0. The fasting blood glucose concentration and the body weight were measured daily for the first week and then once a week for up to 8 wk. Two months after the STZ treatment, an ip glucose tolerance test (2 g/kg body weight) was performed after 14 h of fasting. The experimental protocol was approved by the Animal Care Committee of the Gunma University.
Tissue processing and immunohistochemistry
On d 4 and wk 8, the animals were injected ip with 1 ml bromodeoxyuridine labeling reagent per 100 g body weight (cell proliferation kit; Amersham) and decapitated after 3 h. The pancreas was excised, weighed, and divided into two parts. One portion from the splenic segment was fixed with 4% paraformaldehyde/PBS overnight at 4 C and processed for paraffin embedding. Four series from each pancreas were cut, at intervals of 100 μm in neonates and 300 μm in adults, for immunostaining and histochemistry. The second portion was homogenized in cold acid-ethanol, heated for 5 min in 70 C water bath, and centrifuged, and the supernatant was stored at –20 C before assaying for insulin. Insulin was measured by time-resolved immunofluorometric assay as described previously (9). Immunohistochemistry was performed as described previously (12, 13). Quantitation of the -cell mass was performed on insulin-stained sections using image analysis software (National Institutes of Health image) by means of an AX70 Epifluorescence microscope (Olympus Corp., Tokyo, Japan) equipped with a PXL 1400 cooled-charge-coupled device camera system (Photometrics, Tucson, AZ) operated with IP Lab Spectrum software (Signal Analysis, Vienna, VA). At least 40 random fields (magnification, x200) from one section (three sections from different series per block) were measured for the area of insulin-positive cells. The -cell mass was calculated as described elsewhere (12, 13). To quantify -cell neogenesis, the number of islet-like cell clusters (ICCs) (less than five cells across) was determined. The number of ICCs and islets was counted in the section stained with the antiinsulin antibody, and the area in these sections was measured. At least five sections were analyzed per animal. Results were expressed as means ± SE. For comparisons between two groups, the unpaired t test was used.
Results
BTC-4 is a secreted protein
Like other members of the EGF family, BTC is synthesized as a transmembrane-anchored precursor protein (pro-BTC), which can be proteolytically cleaved to yield soluble BTC containing the EGF-motif (Asp32-Tyr111). Retention of the hydrophobic signal peptide and the absence of the transmembrane domain suggest that BTC-4 may be a secreted protein. To test this, the complete ORF of either BTC or BTC-4 (additionally engineered to incorporate a FLAG epitope between amino acids Ser35 and Thr36 for subsequent detection) was cloned into the expression vector pcDNA3.1 and secretion monitored after transient transfection in COS-7 cells. Seventy-two hours post transfection, BTC-4-FLAG was clearly detected in the culture supernatant by either Western-blot (14 kDa) or enzyme immunoassay, with very little present in the cell lysate or on the cell surface (Fig. 2, A and B). In contrast, BTC-FLAG was predominantly present in the cell lysate (20 kDa) and on the cell surface (Fig. 2, A and B).
Recombinant production of BTC-4
To characterize the biological activity of BTC-4, we cloned the cDNA (Asp32-Ala129; without the signal peptide) into the pET-32a expression vector and expressed the protein as a thioredoxin fusion in Escherichia coli BL21trxB (DE3) (Fig. 3A). This strain is deficient in thioredoxin reductase, allowing for disulfide bond formation in the cytoplasm. The purified thioredoxin fusion protein was cleaved with enterokinase to release BTC-4, which was purified to homogeneity by RP-HPLC (Fig. 3B). As a control, we also generated BTC (Asp32-Tyr111) with this system and purified it in the same manner. The purity of both the BTC and BTC-4 preparations was further confirmed by N-terminal sequence analysis (five cycles), which gave the expected N-terminal sequence with an approximate purity of more than 95% (data not shown). The molecular mass of recombinant BTC and BTC-4 determined by electrospray ionization mass spectrometry was 9,211.35 ± 0.29 Da and 11,450.26 ± 0.23 Da, respectively, which is consistent with the calculated theoretical masses of 9,249 and 11,452 Da (data not shown). After SDS-PAGE and silver staining, a single band at approximately 9 kDa and 11.5 kDa was obtained for BTC and BTC-4 detected under reducing or nonreducing conditions (data not shown).
BTC-4 does not bind or activate either ErbB1 or ErbB4
BTC stimulated the proliferation of Balb/c 3T3 fibroblasts in a dose-dependent fashion (Fig. 4A). In contrast, BTC-4 did not stimulate cell proliferation, even at concentrations as high as 100 nmol/liter (Fig. 4A), or antagonize the effects of BTC in this cell line (Fig. 4B). Consistent with these results, BTC, but not BTC-4, could displace [125I]-BTC from ErbB1 receptors present on human lung fibroblasts (AG2804) (Fig. 4, C and D) and induce ErbB1 receptor autophosphorylation (Fig. 5). Similarly, BTC, but not BTC-4, competed for ErbB4 receptor binding (Fig. 4, E and F) and induced ErbB4 receptor tyrosine phosphorylation (Fig. 5). These results indicate that, unlike BTC, BTC-4 displays no affinity toward either ErbB1 or ErbB4. We also examined the effect of BTC and BTC-4 on DNA synthesis in INS-1 cells. As shown in Fig. 4G, BTC stimulated DNA synthesis in INS-1 cells, whereas BTC-4 did not.
BTC-4 induces the differentiation of pancreatic -cells
BTC is known to stimulate the differentiation of pancreatic -cells (9, 10, 11, 12, 13, 14), and there is some evidence to suggest that this may occur through a unique non-ErbB cell surface receptor (22). To examine the effect of BTC-4 on the differentiation of pancreatic -cells, we used the model cell line AR42J-B20; a subclone of AR42J, an amylase-secreting pancreatic tumor cell line. In these cells, BTC acts coordinately with activin A and converts them to insulin-secreting cells (9). Activin A also induces apoptosis in these cells and, in the absence of a survival factor, many of the activin-treated cells die by apoptosis after the conversion to pancreatic polypeptide (PP)-producing cells (23). BTC exerts two effects in AR42J-B20 cells: BTC inhibits apoptosis induced by activin A; and, second, converts them to insulin-producing cells (9, 23). As shown in Fig. 6A, AR42J-B20 cells differentiated into insulin-producing cells by a combination of activin A and BTC. Quantitatively, 78 ± 6.5% of the cells (mean ± SE; n = 4) became insulin-positive after 48 h. Cells treated with activin A and BTC exhibited extended processes and expressed immunoreactive insulin (Fig. 6C). Cells treated with a combination of activin A and BTC-4 also became insulin-positive; 72 ± 4.5% (n = 4) of the cells became insulin-positive after 48 h (Fig. 6B). It should be noted that cells treated with activin A and BTC-4 were also positive for C-peptide (Fig. 6G). In contrast to cells treated with activin A and BTC, the morphology of cells treated with activin A and BTC-4 was quite different. Some of the insulin-positive cells displayed extended processes, but most of them remained circular in appearance. The nuclei of these cells were condensed and, in some instances, were either fragmented or absent (Fig. 6, B and D), characteristic of cell death by apoptosis. Consistent with this notion, many of the cells treated with activin A and BTC-4 were TUNEL-positive (Fig. 6F), compared with only a small fraction of cells treated with activin A and BTC (Fig. 6E). To further confirm the effect of BTC-4 on the survival of AR42J-B20 cells, we measured the changes in the number of viable cells after treatment with activin A and either BTC or BTC-4. After 48 h, the number of viable cells treated with activin A and BTC was 78.8 ± 4.2% of the number of cells seeded. In contrast, the number of viable cells treated with activin A and BTC-4 was significantly lower (21.8 ± 3.2%) (P < 0.005). Taken together, BTC-4 was as effective as BTC in stimulating the differentiation of AR42J-B20 cells; however, BTC-4 was much less potent in promoting survival of these cells. Note that BTC-4 alone did not induce differentiation of AR42J-B20 cells. Also, BTC-4 did not affect survival of these cells. When both BTC and BTC-4 were administered together with activin A, there was a small additivity in the actions of BTC and BTC-4, and 88 ± 4.5% of the cells became insulin-positive.
