0013-7227/91/1295-2530$03.00/0 Endocrinology Copyright © 1991 by The Endocrine Society
Vol. 129, No. 5 Printed in U.S.A.
Calcitonin and Calcitonin Gene-Related Peptide Are Chemotactic for F9 Embryonal Carcinoma Cells* PASCALE GERBAUD, NADINE SEGOND, MOHSEN M. MOUKHTAR, AND DANIELE EVAIN-BRION Laboratoire de Physiopatholologie du Developpement, CNRS URA 1337, ENS, 75230 Paris Cedex 05; and the Laboratoire d'Endocrinologie Moleculaire, INSERM U113, CHU St. Antoine, 75571 Paris Cedex, France
ABSTRACT. The chemotactic effect of calcitonin (CT) gene products was tested on F9 teratocarcinoma cells, which are an in vitro model of early embryonic development. CT and CT gene-related peptide (CGRP) induce a significant chemotactic response (chemotactic index, 40-50). The order of potency is: chicken CGRP > salmon CT > human CGRP. Human CT is a less potent chemotactic agent (chemotactic index, 15). Compared to other well known peptides with chemotactic activity, such as platelet-derived growth factor (no activity) and transforming growth factor-^ (chemotactic index, 5), CGRP and CT appear to be very active in attracting F9 cells in the Boyden chamber
assay. Interestingly, CT and CGRP exhibit little chemotactic effect toward differentiated teratocarcinoma cells (i.e. retinoic acid-treated F9 cells or parietal endodermal PYS cells). While salmon CT and chicken CGRP activate adenylate cyclase activity in F9 cell membranes by 7- to 8-fold, higher concentrations (>1010 M) of these peptides are required to stimulate cAMP formation than are required to mediate the chemotactic effect of these peptides. These data imply the possible involvement of CT gene products in regulating cell migration during early embryonic development. (Endocrinology 129: 2530-2534,1991)
E
MBRYONIC development is mainly the result of five coordinated events: proliferation, adhesion, migration, differentiation, and death of cells. Polypeptide hormones may play a role in regulating these cell functions during early embryogenesis. Teratocarcinoma cells in culture offer an in vitro system to study early biochemical events involved in embryonic development (1). F9 cells derived from OTT 6050 embryonal carcinoma cells resemble the inner cell mass of the early (days 4-5) postimplantation mouse embryo (2). We have previously shown that salmon calcitonin (sCT) appeared to be the most potent stimulator of F9 adenylate cyclase activity (3) and that F9 cells secreted immunoreactive CT (4), suggesting that this hormone may have autocrine effects on these cells. In addition, during the past few years transfection studies in F9 cells have established that splicing mechanisms leading to the production of CT gene-related peptide (CGRP) predominate in F9 cells (5). Cell migrations play a major role during early embryogenesis and involve an adhesion process as well as Received May 20,1991. Address all correspondence and requests for reprints to: Daniele Evain Brion, Laboratoire de Physiopathologie du Developpement, 8eme etage, 46 rue d'Ulm, 75230 Paris Cedex 05, France. * This work was supported by a grant from le Comite de Paris de la ligue Nationale contre le Cancer.
soluble factors that may act as chemotactic agents. Indeed, some growth factors, such as platelet-derived growth factor (PDGF) or transforming growth factor-/? (TGFjS), act to regulate cell growth and differentiation, but also act as chemotactic agents (6, 7). Therefore, it was of interest to study whether the CT gene-related products may have a chemotactic effect on F9 cells. In this paper we demonstrate that different peptides related to CT [i.e. salmon and human (h) CT and chicken and human CGRP] rank in a particular order of potency with respect to their chemotactic effects in F9 cells and their ability to stimulate the adenylate cyclase activity of these cells. These results suggest the possible involvement of CT gene-related products in cell migration during early embryonic development.
Materials and Methods Peptides The peptides human CGRP (hCGRP), chicken CGRP (cCGRP), and hCT were obtained from the Peninsula Laboratory (Belmont, CA). sCT was a generous gift from Sandoz (Paris, France). Higly purified PDGF was purchased from Collaborative Research (Lexington, MA), and TGF/3 was a generous gift from Dr. G. Grotensdorst (Charleston, SC). Cell culture The teratocarcinoma stem cells, F9, PCC4, and the parietal endoderm-like PYS2, kindly provided by Dr. Jetten (NIH,
2530
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
CHEMOTACTIC EFFECT OF CT AND CGRP Bethesda, MD), were grown on 90-mm tissue culture dishes (Costar, Cambridge, MA) at 37 C in Dulbecco-Vogt Modified Eagle's Medium (DMEM; Gibco, Grand Island, NY) containing 10% fetal calf serum (Seromed, Noisy le Grand, France), antibiotics (50 U/ml penicillin and 50 Mg/ml streptomycin), and 4 mM glutamine (Flow, Les Ulis, France) in a 5% CO2 humidified atmosphere. Growth media were changed every 48 h. When indicated, the F9 cells were treated 1 day after plating and for 4 days with a final concentration of 10"7 M all-trans-retinoic acid. The clonal mesodermal line MES-1 P19 (a generous gift of C. Mummery, Hubrecht Laboratory, Utrecht, The Netherlands) was cultured in a 1:1 mixture of DMEM and Ham's F12 medium plus 8% FCS. Human fibroblasts were obtained from skin biopsies, as previously described (8) and cultured in DMEM plus 8% FCS. NIH 3T3 cells were a generous gift of F. Chany (Unite INSERM 43, Paris, France) and were cultured in DMEM plus 8% FCS.
