- Email: [email protected]
Effects of somatostatin, somatostatin analogs, and endothelial cell somatostatin gene transfer on smooth muscle cell proliferation in vitro
Effects of somatostatin, somatostatin analogs, and endothelial cell somatostatin gene transfer on smooth muscle cell proliferation in vitro Rajabrata Sarkar, MD, PhD, Chris J. Dickinson, MD, and James C. Stanley, MD, Ann Arbor, Mich Objective: Somatostatin analogs inhibit neointimal hyperplasia and smooth muscle cell (SMC) proliferation in vivo. The gene transfer of somatostatin to endothelial cells (ECs) represents a potential means of local delivery of somatostatin to areas of arterial injury. This study tested the hypothesis that the retroviral gene transfer of somatostatin to ECs would inhibit SMC proliferation in vitro and evaluated the effects of somatostatin analogs on DNA synthesis and the growth of SMCs. Methods: Media transfer and coculture were used to determine the effects of somatostatin-producing ECs on SMC proliferation in vitro. The effects of a variety of somatostatin isoforms and analogs on the proliferation of SMCs, mitogenesis of serum-restimulated quiescent SMCs, and arterial explants were measured. Results: Despite the production of biologically relevant concentrations of somatostatin by ECs, no inhibition of SMC proliferation was noted. Somatostatin analogs inhibited DNA synthesis in arterial explants but did not inhibit either DNA synthesis or growth of cultured SMCs, which showed a likely effect of somatostatin on the initial transition in SMC phenotype. Conclusion: Somatostatin exerts inhibitory effects on SMC proliferation only during the early transition to a proliferative phenotype. There are significant differences between this in vivo transition and the standard serum-restimulated model of cultured SMCs. These differences may account for the failure of somatostatin to inhibit SMC proliferation in the standard in vitro models. (J Vasc Surg 1999;29:685-93.)
Smooth muscle cell (SMC) proliferation plays an important role in the pathogenesis of a number of vascular disorders. In atherosclerosis, SMC proliferation is present at low but persistent levels and is associated with plaque neovascularization.1 The proFrom the Conrad Jobst Vascular Research Laboratories, Section of Vascular Surgery, Department of Surgery (Drs Sarkar and Stanley) and Division of Gastroenterology, Department of Pediatrics, University of Michigan Medical Center. Supported by a National Research Service Award (F-32Hl08677) from the National Heart, Lung and Blood Institute, National Institutes of Health (R.S.), and NIH grant HL02816 (J.C.S.). Presented at the Poster Session of the 1997 Joint Annual Meeting of the Society for Vascular Society/International Society for Cardiovascular Surgery (North American chapter), June 1–4, 1997, Boston, Mass. Reprint requests: Dr James C. Stanley, University Hospital, 2210 THC, 1500 East Medical Center Dr, Ann Arbor, MI 48109. Copyright © 1999 by the Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/99/$8.00 + 0 24/1/95743
liferation of SMCs is more pronounced in experimental models of arterial balloon injury2 and vein grafting,3 and the pharmacologic inhibition of SMC proliferation is associated with a decrease in neointimal hyperplasia.4 The variety of pharmocologic agents that can inhibit SMC proliferation and neointimal hyperplasia includes the peptide analogs of the neuroendocrine hormone somatostatin. Somatostatin analogs, such as octreotide and angiopeptin, inhibit neointimal hyperplasia and SMC proliferation in experimental balloon angioplasty,5,6 vein graft models,7 and transplant arteriosclerosis.8 Somatostatin is an endogenous peptide and thus could be produced with genetically modified vascular cells. The introduction of either previously transduced cells onto sites of arterial injury or direct gene transfer to the arterial wall then could be used as a potential therapy for neointimal hyperplasia. Retroviral gene transfer to both endothelial cells9 (ECs) and SMCs10 in culture followed by the seed685
686 Sarkar, Dickinson, and Stanley
JOURNAL OF VASCULAR SURGERY April 1999
noted that correlated with the in vitro and in vivo effects of somatostatin.
Fig 1. Somatostatin production by β-galactosidase gene–transfected endothelial cells (BAG EC) and human somatostatin-transfected endothelial cells (HSS EC) in various serum concentrations. Transfected and selected endothelial cells were grown to confluence, and media were changed to indicated serum concentrations. After 24 hours, media were removed and assayed for somatostatin peptide. *P < .05 versus β-galactosidase gene media (n = 2 to n = 3). #P < .05 versus β-galactosidase gene media and versus human somatostatin 2.5% serum.
ing of these cells onto sites of arterial injury has resulted in the long-term expression of marker proteins, such as galactosidase and human adenosine deaminase in vivo. The seeding of retrovirally transduced ECs onto prosthetic grafts also results in the long-term coverage of the graft surface and the production of the transgene of interest.11 The production of biologically active somatostatin after gene transfer necessitates the presence of multiple separate prohormone processing endopeptidase enzymes, and recent studies from our laboratory have shown that the retroviral gene transfer of the human somatostatin complementary DNA to cultured ECs results in the secretion of significant concentrations of properly processed somatostatin peptide.12 The purpose of this study was to determine whether the retroviral gene transfer of somatostatin to ECs would inhibit the proliferation of cultured SMCs. Additional studies were performed to determine the effects of various somatostatin analogs on the proliferation and DNA synthesis of cultured SMCs. Three different in vitro models of SMC proliferation were used, with significant differences
MATERIALS AND METHODS Cell derivation and culture. Canine venous ECs were derived from the external jugular veins of adult mongrel dogs as previously described13 and were maintained in EC growth media (M199 media, Gibco, Long Island, NY), with 5% bovine calf serum (Hyclone, Logan, Utah), EC growth supplement 10 µg/mL (Collaborative Biochemical, Bedford, Mass), heparin 15 U/mL, penicillin-streptomycin 1 mmol/L L-glutamine, in a 5% CO2 humidified 37°C incubator on gelatin-coated plates. The cultures were fed every 2 to 3 days and were split when confluent. EC type was confirmed by means of cell morphology, lack of staining with an antibody to α-actin, development of a confluent monolayer, and uptake of fluorescent acetylated-low density lipoprotein ligand. Rat SMCs and canine SMCs were derived with enzymatic dissociation from minced segments of rat thoracic aorta or canine femoral artery. The segments were incubated in Dulbecco’s serum-free media that contained 0.1% (wt/vol) collagenase (Worthington Chemicals, Philadelphia, Pa) and 0.01% (wt/vol) elastase (Sigma Chemical, St Louis, Mo) for 2 to 3 hours at 37°C with constant agitation. Detached cells in the supernatant were pelleted and plated into media that consisted of Dulbecco’s minimal essential media (DMEM) with 10% (vol/vol) fetal bovine (rat cells) or bovine calf (canine cells) serum, 1 mmol/L glutamine, and antibiotics. The SMC identity was confirmed by means of morphology and immunoflourescent staining with an α-actin antibody (HF-35, Sigma Chemical). For growth experiments, SMCs were used in passages 2 to 10. Retroviral transfection. Retroviruses with selectable neomycin resistance genes that contained either the human somatostatin (HSS) complementary DNA or a control β-galactosidase gene (BAG) were used to create pure populations of ECs that expressed either β-galactosidase (BAG EC) or human somatostatin (HSS EC). Previous studies have shown that canine BAG ECs have uniform βgalactosidase expression and that HSS ECs produce the somatostatin-14 isoform of somatostatin at concentrations of 400 pmol/L in growth media.12 Rat ECs display nonuniform β-galactosidase expression and do not produce significant concentrations of somatostatin after retroviral transfection.14 For transfection, the ECs in passages 2 or 3 were split 1:5 the day before transfection to maximize the number of replicating cells capable of undergoing
JOURNAL OF VASCULAR SURGERY Volume 29, Number 4
Sarkar, Dickinson, and Stanley
687
Fig 2. Effect of transfer of media conditioned by β-galactosidase gene–transfected endothelial cells (BAG EC) and human somatostatin-transfected endothelial cells (HSS EC) on growth of smooth muscle cells (SMCs). Media containing indicated concentrations of serum were conditioned with canine BAG ECs and HSS ECs every 24 hours and then were transferred to rat or canine SMCs daily. SMC cell number was determined after 72 hours (rat, n = 3; canine, n = 4). *P < .05 versus BAG media.