BTC-4 administration to STZ-treated rats reduces the plasma glucose concentration and improves glucose tolerance
To investigate further the -cell differentiating activity of BTC-4, we administered BTC-4 to neonatal STZ-treated rats. In STZ-treated neonatal rats, the plasma glucose concentration increased markedly and peaked on d 2 (>400 mg/dl) (Fig. 7A), thereafter declining gradually although still significantly higher compared with control rats after 2 months (Table 1). In STZ-treated rats administered BTC-4, the plasma glucose concentration was significantly lower on d 1, and the peak value was markedly reduced (Fig. 7A). The plasma glucose concentration was also reduced thereafter, until 2 months (Table 1). In contrast, BTC was less potent in improving hyperglycemia (Fig. 8A and Table 1). We also administered BTC and BTC-4 simultaneously, but the effect was not different from that obtained by BTC-4 alone (data not shown). At 2 months of age, ip glucose tolerance tests were performed. In STZ-treated rats, the glucose response to ip glucose-loading was markedly impaired; whereas in STZ-treated rats administered BTC-4, the glucose response was significantly improved (Fig. 7B). BTC-4 also increased the insulin response; but, compared with normal rats, the insulin response to ip glucose-loading was still delayed. In BTC-treated rats, glucose tolerance was improved, but the effect of BTC was less than that of BTC-4 (Fig. 7B). Moreover, the insulin content and the -cell mass were significantly increased in BTC-4-treated rats (Table 1), and histological analysis of pancreatic tissue on d 4 indicated that BTC-4 significantly increased the numbers of PDX-1-positive ductal cells and ICCs (Fig. 8). We also examined the effect of BTC and BTC-4 on replication of insulin-positive cells. On d 4, the numbers of bromodeoxyuridine/insulin double-positive cells in saline-, BTC-, and BTC-4-treated rats were 5.4 ± 0.6, 8.3 ± 0.8, and 7.4 ± 1.0%, respectively (n = 5). The effects of BTC-4 and BTC were significant (P < 0.05) compared with saline-treated control. We also counted the number of apoptotic cells. On d 4, the number of TUNEL-positive cells was very low (less than 1 x 10–6/μm2).
Discussion
The present study was conducted to investigate the biological activity of the BTC splice variant, BTC-4. BTC-4 is unique in that it lacks the C-loop of the EGF motif and the transmembrane domain yet retains the signal peptide (15), suggesting that the protein produced from this transcript is a secreted protein. In contrast, BTC is synthesized as a transmembrane precursor, which can undergo ‘ectodomain shedding’ to release the soluble biologically active mature molecule. BTC-4 was found predominantly in the culture supernatant after expression in COS-7 cells as an approximately 14-kDa protein. In contrast, little, if any, BTC was found in the culture supernatant, the majority being present in the cell lysate (Fig. 2A) and on the cell surface (Fig. 2B). Thus, although BTC-4 is a protein product of the splicing variant of the BTC gene, the production of BTC-4 and BTC is differentially regulated. The three disulfide loops and certain key residues in the two -sheet structural regions of EGF ligands (for example, Gly49, Gly70, Arg72, and Leu78; Leu47 and Leu48 in EGF and TGF-, respectively) are crucial for ErbB receptor binding (1, 2, 24). Because BTC-4 lacks the C-loop of the EGF motif encompassing Gly70, Arg72, and Leu78 (see Fig. 1), we were interested to determine whether BTC-4 was able to bind and activate ErbB-1 and ErbB-4, the principal BTC ErbB receptors (4, 5). BTC-4 was unable to bind or induce tyrosine phosphorylation of either ErbB-1 or ErbB-4 (Figs. 4 and 5), indicating that BTC-4 is not a ligand for either ErbB1 or ErbB4. Moreover, BTC-4 did not stimulate DNA synthesis in Balb/c 3T3 cells, which express ErbB-1 (Fig. 4), or induce phosphorylation of ErbB2/3 heterodimers (data not shown). In addition to stimulating cell proliferation, principally through ErbB1 and the activation of the Grb2-Sos-Ras-Raf-Mek1/2-Erk1/2 pathway, BTC also may act as a survival factor for various cell types through ErbB receptor-induced activation of the PI3K-Akt pathway (4, 5, 25). Consistent with this, we found that BTC, but not BTC-4, was able to inhibit apoptosis of AR42J-B20 cells induced by activin A (Fig. 6, E and F). These data support the concept that BTC-4 is not a ligand for the ErbB receptors and, hence, is unable to promote cell growth or survival. Nevertheless, both BTC and BTC-4 converted AR42J cells to insulin-producing cells (Fig. 6); and moreover, BTC-4 significantly increased the insulin content, -cell mass, the number of PDX-1-positive ductal cells, and ICCs in neonatal STZ-treated rats (Table 1). Consequently, BTC-4 appears to act as a -cell differentiation factor in vitro and in vivo, and it exerts this differentiation-inducing activity by acting through a unique cell surface receptor independent of either ErbB1 or ErbB4. We have previously postulated that BTC induces the differentiation of AR42J-B20 cells into insulin-secreting cells through a unique non-ErbB receptor of approximately 190 kDa (22). The data presented here strengthen this notion. Moreover, the fact that both BTC-4 and BTC display this differentiation-inducing activity suggests that the structural determinants for binding this unique receptor lie within the region Asp32-Val94. Analysis of C-terminal truncation derivatives of BTC, which display little affinity to ErbB1, are only weakly mitogenic for AR42J cells, yet retain the ability to induce the differentiation of these cells into insulin-secreting cells (26), is further consistent with this hypothesis. Taken together, these results suggest that the differentiation-inducing activity of both BTC and BTC-4 may be exerted through this unique receptor. The identification and characterization of this receptor is currently under investigation in our laboratory.
Previous studies have shown that BTC is able to increase the -cell mass in various experimental rodent models of diabetes, indicating that it may represent a useful therapeutic approach for the treatment of the human condition (11, 12, 13, 14). From a clinical point of view, administration of recombinant therapeutic growth factors to treat chronic disorders is hampered by the potential side effect of tumorigenesis, as a result of promoting cell proliferation. In this regard, because we have found that BTC-4 reproduces the differentiation activity of BTC without promoting cell proliferation, it may represent a more attractive therapeutic alternative for the treatment of diabetes. The observed in vitro apoptosis in cells treated with activin A and BTC-4 may not necessarily occur in vivo because -cell progenitors would be bathed in a milieu of growth factors that may act to reduce apoptosis by virtue of their prosurvival properties. Thus, BTC-4 may be useful to treat certain types of diabetes caused by impaired -cell neogenesis.
Acknowledgments
We thank Mayumi Odagiri for secretarial assistance during the preparation of the manuscript, and Ilka Priebe and Helen Webb for excellent technical assistance.
Footnotes
This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan; a grant from the Project of Realization of Regeneration Medicine; and the Australian Government through the Cooperative Research Centres Programe.
1 This work is dedicated to the memory of Takeki Ogata, who sadly passed away during the completion of this manuscript.
2 T.O. and A.J.D. contributed equally to this paper.
Abbreviations: BTC, Betacellulin; CM, culture medium; DAPI, 4',6-diamidino-2-phenylindole; EGF, epidermal growth factor; HRP, horseradish peroxidase; ICC, islet-like cell cluster; ORF, open-reading frame; RP-HPLC, reverse-phase HPLC; TFA, trifluoroacetic acid; TUNEL, terminal deoxynucleotidyl transferase-mediated deoxyuridine 5-triphosphate nick end labeling.
