Endogenous mono-ADP-ribosylation mediates smooth muscle cell proliferation and migration via protein kinase N-dependent induction of c- fos expression Lorraine Yau 1,2 , Brenda Litchie 1 , Shawn Thomas 1 , Benjamin Storie 1 , Natalia Yurkova 1 and Peter Zahradka 1,2 1 Institute of Cardiovascular Sciences, St. Boniface Research Centre and 2 Department of Physiology, University of Manitoba, Winnipeg, MB, Canada ADP-ribosylation has been coupled to intracellular events associated with smooth muscle cell vasoreactivity, cytoskel- etal integrity and free radical damage. Additionally, there is evidence that ADP-ribosylation is required for smooth muscle cell proliferation. Our investigation employed selective inhibitors to establish that mono-ADP-ribosylation and not poly(ADP-ribosyl)ation was necessary for the sti- mulation of DNA synthesis by mitogens. Mitogen treatment increased concomitantly the activity of both soluble and particulate mono-ADP-ribosyltransferase, as well as the number of modified proteins. Inclusion of meta-iodo- benzylguanidine (MIBG), a selective decoy substrate of arginine-dependent mono-ADP-ribosylation, prevented the modification of these proteins. MIBG also blocked the stimulation of DNA and RNA synthesis, prevented smooth muscle cell migration and suppressed the induction of c-fos and c-myc gene expression. An examination of relevant signal transduction pathways showed that MIBG did not interfere with MAP kinase and phosphatidylinositol 3-kin- ase stimulation; however, it did inhibit phosphorylation of the Rho effector, PRK1/2. This novel observation sug- gests that mono-ADP-ribosylation participates in a Rho- dependent signalling pathway that is required for immediate early gene expression. Keywords: ADP-ribosylation; smooth muscle; DNA syn- thesis; c-fos; MAP kinase. Post-translational modification by ADP-ribosylation, the enzymatic transfer of ADP-ribose from NAD + to an acceptor protein, has been grouped into two distinct classes that are distinguished by their reaction mechanisms [1]. O-linked ADP-ribosylation of glutamate residues is catalyzed by poly(ADP-ribose) polymerase (PARP-1), as is the subsequent formation of polymers containing 10–100 ADP-ribose units. This process occurs primarily in the nucleus, and is responsible for modulating DNA– protein interactions [2]. It has been established that poly(ADP-ribosyl)ation participates in DNA-base excision repair [3], while PARP-1 degradation is a marker for apoptosis [4]. PARP-1 has also been linked to differen- tiation of neutrophilic cells [5], and may be important for chromatin condensation [6], centromere function [7] and transcription [8,9]. Unlike poly(ADP-ribosyl)ation, mono-ADP-ribosyla- tion reactions involve the transfer of a single ADP-ribose to various amino acid (arginine, histidine, diphthamide, cysteine, asparagine) residues by mono-ADP-ribosyltrans- ferases (mART) [10]. Investigations of bacterial toxins (e.g. clostridia, cholera, pertussis, diphtheria) that exhi- bited mART activity foreshadowed the identification of endogenous enzymes capable of catalyzing similar reac- tions [1]. The majority of vertebrate mARTs studied to date have been found to modify either cysteine or arginine. The cellular location of these enzymes has been shown to vary, with enzymes detected in the cytosol, microsomal and nuclear fractions [11]. Many membrane- bound mARTs belong to the family of glycosylphospha- tidylinositol (GPI)-anchored proteins present on the extracellular surface. It has been proposed that the GPI- anchored mARTs modulate transmembrane signalling events, as integrin a7 is one target molecule that has been identified [11]. In contrast, soluble ARTs have been shown to modify cytoskeletal proteins such as actin, thus interfering with polymerization [12]. Other critical medi- ators of cellular function that undergo ADP-ribosylation Correspondence to P. Zahradka, Institute of Cardiovascular Sciences, Boniface Research Centre, 351 Tache Avenue, Winnipeg, MB, Canada., Fax: +1 204 233 6723, Tel.: +1 204 235 3507, E-mail: