Somatostatin induces rapid contraction of neuroendocrine cells

Somatostatin induces rapid contraction of neuroendocrine cells Jan Saras a, *, Malin Grönberg a, Mats Stridsberg b, Kjell E. Öberg a, Eva T. Janson a a Department of Medical Sciences, Section of Endocrine Oncology, Uppsala University Hospital, Uppsala, Sweden b Department of Medical Sciences, Section of Clinical Chemistry, Uppsala University Hospital, Uppsala, Sweden Received 22 December 2006; revised 20 February 2007; accepted 1 April 2007 Available online 18 April 2007 Edited by Lukas Huber FEBS Letters 581 (2007) 1957 1962 Abstract The peptide hormone somatostatin, as well as the somatostatin analog octreotide, induces rapid morphological changes in neuroendocrine cells. The effect can be detected in less than 2 min: retraction fibers are formed, cells round up and cell cell contacts are broken. Somatostatin-dependent cell contraction is inhibited by Y-27632, indicating that this effect is dependent on Rho kinase. In BON1 cells, the somatostatininduced inhibition of forskolin-induced secretion of chromogranin A is not blocked by Y-27632. It is therefore concluded that the inhibitory effect of somatostatin in forskolin-stimulated cells is not dependent on cell contraction. Ó 2007 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Keywords: Somatostatin; Actin cytoskeleton; Cell contraction; Secretion; Rho kinase 1. Introduction Somatostatin is a tetra deca peptide hormone, first isolated from sheep hypothalamic cells and known to be a potent inhibitor of peptide hormone secretion. Subsequently, other functions such as inhibition of cell growth and inhibition of exocrine pancreatic secretion have been described. Somatostatin acts through specific membrane bound receptors belonging to the seven-transmembrane receptor superfamily and until today five different receptor subtypes have been identified [1]. Because of its potent inhibitory effect on hormone secretion, somatostatin was early proposed to be used in clinical disorders related to excessive hormone secretion. In 1977, the first report of the success of such treatment was published [2]. However, the short half-life of the natural hormone made it difficult to use except in selected cases. In 1982 the first long-acting analog, octreotide, was presented and this compound has been widely used in the clinic [3]. Today somatostatin analogs are used to treat various disorders connected to hypersecretion of hormones such as acromegali and various neuroendocrine tumors. The aim of this treatment is mainly to reduce the symptoms induced by this excessive hormone secretion, but other effects such as growth inhibition are also of interest [4]. Somatostatin-induced relaxation of smooth muscle has been observed in the gastrointestinal system and in blood vessels [5 * Corresponding author. Address: Klinisk Forskningsavd. 2, Lab 14, Entrance 70, 3 trp., Uppsala University Hospital, 751 85 Uppsala, Sweden. Fax: +46 18 553601. E-mail address: jan.saras@medsci.uu.se (J. Saras). 7]. However, it has also been reported that somatostatin can induce contraction of smooth muscle [5,8]. These conflicting results are probably depending on the experimental set up. Also, in experiments employing organisms and tissues, indirect effects may influence the results. The present report is focused on detection of morphological responses of somatostatin in neuroendocrine/hormone-producing cells. The model system used is BON1 cells. This cell line is derived from a metastasis of a human pancreatic neuroendocrine tumor [9]. BON1 cells produce and secrete high levels of chromogranin A, a protein located in dense core vesicles. Somatostatin-induced effects in these cells, including inhibition of chromogranin A secretion, have been reported previously [10]. Here, we show that somatostatin induces a morphological response in neuroendocrine cells. The cells contract rapidly and the effect is shown to be dependent on the Rho kinase. Somatostatin-mediated inhibition of secretion from BON1 cells is not dependent on cell contraction; however, the contraction shows a negative correlation with secretion, suggesting that these events are related. Furthermore, it is shown that lysophosphatidic acid (LPA) is a co-stimulator of forskolin-induced hormone secretion. 