- Email: [email protected]
Bombesin-like peptides in small cell lung cancer: Biochemical characterization and secretion from a cell line
Pergamon Press
Life Sciences, Vol. 32, pp. 487-493 Printed in the U.S.A.
BOMBESIN-LIKE PEPTIDES IN SMALL CELL LUNG CANCER: BIOCHEMICAL CHARACTERIZATION AND SECRETION FROM A CELL LINE Terry W. Moody', Edward K. Russell', Thomas L. O'Donohue3, Carol D. Linden1 and Adi F. Gazdar2 'Department of Biochemistry, The George Washington University School of Medicine and Health Sciences, Washington, D.C. 20037; 'NCI-Navy Medicine Oncology Branch, National Cancer Institute and National Navy Medical Center, Bethesda, MD 20814; and 3Laboratory of Clinical Science, National Institute of Mental Health, Bethesda, MD 20205 (Received in final form October 11, 1982) Summary High intracellular levels of BN-like peptides are present in tumors and cell lines of small cell carcinoma of the lung (SCCL) as well as the putative precursor cells of this tumor, the pulmonary endocrine cell. In cell line NCI-HZ09 the density of bombesin-like peptides was 8.9 f 1.1 pmol/mg total protein. Gel filtration chromatography of an extract of these cells revealed one major peak of immunoreactivity which coeluted with synthetic bombesin (1620 daltons). Also, high pressure liquid chromatography revealed one major peak of immunoreactivity was present which eluted before synthetic peptide. Therefore, SCCL bombesin-like peptides may be of similar size but are more hydrophilic than synthetic peptide. Cells maintained in culture continuously release bombesin-like peptides into the growth medium. Also, high concentrations of K+ stimulated the secretion of immunoreactive bombesin from cell lines in a Ca*-dependent manner. These SCCL bombesin-like peptides may function as important regulatory agents in the malignant lung. Bombesin (BN), a tetradecapeptide isolated from frog skin (l), is biologically active in the central nervous system (CNS) and periphery. In the rat, BN is a potent hypothermic (2) and hyperglycemic agent (3) after CNS administration whereas after intraperitoneal injection BN functions as a potent satiety agent (4). In the dog, intravenous infusion of BN stimulates gastrin release from the antral mucosa (5) and hypertension via a noncatecholaminergic mechanism (6). Thus, BN has a broad spectrum of physiological activities. Some of the actions of BN may be mediated by the receptors for BN which have been detected in the CNS (7) and other organs (8). In rat brain, the Cterminal of BN is essential for high affinity binding and ability to induce hypothermia. Also, the C-terminal octapeptide of BN is essential for high affinity binding and ability to stimulate Ca*-flux, elevate intracellular c-GMP levels and cause exocytosis of granules which contain amylase in guinea pig pancreatic acini. Therefore, the pharmacology of central and peripheral receptors for BN may be similar and the C-terminal of BN is essential for high affinity binding and biological activity. These receptors may be activated by endogenous BN-like peptides which are present in the CNS and periphery. By radioimmunoassay BN-like peptides have Address all correspondence to Dr. Terry W. Moody 0024-3205/83/050487-07$03.00/o Copyright (c) 1983 Pergamon Press Ltd.
488
SCCL Bombesin-like Peptides
Vol. 32, No. 5, 1983
been detected in the gastrointestinal (GI) tract (9,10), CNS (11,12) and fetal lung (13). While the density of immunoreactive BN is low in the normal adult lung, it is present in high concentration in SCCL tumors (14) and cell lines (15). This communication presents the biochemical characterization of BN-like peptides in SCCL cell line NCI-H209. Also, the mechanism of release of BNlike peptides from this cell line was investigated. Materials and Methods Cell Culture. Lung cell line NCI-H209 was established and maintained as described (16). SCCL cultures are continuous, clonable, aneuploid and tumorigenic. They have the properties of APUD cells and frequently elaborate multiple peptide hormones (16,17). The cultures were propagated in growth medium RPM1 1640 (which has 5 mM KC1 and 0.4 mM CaNOs), supplemented with 10% fetal bovine serum. The stock cultures were passaged weekly and had an average population doubling time of 3 days. Cells were used when they were in exponential growth phase. One day prior to use, the cells were counted and allquoted into flasks (25 sq. mm) which contained 5 ml of RPM1 1640 medium without fetal bovine serum. The following day, cells were centrifuged at 1,000 x g for 5 min. The supernatants were frozen at -8OOC and an aliquot of cells recounted and assayed for viability. The cells were resuspended in phosphate buffered saline and centrifuged at 1,000 x g. Then the cells were either frozen at -8OOC or used in release experiments. Secretion of BN-like Peptides. Cells (1~10~) were placed in 1 ml of a specially formulated EMEM buffer which was free of I(+. Every 4 min the samples were centrifuged for 20 set using a Beckman microfuge B and the supernatants removed and saved. One ml of fresh buffer was added, the sample vortexed and incubated. Routinely, the cells were treated with normal buffer 4x, then treated with buffer which contained depolarizing stimuli such as 75 mM KC1 twice. Lastly, the cells were treated with normal buffer twice to verify that the peptide release rate returned to basal levels. Routinely, the release