Prolonged Activation of Extracellular Signal-regulated Kinase by a Protein Kinase C-dependent and N17Ras-insensitive Mechanism Mediates the Proliferative Response of Gi/o-coupled Somatostatin sst4 Receptors*

The human sst4 receptor, recombinantly expressed in Chinese hamster ovary cells, mediates proliferative activity of the peptide hormone somatostatin. This effect was shown to involve activation of pertussis toxin-sensitive G proteins and was inhibited by overexpression of the βγ-sequestrant, transducin. Somatostatin-induced proliferation was abolished by the MEK1 inhibitor, PD 98059, whereas the Src inhibitor, PP1, had no effect. A marked increase was observed in the phosphorylation of extracellular signal-regulated kinase 1 and 2 (ERK1 and ERK2) 10 min after sst4 receptor activation, which was blocked by pertussis toxin, decreased by PP1 and the βγ-sequestrant, but unaffected by PD 98059. In contrast, the somatostatin-induced phosphorylation of ERK obtained at 4 h, although sensitive to both pertussis toxin and transducin, was unaffected by PP1 but ablated by PD 98059. Protein kinase C inhibition also abolished this somatostatin-induced sustained phosphorylation of ERK, together with the associated increase in cell proliferation. Expression of dominant negative Ras (N17) failed to significantly reduce the proliferative effect mediated by the sst4 receptor but markedly attenuated the acute phase of the somatostatin-induced phosphorylation of ERK obtained at 10 min. In contrast, the phosphorylation induced at 4 h was unaffected. We conclude that ERK activation by Gi/o-coupled sst4receptors involves a Src and Ras-dependent acute phase, but the proliferative response is dependent upon the prolonged ERK-induced activity, mediated by protein kinase C.

The peptide hormone somatostatin induces numerous biological actions, most of which are inhibitory, by interacting with cell membrane receptors of which five types, named sst [1][2][3][4][5] , have been heterologously expressed in different cells within the last few years (1). The growth inhibitory effects of somatostatin are well documented as it is considered to be the physiological regulator of growth hormone release (2). As well as reducing the circulatory levels of several other potential mitogenic hormones and growth factors (3), somatostatin has also been shown to have a direct action on cellular proliferation and tissue development, with therapeutic potential in retarding the growth of tumor (4) and vascular smooth muscle cells (5).
Numerous reports have demonstrated the expression of a high density of somatostatin receptors on a variety of human cancer cells, including most tumors of neuroendocrine origin (including gastroenteropancreatic tumors), small cell lung carcinomas, brain tumors (glial tumors and meningiomas), lymphomas, and melanomas as well as colorectal, pituitary, kidandney, and breast tumors (6,7). The antiproliferative action of either somatostatin or its analogue octreotide, however, does not correlate with this expression, having inhibitory actions on pancreatic (8) and breast tumors (9), although eliciting no effect on the growth of small cell lung (10) and colon tumors (11). Growth-promoting effects of somatostatin have also been described in vitro on human pancreatic carcinoid (12) and epidermoid carcinoma cells (13), whereas in both rat mesangial cells (14) and human pancreatic MIA-Pa-Ca-2 cells (15), somatostatin stimulates proliferation in the absence of serum but inhibits the growth of proliferating cells.
Little is known as to the identity of the receptor types mediating the proliferative or antiproliferative responses of somatostatin in tissues, and information has been restricted to studies involving partially selective receptor analogues (16). Activation of either mouse recombinant sst 2 or sst 5 receptors, however, has been shown to inhibit serum-induced proliferation (17), whereas stimulation of the human recombinant sst 4 receptor type induces proliferation in the absence of other mitogenic agents (18). Interestingly, the recently cloned rat sst 2(b) receptor splice variant has also been shown to induce a proliferative response, in marked contrast to the antiproliferative property mediated by the rat sst 2(a) receptor following recombinant expression in the same host cell line (19).
The molecular determinants that mediate the proliferative outcome of somatostatin receptors have not yet been fully clarified. All five human receptor types are functionally coupled to inhibition of adenylate cyclase via pertussis toxin-sensitive G proteins (20) and can mediate phospholipase C activation with subsequent calcium mobilization (21). Stimulation of sst 1 and sst 2 receptors has been shown to activate a protein-tyrosine phosphatase activity (22), and it has been suggested that such an activity may counteract the growth-promoting properties of receptors containing an intrinsic tyrosine kinase domain (23). The inhibition of basic fibroblast growth factor-stimulated proliferation by activation of human sst 1 receptors has also been proposed to be due to the induction of the cell cycle inhibitor, p21 cip1/WAF1 , shown in a recombinant system following the synergistic activation of extracellular signal-regulated kinase * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  (ERK) 1 by the growth factor in the presence of somatostatin (24).