Chemotaxis assay As previously described (9), F9 cells harvested from culture dishes by brief exposure to a solution of 0.05% trypsin/0.02% EDTA were suspended in an equal volume of DMEM with 10% fetal calf serum, centrifuged at 1000 rpm for 5 min, and resuspended in the same medium at room temperature for 1 h with occasional agitation. The cells were then washed twice with serum-free DMEM and suspended in the same medium to which 1% BSA was added at a concentration of 1.5 X 106 cells/ml. A blind well modified Boyden chemotaxis chamber and Nucleopore polycarbonate filters (8 /xm diameter pores) were rinsed to measure F9 cell chemotaxis. The filters were treated with 1% gelatin to promote cell adherence to their upper and lower surfaces. Various concentrations of peptides diluted in DMEM were placed in the lower chamber as attractant (150-/ul volume). The upper chamber compartments were loaded with 200 /xl F9 cell suspension (1.5 x 105 cells). Chemotaxis chambers were incubated at 37 C for 120 min in a humidified atmosphere containing 5% CO2. Filters were then removed from the chambers, and cells were fixed and stained using DiffQuick stain (Ames, Louvreciennes, France). The number of cells on the lower surface of the filter was determined by counting the number of nuclei in 10 fields under a phase contrast microscope. The chemotaxis assay was performed in duplicate and repeated at least five times for each tested peptide. Data are expressed as the chemotactic index, that is the ratio between migration toward test attractants and that toward serum-free DMEM alone. The number of migrating cells in the DMEM alone generally ranged from 2-4 cells/10 fields. Checkerboard analysis has shown that each of the peptides had a specific chemotactic effect and did not increase random motility. For these studies, 10"11 M of each peptide was placed on both side of the filter or in the upper chamber compartment. The number of migrating cells was the
2531
same as that in presence of DMEM alone. Student's t test was used when comparing each concentration of the peptides. Adenylate cyclase assay Cells (day 2 of culture on 90-mm cultured dish) were gently washed with ice-cold homogenization buffer [50 mM Tris-HCl (pH 7.8), 0.33 M sucrose, and 1 mM MgCl2], harvested by scraping, and homogenized by 20 strokes with a Dounce homogenizer (Konteu Co., Vineland, NJ). Microscopic examination revealed essentially complete disruption of each cell type under these conditions. Homogenates were centrifuged at 12,000 X g for 10 min. The pellets were washed twice in the same buffer, and protein content was determined by fluorometric assay (10). Less than 3% of the total adenylate cyclase activity remained in the supernatant. Adenylate cyclase activity was assayed by measuring the conversion of [a-32P]ATP to [32P]cAMP in crude membranes (12,000 X g pellet) at 37 C, as described previously (11). The standard incubation mixture (final volume, 50 pi) contained 25 mM MgCl2, 0.2 mM ATP, 0.5 mM EGTA, 4 X 106 cpm [a-32P]ATP, and 30-50 ng membrane protein. In addition, the reaction mixture contained GTP (10 /XM) and CT or CGRP at the concentrations indicated. Each experiment was repeated at least twice, and the results of triplicate determinations within each experiment agreed within ±10%. Results are expressed as the mean and SEM of two experiments, run in triplicate.
Results Figures 1 and 2 show that cCGRP, hCGRP, and sCT are highly chemotactic for F9 cells, whereas hCT is less effective. In fact, cCGRP reaches the peak chemotactic index of 50 at a concentration of 5 X 10"11 M, although a chemotactic response with this peptide is evident at a concentration as low as 10"15 M (chemotactic index, 10). hCGRP reaches a peak index of 40 at 10"11 M. Similar to that noted with chicken CGRP, chemotactic activity is observed at low concentrations of hCGRP, but the response is less pronounced (chemotactic index, 4 at 10"15 M). SCT exhibits maximal chemotactic activity (chemotactic index, 45) at 10"11 M, while the effective concentration range is relatively narrow (from 10"12-10'9 M). hCT was significantly less active than the other peptides tested in mediating a chemotactic response with F9 cells. As shown in Table 1, CT gene products also were tested on other teratocarcinoma cell lines. cCGRP, hCGRP, and sCT show a chemotactic response toward PCC4 stem cells, but have no chemotactic effects toward the more highly differentiated cells, such as retinoic aciddifferentiated F9 cells, PYS parietal endodermal cells, and MES-1 P19 cells. In F9 cells, CT gene products are much more effective chemotactic agents, than is TGF/3 (chemotactic index, 10 at the optimal concentration of 1010 M). PDGF has no chemotactic activity in F9 cells. As previously shown (9) and confirmed in this study,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
CHEMOTACTIC EFFECT OF CT AND CGRP
2532
Endo • 1991 Voll29«No5
Salmon Calcitonin
Chicken CGRP Chemotaclic index
Chemotactic index 60
60 -i
-16 -15 -14-13-12-11-10-9
-8 -7 -6
-16 -15 -14-13-12-11-10-9
-8 -7 -6
Concentration (log M )
Concentration (log M )
Human Calcitonin Human CGRP
Chemotaclic index 60
Chemotactic index 60
-16 -15 -14-13-12-11-10-9 -16 -15 -14-13-12-11-10-9
-8 -7
-6
-8 -7 -6 Concentration (log M )
Concentration (log M )
FIG. 1. Chemotactic response of F9 cells to CGRP. cCGRP (upper panel) and hCGRP (lower panel) at the concentrations indicated were added to the lower compartment of modified Boyden chambers. The chemotaxis assay was performed as described in Materials and Methods. Chemotactic index = F9 cells migrated toward attractant/F9 cells migrated toward DME1M alone. The results are the mean ± SEM of six experiments.
