Synthesis, Expression and Biological Activity of the Prohormone for Gastrin Releasing Peptide (ProGRP)
http://www.100md.com
《内分泌学杂志》
Department of Surgery, Austin Health (C.D., O.P., S.L., G.S.B., A.S.) and Department of Medicine (A.S.G.), Western Hospital, University of Melbourne, Victoria 3084, Australia
Abstract
Gastrin-releasing peptide (GRP) has a widespread distribution and multiple stimulating effects on endocrine and exocrine secretions and metabolism. The prohormone for GRP (ProGRP, 125 amino acids) is processed to the amidated, biologically active end products GRP1–27 and GRP18–27. Amidated forms of GRP are putative autocrine or paracrine growth factors in a number of cancers including colorectal cancer. However, the potential role and biological activity of proGRP has not been investigated. Using a newly developed antisera directed to the N terminus of human proGRP, proGRP immunoreactivity was detected in all of the endometrial, prostate, and colon cancer cell lines tested and in nine of 10 resected colorectal carcinomas. However, no amidated forms were detected, suggesting an attenuation of processing in tumors. Recombinant proGRP was expressed as a His-tag fusion protein and purified by metal affinity chromatography and HPLC. ProGRP stimulated proliferation of a colon cancer cell line and activated MAPK, but unlike GRP18–27amide had no effect on inositol phosphate production. ProGRP did not compete with iodinated bombesin in binding assays on Balb-3T3 cells transfected with the known GRP receptors, GRP-R or BRS-3. We conclude that proGRP is present in a number of cancer cell lines and in resected colorectal tumors and is biologically active. Our results suggest that antagonists to GRP precursors rather than the amidated end products should be developed as a treatment for colorectal and other cancers that express proGRP-derived peptides.
Introduction
GASTRIN-RELEASING PEPTIDE (GRP) was isolated from the porcine stomach and named for its first identified bioactivity, gastrin release (1). However, it is now known to perform many other functions including stimulation of the secretion of a variety of gastrointestinal hormones and pancreatic enzymes, and control of intestinal transit, metabolism and behavior (reviewed in Refs.2 and 3). GRP is a potent mitogen for a number of tumor types including pancreatic, small cell lung, prostate, renal, breast, and colon cancers (4, 5, 6). In agreement with these multiple roles, GRP has a widespread distribution, and significant amounts have been identified in the central nervous system and throughout the gastrointestinal tract, and in tumors of the lung, prostate and colon (2, 4, 5).
There are two known mammalian receptors for GRP (7, 8). Both GRP-R and BRS-3 are members of the G protein seven-transmembrane receptor superfamily. GRP-R has a high affinity, whereas BRS-3 has a low affinity, for GRP. No naturally occurring ligand for the BRS-3 receptor has been identified (reviewed in Ref.9).
GRP is synthesized as a large precursor molecule that is converted to proGRP1–125 by cleavage of the N-terminal signal sequence (Fig. 1). ProGRP is processed further by endoproteolytic cleavage after K29K30 by a yet unidentified endopeptidase; however, it has been postulated that this processing is mediated by prohormone convertase PC2 (10). Carboxypeptidase B-like activity removes the basic residues allowing peptidyl -amidating mono-oxygenase to act on the glycine-extended intermediates to yield C-terminally amidated GRP1–27. Further endoproteolytic cleavage after R17 releases the decapeptide GRP18–27 (2). The C terminus of GRPamide is identical with the C-terminal sequence of the frog peptide bombesin, which has full agonist activity at the GRP-R.
It was originally thought that only amidated forms of GRP were biologically active. However, we recently reported that GRP18–27amide and GRP18–27gly have comparable potencies for stimulating migration and proliferation in a colorectal cancer (CRC) cell line. Both GRPamide and GRPgly acted through the same receptor, GRP-R (6). Similar results were reported with a glycine-extended bombesin derivative in fibroblasts and a pancreatic cell line (11).
Although these studies with GRPgly establish the principle that nonamidated forms of GRP are biologically active, the question of biological activity of high molecular weight intermediates like proGRP has not been addressed. This is an important issue because larger nonamidated forms of GRP have been identified in the reproductive tracts of human and sheep (12, 13) and are also present in the tumors and circulation of patients with small cell lung carcinoma (SCLC) (14, 15, 16, 17). Assays for circulating proGRP have been developed for diagnosis and treatment monitoring of SCLC (18) and more recently as a tumor marker for prostate and medullary thyroid cancer (19, 20). To investigate the possibility that proGRP may be biologically active, we have prepared recombinant human proGRP1–125 and tested this peptide in a number of biological assays. We have also generated an antibody directed against the N terminal of proGRP to define the pattern of expression of proGRP in cell lines derived from prostate, endometrial and colon cancers, and in CRC resected from patients.
Materials and Methods
RIA
proGRP.
A new antiserum was generated to measure proGRP. The peptide (VPLPAGGGTVY), which corresponds to human proGRP1–10 with an additional C-terminal tyrosine (tyr11-proGRP1–11), was custom synthesized by Auspep Pty Ltd. (Melbourne, Australia) with sequence and purity (>95%) determined by mass spectrometry and HPLC. This peptide was used for conjugation and radiolabeling. For antibody production, the C terminus of the peptide was conjugated to keyhole limpet hemocyanin using bis-diazotized-benzidine (performed by Auspep). The conjugate was emulsified in Freund’s complete adjuvant (Sigma-Aldrich, Castle Hill, Australia) in a 1:1 ratio (vol/vol), and 2 ml of the emulsified conjugate (which contained about 100 nmol of peptide) was injected sc into each of four New Zealand White Cross rabbits. Rabbit immunization and animal care was approved by the Austin Health Animal Ethics committee (Ethics no. 2001/1114). For subsequent booster injections (six to eight weekly intervals), the conjugate was emulsified using Freund’s incomplete adjuvant (Sigma-Aldrich) in a 1:1 ratio (vol/vol). After several booster injections, one rabbit (N798) generated useful antibodies against the proGRP with maximum binding achieved after 42 wk and four booster injections. The specificity of the antiserum was determined by constructing a number of standard curves. proGRP1–125 was detected to a similar extent as proGRP1–11, whereas proGRP31–125 and GRP18–27amide were undetectable confirming that N798 was directed to the N terminus of proGRP. Because this antiserum detects both intact proGRP and N-terminal fragments of proGRP, the results are expressed as proGRP N-terminal immunoreactivity (proGRP NTI). The assay was performed using 0.02 M veronal buffer (1 ml) containing 0.1% BSA and 2 μM NaN3 (pH 8.7). The antibody was used at a final dilution of 1:6000 with 125I-Tyr11-proGRP1–11, prepared by the iodogen method, and purified by reverse-phase HPLC, as the label. Charcoal (2.5%) was added and bound and free peptide were separated by centrifugation. The standard curve was constructed using tyr11-proGRP1–11 (2–2000 fmol/ml) with an ID50 of 200 fmol/tube. The intra- and interassay coefficients of variation were 4% and 14%, respectively.
GRPamide.
Details of the RIA have been published (12). The assay used antiserum R40, which cross-reacts equally with the amidated forms of bombesin, GRP1–27 and GRP18–27 but not with nonamidated forms such GRPgly. This antisera detects GRP from sheep, rat, and human. 125I-Tyr4-bombesin was prepared using iodogen (Pierce Chemical Co., Rockford, IL) followed by reduction with dithiothreitol and purification by reverse-phase HPLC. The standard curve was constructed using bombesin (2–2000 fmol/ml) with an ID50 of 100 fmol/tube.
Preparation of cancer cells and CRC for RIA
Prostate, endometrial, and colon cancer cell lines (detailed in Table 1) were grown in DMEM (Invitrogen, Mulgrave, Australia) supplemented with 10% fetal calf serum (FCS). The cells were seeded into a Petri dish and allowed to grow for 24 h before being serum starved for 24 h. The cells were then lysed with boiling water and the lysate spun down to remove cell debris (12, 21). This lysate was concentrated with a Sep-Pak (Millipore-Waters, Lane Cove, Australia) and the elution fraction from the Sep-Pak was evaporated under a stream of air. After resuspension in assay buffer concentrations of both N-terminal proGRP and amidated GRPs were measured by RIA. Resected CRC which had been stored at –70 C were extracted with boiling water and the pellet was reextracted with boiling acetic acid (3%). Both water and acid extracts were assayed by RIA.
Cell lines for biological activity
DLD-1 cells (ATCC, Manassas, VA), and GRP-R- and BRS-3-transfected Balb-3T3 cells (donated by Dr. J. Battey, National Institutes of Health, Bethesda, MD) were grown in monolayer cultures in DMEM supplemented with 10% FCS and 300 μg/ml of G418 for the transfected cells in an atmosphere of 95% air and 5% carbon dioxide at 37 C. Cultures were passaged at 2- to 3-d intervals to maintain the cells at subconfluent densities.