was conducted+in the presence and absence of 2 mM CaC12 to determine if the release was Ca -dependent. For all supernatants acetic acid was added at the end of the experiment to make a final concentration of 2N. The acidified samples were vortexed, frozen, lyophilized and assayed for BN-like peptides in the radioimmunoassay. Peptide Extraction and Radioimmunoassay. BN-like peptides were extracted from SCCL cell lines using boiling 2N acetic acid (10 min). The extracts were homogenized using a Kontes cell disrupter. An aliquot of the homogenate was saved for protein determination (18). The remainder was centrifuged at 10,000 x g for 10 min,. the supernatants frozen at -800C and lyophilized. The sample was then resuspended in radioimmunoassay buffer and assayed for immunoreactive BN. The radioimmunoassay was conducted in phosphate buffered saline containing 0.25% bovine serum albumin (BSA). Routinely, a rabbit anti-BN serum at l:lOO,OOO dilution was incubated with experimental sample for 1 hr at 4'C. Then 5,000 cpm (4 fmol) of (lz51-Tyr4) BN was added and the solution incubated for 16 hrs at 4'C; the total volume was 0.4 ml. Then 200 ul of normal rabbit sera was added at 1:200 dilution followed by 100 ul of goat anti-rabbit serum at 1:20 dilution. After 30 min, 200 ul of 12% polyethylene glycol was added to enhance precipitation and the samples centrifuged at 1,000 x g for 20 min. The supernatant was removed and the pellet assayed for antibody tracer complex in an LKB gamma counter. Biochemical Characterization. The lyophilized cell extract was resuspended in 500 ul of 2N acetic acid and centrifuged at 1,000 x g for 10 nrtn to remove
Vol. 32, No. 5, 1983
SCCL Bombesin-like Peptides-
489
residual tissue debris. For the gel filtration studies, the supernatant was applied to a Sephadex L&O column (50 x 1.5 cm) and eluted with methanol/acetic acid/water (10/2/l); fraction size was 1.2 ml. Fractions were assayed for optical density at 280 nm, then diluted in water, frozen and lyophilized. The lyophilized material was resuspended and aliquots from each fraction assayed for immunoreactive BN. For the HPLC studies the supernatant was applied to a Cl8 Sep-pak for partial peptide purification. The Sep-pak was then washed with 6 ml of water and the peptides eluted with 6 ml ethanol/O.2 N HCl (l/l). The peptide elute was frozen, lyophilized and resuspended in 500 pl 0.25 N triethylammonium formate (pH 3.0). The sample was injected into the HPLC (Waters Associates, Model 204) and fractionated using a 4 mm x 30 cm Bondapak/C18 column and 18-33% acetonitrile gradient in triethylammonium formate buffer. One min fractions were collected over a period of 60 min. Aliquots were removed from each fraction and assayed for BN-like peptides in the radioimmunoassay. Results Previously, BN-like peptides were detected in 17 SCCL but not other lung cancer cell lines (15). Because the density of immunoreactive BN was highest in SCCL line NCI-H209, the BN-like peptides in this cell line were characterized further. Figure 1 shows that the protein and BN-like peptide content was a linear function of cell number. The mean density of BN-like peptides was 8.9 f 1.1 pmollmg total protein. Also, Figure 1 shows that BN-like peptides were secreted into the cell medium. The peptide density in the medium was a function of cell concentration and ranged from 0.2 nmollml to a maximum of 1.2 pmollml. Therefore BN-like peptides are not only associated with SCCL cell lines, but are present in the supernatant fluids.
FIG. 1 SCCL BN-like peptide density. NCI-H209 cells were counted and placed in 5 ml RPM1 1640 medium as described in Methods. The following day, the medium was saved and assayed for BN-like peptide concentration (0). Also, the cells were extracted and the total BN-like peptide (0) and protein (eP)content determined.
490
SCCL Bombesin-like Peptides
Vol. 32, No. 5, 1983
The ability of various stimuli to induce secretion of BN-like peptides from SCCL was investigated. Because KPMI 1640 medium, which has 5 mM d and 0.4 mM Ca*, may induce secretion of BN-like peptides -in vitro release studies were conducted using a specially formulated EMEM buffer which was K+ free. Figure 2 shows that as the cells were exposed to this buffer, initially moderate levels of immunoreactive BN were present in the buffer (120 fmol, fraction 1) and with successive washes, this value declined to 90 fmol. When 75 mM KC1 was added to the buffer, the peptide release rate increased 3-fold initially to 270 fmol in the presence but not in the absence of Ca* (fraction 5) then declined to 120 fmol (fraction 6). When the cells were treated with normal buffer, the release rate declined to basal levels (80 fmol, fraction 7 and 8). Because KC1 and veratridine were demonstrated to stimulate release BN-like peptides in a Cau-dependent manner from brain and spinal cord neurons (19,20), the effect of veratrldine on the release of SCCL BN-like peptides esjtigated. Figure 3 shows that moderate levels of immunoreactivity were
FIG. 2 Effect of KC1 on secretion of BN-like peptides. SCCL cell line NCl-HZ09 was treated with buffer in the presence (0) and absence (8) of 2 mM CaC12;75 mM KC1 was added to fractions 5 and 6.The samples were processed as described in methods and 40% of the sample was assayed for BN-like peptides in the radioimmunoassay. Each value represents the mean f S.E. of 4 determinations.
