Eur. J . Biochem. 180,429-433 (1989) 0 FEBS 1989

A human melanoma-derived cell line (IGR39) with a very high number of vasoactive-intestinal-peptide (VIP) receptors 1. Molecular characterization of the binding site

Jost- LUIS, Jean-Michel MARTIN, Assou EL BATTARI, Max REYNIER, Jacques MARVALDI and Jacques PICHON Institut de Chimie Biologique, Unite Associee au Centre National de la Recherche Scientifique no. 202, Universite de Provence, Marseille

(Received July 26/December 1, 1988) - EJB 88 0898

Using m~no['~~I]iodinated vasoactive intestinal peptide ( I 251-VIP), a very high number of specific binding sites for VIP were identified at the surface of the human melanoma cell line IGR39. The Scatchard analysis of competitive displacement experiments between native VIP and '251-VIP was consistent with the existence of two classes of VIP-binding sites. IGR39 cells possess 0.54 x lo6 high-affinity sites with a dissociation constant (Kd) of 0.66 nM and 1.3 x lo6 sites of moderate affinity with a Kd of 4.7 nM. Pharmacological studies indicated that the order of potency in inhibiting '251-VIP binding of the VIP/secretin family peptides was VIP >> peptide histidine methioninamide > human growth-hormone-releasing factor(1-44) > secretin. Glucagon has no effect on the binding of the labelled peptide.

By means of photoaffinity labelling a polypeptide of M , 63000 was characterized. The labelling of this species was completely abolished by native VIP. The order of potency of VIP-related peptides in inhibiting 1251-VIP cross-linking to its receptor was the same as in the competition experiments. The glycoprotein nature of the VIP- binding sites of ICR39 cells has been investigated by affinity chromatography on wheat-germ-agglutinin - Sepharose.

Vasoactive intestinal peptide (VIP), a 28-amino-acid poly- peptide originally isolated from porcine intestine [I], is a mem- ber of the glucagon/secretin family. This family of peptides also includes gastric inhibitory peptide (GIP), growth- hormone-releasing factor (GRF), peptide histidine isoleu- cinamide (PHI) and two peptides purified from Gila monster venom: helodermin and helospectrin [2 - 51. Although first purified from the gut of vertebrates, VIP is now considered as a neuromediator [6]. lmmunohistocheniical studies have shown that neurons containing VIP-like immunoreactivity have a wide distribution in the central and peripheral nervous system of mammals. VIP regulates many biological activities on various organ systems. It stimulates the secretion of water and electrolytes from the gastrointestinal tract, relaxes the smooth muscles and displays several other important biologi- cal activities (for a review see [7]).

Specific VIP-binding sites, with a localization that gener- ally correlates with a VIP-like immunoreactivity, have been characterized in various tissues. VIP receptors have been de-

Correspondence to J. Pichon, Institut de Chimie Biologique, Unit6 Associee au Centrc National de la Recherche Scientifique no. 202, Universite de Provence, 3 Place Victor Hugo, F-13331 Marseille Cedex 3, France

Abbreviations. VIP, vasoactive intestinal peptide; '2s1-VIP, mono[12sI]iodinated vasoactive inlestinal peptide; PHI, peptide histi- dine - isoleucinamide, i.e. a 27-residue peptide with N-terminal histi- dine and C-terminal isoleucinamide; PHM, peptide histidine - methioninamide, i.e. a 27-residue peptide with N-terminal histidine and C-terminal methioninamide; BSA, bovine serum albumin; DME medium, Dulbecco's modified Eagle's medium; hGRF, human growth-hormone-releasing factor (1 - 44); NaC1/Pi, phosphate- buffered saline; NP40, Nonidet P40; PhMeS02F, phenylmethyl- sulfonyl fluoride; WGA, wheat germ agglutinin.

scribed in the gastrointestinal, nervous, urogenital, respira- tory, and immune systems [8 - 121. In numerous tissues, two classes of VIP-binding sites, one of high affinity and low capacity the other of low affinity and high capacity, have been characterized by Scatchard analysis of data from competitive displacement experiments. However, in some other tissues a single class of VIP-binding sites was observed. The dis- sociation constant and the number of binding sites per cell are close to those of the high-affinity binding sites in the two- class systems.