Construction of proGRP fusion proteins
A pGEX2TH expression plasmid (Amersham, Baulkham Hills, Australia) encoding a fusion protein between glutathione S-transferase and human proGRP was constructed and expressed in BL21 Escherichia coli cells and purified from bacterial lysates by binding to glutathione-agarose. However, treatment of the fusion protein with enterokinase or thrombin resulted in multiple cleavages within the proGRP sequence (data not shown).
To produce a full-length proGRP1–125, the proGRP sequence was subcloned from the pGEX2TH plasmid into the His tag expression vector proEXHtb (Invitrogen) by restriction digestion with BamHI and HindIII. This DNA was then electroporated into electrocompetent E. coli strain DH5. Clones with the correct insert were selected by restriction digestion. The sequence consisted of the His tag joined to proGRP1–125 by a seven-amino acid linker and a recombinant tobacco etch virus protease (rTEV) cleavage site and an enterokinase cleavage site (Fig. 2). DNA sequencing of positive clones revealed a mutation in codon 14 (ATG to GTG), which resulted in a conservative mutation of amino acid 14 from Met to Val.
Expression of proGRP His tag fusion protein
The His-proGRP1–125 fusion protein was expressed in E. coli BL21 cells and expression was induced by treatment with isopropylthiogalactoside for 4 h. The His-proGRP1–125 fusion protein was purified from sarkosyl lysates by binding to nickel-iminodiacetic acid (Ni-IDA) agarose. The protein content of samples at various stages of the purification was analyzed by SDS-PAGE on 15% gels.
Reverse-phase HPLC
The recombinant human proGRP1–125 prepared above was applied to a C18 μBondapak column (8 x 100 mm; Waters Associates, Milford, MA), which had been equilibrated with 0.05% trifluoroacetic acid. The His-proGRP1–125 fusion protein was eluted with a gradient from 0–70% acetonitrile in 0.05% trifluoroacetic acid at a flow rate of 1 ml/min. Fractions of 1 ml were collected and dried on a Speed Vac (Savant, Hicksville, NY) for RIA, mass spectrometry, and biological activity assays. In a separate chromatogram, a water extract of a CRC was run under similar conditions and the concentrations in each fraction measured by RIA.
Mass spectrometry and N-terminal sequencing
Mass spectrometry (Bruker AutoFlex MALDI-TOF, Billerica, MA) was performed on the peak immunoreactive HPLC fraction to determine the molecular weight of the recombinant protein. N-terminal sequencing was done on the peak HPLC fraction to confirm the amino acid sequence.
Proliferation studies
DLD-1 cells were grown in media supplemented with 10% FCS, seeded in 24-well plates at a density of 50,000 cells/well, and grown for 24 h. The cells were then serum starved for 24 h, before being treated with different concentrations of HPLC-purified His-proGRP1–125, GRP18–27amide (Auspep) or the GRP-R antagonist, (D-Phe6, Leu-NHEt13, des-Met14)-bombesin6–14 (Bachem AG, Bubendorf, Switzerland), in media supplemented with 0.2% FCS for 16 h. As determined by HPLC, His-proGRP1–125 was stable with only a single form eluting in the position of His-proGRP1–125 detected at the end of the incubation. 3H-thymidine was added at a concentration of 0.5 μCi /well and incubated for 4 h at 37 C before the medium was discarded. The cells were washed and then incubated with 5% trichloroacetic acid at 4 C for 30 min, after which time the cells were washed with 95% ethanol and dried. The cells were lysed with 1M sodium hydroxide and the lysate counted in a counter.
Inositol phosphate production
DLD-1 cells were grown in media supplemented with 10% FCS, seeded in a 24-well plate at a density of 50,000 cells/well and grown for 24 h. The cells were serum starved and labeled with 3H myo-inositol (Amersham) for 24 h and then incubated with 20 mM LiCl for 1 h before being treated with different concentrations of His-proGRP1–125, GRP18–27amide, and GRP-R antagonist for 1 h. A stop solution (1:2000 hydrochloric acid in ethanol) was added to the cells before the lysate was loaded onto an AG1X-8 (Bio-Rad, Regents Park, Australia) column. The lysate was allowed to run through, and the column was washed with distilled water and then 40 mM ammonium formate. Inositol phosphates were eluted with 5 ml 1 M ammonium formate. Scintillation fluid (5 ml PicoFluor 40; PerkinElmer, Rowville, Australia) was added to the eluant and the mixture was counted in a -counter.
MAPK activation
Antibodies against total and phosphorylated p42/44 MAPKs were obtained from Cell Signaling Technology (Genesearch, Arundel, Australia). Cells were grown in 10-cm Petri dishes in DMEM containing 10% FCS until 90% confluent. After serum starvation, the cells were stimulated for 0, 3, 5, and 10 min with 10 nM His-proGRP1–125. Maximum stimulations was observed at 5 min and subsequent experiments with 10 nM His-proGRP1–125 or GRP18–27amide in the presence or absence of the GRP-R antagonist, all diluted in serum-free DMEM, were performed at the 5-min time point. Cells were lysed with RIPA buffer [1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 0.2 mM sodium orthovanadate, 0.5 mM dithiothreitol, and protease inhibitors (Sigma-Aldrich) in 20 mM Tris, 150 mM NaCl (pH 7.6)]. Thirty micrograms of total protein lysates were then mixed with loading buffer, denatured, and separated by electrophoresis (10% SDS-PAGE). Proteins were transferred onto a nitrocellulose membrane using a wet transfer system (Bio-Rad). Membranes were then incubated with the appropriate primary antibodies to phosphorylated or total p42/44 MAPK, and detection performed with alkaline phosphatase-coupled antirabbit IgG followed by incubation with a 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium solution (pH 9.2) (Sigma-Aldrich). Membranes were scanned using a Hewlett-Packard ScanJet 5200C, and densitometric analysis of protein bands was performed with Fuji BAS software. Band densities were determined by taking the ratio of densities of phosphorylated to total MAPK.
Binding assays
GRP-R- and BRS-3-transfected Balb-3T3 cells were grown in media supplemented with 10% FCS, seeded in a 12-well plate at a density of 200,000 cells/well, and grown for 24 h. The cells were then serum starved for 24 h before being treated with GRP18–27amide, His-proGRP1–125, or a synthetic BRS-3 agonist, Tyr6-ala11Phe13Nle14Bn6–14, at a concentration of 1 μM. The cells were preincubated for 15 min at 37 C, and approximately 50,000 cpm of 125I Tyr-4-bombesin (Amersham) or 125I-BRS-3 agonist (prepared by the iodogen method without the reduction step, followed by reverse-phase HPLC, as detailed above for the bombesin label) was added to each well. The cells were incubated with mixing at 37 C for 45 min, washed twice with ice-cold PBS/2% BSA, and then lysed with 1 M sodium hydroxide. The lysate was collected and counted in a -counter.
Statistics
Results are expressed as the means ± SE of at least three separate experiments. Results were analyzed by one-way ANOVA followed by Dunnett’s or Bonferroni methods. Differences with P values of < 0.05 were considered significant.
Results
Detection of proGRP NTI in cancer cell lines and CRC
Prostate, endometrial, and colon cancer cell lines and resected CRC were assayed for proGRP and GRPamide. Other than the Ishikawa endometrial cell line, GRPamide was not detected in any of the cell lines or tumors (Tables 1 and 2). In contrast, immunoreactivity measured with the antisera directed to the N terminus of proGRP1–125 (N798) was present in all cell lines tested, at concentrations ranging from 17–94 fmol/106 cells. Nine of 10 of the resected CRC also contained substantial amounts of proGRP NTI with a maximum concentration of 16.1 pmol/g (Table 2). Reverse-phase HPLC was performed on a CRC water extract and the fractions measured with the N-terminal antisera (Fig. 3). Based on coelution with GRP1–27amide standard and the absence of GRPamide immunoreactivity, the N-terminal immunoreactivity eluting at fraction 18 is nonamidated GRP1–27. The nature of the more hydrophobic species eluting at fraction 64 remains to be determined but based on its elution position is not proGRP1–125.
Purification and characterization of recombinant proGRP his tag fusion protein
To generate full-length proGRP1–125, a plasmid encoding a His6-tagged-proGRP1–125 fusion protein was constructed as described in Materials and Methods. As well as the His6 tag, the fusion protein contained a seven-amino acid linker, an rTEV cleavage site and an enterokinase cleavage site before the proGRP sequence (Fig. 2). The fusion protein was expressed in E. coli and purified from bacterial lysates by binding to Ni-IDA-agarose. No cleavage was observed when the fusion protein was treated with enterokinase while bound to the beads. After elution with imidazole, and removal of the imidazole by dialysis enterokinase treatment resulted in incomplete cleavage at multiple sites within the proGRP sequence (data not shown). No cleavage was observed when rTEV protease, which was specific for the rTEV cleavage site, was used.