despite a high chemotactic response (chemotactic index, 24) to PDGF with differentiated MES-1 P19 teratocarcinoma cells, no chemotactic effect of CGRP or CT was observed in these cells. CGRP and CT were also tested on human fibroblasts and NIH 3T3 mouse fibroblasts; they had no chemotactic effect, whereas PDGF was a chemotactic factor toward these cells. As stated in the methodology, the F9 cells migrate only in response to a positive gradient of effective peptide {i.e. sCT, hCGRP, or cCGRP), indicating a true chemotactic effect of these peptides. The checkerboard analysis used helped to eliminate any chemokinetic effect of CT or CGRP, in which cells are stimulated to move more rapidly with a random migration. Since we previously have shown that sCT is the most potent stimulator of F9 adenylate cyclase activity (3) and the presence of CT receptors can serve as a marker of teratocarcinoma stem cells (12), we studied the effect
FIG. 2. Chemotactic response of F9 cells to CT. sCT (upper panel) and hCT (lower panel) at the concentrations indicated were added to the lower compartment of modified Boyden chambers. See Fig. 1 for details.
of CGRP on F9 cell adenylate cyclase activity. As shown in Fig. 3, increasing concentrations of sCT and cCGRP between 10"10-10"6 M show progressive stimulation of adenylate cyclase activity. Both hCGRP and hCT were observed to be less potent in stimulating cAMP synthesis. Interestingly, in the concentration range (1012-10'10 M) that caused the maximal chemotactic effect, we were unable to detect, under the experimental conditions employed, any stimulation of F9 cell adenylate cyclase activity by CT gene products.
Discussion F9 cells are teratocarcinoma cells with a limited capacity to differentiate spontaneously and resemble the inner cell mass of the blastocyst. Previous work in our laboratory suggested a possible role for CT gene-related products in early embryonic development (13, 14). In this study we demonstrate that CT gene products are very potent chemotactic agents in these cells. CT is considered to be a hypocalcemic factor (15). CT and CGRP are derived from an alternative splicing of the same RNA primary transcript. CGRP appears to act
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
CHEMOTACTIC EFFECT OF CT AND CGRP
2533
TABLE 1. Effects of CT gene products on teratocarcinoma cell lines Peptides Cell type Teratocarcinoma cells Stem cells F9 PCC4 Differentiated cells F9 + RA PYS MES-1 P19 Fibroblasts Human NIH 3T3
cCGRP
hCGRP
TGF/3
(lO" 11 M)
(10"n M)
PDGF
(10- 11 M)
(10-10 M)
(lO" 1 0 M)
49 ±4 30 ±6
40 ± 10 25 ± 5
45 ±5 25 ±8
1
ND
5±3 ND
1 1 1
1 1 1
1 1 1
1 1 24 ± 5
ND ND ND
1 1
1 1
1 1
3±2 12 ± 4
4±2 ND
sCT
ND, Not determined.
0
10-12
10-11
10-10
PEPTIDE
10-9
10-8
10-7
10-6
(M)
10-7
10-6
FiG. 3. Effect of increasing hormonal concentration on the adenylate cyclase activity of F9 cell crude membranes. Enzyme activity was determined with 10'6 M GTP and the indicated concentrations of sCT and hCT (A), and cCGRP and hCGRP (B). Each experiment was repeated at least twice, and the results of triplicate determinations within each experiment agreed within ±10%. Results are expressed as the mean and SEM of two experiments, run in triplicate.
principally on the central and peripheral nervous system (16) and can be detected by RIA in the human plasma at a concentration of 10'11 M (17). These different peptides rank in a precise order of potency relative to their effect on the nervous system and bone metabolism, with sCT being more effective than hCT as a hypocalcemic
factor (18). Recently, CT, but not CGRP, has been shown to act as a chemotactic agent for human monocytes (19). In these cells the potency of sCT, as measured by chemotactic index, was approximately 3. In F9 cells the CT gene-related products have a more dramatic chemotactic activity, where a chemoractic index of 50 is observed. As shown in this work, the chemotactic effect of cCGRP is much higher in F9 cells than that of TGFft while PDGF is not effective in these cells. As shown for other chemotactic agents, such as cytokines or growth factors (2022), the chemotactic activity of the peptides studied here is observed only in a restricted range of concentrations (10"13-109 M), with higher concentrations being ineffective. Interestingly, the effective chemotactic concentrations of CGRP or CT are within the physiological range. The CT gene-related products rank in a special order of chemotactic potency, with cCGRP = sCT > hCGRP » hCT. As the sequences of mouse CGRP and CT are not yet established, we were unable to test mouse CGRP and CT. It is, however, interesting to note that, as in the case of CT, nonmammalian CGRP is as active or even more active than its mammalian counterpart, suggesting that they are potent analogs. hCT is a relatively poor chemotactic agent in F9 cells; a 10-fold difference is observed in the chemotactic effect of sCT and hCT. Such a difference between sCT and hCT is also observed in other bioactivities of these hormones, such as the hypocalcemic effect in young rats (23) or in activating adenylate cyclase activity in target cells of the hormone, such as from kidney or bone membranes (24). Similarly, as previously shown (3) and as confirmed in this work, sCT is 5-fold more active than hCT in stimulating the adenylate cyclase activity of the F9 cell membranes. In addition, we demonstrate that cCGRP stimulates cAMP formation, with cCGRP being as effective as sCT. However, this stimulation of cAMP formation occurred at concentrations of the peptides higher than those involved in mediating chemotactic activity. Furthermore, hCT
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
CHEMOTACTIC EFFECT OF CT AND CGRP
2534
and hCGRP, which are almost equipotent in stimulating adenylate cyclase activity, have very different chemotactic effects on F9 cells. This suggests that other signal transduction pathways, such as phospholipid metabolism or Ca2+ mobilization, also may be involved in mediating the chemotactic effect. Following this hypothesis, a cell cycle-dependent coupling of the CT receptor to different G-proteins was demonsrated in a pig kidney cell line (25). However, in endothelial cells, CGRP stimulates adenylate cyclase activity, but has no effect on inositol phosphate formation (26) In conclusion, our study demonstrates that CT gene products are higly chemotactic toward undifferentiated F9 embryonal carcinoma cells and, therefore, may act to modulate cell adhesion and migration. These results suggest a role for these peptides in early embryonic development.