retroviral gene transfer. The ECs were exposed to the HSS viral supernatant with the cationic agent Polybrene (8 µg/mL) for 8 hours and then were refed with appropriate fresh media. After a 48-hour period to allow the reverse transcription of the viral RNA genome, the incorporation of the DNA provirus into a host chromosome, and the expression of the transgenes, the cells were split 1:8 into appropriate complete media supplemented with Geneticin G418 (Sigma Chemical, St Louis, Mo) at 1 mg/mL. Nontransduced ECs from the same passage were split simultaneously into G418-containing media to serve as controls. All the cells were maintained in G418-containing media for approximately 14 days, or 3 days after all the control nontransduced cells had died. Control ECs for all the experiments were transduced with the BAG marker retrovirus with identical procedures as described previously except with the BAG retroviral supernatant used for transduction. Analysis of gene expression. Somatostatin production was assayed with the measurement its level in media from confluent monolayers of BAG-transduced and HSS-transduced ECs that had been collected after 24-hour incubations and stored at –70°C for subsequent radioimmunoassay. The radioimmunoassay used synthetic somatostatin-14 as a standard and an antibody (antibody 1001) that binds to the carboxyl terminus of somatostatin.15 Media transfer and coculture studies. To
determine the effects of somatostatin produced by HSS-transduced ECs on SMC growth, media conditioned with HSS ECs was transferred to growing canine SMCs. Canine or rat SMCs were plated at a density of 5000 cells/cm2 in growth media. The following day, EC media, which had been placed on confluent flasks of BAG ECs or HSS ECs, were removed and transferred to the SMC wells. The EC monolayers were refed with fresh EC media, which then were transferred again to the SMCs after 24 hours of conditioning with the ECs. After a total incubation period of 72 hours in conditioned media, the SMCs were counted as described subsequently. For coculture studies, the Transwell insert system (a polycarbonate filter with a 0.22-micron pore size, Corning Costar Inc, Corning, NY) was used with ECs grown in the Transwell insert and SMCs grown on the underlying well. The ECs were plated onto the insert, which was placed in separate wells that contained EC media. These media were changed until the ECs were confluent. The inserts then were removed and placed into separate wells in which canine SMCs had been plated 24 hours earlier at a density of 5000 cells/cm2. Both the cell types then were maintained in coculture (in EC media) for 96 hours with one change of the media at 48 hours. In selected groups, the somatostatin analog octreotide (10-10 mol/L, Sigma Chemical) was added daily. The integrity of the EC monolayer was verified with daily phase contrast microscopy with-
688 Sarkar, Dickinson, and Stanley
JOURNAL OF VASCULAR SURGERY April 1999
Fig 3. Canine smooth muscle cell (SMC) growth in presence of various endothelial cell (EC) types or somatostatin analog. Canine SMCs were grown in coculture for 96 hours with indicated types of canine ECs. Octreotide (10–10 mol/L) was added in indicated groups. No significant differences were noted among four groups that were cocultured with ECs (SMC + EC, SMC + β-galactosidase gene (BAG) EC, SMC + human somatostatin (HSS) EC, SMC + EC + octreotide). All groups, n = 4. *P < .05 versus SMC alone.
out disturbing the coculture system. At the conclusion of the experiment, the inserts were removed and the SMCs were counted as subsequently described. DNA synthesis and cell proliferation assays. DNA synthesis was analyzed in cells that were initially rendered quiescent and then was triggered to reenter the cell cycle with serum restimulation. Canine or rat SMCs were treated with serum-free media (DMEM plus 1% (wt/vol) albumin) for 48 hours, and then media that contained serum were added with experimental agents as indicated. In selected experiments, the somatostatin analogs were added to the serum-free media to pretreat the cells before the mitogenic stimulus. DNA synthesis was measured 24 hours after the readdition of the serum with the incorporation of tritiated thymidine as previously described.16,17 In brief, 0.37 MBq/mL 3Hthymidine (Dupont, Boston, Mass) was added to the media, and, after 2 hours, the amount of radiotracer incorporation into the cellular DNA was determined by means of precipitation with 10% ice-cold trichloroacetic acid and scintillation counting. Cell proliferation was measured by counting cell numbers with a Coulter counter (Coulter Electronics, Hialeah, Fla) after the enzymatic detachment of the cells. The cells were washed twice with phosphate-buffered saline solution and were treated with 0.05% trypsin/ethylenediamine tetraacetic acid until complete cell detachment was confirmed with phase microscopy. The cell mixture then
was added to 20 mL of counting solution, and triplicate counts were taken of 1-mL aspirations of each sample. The percent inhibition of cell proliferation in experimental groups was determined with the following formula: (control count – experimental count)/control count × 100. For DNA synthesis studies in fresh arterial explants, the method of Vargas was used. 18 Segments of canine femoral artery were removed after the deaths of mongrel adult male dogs and were placed in sterile phosphate-buffered saline solution. Under sterile conditions, the adventitia was dissected free and the endothelium was removed with mechanical denudation. With a ruler, 3-mm square segments then were cut by means of surgical magnification. These segments were placed in 1 mL of DMEM media (without serum) that contained antibiotics, experimental agents, and 0.37 MBq/mL 3H-thymidine. After a 24-hour incubation period, each segment was incubated with cold thymidine (1 mmol/L) for 1 hour, washed with phosphatebuffered saline solution repeatedly, and dissolved in 1 mol/L of sodium hydroxide at 37°C overnight. The resulting solution was neutralized with an equivalent volume of 1 mol/L of hydrochloric acid before the addition of scintillation cocktail and scintillation counting. Statistical analysis. An unpaired t test was used for comparisons between two groups. Analysis of variance with correction for multiple comparisons was used for comparisons between several groups.
JOURNAL OF VASCULAR SURGERY Volume 29, Number 4
Sarkar, Dickinson, and Stanley
689
All the data are presented as the mean ± the standard error of the mean. Differences were considered significant at a P value of less than .05. RESULTS The ability of somatostatin-transduced ECs to produce somatostatin peptide first was assessed in the presence of varying concentrations of serum. We previously have documented that HSS ECs produce approximately 300 to 400 pmol/L of somatostatin concentration in the surrounding media in full growth media (ie, 10% serum).12 The growth of cultured SMCs is dependent on growth factors present in serum, unlike ECs, which necessitate exogenous acidic fibroblast growth factor in the form of EC growth supplement. A range of serum concentrations was used to ensure that a mild antimitogenic effect of the somatostatin peptide would not be masked by an overwhelming mitogenic stimuli from serum. Somatostatin production increased with the concentration of serum used, although the concentrations were in the range of 300 to 500 pmol/L over the range of serum concentrations from 2.5% to 10% (Fig 1). These concentrations of somatostatin have been reported to inhibit the in vitro proliferation of prostate cancer cells,19 breast carcinoma,20 and T lymphocytes.21 There was no significant inhibition of either canine or rat SMC proliferation noted with the transfer of conditioned somatostatin media (Fig 2). As expected, the proliferation of SMCs was dependent on the serum concentration in the transferred media, with greater cell growth in 10% serum. There are several possible reasons for this failure of growth inhibition, including the possibility that the transfer of conditioned media every 24 hours may allow significant degradation of somatostatin peptide in the EC media before the transfer to the SMCs. A loss of biologic activity as a result of peptide degradation is of particular concern with somatostatin because it has a half-life of less than 2 hours in media.22 To minimize the degradation of somatostatin between the time of secretion from the ECs to the contact with the SMCs, we then grew SMCs continuously in the presence of somatostatinproducing ECs (Fig 3). The presence of ECs in the coculture system, either nontransduced or transduced with either the HSS or the BAG control vector, increased SMC proliferation regardless of the EC type (Fig 3), as has been reported by other investigators.23 Octreotide (10-10 mol/L) did not alter the growth of SMCs, with or without the presence of ECs (Fig 3). To
Fig 4. Effect of somatostatin (SS) analogs on DNA synthesis and cell proliferation of canine and rat smooth muscle cells (SMCs). Nitric oxide donor S-nitroso-Nacetylpenicillamine (SNAP; 0.1 mmol/L) was used as positive control for inhibition of proliferation. DNA synthesis (thymidine incorporation) was measured in serum-restimulated quiescent SMCs. Cell number was determined after 72 hours in presence of indicated concentrations of somatostatin analogs (n = 3 to n = 4). *P < .05 versus control.
ensure that this lack of proliferation was not simply caused by the concentrations of somatostatin or octreotide, the effects of a range of concentrations of different somatostatin isoforms and analogs on
690 Sarkar, Dickinson, and Stanley
JOURNAL OF VASCULAR SURGERY April 1999
Fig 5. Effect of angiopeptin (150 ng/mL, 75 µmol/L) on DNA synthesis in canine arterial explants (clear bars) and 72-hour cell proliferation of subcultured canine smooth muscle cells (SMCs; solid bars; n = 4). *P < .05 versus control.