The intact His-proGRP1–125 fusion protein was therefore isolated from bacterial lysates (Fig. 4) by binding to Ni-IDA agarose. The fusion protein was eluted from the Ni-IDA agarose beads with 250 mM imidazole and purified by reverse-phase HPLC (Fig. 5). The absorbance peak in fraction 24 matched very well with the peak of immunoreactivity measured with the N terminally directed antiserum N798.
The identity of the peptide that corresponded to the absorbance and immunoreactivity peak was confirmed by mass spectrometry. Fraction 24 from reverse-phase HPLC contained a species with a molecular mass of 17,495 Da, which does not correspond with the theoretical predicted mass of 17,641 Da. It appeared from fragmentation in the low molecular weight region of the spectra that the N-terminal methionine was absent from the recombinant molecule (Fig. 2). This conclusion was confirmed by N-terminal sequencing. The small discrepancy remaining between the predicted (17,510) and experimental masses (17,495) could be due to a modification of the N-terminal serine after the methionine is cleaved. We conclude that the major species of recombinant proGRP purified by reverse-phase HPLC consists of residues 1–125 together with the His6 tag and linker sequence shown in Fig. 2.
Biological activity
Proliferation assays.
Recombinant human His-proGRP1–125 at a concentration of 10 nM stimulated proliferation of the colon cancer cell line DLD-1 (Fig. 6). There was no significant difference in the stimulation of proliferation between His-proGRP1–125 and GRP18–27amide. The GRP-R antagonist (D-Phe6, Leu-NHEt13, des-Met14)-bombesin6–14 blocked the proliferative effects of both His-proGRP1–125 and GRPamide.
Inositol phosphate production.
The production of inositol phosphates by the colon cancer cell line DLD-1 was monitored by ion-exchange chromatography. Recombinant human His-proGRP1–125 at concentrations of 1 or 10 nM did not stimulate inositol phosphate production significantly (Fig. 7). Addition of the GRP-R antagonist (D-Phe6, Leu-NHEt13, des-Met14)-bombesin6–14 had no effect. In contrast, GRP18–27amide at a concentration of 1 nM significantly stimulated inositol phosphate production, and the stimulation was reversed by the addition of the GRP-R antagonist (Fig. 7).
MAPK activation.
Phosphorylation of both the 42- and 44-kDa forms of MAPK was significantly stimulated (range 30- to 80-fold) by treatment with 10 nM GRP or 10 nM His-proGRP1–125 in DLD-1 colorectal carcinoma cells. This stimulation was reversed by the addition of 50 nM GRP-R antagonist in the case of GRP, but not His-proGRP1–125 (Fig 8).
Binding assays
Recombinant human His-proGRP1–125 at a concentration of 1 μM did not compete with radiolabeled bombesin for binding to the GRP-R on GRP-R-transfected Balb-3T3 cells (Fig. 9A). GRP18–27 (1 μM) used as a positive control nearly abolished binding. Recombinant human His-proGRP1–125 at a concentration of 1 μM did not compete with a radiolabeled BRS-3 agonist for binding to BRS-3 on BRS-3-transfected Balb-3T3 cells, whereas unlabeled BRS-3 agonist abolished binding (Fig. 9B).
Discussion
The autocrine hypothesis for cancer growth postulates that a cell produces both a growth factor and its cognate receptor, which interact resulting in proliferation. GRPamide is the prototypical autocrine growth factor. This description was originally based on the detection of GRPamide and its cognate receptor, and on the antiproliferative effect of antibodies directed against GRPamide, in SCLC (22). However, GRPamide is also a potent mitogen for several other types of carcinomas including colon, pancreas, prostate, and breast (reviewed in Refs.4 and 5). With the exception of recent studies using GRPgly (6, 11), analysis of proGRP-derived peptides has been confined to the amidated forms GRP1–27 and GRP18–27. We describe here the production of recombinant His-proGRP1–125, the demonstration that it is biologically active, and the observation that nonamidated proGRP-derived peptides are produced in various cancer cell lines and resected CRC.
A RIA for the measurement of proGRP and proGRP N-terminal fragments was developed. The assay used a new antisera (N798) produced by injection of tyr11-proGRP1–11 conjugated to KLH into New Zealand White rabbits. We showed significant concentrations of proGRP NTI in cell lines derived from prostate, endometrial, and colon cancers as well as in resected CRC. However, GRPamide was detected in only one endometrial cancer cell line but not in prostate, CRC cell lines or resected CRC (Tables 1 and 2). The absence of GRPamide was unexpected because GRP mRNA is expressed in the majority of CRC cell lines and tumors (23, 24, 25), and it had been assumed that the proGRP would be processed at least in part to amidated forms as occurs in SCLC (17, 22, 26). Indeed immunohistochemical studies of resected CRC using an antisera against GRPamide gave positive results (24). However, this antisera may also detect nonamidated forms at the high antisera concentrations required for immunohistochemistry. To our knowledge, the present report is the first to use RIA to quantify the amount of proGRP-derived peptides in CRC.
In contrast to GRPamide, significant amounts of proGRP NTI were measured in all cell lines and in resected CRC. Reverse-phase HPLC of a resected CRC extract and N-terminal RIA demonstrated two major species, one probably being nonamidated GRP1–27 and a second more hydrophobic form that did not coelute with proGRP1–125 or His-proGRP1–125. The precise nature of the proGRP NTI in the tumors and cell lines will require further study using larger amounts of material. Until the current study, characterization of nonamidated forms of proGRP-derived peptides has been confined to SCLC tumors and cell lines. Using a similar N-terminal-directed antiserum and Western blotting, Cuttitta et al. (26) reported the presence of multiple forms of proGRP peptides with a molecular weight range of 14–18K, consistent with the presence of proGRP1–125 and N-terminal cleavage products. Substantial concentrations of the C-terminal of proGRP in SCLC have also been detected using a variety of C-terminal-directed antisera (14, 15, 16, 17). The presence of C-terminal immunoreactive forms of proGRP in CRC has not been reported. Taken together, we can conclude that proGRP and processing intermediates are expressed in a variety of tumor types and that processing to the amidated forms is incomplete as has been reported for a number of other peptides in tumors (27, 28).
Having established that proGRP NTI is expressed in tumors and tumor-derived cell lines, it was important to determine whether proGRP was bioactive. We therefore produced recombinant human proGRP1–125 in E. coli. In initial studies, proGRP was synthesized as a glutathione S-transferase fusion protein; however, this fusion protein was not cleaved appropriately to the full-length protein. Experiments were therefore performed using the purified His-proGRP1–125 fusion protein. His-tagged recombinant proteins have been shown in many instances to retain their normal biological function (29, 30).
Characterization of the His-proGRP1–125 fusion protein included DNA sequencing of the vector, RIA, mass spectroscopy, and N-terminal sequencing. DNA sequencing of the vector showed a mutation that resulted in an amino acid change from Met to Val at position 14. Because this is a conservative change, it is unlikely to be significant. Although a definitive answer will require comparison with the wild type, we have recently shown that biological activity of proGRP18–125 is similar to His-proGRP1–125 (our unpublished observations). Mass spectroscopic analysis revealed a small discrepancy between the expected theoretical mass and the experimental mass. Careful examination of the spectra suggested that the N-terminal Met of the fusion protein had been removed, a conclusion confirmed by N-terminal sequencing. Several studies have shown that in a significant fraction of E. coli cytosolic proteins, the N-terminal methionine is removed when the side chain of the adjacent amino acid residue is relatively small as is the case with Ser in the present instance (Fig. 2) (30, 31).
Recombinant His-proGRP1–125 was biologically active. Proliferation of the colon cancer cell line DLD-1 was stimulated significantly by His-proGRP1–125 at a concentration of 10 nM. A similar effect was seen with GRP18–27amide. Both GRP18–27amide and His-proGRP1–125 stimulated phosphorylation of MAPK. Interestingly, His-proGRP1–125 had no effect on inositol phosphate production, although GRP18–27amide stimulated production significantly.
The identity of the receptor involved in the biological activity of His-proGRP1–125 has not been established. Neither GRP-R or BRS-3 appears to be involved because His-proGRP1–125 did not compete with labeled bombesin or a labeled BRS-3 agonist for binding to Balb-3T3 cells transfected with GRP-R or BRS-3, respectively. Further evidence against the involvement of the GRP-R is the observation that the GRP-R antagonist (D-Phe6, Leu-NHEt13, des-Met14)-bombesin6–14 did not affect stimulation of MAPK by His-proGRP1–125, but inhibited the effect of GRP18–27amide in the same assay. The fact that His-proGRP1–125 had no effect on inositol phosphate production, whereas GRP18–27amide significantly stimulated production, is consistent with the same conclusion. On the other hand the GRP-R antagonist reversed the proliferative effect of His-proGRP1–125. Resolution of this discrepancy will require isolation and characterization of the proGRP receptor, and further elucidation of its downstream signaling pathways.