Acknowledgment We wish to thank Dr. W. B. Anderson, NCI, NIH (Bethesda, MD), for constructive comments.
References 1. Martin GR 1975 Teratocarcinoma as a model system for the study of embryogenesis and neoplasia. Cell 5:229-243 2. Martin GR 1980 Teratocarcinomas and mammalian embryogenesis. Science 209:768-771 3. Binet E, Evain D, Anderson WB 1980 Calcitonin-responsive adenylate cyclase in cultured F9 embryonal carcinoma cells. Exp Cell Res 129:466-469 4. Evain Brion D, Binet E, Donnadieu M, Laurent P, Anderson WB 1984 Production of immunoreactive calcitonin and parathyroid hormone by embryonal carcinoma cells: alteration with retinoic acid-induced differentiation. Dev Biol 104:406-412 5. Leff SE, Evans RM, Rosenfeld MG 1987 Splice commitment dictates neuron specific alternative RNA processing in calcitoninl CGRP gene expression. Cell 48:517-524 6. Seppa H, Grotensdort G, Seppa S, Schifmann E, Martin G 1982 Platelet derived growth factor is chemotactic for fibroblasts. J Cell Biol 92:584-588 7. Sporn MB, Roberts AB, Wakerfield LM, De Combrugge B 1987 Some recent advences in the chemistry and biology of transforming growth factor beta. J Cell Biol 105:1039-1045 8. Evain Brion D, Raynaud F, Plet A, Laurent P, Leduc B, Ancerson WB 1986 Deficiency of cyclic AMP dependent protein kinases in human psoriasis. Proc Natl Acad Sci USA 83:5272-5276 9. Liapi C, Raynaud F, Anderson WB, Evain Brion D 1990 High chemotactic response to platelet derived growth factor of a teratocarcinoma differentiated mesodermal cell line. In Vitro Cell Dev
Endo • 1991 Voll29«No5
Biol 26:388-392 10. Bohlen P, Stein S, Dairman W, Unfriend S 1973 Flurometric assay of proteins in the nanogram range. Arch Biochem Biophys 155:213220 11. Anderson WB, Gallo M, Pastan I 1974 Adenylate cyclase activity in fibroblasts transformed by Kirsten or Moloney sarcoma viruses: decreased activity and loss of response to prostaglandin El. J Biol Chem 249:7041-7048 12. Liapi C, Gerbaud P, Anderson WB, Evain Brion D 1987 Altered hormonal responses: markers for embryonal carcinoma stem cells and their differentiated derivative. J Cell Physiol 133:405-410 13. Evain Brion D, Binet E, Anderson WB 1981 Alterations in calcitonin and parathyroid hormone responsiveness of adenylatecyclase in F9 embryonal carcinoma cells treated with retinoic acid and dibutyryl cyclic AMP. J Cell Physiol 109:453-459 14. Binet E, Laurent P, Evain Brion D 1985 F9 embryonal carcinoma cell calcitonin autocrine system: correlation between immunoreactive calcitonin secretion and calcitonin receptor number. J Cell Physiol 124:288-292 15. Milhaud G, Moukhtar MS 1966 Antagonist and synergic action of thyrocalcitonin and parathyroid hormone on the levels of calcium and phosphate in the rat. Nature 211:1186-1187 16. Breimer LH, Maclntyre I, Zaid M 1988 Peptides from the calcitonin genes: molecular genetics, structure and function. Biochem J 255:377-390 17. Mason RT, Shulkes K, Zajac JD, Fletcher AE, Hardys KJ, Martin TJ 1986 Basal and stimulated release of calcitonin gene related peptide in patients with medullary thyroid carcinoma. J Clin Endocrinol Metab 25:675-686 18. Potts JT, Niall HD, Keutmann HT, Destos MJ, Parson JA 1969 Calcitonin: recent chemical and immunological studies. In: Heinemann W (ed) Calcitonin 1969. Heinemann, London, pp 56-73 19. Sacerdote P, Bianchi M, Panerai A 1989 Human monocyte chemotactic activity of calcitonin and somatostatin related peptides: modulation by chronic peptide treatment. J Clin Endocrinol Metab 70:141-148 20. Ye S, Cheug HT 1989 Regulation of lymphocyte motility by macrophages: characreization of a lymphocyte migration inhibitory factor derived from a macrophage-like cell line. Cell Immunol 122:231-243 21. Bussulino F, Ming Wang J, Defilippi P, Turrini F, Savanio F, Edgell CJS, Aglietta M, Arese P, Mantovani A 1989 Granulocyte and granulocyte macrophage colony stimulating factors induce human endothelial cells to migrate and proliferate. Nature 337:471473 22. Wilkinson P, Haston W 1988 Chemotaxis: an overview. Methods Enzymol 162:3-16 23. Hirsch PS, Voelkel ES, Munson PL 1964 Thyrocalcitonin hypocalcemic hypophosphatemic principal of the thyroid gland. Science 146:412-413 24. Marx SJ, Woodard CJ, Aurbach GD 1972 Calcitonin receptors of kidney and bone. Science 176:999-1000 25. Chakraborty M, Chatterjee D, Kellokumpu S, Rasmussen H, Baron R 1991 Cell cycle-dependent coupling of the calcitonin receptor to different G proteins. Science 251:1078-1082 26. Haegerstrand A, Dalsgaard CG, Jonzon B, Larsson O, Nilsson J 1990 Calcitonin gene-related peptide stimulates proliferation of human endothelial cells. Proc Natl Acad Sci USA 87:3299-3303
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
Vol. 129, No. 5 Printed in U.S.A.