growth and DNA synthesis of several species of cultured SMCs then were determined. Octreotide did not inhibit the DNA synthesis of canine SMCs over a wide range of concentrations (Fig 4), and somatostatin-14 failed to inhibit the proliferation of both rat and canine SMCs. We used the nitric oxide donor Snitroso-N-acetylpenicillamine (0.1 mmol/L), a known inhibitor of SMC growth,16,17 as a positive control to confirm that our experimental system could detect changes in cell proliferation. Similar negative results were found with various concentrations of octreotide and cell proliferation and with either somatostatin-14 or somatostatin-28 and DNA synthesis in synchronized cells and with cultured human SMCs (all studies, n = 3 to n = 4; P > .05 vs control for concentrations 10–6 to 10–10 mol/L). The pretreatment of quiescent SMCs with somatostatin analogs before restimulation with serum also did not result in the inhibition of subsequent DNA synthesis (n = 4; P > .05 vs control for 10-7 to 10-10 mol/L). Previous studies have shown that the somatostatin analog angiopeptin inhibits DNA synthesis in arterial explants ex vivo.18 To examine the discrepancy between this finding and our uniformly negative findings with somatostatin analogs (Fig 4), we compared the effects of angiopeptin on DNA synthesis in cultured SMCs and fresh arterial explants. Angiopeptin (150 ng/mL, 75 µmol/L) significantly inhibited DNA synthesis in arterial segments within the first 24 hours after explantation but failed to inhibit DNA synthesis in serum-restimulated quiescent cultured SMCs (Fig 5). Experiments that used the transfer of somatostatin-conditioned media produced with ECs to inhibit DNA synthesis in arterial
explants were not feasible because somatostatin analogs only inhibit DNA synthesis in arterial explants under serum-free conditions.18 DISCUSSION The significant findings in this study were: (1) the production of somatostatin after the gene transfer of human somatostatin to ECs does not inhibit the proliferation of cultured SMCs; (2) neither exogenous somatostatin nor its analogs inhibit mitogenesis or proliferation of several species of SMCs; and (3) the somatostatin analog angiopeptin inhibits DNA synthesis in fresh arterial explants but not in cultured SMCs, which thus shows a potential mechanism for the failure of gene transfer of somatostatin to inhibit SMC growth. To our knowledge, this report is the first description of the effects of somatostatin on the growth of cultured SMCs. Our findings in cultured SMCs and arterial explants correlate not only with the molecular mechanism of action of somatostatin but also with the efficacy of somatostatin analogs in decreasing neointimal hyperplasia in vivo. The molecular mechanism by which somatostatin inhibits cell proliferation has been defined in several different cell types. Somatostatin binds to cell surface receptors and activates tyrosine phosphatases, which dephosphorylate the signaling proteins that are the substrates for mitogen-activated tyrosine kinases. 24 Consistent with this early inhibition of mitogenic signal transduction is the finding that the administration of angiopeptin must be within the first hour after arterial injury for the subsequent inhibition of proliferation to occur.25,26
JOURNAL OF VASCULAR SURGERY Volume 29, Number 4
Our media transfer and coculture experiments were performed with the SMCs growing continuously in serum-containing media. The cell cycle events of continuously proliferating SMCs are different from both those of cultured SMCs synchronized with serum starvation27 and of quiescent SMCs in uninjured arteries.28 It is possible that the actions of somatostatin analogs noted in vivo are caused by the inhibition of SMC proliferative events during the initial transition from quiescence to proliferation, and thus these effects would be missed in our experiments with continuously proliferating SMCs (Figs 2, 3). We thus used quiescent synchronized SMCs to which somatostatin analogs were added either with or before the serum restimulation, and we were not able to detect either inhibition of DNA synthesis or cell number (Fig 4). This suggests that the inhibition of neointimal hyperplasia with somatostatin analogs in vivo is either independent of cell proliferation or is acting on elements of SMC proliferation that are not present in either of these commonly used in vitro models. In the media transfer experiment, the rat SMCs that were treated with HSS media showed greater proliferation than did the control cells (Fig 2). This greater proliferation was not seen with the canine cells (Fig 2) nor with the rat SMCs at higher or lower serum concentrations. Although somatostatin has been shown to increase the proliferation of certain cell types,29 the isolated finding with rat SMCs at 5% serum is not reproduced with exogenous somatostatin (Fig 4). Thus, although we do not have an explanation for this result, the effect appears to be an isolated event that is particular to the media transfer assay with rat SMCs. There are other mechanisms by which somatostatin may decrease neointimal hyperplasia besides the direct inhibition of SMC proliferation. Somatostatin inhibits the migration of SMCs30 and thus may inhibit the migration of SMCs from the media into the intima, which is critical to the formation of intimal lesions.31 Somatostatin inhibits the adhesion of inflammatory cells,32 and recent studies33 have shown that a significant fraction of the neointimal cells traditionally thought of as SMCs2,34 are actually inflammatory cells. Thus, the inhibition of inflammatory cell adhesion with somatostatin may inhibit net intimal lesion size independent of direct effects on SMC proliferation. The inhibition of inflammatory cell adhesion also inhibits SMC proliferation via the decreased production of SMC mitogens with inflammatory cells.33 Despite these other potential confounding mechanisms, the finding that angiopeptin inhibits
Sarkar, Dickinson, and Stanley
691
SMC proliferation in rat18 and canine arterial explants (Fig 5) suggests that the direct inhibition of SMC proliferation may be a contributory mechanism to the overall inhibition of neointimal hyperplasia with somatostatin analogs. This study shows that there is a lack of correlation between the in vivo (or fresh arterial explant model) findings and the standard cell cycle model of serum-starved cultured SMCs, which reenter the cell cycle on readdition of serum. Serum-restimulated quiescent SMCs recapitulate many of the cell cycle events that are noted after in vivo arterial injury, including time-dependent expression of immediate response genes, such as fos and myc, induction of polyamine synthesis, and DNA synthesis after 24 to 48 hours.27 This study did not evaluate the effect of EC-derived somatostatin on DNA synthesis in fresh arterial explants because the serum-free assay conditions for explants were not compatible with the ECconditioned media. However, given our previous characterization of EC-derived somatostatin as somatostatin-14,12 experiments with EC-derived somatostatin would add little to the results that were obtained with exogenous somatostatin analogs (Figs 4, 5). The use of the cultured SMC model in understanding in vivo responses to antiproliferative agents is shown best by the case of heparin, which inhibits cell cycle progression in the late G1 phase, with a loss of inhibition of subsequent DNA synthesis when heparin is added more than 12 hours after the serum restimulation of quiescent rat SMCs. 35 Similarly, the delayed administration of heparin after arterial injury causes a loss of inhibition of subsequent proliferation.4 This correlation is not borne out by the somatostatin analogs in our study because the ability of angiopeptin to inhibit DNA synthesis in vivo5 and in fresh arterial explants18 is not reflected in the inhibition of DNA synthesis in serumrestimulated SMCs (Fig 5). Thus, the somatostatin analogs act on an early aspect of the phenotype change of in vivo SMCs, which occurs after arterial explantation18 or balloon injury25 but which is not mimicked by the serum restimulation of cultured SMCs. This discrepancy between in vivo and in vitro findings appears to be specific to the SMC cell type because other types of passaged cells show inhibition of growth in vitro on treatment with somatostatin.19-21 Another possible factor in evaluating the antiproliferative effects of somatostatin may be gender differences in SMC response. However, angiopeptin inhibits proliferation in male animals25 and all the cell types used in this
JOURNAL OF VASCULAR SURGERY April 1999
692 Sarkar, Dickinson, and Stanley
study, as well as the explants in previous studies,18 were from male animals. Attempts to define the molecular mechanisms that distinguish the in vivo and in vitro activities of somatostatin analogs are limited by the lack of effect in cultured SMCs. The critical proliferative signals in SMCs that somatostatin analogs inhibit occur almost immediately after arterial injury because angiopeptin pretreatment inhibits the expression of the proto-oncogenes fos and jun 30 minutes after injury.25 These in vivo signaling events are presumably also inhibited ex vivo in arterial explants but not in serum-restimulated quiescent cultured SMCs. Unfortunately, the model of serum-restimulated cultured cells is most suited to the analysis of molecular signaling events,19-21 and it is this particular model in which somatostatin analogs do not display activity on SMCs. Thus, a limitation of this study, and of future efforts to define the mechanisms of action of somatostatin on SMCs, is the lack of a simple in vitro model in which these mechanisms can be characterized, as has been done with other cell types.19-21 These findings have three consequences for the development of antiproliferative SMC therapy. The study shows for the first time, to the best of our knowledge, that there are undefined but critical cell cycle events that occur after arterial injury in vivo and that are not present in the widely used in vitro model of serum-restimulated quiescent SMCs. Secondly, this finding implies that the use of the serum-restimulated (Fig 5) or the continuously proliferating cultured SMC model (Fig 4) to screen potential antiproliferative compounds may miss those, such as angiopeptin, that do not inhibit DNA synthesis in culture but do have significant in vivo effects. Finally, our findings suggest that the reseeding of arterial injury sites with somatostatin-producing ECs may not inhibit subsequent neointimal hyperplasia. Both the lack of inhibition of SMC proliferation after the phenotype change in culture and the loss of effect of angiopeptin when delayed 1 hour after injury25,26 suggest that critical somatostatin-sensitive events in SMC proliferation will probably occur well before significant somatostatin production with reseeded ECs occurs. We thank the Henri Beaufour Institute USA, Inc, who supplied the angiopeptin, Susan Finniss, who assisted with the somatostatin RIA assay, Abraham Jacobs, MD, and Eric Meinberg, MD, who assisted with cell culture, Thomas Wakefield, MD, who supplied the canine femoral arteries, and Peter Ramwell, PhD, for helpful discussions.