2. Materials and methods 2.1. Cell cultivation Human neuroendocrine pancreatic tumour cells, BON1, were cultivated in 50% Dulbecco s modified Eagle s medium and 50% F12 supplemented with 10% foetal bovine serum (FBS) and penicillin/ streptomycin. Human neuroblastoma cells IMR-32 was cultivated in Eagles modified essential medium supplemented with 10% FBS, 1% non-essential amino acids and penicillin/streptomycin. All cells were cultivated at 37 C in an atmosphere containing 5% CO 2. 2.2. Cell stimulations and fluorescence microscopy Cells were seeded in chamber slides and cultivated for 48 h. Prior to stimulation, the cells were incubated in medium devoid of FBS over night. Stimulation with somatostatin-14 (Sigma), Octreotide (Novartis), L-A-lysophosphatidic acid (LPA, Sigma), Y-27632 (Sigma) and forskolin (Sigma) were performed as indicated in the figure legends. After stimulation, the chamber slides were fixed in 3.8% formaldehyde in phosphate-buffered saline (PBS) for 30 min at 25 C and washed in PBS. The cells were permeabilized in 0.2% Triton X-100 in PBS for 5 min and washed in PBS. Blocking was done using 5% FBS in PBS for 1 h at 25 C. Cells were then incubated with a rabbit antibody directed against chromogranin A [19] in 5% FCS in PBS for 1 h at 25 C and washed again in PBS. Incubation with FITC-labeled secondary anti-rabbit antibody (DAKO) and TRITC-labeled phalloidin (Sigma) in 5% FCS in PBS was performed for 1 h at 25 C and subsequently the cells were washed in PBS. The slides were mounted with coverslips by the use of Vectashield (Vector Laboratories). Cells were photographed by an Axiocam HRm camera employing the Axiovision 0014-5793/$32.00 Ó 2007 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2007.04.019

1958 J. Saras et al. / FEBS Letters 581 (2007) 1957 1962 imaging software using a 63 plan-apochromat objective and a Zeiss Axioplan2 microscope. 2.3. Time lapse studies of living cells Cells were seeded in chamber slides and cultivated for 48 h. Prior to stimulation with somatostatin, the cells were incubated in medium devoid of FBS over night. Time lapse series of images were started before stimulation and images were recovered every 10th second using a Canon Powershot digital camera employing the Remote Capture imaging software and a Axiovert 25 microscope (Zeiss) equipped with a phase contrast 40 objective. 2.4. Chromogranin A secretion assay Confluent BON1 cells in 24 well plates were washed with serumfree medium and thereafter stimulated with somatostatin, forskolin, LPA and Y-27632 for 60 min as indicated in the figure legends. The conditioned medium was then aspirated, centrifuged and subjected to a chromogranin A radioimmuno assay (EuroDiagnostica). All experiments were performed in triplicates and repeated at least three times. Variations between samples in an experiment were less than 10%. 3. Results 3.1. Somatostatin induces rapid rearrangement of the actin filament system The effect of somatostatin treatment on neuroendocrine cell morphology was examined using BON1 cells. These cells are derived from a metastasis of a neuroendocrine pancreatic tumor [9]. Previous studies, based on northern blot analysis, have identified mrna for somatostatin receptors 1, 2 and 5 in BON1 cells [10]. In contrast, by using real time polymerase chain reaction, we can detect the expression of all five somatostatin receptors in BON1 cells (data not shown). In our experiments, the expression level of receptor subtype 4 is very low and it is currently unclear whether BON1 cells should be regarded as somatostatin receptor 4 positive. We have previously shown that somatostatin, as well as synthetic somatostatin analogs, induce responses in this cell line [10]. Using fluorescently labeled phalloidin, the actin cytoskeleton was visualized. Fig. 1. BON1 cells were stimulated with 500 nm somatostatin, 100 nm forskolin and 10 lm Y-27632, as indicated in the figure, for 10 min and thereafter fixed and stained with fluorescent phalloidin in order to visualize the actin filament system. Scale bar represents 50 lm.