peterz@sbrc.ca Abbreviations: 3AB, 3-aminobenzamide; AngII, angiotensin II; DMEM, Dulbecco’s modified Eagle’s medium; GPI, glycosylphos- phatidylinositol; MAP kinase, mitogen-activated protein kinase; MIBA, meta-iodobenzylamine; MIBG, meta-iodobenzylguanidine; PD128763, 3,4-dihydro-5-methylisoquinoline; PGE 2 , prostaglandin E 2 ; PI3-kinase, phosphatidylinositol 3-kinase; PtdInsP 3 , phos- phatidylinositol-3,4,5-trisphosphate; PRK, protein kinase C-related kinase; PVDF, poly(vinylidene difluoride); SMC, smooth muscle cell. Enzymes: poly(ADP-ribose) polymerase (EC 2.4.2.30); mono-ADP- ribosyltransferase, arginine-dependent (EC 2.4.2.31); protein kinase N/PRK1 (EC 2.7.1.37); extracellular signal-regulated (MAP) kinase (EC 2.7.1.37); 1-phosphatidylinositol 3-kinase (EC 2.7.1.137); Rho (EC 3.6.1.47). (Received 21 August 2002, revised 21 October 2002, accepted 14 November 2002) Eur. J. Biochem. 270, 101–110 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03366.x include various GTP-binding proteins, including G s ,G i and Rho [13–15]. Although these target proteins suggest a role for mono-ADP-ribosylation in signal transduction, the functional significance of these modifications in normal cell physiology, including cell proliferation and differentiation, remains poorly understood. Vascular smooth muscle cells (SMC) are required to proliferate under pathological conditions and following revascularization procedures that damage the vessel wall [16]. Furthermore, this change in proliferation status is preceded by a switch in cell phenotype [17]. Phenotypic modulation, or the conversion of SMC to the less mature phenotype, is reversible once the injury has been repaired [18,19]. However, failure of SMC to revert to the mature phenotype is associated with conditions such as athero- sclerosis, hypertension and restenosis. For this reason, interventions to treat these conditions have focused primar- ily on the ability to interfere with either SMC migration or proliferation. Thyberg et al. [20] and Grainger et al. [21] have reported that inhibitors of ADP-ribosylation prevent phenotypic modulation, and consequently proliferation of SMC. Furthermore, Thyberg et al. [20] concluded that poly(ADP-ribosyl)ation and mono-ADP-ribosylation were essential for the SMC response to mitogenic stimulation based on evidence that inhibitors of both PARP-1 and arginine-dependent mART prevented PDGF-stimulated thymidine incorporation. Although a number of cell processes were investigated, no definitive mechanism of action was identified in these studies. We therefore investigated the role of ADP-ribosylation in the SMC response to mitogen stimulation in greater detail. Over the course of this study, we established that activation of an arginine-dependent mART, but not PARP-1, was required for SMC proliferation, and identified a novel link between mono-ADP-ribosylation and c-fos gene expression. Materials and methods Materials Cell culture materials (DMEM, fetal bovine serum, trypsin, culture dishes) were obtained from Gibco/BRL, as were oligonucleotide primers used for RT-PCR ampli- fication. Poly(vinylidene difluoride) (PVDF) membrane and DNA molecular mass markers were supplied by Roche. The Boyden chamber and Track-Etch Membrane polycarbonate filters used for cell migration assays were purchased from Neuroprobe and Nucleopore, respectively. The GeneAmp RT-PCR kit and radiolabelled chemicals ([ 3 H]uridine, [ 3 H]thymidine, [ 3 H]NAD + ,[ 14 C]adenosine, [ 32 P]orthophosphate; manufactured by New England Nuclear) were from Perkin Elmer-Cetus. SYBR Green I was provided by Molecular Probes, while the BCA Protein assay kit was supplied by Pierce-Endogen. PD128763 was a generous gift from Parke-Davis. Other chemicals, including mitogens and ADP-ribosylation in- hibitors, were purchased from Sigma-Aldrich. Silica G thin layer chromatography plates were from Whatman. Phospho-specific antibodies (Elk1, PRK1/2) were obtained from Cell Signaling Technology. Smooth muscle cell culture Primary cultures of porcine coronary artery SMC were