FIG. 3 Effect of veratridine on secretion of BN-like peptides. Cells were treated with buffer in the presence (0) and absence (0) of 2 mM CaC12; 100 uM veratridine was added to fractions 5 and 6. 20% of the sample was assayed for immunoreactive BN and the mean value f S.E. is indicated.
I
I
I
3
6 FRACTION
SCCL Bombesin-like Peptides
vol. 32, No. 5, 1983
491
released initially (90 fmol, fraction 1) and this value declined to approximately 20 fmol. One hundred micromolar veratridine did not significantly increase the peptide release rate (fractions 5 and 6). Therefore, KU, but not veratridine, stimulates secretion of BN-like peptides from SCCL cell line NCI-H209 in a Ca*-dependent manner. Similar data were obtained for SCCL cell lines NCI-H182 and NCI-H250. Previously, techniques were developed to characterize CNS BN-like peptides (12,20). Figure 4 shows that SCCL extracts were fractionated using gel filtration techniques. The SCCL BN-like peptides eluted in fraction 25, at approximately the same position as did synthetic BN. In comparison, very little U.V. absorbing material was present in this fraction; large molecular weight proteins eluted in the void volumn (fraction 21) and biogenic amines eluted in the included volumn (fraction 52). Therefore, BN-like peptides may have a molecular weight similar to that of BN (1620 daltons). ENK
BSA
Nd i
0.3 b 8 1 02
; P 0.1
7 3
F
25
50
FIG. 4
g
Gel filtration profile of SCCL extracts. 2 x lo8 cells were extracted and fractionated on Sephadex LH20 as described in Methods. The total optical density and content of BN-like peptides was determined. The elution positions of BSA, BN, enkephalin (ENK) and NaI are indicated.
FIG. 5 %LC profile of SCCL extracts. 1 x 107 cells were extracted and fractionated as described in Methods. 10% of the sample was assayed for BNlike peptides. The elution positions for BN and BN-SO are indicated.
;c U
20
n
40
TIME (MIN)
60
492
SCCL Bombesin-like Peptides
Vol. 32, No. 5, 1983
Figure 5 shows that using HPLC techniques, one major peak of immunoreactive BN eluted 21 min after injection into that HPLC. In comparison, synthetic BN eluted at 37 min and BN-sulfoxide (BN-SO) at 21 min. Therefore, SCCL BNlike peptides, similar to BN-SO, may be more hydrophilic than is BN. Discussion It is generally accepted that neurons and neuroendocrine cells contain transnitters as well as peptides. These biologically active molecules are usually stored in intracellular granules. In this regard, SCCL cells have dense core neurosecretory granules (16) and amine precursor uptake and decarboxylation properties (16,17,21). Similarly, high levels of BN-like peptides are present in SCCL cell lines (15), lung tumors (14) and xenographs in nude mice (22). Because neuronal BN-like peptides are released into the extracellular fluids when the vesicles that they are stored in undergo exocytosis (19,20), the secretion of BN-like peptides from SCCL cell lines was investigated. Our studies demonstrate that the tissue culture medium and high K+ stimulate the secretion of SCCL BN-like peptides. Figure 1 shows that culture medium exposed to SCCL cells has appreciable amounts of immunoreactive BN (0.2-1.2 pmol/ml depending on the cell concentration) whereas medium alone was devoid of activity. Further, the peptide concentration in the medium was linear at low cell concentrations, but did not exceed 1.2 pmol/ml even at high cell concentrations; cell viability was 80-90% in all samples based on ability to exclude trypan blue. Therefore, the medium RPM1 1640, which contains 5 mM K+ and 0.4 mM Ca* may induce secretion of BN-like peptides from SCCL cells. Similar results have been obtained for ACTH secretion from DIS cells (23) and neurotensin secretion from adrenal medullary thyroid cancer cells (24). Therefore, SCCL cells, similar to other cancer cells which secrete peptides, may have a feedback mechanism whereby peptide secretion is carefully regulated. SCCL cells may secrete transmitters and peptides in response to various depolarizing stimuli (25). Figures 2 and 3 show that KC1 but not veratridine stimulate increased release of BN-like peptides. Therefore, SCCL cells may be depolarized by KCl, as well as other stimuli, but not veratridine. In comparison, brain and spinal cord neurons are depolarized by high K+ and veratridine resulting in the increased release of BN-like peptides (19,20). Therefore, brain and spinal cord neurons, but not SCCL cells, may contain Na+ channels which are activated by the alkaloid veratridine. Secretion of BNlike peptides from patients with extensive SCCL may account for some of the paraneoplastic syndromes associated with this disease, in particular, impaired ability to regulate body temperature and anorexia. In this regard, elevated plasma levels of BN-like peptides have been detected in patients with extensive SCCL (26,27). The high concentration of BN-like peptides in SCCL cell lines facilitated biochemical characterization of these peptides. Figure 4 shows that SCCL BNlike peptides, similar to rat brain (12), have a molecular weight approximately that of BN (1620 daltons). In comparison, BN-like peptides in porcine stomach (gastrin releasing peptides) are composed of 27 amino acid residues and thus are much larger than the tetradecapeptide BN (9). Also, Figure 5 shows that using HPLC techniques SCCL immunoreactive BN elutes before BN and coelutes with the more hydrophilic peptide BN-SO which has the C-terminal methionine amino acid oxidized to the sulfoxide. Therefore, SCCL BN:like peptides are more hydrophilic than rat brain and spinal cord peptides or synthetic standard (20). While the structure of these SCCL BN-like peptides remains to be determined, they may function as important regulatory agents in the normal and malignant lung.