References
Van Zoelen EJJ, Stortelers C, Lenferink AEG, Van de Poll MLM 2000 The EGF domain: requirements for binding to receptors of the ErbB family. Vitam Horm 59:99–131
Harris RC, Chung E, Coffey RJ 2003 EGF receptor ligands. Exp Cell Res 284:2–13
van der Greer P, Hunter T, Lindberg RA 1994 Receptor-tyrosine kinases and their signal transduction pathways. Annu Rev Cell Biol 10:251–337
Alimandi M, Wang LM, Bottaro D, Lee CC, Kuo A, Frankel M, Fedi P, Tang C, Lippman M, Pierce JH 1997 Epidermal growth factor and betacellulin mediate signal transduction through co-expressed ErbB2 and ErbB3 receptors. EMBO J 16:5608–5617
Pinkas-Kramarski R, Leferink AEG, Bacus SS, Lyass L, an de Poll MLM, Klapper LN, Tzahar E, Sela M, van Zoelen EJJ, Yarden Y 1998 The oncogenic ErbB2-ErbB-3 heterodimer is a surrogate receptor of the epidermal growth factor and betacellulin. Oncogene 16:1249–1258
Shing Y, Christofori G, Hanahan D, Ono Y, Sasada R, Igarashi K, Folkman J 1993 Betacellulin: a mitogen from pancreatic -cell tumors. Science 259:1604–1607
Seno M, Tada H, Kosaka M, Sasada R, Igarashi K, Shing Y, Folkman J, Ueda M, Yamada H 1996 Human betacellulin, a member of the EGF family, dominantly expressed in pancreas and small intestine, is fully active in a monomeric form. Growth Factors 13:181–191
Miyagawa J, Hanafusa T, Sasada R, Yamamoto K, Igarashi K, Yamamoto K, Seno M, Tada R, Nammo T, Li M, Yamagata K, Nakajima H, Namba M, Kuwajima M, Matsuzawa Y 1999 Immunohistochemical localization of pancreas and islet tumor cells. Endocr J 46:755–764
Mashima H, Ohnishi H, Wakabayashi K, Miyagawa J, Hanafusa T, Seno M, Yamada H, Kojima I 1996 Betacellulin and activin A coordinately convert amylase-secreting pancreatic AR42J cells into insulin-secreting cells. J Clin Invest 97:1647–1654
Watada H, Kajimoto Y, Miyagawa J, Hanafusa T, Hamaguchi K, Matsuoka T, Yamamoto K, Matsuzawa Y, Kawamori R, Yamasaki Y 1996 PDX-1 induces insulin and glucokinase gene expression in TC clone 6 cells in the presence of betacellulin. Diabetes 45:1826–1831
Yamamoto K, Miyagawa J, Waguri M, Sasada R, Igarashi K, Li M, Nammo T, Moriwaki T, Imagawa A, Yamagata K, Nakajima H, Namba M, Tochino Y, Hanafusa T, Matsuzawa Y 2000 Recombinant betacellulin promotes neogenesis of -cells and ameliorates glucose intolerance in mice with diabetes induced by selective alloxan perfusion. Diabetes 49:2021–2027
Li L, Seno M, Yamada H, Kojima I 2001 Promotion of -cell regeneration by betacellulin in ninety percent pancreatectomized rats. Endocrinology 142:5379–5385
Li L, Seno M, Yamada H, Kojima I 2003 Betacellulin improves glucose metabolism by promoting conversion of intraislet precursor cells to -cells in streptozotocin-treated mice. Am J Physiol 285:E577–E583
Li L, Yi Z, Seno M, Kojima I 2004 Activin A and betacellulin: effect on -cell regeneration in neonatal streptozotocin-treated rats. Diabetes 53:608–615
Dunbar A, Goddard C 2000 Identification of an alternatively spliced mRNA branscript of human betacellulin lacking the C-loop of the EGF motif and the transmembrane domain. Growth Factors 18:169–176
Loukianov E, Loukinanova T, Wiedlocha A, Olnes S 1997 Expression of mRNA for a short form of heparin-binding EGF-like growth factor. Gene 195:81–86
Maeda T, Kitazoe M, Tada H, de Llorens R, Salomon DS, Ueda M, Yamada H, Seno M 2002 Growth inhibition of mammalian cells by eosinophil cationic protein. Eur J Biochem 269:307–316
Dunbar AJ, Priebe IK, Belford DA, Goddard C 1999 Identification of betacellulin as a major peptide growth factor in milk. Biochem J 344:713–721
Tzahar E, Waterman H, Chen X, Levkowitz G, Karunagaran D, Lavi S, Ratzkin BJ, Yarden Y 1996 A hierarchical network of interreceptor interactions determines signal transduction by Neu differentiation factor/neuregulin and epidermal growth factor. Mol Cell Biol 16:5276–5287
Ogata T, Park KY, Seno M, Kojima I 2004 Reversal of streptozotocin-induced hyperglycemia by transplantation of pseudoislets consisting of -cells derived from ductal cells. Endocr J 51:381–386
Carmichael J, DeGraff WC, Gazdar AF, Minnam JD, Mitchell JB 1987 Evaluation of tetrazolium-based semiautomated colorimetric assay. Cancer Res 47:936–942
Ishiyama N, Kanzaki M, Seno M, Yamada H, Kobayashi I, Kojima I 1998 Studies in the betacellulin receptor in pancreatic AR42J cells. Diabetologia 41:623–628
Furukawa M, Zhang YQ, Nie L, Shibata H, Kojima I 1999 Role of MAP kinase and PI 3-kianse in differentiation of pancreatic AR42J cells induced by HGF. Diabetologia 42:450–456
Groenen LC, Nice EC, Burgess AW 1994 Structure-function relationships for the EGF/TGF-a family mitogens. Growth Factors 11:235–257
Daly JM, Olayioye MA, Wong AM, Neve R, Lane HA, Maurer FG, Hynes NE 1999 NDF/heregulin-induced cell cycle changes and apoptosis in breast tumor cells: role of PI3 kinase and p38 MAP kinase pathways. Oncogene 18:3440–3451
Ito T, Kondo M, Tanaka Y, Kobayashi M, Sasada R, Igarashi K, Suenaga M, Koyama N, Nishimura O, Fujino M 2001 Novel betacellulin derivatives. J Biol Chem 276:40698–40703(Takeki Ogata1,2, Andrew J)
Third Department of Medicine (T.O., Y.Y., Y.T.), National Defense Medical College, Tokorozawa 359-8513, Japan
GroPep Ltd. (A.J.D.), Adelaide SA 5000, Australia
Department of Bioscience and Biotechnology (M.S.), Graduate School of Natural Science and Technology, Okayama University, Okayama 700-8530, Japan
Abstract
We previously described a novel alternatively spliced mRNA transcript of the betacellulin (BTC) gene. This splice isoform, termed BTC-4, lacks the C-loop of the epidermal growth factor motif and the transmembrane domain as a result of exon 4 ‘skipping’. In this study, we expressed BTC-4 recombinantly to explore its biological function. When BTC-4 was expressed in COS-7 cells, it was secreted largely into the culture medium, in contrast to BTC. Unlike BTC, highly purified recombinant BTC-4 produced in Escherichia coli failed to bind or induce tyrosine phosphorylation of either ErbB1 or ErbB4, nor did it antagonize the binding of BTC to these receptors. Consistent with this, BTC-4 failed to stimulate DNA synthesis in Balb/c 3T3 and INS-1 cells. However, BTC-4 induced differentiation of pancreatic -cells; BTC-4 converted AR42J cells to insulin-producing cells. When recombinant BTC-4 was administered to streptozotocin-treated neonatal rats, it reduced the plasma glucose concentration and improved glucose tolerance. Importantly, BTC-4 significantly increased the insulin content, the -cell mass, and the numbers of islet-like cell clusters and PDX-1-positive ductal cells. Thus, BTC-4 is a secreted protein that stimulates differentiation of -cells in vitro and in vivo in an apparent ErbB1- and ErbB4-independent manner. The mechanism by which BTC-4 exerts this action on -cells remains to be defined but presumably involves an, as yet, unidentified unique receptor.