We have previously demonstrated that stimulation of recombinantly expressed sst 4 receptors by somatostatin gives rise to both a marked transient increase as well as a sustained period of ERK1 and ERK2 phosphorylation (18). ERK phoshorylation at threonine 183 and tyrosine 185 are widely used indices of mitogen-activated protein (MAP) kinase activation by the ERK kinase, MEK1. We have also provided evidence using a somatostatin analogue that induces only transient phosphorylation of ERK following sst 4 receptor activation that it is the sustained component of MAP kinase activity that is critical for the induced proliferative response. Other G i protein-coupled receptors (25) have recently been shown to utilize the MAP kinase cascade through a Src-dependent mechanism following release of ␤␥ subunits. In this study therefore, we have examined the ability of the human sst 4 receptor to activate intracellular signaling components that converge on the MAP kinase cascade and in particular to see if a differential requirement can be demonstrated for their involvement in mediating the acute or prolonged phases of ERK phosphorylation. In addition, the resultant effect on cell proliferation of any change detected in the somatostatin-induced ERK phosphorylation following effector inhibition was evaluated by direct cell counting using a model to determine the re-population of denuded areas in a previously confluent monolayer (16). The effect of somatostatin on basal proliferation in the absence of exogenously added mitogenic agents was examined, and responses induced by basic fibroblast growth factor (bFGF) were used as a comparison.

EXPERIMENTAL PROCEDURES
Materials-Chinese hamster ovary (CHO K1) cells were obtained from The European Collection of Animal Cell Cultures. Geneticin, Li-pofectAMINE TM , and culture reagents were supplied by Life Technologies, Inc. Thermanox coverslips were manufactured by Nunc; otherwise Costar tissue culture plastic ware was used. Platelet-derived growth factor (PDGF-BB), bFGF, and anti-␤ actin antibodies were supplied by Sigma. Somatostatin was custom-synthesized by Peninsula Laboratories. Genistein, lavendustin A, lavendustin B, PD 98059, Ro 31-8220, phorbol 12-myristate 13-acetate, and Bordetella pertussis toxin were purchased from Calbiochem. PP1 was supplied by Biomol Research Laboratories, Inc. Antibodies to ERK1 (C-16) and ERK2 (C-14) were obtained from Santa Cruz Biotechnology, whereas that specific to the phosphoforms was supplied by New England Biolabs. Antibodies specific for Ras (L2 region) and human Ha-ras cDNA (dominant negative S17N) in pUSEamp together with the empty vector were supplied by Upstate Biotechnology. The eukaryotic expression vector pCDNA3 and that incorporating transducin cDNA were kind gifts of Alan Wise, Receptor Systems, Glaxo Wellcome Medicines Research Center, Stevenage, UK. An appropriate antibody for monitoring expression levels of transducin was purchased from NEN Life Science Products.
Stable Expression of Somatostatin sst 4 Receptors in Chinese Hamster Ovary Cells-The cDNA encoding the human sst 4 receptor was subcloned into the mammalian expression vector pCIN4 harboring a neomycin-resistant gene as a selection marker. CHO K1 cells were cultured in Dulbecco's modified Eagle's medium/Ham's F-12 (1:1) containing 10% fetal calf serum and 1 mM Glutamax I, and clonal lines stably overexpressing the sst 4 receptor were prepared as described previously (18). Levels of receptor density remained constant over the time course of the study, estimated at 2.07 Ϯ 0.43 pmol/mg of membrane protein (n ϭ 3). All cultures were routinely maintained in the presence of the selection agent G418 sulfate (specific activity 500 g/ml) at 37°C in humidified air containing 5% carbon dioxide and passaged when 95% confluence was reached.
Partial Denudation of Confluent Cell Monolayers and Assessment of Proliferation-Cells were grown to confluence in complete media on Thermanox coverslips. Multiple parallel areas (400-m wide) were denuded of cells, by dragging a Perspex comb across the surface of each coverslip according to the method described previously (18,26). Coverslips were washed in phosphate-buffered saline and placed in a fresh well containing drug or vehicle in media without serum. Cells were harvested following incubation for 24 h by washing the coverslip as above and adding 0.05% trypsin, 0.02% EDTA solution for 2-5 min. The digestion process was terminated by adding complete media and the single cell suspension was counted using a Coulter Counter TM model Z1. Results were calculated from a minimum of three experiments with four replicates per test group and expressed as the arithmetic mean Ϯ S.E. of the mean. Statistical analysis was by analysis of variance followed by Tukey's test (SigmaStat version 2).