Vol. 32, No. 5, 1983
SCCL Bombesin-like Peptides
493
Acknowledgements The authors sincerely thank Drs. D.N. Carney, F.Cuttitta and J.D. Minna for helpful discussions. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. .18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
A. ANASTASI, V. ERSPAMER and M. BUCCI, Experientia 7, 166-167 (1971). M. BROWN, J. RIVIER and W. VALE, Science 196, 988-1000 (1977). M.R. BROWN, .J. RIVIER and W. VALE, Life Sci. 1, 1729-1734 (1977). J. GIBBS, J.D. FAUSER, E.A. ROWE, B.J. ROLLS, E.T. ROLLS and S.P. MADISON, Nature 282, 208-210 (1979). G. BERTACINI, V. ERSPAMER, P. MELCHIORI and N. SOPRANZI, Br. J. Pharm. 52, 219-225 (1974). V. ERSPAMER, P. MELCHIORRI and N. SOPRANZI, Br. J. Pharm. 9, 442-460 (1972). T.W. MOODY, C.B. PERT, J. RIVIER and M.R. BROWN, Proc. Natl. Acad. Sci. USA 11, 5372-5376 (1978). R.T. JENSEN, T.W. MOODY, C.B. PERT, J.E. RIVIER and J.D. GARDNER, Proc. Natl. Acad. Sci. USA 75, 6139-6143 (1978). T.J. MCDONALD, H. TORNVALL, G. NILSSON, M. VAGNE, M. GHATEI, S.R. BLOOM and V. MUTT, Biochem. Biophys. Res. Commun. 90, 227-233. J.H. WALSH, H.C. WONG and G.J. DOCKRAY, Fed. Proc. 2, 2315-2319 (1979). M.R. BROWN, R. ALLEN, J. VILLAREAL, J. RIVIER and W. VALE, Life Sci. 2, 2321-2728 (1978) T.W. MOODY and C.B. PERT, Biochem. Biophys. Res. Commun. 90, 7-14 (1979). J. WHARTON, J.M. POLAK, S.R. BLOOM and A.G.E. PEARSE, Nature 273, 769770 (1978). S.M. WOOD, J.R. WOOD, M.A. GHATEI, D. O'SHAUGHNESSY and S.R. BLOOM, J. Clin. Endocrinology 12, 1310-1312 (1981). T.W. MOODY, C.B. PERT, A.F. GAZDAR, D.N. CARNN and J.D. MINNA, Science 2&, 1246-1248 (1981). A.F. GAZDAR, D.N. CARNEY, E.K. REDDEL, H.L. SIMS, S.B. BAYLIN, P.A. BUNN, J.G. GUCCION and J.D. MINNA, Cancer Res. 40, 3502-3507 (1980). G.D. SORENSON, O.S. PETTENGILL, T. BRINCK-JOHNSEN, C.C. CATE and L.H. MAUER, Cancer 47, 1289-1296 (1981). O.H. LOWRY, N.J. ROSEBROUGH, L. FARR and R.J. RANDELL, J. Biol. Chem. 193, 165-175 (1951). T.W. MOODY, N.B. THOA, T.L. O'DONOHUE and C.B. PERT, Life Sci. 2, 17071712 (1980). T.W. MOODY, N.B. THOA, T.L. O'DONOHUE and D.M. JACOBOWITZ, Life Sci. 29, 2273-2279 (1981). O.S. PETTENGILL, N.G. BACOPOULOS and G.D. SORENSON, Life Sci. 30, 13551360 (1982). M.D. ERISMAN, R.I. LINNOILA, 0. HERNANDIZ, R.P. DIAUGISTINE and L.H. LAZARUS, Proc. Natl. Acad. Sci. USA 2, 2379-2383 (1982). U.I. RICHARDSON, Endocrinology 102, 910-917 (1978). F.N. ZEYTINOGLU, R.F. GAGEL, A.H. TASHJIAN, R.A. HAMMER and S.E. LEEMAN, Proc. Natl. Acad. Sci. USA z, 3741-3745 (1980). F.V. MCCANN, O.S. PETTENGILL, J.S. COLE, J.A.G. RUSSEL and J.D. SORENSON, Science 212, 1155-1157 (1981). C.B. PERT AND U.K SCHUMACHER, Lancet 1, 509 (1982). S.M. WOOD, J.R. WOOD, M.A. GHATEI, G.D. SORENSON AND S.R. BLOOM, Lancet 1, 690-691 (1982).