Calcitonin and Calcitonin Gene-Related Peptide Are Chemotactic for F9 Embryonal Carcinoma Cells* PASCALE GERBAUD, NADINE SEGOND, MOHSEN M. MOUKHTAR, AND DANIELE EVAIN-BRION Laboratoire de Physiopatholologie du Developpement, CNRS URA 1337, ENS, 75230 Paris Cedex 05; and the Laboratoire d'Endocrinologie Moleculaire, INSERM U113, CHU St. Antoine, 75571 Paris Cedex, France
ABSTRACT. The chemotactic effect of calcitonin (CT) gene products was tested on F9 teratocarcinoma cells, which are an in vitro model of early embryonic development. CT and CT gene-related peptide (CGRP) induce a significant chemotactic response (chemotactic index, 40-50). The order of potency is: chicken CGRP > salmon CT > human CGRP. Human CT is a less potent chemotactic agent (chemotactic index, 15). Compared to other well known peptides with chemotactic activity, such as platelet-derived growth factor (no activity) and transforming growth factor-^ (chemotactic index, 5), CGRP and CT appear to be very active in attracting F9 cells in the Boyden chamber
assay. Interestingly, CT and CGRP exhibit little chemotactic effect toward differentiated teratocarcinoma cells (i.e. retinoic acid-treated F9 cells or parietal endodermal PYS cells). While salmon CT and chicken CGRP activate adenylate cyclase activity in F9 cell membranes by 7- to 8-fold, higher concentrations (>1010 M) of these peptides are required to stimulate cAMP formation than are required to mediate the chemotactic effect of these peptides. These data imply the possible involvement of CT gene products in regulating cell migration during early embryonic development. (Endocrinology 129: 2530-2534,1991)
E
MBRYONIC development is mainly the result of five coordinated events: proliferation, adhesion, migration, differentiation, and death of cells. Polypeptide hormones may play a role in regulating these cell functions during early embryogenesis. Teratocarcinoma cells in culture offer an in vitro system to study early biochemical events involved in embryonic development (1). F9 cells derived from OTT 6050 embryonal carcinoma cells resemble the inner cell mass of the early (days 4-5) postimplantation mouse embryo (2). We have previously shown that salmon calcitonin (sCT) appeared to be the most potent stimulator of F9 adenylate cyclase activity (3) and that F9 cells secreted immunoreactive CT (4), suggesting that this hormone may have autocrine effects on these cells. In addition, during the past few years transfection studies in F9 cells have established that splicing mechanisms leading to the production of CT gene-related peptide (CGRP) predominate in F9 cells (5). Cell migrations play a major role during early embryogenesis and involve an adhesion process as well as Received May 20,1991. Address all correspondence and requests for reprints to: Daniele Evain Brion, Laboratoire de Physiopathologie du Developpement, 8eme etage, 46 rue d'Ulm, 75230 Paris Cedex 05, France. * This work was supported by a grant from le Comite de Paris de la ligue Nationale contre le Cancer.
soluble factors that may act as chemotactic agents. Indeed, some growth factors, such as platelet-derived growth factor (PDGF) or transforming growth factor-/? (TGFjS), act to regulate cell growth and differentiation, but also act as chemotactic agents (6, 7). Therefore, it was of interest to study whether the CT gene-related products may have a chemotactic effect on F9 cells. In this paper we demonstrate that different peptides related to CT [i.e. salmon and human (h) CT and chicken and human CGRP] rank in a particular order of potency with respect to their chemotactic effects in F9 cells and their ability to stimulate the adenylate cyclase activity of these cells. These results suggest the possible involvement of CT gene-related products in cell migration during early embryonic development.
Materials and Methods Peptides The peptides human CGRP (hCGRP), chicken CGRP (cCGRP), and hCT were obtained from the Peninsula Laboratory (Belmont, CA). sCT was a generous gift from Sandoz (Paris, France). Higly purified PDGF was purchased from Collaborative Research (Lexington, MA), and TGF/3 was a generous gift from Dr. G. Grotensdorst (Charleston, SC). Cell culture The teratocarcinoma stem cells, F9, PCC4, and the parietal endoderm-like PYS2, kindly provided by Dr. Jetten (NIH,
2530
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
CHEMOTACTIC EFFECT OF CT AND CGRP Bethesda, MD), were grown on 90-mm tissue culture dishes (Costar, Cambridge, MA) at 37 C in Dulbecco-Vogt Modified Eagle's Medium (DMEM; Gibco, Grand Island, NY) containing 10% fetal calf serum (Seromed, Noisy le Grand, France), antibiotics (50 U/ml penicillin and 50 Mg/ml streptomycin), and 4 mM glutamine (Flow, Les Ulis, France) in a 5% CO2 humidified atmosphere. Growth media were changed every 48 h. When indicated, the F9 cells were treated 1 day after plating and for 4 days with a final concentration of 10"7 M all-trans-retinoic acid. The clonal mesodermal line MES-1 P19 (a generous gift of C. Mummery, Hubrecht Laboratory, Utrecht, The Netherlands) was cultured in a 1:1 mixture of DMEM and Ham's F12 medium plus 8% FCS. Human fibroblasts were obtained from skin biopsies, as previously described (8) and cultured in DMEM plus 8% FCS. NIH 3T3 cells were a generous gift of F. Chany (Unite INSERM 43, Paris, France) and were cultured in DMEM plus 8% FCS.