generated from the left anterior descending coronary artery by an explant organ culture method [22] and propogated in DMEM containing 20% fetal bovine serum. Cells were used only after the second passage to maintain consistency between cultures. Quiescence was achieved by incubating in serum-free DMEM supplemented with 11 lgÆmL )1 pyru- vate, 5 lgÆmL )1 transferrin, 1 n M selenium, 0.2 m M ascor- bate and 10 n M insulin for 5 days. DNA and RNA synthesis Quiescent cells were prepared in 24-well dishes and stimu- lated by direct addition of the indicated compounds without replacing the media. When inhibitors were used, they were added 10–15 min prior to the stimulating agents. To measure RNA synthesis, cells were incubated for 6 h with 2 lCi [ 3 H]uridine, added concomitantly with the stimulating agent. Similarly, DNA synthesis was measured by incuba- ting the cells with 2 lCi [ 3 H]thymidine, added 24 h after mitogen stimulation, for 48 h. Incorporation of radiola- belled precursors into trichloroacetic acid-insoluble nucleic acids was measured as described previously [23]. Western blotting Extracts prepared by addition of 150 lL2· SDS/gel loading buffer to cells in 12-well culture dishes were loaded onto 7.5% polyacrylamide gels and the proteins were subsequently transferred to PVDF membrane. The mem- brane was probed with antibodies as described previously [24]. Ponceau S staining was used to ensure equal protein loading. Cell migration SMC were collected by trypsinization and 1 500 cells were added to each well of a Boyden chamber (48-well unit) containing a polycarbonate filter with 5 lm pores. Serum- free DMEM with chemoattractant [1 l M angiotensin II (AngII)] was added to the lower compartment while inhibitors were added to the upper compartment. After 48hinastandardCO 2 incubator at 37 °C, the membrane was removed from the chamber and placed into methanol for 5 min Cells were scraped from the upper side of the membrane and the membrane was subsequently placed into Giemsa stain for 60 min The membrane was mounted on a slide and the cells present on the underside of the membrane were counted (n ¼ 6 per treatment). Subcellular fractionation SMC prepared and treated in 150-mm diameter tissue culture dishes (Nunc) were washed twice in NaCl/P i (0.225 M NaCl, 0.1 M NaP i , pH 7.1), harvested by scraping in 3.0 mL of NaCl/P i and collected by centifugation (5 min, 3 000 g at 4 °C). The cells were disrupted in 2.5 cell pellet volumes of homogenization buffer (0.25 M sucrose, 5 m M Tris/HCl, pH 8.0, 3 m M CaCl 2 ,1 m M EDTA, 0.5 m M EGTA, 0.2 m M phenylmethanesulfonyl fluoride, 25 kUÆmL )1 aprotinin, 102 L. Yau et al. (Eur. J. Biochem. 270) Ó FEBS 2003 25% glycerol) using a Pro200 homogenizer (Pro Scientific Inc.) fitted with a 5-mm generator. Nuclei were removed by low-speed centrifugation (10 min at 8 000 g at 4 °C) and the supernatant subsequently centrifuged (70 000 g for 60 min at 4 °C) to separate the membrane and cytoplasmic fractions. The microsomal pellet was resuspended with 200 lL homogenization buffer. Protein content was deter- mined with the BCA protein assay kit was modified for a 96-well format to measure a 10-lL sample volume. MAP kinase activity gel assay Samples prepared from treated SMC by detergent lysis [50 m M a-glycerphosphate pH 7.4, 0.5% (v/v) Triton X-100, 25% (v/v) glycerol, 2 m M EGTA, 1 m M orthovana- date, 1 m M dithiothreitol, 0.5 m M phenylmethanesulfonyl fluoride, 0.1 m M bacitracin and 20 lgÆmL )1 aprotinin] were loaded (without heating) onto a 10% polyacrylamide gel containing 0.5 mgÆmL )1 myelin basic protein. Following electrophoresis, phosphorylation of myelin basic protein was assayed by incubating the gel with 25 lCi [c- 32 P]ATP in 50 m M Tris/HCl, pH 8.0, 2 m M dithiothreitol, 0.1 m M EGTA, 5 m M MgCl 2 and 100 l M ATP for 1 h at room temperature [25]. The gel was then washed 5 · in 1% sodium pyrophosphate/5% trichloroacetic acid, dried and exposed to reflection film (Dupont) at )80 °C with one intensifying screen. Phosphatidylinositol 3-kinase assay Quiescent cells