Life Sciences, Vol. 32, pp. 487-493 Printed in the U.S.A.
BOMBESIN-LIKE PEPTIDES IN SMALL CELL LUNG CANCER: BIOCHEMICAL CHARACTERIZATION AND SECRETION FROM A CELL LINE Terry W. Moody', Edward K. Russell', Thomas L. O'Donohue3, Carol D. Linden1 and Adi F. Gazdar2 'Department of Biochemistry, The George Washington University School of Medicine and Health Sciences, Washington, D.C. 20037; 'NCI-Navy Medicine Oncology Branch, National Cancer Institute and National Navy Medical Center, Bethesda, MD 20814; and 3Laboratory of Clinical Science, National Institute of Mental Health, Bethesda, MD 20205 (Received in final form October 11, 1982) Summary High intracellular levels of BN-like peptides are present in tumors and cell lines of small cell carcinoma of the lung (SCCL) as well as the putative precursor cells of this tumor, the pulmonary endocrine cell. In cell line NCI-HZ09 the density of bombesin-like peptides was 8.9 f 1.1 pmol/mg total protein. Gel filtration chromatography of an extract of these cells revealed one major peak of immunoreactivity which coeluted with synthetic bombesin (1620 daltons). Also, high pressure liquid chromatography revealed one major peak of immunoreactivity was present which eluted before synthetic peptide. Therefore, SCCL bombesin-like peptides may be of similar size but are more hydrophilic than synthetic peptide. Cells maintained in culture continuously release bombesin-like peptides into the growth medium. Also, high concentrations of K+ stimulated the secretion of immunoreactive bombesin from cell lines in a Ca*-dependent manner. These SCCL bombesin-like peptides may function as important regulatory agents in the malignant lung. Bombesin (BN), a tetradecapeptide isolated from frog skin (l), is biologically active in the central nervous system (CNS) and periphery. In the rat, BN is a potent hypothermic (2) and hyperglycemic agent (3) after CNS administration whereas after intraperitoneal injection BN functions as a potent satiety agent (4). In the dog, intravenous infusion of BN stimulates gastrin release from the antral mucosa (5) and hypertension via a noncatecholaminergic mechanism (6). Thus, BN has a broad spectrum of physiological activities. Some of the actions of BN may be mediated by the receptors for BN which have been detected in the CNS (7) and other organs (8). In rat brain, the Cterminal of BN is essential for high affinity binding and ability to induce hypothermia. Also, the C-terminal octapeptide of BN is essential for high affinity binding and ability to stimulate Ca*-flux, elevate intracellular c-GMP levels and cause exocytosis of granules which contain amylase in guinea pig pancreatic acini. Therefore, the pharmacology of central and peripheral receptors for BN may be similar and the C-terminal of BN is essential for high affinity binding and biological activity. These receptors may be activated by endogenous BN-like peptides which are present in the CNS and periphery. By radioimmunoassay BN-like peptides have Address all correspondence to Dr. Terry W. Moody 0024-3205/83/050487-07$03.00/o Copyright (c) 1983 Pergamon Press Ltd.
488
SCCL Bombesin-like Peptides
Vol. 32, No. 5, 1983
been detected in the gastrointestinal (GI) tract (9,10), CNS (11,12) and fetal lung (13). While the density of immunoreactive BN is low in the normal adult lung, it is present in high concentration in SCCL tumors (14) and cell lines (15). This communication presents the biochemical characterization of BN-like peptides in SCCL cell line NCI-H209. Also, the mechanism of release of BNlike peptides from this cell line was investigated. Materials and Methods Cell Culture. Lung cell line NCI-H209 was established and maintained as described (16). SCCL cultures are continuous, clonable, aneuploid and tumorigenic. They have the properties of APUD cells and frequently elaborate multiple peptide hormones (16,17). The cultures were propagated in growth medium RPM1 1640 (which has 5 mM KC1 and 0.4 mM CaNOs), supplemented with 10% fetal bovine serum. The stock cultures were passaged weekly and had an average population doubling time of 3 days. Cells were used when they were in exponential growth phase. One day prior to use, the cells were counted and allquoted into flasks (25 sq. mm) which contained 5 ml of RPM1 1640 medium without fetal bovine serum. The following day, cells were centrifuged at 1,000 x g for 5 min. The supernatants were frozen at -8OOC and an aliquot of cells recounted and assayed for viability. The cells were resuspended in phosphate buffered saline and centrifuged at 1,000 x g. Then the cells were either frozen at -8OOC or used in release experiments. Secretion of BN-like Peptides. Cells (1~10~) were placed in 1 ml of a specially formulated EMEM buffer which was free of I(+. Every 4 min the samples were centrifuged for 20 set using a Beckman microfuge B and the supernatants removed and saved. One ml of fresh buffer was added, the sample vortexed and incubated. Routinely, the cells were treated with normal buffer 4x, then treated with buffer which contained depolarizing stimuli such as 75 mM KC1 twice. Lastly, the cells were treated with normal buffer twice to verify that the peptide release rate returned to basal levels. Routinely, the release was conducted+in the presence and absence of 2 mM CaC12 to determine if the