In summary, we have described for the first time the production and purification of proGRP and shown it to be biologically active in vitro. We have also shown the presence of nonamidated forms of proGRP in a panel of cancer cell lines and resected colorectal tumors. The results support the possibility of using antisera or antagonists to precursor forms of GRP as a treatment for colorectal and other cancers that express proGRP-derived peptides. Finally, we provide another example for the production from a single precursor of multiple peptides with independent receptors and different bioactivities (32, 33, 34).
Footnotes
This work was supported by the National Health and Medical Research Council of Australia, the Austin Hospital Medical Research Foundation, and Western Hospital. We thank Dr. Mustafa Ayan from LaTrobe University (Bundoora, Victoria, Australia) for the mass spectroscopy, and Dr. Lindsay Sparrow from Commonwealth Scientific and Industrial Research Organisation (Parkville, Victoria, Australia) for the N-terminal sequencing. Tyr6-ala11Phe13Nle14Bn6–14 was donated by Professor David Coy (Tulane University, New Orleans, LA), and the proGRP-containing plasmid by Professor Jim Battey (National Institutes of Health, Bethesda, MD).
First Published Online October 13, 2005
Abbreviations: CRC, Colorectal cancer; FCS, fetal calf serum; GRP, gastrin-releasing peptide; Ni-IDA, nickel-iminodiacetic acid; NTI, N-terminal immunoreactivity; ProGRP, prohormone for GRP; rTEV, recombinant tobacco etch virus protease; SCLC, small cell lung carcinoma; SDS, sodium dodecyl sulfate.
Accepted for publication October 3, 2005.
References
McDonald TJ, Jornvall H, Nilsson G, Vagne M, Ghatei M, Bloom SR, Mutt V 1979 Characterization of a gastrin releasing peptide from porcine non-antral gastric tissue. Biochem Biophys Res Commun 90:227–233
Bunnett N 1994 Gastrin releasing peptide. In: Walsh JH, Dockray GJ, eds. Gut peptides. New York: Raven Press; 423–455
Yamada K, Wada E, Wada K 2000 Bombesin-like peptides: studies on food intake and social behaviour with receptor knock-out mice. Ann Med 32:519–529
Preston SR, Miller GV, Primrose JN 1996 Bombesin-like peptides and cancer. Crit Rev Oncol Hematol 23:225–238
Moody TW, Chan D, Fahrenkrug J, Jensen RT 2003 Neuropeptides as autocrine growth factors in cancer cells. Curr Pharm Des 9:495–509
Patel O, Dumesny C, Giraud AS, Baldwin GB, Shulkes A 2004 Stimulation of proliferation and migration of a colorectal cancer cell line by amidated and glycine-extended gastrin-releasing peptide via the same receptor. Biochem Pharmacol 68:2129–2142
Sano H, Feighner SD, Hreniuk DL, Iwaasa H, Sailer AW, Pan J, Reitman ML, Kanatani A, Howard AD, Tan CP 2004 Characterization of the bombesin-like peptide receptor family in primates. Genomics 84:139–146
Moody TW, Merali Z 2004 Bombesin-like peptides and associated receptors within the brain: distribution and behavioral implications. Peptides 25:511–520
Kroog GS, Jensen RT, Battey JF 1995 Mammalian bombesin receptors. Med Res Rev 15:389–417
Rounseville MP, Davis TP 2000 Prohormone convertase and autocrine growth factor mRNAs are coexpressed in small cell lung carcinoma. J Mol Endocrinol 25:121–128
Oiry C, Pannequin J, Bernad N, Artis AM, Galleyrand JC, Devin C, Cristau M, Fehrentz JA, Martinez J 2000 A synthetic glycine-extended bombesin analogue interacts with the GRP/bombesin receptor. Eur J Pharmacol 403:17–25
Giraud A, Whitley J, Shulkes A, Parker L 1996 The pregnant ovine endometrium constitutively expresses and secretes a highly stable bombesin-like peptide, which shares C-terminal sequence but differs structurally from gastrin-releasing peptide. Biochim Biophys Acta 5:189–197
Whitley JC, Giraud AS, Shulkes A 1996 Expression of gastrin-releasing peptide (GRP) and GRP receptors in the pregnant human uterus at term. J Clin Endocrinol Metab 81:3944–3950
Reeve Jr JR, Cuttitta F, Vigna SR, Heubner V, Lee TD, Shively JE, Ho FJ, Fedorko J, Minna JD, Walsh JH 1989 Multiple gastrin-releasing peptide gene-associated peptides are produced by a human small cell lung cancer line. J Biol Chem 264:1928–1932
Hamid QA, Addis BJ, Springall DR, Ibrahim NB, Ghatei MA, Bloom SR, Polak JM 1987 Expression of the C-terminal peptide of human pro-bombesin in 361 lung endocrine tumours, a reliable marker and possible prognostic indicator for small cell carcinoma. Virchows Arch A Pathol Anat Histopathol 411:185–192
Holst JJ, Hansen M, Bork E, Schwartz TW 1989 Elevated plasma concentrations of C-flanking gastrin-releasing peptide in small-cell lung cancer. J Clin Oncol 7:1831–1838
Vangsted AJ, Schwartz TW 1990 Production of gastrin-releasing peptide-(18–27) and a stable fragment of its precursor in small cell lung carcinoma cells. J Clin Endocrinol Metab 70:1586–1593
Sunaga N, Tsuchiya S, Minato K, Watanabe S, Fueki N, Hoshino H, Makimoto T, Ishihara S, Saito R, Mori M 1999 Serum pro-gastrin-releasing peptide is a useful marker for treatment monitoring and survival in small-cell lung cancer. Oncology 57:143–148
Ide A, Ashizawa K, Ishikawa N, Ishii R, Ando T, Abe Y, Sera N, Usa T, Tominaga T, Ejima E, Nakashima M, Ito K, Eguchi K 2001 Elevation of serum pro-gastrin-releasing peptide in patients with medullary thyroid carcinoma and small cell lung carcinoma. Thyroid 11:1055–1061
Yashi M, Muraishi O, Kobayashi Y, Tokue A, Nanjo H 2002 Elevated serum progastrin-releasing peptide (31–98) in metastatic and androgen-independent prostate cancer patients. Prostate 51:84–97
Dumesny C, Whitley JC, Baldwin GS, Giraud AS, Shulkes A 2004 Developmental expression and biological activity of gastrin-releasing peptide and its receptors in the kidney. Am J Physiol Renal Physiol 287:F578–F585
Cuttitia F, Carney DJM 1985 Bombesin-like peptides can function as autocrine growth factors in human lung carcinoma cells. Nature 316:823–826
Carroll RE, Ostrovskiy D, Lee S, Danilkovich A, Benya RV 2000 Characterization of gastrin-releasing peptide and its receptor aberrantly expressed by human colon cancer cell lines. Mol Pharmacol 58:601–607
Carroll RE, Matkowskyj KA, Chakrabarti S, McDonald TJ, Benya RV 1999 Aberrant expression of gastrin-releasing peptide and its receptor by well-differentiated colon cancers in humans. Am J Physiol 276:G655–G665
Chave HS, Gough AC, Palmer K, Preston SR, Primrose JN 2000 Bombesin family receptor and ligand gene expression in human colorectal cancer and normal mucosa. Br J Cancer 82:124–130
Cuttitta F, Fedorko J, Gu JA, Lebacq-Verheyden AM, Linnoila RI, Battey JF 1988 Gastrin-releasing peptide gene-associated peptides are expressed in normal human fetal lung and small cell lung cancer: a novel peptide family found in man. J Clin Endocrinol Metab 67:576–583
Rehfeld JF, Bardram L, Blanke S, Bundgaard JR, Friis-Hansen L, Hilsted L, Johnsen AH, Kofod M, Luttichau HR, Monstein HJ, Nielson C, Nielson FC, Paloheimo LI, Pederson K, Pildal J, Ramlau J, Van Solinge WW, Thorup U, Odum L 1993 Peptide hormone processing in tumours: biogenetic and diagnostic implications. Tumour Biol 14:174–183
Ciccotosto GD, McLeish A, Hardy KJ, Shulkes A 1995 Expression, processing, and secretion of gastrin in patients with colorectal carcinoma. Gastroenterology 109:1142–1153
Hochuli E 1990 Purification of recombinant proteins with metal chelate adsorbent. Genet Eng (NY) 12:87–98
Taussig R, Quarmby LM, Gilman AG 1993 Regulation of purified type I and type II adenylylcyclases by G protein subunits. J Biol Chem 268:9–12
Boissel JP, Kasper TJ, Bunn HF 1988 Cotranslational amino-terminal processing of cytosolic proteins. Cell-free expression of site-directed mutants of human hemoglobin. J Biol Chem 263:8443–8449
Raffin-Sanson ML, de Keyzer Y, Bertagna X 2003 Proopiomelanocortin, a polypeptide precursor with multiple functions: from physiology to pathological conditions. Eur J Endocrinol 149:79–90
Drucker DJ 2002 Biological actions and therapeutic potential of the glucagon-like peptides. Gastroenterology 122:531–544
Aly A, Shulkes A, Baldwin GS 2004 Gastrins, cholecystokinins and gastrointestinal cancer. Biochim Biophys Acta 1704:1–10(Chelsea Dumesny, Oneel Patel, Shamilah L)
Abstract
Gastrin-releasing peptide (GRP) has a widespread distribution and multiple stimulating effects on endocrine and exocrine secretions and metabolism. The prohormone for GRP (ProGRP, 125 amino acids) is processed to the amidated, biologically active end products GRP1–27 and GRP18–27. Amidated forms of GRP are putative autocrine or paracrine growth factors in a number of cancers including colorectal cancer. However, the potential role and biological activity of proGRP has not been investigated. Using a newly developed antisera directed to the N terminus of human proGRP, proGRP immunoreactivity was detected in all of the endometrial, prostate, and colon cancer cell lines tested and in nine of 10 resected colorectal carcinomas. However, no amidated forms were detected, suggesting an attenuation of processing in tumors. Recombinant proGRP was expressed as a His-tag fusion protein and purified by metal affinity chromatography and HPLC. ProGRP stimulated proliferation of a colon cancer cell line and activated MAPK, but unlike GRP18–27amide had no effect on inositol phosphate production. ProGRP did not compete with iodinated bombesin in binding assays on Balb-3T3 cells transfected with the known GRP receptors, GRP-R or BRS-3. We conclude that proGRP is present in a number of cancer cell lines and in resected colorectal tumors and is biologically active. Our results suggest that antagonists to GRP precursors rather than the amidated end products should be developed as a treatment for colorectal and other cancers that express proGRP-derived peptides.