The molecular characterization of the VIP-binding sites of numerous tissues has been carried out mainly by covalent crosslinking experiments [I 1, 13 - 151 or photoaffinity label- ling [16, 171. Data show an important heterogeneity in molec- ular masses of the VIP-binding components thus charac- terized for the same tissue type from different species as well as for different tissues from the same species. This diversity could reflect the various functions of VIP according to the tissue [18] and could explain the different pharmacological properties observed in human or rat [19]. It has been suggested recently that the heterogeneity of the VIP-binding site could be due to a different pattern of glycosylation of an identical polypeptide backbone [20]. The glycoprotein nature of the VIP receptor has indeed been demonstrated in rat liver [21], in mammalian lung [I l l and in the human colonic adenocarcinoma cell line, HT29 122, 231.

In the present paper, we report on the characterization of the VIP receptor of IGR39, a human melanoma-derived cell line [24]. This cell possesses a very high number of VIP-binding sites at its surface. The molecular identification of these VIP- binding sites has been performed by photoaffinity labelling. Evidence for the glycoprotein nature of this molecule is also

430

presented. In contrast, IGR37 cells isolated from a metastasic tumor of the same patient [24] present a very weak binding of ‘2’1-VIP.

MATERIALS AND METHODS

Materiuls

Purified porcine VIP was provided by Professor V. Mutt (Karolinska Institutet, Stockholm, Sweden). Synthetic por- cine VIP, peptide histidine methioninamide (PHM), human growth-hormone-releasing factor(1- 44) (hGRF), porcine se- cretin and glucagon were purchased from Sigma (St Louis, MO, USA). Electrophoretic chemicals and M , standards were from BioRad Laboratories (Richmond, CA, USA). Dul- becco’s modified Eagle’s medium (DME medium) was from Gibco (Cergy Pontoise, France). Fetal calf serum was from Flow Laboratories (Puteaux, France). Wheat-germ- agglutinin - Sepharose (WGA-Sepharose) was purchased from Vector Laboratories (Burlingame, CA, USA).

VIP was iodinated with 1251Na (Amersham, Bucking- hamshire, England) by the chloramine-T method [25]. Purifi- cation of ‘2’I-V1P was performed as previously described [17].

Cell culture

The human melanoma-derived cell lines 1GR37 and IGR39 [24] and the human colonic adenocarcinoma cell line HT29-D4 [26] were routinely cultured in DME medium con- taining 4.5 g . I - ’ glucose, 10% (by vol.) fetal calf serum in a humidified atmosphere of air/C02 (19: 1). Subcultures were obtained by harvesting exponentially growing cells with 0.53 mM EDTA containing 0.05% trypsin in Dulbecco phosphate-buffered saline (NaCI/Pi).

’I- VIP-binding procedure

IGR39 cell monolayers were harvested with 0.53 mM EDTA in NaC1/Pi and washed twice with ice-cold NaCl/Pi containing 0.1 % (massivol.) bovine serum albumin (NaCI/P,/ BSA). Cells (200000/ml) were incubated with 0.05 nM 1251- VIP in 0.5 ml binding medium [DME medium containing 15 mM Hepes pH 7.4, 150 pM phenylmethylsulphonyl fluo- ride (PhMeSO,F), 1 % (massivol.) bovine serum albumin and 0.1 % (mass/vol.) bacitracin] for 3 h at 13 “C. The cells were then washed three times with ice-cold NaCl/P,/BSA and the radioactivity bound to the cells was counted in a y-radiation spectrometer (Kontron). Specific binding was calculated as the difference between the mean of determinations of total binding and binding in the presence of 0.3 pM native VIP. Non-specific binding never exceeds 10- 15% of total binding.