Chemotaxis assay As previously described (9), F9 cells harvested from culture dishes by brief exposure to a solution of 0.05% trypsin/0.02% EDTA were suspended in an equal volume of DMEM with 10% fetal calf serum, centrifuged at 1000 rpm for 5 min, and resuspended in the same medium at room temperature for 1 h with occasional agitation. The cells were then washed twice with serum-free DMEM and suspended in the same medium to which 1% BSA was added at a concentration of 1.5 X 106 cells/ml. A blind well modified Boyden chemotaxis chamber and Nucleopore polycarbonate filters (8 /xm diameter pores) were rinsed to measure F9 cell chemotaxis. The filters were treated with 1% gelatin to promote cell adherence to their upper and lower surfaces. Various concentrations of peptides diluted in DMEM were placed in the lower chamber as attractant (150-/ul volume). The upper chamber compartments were loaded with 200 /xl F9 cell suspension (1.5 x 105 cells). Chemotaxis chambers were incubated at 37 C for 120 min in a humidified atmosphere containing 5% CO2. Filters were then removed from the chambers, and cells were fixed and stained using DiffQuick stain (Ames, Louvreciennes, France). The number of cells on the lower surface of the filter was determined by counting the number of nuclei in 10 fields under a phase contrast microscope. The chemotaxis assay was performed in duplicate and repeated at least five times for each tested peptide. Data are expressed as the chemotactic index, that is the ratio between migration toward test attractants and that toward serum-free DMEM alone. The number of migrating cells in the DMEM alone generally ranged from 2-4 cells/10 fields. Checkerboard analysis has shown that each of the peptides had a specific chemotactic effect and did not increase random motility. For these studies, 10"11 M of each peptide was placed on both side of the filter or in the upper chamber compartment. The number of migrating cells was the
2531
same as that in presence of DMEM alone. Student's t test was used when comparing each concentration of the peptides. Adenylate cyclase assay Cells (day 2 of culture on 90-mm cultured dish) were gently washed with ice-cold homogenization buffer [50 mM Tris-HCl (pH 7.8), 0.33 M sucrose, and 1 mM MgCl2], harvested by scraping, and homogenized by 20 strokes with a Dounce homogenizer (Konteu Co., Vineland, NJ). Microscopic examination revealed essentially complete disruption of each cell type under these conditions. Homogenates were centrifuged at 12,000 X g for 10 min. The pellets were washed twice in the same buffer, and protein content was determined by fluorometric assay (10). Less than 3% of the total adenylate cyclase activity remained in the supernatant. Adenylate cyclase activity was assayed by measuring the conversion of [a-32P]ATP to [32P]cAMP in crude membranes (12,000 X g pellet) at 37 C, as described previously (11). The standard incubation mixture (final volume, 50 pi) contained 25 mM MgCl2, 0.2 mM ATP, 0.5 mM EGTA, 4 X 106 cpm [a-32P]ATP, and 30-50 ng membrane protein. In addition, the reaction mixture contained GTP (10 /XM) and CT or CGRP at the concentrations indicated. Each experiment was repeated at least twice, and the results of triplicate determinations within each experiment agreed within ±10%. Results are expressed as the mean and SEM of two experiments, run in triplicate.
Results Figures 1 and 2 show that cCGRP, hCGRP, and sCT are highly chemotactic for F9 cells, whereas hCT is less effective. In fact, cCGRP reaches the peak chemotactic index of 50 at a concentration of 5 X 10"11 M, although a chemotactic response with this peptide is evident at a concentration as low as 10"15 M (chemotactic index, 10). hCGRP reaches a peak index of 40 at 10"11 M. Similar to that noted with chicken CGRP, chemotactic activity is observed at low concentrations of hCGRP, but the response is less pronounced (chemotactic index, 4 at 10"15 M). SCT exhibits maximal chemotactic activity (chemotactic index, 45) at 10"11 M, while the effective concentration range is relatively narrow (from 10"12-10'9 M). hCT was significantly less active than the other peptides tested in mediating a chemotactic response with F9 cells. As shown in Table 1, CT gene products also were tested on other teratocarcinoma cell lines. cCGRP, hCGRP, and sCT show a chemotactic response toward PCC4 stem cells, but have no chemotactic effects toward the more highly differentiated cells, such as retinoic aciddifferentiated F9 cells, PYS parietal endodermal cells, and MES-1 P19 cells. In F9 cells, CT gene products are much more effective chemotactic agents, than is TGF/3 (chemotactic index, 10 at the optimal concentration of 1010 M). PDGF has no chemotactic activity in F9 cells. As previously shown (9) and confirmed in this study,
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
CHEMOTACTIC EFFECT OF CT AND CGRP
2532
Endo • 1991 Voll29«No5
Salmon Calcitonin
Chicken CGRP Chemotaclic index
Chemotactic index 60
60 -i
-16 -15 -14-13-12-11-10-9
-8 -7 -6
-16 -15 -14-13-12-11-10-9
-8 -7 -6
Concentration (log M )
Concentration (log M )
Human Calcitonin Human CGRP
Chemotaclic index 60
Chemotactic index 60
-16 -15 -14-13-12-11-10-9 -16 -15 -14-13-12-11-10-9
-8 -7
-6
-8 -7 -6 Concentration (log M )
Concentration (log M )
FIG. 1. Chemotactic response of F9 cells to CGRP. cCGRP (upper panel) and hCGRP (lower panel) at the concentrations indicated were added to the lower compartment of modified Boyden chambers. The chemotaxis assay was performed as described in Materials and Methods. Chemotactic index = F9 cells migrated toward attractant/F9 cells migrated toward DME1M alone. The results are the mean ± SEM of six experiments.