Introduction
GASTRIN-RELEASING PEPTIDE (GRP) was isolated from the porcine stomach and named for its first identified bioactivity, gastrin release (1). However, it is now known to perform many other functions including stimulation of the secretion of a variety of gastrointestinal hormones and pancreatic enzymes, and control of intestinal transit, metabolism and behavior (reviewed in Refs.2 and 3). GRP is a potent mitogen for a number of tumor types including pancreatic, small cell lung, prostate, renal, breast, and colon cancers (4, 5, 6). In agreement with these multiple roles, GRP has a widespread distribution, and significant amounts have been identified in the central nervous system and throughout the gastrointestinal tract, and in tumors of the lung, prostate and colon (2, 4, 5).
There are two known mammalian receptors for GRP (7, 8). Both GRP-R and BRS-3 are members of the G protein seven-transmembrane receptor superfamily. GRP-R has a high affinity, whereas BRS-3 has a low affinity, for GRP. No naturally occurring ligand for the BRS-3 receptor has been identified (reviewed in Ref.9).
GRP is synthesized as a large precursor molecule that is converted to proGRP1–125 by cleavage of the N-terminal signal sequence (Fig. 1). ProGRP is processed further by endoproteolytic cleavage after K29K30 by a yet unidentified endopeptidase; however, it has been postulated that this processing is mediated by prohormone convertase PC2 (10). Carboxypeptidase B-like activity removes the basic residues allowing peptidyl -amidating mono-oxygenase to act on the glycine-extended intermediates to yield C-terminally amidated GRP1–27. Further endoproteolytic cleavage after R17 releases the decapeptide GRP18–27 (2). The C terminus of GRPamide is identical with the C-terminal sequence of the frog peptide bombesin, which has full agonist activity at the GRP-R.
It was originally thought that only amidated forms of GRP were biologically active. However, we recently reported that GRP18–27amide and GRP18–27gly have comparable potencies for stimulating migration and proliferation in a colorectal cancer (CRC) cell line. Both GRPamide and GRPgly acted through the same receptor, GRP-R (6). Similar results were reported with a glycine-extended bombesin derivative in fibroblasts and a pancreatic cell line (11).
Although these studies with GRPgly establish the principle that nonamidated forms of GRP are biologically active, the question of biological activity of high molecular weight intermediates like proGRP has not been addressed. This is an important issue because larger nonamidated forms of GRP have been identified in the reproductive tracts of human and sheep (12, 13) and are also present in the tumors and circulation of patients with small cell lung carcinoma (SCLC) (14, 15, 16, 17). Assays for circulating proGRP have been developed for diagnosis and treatment monitoring of SCLC (18) and more recently as a tumor marker for prostate and medullary thyroid cancer (19, 20). To investigate the possibility that proGRP may be biologically active, we have prepared recombinant human proGRP1–125 and tested this peptide in a number of biological assays. We have also generated an antibody directed against the N terminal of proGRP to define the pattern of expression of proGRP in cell lines derived from prostate, endometrial and colon cancers, and in CRC resected from patients.
Materials and Methods
RIA
proGRP.
A new antiserum was generated to measure proGRP. The peptide (VPLPAGGGTVY), which corresponds to human proGRP1–10 with an additional C-terminal tyrosine (tyr11-proGRP1–11), was custom synthesized by Auspep Pty Ltd. (Melbourne, Australia) with sequence and purity (>95%) determined by mass spectrometry and HPLC. This peptide was used for conjugation and radiolabeling. For antibody production, the C terminus of the peptide was conjugated to keyhole limpet hemocyanin using bis-diazotized-benzidine (performed by Auspep). The conjugate was emulsified in Freund’s complete adjuvant (Sigma-Aldrich, Castle Hill, Australia) in a 1:1 ratio (vol/vol), and 2 ml of the emulsified conjugate (which contained about 100 nmol of peptide) was injected sc into each of four New Zealand White Cross rabbits. Rabbit immunization and animal care was approved by the Austin Health Animal Ethics committee (Ethics no. 2001/1114). For subsequent booster injections (six to eight weekly intervals), the conjugate was emulsified using Freund’s incomplete adjuvant (Sigma-Aldrich) in a 1:1 ratio (vol/vol). After several booster injections, one rabbit (N798) generated useful antibodies against the proGRP with maximum binding achieved after 42 wk and four booster injections. The specificity of the antiserum was determined by constructing a number of standard curves. proGRP1–125 was detected to a similar extent as proGRP1–11, whereas proGRP31–125 and GRP18–27amide were undetectable confirming that N798 was directed to the N terminus of proGRP. Because this antiserum detects both intact proGRP and N-terminal fragments of proGRP, the results are expressed as proGRP N-terminal immunoreactivity (proGRP NTI). The assay was performed using 0.02 M veronal buffer (1 ml) containing 0.1% BSA and 2 μM NaN3 (pH 8.7). The antibody was used at a final dilution of 1:6000 with 125I-Tyr11-proGRP1–11, prepared by the iodogen method, and purified by reverse-phase HPLC, as the label. Charcoal (2.5%) was added and bound and free peptide were separated by centrifugation. The standard curve was constructed using tyr11-proGRP1–11 (2–2000 fmol/ml) with an ID50 of 200 fmol/tube. The intra- and interassay coefficients of variation were 4% and 14%, respectively.
GRPamide.
Details of the RIA have been published (12). The assay used antiserum R40, which cross-reacts equally with the amidated forms of bombesin, GRP1–27 and GRP18–27 but not with nonamidated forms such GRPgly. This antisera detects GRP from sheep, rat, and human. 125I-Tyr4-bombesin was prepared using iodogen (Pierce Chemical Co., Rockford, IL) followed by reduction with dithiothreitol and purification by reverse-phase HPLC. The standard curve was constructed using bombesin (2–2000 fmol/ml) with an ID50 of 100 fmol/tube.
Preparation of cancer cells and CRC for RIA
Prostate, endometrial, and colon cancer cell lines (detailed in Table 1) were grown in DMEM (Invitrogen, Mulgrave, Australia) supplemented with 10% fetal calf serum (FCS). The cells were seeded into a Petri dish and allowed to grow for 24 h before being serum starved for 24 h. The cells were then lysed with boiling water and the lysate spun down to remove cell debris (12, 21). This lysate was concentrated with a Sep-Pak (Millipore-Waters, Lane Cove, Australia) and the elution fraction from the Sep-Pak was evaporated under a stream of air. After resuspension in assay buffer concentrations of both N-terminal proGRP and amidated GRPs were measured by RIA. Resected CRC which had been stored at –70 C were extracted with boiling water and the pellet was reextracted with boiling acetic acid (3%). Both water and acid extracts were assayed by RIA.
Cell lines for biological activity
DLD-1 cells (ATCC, Manassas, VA), and GRP-R- and BRS-3-transfected Balb-3T3 cells (donated by Dr. J. Battey, National Institutes of Health, Bethesda, MD) were grown in monolayer cultures in DMEM supplemented with 10% FCS and 300 μg/ml of G418 for the transfected cells in an atmosphere of 95% air and 5% carbon dioxide at 37 C. Cultures were passaged at 2- to 3-d intervals to maintain the cells at subconfluent densities.
Construction of proGRP fusion proteins
A pGEX2TH expression plasmid (Amersham, Baulkham Hills, Australia) encoding a fusion protein between glutathione S-transferase and human proGRP was constructed and expressed in BL21 Escherichia coli cells and purified from bacterial lysates by binding to glutathione-agarose. However, treatment of the fusion protein with enterokinase or thrombin resulted in multiple cleavages within the proGRP sequence (data not shown).