Photouflinity labelling of’the VIP-binding site from IGR39 cells

4-Azidobenzoate derivatives of 12’l-VIP were synthesized as already described [17]. The IGR39 cell monolayers were incubated in the presence of these photoactivatable lZ5I-VIP derivatives in binding medium for 3 h at 13°C. The mono- layers were washed twice with NaCl/Pi and irradiated under a long-wavelength ultraviolet lamp (Philips HP3108) for 5 min. The cells were then washed with NaCl/Pi and solubilized in 0.1 M sodium phosphate buffer pH 6.1 containing 1% (mass/ vol.) Nonidet P40 (NP40), 5 mM MgC12, 5 mM EDTA, 1% (by vol.) 2-mercaptoethanol and 0.1 % (mass/vol.) SDS for 20 min on ice. The whole mixture was then analysed by SDS/ PAGE.

WG A-Sepharose affinity chromutogruphy

Photoaffinity labelled cells were incubated for 45 min at 4°C in the presence of WGA buffer (50 mM Hepes, pH 7.5, 0.16 M NaC1, 1 mM MgC12, 1 mM CaCI2, 150 pM PhMeSO2F, 2 mM benzamidine and 5 pg . ml-l leupeptin) containing 1 % (massivol.) NP40. The soluble fraction was then applied to a WGA-Sepharose column prepared in a I-ml single-use syringe. WGA buffer containing 0.1 YO (mass/vol.) NP40 was run through the column until the radioactivity returned to the baseline. WGA buffer containing 0.5 M N-acetylglucosamine and 0.1 YO (mass/vol.) NP40 was added and column flow was stopped for 30 min. Fractions (0.2 ml) were then collected and monitored for their radioactivity con- tent. Flow-through and specifically eluted fractions were pooled separately and precipitated with 10% (mass/vol.) tri- chloroacetic acid. Precipitated material was washed once with acetone (SO%, by vol., in water), solubilized in SDS-contain- ing sample buffer and analysed by SDSIPAGE.

SDS/ PAGE

12.5% SDSjPAGE was run according to Laemmli [27] under reducing conditions in slab gels 3.5 mm thick. After electrophoresis the gels were stained with Coomassie blue R-250, destained and then dried. Gels were then exposed to X-ray film (Fuji) for 1 week. Calibrations proteins were: myosin ( M , 200000), fl-galactosidase ( M , 116250), phos- phorylase b ( M , 92000), bovine serum albumin ( M , 66200), ovalbumin ( M , 45000), carbonic anhydrase ( M , 31 000), soybean trypsin inhibitor ( M , 21 500) and lysozyme (Mr 14400).

RESULTS

Specific binding oft2’I- VIP on IGR39 cells

The ability of human cell lines HT29-D4, IGR37 and IGR39 to bind 1251-VIP has been investigated. Table 1 shows that a high level of binding could be detected on IGR39 cells when compared to HT29-D4. In contrast, the binding of I2’I- VIP to IGR37 cells was very low.

Table I . ’251-VIP specijic binding to HT29-D4, IGR37 and IGR39 cell lines Cells were cultured and harvested as described in Materials and Methods. The mean amount of specific binding i s indicated for three separate determinations

Cell line lZ5I-VIP specific binding

cpm/l O6 cells cpm/mg protein HT29-D4 25000 k 12000 48000 15000 IGR37 8500+_ 800 9200+_ 900 IGR39 170 000 f 28 000 206000 35000

Results from competition experiments between 12sI-VIP and native VIP are shown in Fig. 1. Native VIP readily inhibit- ed 12’I-VIP binding to IGR39 cells in the 0.03 - 300-nM range. The concentration of native VIP giving half-maximal inhibition of binding (ICs0) was 1.16 f 0.24 nM (mean

SEM from four separate experiments). The Scatchard analysis of the data was curvilinear with upward concavity (Fig. 1, inset), suggesting the presence of two classes of VIP-

43 1

0.30 1

0 50 100 I50 200 250 300 350

B (pM)