were incubated with 200 lCiÆmL )1 [ 32 P] orthophosphate for 4 h in phosphate-free media after a 1-h preincubation in phosphate-free serum-free DMEM, as described previously [26]. Cells were then stimulated for 15 min, with inhibitors added 10 min prior to addition of agonist. Phosphatidylinositides were extracted after 5% perchloric acid treatment and analyzed by thin layer chro- matography on silica G plates developed in chloroform/ acetone/methanol/acetic acid/water (80 : 30 : 26 : 24 : 14, v/v), and visualized by autoradiography. Arginine-dependent mono-ADP-ribosyltransferase assay Arginine-dependent mART was assayed by combining 70 lL assay buffer (70 m M Tris/HCl pH 7.5, 100 l M [ 3 H]NAD + (0.25 lCi), 0.1 m M phenylmethanesulfonyl fluoride), 10 lL polyarginine (2 mgÆmL )1 )and20lL cell extract (microsomal or cytosolic fraction) and incubating this mixture for 30 min at 30 °C [27]. The reaction was stopped by the addition of 1 mL cold 20% trichloroacetic acid and the precipitate was captured on GF/C glass fibre filters. The filters were washed extensively with 5% trichlo- roacetic acid and the radioactivity was quantified by liquid scintillation counting. Metabolic labeling of mono-ADP-ribosylated proteins Cells were incubated in 24-well dishes with 40 lCi [ 14 C]adenosine for 16 h in the presence or absence of inhibitor [24]. The cells were washed once with ice-cold NaCl/P i , and solubilized directly with 100 lL2· SDS/ gel loading buffer. The sample was then clarified by centrifugation (12 000 g, 10 min) and an aliquot of 25 lL was loaded onto a 10% polyacrylamide gel. The radio- activity present in the gels was detected with a Molecular Devices Storm phosphorimager. Reverse transcription-polymerase chain reaction amplification Total RNA (1 lg) was amplified according to the protocol recommended for the GeneAmp kit with oligodeoxynucle- otide primers specific for GAPDH (sense: 5¢-CGGTGTG AACGGATTTGGCCGTAT-3¢,antisense:5¢-AGCCTTC TCCATGGTGGTGAAGAC-3¢); c-fos (sense: 5¢-GAATA AGATGGCTGCAGCCAAGTGC-3¢,antisense:5¢-AAG GAAGACGTGTAAGCAGTGCAGC-3¢), and c-myc (sense: 5¢-AAGTTGGACAGTGGCAGGGT-3¢,antisense: 5¢-TTGCTCCTCTGCTTGGACAG-3¢). Amplification was conducted over 35 cycles using a three-step program as described previously [28]. Samples were analyzed by electrophoresis in 1.7% agarose gels and visualized with SYBR Green I. Data measurement and statistical analysis Radiotracer, cell number and enzyme assay data were quantified and plotted as means ± SEM of individual experiments (n ¼ 3–6). Student’s t-test was used to compare treatment means vs. controls. Statistical significance was set at P < 0.05. Quantification of data obtained on film or autoradiographs was accomplished with a Bio-Rad Model-670 Imaging Densitometer under nonsaturating conditions. Results meta -Iodobenzylguanidine inhibits SMC proliferation The incorporation of [ 3 H]uridine and [ 3 H]thymidine was used to quantify the relative rates of RNA and DNA synthesis, respectively. Quiescent SMC were pretreated with meta-iodobenzylguanidine (MIBG), a decoy substrate of arginine-dependent mono-ADP-ribosylation that conse- quently reduces modification of endogenous protein targets [29], for 10 min prior to addition of various growth stimulating agents. MIBG exhibited a concentration- dependent inhibition of both RNA and DNA synthesis in SMC stimulated with AngII (1 l M ), with complete inhibi- tion observed at 25 l M (Fig 1A and C). Similarly, MIBG inhibited both RNA and DNA synthesis in fetal bovine serum-stimulated (2% v/v) SMC (Fig 1B and D), although a decrease to basal levels was not achieved. This is likely due to the magnitude of the SMC growth response elicited by the multiple growth stimulating agents present in fetal bovine serum. However, when MIBG was tested with prostaglandin E 2 (PGE 2 )-stimulated (1 l M )SMC,the degree of inhibition (Fig. 1E) was comparable to that seen with AngII. The inability of meta-iodobenzylamine (MIBA), an inactive analogue of MIBG, to inhibit DNA synthesis, even at a concentration of 50 l M (Fig. 1F), confirmed the specificity of MIBG. Confirmation that cell death is not the mechanism by which MIBG inhibits SMC proliferation was obtained by examining the effect of MIBG Ó FEBS 2003 ADP-ribosylation and SMC proliferation (Eur. J. Biochem. 