Determination of Change in the Phosphorylation Status of ERK1 and ERK2-To analyze changes in the phosphorylation status of ERK1 and ERK2 at various stages during the proliferative processes following partial denudation, whole cell protein extract was combined from four coverslips for each treatment group. Proliferation was terminated by washing the transfected CHO K1 cell monolayers in ice-cold phosphatebuffered saline before applying SDS-polyacrylamide gel electrophoresis sample buffer (50 l of 3ϫ strength) to each test well (1ϫ sample buffer: 4% SDS, 5% glycerol, 60 mM Tris, and 0.01% bromphenol blue, pH 6.8) under reducing conditions (50 mM 2-mercaptoethanol). After solubilization of cellular protein by rapid mixing, the well contents were transferred to a separate tube and combined with two further washings of the well with deionized water (50 l). Samples were vortexed, clarified by centrifugation at 10,000 g for 2 min, and heated at 95°C for 5 min. Total cell protein for each of the extracts was measured by microBCA (Pierce), and equivalent amounts of protein were electrophoretically resolved on 10% polyacrylamide gels.
Following electrophoretic transfer onto nitrocellulose (0.22 m) using a semi-dry blotter, the membrane was washed briefly in Trisbuffered saline (TBS: 50 mM Tris, 250 mM NaCl, pH 7.5) and saturated overnight in TBS supplemented with 0.1% Tween 20 and 5% dried milk. For detection with the antibodies to ERK1 and ERK2, the membranes were incubated with a 1:2,000 dilution (1:1 mix of ERK1 and 2) or a 1:1,000 dilution of the anti-phospho-ERK antibody. When used, the antibody to ␤ actin was at a 1:5,000 dilution. All primary incubations were for 1 h at 22°C in TBS containing 0.1% Tween 20 followed by washing five times for 10 min each in TBS containing 0.1% Tween 20. Membranes were incubated for 1 h at 22°C with a 1:3,000 dilution of the appropriate horseradish peroxidase-conjugated secondary antibody in TBS containing 0.1% Tween 20 and 5% dried milk. Excess antibody was removed by washing as above, and immunocomplexes were visualized using enhanced chemiluminescence (ECL) detection according to the manufacturer's instructions (Amersham Pharmacia Biotech).
Transient Expression of Dominant Negative Ras or Transducin-Human Ha-ras (S17N) cDNA was inserted as an EcoRI fragment into pUSEamp under the control of the cytomegalovirus promoter and used to transiently transfect CHO K1 cells expressing the human recombinant sst 4 receptor. Briefly, cells at 50% confluence in serum-free media were transfected with 2 g of DNA following complex formation with LipofectAMINE TM reagent, according to the manufacturer's instructions. The DNA-containing media was removed following incubation for 3 h at 37°C, and the cells were incubated for an additional 24 h in complete media before transferring onto coverslips. Gene expression using immunoblot analysis as described above was determined immediately before partial denudation, approximately 48 h post-transfection using a primary antibody concentration of 1:1,000. Transfection with transducin was carried out by the same methodology using 1 g of cDNA cloned in pCDNA3. 4 Receptors-Following partial denudation of a confluent monolayer, the total number of CHO K1 cells recombinantly expressing human sst 4 receptors that remained on a single coverslip was 152 Ϯ 3 ϫ 10 3 . After 24 h in the presence of incomplete media, this number had slightly increased to 166 Ϯ 5 ϫ 10 3 , with less than 0.6% of the cells detaching from the coverslip over the time course examined. Application of somatostatin (100 nM) immediately following denudation in the absence of other exogenously added mitogenic factors caused a significant increase in cell number ( Table I) that was comparable to that induced by bFGF (Table I) using a concentration (10 ng/ml) that produced 80% of its maximal response.

Effect of Somatostatin on the Proliferation of Chinese Hamster Ovary Cells Recombinantly Expressing Human sst
Effect of Pertussis Toxin Pretreatment-Pretreatment for 20 h with pertussis toxin (100 ng/ml) had no significant effect on either basal or bFGF-induced (10 ng/ml) proliferation (Table  I). However, the increased proliferation induced by somatostatin (100 nM) was abolished following pretreatment with the toxin to values not significantly different from basal ( Table I).
Effect of Protein-tyrosine Kinase Inhibitors-Neither tyrosine kinase inhibitor, genistein (50 M) nor lavendustin A (11 nM), had any significant effect on basal proliferation (Table I) in the absence of exogenous growth factors. However, both inhibitors abolished the increase in proliferation induced by either somatostatin (100 nM) or bFGF (10 ng/ml) (Table I). Importantly, the inactive enantiomer of lavendustin A, lavendustin B (11 nM), had no effect on basal cell numbers or on the proliferation induced by either somatostatin or bFGF ( Table I).
The Src-family inhibitor, PP1 (200 nM), had no significant effect on basal proliferation (Table I). The increased cell counts induced by either somatostatin or bFGF were also unaffected by incubation with PP1 (Table I). Pretreatment of the cells with PP1 for 1 h before partial denudation, again elicited no significant effect on proliferation induced by somatostatin (237 Ϯ 13 ϫ 10 3 and 242 Ϯ 5 ϫ 10 3 in the presence and absence of PP1, respectively) or bFGF (240 Ϯ 10 ϫ 10 3 and 247 Ϯ 4 ϫ 10 3 ). As a positive control for PP1, proliferation induced by plateletderived growth factor (PDGF-BB) was examined. The PDGFevoked increase in cell number (using 5 ng/ml) was significantly reduced by PP1 from 268 Ϯ 7 ϫ 10 3 to 185 Ϯ 10 ϫ 10 3 .