.. - w - 1 1 -10 -9 -8 -7 -6

log [ V I P ] (MI

Fig. 1. Competitive displucement of 1z51-VIP from IGR39 cells by nulive VIP. IGR39 cells were incubated with '251-VIP as described in Materials and Methods in the presence of increasing concentrations of native VIP. Results are from one experiment representative of four performed in triplicate and are expressed as the percentage of lZ5I-VIP specifically bound. A Scatchard analysis of the data is shown in the inset: B, bound ligand, F, free ligand

120 1

-00 - 1 1 -10 -9 -8 -7 -6 -5

log [peptidel (MI Fig. 2. Competition for '251-VIP binding by VIP-reluted peptides. IGR39 cells were incubated with '"I-VIP in the presence ofincreasing concentrations of VIP (+), PHM (0), hGRF (W) , porcine secretin (0) or glucagon (A). Results are expressed as the percentage of specifically bound radioactivity. Data are the mean of three indepen- dent experiments performed in triplicate. SEM values are < 10%

binding sites on ICR39 cells: a high-affinity site with a dis- sociation constant, Kd = 0.66 f 0.1 1 nM and a moderate-af- finity site with a Kd = 4.7 0.3 nM. The number of binding sites per cell was (0.54 0.09) x lo6 and (1.3 2 0.2) x lo6 re- spectively.

The specificity of VIP binding to its receptor has been investigated by competition experiments between '251-VIP and peptides structurally related to VIP. Results are shown in Fig. 2. The order of potency of the different peptides in inhibiting 1251-VIP binding was: VIP (1CsO = 1.2 nM) 9 PHM (IC50 = 400 nM) > hCRF (ICSO = 790 nM) > secretin (ICSO > 1000 nM). Glucagon wds not able to inhibit 1251-VIP binding at all for concentrations up to 10 pM.

Molecular churacterizution of the VIP-binding sites of IGR39 cells

The molecular identification of the VIP-binding sites of IGR39 cells has been carried out by photoaffinity labelling

a b c d e f

Fig. 3. Photoajjinity labelling of the VIP-binding site of ICR39 cells. IGR39 cells were incubated with: 4-a~ido-benzoyl-'~~I-VIP deriva- tives for 3 h at 13°C in the absence (lane a) or in the presence of 0.3 pM of native VIP (lane b), 0.3 pM PHM (lane c), 0.3 pM hGRF (lane d), 0.3 pM secretin (lane e) or 3 pM glucagon (lane f) and washed twice with NaCI/Pi. After ultraviolet irradiation, the labelled polypeptides were solubilized and analysed by SDS/PAGE. The gel was stained, destained, dried and then subjected to autoradiography. Reference protein size markers are indicated on the right of the figure

experiments. Monolayers of 1GR39 cells were incubated in the presence of azidobenzoyl derivatives of 251-VIP. After removal of the ligand, the cells were ultraviolet-irradiated and solubilized. The solubilized material was then analysed by SDSjPAGE and autoradiography. A single photoaffinity labelled complex with an M , of 63000 2000 (n = 5 ) was detected on the gel (Fig. 3, lane a). The specificity of the

432

35

m 30

7 25 0

F 20 a 7 15

- 10

5

0

- -

ln N

0 5 10 15 20 25 30 35 40 45 50

F r a c t i o n no Fig. 4. Aj'jfinity chromatography qf the VIP-binding sites on WGA-Sepharose. Conditions are described in Materials and Methods. (A) Bound material was eluted with 0.5 M N-acetylglucosamine in WGA buffer containing 0.1 % (massjvol.) NP40. (B) Unretained material (fraction 14, lane b) and specifically eluted material (rractions 35 -42, lane a) were precipitated with trichloroacetic acid and analysed by SDS/PAGE and autoradiography. Reference protein size markers are indicated on the right of the figure

labelling has been investigated. On the one hand, the labelling of the M,-63000 species was completely abolished when the incubation step was performed in the presence of 0.3 pM native VIP (Fig. 3, lane b). Incubation in the presence of 0.3 pM PHM or hGRF reduced significantly the extent of the labelling of the Mr-63000 complex (Fig. 3, lanes c and d respectively). On the other hand, incubation in the presence of 0.3 pM porcine secretin or 3 pM glucagon did not prevent the binding and linking of VIP derivatives (Fig. 3, lane e and f respectively) and the labelling of the M,-63 000 polypeptide was fully recovered.