release was Ca -dependent. For all supernatants acetic acid was added at the end of the experiment to make a final concentration of 2N. The acidified samples were vortexed, frozen, lyophilized and assayed for BN-like peptides in the radioimmunoassay. Peptide Extraction and Radioimmunoassay. BN-like peptides were extracted from SCCL cell lines using boiling 2N acetic acid (10 min). The extracts were homogenized using a Kontes cell disrupter. An aliquot of the homogenate was saved for protein determination (18). The remainder was centrifuged at 10,000 x g for 10 min,. the supernatants frozen at -800C and lyophilized. The sample was then resuspended in radioimmunoassay buffer and assayed for immunoreactive BN. The radioimmunoassay was conducted in phosphate buffered saline containing 0.25% bovine serum albumin (BSA). Routinely, a rabbit anti-BN serum at l:lOO,OOO dilution was incubated with experimental sample for 1 hr at 4'C. Then 5,000 cpm (4 fmol) of (lz51-Tyr4) BN was added and the solution incubated for 16 hrs at 4'C; the total volume was 0.4 ml. Then 200 ul of normal rabbit sera was added at 1:200 dilution followed by 100 ul of goat anti-rabbit serum at 1:20 dilution. After 30 min, 200 ul of 12% polyethylene glycol was added to enhance precipitation and the samples centrifuged at 1,000 x g for 20 min. The supernatant was removed and the pellet assayed for antibody tracer complex in an LKB gamma counter. Biochemical Characterization. The lyophilized cell extract was resuspended in 500 ul of 2N acetic acid and centrifuged at 1,000 x g for 10 nrtn to remove
Vol. 32, No. 5, 1983
SCCL Bombesin-like Peptides-
489
residual tissue debris. For the gel filtration studies, the supernatant was applied to a Sephadex L&O column (50 x 1.5 cm) and eluted with methanol/acetic acid/water (10/2/l); fraction size was 1.2 ml. Fractions were assayed for optical density at 280 nm, then diluted in water, frozen and lyophilized. The lyophilized material was resuspended and aliquots from each fraction assayed for immunoreactive BN. For the HPLC studies the supernatant was applied to a Cl8 Sep-pak for partial peptide purification. The Sep-pak was then washed with 6 ml of water and the peptides eluted with 6 ml ethanol/O.2 N HCl (l/l). The peptide elute was frozen, lyophilized and resuspended in 500 pl 0.25 N triethylammonium formate (pH 3.0). The sample was injected into the HPLC (Waters Associates, Model 204) and fractionated using a 4 mm x 30 cm Bondapak/C18 column and 18-33% acetonitrile gradient in triethylammonium formate buffer. One min fractions were collected over a period of 60 min. Aliquots were removed from each fraction and assayed for BN-like peptides in the radioimmunoassay. Results Previously, BN-like peptides were detected in 17 SCCL but not other lung cancer cell lines (15). Because the density of immunoreactive BN was highest in SCCL line NCI-H209, the BN-like peptides in this cell line were characterized further. Figure 1 shows that the protein and BN-like peptide content was a linear function of cell number. The mean density of BN-like peptides was 8.9 f 1.1 pmollmg total protein. Also, Figure 1 shows that BN-like peptides were secreted into the cell medium. The peptide density in the medium was a function of cell concentration and ranged from 0.2 nmollml to a maximum of 1.2 pmollml. Therefore BN-like peptides are not only associated with SCCL cell lines, but are present in the supernatant fluids.
FIG. 1 SCCL BN-like peptide density. NCI-H209 cells were counted and placed in 5 ml RPM1 1640 medium as described in Methods. The following day, the medium was saved and assayed for BN-like peptide concentration (0). Also, the cells were extracted and the total BN-like peptide (0) and protein (eP)content determined.
490
SCCL Bombesin-like Peptides
Vol. 32, No. 5, 1983
The ability of various stimuli to induce secretion of BN-like peptides from SCCL was investigated. Because KPMI 1640 medium, which has 5 mM d and 0.4 mM Ca*, may induce secretion of BN-like peptides -in vitro release studies were conducted using a specially formulated EMEM buffer which was K+ free. Figure 2 shows that as the cells were exposed to this buffer, initially moderate levels of immunoreactive BN were present in the buffer (120 fmol, fraction 1) and with successive washes, this value declined to 90 fmol. When 75 mM KC1 was added to the buffer, the peptide release rate increased 3-fold initially to 270 fmol in the presence but not in the absence of Ca* (fraction 5) then declined to 120 fmol (fraction 6). When the cells were treated with normal buffer, the release rate declined to basal levels (80 fmol, fraction 7 and 8). Because KC1 and veratridine were demonstrated to stimulate release BN-like peptides in a Cau-dependent manner from brain and spinal cord neurons (19,20), the effect of veratrldine on the release of SCCL BN-like peptides esjtigated. Figure 3 shows that moderate levels of immunoreactivity were
FIG. 2 Effect of KC1 on secretion of BN-like peptides. SCCL cell line NCl-HZ09 was treated with buffer in the presence (0) and absence (8) of 2 mM CaC12;75 mM KC1 was added to fractions 5 and 6.The samples were processed as described in methods and 40% of the sample was assayed for BN-like peptides in the radioimmunoassay. Each value represents the mean f S.E. of 4 determinations.