J. Saras et al. / FEBS Letters 581 (2007) 1957 1962 1959 Five minutes stimulation with 500 nm somatostatin induced numerous thin filopodia-like protrusions in a majority of the cells (Fig. 1, upper panel). Time course studies revealed that the protrusions are actually formed very fast and appear after less than 2 min, peaks at 10 min and is virtually absent after 60 min (data not shown), indicating that this is a rapid and transient somatostatin-mediated response. Octreotide is a long lived and potent somatostatin-analog used in clinical situations where a reduction of hormone circulation is wanted. In BON1 cells, Octreotide had similar effects on the actin cytoskeleton as somatostatin (Fig. 2). 3.2. Morphological effects detected in living cells Next, we performed time lapse studies on living cells using phase contrast microscopy. An image was captured every 10th second. In these experiments it was clear that the main morphological effect of somatostatin in BON1 cells is actually cell contraction. The cells retract, round up, cell cell contacts are broken and membrane blebs are formed (Fig. 3). In Fig. 3, it is possible to detect the contraction of individual cells loosing contact with other cells and cell surface contact, while in the film-like sequence of images in Supplementary material 1 it is clear that most of the cells contract. This phenomenon can be detected at 10 nm somatostatin but is stronger at higher concentrations. Thus, the filopodia-like structures seen in phalloidin-labeled fixed cells are not active protrusions but rather retraction fibers. This finding explains the extremely rapid formation of actin-containing protrusions seen in fixed cells. The rapid morphological effects of somatostatin on hormone-producing cells are not restricted to BON1 cells; contraction and retraction was seen also in the human neuroblastoma cell lines IMR-32 (Fig. 4), SKNAS and Kelly cells (not shown). Thus, this effect of somatostatin is probably present in various hormone-producing cells. Fig. 2. BON1 cells were stimulated with 1 lm somatostatin or 1 lm octreotide for 10 min and thereafter fixed and stained with fluorescent phalloidin in order to visualize the actin filament system. Scale bar represents 50 lm. Fig. 3. BON1 cells were stimulated with 500 nm somatostatin and the morphological response was studied by phase contrast microscopy. Images were captured every 10th second. Arrows indicate cells that contract, loose adhesion and round up. Scale bar represents 50 lm. In the film-like sequence of Supplementary material 1, it is clear that most cells contract in response to somatostatin.

1960 J. Saras et al. / FEBS Letters 581 (2007) 1957 1962 3.3. Somatostatin-induced cell contraction is Rock-dependent Activation of the Rho kinase (Rock) is tightly linked to cell contraction. Rock phosphorylates myosin and myosin phosphatase, both events leading to the activation of myosin [11,12]. The activated myosin contracts actin filaments leading to cell contraction. BON1 cells treated with the low molecular weight compound Y-27632, a specific inhibitor of Rock [13], had similar morphology as unstimulated cells (Fig. 1, lower panel). When cells were stimulated with somatostatin in presence of Y-27632 the contractive effect was inhibited (Fig. 1, lower panel), indicating that the somatostatin-mediated cell contraction is Rock-dependent. This interesting observation raises the question whether Rock is involved also in the regulation of secretion seen in neuroendocrine cells stimulated with somatostatin. 3.4. Somatostatin induces a redistribution of chromogranin A-containing vesicles In order to visualize hormone-containing dense core vesicles, BON1 cells were subjected to immunofluorescence staining of chromogranin A. In Fig. 5, it is shown that a majority of the dense core vesicles in unstimulated cells are located at the periphery of the cells. When cells were stimulated with somatostatin for 10 min, this pattern was altered and most chromogranin A positive vesicles were found in the perinuclear region. This response correlates well in time with the morphological effects and it is likely that the hormone-containing vesicles are translocated to the perinuclear region as the cells round up and retract protrusions. 3.5. Is there a connection between actin filament rearrangements and hormone secretion? Chromogranin A is good marker for exocytosis of dense core vesicles of neuroendocrine cells [14,15]. Since FBS contains chromogranin A, the experiments were performed in serum-free conditions. The amount of chromogranin A in the conditioned media was determined using a radio-immuno assay. Upon stimulation with 1 lm somatostatin the secretion of chomogranin A was decreased by approximately 20%. When cells are incubated in serum-free conditions the level of secretion of dense core vesicles is low. In order to study BON1 cells in situations where the secretion is stimulated a new forskolin-based assay was developed. Forskolin is a strong inducer of regulated exocytosis [16]. In Fig. 6, it is shown that forskolin enhances the secretion of chromogranin A in BON1 cells. Furthermore, the somatostatin-mediated inhibition was significantly increased in forskolin-treated cells, Fig. 4. The morphological response of IMR-32 cells stimulated with 500 nm somatostatin was studied by phase contrast microscopy. Images were captured every 10th second. Scale bar represents 50 lm. The film-like sequence of images in Supplementary material 2 covers the process during 25 min. Note that IMR-32 cells grow in clusterscontaining several cells. Fig. 5. Somatostatin-induced translocation of chromogranin A-containing vesicles was detected by immunofluorscence microscopy. BON1 cells, unstimulated (control) or stimulated for 10 min with 500 nm somatostatin were fixed and subjected to immunofluorescence staining. Cells were visualized by TRITC-labeled (red) phalloidin and chomogranin A was detected by an antibody directed against this protein followed by a FITC-labeled (green) secondary anti-rabbit antibody. Scale bar represents 50 lm.