Chromutogruphy of VZP-binding proteins on WGA-Sephurose column

In order to determine whether the photoaffinity labelled M,-63 000 polypeptide was glycosylated, we chose to test its ability to be specifically retained on immobilized WGA lectin. For this purpose, photoaffinity labelled VIP receptor was solubilized as described in Materials and Methods and applied on a WGA-Sepharose column. Lectin was able to retain 10 - 12% of the applied radioactivity. When 0.5 M of the com- plementary sugar (i.e N-acetylglucosamine) was run through the column, all of the radioactivity adsorbed on the lectin was eluted (Fig. 4A). SDS/PAGE analysis of the specifically eluted material showed the presence of a single labelled species of M , 63000 (Fig. 4B, lane a). N o crosslinked polypeptides were detected in the flow-through fractions (Fig. 4B, lane b).

DISCUSSION Specific VIP receptors have been characterized in a wide

variety of tissues and cell lines of different species. It appears that the receptor number of the few human cell lines which are known at the present time to have specific VIP-binding sites is rather low compared to isolated normal gastrointesti- nal cells or membrane extracts from other tissues [8, 14, 15, 28, 291.

In the present study we show that IGR39 cells derived from a human superficial melanoma display a high level of

'251-VIP binding. They were able to bind 1251-VIP with a low non-specific binding (10- 15% of total binding). Scatchard analysis of data from competitive displacement experiments between 12'I-VIP and native VIP showed a curvilinear plot with upward curvature. The graph did not present a smooth bend but a rather sharp break around 100 - 150 pM bound ligand. This allowed us to resolve the graph into two straight lines. Therefore, it appears that IGR39 cells possess two classes of VIP binding sites: one of high affinity (& = 0.66 nM; 0.54 x lo6 sites per cell), the other one of low- medium affinity (& = 4.7 nM; 1.3 x lo6 sites per cell). Depending on the material used for the studies, either a single class or two classes of VIP-binding sites have been reported. However until the present study, only one class of high-affinity VIP-binding sites had been characterized with established hu- man cell lines HeLa [28], Molt 4b [15] or HT29 [14]. If the dissociation constants observed with IGR39 cells are in the range of those of other cells and tissues studied so far, the number of binding sites per cell is much higher, especially when compared to other established cell lines.

The competition experiments performed between 1251-V1P and different peptides that belong to the VIP/secretin family showed an order of efficacy similar to that observed with other human cells (i.e. VIP 9 PHM 2 hGRF > secretin). Glucagon had no effect at all on the '251-VIP binding. This is in good agreement with the results obtained with human intestinal epithelial cell plasma membrane [30] and the human colonic adenocarcinoma cell line HT29 [18]. From this point of view it seems that all the human VIP receptors behave similarly and that the human VIP receptor shows a greater specificity for VIP than the VIP receptor of other species. In rat intestinal cell plasma membrane the following order of potency is observed: VIP > PHI > secretin > rat GRF [31].

The molecular characterization of the VIP-binding sites of IGR39 cells has been carried out by photoaffinity labelling with azidobenzoyl derivatives of '251-VIP. A single polypep- tide of M , 63 000 was characterized. Assuming one molecule of VIP linked to the receptor, the molecular mass of the VIP- binding component can be estimated to be 60000. Thus the molecular mass of the VIP-binding polypeptide is significantly

433

lower in IGR39 cells than in HT29 cells or human epithelial intestinal cells ( M , 63000-67000) [14, 17, 181, but still higher than that of the human lymphoblastoid Molt4b T cells ( M , 47000) [15].