270) 103 on cell morphology, which revealed no visible changes at concentrations below 500 l M (Fig. 2). Activation of arg-mART by mitogens While the ability of MIBG to inhibit cell proliferation suggests mono-ADP-ribosylation is required for cell cycle progression, direct evidence to support this correlation is limited. Furthermore, there is little experimental support to show that mono-ADP-ribosylation can be activated in response to mitogens. Therefore, arginine-dependent mART activity was measured in extracts of SMC that had been stimulated with either AngII or PGE 2 .Itwas observed that both cytosolic and microsomal arg-mART activity were increased transiently following AngII (1 l M ) stimulation (Fig. 3A), with cytosolic activity 1.78-fold over basal levels at 30 min (Fig. 3B). Likewise, PGE 2 (1 l M ) treatment stimulated microsomal arg-mART (1.35-fold), however, the increase in cytosolic activity was only marginal (Fig. 3B). Activation of arginine-dependent mART would be expected to lead to an increase in the post-translational modification of select proteins. Quiescent (5 days under serum-free conditions) and growing cells (continuously in fetal bovine serum) were therefore incubated for 16 h with [ 14 C]-adenosine. This approach was employed because (a) cells are impermeable to NAD + ,and(b) Fig. 2. Effect of MIBG and MIBA on the Morphology of Quiescent SMCs. Quiescent SMC were treated with MIBG or MIBA for 72 h. Photomicrographs were used to record cell morphology. Representa- tive micrographs are shown. Specific treatments are: untreated control (A), 50 l M MIBA (B) and MIBG (C), 100 l M MIBA (D) and MIBG (E), 200 l M MIBA (F) and MIBG (G), and 500 l M MIBA (H) and MIBG (I). Magnification, 120 ·. Fig. 1. Sensitivity of mitogen-stimulated SMC growth to MIBG. Qui- escent SMC were pretreated with MIBG for 10 min prior to addition of AngII (1 l M ) (A and B), serum (2% v/v fetal bovine serum) (C and D) or PGE 2 (1 l M ) (E). The incorporation of [ 3 H]uridine or [ 3 H]thymidine into trichloroacetate-precipitable material after addition of AngII and fetal bovine serum was used to monitor RNA and DNA synthesis, respectively. In panel F, [ 3 H]thymidine incorporation by quiescent SMC treated with PGE 2 (1 l M ) was measured after pre- treatment with MIBA rather than MIBG. The incorporation rate of untreated cells was set to 100%. Bars show means + SEM (n ¼ 3) of separate experiments conducted with three different SMC isolations. Statistically significant differences relative to maximally stimulated cells(minusMIBG)areindicated(*P <0.05). 104 L. Yau et al. (Eur. J. Biochem. 270) Ó FEBS 2003 permeabilization techniques stimulate poly(ADP-ribo- syl)ation that masks the less abundant mono-ADP- ribosylation. Numerous labeled proteins were detected in quiescent SMC (Fig. 4), however, both the number of bands and the labeling intensity were increased in growing cells. The most prominent increases were in bands of 140, 133, 116, 95, 82, 73 and 56 kDa (Fig. 4). Other bands also showed differences between the quiescent and growing populations, but these were not as pronounced. Interestingly, a 60-kDa band detected in quiescent cells was absent in the growing cell population. Inclusion of 50 l M MIBG during the incubation period with the growing cells caused a reduction in labeling of the 140, 133, 116, 95, 82 and 73 kDa proteins, while the labeling of several proteins (66, 56 kDa) was unchanged. The insensitivity of the 56 and 66 kDa bands to MIBG can most likely be attributed to modification of amino acids other than arginine [10]. Alternatively, these proteins may have been subjected to another form of post-translational modifi- cation that leads to attachment of labeled adenosine (e.g. adenylylation). Nevertheless, these data establish for the first time that arginine-dependent