J. Saras et al. / FEBS Letters 581 (2007) 1957 1962 1961 altered by Y-27632. Therefore, it appears that inhibition of secretion is not dependent on cell contraction. 3.6. Lysophosphatidic acid is a co-stimulator of hormone secretion The small GTPase Rho and its effector Rock are main activators of myosin/actin induced cell contraction [17]. Lysophosphatidic acid (LPA) is a well-known activator of Rho and Rock acting via a G-protein coupled receptor (reviewed in Ref. [18]). In order to determine if Rho/Rock have positive effects on secretion, BON1 cells were stimulated with LPA. This compound alone had no detectable effect on chromogranin A secretion (Fig. 7). In the presence of forskolin, however, LPA had a prominent stimulatory effect on secretion. 4. Discussion Fig. 6. BON1 cells were stimulated with 100 nm forskolin together with somatostatin at concentrations indicated in the figure. Secretion of chromogranin A into the conditional media was determined. Level of secretion is expressed as arbitrary units where the unstimulated sample (no forskolin) is 1. Error bars represent two S.D. leading to a more sensitive assay. The inhibitory effect of somatostatin on chromogranin A secretion could be detected at 10 nm (Fig. 6) which is similar to the lower limit of detection of somatostatin-dependent cell contraction. As most documented effects of somatostatin are related to regulation of secretion, we wanted to study the relation between the somatostatin-induced cytoskeletal and secretory effects. The actin cytoskeleton of BON1 cells stimulated by forskolin appeared as similar to untreated cells (Fig. 1, middle panel). In presence of forskolin the somatostatin-induced contraction was totally absent. In the presence of Y-27632, the secretion from unstimulated cells was reduced by 50% (Fig. 7). This result suggests that Rock actually has a positive effect on exocytosis. However, forskolin could overcome the Y-27632-related inhibition (Fig. 7). The inhibitory action of somatostatin in forskolin stimulated cells was not significantly We have shown that somatostatin can induce cell contraction in various hormone-producing cells. The response is detected at similar concentrations as the inhibitory action of somatostatin on secretion, indicating that somatostatin-induced cell contraction has physiological relevance. That octreotide has similar effects on cell morphology as somatostatin indicates that further studies of this phenomenon are of clinical importance. Also, the finding that chromogranin A-containing vesicles are translocated from the cell periphery to the perinuclear region in response to somatostatin should be further investigated. It is possible that somatostatin-induced inhibition of secretion, in part, is dependent of the observed translocation of secretory vesicles, as the vesicles after stimulation have less contact with the plasma membrane. One should bear in mind though that only a small fraction of the dense core vesicles in an endocrine cell are mature (ready for exocytosis). From this experiment it cannot be concluded that the mature secretory vesicles have reduced contact with the plasma membrane after somatostatin stimulation. It is shown that the somatostatin-induced cell contraction is dependent on Rock activity whereas inhibition of Rock cannot inhibit somatostatin-induced inhibition of secretion. Fig. 7. Confluent BON1 cells were stimulated with 500 nm somatostatin (sst), 100 nm forskolin (for), 2 lm LPA and 10 lm Y-27632 (Y), in combinations as indicated in the figure, for 1 h and the conditioned media was subjected to a chromogranin A radio immune assay. Level of secretion is expressed as arbitrary units where the unstimulated sample is 1. Error bars represent two S.D.