REFERENCES 1. Rekhter MD, Gordon D. Does platelet-derived growth factor-A chain stimulate proliferation of arterial mesenchymal cells in human atherosclerotic plaques? Circ Res 1994;75:410-7. 2. Clowes AW, Reidy MA, Clowes MM. Kinetics of cellular proliferation after arterial injury. I. Smooth muscle growth in the absence of endothelium. Lab Invest 1983;49:327-33. 3. Zwolak RM, Adams MC, Clowes AW. Kinetics of vein graft hyperplasia: association with tangential stress. J Vasc Surg 1987;5:126-36. 4. Majesky MW, Schwartz SM, Clowes MM, Clowes AW. Heparin regulates smooth muscle S phase entry in the injured rat carotid artery. Circ Res 1987;61:296-300. 5. Lundergan C, Foegh ML, Vargas R, et al. Inhibition of myointimal proliferation of the rat carotid artery by the peptides, angiopeptin and BIM 23034. Atherosclerosis 1989;80:49-55. 6. Yumi K, Fagin JA, Yamashita M, et al. Direct effects of somatostatin analog octreotide on insulin-like growth factorI in the arterial wall. Lab Invest 1997;76:329-38. 7. Calcagno D, Conte JV, Howell MH, Foegh ML. Peptide inhibition of neointimal hyperplasia in vein grafts. J Vasc Surg 1991;13:475-9. 8. Foegh ML. Angiopeptin: a treatment for accelerated myointimal hyperplasia? J Heart Lung Transplant 1992;11:S28-31. 9. Nabel EG, Plautz G, Boyce FM, Stanley JC, Nabel GJ. Recombinant gene expression in vivo within endothelial cells of the arterial wall. Science 1989;244:1342-4. 10. Lynch CM, Clowes MM, Osborne WR, Clowes AW, Miller AD. Long-term expression of human adenosine deaminase in vascular smooth muscle cells of rats: a model for gene therapy. Proc Nat Acad Sci USA 1992;89:1138-42. 11. Baer RP, Whitehill TE, Sarkar R, Sarkar M, Messina LM, Komorowski TA, et al. Retroviral-mediated transduction of endothelial cells with the lac Z gene impairs cellular proliferation in vitro and graft endothelialization in vivo. J Vasc Surg 1996;24:892-9. 12. Sarkar R, Finniss S, Dickinson CJ, Stanley JC. Somatostatin gene transfer and expression in endothelial cells. J Vasc Surg 1998;27:955-62. 13. Graham LM, Burkel WE, Ford JW, Vinter DW, Kahn RH, Stanley JC. Immediate seeding of enzymatically derived autologous canine endothelium in dacron velour vascular grafts. Arch Surg 1980;115:1289-94. 14. Watts CKW, Brady A, Sarcevic B, deFazio A, Musgrove EA, Sutherland RL. Antiestrogen inhibition of cell cycle progression in breast cancer cells is associated with inhibition of cyclin-dependent kinase activity and decreased retinoblastoma protein phosphorylation. Mol Endocrinol 1995;9: 1804-13. 15. Yamada T, Marshak D, Basinger S, Walsh J, Morley J, Stell W. Somatostatin-like immunoreactivity in the retina. Proc Nat Acad Sci USA 1980;77:1691-5. 16. Sarkar R, Webb RC, Stanley JC. Nitric oxide inhibition of endothelial cell mitogenesis and proliferation. Surgery 1995;118:274-9. 17. Sarkar R, Gordon D, Stanley JC, Webb RC. Dual cell cyclespecific mechanisms mediate the antimitogenic effects of nitric oxide in vascular smooth muscle cells. J Hypertens 1997;15:275-83. 18. Vargas R, Bormes GW, Wroblewska B, Foegh ML, Kot PA, Ramwell PW. Angiopeptin inhibits thymidine incorporation
JOURNAL OF VASCULAR SURGERY Volume 29, Number 4
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
in rat carotid artery in vitro. Transplant Proc 1989;21: 3702-4. Brevini TA, Bianchi R, Motta M. Direct inhibitory effect of somatostatin on the growth of the human prostatic cancer cell line LNCaP: possible mechanism of action. J Clin Endocrinol Metab 1993;77:626-31. Setyono-Han B, Henkelman MS, Foekens JA, Klijn JGM. Direct inhibitory effects of somatostatin (analogues) on the growth of human breast cancer cells. Cancer Res 1987; 47:1566-70. Payan DG, Hess CA, Goetzl EJ. Inhibition by somatostatin of the proliferation of T-lymphocytes and Molt-4 lymphoblasts. Cell Immunol 1984;84:433-8. Chou CK, Ho LT, Ting LP, et al. Selective suppression of insulin-induced proliferation of cultured human hepatoma cells by somatostatin. J Clin Invest 1987;79:175-8. Fillinger MF, O’Connor SE, Wagner RJ, Cronenwett JL. The effect of endothelial cell coculture on smooth muscle cell proliferation. J Vasc Surg 1993;17:1058-67. Hierowski MT, Liebow C, du Sapin K, Schally AV. Stimulation by somatostatin of dephosphorylation of membrane proteins in pancreatic cancer MIA PaCa-2 cell line. FEBS Lett 1985;179:252-6. Bauters C, Van Belle E, Werner N, et al. Angiopeptin inhibits oncogene induction in rabbit aorta after balloon denudation. Circulation 1994;89:2327-31. Foegh ML, Asotra S, Conte JV, Howell M, Kagan E, Verma K, et al. Early inhibition of myointimal proliferation by angiopeptin after balloon catheter injury in the rabbit. J Vasc Surg 1994;19:1084-91. Campan M, Desgranges C, Gadeau A, Millet D, Belloc F. Cell cycle dependent gene expression in quiescent stimulated and asynchronously cycling arterial smooth muscle cells in culture. J Cell Physiol 1992;150:493-500. Miano JM, Tota RR, Vlasic N, Danishefsky KJ, Stemerman
Sarkar, Dickinson, and Stanley
29.
30.
31.
32.
33.
34.
35.
693
MB. Early proto-oncogene expression in rat aortic smooth muscle cells following endothelial removal. Am J Pathol 1990;137:761-5. Kamiya Y, Ohmura E, Arai M, et al. Effect of somatostatin and its analogue on proliferation of human epidermoid carcinoma cells in vitro. Biochem Biophys Res Commun 1993;191:302-7. Mooradian DL, Fernandes B, Diglio CA, Lester BR. Angiopeptin (BIM23014C) inhibits vascular smooth muscle cell migration in vitro through a G-protein-mediated pathway and is associated with inhibition of adenylyl cyclase and cyclic AMC accumulation. J Cardiovasc Pharmacol 1995; 25:611-8. Clowes AW, Schwartz SM. Significance of quiescent smooth muscle migration in the injured rat carotid artery. Circ Res 1985;56:139-45. Leszczynski D, Josephs MD, Fournier RS, Foegh ML. Angiopeptin, the octapeptide analogue of somatostatin, decreases rat heart endothelial cell adhesiveness for mononuclear cells. Regul Pept 1993;43:131-40. Hancock WW, Adams DH, Wyner LR, Sayegh MH, Karnovsky MJ. CD4+ mononuclear cells induce cytokine expression, vascular smooth muscle cell proliferation, and arterial occlusion after endothelial injury. Am J Pathol 1994;145:1008-14. Ferns GA, Reidy MA, Ross R. Balloon catheter de-endothelialization of the nude rat carotid. Response to injury in the absence of functional T lymphocytes. Am J Pathol 1991; 138:1045-57. Castellot JJ, Pukac LA, Caleb BL, Wright TC Jr, Karnovsky MJ. Heparin selectively inhibits a protein kinase C-dependent mechanism of cell cycle progression in calf aortic smooth muscle cells. J Cell Biol 1989;109:3147-55.