Introduction
MEMBERS OF THE epidermal growth factor (EGF) family are characterized by a high degree of sequence similarity, particularly with respect to a common six-cysteine 36- to 40-amino-acid residue EGF-motif with a spacing of CX7CX4–5CX10–13CX1CX8C that forms three intramolecular disulfide bonds (C1-C3, C2-C4, and C5-C6) and a characteristic three-loop structure (1). The EGF-motif of growth factor peptides belonging to this family all appear to be encoded by two exons with a precisely located intervening intron, corresponding to the border separating the first two disulfide loops (A loop, C1-C3; B loop, C2-C4) from the third loop (C loop, C5-C6) (2). A common feature of these molecules is that they are synthesized as larger transmembrane precursors that are proteolytically cleaved to release the soluble biologically active form of the growth factor. The consensus EGF-motif is crucial for binding to and activating members of the ErbB receptor tyrosine kinase family (1, 2). Four receptors belonging to this family have been identified (ErbB1/EGFR, ErbB2/HER2/Neu, ErbB3, and ErbB4). Ligand binding to ErbB receptors induces receptor homo- or heterodimerization, autophosphorylation, and subsequent activation of downstream signaling pathways, resulting in diverse physiological processes, including cell proliferation, differentiation, migration, and survival (3). Ligands belonging to this family include EGF; TGF-; heparin-binding EGF; epiregulin; amphiregulin; the neuregulin subfamily, which includes the products of four genes (NRG1–4); and betacellulin (BTC). BTC binds and activates ErbB1 and ErbB4 homodimers and is further characterized by its ability to activate, albeit weakly, the highly oncogenic ErbB2/ErbB3 complex (4, 5). BTC was originally isolated as a growth factor synthesized in an insulinoma cell line (6). BTC is expressed abundantly in the intestine and pancreas (7). In the pancreas, BTC is expressed in non--cells of the islet in adults and is expressed in endocrine precursor cells in the fetus (8). This suggests the possibility that BTC regulates the growth and/or differentiation of pancreatic -cells and their precursors. In accordance with this notion, BTC is able to induce differentiation of pancreatic AR42J cells and convert them into insulin-producing cells (9) and to convert glucagon-producing -cells into -cells (10). Furthermore, in vivo studies have shown that BTC promotes regeneration of pancreatic -cells in various animal models of experimental diabetes (11, 12, 13, 14). Collectively, BTC is a growth or differentiation factor for pancreatic -cells.
We recently identified an alternately spliced mRNA transcript encoding a novel isoform of human BTC (15). This isoform, originally termed BTC- but now referred to as BTC-4, lacks 147 bp encoding exon 4 of the BTC gene, leading to the generation of an mRNA encoding an unusual BTC precursor in which the C-loop of the EGF domain and the transmembrane domain are deleted while the remainder of the mature molecule, including loops A and B and the ‘hinge’ valine, is fused in frame to the truncated C-terminal cytoplasmic tail (Fig. 1). Retention of the hydrophobic signal sequence and the absence of the transmembrane domain suggests that BTC-4 may be a secreted protein. Interestingly, a similar isoform lacking the third disulfide loop of the EGF domain has been reported for heparin-binding EGF (16). In this case, a 94-bp insertion between exons III and IV causes a frameshift and premature termination generating a protein that retains the signal peptide, proregion, heparin-binding domain and the first two conserved disuphide loops of the EGF motif, while a short nine-amino-acid tail replaces the third disulfide loop, transmembrane, and cytoplasmic domains. In this report, we have produced BTC-4 recombinantly to characterize it’s biological activity and, in particular, to determine whether it may act as a naturally occurring BTC antagonist. Our results indicate that whereas BTC-4 does not bind to either ErbB1, ErbB4, or the ErbB2–3 heterodimeric complex and shows no evidence of acting as an BTC antagonist, BTC-4 is biologically active and acts as an inducer of differentiation of pancreatic -cells in vitro and in vivo, possibly through a novel cell surface receptor.
Materials and Methods
Reagents
Recombinant human activin A was a generous gift from Dr. Y. Eto of Central Research Laboratory, Ajinomoto Inc. (Kawasaki, Japan). Streptozotocin (STZ) was purchased from Wako Pure Chemicals (Osaka, Japan). Mouse fibroblast Balb/c 3T3 cells were from the American Type Culture Collection (Manassas, VA). Human lung fibroblast AG2804 cells and CHO cells transfected with either ErbB 4 or ErbB2/3 were kindly provided by Dr. J. Gunn (Texas A&M University, College Station, TX) and Prof. Y. Yarden (Weizmann Institute, Rehovet, Israel), respectively. INS-1 cells were generously provided by Prof. C. Wollheim (University of Geneva, Geneva, Switzerland). Anti-FLAG M2 antibody was from Sigma (St. Louis, MO). Anti-ErbB1 (1005), anti-ErbB4 (C-18), and antiphosphotyrosine (PY20) antibodies were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) and horseradish peroxidase (HRP)-conjugated rabbit antisheep antibody from Zymed Laboratories (San Francisco, CA). HRP-conjugated sheep antimouse antibody was from Silenus Laboratories (Hawthorn, Australia). Lipofectamine 2000, Enterokinase, Escherichia coli TOP10 cells, Optimem-1, 10–20% Tris Tricine gels, and pcDNA3.1 were from Invitrogen (Carlsbad, CA). West Pico SuperSignal substrate was from Pierce Biotechnology, Inc. (Rockford, IL) pET32, BL21trxB (DE3) cells, and BugBuster Protein Extraction Reagent were from Novagen; Ni-NTA agarose was from QIAGEN (Valencia, CA). Recombinant human BTC and goat antihuman BTC ectodomain (Asp32-Tyr111) antibody were from R&D Systems, Inc. (Minneapolis, MN). Complete protease inhibitors and protein G-Sepharose were from Roche Molecular Biochemicals (Basel, Switzerland).
Cloning of BTC and BTC-4 FLAG-tagged constructs
Full-length BTC (amino acids Met1-Ala178) and BTC-4 (amino acids Met1-Ala129) were cloned into pcDNA3.1 to generate expression vectors for the mammalian production of BTC and BTC-4 as follows. The open-reading frames (ORFs) of BTC and BTC-4 were excised from pBlue-BTC-4 and pBlue-BTC (15) with ApaI and BamHI and recloned into ApaI/BamHI digested pcDNA3.1. To generate ‘Flag-tagged’ pcDNA3.1-BTC-4 and pcDNA3.1-BTC constructs, in which the Flag epitope DYKDDDDK is inserted between amino acids S35–T36, pcDNA3.1-BTC-4 or pcDNA3.1-BTC was used as a template in PCR using the primers; 5'-CTCGGGAATTCCGACTACAAGGACGACGATGACAAGACCAGAAGTCCTGAA-3' (sense) (underlined nucleotides correspond to an EcoRI restriction site; double underlined nucleotides encode the FLAG epitope tag) and 5'-CTCCTGCAGTTAAGCAATATTTGTCTCTTC-3' (antisense) (underlined nucleotides correspond to a PstI restriction site). The resulting PCR products were purified and digested with EcoRI and PstI and cloned into EcoRI/Pst1 digested pBlue-BTC or pBlue-BTC-4. Subsequently, the resultant Flag-tagged pBlue-BTC and pBlue-BTC-4 constructs were digested with ApaI/BamH1 and cloned into ApaI/BamH1 digested pcDNA3.1 to generate pcDNA3.1-FLAG-BTC-4 and pcDNA3.1-FLAG-BTC.
Expression and analysis of BTC and BTC-4 FLAG-tagged constructs in COS-7 cells
For transient expression of pcDNA3.1-FLAG-BTC-4 and pcDNA3.1-FLAG-BTC, COS-7 cells were plated into 12-well plates, at 2 x 105 cells/well, in DMEM/10% fetal bovine serum. After overnight incubation, cells were transfected with 2 μg construct DNA using Lipofectamine 2000 and Optimem-1 media according to the manufacturer’s instructions. Twelve hours later, the cells were washed twice and replenished with fresh Optimem-1 media. Seventy-two hours post transfection, culture medium (CM) was collected and cell lysates prepared. CM was clarified by centrifugation, and the presence of BTC or BTC-4 in the media was analyzed by ELISA or Western blotting using an anti-FLAG-M2 antibody (Sigma). BTC or BTC-4 present in the cell lysate was analyzed by Western blotting with the anti-FLAG-M2 antibody. Cell lysates were prepared by washing the cells twice with PBS after the removal of CM and then lysing the cells directly in SDS-PAGE sample buffer and heating to 95 C for 5 min. For ELISA analysis of the CM, 90 μl of CM was mixed with 10 μl of 10x coating buffer (0.15 mol/liter Na2CO3, 0.35 mol/liter NaHCO3, pH 9.3) and loaded into 96-well immunosorbent plates. Plates were coated overnight at 4 C, then blocked in 2% BSA in PBS/0.1% Tween (PBS-T). Plates were washed four times with PBS-T and then incubated for 1 h at 37 C with anti-FLAG antibody diluted in PBS-T (2.5 μg/ml). After incubation, plates were washed as above and incubated with HRP-conjugated sheep antimouse antibody (1:2000) diluted in PBS-T. Plates were then washed four times with PBS-T and developed with o-phenylamine diamine substrate and stopped with 2 mol/liter H2SO4. Absorbance was read at 490 nm. Results are expressed as the mean ± SD of triplicate determinations. To analyze the presence of BTC or BTC-4 on the cell surface by ELISA, CM was removed and the cells washed three times with PBS. The cells were then fixed with 4% paraformaldehyde and stained with anti-FLAG M2 antibody as described above. For Western blot analysis with the anti-FLAG-M2 antibody, CM or cell lysates were resolved by SDS-PAGE (10–20% Tris-Tricine gels). Proteins were then transferred to nitrocellulose (Hybond-C extra; Amersham Pharmacia Biotech, Little Chalfont, UK) and membranes probed with mouse anti-FLAG-M2 antibody (2.5 μg/ml) and then HRP-conjugated sheep antimouse antibody (1:10,000). HRP-labeled proteins were visualized using SuperSignal West Dura Extended Duration Substrate.