despite a high chemotactic response (chemotactic index, 24) to PDGF with differentiated MES-1 P19 teratocarcinoma cells, no chemotactic effect of CGRP or CT was observed in these cells. CGRP and CT were also tested on human fibroblasts and NIH 3T3 mouse fibroblasts; they had no chemotactic effect, whereas PDGF was a chemotactic factor toward these cells. As stated in the methodology, the F9 cells migrate only in response to a positive gradient of effective peptide {i.e. sCT, hCGRP, or cCGRP), indicating a true chemotactic effect of these peptides. The checkerboard analysis used helped to eliminate any chemokinetic effect of CT or CGRP, in which cells are stimulated to move more rapidly with a random migration. Since we previously have shown that sCT is the most potent stimulator of F9 adenylate cyclase activity (3) and the presence of CT receptors can serve as a marker of teratocarcinoma stem cells (12), we studied the effect
FIG. 2. Chemotactic response of F9 cells to CT. sCT (upper panel) and hCT (lower panel) at the concentrations indicated were added to the lower compartment of modified Boyden chambers. See Fig. 1 for details.
of CGRP on F9 cell adenylate cyclase activity. As shown in Fig. 3, increasing concentrations of sCT and cCGRP between 10"10-10"6 M show progressive stimulation of adenylate cyclase activity. Both hCGRP and hCT were observed to be less potent in stimulating cAMP synthesis. Interestingly, in the concentration range (1012-10'10 M) that caused the maximal chemotactic effect, we were unable to detect, under the experimental conditions employed, any stimulation of F9 cell adenylate cyclase activity by CT gene products.
Discussion F9 cells are teratocarcinoma cells with a limited capacity to differentiate spontaneously and resemble the inner cell mass of the blastocyst. Previous work in our laboratory suggested a possible role for CT gene-related products in early embryonic development (13, 14). In this study we demonstrate that CT gene products are very potent chemotactic agents in these cells. CT is considered to be a hypocalcemic factor (15). CT and CGRP are derived from an alternative splicing of the same RNA primary transcript. CGRP appears to act
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
CHEMOTACTIC EFFECT OF CT AND CGRP
2533
TABLE 1. Effects of CT gene products on teratocarcinoma cell lines Peptides Cell type Teratocarcinoma cells Stem cells F9 PCC4 Differentiated cells F9 + RA PYS MES-1 P19 Fibroblasts Human NIH 3T3
cCGRP
hCGRP
TGF/3
(lO" 11 M)
(10"n M)
PDGF
(10- 11 M)
(10-10 M)
(lO" 1 0 M)
49 ±4 30 ±6
40 ± 10 25 ± 5
45 ±5 25 ±8
1
ND
5±3 ND
1 1 1
1 1 1
1 1 1
1 1 24 ± 5
ND ND ND
1 1
1 1
1 1
3±2 12 ± 4
4±2 ND
sCT
ND, Not determined.
0
10-12
10-11
10-10
PEPTIDE
10-9
10-8
10-7
10-6
(M)
10-7
10-6
FiG. 3. Effect of increasing hormonal concentration on the adenylate cyclase activity of F9 cell crude membranes. Enzyme activity was determined with 10'6 M GTP and the indicated concentrations of sCT and hCT (A), and cCGRP and hCGRP (B). Each experiment was repeated at least twice, and the results of triplicate determinations within each experiment agreed within ±10%. Results are expressed as the mean and SEM of two experiments, run in triplicate.
principally on the central and peripheral nervous system (16) and can be detected by RIA in the human plasma at a concentration of 10'11 M (17). These different peptides rank in a precise order of potency relative to their effect on the nervous system and bone metabolism, with sCT being more effective than hCT as a hypocalcemic
factor (18). Recently, CT, but not CGRP, has been shown to act as a chemotactic agent for human monocytes (19). In these cells the potency of sCT, as measured by chemotactic index, was approximately 3. In F9 cells the CT gene-related products have a more dramatic chemotactic activity, where a chemoractic index of 50 is observed. As shown in this work, the chemotactic effect of cCGRP is much higher in F9 cells than that of TGFft while PDGF is not effective in these cells. As shown for other chemotactic agents, such as cytokines or growth factors (2022), the chemotactic activity of the peptides studied here is observed only in a restricted range of concentrations (10"13-109 M), with higher concentrations being ineffective. Interestingly, the effective chemotactic concentrations of CGRP or CT are within the physiological range. The CT gene-related products rank in a special order of chemotactic potency, with cCGRP = sCT > hCGRP » hCT. As the sequences of mouse CGRP and CT are not yet established, we were unable to test mouse CGRP and CT. It is, however, interesting to note that, as in the case of CT, nonmammalian CGRP is as active or even more active than its mammalian counterpart, suggesting that they are potent analogs. hCT is a relatively poor chemotactic agent in F9 cells; a 10-fold difference is observed in the chemotactic effect of sCT and hCT. Such a difference between sCT and hCT is also observed in other bioactivities of these hormones, such as the hypocalcemic effect in young rats (23) or in activating adenylate cyclase activity in target cells of the hormone, such as from kidney or bone membranes (24). Similarly, as previously shown (3) and as confirmed in this work, sCT is 5-fold more active than hCT in stimulating the adenylate cyclase activity of the F9 cell membranes. In addition, we demonstrate that cCGRP stimulates cAMP formation, with cCGRP being as effective as sCT. However, this stimulation of cAMP formation occurred at concentrations of the peptides higher than those involved in mediating chemotactic activity. Furthermore, hCT
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
CHEMOTACTIC EFFECT OF CT AND CGRP
2534
and hCGRP, which are almost equipotent in stimulating adenylate cyclase activity, have very different chemotactic effects on F9 cells. This suggests that other signal transduction pathways, such as phospholipid metabolism or Ca2+ mobilization, also may be involved in mediating the chemotactic effect. Following this hypothesis, a cell cycle-dependent coupling of the CT receptor to different G-proteins was demonsrated in a pig kidney cell line (25). However, in endothelial cells, CGRP stimulates adenylate cyclase activity, but has no effect on inositol phosphate formation (26) In conclusion, our study demonstrates that CT gene products are higly chemotactic toward undifferentiated F9 embryonal carcinoma cells and, therefore, may act to modulate cell adhesion and migration. These results suggest a role for these peptides in early embryonic development.