Submitted Jun 9, 1998; accepted Nov 9, 1998.
Smooth muscle cell (SMC) proliferation plays an important role in the pathogenesis of a number of vascular disorders. In atherosclerosis, SMC proliferation is present at low but persistent levels and is associated with plaque neovascularization.1 The proFrom the Conrad Jobst Vascular Research Laboratories, Section of Vascular Surgery, Department of Surgery (Drs Sarkar and Stanley) and Division of Gastroenterology, Department of Pediatrics, University of Michigan Medical Center. Supported by a National Research Service Award (F-32Hl08677) from the National Heart, Lung and Blood Institute, National Institutes of Health (R.S.), and NIH grant HL02816 (J.C.S.). Presented at the Poster Session of the 1997 Joint Annual Meeting of the Society for Vascular Society/International Society for Cardiovascular Surgery (North American chapter), June 1–4, 1997, Boston, Mass. Reprint requests: Dr James C. Stanley, University Hospital, 2210 THC, 1500 East Medical Center Dr, Ann Arbor, MI 48109. Copyright © 1999 by the Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/99/$8.00 + 0 24/1/95743
liferation of SMCs is more pronounced in experimental models of arterial balloon injury2 and vein grafting,3 and the pharmacologic inhibition of SMC proliferation is associated with a decrease in neointimal hyperplasia.4 The variety of pharmocologic agents that can inhibit SMC proliferation and neointimal hyperplasia includes the peptide analogs of the neuroendocrine hormone somatostatin. Somatostatin analogs, such as octreotide and angiopeptin, inhibit neointimal hyperplasia and SMC proliferation in experimental balloon angioplasty,5,6 vein graft models,7 and transplant arteriosclerosis.8 Somatostatin is an endogenous peptide and thus could be produced with genetically modified vascular cells. The introduction of either previously transduced cells onto sites of arterial injury or direct gene transfer to the arterial wall then could be used as a potential therapy for neointimal hyperplasia. Retroviral gene transfer to both endothelial cells9 (ECs) and SMCs10 in culture followed by the seed685
686 Sarkar, Dickinson, and Stanley
JOURNAL OF VASCULAR SURGERY April 1999
noted that correlated with the in vitro and in vivo effects of somatostatin.
Fig 1. Somatostatin production by β-galactosidase gene–transfected endothelial cells (BAG EC) and human somatostatin-transfected endothelial cells (HSS EC) in various serum concentrations. Transfected and selected endothelial cells were grown to confluence, and media were changed to indicated serum concentrations. After 24 hours, media were removed and assayed for somatostatin peptide. *P < .05 versus β-galactosidase gene media (n = 2 to n = 3). #P < .05 versus β-galactosidase gene media and versus human somatostatin 2.5% serum.
ing of these cells onto sites of arterial injury has resulted in the long-term expression of marker proteins, such as galactosidase and human adenosine deaminase in vivo. The seeding of retrovirally transduced ECs onto prosthetic grafts also results in the long-term coverage of the graft surface and the production of the transgene of interest.11 The production of biologically active somatostatin after gene transfer necessitates the presence of multiple separate prohormone processing endopeptidase enzymes, and recent studies from our laboratory have shown that the retroviral gene transfer of the human somatostatin complementary DNA to cultured ECs results in the secretion of significant concentrations of properly processed somatostatin peptide.12 The purpose of this study was to determine whether the retroviral gene transfer of somatostatin to ECs would inhibit the proliferation of cultured SMCs. Additional studies were performed to determine the effects of various somatostatin analogs on the proliferation and DNA synthesis of cultured SMCs. Three different in vitro models of SMC proliferation were used, with significant differences
MATERIALS AND METHODS Cell derivation and culture. Canine venous ECs were derived from the external jugular veins of adult mongrel dogs as previously described13 and were maintained in EC growth media (M199 media, Gibco, Long Island, NY), with 5% bovine calf serum (Hyclone, Logan, Utah), EC growth supplement 10 µg/mL (Collaborative Biochemical, Bedford, Mass), heparin 15 U/mL, penicillin-streptomycin 1 mmol/L L-glutamine, in a 5% CO2 humidified 37°C incubator on gelatin-coated plates. The cultures were fed every 2 to 3 days and were split when confluent. EC type was confirmed by means of cell morphology, lack of staining with an antibody to α-actin, development of a confluent monolayer, and uptake of fluorescent acetylated-low density lipoprotein ligand. Rat SMCs and canine SMCs were derived with enzymatic dissociation from minced segments of rat thoracic aorta or canine femoral artery. The segments were incubated in Dulbecco’s serum-free media that contained 0.1% (wt/vol) collagenase (Worthington Chemicals, Philadelphia, Pa) and 0.01% (wt/vol) elastase (Sigma Chemical, St Louis, Mo) for 2 to 3 hours at 37°C with constant agitation. Detached cells in the supernatant were pelleted and plated into media that consisted of Dulbecco’s minimal essential media (DMEM) with 10% (vol/vol) fetal bovine (rat cells) or bovine calf (canine cells) serum, 1 mmol/L glutamine, and antibiotics. The SMC identity was confirmed by means of morphology and immunoflourescent staining with an α-actin antibody (HF-35, Sigma Chemical). For growth experiments, SMCs were used in passages 2 to 10. Retroviral transfection. Retroviruses with selectable neomycin resistance genes that contained either the human somatostatin (HSS) complementary DNA or a control β-galactosidase gene (BAG) were used to create pure populations of ECs that expressed either β-galactosidase (BAG EC) or human somatostatin (HSS EC). Previous studies have shown that canine BAG ECs have uniform βgalactosidase expression and that HSS ECs produce the somatostatin-14 isoform of somatostatin at concentrations of 400 pmol/L in growth media.12 Rat ECs display nonuniform β-galactosidase expression and do not produce significant concentrations of somatostatin after retroviral transfection.14 For transfection, the ECs in passages 2 or 3 were split 1:5 the day before transfection to maximize the number of replicating cells capable of undergoing
JOURNAL OF VASCULAR SURGERY Volume 29, Number 4
Sarkar, Dickinson, and Stanley
687
Fig 2. Effect of transfer of media conditioned by β-galactosidase gene–transfected endothelial cells (BAG EC) and human somatostatin-transfected endothelial cells (HSS EC) on growth of smooth muscle cells (SMCs). Media containing indicated concentrations of serum were conditioned with canine BAG ECs and HSS ECs every 24 hours and then were transferred to rat or canine SMCs daily. SMC cell number was determined after 72 hours (rat, n = 3; canine, n = 4). *P < .05 versus BAG media.