Expression and purification of BTC and BTC-4
BTC-4 [constituting amino acids Asp32-Ala129 (BTC-432–129)] (Fig. 1) was expressed as a thiroredoxin fusion protein in pET32a. The ORF sequence of BTC-432–129 was amplified by PCR using the primer set, 5'-CGTCCATGGCTGATGGGAATTCCACCAGAAGT-3' (sense) and 5'-CGTCTCGAGTCATTAAGCAATATTTGTCTCTTC-3' and pBlue-BTC-4 (15) as template. NcoI and XhoI recognition sites were attached to the sense and antisense primers, respectively (underlined). BTC (constituting amino acids 32–111; BTC32–111) was also expressed using the pET system as a positive control, as described previously (7), or produced as a thioredoxin fusion protein with the plasmid pET32a. For this construct, the ORF sequence of BTC32–111 was amplified by PCR using the primer set, 5'-CGTCCATGGCTGATGGGAATTCCACCAGAAGT-3' (sense) and 5'-CGTCTCGAGTCAGTAAAACAAGTCAACTGT-3' (antisense) and pBlue-BTC (see Ref.18) as template. NcoI and XhoI recognition sites were also attached to the sense and antisense primers, respectively (underlined). The PCR products were digested with NcoI/XhoI and cloned into NcoI/XhoI digested pET32a vector as above. The resultant plasmids, pETBTC-432–129 and pETBTC32–111, were maintained in Escherichia coli JM109 and, for expression, transformed into Escherichia coli BL21trxB (DE3) cells. Large-scale expression of BTC or BTC-4 in BL21trxB (DE3) cells was carried out as described in the pET system manual (Novagen, Madison, WI). BTC or BTC-4 thioredoxin fusion proteins were purified from clarified cell lysates by Ni-NTA agarose affinity, followed by cleavage with Enterokinase to isolate authentic BTC or BTC-4. Authentic BTC or BTC-4 was separated from the thioredoxin fusion partner by additional Ni-NTA agarose affinity chromatography and reverse-phase HPLC (RP-HPLC). Briefly, frozen cell pellets were thawed on ice and resuspended in 18 ml BugBuster Protein Extraction Reagent containing lysozyme (100 μg/ml) and incubated at room temperature with gentle shaking for 10 min. After incubation, the cell lysate was sonicated (3 x 5-sec bursts) and clarified by centrifugation (20 min, 16,000 rpm, 10 min). The clarified cell lysate was adjusted to 10 mmol/liter imidazole and applied to a Ni-NTA agarose affinity column preequilibrated with 50 mmol/liter NaH2PO4, 0.3 mol/liter NaCl, 10 mmol/liter imidazole (pH 8.0). After column washing with 50 mmol/liter NaH2PO4, 0.3 mol/liter NaCl, 40 mmol/liter imidazole (pH 8.0), BTC or BTC-4 was eluted with 50 mmol/liter NaH2PO4, 0.3 mol/liter NaCl, 250 mmol/liter imidazole (pH 8.0) and subsequently dialyzed overnight (Spectra/Por, 3.5 kDa MWCO; Spectrum Laboratories, Houston, TX) against several changes of Enterokinase cleavage buffer (50 mmol/liter Tris-Cl, 1 mmol/liter CaCl2, 0.1% Tween 20, pH 7.4). To remove the thioredoxin fusion partner from BTC or BTC-4, Enterokinase was added to a final concentration of 0.1 U/20 μg protein and incubated overnight at 37 C with gentle mixing. BTC or BTC-4 was separated from the thioredoxin fusion partner by further Ni-NTA agarose affinity chromatography as described above. In this case, the cleaved thioredoxin fusion partner is captured on the resin and BTC or BTC-4 is collected in the flowthrough fraction flow-through fraction. BTC or BTC-4 present in the flow-through fraction was further purified by RP-HPLC. Briefly, the Ni-NTA agarose flow-through fraction was diluted 1:4 (vol/vol) with 0.1% trifluoroacetic acid (TFA) and applied to a C4 Prep-Pak RP-HPLC column (25 mm x 100 mm; 300, 15 μm; Millipore-Waters, Bedford, MA) at a flow rate of 10 ml/min. The column was washed with 0.1% TFA and eluted with a gradient of 8–80% (vol/vol) acetonitrile over 150 min in the presence of 0.08% TFA at a flow rate of 10 ml/min. Twenty-milliliter fractions were collected, and aliquots of each (50 μl) were analyzed by SDS-PAGE and Western blotting using a goat antihuman BTC ectodomain (Asp32-Tyr111) antibody to identify pure BTC or BTC-4 containing fractions. The purity of all BTC or BTC-4 preparations were analyzed by analytical RP-HPLC, electrospray ionization mass spectrometry, and SDS-PAGE. Analytical RP-HPLC was performed using a Brownlee aquopore C4 RP-HPLC column (2.1 x 100 mm) (Alltech, Deerfield, IL) with a linear gradient of 8–80% (vol/vol) acetonitrile and 0.08% TFA over 40 min at a flow rate of 0.5 ml/min. Alternatively, BTC-432–129 (containing an initiation methionine residue) was cloned into the expression vector pET3b to construct pBO604, and the recombinant protein was solubilized and refolded from inclusion bodies, purified through cation exchange column and gel filtration column chromatography as previously described (7, 17). The authenticity of BTC-432–129 was confirmed by amino-terminal sequence analysis and amino acid composition. All the recombinant proteins prepared in this study were lyophilized and stored at –80 C before use.
ErbB-receptor binding and mitogenic activity of BTC-4
The binding affinity of BTC or BTC-4 to ErbB receptors was determined by measuring the ability of BTC or BTC-4 to competitively displace [125I]-labeled recombinant human BTC from ErbB1 or ErbB4 receptors present on AG2804 fibroblasts or CHO cells stably transfected with ErbB4 (18) as described previously (19). The mitogenic activity of BTC and BTC-4 toward Balb/c 3T3 fibroblasts and INS-1 cells was performed as previously described (19).