When covalent crosslinking experiments using the homo- bifunctional reagent disuccinimidyl suberate were performed on IGR39 cells two polypeptides of M , 60000 (major) and M , 130000- 140000 (minor) were specifically labelled [32]. A similar phenomenon has been reported for HT29 cells. With these cells a single polypeptide of M , 67000 was observed after photoaffinity labelling [I71 whereas two species of M , 63000 (major) and M , 3 17 000 (minor) were observed after covalent crosslinking using disuccinimidyl suberate or its cleavable analog dithiobis(succinimidy1 propionate) [14, 331. The label- ling of the high-molecular-mass component was probably artefactual as has already been suggested [17].

The presence of a single photolabelled polypeptide ( M , 60000) suggests that the VIP receptor of IGR39 cells contains a single species of binding protein subunit, despite the characterization of two classes of VIP-binding sites by competitive displacement experiments. This could be ex- plained in three different ways: (a) a low efficiency of forma- tion of the covalent bond between 1251-VIP and its low-affinity binding site, leading to the selection of one homogeneous population; (b) both populations are formed by different entities with similar molecular mass; (c) both populations are formed by a unique molecular entity, but with different localization or surrounding (coupling to the adenylate cyclase system for example) at the cell surface membrane inducing differential accessibility to the ligand or different binding properties.

The glycoprotein nature of the VIP-binding site seems to be a common characteristic of the receptor of various tissues [II, 21 -231. The affinity labelled VIP-binding site of IGR39 cells can be retained on immobilized WGA and specifically eluted with N-acetylglucosamine. This result strongly suggests that the VIP-binding site of IGR39 cells possesses a carbo- hydrate moiety. From this point of view it would be very interesting to determine whether the M,-47 000 deglycosylated form characterized using HT29 cells 1221 is also obtained with IGR39 cells. This point is now under investigation in our laboratory. It may confirm the hypothesis of Luis et al. [20] who postulate the existence for the VIP receptor of a single polypeptide core with different carbohydrate moiety, thus leading to different molecular mass depending on the tissues or the species used for the studies.

In conclusion, we show in the present paper that IGR39, a human tumor cell line, possesses specific VIP-binding sites whose structural and binding characteristics display similari- ties but also differences with other human VIP receptors so far studied. Similarities are: (a) the order of specifity for VIP- related peptides; (b) the glycoprotein nature. Dissimilarities are (a) the molecular mass of the affinity labelled polypeptide; (b) the very high number of receptors available at the cell surface. This last feature is very important in considering further studies such as solubilization in an active form and purification.

We thank Dr Christian Aubcrt and coworkers for IGR37 and IGR39 cell lines. We also thank Miss Jeanine Secchi, Mr Fernand Giannellini and Mr Roger Soirat for expert technical assistance. This work was supported in part by the Centre National de fa Recherche Scientifique (CNRS UA 202), the Institut National de la SuntC. et de la Recherche Mhdicale (INSERM grant 847006) and by the Associ- ution pour le Dhveloppement de f a Recherche sur le Cancer (ARC grant 61 87).

REFERENCES

1. Said, S. I. & Mutt, V. (1970) Science (Wash. DC) 169, 1217- 1218.

2. Brown, J. C. &Dryburgh, J. R. (1971) Can. J . Biochem. 49,867- 872.

3. Guillemin, R., Brazedu, P., Bohlen, P., Esch, F., Ling, N. & Wehrenberg, W. B. (1982) Science (Wash. DC) 218, 585-587.

4. Tatemoto, K . & Mutt, V. (1981) Proc. Natl Acad. Sci. USA 78,

5. Parker, D. S., Raufman, J. P., O’Donohue, T. L., Bledsoe, M., Yoshida, H. & Pisano, J. J. (1984) J . Biol. Chem. 259, 11 751 - 11 755.

6603 - 6607.

6. Rostene, W. H. (1984) Prog. Neurobiol. (0xf.j 22, 103-129. 7. Said, S. I. (1984) Peptides (Fayettvillej 5 , 143-150. 8 . Prieto, J. C., Laburthe, M. & Rosselin, G. (1979) Eur. J . Biochem.