retroviral gene transfer. The ECs were exposed to the HSS viral supernatant with the cationic agent Polybrene (8 µg/mL) for 8 hours and then were refed with appropriate fresh media. After a 48-hour period to allow the reverse transcription of the viral RNA genome, the incorporation of the DNA provirus into a host chromosome, and the expression of the transgenes, the cells were split 1:8 into appropriate complete media supplemented with Geneticin G418 (Sigma Chemical, St Louis, Mo) at 1 mg/mL. Nontransduced ECs from the same passage were split simultaneously into G418-containing media to serve as controls. All the cells were maintained in G418-containing media for approximately 14 days, or 3 days after all the control nontransduced cells had died. Control ECs for all the experiments were transduced with the BAG marker retrovirus with identical procedures as described previously except with the BAG retroviral supernatant used for transduction. Analysis of gene expression. Somatostatin production was assayed with the measurement its level in media from confluent monolayers of BAG-transduced and HSS-transduced ECs that had been collected after 24-hour incubations and stored at –70°C for subsequent radioimmunoassay. The radioimmunoassay used synthetic somatostatin-14 as a standard and an antibody (antibody 1001) that binds to the carboxyl terminus of somatostatin.15 Media transfer and coculture studies. To
determine the effects of somatostatin produced by HSS-transduced ECs on SMC growth, media conditioned with HSS ECs was transferred to growing canine SMCs. Canine or rat SMCs were plated at a density of 5000 cells/cm2 in growth media. The following day, EC media, which had been placed on confluent flasks of BAG ECs or HSS ECs, were removed and transferred to the SMC wells. The EC monolayers were refed with fresh EC media, which then were transferred again to the SMCs after 24 hours of conditioning with the ECs. After a total incubation period of 72 hours in conditioned media, the SMCs were counted as described subsequently. For coculture studies, the Transwell insert system (a polycarbonate filter with a 0.22-micron pore size, Corning Costar Inc, Corning, NY) was used with ECs grown in the Transwell insert and SMCs grown on the underlying well. The ECs were plated onto the insert, which was placed in separate wells that contained EC media. These media were changed until the ECs were confluent. The inserts then were removed and placed into separate wells in which canine SMCs had been plated 24 hours earlier at a density of 5000 cells/cm2. Both the cell types then were maintained in coculture (in EC media) for 96 hours with one change of the media at 48 hours. In selected groups, the somatostatin analog octreotide (10-10 mol/L, Sigma Chemical) was added daily. The integrity of the EC monolayer was verified with daily phase contrast microscopy with-
688 Sarkar, Dickinson, and Stanley
JOURNAL OF VASCULAR SURGERY April 1999
Fig 3. Canine smooth muscle cell (SMC) growth in presence of various endothelial cell (EC) types or somatostatin analog. Canine SMCs were grown in coculture for 96 hours with indicated types of canine ECs. Octreotide (10–10 mol/L) was added in indicated groups. No significant differences were noted among four groups that were cocultured with ECs (SMC + EC, SMC + β-galactosidase gene (BAG) EC, SMC + human somatostatin (HSS) EC, SMC + EC + octreotide). All groups, n = 4. *P < .05 versus SMC alone.
out disturbing the coculture system. At the conclusion of the experiment, the inserts were removed and the SMCs were counted as subsequently described. DNA synthesis and cell proliferation assays. DNA synthesis was analyzed in cells that were initially rendered quiescent and then was triggered to reenter the cell cycle with serum restimulation. Canine or rat SMCs were treated with serum-free media (DMEM plus 1% (wt/vol) albumin) for 48 hours, and then media that contained serum were added with experimental agents as indicated. In selected experiments, the somatostatin analogs were added to the serum-free media to pretreat the cells before the mitogenic stimulus. DNA synthesis was measured 24 hours after the readdition of the serum with the incorporation of tritiated thymidine as previously described.16,17 In brief, 0.37 MBq/mL 3Hthymidine (Dupont, Boston, Mass) was added to the media, and, after 2 hours, the amount of radiotracer incorporation into the cellular DNA was determined by means of precipitation with 10% ice-cold trichloroacetic acid and scintillation counting. Cell proliferation was measured by counting cell numbers with a Coulter counter (Coulter Electronics, Hialeah, Fla) after the enzymatic detachment of the cells. The cells were washed twice with phosphate-buffered saline solution and were treated with 0.05% trypsin/ethylenediamine tetraacetic acid until complete cell detachment was confirmed with phase microscopy. The cell mixture then
was added to 20 mL of counting solution, and triplicate counts were taken of 1-mL aspirations of each sample. The percent inhibition of cell proliferation in experimental groups was determined with the following formula: (control count – experimental count)/control count × 100. For DNA synthesis studies in fresh arterial explants, the method of Vargas was used. 18 Segments of canine femoral artery were removed after the deaths of mongrel adult male dogs and were placed in sterile phosphate-buffered saline solution. Under sterile conditions, the adventitia was dissected free and the endothelium was removed with mechanical denudation. With a ruler, 3-mm square segments then were cut by means of surgical magnification. These segments were placed in 1 mL of DMEM media (without serum) that contained antibiotics, experimental agents, and 0.37 MBq/mL 3H-thymidine. After a 24-hour incubation period, each segment was incubated with cold thymidine (1 mmol/L) for 1 hour, washed with phosphatebuffered saline solution repeatedly, and dissolved in 1 mol/L of sodium hydroxide at 37°C overnight. The resulting solution was neutralized with an equivalent volume of 1 mol/L of hydrochloric acid before the addition of scintillation cocktail and scintillation counting. Statistical analysis. An unpaired t test was used for comparisons between two groups. Analysis of variance with correction for multiple comparisons was used for comparisons between several groups.
JOURNAL OF VASCULAR SURGERY Volume 29, Number 4
Sarkar, Dickinson, and Stanley
689
All the data are presented as the mean ± the standard error of the mean. Differences were considered significant at a P value of less than .05. RESULTS The ability of somatostatin-transduced ECs to produce somatostatin peptide first was assessed in the presence of varying concentrations of serum. We previously have documented that HSS ECs produce approximately 300 to 400 pmol/L of somatostatin concentration in the surrounding media in full growth media (ie, 10% serum).12 The growth of cultured SMCs is dependent on growth factors present in serum, unlike ECs, which necessitate exogenous acidic fibroblast growth factor in the form of EC growth supplement. A range of serum concentrations was used to ensure that a mild antimitogenic effect of the somatostatin peptide would not be masked by an overwhelming mitogenic stimuli from serum. Somatostatin production increased with the concentration of serum used, although the concentrations were in the range of 300 to 500 pmol/L over the range of serum concentrations from 2.5% to 10% (Fig 1). These concentrations of somatostatin have been reported to inhibit the in vitro proliferation of prostate cancer cells,19 breast carcinoma,20 and T lymphocytes.21 There was no significant inhibition of either canine or rat SMC proliferation noted with the transfer of conditioned somatostatin media (Fig 2). As expected, the proliferation of SMCs was dependent on the serum concentration in the transferred media, with greater cell growth in 10% serum. There are several possible reasons for this failure of growth inhibition, including the possibility that the transfer of conditioned media every 24 hours may allow significant degradation of somatostatin peptide in the EC media before the transfer to the SMCs. A loss of biologic activity as a result of peptide degradation is of particular concern with somatostatin because it has a half-life of less than 2 hours in media.22 To minimize the degradation of somatostatin between the time of secretion from the ECs to the contact with the SMCs, we then grew SMCs continuously in the presence of somatostatinproducing ECs (Fig 3). The presence of ECs in the coculture system, either nontransduced or transduced with either the HSS or the BAG control vector, increased SMC proliferation regardless of the EC type (Fig 3), as has been reported by other investigators.23 Octreotide (10-10 mol/L) did not alter the growth of SMCs, with or without the presence of ECs (Fig 3). To
Fig 4. Effect of somatostatin (SS) analogs on DNA synthesis and cell proliferation of canine and rat smooth muscle cells (SMCs). Nitric oxide donor S-nitroso-Nacetylpenicillamine (SNAP; 0.1 mmol/L) was used as positive control for inhibition of proliferation. DNA synthesis (thymidine incorporation) was measured in serum-restimulated quiescent SMCs. Cell number was determined after 72 hours in presence of indicated concentrations of somatostatin analogs (n = 3 to n = 4). *P < .05 versus control.
ensure that this lack of proliferation was not simply caused by the concentrations of somatostatin or octreotide, the effects of a range of concentrations of different somatostatin isoforms and analogs on
690 Sarkar, Dickinson, and Stanley
JOURNAL OF VASCULAR SURGERY April 1999
Fig 5. Effect of angiopeptin (150 ng/mL, 75 µmol/L) on DNA synthesis in canine arterial explants (clear bars) and 72-hour cell proliferation of subcultured canine smooth muscle cells (SMCs; solid bars; n = 4). *P < .05 versus control.