FIG. 3 Effect of veratridine on secretion of BN-like peptides. Cells were treated with buffer in the presence (0) and absence (0) of 2 mM CaC12; 100 uM veratridine was added to fractions 5 and 6. 20% of the sample was assayed for immunoreactive BN and the mean value f S.E. is indicated.
I
I
I
3
6 FRACTION
SCCL Bombesin-like Peptides
vol. 32, No. 5, 1983
491
released initially (90 fmol, fraction 1) and this value declined to approximately 20 fmol. One hundred micromolar veratridine did not significantly increase the peptide release rate (fractions 5 and 6). Therefore, KU, but not veratridine, stimulates secretion of BN-like peptides from SCCL cell line NCI-H209 in a Ca*-dependent manner. Similar data were obtained for SCCL cell lines NCI-H182 and NCI-H250. Previously, techniques were developed to characterize CNS BN-like peptides (12,20). Figure 4 shows that SCCL extracts were fractionated using gel filtration techniques. The SCCL BN-like peptides eluted in fraction 25, at approximately the same position as did synthetic BN. In comparison, very little U.V. absorbing material was present in this fraction; large molecular weight proteins eluted in the void volumn (fraction 21) and biogenic amines eluted in the included volumn (fraction 52). Therefore, BN-like peptides may have a molecular weight similar to that of BN (1620 daltons). ENK
BSA
Nd i
0.3 b 8 1 02
; P 0.1
7 3
F
25
50
FIG. 4
g
Gel filtration profile of SCCL extracts. 2 x lo8 cells were extracted and fractionated on Sephadex LH20 as described in Methods. The total optical density and content of BN-like peptides was determined. The elution positions of BSA, BN, enkephalin (ENK) and NaI are indicated.
FIG. 5 %LC profile of SCCL extracts. 1 x 107 cells were extracted and fractionated as described in Methods. 10% of the sample was assayed for BNlike peptides. The elution positions for BN and BN-SO are indicated.
;c U
20
n
40
TIME (MIN)
60
492
SCCL Bombesin-like Peptides
Vol. 32, No. 5, 1983
Figure 5 shows that using HPLC techniques, one major peak of immunoreactive BN eluted 21 min after injection into that HPLC. In comparison, synthetic BN eluted at 37 min and BN-sulfoxide (BN-SO) at 21 min. Therefore, SCCL BNlike peptides, similar to BN-SO, may be more hydrophilic than is BN. Discussion It is generally accepted that neurons and neuroendocrine cells contain transnitters as well as peptides. These biologically active molecules are usually stored in intracellular granules. In this regard, SCCL cells have dense core neurosecretory granules (16) and amine precursor uptake and decarboxylation properties (16,17,21). Similarly, high levels of BN-like peptides are present in SCCL cell lines (15), lung tumors (14) and xenographs in nude mice (22). Because neuronal BN-like peptides are released into the extracellular fluids when the vesicles that they are stored in undergo exocytosis (19,20), the secretion of BN-like peptides from SCCL cell lines was investigated. Our studies demonstrate that the tissue culture medium and high K+ stimulate the secretion of SCCL BN-like peptides. Figure 1 shows that culture medium exposed to SCCL cells has appreciable amounts of immunoreactive BN (0.2-1.2 pmol/ml depending on the cell concentration) whereas medium alone was devoid of activity. Further, the peptide concentration in the medium was linear at low cell concentrations, but did not exceed 1.2 pmol/ml even at high cell concentrations; cell viability was 80-90% in all samples based on ability to exclude trypan blue. Therefore, the medium RPM1 1640, which contains 5 mM K+ and 0.4 mM Ca* may induce secretion of BN-like peptides from SCCL cells. Similar results have been obtained for ACTH secretion from DIS cells (23) and neurotensin secretion from adrenal medullary thyroid cancer cells (24). Therefore, SCCL cells, similar to other cancer cells which secrete peptides, may have a feedback mechanism whereby peptide secretion is carefully regulated. SCCL cells may secrete transmitters and peptides in response to various depolarizing stimuli (25). Figures 2 and 3 show that KC1 but not veratridine stimulate increased release of BN-like peptides. Therefore, SCCL cells may be depolarized by KCl, as well as other stimuli, but not veratridine. In comparison, brain and spinal cord neurons are depolarized by high K+ and veratridine resulting in the increased release of BN-like peptides (19,20). Therefore, brain and spinal cord neurons, but not SCCL cells, may contain Na+ channels which are activated by the alkaloid veratridine. Secretion of BNlike peptides from patients with extensive SCCL may account for some of the paraneoplastic syndromes associated with this disease, in particular, impaired ability to regulate body temperature and anorexia. In this regard, elevated plasma levels of BN-like peptides have been detected in patients with extensive SCCL (26,27). The high concentration of BN-like peptides in SCCL cell lines facilitated biochemical characterization of these peptides. Figure 4 shows that SCCL BNlike peptides, similar to rat brain (12), have a molecular weight approximately that of BN (1620 daltons). In comparison, BN-like peptides in porcine stomach (gastrin releasing peptides) are composed of 27 amino acid residues and thus are much larger than the tetradecapeptide BN (9). Also, Figure 5 shows that using HPLC techniques SCCL immunoreactive BN elutes before BN and coelutes with the more hydrophilic peptide BN-SO which has the C-terminal methionine amino acid oxidized to the sulfoxide. Therefore, SCCL BN:like peptides are more hydrophilic than rat brain and spinal cord peptides or synthetic standard (20). While the structure of these SCCL BN-like peptides remains to be determined, they may function as important regulatory agents in the normal and malignant lung.