9. Taylor, D. P. & Pert, C. B. (1979) Proc. Nutl Acad. Sci. USA 76,

10. Ottesen, B., Staun-Olsen, P., Gammeltoft, S. & Fahrenkrug, J. (1982) Endocrinology 110, 2037-2043.

11. Dickinson, K. E., Schachter, M., Miles, C. M. M., Coy, D. H. & Sever, P. S. (1986) Peptides (Fuyettville) 7, 791 -800.

12. Ottaway, C. A. & Greenberg, G. R. (1984) J . Immunol. 132,417- 423.

13. Laburthe, M., Breant, B. & Rouyer-Fessard, C. (1984) Eur. J . Biochem. 139, 181 - 187.

14. Muller, J. M., Luis, J., Fantini, J., Abadie, B., Giannellini, F., Marvaldi, J. & Pichon, J. (1985) Eur. J . Biochem. 151, 411- 417.

15. Wood, C. L. &O’Dorisio, M. S. (1985) J . Biol. Chem. 260,1243- 1247.

16. Robichon, A,, Kuks, P. F. M. & Besson, J. (1987) J . Biol. Chem. 262,11539 - 11 545.

17. Martin, J. M., Darbon, H., Luis, J., El Battari, A., Marvaldi, J. 81 Pichon, J. (1988) Biochem. J. 250, 679-685.

18. Couvineau, A. & Laburthe, M. (1985) Biochem. J . 225,473 -479. 19. Couvineau, A. & Laburthe, M. (1985) J . Clin. Endocrinol. Metuh.

20. Luis, J., Martin, J. M., El Battari, A., Marvaldi, J. & Pichon, J.

21. Nguyen, T. D., Williams, J. A. &Gray, G. M. (1986) Biochemistry

22. El Battari, A., Luis, J., Martin, J . M., Fantini, J., Muller, J. M.. Marvaldi, J. &Pichon, J . (1987) Biochem. J . 242, 185-191.

23. El Battari, A., Luis, J., Martin, J . M., Fantini, J., Muller, J . M., Marvaldi, J. & Pichon, J. (1988) Ann. N.Y. Acad. Sci. 527,

24. Aubert, C., Rouge, F. & Galindo, J . R. (1980) J . Nut/ Cancer

25. Hunter, W. M. & Greenwood, F. C. (1962) Nuture (Lond.) 194,

26. Fantini, J., Abadie, B., Tirard, A., Rerny, L., Ripert, J. P., El

27. Laemmli, U. K. (1970) Nature (Lond.) 227, 680-685. 28. Prieto, J. C., Pintado, E., De Miguel, L. C., Guerrero, J . M. &

Goberna, R. (1981) Biochem. Biophys. Res. Commun. 103,

29. Gammeltoft, S., Staun-Olsen, P., Ottensen, B. & Fahrenkrug, J.

30. Laburthe, M., Amiranoff, B., Boige, N., Rouyer-Fessard, C.,

31. Laburthe, M. (1985) Biochimie (Paris) 67, XI-XVII. 32. Luis, J., Martin, J. M., El Battari, A., Marvaldi, J. & Pichon, J .

(1989) Les actes du congr&, 22me ,forum Peptides, Collection INSERM, Paris, in the press.

33. Marvaldi, J., Luis, J., Muller, J . M., El Battari, A., Fantini, J., Martin, J. M., Abadie, B., Tirard, A. & Pichon, J. (1986) Peptides (Fayettville) 7, (Suppl. I ) , 137- 145.

96,229-237.

660 - 664.

61,50- 55.

(1988) Biochimie (Paris) 70, 1311 -1322.

25,361 -368.

667 - 671.

Inst. 64, 1029-1037.

4Y 5 - 496.

Battari, E. & Marvaldi, J. (1986) J . Cell Sci. 83, 235-249.

799 - 805.

(1983) Regul. Pept. 6, 303.

Tatemoto, K. & Moroder, L. (1983) FEBS Lett. 159, 89-92.