Analysis of ErbB-1 and ErbB-4 receptor tyrosine phosphorylation
AG2804 or CHO-ErbB4 cells were grown to confluence in 10-cm dishes and subsequently incubated for 12–14 h in serum-free medium. Cells were then stimulated with 10 nmol/liter BTC, BTC-4, or a combination of both for 10 min at room temperature. After stimulation, cells were washed twice in PBS and then lysed in lysis buffer (0.5 ml) (50 mmol/liter Tris-Cl, pH 7.4; 150 mmol/liter NaCl; 1% deoxycholate; 1% Triton X-100; 0.1% sodium dodecyl sulfate; 5 mmol/liter sodium orthovanadate; 10 mmol/liter sodium fluoride; 1 mmol/liter EGTA; and complete protease inhibitors. Cell lysates were cleared by centrifugation (20 min, 15,000 x g at 4 C) and ErbB1 or ErbB4 immunoprecipitated by incubating the lysate with 1 μg rabbit polyclonal anti-ErbB1 antibody (1005, Santa Cruz) or rabbit polyclonal anti-ErbB4 antibody, respectively, for 2 h at 4 C with gentle shaking. After incubation, protein-G Sepharose was added and the mixture incubated for a further 1 h at 4 C. Immune complexes were collected by centrifugation, washed three times in lysis buffer, and heated (3 min, 95 C) in SDS-PAGE sample buffer. Proteins were separated by SDS-PAGE (10–20% Tricine gels) and transferred to nitrocellulose filters (Hybond C; Amersham). Membranes were probed with antiphosphotyrosine monoclonal antibody (PY20) and then with HRP-conjugated sheep antimouse antibody. HRP-labeled proteins were visualized using enzyme-linked chemiluminescence (Amersham). To confirm equal loading, blots were stripped and reprobed with rabbit polyclonal anti-ErbB1 or rabbit polyclonal anti-ErbB4 antibodies and HRP-conjugated rabbit antisheep antibody.
Measurement of differentiation and apoptosis of AR42J cells
AR42J-B20 cells were cultured in DMEM medium containing 10% fetal calf serum as described previously (9). To assess differentiation into insulin-producing cells, cells were incubated for 48 h with 2 nmol/liter activin A and either 1 nmol/liter BTC or BTC-4. Cells were then fixed and stained with antiinsulin or anti-C peptide antibody as described previously (9, 20), and the number of insulin-positive cells was counted. Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI). Apoptosis was assessed by using a terminal deoxynucleotidyl transferase-mediated deoxyuridine 5-triphosphate nick end labeling (TUNEL) technique (Wako Pure Chemicals.). Changes in the number of viable cells were assessed by using 3-[4,5-dimethylthiazole-2-yl]-2,5,-dipheltetrazodium bromide (MTT) (21). Anti-C-peptide antibody was obtained from Linco Research, Inc. (St. Charles, MS).
Treatment of neonatal rats with streptozotocin
One-day-old Sprague Dawley rats were treated with streptozotocin and tested for blood glucose levels as described previously (14). The animals were included in the study only if the blood glucose concentration was between 200 and 350 mg/dl on the next day (designated as d 0). Animals whose blood glucose was more than 350 mg/dl on d 0 developed severe diabetes, and the blood glucose concentrations in these animals were more than 600 mg/dl on d 1 or 2. Neonatal STZ-treated rats were injected with 3 pmol/g BTC-4, BTC, or saline every day for 5 d starting from d 0. The fasting blood glucose concentration and the body weight were measured daily for the first week and then once a week for up to 8 wk. Two months after the STZ treatment, an ip glucose tolerance test (2 g/kg body weight) was performed after 14 h of fasting. The experimental protocol was approved by the Animal Care Committee of the Gunma University.
Tissue processing and immunohistochemistry
On d 4 and wk 8, the animals were injected ip with 1 ml bromodeoxyuridine labeling reagent per 100 g body weight (cell proliferation kit; Amersham) and decapitated after 3 h. The pancreas was excised, weighed, and divided into two parts. One portion from the splenic segment was fixed with 4% paraformaldehyde/PBS overnight at 4 C and processed for paraffin embedding. Four series from each pancreas were cut, at intervals of 100 μm in neonates and 300 μm in adults, for immunostaining and histochemistry. The second portion was homogenized in cold acid-ethanol, heated for 5 min in 70 C water bath, and centrifuged, and the supernatant was stored at –20 C before assaying for insulin. Insulin was measured by time-resolved immunofluorometric assay as described previously (9). Immunohistochemistry was performed as described previously (12, 13). Quantitation of the -cell mass was performed on insulin-stained sections using image analysis software (National Institutes of Health image) by means of an AX70 Epifluorescence microscope (Olympus Corp., Tokyo, Japan) equipped with a PXL 1400 cooled-charge-coupled device camera system (Photometrics, Tucson, AZ) operated with IP Lab Spectrum software (Signal Analysis, Vienna, VA). At least 40 random fields (magnification, x200) from one section (three sections from different series per block) were measured for the area of insulin-positive cells. The -cell mass was calculated as described elsewhere (12, 13). To quantify -cell neogenesis, the number of islet-like cell clusters (ICCs) (less than five cells across) was determined. The number of ICCs and islets was counted in the section stained with the antiinsulin antibody, and the area in these sections was measured. At least five sections were analyzed per animal. Results were expressed as means ± SE. For comparisons between two groups, the unpaired t test was used.
Results
BTC-4 is a secreted protein
Like other members of the EGF family, BTC is synthesized as a transmembrane-anchored precursor protein (pro-BTC), which can be proteolytically cleaved to yield soluble BTC containing the EGF-motif (Asp32-Tyr111). Retention of the hydrophobic signal peptide and the absence of the transmembrane domain suggest that BTC-4 may be a secreted protein. To test this, the complete ORF of either BTC or BTC-4 (additionally engineered to incorporate a FLAG epitope between amino acids Ser35 and Thr36 for subsequent detection) was cloned into the expression vector pcDNA3.1 and secretion monitored after transient transfection in COS-7 cells. Seventy-two hours post transfection, BTC-4-FLAG was clearly detected in the culture supernatant by either Western-blot (14 kDa) or enzyme immunoassay, with very little present in the cell lysate or on the cell surface (Fig. 2, A and B). In contrast, BTC-FLAG was predominantly present in the cell lysate (20 kDa) and on the cell surface (Fig. 2, A and B).
Recombinant production of BTC-4
To characterize the biological activity of BTC-4, we cloned the cDNA (Asp32-Ala129; without the signal peptide) into the pET-32a expression vector and expressed the protein as a thioredoxin fusion in Escherichia coli BL21trxB (DE3) (Fig. 3A). This strain is deficient in thioredoxin reductase, allowing for disulfide bond formation in the cytoplasm. The purified thioredoxin fusion protein was cleaved with enterokinase to release BTC-4, which was purified to homogeneity by RP-HPLC (Fig. 3B). As a control, we also generated BTC (Asp32-Tyr111) with this system and purified it in the same manner. The purity of both the BTC and BTC-4 preparations was further confirmed by N-terminal sequence analysis (five cycles), which gave the expected N-terminal sequence with an approximate purity of more than 95% (data not shown). The molecular mass of recombinant BTC and BTC-4 determined by electrospray ionization mass spectrometry was 9,211.35 ± 0.29 Da and 11,450.26 ± 0.23 Da, respectively, which is consistent with the calculated theoretical masses of 9,249 and 11,452 Da (data not shown). After SDS-PAGE and silver staining, a single band at approximately 9 kDa and 11.5 kDa was obtained for BTC and BTC-4 detected under reducing or nonreducing conditions (data not shown).
BTC-4 does not bind or activate either ErbB1 or ErbB4
BTC stimulated the proliferation of Balb/c 3T3 fibroblasts in a dose-dependent fashion (Fig. 4A). In contrast, BTC-4 did not stimulate cell proliferation, even at concentrations as high as 100 nmol/liter (Fig. 4A), or antagonize the effects of BTC in this cell line (Fig. 4B). Consistent with these results, BTC, but not BTC-4, could displace [125I]-BTC from ErbB1 receptors present on human lung fibroblasts (AG2804) (Fig. 4, C and D) and induce ErbB1 receptor autophosphorylation (Fig. 5). Similarly, BTC, but not BTC-4, competed for ErbB4 receptor binding (Fig. 4, E and F) and induced ErbB4 receptor tyrosine phosphorylation (Fig. 5). These results indicate that, unlike BTC, BTC-4 displays no affinity toward either ErbB1 or ErbB4. We also examined the effect of BTC and BTC-4 on DNA synthesis in INS-1 cells. As shown in Fig. 4G, BTC stimulated DNA synthesis in INS-1 cells, whereas BTC-4 did not.