Changes to the Phosphorylation Status of ERK1 and ERK2-Activation of MEK1 in mediating the proliferative action of somatostatin was substantiated by assessing change to the phosphorylation status of ERK1 and ERK2 using an antibody that recognizes only the dually phosphorylated, active form of ERK. To distinguish between effects on either the acute or prolonged phases of MAP kinase activation, the phosphorylation status of ERK1 and ERK2 was examined at 10 min and 4 h following partial denudation. For all treatment groups there was no detectable change in the expression of ERK1 or ERK2 protein at the time points investigated (Fig. 1A). An increase in the phosphorylation status over basal was observed for both ERK1 and ERK2 following somatostatin treatment (100 nM), but that induced at 10 min was considerably greater than that observed at 4 h post-denudation (Fig. 1, B and C), consistent with previous observations (18). The enhanced phosphorylation of ERK1 and ERK2 induced by somatostatin at both time points was abolished by pertussis toxin pretreatment (20 h at 100 ng/ml) and unaffected by genistein (50 M) (Fig. 1, B and C). Neither pertussis toxin nor genistein had any detectable effect on the basal level of phosphorylation of ERK1 or ERK2 at either time point (Fig. 1, B and C).
Differential effects of PD 98059 (2 M) and PP1 (200 nM) treatments, however, were observed on the somatostatin-induced increase in ERK phosphorylation at the times examined. PD 98059 had no detectable effect on basal or somatostatininduced phosphorylation obtained 10 min post-denudation (Fig. 1B). In contrast, the somatostatin-mediated increase was reduced following PP1 treatment, with the Src inhibitor showing no detectable effect on basal phosphorylation levels (Fig.  1B). After 4 h of regenerative processes, the somatostatininduced phosphorylation of ERK1 and ERK2 was unaffected by PP1 but abolished by PD 98059, with neither inhibitor having any observable effect on the basal level of phosphorylation (Fig. 1C).
Effect of a Dominant Negative Mutant of Ras on Cell Proliferation and ERK Phosphorylation-To evaluate the involvement of Ras in mediating the activation of ERK by somatostatin sst 4 receptors, transient expression of the dominant negative mutant of Ras (N17) was performed. This Ras mutant, in which amino acid 17 (serine) is changed to asparagine, is thought to function by inhibiting guanine nucleotide exchange factors (27). Transient transfection with the empty vector had no significant effect on the proliferation induced in CHO K1 cells expressing the sst 4 receptor by either somatostatin (100 nM) or bFGF (10 ng/ml), as determined by counting the number of cells forming the regenerating monolayers 24 h after partial denudation ( Fig. 2A). Transient expression of N17Ras also failed to significantly effect the increase in cell number evoked by somatostatin treatment, but in contrast that induced by bFGF was markedly attenuated ( Fig. 2A). Transfection with either the empty vector or that incorporating N17Ras cDNA failed to significantly effect basal cell counts (data not shown). The increase in N17Ras levels after transfection was evaluated by immunoblotting cell extracts immediately before partial denudation with a polyclonal antibody to Ras. The inset in Fig.  2A shows that in mock-transfected cells, the immunoreactivity with the anti-Ras antibody was almost undetectable compared with the intense reactivity obtained from the same number of cells transfected with pUSEamp(ϩ) plasmids containing dominant negative Ha-ras.
The level of phosphorylation of ERK1 and ERK2 induced by somatostatin (100 nM) at either time point examined (10 min and 4 h) was unchanged by transfection with the empty plasmid (Fig. 2B). After transient expression of N17Ras, the somatostatin-induced phosphorylation observed 10 min after partial denudation was substantially decreased, whereas that obtained at 4 h was unaffected (Fig. 2B). Neither transfection with pUSEamp or pUSEamp(N17ras) showed any effect on basal levels of ERK phosphorylation observed at 10 min or 4 h (Fig. 2B). To show consistency in protein loading, detection of ERK1 and ERK2 using phosphorylation state-independent pan antibodies was also made (Fig. 2C). Electrophoretic mobility shifts for both ERK1 and ERK2 could be observed in those treatment groups where a marked change in the phosphorylation status of these proteins had occurred. It has previously been shown using the same model system as employed in this study that bFGF (10 ng/ml) induces a sustained phosphorylation of both ERK1 and ERK2 (18). However, in contrast to that evoked by somatostatin, the time profile for the growth factor-stimulated phosphorylation was biphasic, producing peaks at 10 min and again at 4 h postdenudation. In this study, ERK phosphorylation induced by bFGF (10 ng/ml) at 10 min and 4 h was unaffected by transfection with pUSEamp (Fig. 2D). However, transient expression of N17Ras inhibited the growth factor-induced phosphorylation at both time points examined (Fig. 2D). Neither transfection with pUSEamp or pUSEamp(N17ras) showed any effect on basal levels of ERK phosphorylation observed at 10 min or 4 h (Fig. 2D), and the level of ERK protein expression remained unchanged across treatment groups (data not shown).