To produce a full-length proGRP1–125, the proGRP sequence was subcloned from the pGEX2TH plasmid into the His tag expression vector proEXHtb (Invitrogen) by restriction digestion with BamHI and HindIII. This DNA was then electroporated into electrocompetent E. coli strain DH5. Clones with the correct insert were selected by restriction digestion. The sequence consisted of the His tag joined to proGRP1–125 by a seven-amino acid linker and a recombinant tobacco etch virus protease (rTEV) cleavage site and an enterokinase cleavage site (Fig. 2). DNA sequencing of positive clones revealed a mutation in codon 14 (ATG to GTG), which resulted in a conservative mutation of amino acid 14 from Met to Val.
Expression of proGRP His tag fusion protein
The His-proGRP1–125 fusion protein was expressed in E. coli BL21 cells and expression was induced by treatment with isopropylthiogalactoside for 4 h. The His-proGRP1–125 fusion protein was purified from sarkosyl lysates by binding to nickel-iminodiacetic acid (Ni-IDA) agarose. The protein content of samples at various stages of the purification was analyzed by SDS-PAGE on 15% gels.
Reverse-phase HPLC
The recombinant human proGRP1–125 prepared above was applied to a C18 μBondapak column (8 x 100 mm; Waters Associates, Milford, MA), which had been equilibrated with 0.05% trifluoroacetic acid. The His-proGRP1–125 fusion protein was eluted with a gradient from 0–70% acetonitrile in 0.05% trifluoroacetic acid at a flow rate of 1 ml/min. Fractions of 1 ml were collected and dried on a Speed Vac (Savant, Hicksville, NY) for RIA, mass spectrometry, and biological activity assays. In a separate chromatogram, a water extract of a CRC was run under similar conditions and the concentrations in each fraction measured by RIA.
Mass spectrometry and N-terminal sequencing
Mass spectrometry (Bruker AutoFlex MALDI-TOF, Billerica, MA) was performed on the peak immunoreactive HPLC fraction to determine the molecular weight of the recombinant protein. N-terminal sequencing was done on the peak HPLC fraction to confirm the amino acid sequence.
Proliferation studies
DLD-1 cells were grown in media supplemented with 10% FCS, seeded in 24-well plates at a density of 50,000 cells/well, and grown for 24 h. The cells were then serum starved for 24 h, before being treated with different concentrations of HPLC-purified His-proGRP1–125, GRP18–27amide (Auspep) or the GRP-R antagonist, (D-Phe6, Leu-NHEt13, des-Met14)-bombesin6–14 (Bachem AG, Bubendorf, Switzerland), in media supplemented with 0.2% FCS for 16 h. As determined by HPLC, His-proGRP1–125 was stable with only a single form eluting in the position of His-proGRP1–125 detected at the end of the incubation. 3H-thymidine was added at a concentration of 0.5 μCi /well and incubated for 4 h at 37 C before the medium was discarded. The cells were washed and then incubated with 5% trichloroacetic acid at 4 C for 30 min, after which time the cells were washed with 95% ethanol and dried. The cells were lysed with 1M sodium hydroxide and the lysate counted in a counter.
Inositol phosphate production
DLD-1 cells were grown in media supplemented with 10% FCS, seeded in a 24-well plate at a density of 50,000 cells/well and grown for 24 h. The cells were serum starved and labeled with 3H myo-inositol (Amersham) for 24 h and then incubated with 20 mM LiCl for 1 h before being treated with different concentrations of His-proGRP1–125, GRP18–27amide, and GRP-R antagonist for 1 h. A stop solution (1:2000 hydrochloric acid in ethanol) was added to the cells before the lysate was loaded onto an AG1X-8 (Bio-Rad, Regents Park, Australia) column. The lysate was allowed to run through, and the column was washed with distilled water and then 40 mM ammonium formate. Inositol phosphates were eluted with 5 ml 1 M ammonium formate. Scintillation fluid (5 ml PicoFluor 40; PerkinElmer, Rowville, Australia) was added to the eluant and the mixture was counted in a -counter.
MAPK activation
Antibodies against total and phosphorylated p42/44 MAPKs were obtained from Cell Signaling Technology (Genesearch, Arundel, Australia). Cells were grown in 10-cm Petri dishes in DMEM containing 10% FCS until 90% confluent. After serum starvation, the cells were stimulated for 0, 3, 5, and 10 min with 10 nM His-proGRP1–125. Maximum stimulations was observed at 5 min and subsequent experiments with 10 nM His-proGRP1–125 or GRP18–27amide in the presence or absence of the GRP-R antagonist, all diluted in serum-free DMEM, were performed at the 5-min time point. Cells were lysed with RIPA buffer [1% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 0.2 mM sodium orthovanadate, 0.5 mM dithiothreitol, and protease inhibitors (Sigma-Aldrich) in 20 mM Tris, 150 mM NaCl (pH 7.6)]. Thirty micrograms of total protein lysates were then mixed with loading buffer, denatured, and separated by electrophoresis (10% SDS-PAGE). Proteins were transferred onto a nitrocellulose membrane using a wet transfer system (Bio-Rad). Membranes were then incubated with the appropriate primary antibodies to phosphorylated or total p42/44 MAPK, and detection performed with alkaline phosphatase-coupled antirabbit IgG followed by incubation with a 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium solution (pH 9.2) (Sigma-Aldrich). Membranes were scanned using a Hewlett-Packard ScanJet 5200C, and densitometric analysis of protein bands was performed with Fuji BAS software. Band densities were determined by taking the ratio of densities of phosphorylated to total MAPK.
Binding assays
GRP-R- and BRS-3-transfected Balb-3T3 cells were grown in media supplemented with 10% FCS, seeded in a 12-well plate at a density of 200,000 cells/well, and grown for 24 h. The cells were then serum starved for 24 h before being treated with GRP18–27amide, His-proGRP1–125, or a synthetic BRS-3 agonist, Tyr6-ala11Phe13Nle14Bn6–14, at a concentration of 1 μM. The cells were preincubated for 15 min at 37 C, and approximately 50,000 cpm of 125I Tyr-4-bombesin (Amersham) or 125I-BRS-3 agonist (prepared by the iodogen method without the reduction step, followed by reverse-phase HPLC, as detailed above for the bombesin label) was added to each well. The cells were incubated with mixing at 37 C for 45 min, washed twice with ice-cold PBS/2% BSA, and then lysed with 1 M sodium hydroxide. The lysate was collected and counted in a -counter.
Statistics
Results are expressed as the means ± SE of at least three separate experiments. Results were analyzed by one-way ANOVA followed by Dunnett’s or Bonferroni methods. Differences with P values of < 0.05 were considered significant.
Results
Detection of proGRP NTI in cancer cell lines and CRC
Prostate, endometrial, and colon cancer cell lines and resected CRC were assayed for proGRP and GRPamide. Other than the Ishikawa endometrial cell line, GRPamide was not detected in any of the cell lines or tumors (Tables 1 and 2). In contrast, immunoreactivity measured with the antisera directed to the N terminus of proGRP1–125 (N798) was present in all cell lines tested, at concentrations ranging from 17–94 fmol/106 cells. Nine of 10 of the resected CRC also contained substantial amounts of proGRP NTI with a maximum concentration of 16.1 pmol/g (Table 2). Reverse-phase HPLC was performed on a CRC water extract and the fractions measured with the N-terminal antisera (Fig. 3). Based on coelution with GRP1–27amide standard and the absence of GRPamide immunoreactivity, the N-terminal immunoreactivity eluting at fraction 18 is nonamidated GRP1–27. The nature of the more hydrophobic species eluting at fraction 64 remains to be determined but based on its elution position is not proGRP1–125.
Purification and characterization of recombinant proGRP his tag fusion protein
To generate full-length proGRP1–125, a plasmid encoding a His6-tagged-proGRP1–125 fusion protein was constructed as described in Materials and Methods. As well as the His6 tag, the fusion protein contained a seven-amino acid linker, an rTEV cleavage site and an enterokinase cleavage site before the proGRP sequence (Fig. 2). The fusion protein was expressed in E. coli and purified from bacterial lysates by binding to Ni-IDA-agarose. No cleavage was observed when the fusion protein was treated with enterokinase while bound to the beads. After elution with imidazole, and removal of the imidazole by dialysis enterokinase treatment resulted in incomplete cleavage at multiple sites within the proGRP sequence (data not shown). No cleavage was observed when rTEV protease, which was specific for the rTEV cleavage site, was used.
The intact His-proGRP1–125 fusion protein was therefore isolated from bacterial lysates (Fig. 4) by binding to Ni-IDA agarose. The fusion protein was eluted from the Ni-IDA agarose beads with 250 mM imidazole and purified by reverse-phase HPLC (Fig. 5). The absorbance peak in fraction 24 matched very well with the peak of immunoreactivity measured with the N terminally directed antiserum N798.