mono-ADP-ribosy- lation is responsive to mitogens, and that MIBG directly inhibits this process. Fig. 3. Activation of mono-ADP-ribosyltransferase by angiotensin II and prostaglandin E 2 . Quiescent SMC were treated with AngII (1 l M ) or PGE 2 (1 l M ) over 120 min and subcellular fractions prepared as described in Materials and methods at the specified times. (A) Microsomal (membrane) and cytosolic (cytosol) fractions of AngII- stimulated SMC were assayed for arg-mART activity as described in Materials and methods with polyarginine (2 mgÆmL )1 ) as the acceptor molecule. (B) Mono-ADP-ribosylation was measured in vitro with microsomal (membrane) and cytosolic (cytosol) fractions prepared 30minafterAngIIandPGE 2 (conditions equivalent to panel A) treatment. Activity of the untreated cell extracts was set to 1.0. Bars show means + SEM for three independent experiments. Statistically significant differences relative to the respective unstimulated controls (*P < 0.05) are indicated in both panels. Fig. 4. Protein modification with mono-ADP-ribose. Quiescent (Q) and growing (G) SMC were incubated with [ 14 C]adenosine for 16 h as describedinMaterialsandmethods.MIBG(50l M )wasincludedfor the entire incubation period with a growing cell sample (M). Cellular protein was extracted, separated by SDS/PAGE and tagged proteins visualized as described in Materials and methods. Positions of the molecular mass markers are indicated on the right side, while estimated molecular masses of labeled proteins that have changed intensity in relation to growth state are presented on the left. Data are represen- tative of three independent experiments, each of which exhibited similar results. Ó FEBS 2003 ADP-ribosylation and SMC proliferation (Eur. J. Biochem. 270) 105 Participation of poly(ADP-ribosyl)ation in cell proliferation The incorporation of [ 3 H]thymidine was used to compare the effect of a selective inhibitor of poly(ADP-ribosyl)ation, PD128763 [24], on the stimulation of DNA synthesis by mitogen (2% fetal bovine serum). PD128763 was unable to prevent the increase in thymidine incorporation obtained following mitogen stimulation (Fig. 5). 3-Aminobenzamide (3AB), a weak inhibitor of both PARP-1 and mART, was also tested for comparative purposes [24] and was found to reduce thymidine incorporation only slightly at a concen- tration of 5 m M . In contrast, DNA synthesis was inhibited completely in the presence of 50 l M MIBG (Fig. 5). These data suggest that poly(ADP-ribosyl)ation is not required for SMC proliferation. Effect of MIBG on AngII-mediated SMC migration As MIBG prevents the proliferation of SMC, it was also of interest to determine if MIBG could inhibit SMC migration. AngII (10 l M ) is a potent chemoattractant when tested in a Boyden chamber assay (Fig. 6). Inclusion of MIBG (50 l M ) significantly reduced SMC migration relative to AngII alone. These data demonstrate that MIBG inhibits SMC migration in addition to SMC growth. The ineffectiveness of other inhibitors (3AB, PD128763) indicated that poly (ADP-ribosyl)ation is not connected to SMC migration. MIBG prevents induction of c- fos gene expression by mitogens Increased c-fos gene activity is a hallmark of mitogen stimulation, and RT-PCR of total RNA was therefore used to monitor the effect of MIBG on c-fos expression. As expected, both PGE 2 (Fig. 7A) and AngII (data not shown) transiently elevate c-fos mRNA levels within 15 min post- stimulus. Pretreatment of SMC with MIBG, however, prevented the increase in c-fos mRNA levels obtained in response to PGE 2 in a concentration-dependent manner (Fig. 7A). Quantitative analysis of these data confirmed that 50 l M MIBG was sufficient to inhibit PGE 2 -dependent c-fos gene expression by 82%. Similarly, 50 l M MIBG prevented expression of c-myc at 90 min when measured by RT-PCR (Fig. 7B). These observations indicate that a step required for induction of c-fos gene transcription, and consequently c-myc expression, is sensitive to MIBG treatment and may be regulated by an arginine-dependent mART. Inhibition