1962 J. Saras et al. / FEBS Letters 581 (2007) 1957 1962 The inhibitory effect of the Rock-inhibitor Y-27632 on basal secretion suggests that the activation of this protein might promote exocytosis. This idea led to the finding that LPA can function as a co-stimulator of regulated secretion in endocrine cells, possibly through the activation of Rho/Rock. To our knowledge, LPA has not previously been implicated in regulation of secretion in neuroendocrine cells. Further studies of the effects of LPA on neuroendocrine cells are needed in order to elucidate the involvement of this natural compound in the regulation of hormone secretion in vivo. Forskolin and somatostatin have opposing effects on both secretion and morphology, suggesting that these events are connected. Forskolin can inhibit somatostatin-mediated cell contraction while somatostatin can inhibit forskolin-induced secretion. The inhibition of forskolin-induced secretion by somatostatin appears not to be dependent on cell contraction. Our study does not support a model where cell contraction is involved in the inhibitory effects of somatostatin on secretion. However, if rapid cell contraction is an immediate mechanism for somatostatin-induced inhibition of secretion, such effects would not be detected in our secretory assays. Further studies of cell contraction by somatostatin are needed in order to increase the physiological understanding of this hormone and its clinically used analogs. Acknowledgements: This work was supported by the Swedish Cancer Society and the Lions Foundation for Cancer Research at the Uppsala University Hospital. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.febslet.2007.04.019. References [1] Patel, Y.C. (1999) Somatostatin and its receptor family. Front Neuroendocrinol. 20, 157 198. [2] Thulin, L., Samnegard, H., Tyden, G., Long, D.H. and Efendic, S. (1978) Efficacy of somatostatin in a patient with carcinoid syndrome. Lancet 2, 43. [3] Bauer, W., Briner, U., Doepfner, W., Haller, R., Huguenin, R., Marbach, P., Petcher, T.J. and Pless (1982) SMS 201-995: a very potent and selective octapeptide analogue of somatostatin with prolonged action. Life Sci. 31, 1133 1140. [4] de Herder, W.W. and Lamberts, S.W. (2003) Somatostatin analog therapy in treatment of gastrointestinal disorders and tumors. Endocrine 20, 285 290. [5] Corleto, V.D. et al. (2006) Somatostatin receptor subtypes mediate contractility on human colonic smooth muscle cells. Neurogastroenterol. Motil. 18, 217 225. [6] John, K.D., Ballantyne, G.H. and Modlin, I.M. (1997) Octreotide acetate inhibits motility in the rabbit distal colon. Eur. Surg. Res. 29, 311 318. [7] Ruiz, E., Padilla, E. and Tejerina, T. (2002) Effect of somatostatin on rabbit isolated coronary arteries. Jpn. J. Pharmacol. 90, 51 58. [8] Torrecillas, G., Medina, J., Diez-Marques, M.L., Rodriguez- Puyol, D. and Rodriguez-Puyol, M. (1999) Mechanisms involved in the somatostatin-induced contraction of vascular smooth muscle cells. Peptides 20, 929 935. [9] Townsend Jr., C.M., Ishizuka, J. and Thompson, J.C. (1993) Studies of growth regulation in a neuroendocrine cell line. Acta Oncol. 32, 125 130. [10] Ludvigsen, E., Stridsberg, M., Taylor, J.E., Culler, M.D., Oberg, K. and Janson, E.T. (2004) Subtype selective interactions of somatostatin and somatostatin analogs with sst1, sst2, and sst5 in BON-1 cells. Med. Oncol. 21, 285 295. [11] Kimura, K. et al. (1996) Regulation of myosin phosphatase by Rho and Rho-associated kinase (Rho-kinase). Science 273, 245 248. [12] Takaishi, K., Matozaki, T., Nakano, K. and Takai, Y. (2000) Multiple downstream signalling pathways from ROCK, a target molecule of Rho small G protein, in reorganization of the actin cytoskeleton in Madin-Darby canine kidney cells. Genes Cells 5, 929 936. [13] Narumiya, S., Ishizaki, T. and Uehata, M. (2000) Use and properties of ROCK-specific inhibitor Y-27632. Methods Enzymol. 325, 273 284. [14] Stridsberg, M., Eriksson, B., Oberg, K. and Janson, E.T. (2003) A comparison between three commercial kits for chromogranin A measurements. J. Endocrinol. 177, 337 341. [15] Feldman, S.A. and Eiden, L.E. (2003) The chromogranins: their roles in secretion from neuroendocrine cells and as markers for neuroendocrine neoplasia. Endocr. Pathol. 14, 3 23. [16] Hermansen, K. (1985) Forskolin, an activator of adenylate cyclase, stimulates pancreatic insulin, glucagon, and somatostatin release in the dog: studies in vitro. Endocrinology 116, 2251 2258. [17] Riento, K. and Ridley, A.J. (2003) Rocks: multifunctional kinases in cell behaviour. Nat. Rev. Mol. Cell Biol. 4, 446 456. [18] Moolenaar, W.H., van Meeteren, L.A. and Giepmans, B.N. (2004) The ins and outs of lysophosphatidic acid signaling. Bioessays 26, 870 881. [19] Stridsberg, M., Hellman, U., Wilander, E., Lundqvist, G., Hellsing, K. and Oberg, K. (1993) Fragments of chromogranin A are present in the urine of patients with carcinoid tumours: development of a specific radioimmunoassay for chromogranin A and its fragments. J. Endocrinol. 139, 329 337.