BTC-4 induces the differentiation of pancreatic -cells
BTC is known to stimulate the differentiation of pancreatic -cells (9, 10, 11, 12, 13, 14), and there is some evidence to suggest that this may occur through a unique non-ErbB cell surface receptor (22). To examine the effect of BTC-4 on the differentiation of pancreatic -cells, we used the model cell line AR42J-B20; a subclone of AR42J, an amylase-secreting pancreatic tumor cell line. In these cells, BTC acts coordinately with activin A and converts them to insulin-secreting cells (9). Activin A also induces apoptosis in these cells and, in the absence of a survival factor, many of the activin-treated cells die by apoptosis after the conversion to pancreatic polypeptide (PP)-producing cells (23). BTC exerts two effects in AR42J-B20 cells: BTC inhibits apoptosis induced by activin A; and, second, converts them to insulin-producing cells (9, 23). As shown in Fig. 6A, AR42J-B20 cells differentiated into insulin-producing cells by a combination of activin A and BTC. Quantitatively, 78 ± 6.5% of the cells (mean ± SE; n = 4) became insulin-positive after 48 h. Cells treated with activin A and BTC exhibited extended processes and expressed immunoreactive insulin (Fig. 6C). Cells treated with a combination of activin A and BTC-4 also became insulin-positive; 72 ± 4.5% (n = 4) of the cells became insulin-positive after 48 h (Fig. 6B). It should be noted that cells treated with activin A and BTC-4 were also positive for C-peptide (Fig. 6G). In contrast to cells treated with activin A and BTC, the morphology of cells treated with activin A and BTC-4 was quite different. Some of the insulin-positive cells displayed extended processes, but most of them remained circular in appearance. The nuclei of these cells were condensed and, in some instances, were either fragmented or absent (Fig. 6, B and D), characteristic of cell death by apoptosis. Consistent with this notion, many of the cells treated with activin A and BTC-4 were TUNEL-positive (Fig. 6F), compared with only a small fraction of cells treated with activin A and BTC (Fig. 6E). To further confirm the effect of BTC-4 on the survival of AR42J-B20 cells, we measured the changes in the number of viable cells after treatment with activin A and either BTC or BTC-4. After 48 h, the number of viable cells treated with activin A and BTC was 78.8 ± 4.2% of the number of cells seeded. In contrast, the number of viable cells treated with activin A and BTC-4 was significantly lower (21.8 ± 3.2%) (P < 0.005). Taken together, BTC-4 was as effective as BTC in stimulating the differentiation of AR42J-B20 cells; however, BTC-4 was much less potent in promoting survival of these cells. Note that BTC-4 alone did not induce differentiation of AR42J-B20 cells. Also, BTC-4 did not affect survival of these cells. When both BTC and BTC-4 were administered together with activin A, there was a small additivity in the actions of BTC and BTC-4, and 88 ± 4.5% of the cells became insulin-positive.
BTC-4 administration to STZ-treated rats reduces the plasma glucose concentration and improves glucose tolerance
To investigate further the -cell differentiating activity of BTC-4, we administered BTC-4 to neonatal STZ-treated rats. In STZ-treated neonatal rats, the plasma glucose concentration increased markedly and peaked on d 2 (>400 mg/dl) (Fig. 7A), thereafter declining gradually although still significantly higher compared with control rats after 2 months (Table 1). In STZ-treated rats administered BTC-4, the plasma glucose concentration was significantly lower on d 1, and the peak value was markedly reduced (Fig. 7A). The plasma glucose concentration was also reduced thereafter, until 2 months (Table 1). In contrast, BTC was less potent in improving hyperglycemia (Fig. 8A and Table 1). We also administered BTC and BTC-4 simultaneously, but the effect was not different from that obtained by BTC-4 alone (data not shown). At 2 months of age, ip glucose tolerance tests were performed. In STZ-treated rats, the glucose response to ip glucose-loading was markedly impaired; whereas in STZ-treated rats administered BTC-4, the glucose response was significantly improved (Fig. 7B). BTC-4 also increased the insulin response; but, compared with normal rats, the insulin response to ip glucose-loading was still delayed. In BTC-treated rats, glucose tolerance was improved, but the effect of BTC was less than that of BTC-4 (Fig. 7B). Moreover, the insulin content and the -cell mass were significantly increased in BTC-4-treated rats (Table 1), and histological analysis of pancreatic tissue on d 4 indicated that BTC-4 significantly increased the numbers of PDX-1-positive ductal cells and ICCs (Fig. 8). We also examined the effect of BTC and BTC-4 on replication of insulin-positive cells. On d 4, the numbers of bromodeoxyuridine/insulin double-positive cells in saline-, BTC-, and BTC-4-treated rats were 5.4 ± 0.6, 8.3 ± 0.8, and 7.4 ± 1.0%, respectively (n = 5). The effects of BTC-4 and BTC were significant (P < 0.05) compared with saline-treated control. We also counted the number of apoptotic cells. On d 4, the number of TUNEL-positive cells was very low (less than 1 x 10–6/μm2).
Discussion
The present study was conducted to investigate the biological activity of the BTC splice variant, BTC-4. BTC-4 is unique in that it lacks the C-loop of the EGF motif and the transmembrane domain yet retains the signal peptide (15), suggesting that the protein produced from this transcript is a secreted protein. In contrast, BTC is synthesized as a transmembrane precursor, which can undergo ‘ectodomain shedding’ to release the soluble biologically active mature molecule. BTC-4 was found predominantly in the culture supernatant after expression in COS-7 cells as an approximately 14-kDa protein. In contrast, little, if any, BTC was found in the culture supernatant, the majority being present in the cell lysate (Fig. 2A) and on the cell surface (Fig. 2B). Thus, although BTC-4 is a protein product of the splicing variant of the BTC gene, the production of BTC-4 and BTC is differentially regulated. The three disulfide loops and certain key residues in the two -sheet structural regions of EGF ligands (for example, Gly49, Gly70, Arg72, and Leu78; Leu47 and Leu48 in EGF and TGF-, respectively) are crucial for ErbB receptor binding (1, 2, 24). Because BTC-4 lacks the C-loop of the EGF motif encompassing Gly70, Arg72, and Leu78 (see Fig. 1), we were interested to determine whether BTC-4 was able to bind and activate ErbB-1 and ErbB-4, the principal BTC ErbB receptors (4, 5). BTC-4 was unable to bind or induce tyrosine phosphorylation of either ErbB-1 or ErbB-4 (Figs. 4 and 5), indicating that BTC-4 is not a ligand for either ErbB1 or ErbB4. Moreover, BTC-4 did not stimulate DNA synthesis in Balb/c 3T3 cells, which express ErbB-1 (Fig. 4), or induce phosphorylation of ErbB2/3 heterodimers (data not shown). In addition to stimulating cell proliferation, principally through ErbB1 and the activation of the Grb2-Sos-Ras-Raf-Mek1/2-Erk1/2 pathway, BTC also may act as a survival factor for various cell types through ErbB receptor-induced activation of the PI3K-Akt pathway (4, 5, 25). Consistent with this, we found that BTC, but not BTC-4, was able to inhibit apoptosis of AR42J-B20 cells induced by activin A (Fig. 6, E and F). These data support the concept that BTC-4 is not a ligand for the ErbB receptors and, hence, is unable to promote cell growth or survival. Nevertheless, both BTC and BTC-4 converted AR42J cells to insulin-producing cells (Fig. 6); and moreover, BTC-4 significantly increased the insulin content, -cell mass, the number of PDX-1-positive ductal cells, and ICCs in neonatal STZ-treated rats (Table 1). Consequently, BTC-4 appears to act as a -cell differentiation factor in vitro and in vivo, and it exerts this differentiation-inducing activity by acting through a unique cell surface receptor independent of either ErbB1 or ErbB4. We have previously postulated that BTC induces the differentiation of AR42J-B20 cells into insulin-secreting cells through a unique non-ErbB receptor of approximately 190 kDa (22). The data presented here strengthen this notion. Moreover, the fact that both BTC-4 and BTC display this differentiation-inducing activity suggests that the structural determinants for binding this unique receptor lie within the region Asp32-Val94. Analysis of C-terminal truncation derivatives of BTC, which display little affinity to ErbB1, are only weakly mitogenic for AR42J cells, yet retain the ability to induce the differentiation of these cells into insulin-secreting cells (26), is further consistent with this hypothesis. Taken together, these results suggest that the differentiation-inducing activity of both BTC and BTC-4 may be exerted through this unique receptor. The identification and characterization of this receptor is currently under investigation in our laboratory.