Vol. 32, No. 5, 1983
SCCL Bombesin-like Peptides
493
Acknowledgements The authors sincerely thank Drs. D.N. Carney, F.Cuttitta and J.D. Minna for helpful discussions. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. .18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
A. ANASTASI, V. ERSPAMER and M. BUCCI, Experientia 7, 166-167 (1971). M. BROWN, J. RIVIER and W. VALE, Science 196, 988-1000 (1977). M.R. BROWN, .J. RIVIER and W. VALE, Life Sci. 1, 1729-1734 (1977). J. GIBBS, J.D. FAUSER, E.A. ROWE, B.J. ROLLS, E.T. ROLLS and S.P. MADISON, Nature 282, 208-210 (1979). G. BERTACINI, V. ERSPAMER, P. MELCHIORI and N. SOPRANZI, Br. J. Pharm. 52, 219-225 (1974). V. ERSPAMER, P. MELCHIORRI and N. SOPRANZI, Br. J. Pharm. 9, 442-460 (1972). T.W. MOODY, C.B. PERT, J. RIVIER and M.R. BROWN, Proc. Natl. Acad. Sci. USA 11, 5372-5376 (1978). R.T. JENSEN, T.W. MOODY, C.B. PERT, J.E. RIVIER and J.D. GARDNER, Proc. Natl. Acad. Sci. USA 75, 6139-6143 (1978). T.J. MCDONALD, H. TORNVALL, G. NILSSON, M. VAGNE, M. GHATEI, S.R. BLOOM and V. MUTT, Biochem. Biophys. Res. Commun. 90, 227-233. J.H. WALSH, H.C. WONG and G.J. DOCKRAY, Fed. Proc. 2, 2315-2319 (1979). M.R. BROWN, R. ALLEN, J. VILLAREAL, J. RIVIER and W. VALE, Life Sci. 2, 2321-2728 (1978) T.W. MOODY and C.B. PERT, Biochem. Biophys. Res. Commun. 90, 7-14 (1979). J. WHARTON, J.M. POLAK, S.R. BLOOM and A.G.E. PEARSE, Nature 273, 769770 (1978). S.M. WOOD, J.R. WOOD, M.A. GHATEI, D. O'SHAUGHNESSY and S.R. BLOOM, J. Clin. Endocrinology 12, 1310-1312 (1981). T.W. MOODY, C.B. PERT, A.F. GAZDAR, D.N. CARNN and J.D. MINNA, Science 2&, 1246-1248 (1981). A.F. GAZDAR, D.N. CARNEY, E.K. REDDEL, H.L. SIMS, S.B. BAYLIN, P.A. BUNN, J.G. GUCCION and J.D. MINNA, Cancer Res. 40, 3502-3507 (1980). G.D. SORENSON, O.S. PETTENGILL, T. BRINCK-JOHNSEN, C.C. CATE and L.H. MAUER, Cancer 47, 1289-1296 (1981). O.H. LOWRY, N.J. ROSEBROUGH, L. FARR and R.J. RANDELL, J. Biol. Chem. 193, 165-175 (1951). T.W. MOODY, N.B. THOA, T.L. O'DONOHUE and C.B. PERT, Life Sci. 2, 17071712 (1980). T.W. MOODY, N.B. THOA, T.L. O'DONOHUE and D.M. JACOBOWITZ, Life Sci. 29, 2273-2279 (1981). O.S. PETTENGILL, N.G. BACOPOULOS and G.D. SORENSON, Life Sci. 30, 13551360 (1982). M.D. ERISMAN, R.I. LINNOILA, 0. HERNANDIZ, R.P. DIAUGISTINE and L.H. LAZARUS, Proc. Natl. Acad. Sci. USA 2, 2379-2383 (1982). U.I. RICHARDSON, Endocrinology 102, 910-917 (1978). F.N. ZEYTINOGLU, R.F. GAGEL, A.H. TASHJIAN, R.A. HAMMER and S.E. LEEMAN, Proc. Natl. Acad. Sci. USA z, 3741-3745 (1980). F.V. MCCANN, O.S. PETTENGILL, J.S. COLE, J.A.G. RUSSEL and J.D. SORENSON, Science 212, 1155-1157 (1981). C.B. PERT AND U.K SCHUMACHER, Lancet 1, 509 (1982). S.M. WOOD, J.R. WOOD, M.A. GHATEI, G.D. SORENSON AND S.R. BLOOM, Lancet 1, 690-691 (1982).