of intracellular signalling by MIBG The induction of c-fos gene expression involves multiple pathways, each activating an essential transcription factor that binds to the c-fos promoter. MAP kinase, for instance, is necessary for phosphorylation of Elk1. Therefore, to determine whether the inhibition of growth effect of MIBG is due to an effect on MAP kinase activation, MIBG (50 l M ) was added to SMC 10 min prior to growth factor stimulation. Under these conditions, MIBG did not prevent PGE 2 -stimulated phosphorylation of myelin basic protein by MAP kinase (Fig. 7C). Rather, there was a slight stimulation in the presence of MIBG for which there is no explanation. Nevertheless, as this increase in MAP kinase activity was observed in all experiments, it may be speculated that MIBG inhibits a negative regulator of this Fig. 5. Effect of ADP-ribosylation inhibitors on DNA synthesis. Qui- escentSMCwerepreparedin24-welldishesandstimulatedwithPGE 2 (1 l M ). ADP-ribosylation inhibitors (5 m M 3-aminobenzamide, 10 l M PD128763, 50 l M MIBG) were added 15 min prior to mitogenic sti- mulation. Thymidine incorporation was monitored as described in Materials and methods. Bars show means + SEM (n ¼ 3). The results were reproducible with three different SMC isolations. Statistically significant differences relative to control (*P < 0.05) and PGE 2 -sti- mulation (+P < 0.05) are indicated. Fig. 6. Inhibition of SMC migration by ADP-ribosylation inhibitors. Growing SMC were placed into a Boyden chamber and directional movement of SMC through a membrane with 5 lmporestowards AngII (1 l M ) was monitored as described in Materials and methods. SMC on the underside of the membrane were fixed, stained and counted. Inhibitors (5 m M 3-aminobenzamide, 10 l M PD128763, 50 l M MIBG) were in the upper chamber. Bars show means + SEM (n ¼ 6). The histogram represents one of two independent experiments using different SMC isolations. Statistically significant differences relative to control (*P < 0.05) and to AngII-stimulated cells in the absence of inhibitor (+P < 0.05) are indicated. 106 L. Yau et al. (Eur. J. Biochem. 270) Ó FEBS 2003 cascade. Like MAP kinase, PI3-kinase is also required for c-fos expression [26], and is activated by both PGE 2 (1 l M ) and thromboxane A 2 (0.1 l M ). However, pretreatment of SMC with MIBG (50 l M ) similarly did not inhibit PGE 2 - dependent formation of PtdInsP 3 by PI3-kinase (Fig. 7D). These results suggest pathways leading to Elk1 are not sensitive to MIBG, and this was confirmed by showing MIBG did not inhibit Elk1 phosphorylation in response to mitogen stimulation (Fig. 7E). In contrast with MAP kinase, Rho mediates growth factor-dependent activation of c-fos via the serum response factor (SRF) [30]. Protein kinase C-related kinase 1 (PRK1), also termed protein kinase N (PKN), and PRK2 have been identified as critical intermediates linking Rho to SRF [31]. Mitogen stimulation of SMC resulted in phosphorylation of both PRK1 and PRK2, which is indicative of their activation (Fig. 7F), while MIBG inhibited this phosphorylation event. These data suggest that the growth inhibitory actions of MIBG result from its ability to prevent Rho-dependent activation of SRF. Discussion The results of this study indicate that an arginine-dependent mART is activated upon stimulation of SMC with growth factors such as serum, AngII and PGE 2 ,andthatMIBG,a selective decoy substrate of arginine-dependent mARTs that competes with the arginine moiety of target proteins [29], prevents SMC proliferation and migration following treat- ment with these growth factors. Furthermore, our novel Fig. 7. Effect of MIBG on prostaglandin E 2 -stimulated gene expression and intracellular signalling pathways. (A) Quiescent SMC were treated for 15minwithPGE 2 (1 l M )subsequenttoa10-minpretreatmentwithMIBG(50l M ). RNA was extracted and c-fos mRNA levels assessed by RT- PCR as described in Materials and methods. Molecular mass markers (DNA Marker VI) were used to confirm the size of the PCR products. RNA loading was confirmed