Previous studies have shown that BTC is able to increase the -cell mass in various experimental rodent models of diabetes, indicating that it may represent a useful therapeutic approach for the treatment of the human condition (11, 12, 13, 14). From a clinical point of view, administration of recombinant therapeutic growth factors to treat chronic disorders is hampered by the potential side effect of tumorigenesis, as a result of promoting cell proliferation. In this regard, because we have found that BTC-4 reproduces the differentiation activity of BTC without promoting cell proliferation, it may represent a more attractive therapeutic alternative for the treatment of diabetes. The observed in vitro apoptosis in cells treated with activin A and BTC-4 may not necessarily occur in vivo because -cell progenitors would be bathed in a milieu of growth factors that may act to reduce apoptosis by virtue of their prosurvival properties. Thus, BTC-4 may be useful to treat certain types of diabetes caused by impaired -cell neogenesis.
Acknowledgments
We thank Mayumi Odagiri for secretarial assistance during the preparation of the manuscript, and Ilka Priebe and Helen Webb for excellent technical assistance.
Footnotes
This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan; a grant from the Project of Realization of Regeneration Medicine; and the Australian Government through the Cooperative Research Centres Programe.
1 This work is dedicated to the memory of Takeki Ogata, who sadly passed away during the completion of this manuscript.
2 T.O. and A.J.D. contributed equally to this paper.
Abbreviations: BTC, Betacellulin; CM, culture medium; DAPI, 4',6-diamidino-2-phenylindole; EGF, epidermal growth factor; HRP, horseradish peroxidase; ICC, islet-like cell cluster; ORF, open-reading frame; RP-HPLC, reverse-phase HPLC; TFA, trifluoroacetic acid; TUNEL, terminal deoxynucleotidyl transferase-mediated deoxyuridine 5-triphosphate nick end labeling.
References
Van Zoelen EJJ, Stortelers C, Lenferink AEG, Van de Poll MLM 2000 The EGF domain: requirements for binding to receptors of the ErbB family. Vitam Horm 59:99–131
Harris RC, Chung E, Coffey RJ 2003 EGF receptor ligands. Exp Cell Res 284:2–13
van der Greer P, Hunter T, Lindberg RA 1994 Receptor-tyrosine kinases and their signal transduction pathways. Annu Rev Cell Biol 10:251–337
Alimandi M, Wang LM, Bottaro D, Lee CC, Kuo A, Frankel M, Fedi P, Tang C, Lippman M, Pierce JH 1997 Epidermal growth factor and betacellulin mediate signal transduction through co-expressed ErbB2 and ErbB3 receptors. EMBO J 16:5608–5617
Pinkas-Kramarski R, Leferink AEG, Bacus SS, Lyass L, an de Poll MLM, Klapper LN, Tzahar E, Sela M, van Zoelen EJJ, Yarden Y 1998 The oncogenic ErbB2-ErbB-3 heterodimer is a surrogate receptor of the epidermal growth factor and betacellulin. Oncogene 16:1249–1258
Shing Y, Christofori G, Hanahan D, Ono Y, Sasada R, Igarashi K, Folkman J 1993 Betacellulin: a mitogen from pancreatic -cell tumors. Science 259:1604–1607
Seno M, Tada H, Kosaka M, Sasada R, Igarashi K, Shing Y, Folkman J, Ueda M, Yamada H 1996 Human betacellulin, a member of the EGF family, dominantly expressed in pancreas and small intestine, is fully active in a monomeric form. Growth Factors 13:181–191
Miyagawa J, Hanafusa T, Sasada R, Yamamoto K, Igarashi K, Yamamoto K, Seno M, Tada R, Nammo T, Li M, Yamagata K, Nakajima H, Namba M, Kuwajima M, Matsuzawa Y 1999 Immunohistochemical localization of pancreas and islet tumor cells. Endocr J 46:755–764
Mashima H, Ohnishi H, Wakabayashi K, Miyagawa J, Hanafusa T, Seno M, Yamada H, Kojima I 1996 Betacellulin and activin A coordinately convert amylase-secreting pancreatic AR42J cells into insulin-secreting cells. J Clin Invest 97:1647–1654
Watada H, Kajimoto Y, Miyagawa J, Hanafusa T, Hamaguchi K, Matsuoka T, Yamamoto K, Matsuzawa Y, Kawamori R, Yamasaki Y 1996 PDX-1 induces insulin and glucokinase gene expression in TC clone 6 cells in the presence of betacellulin. Diabetes 45:1826–1831
Yamamoto K, Miyagawa J, Waguri M, Sasada R, Igarashi K, Li M, Nammo T, Moriwaki T, Imagawa A, Yamagata K, Nakajima H, Namba M, Tochino Y, Hanafusa T, Matsuzawa Y 2000 Recombinant betacellulin promotes neogenesis of -cells and ameliorates glucose intolerance in mice with diabetes induced by selective alloxan perfusion. Diabetes 49:2021–2027
Li L, Seno M, Yamada H, Kojima I 2001 Promotion of -cell regeneration by betacellulin in ninety percent pancreatectomized rats. Endocrinology 142:5379–5385
Li L, Seno M, Yamada H, Kojima I 2003 Betacellulin improves glucose metabolism by promoting conversion of intraislet precursor cells to -cells in streptozotocin-treated mice. Am J Physiol 285:E577–E583
Li L, Yi Z, Seno M, Kojima I 2004 Activin A and betacellulin: effect on -cell regeneration in neonatal streptozotocin-treated rats. Diabetes 53:608–615
Dunbar A, Goddard C 2000 Identification of an alternatively spliced mRNA branscript of human betacellulin lacking the C-loop of the EGF motif and the transmembrane domain. Growth Factors 18:169–176
Loukianov E, Loukinanova T, Wiedlocha A, Olnes S 1997 Expression of mRNA for a short form of heparin-binding EGF-like growth factor. Gene 195:81–86
Maeda T, Kitazoe M, Tada H, de Llorens R, Salomon DS, Ueda M, Yamada H, Seno M 2002 Growth inhibition of mammalian cells by eosinophil cationic protein. Eur J Biochem 269:307–316
Dunbar AJ, Priebe IK, Belford DA, Goddard C 1999 Identification of betacellulin as a major peptide growth factor in milk. Biochem J 344:713–721
Tzahar E, Waterman H, Chen X, Levkowitz G, Karunagaran D, Lavi S, Ratzkin BJ, Yarden Y 1996 A hierarchical network of interreceptor interactions determines signal transduction by Neu differentiation factor/neuregulin and epidermal growth factor. Mol Cell Biol 16:5276–5287
Ogata T, Park KY, Seno M, Kojima I 2004 Reversal of streptozotocin-induced hyperglycemia by transplantation of pseudoislets consisting of -cells derived from ductal cells. Endocr J 51:381–386
Carmichael J, DeGraff WC, Gazdar AF, Minnam JD, Mitchell JB 1987 Evaluation of tetrazolium-based semiautomated colorimetric assay. Cancer Res 47:936–942
Ishiyama N, Kanzaki M, Seno M, Yamada H, Kobayashi I, Kojima I 1998 Studies in the betacellulin receptor in pancreatic AR42J cells. Diabetologia 41:623–628
Furukawa M, Zhang YQ, Nie L, Shibata H, Kojima I 1999 Role of MAP kinase and PI 3-kianse in differentiation of pancreatic AR42J cells induced by HGF. Diabetologia 42:450–456
Groenen LC, Nice EC, Burgess AW 1994 Structure-function relationships for the EGF/TGF-a family mitogens. Growth Factors 11:235–257
Daly JM, Olayioye MA, Wong AM, Neve R, Lane HA, Maurer FG, Hynes NE 1999 NDF/heregulin-induced cell cycle changes and apoptosis in breast tumor cells: role of PI3 kinase and p38 MAP kinase pathways. Oncogene 18:3440–3451
Ito T, Kondo M, Tanaka Y, Kobayashi M, Sasada R, Igarashi K, Suenaga M, Koyama N, Nishimura O, Fujino M 2001 Novel betacellulin derivatives. J Biol Chem 276:40698–40703(Takeki Ogata1,2, Andrew J)