Effect of Protein Kinase C Inhibition on Cell Proliferation and ERK Phosphorylation-The protein kinase C inhibitor, Ro 31-8220 (50 nM), had no significant effect on the number of CHO K1 cells (expressing the human recombinant sst 4 receptor) maintained in the absence of serum for 24 h after partial denudation (Fig. 3A). However, Ro 31-8220 abolished the increase in cell number induced by either somatostatin (100 nM) or bFGF (10 ng/ml) (Fig. 3A). Down-regulation of protein kinase C following pretreatment of cells with the phorbol ester phorbol 12-myristate 13-acetate (100 ng/ml for 24 h) also reduced the somatostatin-induced increase in cell counts from 244 Ϯ 6 ϫ 10 3 to 172 Ϯ 11 ϫ 10 3 . Incubation with Ro 31-8220 abolished the induced phosphorylation of ERK1 and ERK2 observed following somatostatin treatment for 4 h (Fig. 3B). However, there was no detectable change in the level of somatostatin-induced immunoreactivity detected with the anti-phosphospecific antibody obtained 10 min after partial denudation (Fig. 3B). Similarly, bFGF-induced ERK phosphorylation at 10 min was unaffected by Ro 31-8220, whereas that obtained at 4 h was markedly attenuated (Fig. 3C). Ro 31-8220 had no apparent effect on the basal level of ERK phosphorylation observed at either time point (Fig. 3, B and C) or on the expression levels of the kinases (data not shown).
Effect of Transient Transfection with Transducin on Cell Proliferation and ERK Phosphorylation-Neither mock transfection with pCDNA3 nor pCDNA3 incorporating transducin cDNA had any significant effect on basal proliferation of CHO K1 cells recombinantly expressing the sst 4 receptor, determined 24 h following partial denudation in the absence of exogenous growth factors (data not shown). The increased cell number induced by the submaximal concentration of bFGF (10 ng/ml) was also unaffected following transfection with either plasmid (Fig. 4A). However, the proliferation induced by somatostatin (100 nM) was significantly reduced following transient expression of transducin, whereas that following mock transfection was unaffected (Fig. 4A). The inset in Fig. 4A shows a representative immunoblot of protein extract from cells transfected with either the empty vector or that incorporating transducin cDNA, determined immediately before denudation. For the same number of cells as indicated by the consistency in ␤ actin levels, those transfected with transducin cDNA showed marked immunoreactivity as detected by an anti-transducin antibody compared with those transfected with the empty vector.
There was no detectable change in the basal level or somatostatin-induced phosphorylation of ERK1 or ERK2 in samples allowed to regenerate for either 10 min or 4 h after mock transfection (data not shown). However, overexpression of transducin reduced the somatostatin-induced phosphorylation of ERK1 and ERK2 observed at both 10 min and 4 h after denudation compared with mock-transfected cells (Fig. 4B). Transducin overexpression had no apparent effect on basal phosphorylation levels at either time point examined (Fig. 4B).

FIG. 1. Changes in the phosphorylation status of ERK1 and ERK2 in CHO K1 cells expressing human recombinant sst 4 receptors induced by somatostatin at the onset of proliferative events. Analysis at 10 min (panels A and B) and 4 h (panel C)
following partial denudation of confluent monolayers was determined by Western detection using antibodies specific to ERK1 and ERK2 (panel A) or that recognizing the phosphorylated, active forms (panels B and C). The immunoreactivity detected with antibodies for both ERK1 (1:2,000 dilution) and ERK2 (1:4,000) had the correct electrophoretic mobility on 10% polyacrylamide gels, as compared with molecular weight standards, and panel A shows that expression of the kinases remained unaffected by the various treatments or between the time points examined. Whole cell protein extracts were prepared from partially denuded monolayers incubated in the presence of incomplete media (CON) or somatostatin (100 nM; SRIF) after pertussis toxin pretreatment ( The treatment group labeled with # is significantly different from that incubated in the presence of bFGF without transfection (p Ͻ 0.01). The inset in panel A shows an immunoblot of protein from cell samples extracted immediately before partial denudation that had been transfected 48 h previously with either pUSEamp (Mock) or that incorporating dominant negative Ha-ras (-Ras). After separation by 15% polyacrylamide gel electrophoresis and transfer onto nitrocellulose, detection was made with an anti-␤ actin antibody to demonstrate consistency of protein loading as well as with an antibody to Ras. The effect of transient transfection with N17Ras (-Ras) on the somatostatin-induced (100 nM; SRIF) phosphorylation of ERK1 and ERK2, as detected by the phosphospecific antibody using Western analysis, is shown in panel B. Samples from cells transfected with the empty plasmid 48 h characterized model to determine the re-population of denuded areas in an otherwise confluent monolayer, we have previously demonstrated that the somatostatin-induced proliferative response of this receptor type depends on the sustained activation of ERK1 and ERK2 and independent of a robust transient phase (18). One of the advantages of this model is that cells are synchronized in G 0 or early G 1 of the cell cycle at the onset of the investigative period, and to substantiate that effects are on a proliferative rather than a motogenic process, the total number of cells forming the regenerating monolayer after 24 h in the presence of test agents was determined in the current study. The aim of this investigation was to attempt to identify the transduction effectors involved in mediating the sustained activation of ERK1 and ERK2 by somatostatin sst 4 receptors and, hence, those responsible for inducing the proliferative response.