Acknowledgment We wish to thank Dr. W. B. Anderson, NCI, NIH (Bethesda, MD), for constructive comments.
References 1. Martin GR 1975 Teratocarcinoma as a model system for the study of embryogenesis and neoplasia. Cell 5:229-243 2. Martin GR 1980 Teratocarcinomas and mammalian embryogenesis. Science 209:768-771 3. Binet E, Evain D, Anderson WB 1980 Calcitonin-responsive adenylate cyclase in cultured F9 embryonal carcinoma cells. Exp Cell Res 129:466-469 4. Evain Brion D, Binet E, Donnadieu M, Laurent P, Anderson WB 1984 Production of immunoreactive calcitonin and parathyroid hormone by embryonal carcinoma cells: alteration with retinoic acid-induced differentiation. Dev Biol 104:406-412 5. Leff SE, Evans RM, Rosenfeld MG 1987 Splice commitment dictates neuron specific alternative RNA processing in calcitoninl CGRP gene expression. Cell 48:517-524 6. Seppa H, Grotensdort G, Seppa S, Schifmann E, Martin G 1982 Platelet derived growth factor is chemotactic for fibroblasts. J Cell Biol 92:584-588 7. Sporn MB, Roberts AB, Wakerfield LM, De Combrugge B 1987 Some recent advences in the chemistry and biology of transforming growth factor beta. J Cell Biol 105:1039-1045 8. Evain Brion D, Raynaud F, Plet A, Laurent P, Leduc B, Ancerson WB 1986 Deficiency of cyclic AMP dependent protein kinases in human psoriasis. Proc Natl Acad Sci USA 83:5272-5276 9. Liapi C, Raynaud F, Anderson WB, Evain Brion D 1990 High chemotactic response to platelet derived growth factor of a teratocarcinoma differentiated mesodermal cell line. In Vitro Cell Dev
Endo • 1991 Voll29«No5
Biol 26:388-392 10. Bohlen P, Stein S, Dairman W, Unfriend S 1973 Flurometric assay of proteins in the nanogram range. Arch Biochem Biophys 155:213220 11. Anderson WB, Gallo M, Pastan I 1974 Adenylate cyclase activity in fibroblasts transformed by Kirsten or Moloney sarcoma viruses: decreased activity and loss of response to prostaglandin El. J Biol Chem 249:7041-7048 12. Liapi C, Gerbaud P, Anderson WB, Evain Brion D 1987 Altered hormonal responses: markers for embryonal carcinoma stem cells and their differentiated derivative. J Cell Physiol 133:405-410 13. Evain Brion D, Binet E, Anderson WB 1981 Alterations in calcitonin and parathyroid hormone responsiveness of adenylatecyclase in F9 embryonal carcinoma cells treated with retinoic acid and dibutyryl cyclic AMP. J Cell Physiol 109:453-459 14. Binet E, Laurent P, Evain Brion D 1985 F9 embryonal carcinoma cell calcitonin autocrine system: correlation between immunoreactive calcitonin secretion and calcitonin receptor number. J Cell Physiol 124:288-292 15. Milhaud G, Moukhtar MS 1966 Antagonist and synergic action of thyrocalcitonin and parathyroid hormone on the levels of calcium and phosphate in the rat. Nature 211:1186-1187 16. Breimer LH, Maclntyre I, Zaid M 1988 Peptides from the calcitonin genes: molecular genetics, structure and function. Biochem J 255:377-390 17. Mason RT, Shulkes K, Zajac JD, Fletcher AE, Hardys KJ, Martin TJ 1986 Basal and stimulated release of calcitonin gene related peptide in patients with medullary thyroid carcinoma. J Clin Endocrinol Metab 25:675-686 18. Potts JT, Niall HD, Keutmann HT, Destos MJ, Parson JA 1969 Calcitonin: recent chemical and immunological studies. In: Heinemann W (ed) Calcitonin 1969. Heinemann, London, pp 56-73 19. Sacerdote P, Bianchi M, Panerai A 1989 Human monocyte chemotactic activity of calcitonin and somatostatin related peptides: modulation by chronic peptide treatment. J Clin Endocrinol Metab 70:141-148 20. Ye S, Cheug HT 1989 Regulation of lymphocyte motility by macrophages: characreization of a lymphocyte migration inhibitory factor derived from a macrophage-like cell line. Cell Immunol 122:231-243 21. Bussulino F, Ming Wang J, Defilippi P, Turrini F, Savanio F, Edgell CJS, Aglietta M, Arese P, Mantovani A 1989 Granulocyte and granulocyte macrophage colony stimulating factors induce human endothelial cells to migrate and proliferate. Nature 337:471473 22. Wilkinson P, Haston W 1988 Chemotaxis: an overview. Methods Enzymol 162:3-16 23. Hirsch PS, Voelkel ES, Munson PL 1964 Thyrocalcitonin hypocalcemic hypophosphatemic principal of the thyroid gland. Science 146:412-413 24. Marx SJ, Woodard CJ, Aurbach GD 1972 Calcitonin receptors of kidney and bone. Science 176:999-1000 25. Chakraborty M, Chatterjee D, Kellokumpu S, Rasmussen H, Baron R 1991 Cell cycle-dependent coupling of the calcitonin receptor to different G proteins. Science 251:1078-1082 26. Haegerstrand A, Dalsgaard CG, Jonzon B, Larsson O, Nilsson J 1990 Calcitonin gene-related peptide stimulates proliferation of human endothelial cells. Proc Natl Acad Sci USA 87:3299-3303
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 03:35 For personal use only. No other uses without permission. . All rights reserved.