The identity of the peptide that corresponded to the absorbance and immunoreactivity peak was confirmed by mass spectrometry. Fraction 24 from reverse-phase HPLC contained a species with a molecular mass of 17,495 Da, which does not correspond with the theoretical predicted mass of 17,641 Da. It appeared from fragmentation in the low molecular weight region of the spectra that the N-terminal methionine was absent from the recombinant molecule (Fig. 2). This conclusion was confirmed by N-terminal sequencing. The small discrepancy remaining between the predicted (17,510) and experimental masses (17,495) could be due to a modification of the N-terminal serine after the methionine is cleaved. We conclude that the major species of recombinant proGRP purified by reverse-phase HPLC consists of residues 1–125 together with the His6 tag and linker sequence shown in Fig. 2.
Biological activity
Proliferation assays.
Recombinant human His-proGRP1–125 at a concentration of 10 nM stimulated proliferation of the colon cancer cell line DLD-1 (Fig. 6). There was no significant difference in the stimulation of proliferation between His-proGRP1–125 and GRP18–27amide. The GRP-R antagonist (D-Phe6, Leu-NHEt13, des-Met14)-bombesin6–14 blocked the proliferative effects of both His-proGRP1–125 and GRPamide.
Inositol phosphate production.
The production of inositol phosphates by the colon cancer cell line DLD-1 was monitored by ion-exchange chromatography. Recombinant human His-proGRP1–125 at concentrations of 1 or 10 nM did not stimulate inositol phosphate production significantly (Fig. 7). Addition of the GRP-R antagonist (D-Phe6, Leu-NHEt13, des-Met14)-bombesin6–14 had no effect. In contrast, GRP18–27amide at a concentration of 1 nM significantly stimulated inositol phosphate production, and the stimulation was reversed by the addition of the GRP-R antagonist (Fig. 7).
MAPK activation.
Phosphorylation of both the 42- and 44-kDa forms of MAPK was significantly stimulated (range 30- to 80-fold) by treatment with 10 nM GRP or 10 nM His-proGRP1–125 in DLD-1 colorectal carcinoma cells. This stimulation was reversed by the addition of 50 nM GRP-R antagonist in the case of GRP, but not His-proGRP1–125 (Fig 8).
Binding assays
Recombinant human His-proGRP1–125 at a concentration of 1 μM did not compete with radiolabeled bombesin for binding to the GRP-R on GRP-R-transfected Balb-3T3 cells (Fig. 9A). GRP18–27 (1 μM) used as a positive control nearly abolished binding. Recombinant human His-proGRP1–125 at a concentration of 1 μM did not compete with a radiolabeled BRS-3 agonist for binding to BRS-3 on BRS-3-transfected Balb-3T3 cells, whereas unlabeled BRS-3 agonist abolished binding (Fig. 9B).
Discussion
The autocrine hypothesis for cancer growth postulates that a cell produces both a growth factor and its cognate receptor, which interact resulting in proliferation. GRPamide is the prototypical autocrine growth factor. This description was originally based on the detection of GRPamide and its cognate receptor, and on the antiproliferative effect of antibodies directed against GRPamide, in SCLC (22). However, GRPamide is also a potent mitogen for several other types of carcinomas including colon, pancreas, prostate, and breast (reviewed in Refs.4 and 5). With the exception of recent studies using GRPgly (6, 11), analysis of proGRP-derived peptides has been confined to the amidated forms GRP1–27 and GRP18–27. We describe here the production of recombinant His-proGRP1–125, the demonstration that it is biologically active, and the observation that nonamidated proGRP-derived peptides are produced in various cancer cell lines and resected CRC.
A RIA for the measurement of proGRP and proGRP N-terminal fragments was developed. The assay used a new antisera (N798) produced by injection of tyr11-proGRP1–11 conjugated to KLH into New Zealand White rabbits. We showed significant concentrations of proGRP NTI in cell lines derived from prostate, endometrial, and colon cancers as well as in resected CRC. However, GRPamide was detected in only one endometrial cancer cell line but not in prostate, CRC cell lines or resected CRC (Tables 1 and 2). The absence of GRPamide was unexpected because GRP mRNA is expressed in the majority of CRC cell lines and tumors (23, 24, 25), and it had been assumed that the proGRP would be processed at least in part to amidated forms as occurs in SCLC (17, 22, 26). Indeed immunohistochemical studies of resected CRC using an antisera against GRPamide gave positive results (24). However, this antisera may also detect nonamidated forms at the high antisera concentrations required for immunohistochemistry. To our knowledge, the present report is the first to use RIA to quantify the amount of proGRP-derived peptides in CRC.
In contrast to GRPamide, significant amounts of proGRP NTI were measured in all cell lines and in resected CRC. Reverse-phase HPLC of a resected CRC extract and N-terminal RIA demonstrated two major species, one probably being nonamidated GRP1–27 and a second more hydrophobic form that did not coelute with proGRP1–125 or His-proGRP1–125. The precise nature of the proGRP NTI in the tumors and cell lines will require further study using larger amounts of material. Until the current study, characterization of nonamidated forms of proGRP-derived peptides has been confined to SCLC tumors and cell lines. Using a similar N-terminal-directed antiserum and Western blotting, Cuttitta et al. (26) reported the presence of multiple forms of proGRP peptides with a molecular weight range of 14–18K, consistent with the presence of proGRP1–125 and N-terminal cleavage products. Substantial concentrations of the C-terminal of proGRP in SCLC have also been detected using a variety of C-terminal-directed antisera (14, 15, 16, 17). The presence of C-terminal immunoreactive forms of proGRP in CRC has not been reported. Taken together, we can conclude that proGRP and processing intermediates are expressed in a variety of tumor types and that processing to the amidated forms is incomplete as has been reported for a number of other peptides in tumors (27, 28).
Having established that proGRP NTI is expressed in tumors and tumor-derived cell lines, it was important to determine whether proGRP was bioactive. We therefore produced recombinant human proGRP1–125 in E. coli. In initial studies, proGRP was synthesized as a glutathione S-transferase fusion protein; however, this fusion protein was not cleaved appropriately to the full-length protein. Experiments were therefore performed using the purified His-proGRP1–125 fusion protein. His-tagged recombinant proteins have been shown in many instances to retain their normal biological function (29, 30).
Characterization of the His-proGRP1–125 fusion protein included DNA sequencing of the vector, RIA, mass spectroscopy, and N-terminal sequencing. DNA sequencing of the vector showed a mutation that resulted in an amino acid change from Met to Val at position 14. Because this is a conservative change, it is unlikely to be significant. Although a definitive answer will require comparison with the wild type, we have recently shown that biological activity of proGRP18–125 is similar to His-proGRP1–125 (our unpublished observations). Mass spectroscopic analysis revealed a small discrepancy between the expected theoretical mass and the experimental mass. Careful examination of the spectra suggested that the N-terminal Met of the fusion protein had been removed, a conclusion confirmed by N-terminal sequencing. Several studies have shown that in a significant fraction of E. coli cytosolic proteins, the N-terminal methionine is removed when the side chain of the adjacent amino acid residue is relatively small as is the case with Ser in the present instance (Fig. 2) (30, 31).
Recombinant His-proGRP1–125 was biologically active. Proliferation of the colon cancer cell line DLD-1 was stimulated significantly by His-proGRP1–125 at a concentration of 10 nM. A similar effect was seen with GRP18–27amide. Both GRP18–27amide and His-proGRP1–125 stimulated phosphorylation of MAPK. Interestingly, His-proGRP1–125 had no effect on inositol phosphate production, although GRP18–27amide stimulated production significantly.
The identity of the receptor involved in the biological activity of His-proGRP1–125 has not been established. Neither GRP-R or BRS-3 appears to be involved because His-proGRP1–125 did not compete with labeled bombesin or a labeled BRS-3 agonist for binding to Balb-3T3 cells transfected with GRP-R or BRS-3, respectively. Further evidence against the involvement of the GRP-R is the observation that the GRP-R antagonist (D-Phe6, Leu-NHEt13, des-Met14)-bombesin6–14 did not affect stimulation of MAPK by His-proGRP1–125, but inhibited the effect of GRP18–27amide in the same assay. The fact that His-proGRP1–125 had no effect on inositol phosphate production, whereas GRP18–27amide significantly stimulated production, is consistent with the same conclusion. On the other hand the GRP-R antagonist reversed the proliferative effect of His-proGRP1–125. Resolution of this discrepancy will require isolation and characterization of the proGRP receptor, and further elucidation of its downstream signaling pathways.
In summary, we have described for the first time the production and purification of proGRP and shown it to be biologically active in vitro. We have also shown the presence of nonamidated forms of proGRP in a panel of cancer cell lines and resected colorectal tumors. The results support the possibility of using antisera or antagonists to precursor forms of GRP as a treatment for colorectal and other cancers that express proGRP-derived peptides. Finally, we provide another example for the production from a single precursor of multiple peptides with independent receptors and different bioactivities (32, 33, 34).