The protein-tyrosine kinase inhibitors, genistein and lavendustin A, had no effect on basal cell numbers at concentrations that abolished the increased proliferation induced by bFGF. The signaling cascades activated by bFGF receptors would thus seem dependent on this type of phosphate transfer process for transduction of the proliferative function and is compatible with the well characterized mechanism through which this family of receptors mediate their mitogenic effects (29). It would also appear that transmission of the growth-promoting activity of sst 4 receptors is similarly dependent on a proteintyrosine kinase activity. However, in contrast to the growth factor receptor, which contains an intrinsic tyrosine kinase domain within the COOH terminus of each subunit forming the active dimer, the site of intervention of these kinase blockers in the transduction process for sst 4 receptor-mediated growth must be localized to secondary effectors. The lack of effect of genistein on the somatostatin-induced ERK phosphorylation determined at time points representative of the acute and sustained phases of MAP kinase activation suggests that either a parallel cooperative pathway utilizing a tyrosine kinase is essential for growth or, alternatively, that the kinase lies exclusively downstream from MAP kinase. Evidence for a pertussis toxin-insensitive pathway mediating tyrosine phosphorylation of the transcription factor STAT3 has recently been provided for the sst 4 receptor (18), and it may be that the genistein-sensitive effector is situated within this particular cascade (see Fig. 5). Interestingly, it has also been shown in this previous study that only after the additional phosphorylation on serine residues of this same transcription factor as a consequence of prolonged MAP kinase activation, could a proliferative response be induced by somatostatin.
The growth-promoting effect of somatostatin is additionally dependent on a pertussis toxin-sensitive pathway that distinguishes its proliferative mechanism from that of bFGF. Other G i protein-coupled receptors have been shown to mediate rapid tyrosine phosphorylation of several proteins that participate in mitogenic signal transduction such as the adapter protein Shc (30), which is a major substrate for Src kinase. The mechanism whereby these receptors stimulate tyrosine phosphorylation is poorly understood, although activation of the Src-family kinases by several G protein-coupled receptors has been reported (31). In addition, activation of Src seems to account for the G i -mediated tyrosine phosphorylation events that direct recruitment of the Shc and Grb2 adaptor proteins to the membrane (32), thus providing a route into the Ras-ERK cascade (Fig. 5).
To determine whether activation of the MAP kinase cascade was a prerequisite for processing the growth effects induced by somatostatin, the selective MEK1 inhibitor, PD 98059 (33), was used in the proliferation model. It is well documented that the dual-specific kinase MEK stimulates ERK by phosphorylation on threonine (Thr-183) and tyrosine (Tyr-185) residues, which following subsequent translocation into the nucleus, activates transcription factors, resulting in enhanced cell growth (34). In this study, the proliferative effect of somatostatin was abolished by PD 98059, confirming that the MAP kinase cascade is critical for the growth-promoting effect of somatostatin by the sst 4 receptor. In contrast, bFGF-stimulated proliferation in the The vertical bars represent the S.E., and those groups labeled * are significantly different from basal (p Ͻ 0.001). The treatment group labeled # is significantly different from that incubated in the presence of somatostatin without transfection (p Ͻ 0.01). The inset in panel A shows an immunoblot of protein from cell samples extracted immediately before partial denudation that had been transfected 48 h previously with either pCDNA3 (Mock) or that incorporating transducin cDNA (Trans). Western detection was made with an anti-␤ actin antibody to demonstrate consistency of protein loading as well as by the anti-transducin antibody (Trans). The effect of transient expression of transducin (Trans) on the phosphorylation of ERK1 and ERK2 induced by somatostatin (100 nM; SRIF) as well as that obtained after incubation with incomplete media (CON), as detected by the phosphospecific antibody using Western analysis, is shown in panel B. Samples from cells transfected with the empty plasmid 48 h before partial denudation are labeled Mock, and a comparison is shown of the phosphorylation status of ERK1 and ERK2 determined at both 10 min and 4 h post-denudation. The Western blot shown is a representative from at least three separate experiments, and the panel has been taken from a single immunoblot. same host cell was only partially inhibited on co-application with PD 98059, which is consistent with the ability of this receptor type to recruit a multitude of secondary effectors and initiate a number of distinct, yet parallel signaling pathways. An involvement of MEK1 in the sst 4 receptor-mediated proliferative response was further supported by the demonstration of increased phosphorylation of ERK1 and ERK2 following somatostatin treatment. However, although both the acute and sustained phases of MAP kinase activation were abolished by pertussis toxin, a differential effect on the temporally distinct activities was observed following MEK1 inhibition. Abolition of the sustained phase with PD 98059, although having no observable effect on the transient activity, is supportive evidence for the requirement of the prolonged activation of MAP kinase in mediating cell growth. The lack of effect of the MEK1 inhibitor on the marked transient phosphorylation of ERK is possibly due to the ineffectiveness of the concentration of PD 98059 administered with somatostatin and is in keeping with other reports showing that high levels of MAP kinase activity are PD 98059-insensitive (33).