growth and DNA synthesis of several species of cultured SMCs then were determined. Octreotide did not inhibit the DNA synthesis of canine SMCs over a wide range of concentrations (Fig 4), and somatostatin-14 failed to inhibit the proliferation of both rat and canine SMCs. We used the nitric oxide donor Snitroso-N-acetylpenicillamine (0.1 mmol/L), a known inhibitor of SMC growth,16,17 as a positive control to confirm that our experimental system could detect changes in cell proliferation. Similar negative results were found with various concentrations of octreotide and cell proliferation and with either somatostatin-14 or somatostatin-28 and DNA synthesis in synchronized cells and with cultured human SMCs (all studies, n = 3 to n = 4; P > .05 vs control for concentrations 10–6 to 10–10 mol/L). The pretreatment of quiescent SMCs with somatostatin analogs before restimulation with serum also did not result in the inhibition of subsequent DNA synthesis (n = 4; P > .05 vs control for 10-7 to 10-10 mol/L). Previous studies have shown that the somatostatin analog angiopeptin inhibits DNA synthesis in arterial explants ex vivo.18 To examine the discrepancy between this finding and our uniformly negative findings with somatostatin analogs (Fig 4), we compared the effects of angiopeptin on DNA synthesis in cultured SMCs and fresh arterial explants. Angiopeptin (150 ng/mL, 75 µmol/L) significantly inhibited DNA synthesis in arterial segments within the first 24 hours after explantation but failed to inhibit DNA synthesis in serum-restimulated quiescent cultured SMCs (Fig 5). Experiments that used the transfer of somatostatin-conditioned media produced with ECs to inhibit DNA synthesis in arterial
explants were not feasible because somatostatin analogs only inhibit DNA synthesis in arterial explants under serum-free conditions.18 DISCUSSION The significant findings in this study were: (1) the production of somatostatin after the gene transfer of human somatostatin to ECs does not inhibit the proliferation of cultured SMCs; (2) neither exogenous somatostatin nor its analogs inhibit mitogenesis or proliferation of several species of SMCs; and (3) the somatostatin analog angiopeptin inhibits DNA synthesis in fresh arterial explants but not in cultured SMCs, which thus shows a potential mechanism for the failure of gene transfer of somatostatin to inhibit SMC growth. To our knowledge, this report is the first description of the effects of somatostatin on the growth of cultured SMCs. Our findings in cultured SMCs and arterial explants correlate not only with the molecular mechanism of action of somatostatin but also with the efficacy of somatostatin analogs in decreasing neointimal hyperplasia in vivo. The molecular mechanism by which somatostatin inhibits cell proliferation has been defined in several different cell types. Somatostatin binds to cell surface receptors and activates tyrosine phosphatases, which dephosphorylate the signaling proteins that are the substrates for mitogen-activated tyrosine kinases. 24 Consistent with this early inhibition of mitogenic signal transduction is the finding that the administration of angiopeptin must be within the first hour after arterial injury for the subsequent inhibition of proliferation to occur.25,26
JOURNAL OF VASCULAR SURGERY Volume 29, Number 4
Our media transfer and coculture experiments were performed with the SMCs growing continuously in serum-containing media. The cell cycle events of continuously proliferating SMCs are different from both those of cultured SMCs synchronized with serum starvation27 and of quiescent SMCs in uninjured arteries.28 It is possible that the actions of somatostatin analogs noted in vivo are caused by the inhibition of SMC proliferative events during the initial transition from quiescence to proliferation, and thus these effects would be missed in our experiments with continuously proliferating SMCs (Figs 2, 3). We thus used quiescent synchronized SMCs to which somatostatin analogs were added either with or before the serum restimulation, and we were not able to detect either inhibition of DNA synthesis or cell number (Fig 4). This suggests that the inhibition of neointimal hyperplasia with somatostatin analogs in vivo is either independent of cell proliferation or is acting on elements of SMC proliferation that are not present in either of these commonly used in vitro models. In the media transfer experiment, the rat SMCs that were treated with HSS media showed greater proliferation than did the control cells (Fig 2). This greater proliferation was not seen with the canine cells (Fig 2) nor with the rat SMCs at higher or lower serum concentrations. Although somatostatin has been shown to increase the proliferation of certain cell types,29 the isolated finding with rat SMCs at 5% serum is not reproduced with exogenous somatostatin (Fig 4). Thus, although we do not have an explanation for this result, the effect appears to be an isolated event that is particular to the media transfer assay with rat SMCs. There are other mechanisms by which somatostatin may decrease neointimal hyperplasia besides the direct inhibition of SMC proliferation. Somatostatin inhibits the migration of SMCs30 and thus may inhibit the migration of SMCs from the media into the intima, which is critical to the formation of intimal lesions.31 Somatostatin inhibits the adhesion of inflammatory cells,32 and recent studies33 have shown that a significant fraction of the neointimal cells traditionally thought of as SMCs2,34 are actually inflammatory cells. Thus, the inhibition of inflammatory cell adhesion with somatostatin may inhibit net intimal lesion size independent of direct effects on SMC proliferation. The inhibition of inflammatory cell adhesion also inhibits SMC proliferation via the decreased production of SMC mitogens with inflammatory cells.33 Despite these other potential confounding mechanisms, the finding that angiopeptin inhibits
Sarkar, Dickinson, and Stanley
691
SMC proliferation in rat18 and canine arterial explants (Fig 5) suggests that the direct inhibition of SMC proliferation may be a contributory mechanism to the overall inhibition of neointimal hyperplasia with somatostatin analogs. This study shows that there is a lack of correlation between the in vivo (or fresh arterial explant model) findings and the standard cell cycle model of serum-starved cultured SMCs, which reenter the cell cycle on readdition of serum. Serum-restimulated quiescent SMCs recapitulate many of the cell cycle events that are noted after in vivo arterial injury, including time-dependent expression of immediate response genes, such as fos and myc, induction of polyamine synthesis, and DNA synthesis after 24 to 48 hours.27 This study did not evaluate the effect of EC-derived somatostatin on DNA synthesis in fresh arterial explants because the serum-free assay conditions for explants were not compatible with the ECconditioned media. However, given our previous characterization of EC-derived somatostatin as somatostatin-14,12 experiments with EC-derived somatostatin would add little to the results that were obtained with exogenous somatostatin analogs (Figs 4, 5). The use of the cultured SMC model in understanding in vivo responses to antiproliferative agents is shown best by the case of heparin, which inhibits cell cycle progression in the late G1 phase, with a loss of inhibition of subsequent DNA synthesis when heparin is added more than 12 hours after the serum restimulation of quiescent rat SMCs. 35 Similarly, the delayed administration of heparin after arterial injury causes a loss of inhibition of subsequent proliferation.4 This correlation is not borne out by the somatostatin analogs in our study because the ability of angiopeptin to inhibit DNA synthesis in vivo5 and in fresh arterial explants18 is not reflected in the inhibition of DNA synthesis in serumrestimulated SMCs (Fig 5). Thus, the somatostatin analogs act on an early aspect of the phenotype change of in vivo SMCs, which occurs after arterial explantation18 or balloon injury25 but which is not mimicked by the serum restimulation of cultured SMCs. This discrepancy between in vivo and in vitro findings appears to be specific to the SMC cell type because other types of passaged cells show inhibition of growth in vitro on treatment with somatostatin.19-21 Another possible factor in evaluating the antiproliferative effects of somatostatin may be gender differences in SMC response. However, angiopeptin inhibits proliferation in male animals25 and all the cell types used in this
JOURNAL OF VASCULAR SURGERY April 1999
692 Sarkar, Dickinson, and Stanley
study, as well as the explants in previous studies,18 were from male animals. Attempts to define the molecular mechanisms that distinguish the in vivo and in vitro activities of somatostatin analogs are limited by the lack of effect in cultured SMCs. The critical proliferative signals in SMCs that somatostatin analogs inhibit occur almost immediately after arterial injury because angiopeptin pretreatment inhibits the expression of the proto-oncogenes fos and jun 30 minutes after injury.25 These in vivo signaling events are presumably also inhibited ex vivo in arterial explants but not in serum-restimulated quiescent cultured SMCs. Unfortunately, the model of serum-restimulated cultured cells is most suited to the analysis of molecular signaling events,19-21 and it is this particular model in which somatostatin analogs do not display activity on SMCs. Thus, a limitation of this study, and of future efforts to define the mechanisms of action of somatostatin on SMCs, is the lack of a simple in vitro model in which these mechanisms can be characterized, as has been done with other cell types.19-21 These findings have three consequences for the development of antiproliferative SMC therapy. The study shows for the first time, to the best of our knowledge, that there are undefined but critical cell cycle events that occur after arterial injury in vivo and that are not present in the widely used in vitro model of serum-restimulated quiescent SMCs. Secondly, this finding implies that the use of the serum-restimulated (Fig 5) or the continuously proliferating cultured SMC model (Fig 4) to screen potential antiproliferative compounds may miss those, such as angiopeptin, that do not inhibit DNA synthesis in culture but do have significant in vivo effects. Finally, our findings suggest that the reseeding of arterial injury sites with somatostatin-producing ECs may not inhibit subsequent neointimal hyperplasia. Both the lack of inhibition of SMC proliferation after the phenotype change in culture and the loss of effect of angiopeptin when delayed 1 hour after injury25,26 suggest that critical somatostatin-sensitive events in SMC proliferation will probably occur well before significant somatostatin production with reseeded ECs occurs. We thank the Henri Beaufour Institute USA, Inc, who supplied the angiopeptin, Susan Finniss, who assisted with the somatostatin RIA assay, Abraham Jacobs, MD, and Eric Meinberg, MD, who assisted with cell culture, Thomas Wakefield, MD, who supplied the canine femoral arteries, and Peter Ramwell, PhD, for helpful discussions.