Footnotes
This work was supported by the National Health and Medical Research Council of Australia, the Austin Hospital Medical Research Foundation, and Western Hospital. We thank Dr. Mustafa Ayan from LaTrobe University (Bundoora, Victoria, Australia) for the mass spectroscopy, and Dr. Lindsay Sparrow from Commonwealth Scientific and Industrial Research Organisation (Parkville, Victoria, Australia) for the N-terminal sequencing. Tyr6-ala11Phe13Nle14Bn6–14 was donated by Professor David Coy (Tulane University, New Orleans, LA), and the proGRP-containing plasmid by Professor Jim Battey (National Institutes of Health, Bethesda, MD).
First Published Online October 13, 2005
Abbreviations: CRC, Colorectal cancer; FCS, fetal calf serum; GRP, gastrin-releasing peptide; Ni-IDA, nickel-iminodiacetic acid; NTI, N-terminal immunoreactivity; ProGRP, prohormone for GRP; rTEV, recombinant tobacco etch virus protease; SCLC, small cell lung carcinoma; SDS, sodium dodecyl sulfate.
Accepted for publication October 3, 2005.
References
McDonald TJ, Jornvall H, Nilsson G, Vagne M, Ghatei M, Bloom SR, Mutt V 1979 Characterization of a gastrin releasing peptide from porcine non-antral gastric tissue. Biochem Biophys Res Commun 90:227–233
Bunnett N 1994 Gastrin releasing peptide. In: Walsh JH, Dockray GJ, eds. Gut peptides. New York: Raven Press; 423–455
Yamada K, Wada E, Wada K 2000 Bombesin-like peptides: studies on food intake and social behaviour with receptor knock-out mice. Ann Med 32:519–529
Preston SR, Miller GV, Primrose JN 1996 Bombesin-like peptides and cancer. Crit Rev Oncol Hematol 23:225–238
Moody TW, Chan D, Fahrenkrug J, Jensen RT 2003 Neuropeptides as autocrine growth factors in cancer cells. Curr Pharm Des 9:495–509
Patel O, Dumesny C, Giraud AS, Baldwin GB, Shulkes A 2004 Stimulation of proliferation and migration of a colorectal cancer cell line by amidated and glycine-extended gastrin-releasing peptide via the same receptor. Biochem Pharmacol 68:2129–2142
Sano H, Feighner SD, Hreniuk DL, Iwaasa H, Sailer AW, Pan J, Reitman ML, Kanatani A, Howard AD, Tan CP 2004 Characterization of the bombesin-like peptide receptor family in primates. Genomics 84:139–146
Moody TW, Merali Z 2004 Bombesin-like peptides and associated receptors within the brain: distribution and behavioral implications. Peptides 25:511–520
Kroog GS, Jensen RT, Battey JF 1995 Mammalian bombesin receptors. Med Res Rev 15:389–417
Rounseville MP, Davis TP 2000 Prohormone convertase and autocrine growth factor mRNAs are coexpressed in small cell lung carcinoma. J Mol Endocrinol 25:121–128
Oiry C, Pannequin J, Bernad N, Artis AM, Galleyrand JC, Devin C, Cristau M, Fehrentz JA, Martinez J 2000 A synthetic glycine-extended bombesin analogue interacts with the GRP/bombesin receptor. Eur J Pharmacol 403:17–25
Giraud A, Whitley J, Shulkes A, Parker L 1996 The pregnant ovine endometrium constitutively expresses and secretes a highly stable bombesin-like peptide, which shares C-terminal sequence but differs structurally from gastrin-releasing peptide. Biochim Biophys Acta 5:189–197
Whitley JC, Giraud AS, Shulkes A 1996 Expression of gastrin-releasing peptide (GRP) and GRP receptors in the pregnant human uterus at term. J Clin Endocrinol Metab 81:3944–3950
Reeve Jr JR, Cuttitta F, Vigna SR, Heubner V, Lee TD, Shively JE, Ho FJ, Fedorko J, Minna JD, Walsh JH 1989 Multiple gastrin-releasing peptide gene-associated peptides are produced by a human small cell lung cancer line. J Biol Chem 264:1928–1932
Hamid QA, Addis BJ, Springall DR, Ibrahim NB, Ghatei MA, Bloom SR, Polak JM 1987 Expression of the C-terminal peptide of human pro-bombesin in 361 lung endocrine tumours, a reliable marker and possible prognostic indicator for small cell carcinoma. Virchows Arch A Pathol Anat Histopathol 411:185–192
Holst JJ, Hansen M, Bork E, Schwartz TW 1989 Elevated plasma concentrations of C-flanking gastrin-releasing peptide in small-cell lung cancer. J Clin Oncol 7:1831–1838
Vangsted AJ, Schwartz TW 1990 Production of gastrin-releasing peptide-(18–27) and a stable fragment of its precursor in small cell lung carcinoma cells. J Clin Endocrinol Metab 70:1586–1593
Sunaga N, Tsuchiya S, Minato K, Watanabe S, Fueki N, Hoshino H, Makimoto T, Ishihara S, Saito R, Mori M 1999 Serum pro-gastrin-releasing peptide is a useful marker for treatment monitoring and survival in small-cell lung cancer. Oncology 57:143–148
Ide A, Ashizawa K, Ishikawa N, Ishii R, Ando T, Abe Y, Sera N, Usa T, Tominaga T, Ejima E, Nakashima M, Ito K, Eguchi K 2001 Elevation of serum pro-gastrin-releasing peptide in patients with medullary thyroid carcinoma and small cell lung carcinoma. Thyroid 11:1055–1061
Yashi M, Muraishi O, Kobayashi Y, Tokue A, Nanjo H 2002 Elevated serum progastrin-releasing peptide (31–98) in metastatic and androgen-independent prostate cancer patients. Prostate 51:84–97
Dumesny C, Whitley JC, Baldwin GS, Giraud AS, Shulkes A 2004 Developmental expression and biological activity of gastrin-releasing peptide and its receptors in the kidney. Am J Physiol Renal Physiol 287:F578–F585
Cuttitia F, Carney DJM 1985 Bombesin-like peptides can function as autocrine growth factors in human lung carcinoma cells. Nature 316:823–826
Carroll RE, Ostrovskiy D, Lee S, Danilkovich A, Benya RV 2000 Characterization of gastrin-releasing peptide and its receptor aberrantly expressed by human colon cancer cell lines. Mol Pharmacol 58:601–607
Carroll RE, Matkowskyj KA, Chakrabarti S, McDonald TJ, Benya RV 1999 Aberrant expression of gastrin-releasing peptide and its receptor by well-differentiated colon cancers in humans. Am J Physiol 276:G655–G665
Chave HS, Gough AC, Palmer K, Preston SR, Primrose JN 2000 Bombesin family receptor and ligand gene expression in human colorectal cancer and normal mucosa. Br J Cancer 82:124–130
Cuttitta F, Fedorko J, Gu JA, Lebacq-Verheyden AM, Linnoila RI, Battey JF 1988 Gastrin-releasing peptide gene-associated peptides are expressed in normal human fetal lung and small cell lung cancer: a novel peptide family found in man. J Clin Endocrinol Metab 67:576–583
Rehfeld JF, Bardram L, Blanke S, Bundgaard JR, Friis-Hansen L, Hilsted L, Johnsen AH, Kofod M, Luttichau HR, Monstein HJ, Nielson C, Nielson FC, Paloheimo LI, Pederson K, Pildal J, Ramlau J, Van Solinge WW, Thorup U, Odum L 1993 Peptide hormone processing in tumours: biogenetic and diagnostic implications. Tumour Biol 14:174–183
Ciccotosto GD, McLeish A, Hardy KJ, Shulkes A 1995 Expression, processing, and secretion of gastrin in patients with colorectal carcinoma. Gastroenterology 109:1142–1153
Hochuli E 1990 Purification of recombinant proteins with metal chelate adsorbent. Genet Eng (NY) 12:87–98
Taussig R, Quarmby LM, Gilman AG 1993 Regulation of purified type I and type II adenylylcyclases by G protein subunits. J Biol Chem 268:9–12
Boissel JP, Kasper TJ, Bunn HF 1988 Cotranslational amino-terminal processing of cytosolic proteins. Cell-free expression of site-directed mutants of human hemoglobin. J Biol Chem 263:8443–8449
Raffin-Sanson ML, de Keyzer Y, Bertagna X 2003 Proopiomelanocortin, a polypeptide precursor with multiple functions: from physiology to pathological conditions. Eur J Endocrinol 149:79–90
Drucker DJ 2002 Biological actions and therapeutic potential of the glucagon-like peptides. Gastroenterology 122:531–544
Aly A, Shulkes A, Baldwin GS 2004 Gastrins, cholecystokinins and gastrointestinal cancer. Biochim Biophys Acta 1704:1–10(Chelsea Dumesny, Oneel Patel, Shamilah L)