Several G i -coupled receptors have been shown to mediate MAP kinase activation through the ␤␥-component of the G protein possibly through the activation of the Src-family of tyrosine kinases (35). The proliferative response induced by somatostatin was inhibited following overexpression of the ␤␥sequestering protein, transducin, in contrast to the lack of effect on bFGF-induced growth. However, the Src-family inhibitor, PP1 (36), failed to reduce the proliferative effect induced by either mitogen in this cell line. The bFGF results were somewhat unexpected since it is well known that Src is a co-transducer of mitogenic signals arising from a number of tyrosine kinase growth factor receptors, such as platelet-derived growth factor or epidermal growth factor receptors (37). However, the association of Src with bFGF receptors appears to be cell-specific (38), and in this respect, bFGF-induced proliferation in vascular smooth muscle cells has been shown to be partially inhibited by the Src inhibitor PP1, as determined by the same model system employed in this current study. 2 PP1 in CHO K1 cells transfected with the somatostatin sst 4 receptor was shown to reduce platelet-derived growth factor-stimulated growth.
The inability of PP1 to inhibit somatostatin-induced proliferation suggests Src is not involved in this response mechanism, and therefore the process through which MEK is activated appears to be very different to that employed by other G i protein-coupled receptors (35). However, further examination of the somatostatin-induced ERK phosphorylation showed that the transient phase was sensitive to the Src inhibitor, in contrast to the PP1-independent prolonged phosphorylation. The attenuation of the transient phosphorylation of ERK by the Src inhibitor without any resultant effect on the proliferative response again suggests that it is the sustained activation of MAP kinase that is critical for proliferation. In addition, the sensitivity of the transient and not the prolonged phase of ERK phosphorylation to Src-inhibition provides evidence that different transduction events are involved in mediating the temporally distinct MAP kinase activities. The involvement of Src in mediating the acute phase of ERK phosphorylation and the insensitivity of this component to genistein seem incompatible. However, genistein is a nonselective tyrosine kinase inhibitor, and it may be that in this system, Src activity is unaffected by the concentration of genistein used.
The mechanisms by which G i and G q -coupled receptors typically activate MAP kinase are through Ras-dependent or protein kinase C-dependent pathways, respectively. However, a 2 L. A. Sellers, unpublished data.
FIG. 5. Proposed mechanism that transduces the proliferative response of somatostatin in CHO K1 cells recombinantly expressing the human sst 4 receptor. G i -coupled receptors typically converge at or before Src to mediate Ras-dependent MAP kinase activation. In contrast, receptors coupled to pertussis toxin-insensitive G q/11 activate protein kinase C, which in turn can mediate a Ras-independent MAP kinase activation. The proliferative response mediated via the sst 4 receptor requires G␤␥ release, consistent with mitogenic pathways of other G i/o -coupled receptors but appears independent of the transient activation of ERK, which is Src-and Ras-dependent. By contrast, the sustained activation of ERK (known to be associated with nuclear translocation) and the proliferative response of somatostatin are both dependent on protein kinase C (PKC) activity. Activation of protein kinase C by sst 4 receptors has not been determined in this study but could be via ␤␥-mediated phospholipase C-␤ (PLC␤) stimulation (21) or the ␤␥-dependent phosphoinositide 3-OH kinase (PI-3) isoform. Sustained MAP kinase activation is critical for the somatostatin-induced proliferative response, but an additional, as yet uncharacterized tyrosine kinase activity is also required in a cooperative pathway. This could possibly involve the tyrosine phosphorylation of STAT3, shown to be mediated through a pertussis toxin-insensitive mechanism following sst 4 receptor activation (18). The site of action of the inhibitors used in this study are shown by the shaded boxes. SOS, son of seven less. GRB, adaptor protein Grb2. few exceptions to this rule have been recently reported for G q -coupled receptors in that MAP kinase can be activated through a pertussis toxin-insensitive but protein kinase Cindependent pathway (39). In this study, we have demonstrated that the transient phosphorylation of ERK by sst 4 receptors is sensitive to both transducin and dominant negative Ras (N17) but unaffected following protein kinase C inhibition. These results are also in accord with the ␤␥-mediated Src stimulation utilized by other G i -coupled receptors to activate MAP kinase through a Ras-dependent mechanism (Fig. 5). In addition, these data are also consistent with the acute phase of MAP kinase activity not being involved in mediating a growth response, as expression of N17Ras had no effect on somatostatin-induced proliferation or the prolonged activation of MAP kinase.