REFERENCES 1. Rekhter MD, Gordon D. Does platelet-derived growth factor-A chain stimulate proliferation of arterial mesenchymal cells in human atherosclerotic plaques? Circ Res 1994;75:410-7. 2. Clowes AW, Reidy MA, Clowes MM. Kinetics of cellular proliferation after arterial injury. I. Smooth muscle growth in the absence of endothelium. Lab Invest 1983;49:327-33. 3. Zwolak RM, Adams MC, Clowes AW. Kinetics of vein graft hyperplasia: association with tangential stress. J Vasc Surg 1987;5:126-36. 4. Majesky MW, Schwartz SM, Clowes MM, Clowes AW. Heparin regulates smooth muscle S phase entry in the injured rat carotid artery. Circ Res 1987;61:296-300. 5. Lundergan C, Foegh ML, Vargas R, et al. Inhibition of myointimal proliferation of the rat carotid artery by the peptides, angiopeptin and BIM 23034. Atherosclerosis 1989;80:49-55. 6. Yumi K, Fagin JA, Yamashita M, et al. Direct effects of somatostatin analog octreotide on insulin-like growth factorI in the arterial wall. Lab Invest 1997;76:329-38. 7. Calcagno D, Conte JV, Howell MH, Foegh ML. Peptide inhibition of neointimal hyperplasia in vein grafts. J Vasc Surg 1991;13:475-9. 8. Foegh ML. Angiopeptin: a treatment for accelerated myointimal hyperplasia? J Heart Lung Transplant 1992;11:S28-31. 9. Nabel EG, Plautz G, Boyce FM, Stanley JC, Nabel GJ. Recombinant gene expression in vivo within endothelial cells of the arterial wall. Science 1989;244:1342-4. 10. Lynch CM, Clowes MM, Osborne WR, Clowes AW, Miller AD. Long-term expression of human adenosine deaminase in vascular smooth muscle cells of rats: a model for gene therapy. Proc Nat Acad Sci USA 1992;89:1138-42. 11. Baer RP, Whitehill TE, Sarkar R, Sarkar M, Messina LM, Komorowski TA, et al. Retroviral-mediated transduction of endothelial cells with the lac Z gene impairs cellular proliferation in vitro and graft endothelialization in vivo. J Vasc Surg 1996;24:892-9. 12. Sarkar R, Finniss S, Dickinson CJ, Stanley JC. Somatostatin gene transfer and expression in endothelial cells. J Vasc Surg 1998;27:955-62. 13. Graham LM, Burkel WE, Ford JW, Vinter DW, Kahn RH, Stanley JC. Immediate seeding of enzymatically derived autologous canine endothelium in dacron velour vascular grafts. Arch Surg 1980;115:1289-94. 14. Watts CKW, Brady A, Sarcevic B, deFazio A, Musgrove EA, Sutherland RL. Antiestrogen inhibition of cell cycle progression in breast cancer cells is associated with inhibition of cyclin-dependent kinase activity and decreased retinoblastoma protein phosphorylation. Mol Endocrinol 1995;9: 1804-13. 15. Yamada T, Marshak D, Basinger S, Walsh J, Morley J, Stell W. Somatostatin-like immunoreactivity in the retina. Proc Nat Acad Sci USA 1980;77:1691-5. 16. Sarkar R, Webb RC, Stanley JC. Nitric oxide inhibition of endothelial cell mitogenesis and proliferation. Surgery 1995;118:274-9. 17. Sarkar R, Gordon D, Stanley JC, Webb RC. Dual cell cyclespecific mechanisms mediate the antimitogenic effects of nitric oxide in vascular smooth muscle cells. J Hypertens 1997;15:275-83. 18. Vargas R, Bormes GW, Wroblewska B, Foegh ML, Kot PA, Ramwell PW. Angiopeptin inhibits thymidine incorporation
JOURNAL OF VASCULAR SURGERY Volume 29, Number 4
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
in rat carotid artery in vitro. Transplant Proc 1989;21: 3702-4. Brevini TA, Bianchi R, Motta M. Direct inhibitory effect of somatostatin on the growth of the human prostatic cancer cell line LNCaP: possible mechanism of action. J Clin Endocrinol Metab 1993;77:626-31. Setyono-Han B, Henkelman MS, Foekens JA, Klijn JGM. Direct inhibitory effects of somatostatin (analogues) on the growth of human breast cancer cells. Cancer Res 1987; 47:1566-70. Payan DG, Hess CA, Goetzl EJ. Inhibition by somatostatin of the proliferation of T-lymphocytes and Molt-4 lymphoblasts. Cell Immunol 1984;84:433-8. Chou CK, Ho LT, Ting LP, et al. Selective suppression of insulin-induced proliferation of cultured human hepatoma cells by somatostatin. J Clin Invest 1987;79:175-8. Fillinger MF, O’Connor SE, Wagner RJ, Cronenwett JL. The effect of endothelial cell coculture on smooth muscle cell proliferation. J Vasc Surg 1993;17:1058-67. Hierowski MT, Liebow C, du Sapin K, Schally AV. Stimulation by somatostatin of dephosphorylation of membrane proteins in pancreatic cancer MIA PaCa-2 cell line. FEBS Lett 1985;179:252-6. Bauters C, Van Belle E, Werner N, et al. Angiopeptin inhibits oncogene induction in rabbit aorta after balloon denudation. Circulation 1994;89:2327-31. Foegh ML, Asotra S, Conte JV, Howell M, Kagan E, Verma K, et al. Early inhibition of myointimal proliferation by angiopeptin after balloon catheter injury in the rabbit. J Vasc Surg 1994;19:1084-91. Campan M, Desgranges C, Gadeau A, Millet D, Belloc F. Cell cycle dependent gene expression in quiescent stimulated and asynchronously cycling arterial smooth muscle cells in culture. J Cell Physiol 1992;150:493-500. Miano JM, Tota RR, Vlasic N, Danishefsky KJ, Stemerman
Sarkar, Dickinson, and Stanley
29.
30.
31.
32.
33.
34.
35.
693
MB. Early proto-oncogene expression in rat aortic smooth muscle cells following endothelial removal. Am J Pathol 1990;137:761-5. Kamiya Y, Ohmura E, Arai M, et al. Effect of somatostatin and its analogue on proliferation of human epidermoid carcinoma cells in vitro. Biochem Biophys Res Commun 1993;191:302-7. Mooradian DL, Fernandes B, Diglio CA, Lester BR. Angiopeptin (BIM23014C) inhibits vascular smooth muscle cell migration in vitro through a G-protein-mediated pathway and is associated with inhibition of adenylyl cyclase and cyclic AMC accumulation. J Cardiovasc Pharmacol 1995; 25:611-8. Clowes AW, Schwartz SM. Significance of quiescent smooth muscle migration in the injured rat carotid artery. Circ Res 1985;56:139-45. Leszczynski D, Josephs MD, Fournier RS, Foegh ML. Angiopeptin, the octapeptide analogue of somatostatin, decreases rat heart endothelial cell adhesiveness for mononuclear cells. Regul Pept 1993;43:131-40. Hancock WW, Adams DH, Wyner LR, Sayegh MH, Karnovsky MJ. CD4+ mononuclear cells induce cytokine expression, vascular smooth muscle cell proliferation, and arterial occlusion after endothelial injury. Am J Pathol 1994;145:1008-14. Ferns GA, Reidy MA, Ross R. Balloon catheter de-endothelialization of the nude rat carotid. Response to injury in the absence of functional T lymphocytes. Am J Pathol 1991; 138:1045-57. Castellot JJ, Pukac LA, Caleb BL, Wright TC Jr, Karnovsky MJ. Heparin selectively inhibits a protein kinase C-dependent mechanism of cell cycle progression in calf aortic smooth muscle cells. J Cell Biol 1989;109:3147-55.
Submitted Jun 9, 1998; accepted Nov 9, 1998.