Since it appears that the sustained activation of MAP kinase, required for the somatostatin-induced proliferative effect, utilizes a distinct but convergent pathway to that mediating the transient Ras-dependent ERK phosphorylation, we examined the involvement of protein kinase C, which can activate the Ras-ERK cascade at the point of Raf (40) (Fig. 5). Both the proliferation and sustained phosphorylation of ERK1 and ERK2 induced by somatostatin were abolished following protein kinase C blockade. This suggests that protein kinase C involvement is critical for the growth response and is placed upstream to ERK activation, consistent with other reports investigating MAP kinase stimulation through G q -coupled receptors (39). Receptor tyrosine kinase-mediated activation of Raf-1 is coupled to Ras, and bFGF-induced proliferation of CHO K1 cells used in this study was N17Ras-sensitive together with both the acute and sustained phases of ERK phosphorylation. By contrast, protein kinase C-mediated activation of Raf-1 is thought to be Ras-independent and is in keeping with the lack of effect on the prolonged MAP kinase phosphorylation and the induced proliferative response observed in this study following application of somatostatin to cells overexpressing N17Ras. Activation of Raf-1 by protein kinase C has been shown to be insensitive to dominant negative Ras (41), indicating that protein kinase C activates Raf by a mechanism distinct from that initiated by activation of receptor tyrosine kinases. Although in this study the acute phase of ERK phosphorylation induced by bFGF was unaffected by a protein kinase C inhibitor, the sustained component was reduced, and the proliferative effect was abolished. It thus appears that it is the sustained ERK activity that is also critical for bFGF to induce cell proliferation. However, in contrast to that following sst 4 receptor activation, the sustained phase of ERK phosphorylation induced by bFGF appears to involve both Ras-and protein kinase C-dependent mechanisms, and both seem to be required for the proliferative effect.
The protein kinase C family has at least 11 members, 6 of which (␦, ⑀, ␥, , , ) have been shown to be expressed in the CHO K1 cells used in this study. 2 Both the typical and atypical protein kinase C isozymes are activated by diacylglycerol, which is produced by the metabolism of phosphatidylinositols. Although sst 4 receptors have been shown to mediate inositol 1,4,5,-trisphosphate production (21), it has not been determined here if through ␤␥ release, the subsequent stimulation of phospholipase C-␤ is mandatory for the sustained ERK phosphorylation induced by somatostatin. Activation of MAP kinase after sst 4 receptor stimulation has been shown to be dependent on phosphoinositide 3-OH kinase (42). Signaling targets of the lipid products of this kinase activity include the calcium-independent protein kinase C isoforms (43), raising the possibility that ERK activation through protein kinase C in this study could be via a calcium-independent pathway. Involvement of phosphoinositide 3-OH kinase in the prolonged activation of ERK would also be in keeping with the G␤␥ dependence of both this and the proliferative events observed after somatostatin application, as a phosphoinositide 3-OH kinase responsive to ␤␥ subunits has recently been cloned (44).
Very little evidence is currently available as to the identity of the molecular determinants responsible for the sustained activation of MAP kinase. Recently it has been shown that in PC12 cells, Ras must be activated for the initial phase of ERK activation following stimulation of the nerve growth factor receptor (TrkA), but the sustained phase involves another small GTPase, Rap1 (45). In this study we have demonstrated that sst 4 receptors can stimulate cellular proliferation through transduction mechanisms with a critical requirement for a sustained, protein kinase C-dependent activation of MAP kinase. Stimulation of MAP kinase has been shown to regulate a diverse range of functional responses, sometimes with opposing effects. For example, although we have shown a critical requirement for ERK activity in the sst 4 receptor-mediated proliferative effect, this activity also appears to be necessary for the growth inhibitory response of sst 1 receptor types (24). These apparent conflicting functional processes following activation of the MAP kinase cascade will be better explained once the kinetics of ERK activation as well as the strength of the stimulus for a given receptor type have been fully evaluated and elements of the transduction machinery required for the temporally distinct activities have been identified.