by concurrent amplification of GAPDH. One representative agarose gel is shown. (B) Quiescent SMC were treated as described in panel A, and total RNA was harvested at 90 min and RT-PCR amplification was used to determine c-myc expression. RNA loading was confirmed by concurrent amplification of GAPDH. One representative agarose gel is shown. The results presented in panels A and B were confirmed in two independent experiments using different SMC isolations. (C) MAP kinase activity was measured by activity gel assay over 20 min after treatment with PGE 2 (1 l M ). Cells were pretreated with MIBG (50 l M ) for 10 min prior to PGE 2 stimulation. Specific phosphorylation of myelin basic protein by p42 MAPK (p42) and p44 MAPK (p44) is shown. One of three independent experiments with different SMC isolations is presented. (D) Phosphate pools in quiescent SMC were labelled with 200 lCi [ 32 P]orthophosphate for 4 h prior to treatment with MIBG (50 l M ) for 10 min followed by addition of PGE 2 (1 l M )orTxA 2 (0.1 l M ) for 15 min. Phosphoinositides were extracted from the cells and the phos- phorylated forms of phosphoinositol were resolved by thin layer chromatography (TLC) as described in Materials and methods. A representative autoradiogram of a TLC plate is shown, with PtdInsP 1 ,PtdInsP 2 and PtdInsP 3 indicated. These results were confirmed in two independent experiments. (E) Quiescent SMC were treated for 15 min with PGE 2 (1 l M )subsequenttoa10-minpretreatmentwithMIBG(50l M ). Proteins were extracted and analyzed by Western blotting. Detection of phosphorylated Elk1 is shown. One of three independent experiments with different SMC isolations is presented. (F) Phosphorylation of PRK1/2 was measured by Western blotting as described for panel E. Three independent experiments employing different SMC isolations were tested, and one result is presented in this panel. Ó FEBS 2003 ADP-ribosylation and SMC proliferation (Eur. J. Biochem. 270) 107 findings suggest that mono-ADP-ribosylation is essential for the induction of c-fos, and subsequently c-myc,gene expression in response to certain mitogenic stimuli, possibly via a Rho-dependent process. These observations therefore implicate arginine-dependent mono-ADP-ribosylation in the transduction of signals from cell surface receptors to the nucleus. In addition to establishing a role for mono-ADP- ribosylation in SMC proliferation and migration, our data also confirm that poly(ADP-ribosyl)ation does not partici- pate in these events. While consistent with our study of insulin-dependent proliferation of H4IIE hepatoma cells [24], this conclusion does not concur with Thyberg et al.[20] who found poly(ADP-ribosyl)ation is necessary for SMC growth. Their reasoning was based on evidence that HMBA (hexamethylenebisacetamide) blocked the transition of rat aortic SMC from a contractile to a synthetic phenotype, in addition to inhibiting DNA synthesis, protein synthesis and the expression of several genes. Although inhibition of SMC proliferation by HMBA has been reported in two inde- pendent studies [20,21], the relationship between these actions and the inhibition of PARP-1 remains tenuous. In fact, numerous other intracellular targets for HMBA have been proposed [32], including SMC-derived TGF-a [33]. As well, the inability of 3AB to effectively inhibit cellular proliferation [29,34], as we have shown for SMC (Fig. 5), further supports the inference that poly(ADP-ribosyl)ation does not participate in this process. Finally, the ineffective- ness of PD128763 (Fig. 5), a specific PARP-1 inhibitor at the concentrations employed in this study [24], shows convincingly that poly(ADP-ribosyl)ation has no role in either SMC proliferation or migration. Uncoupling poly(ADP-ribosyl)ation from SMC prolifer- ation was anticipated given the results of our previous study [24]. Similarly, as MIBG can prevent the proliferation of multiple cell types [35], its effect on SMC was also expected. However, the mechanism by which MIBG inhibits cell growth has yet to be resolved, although it has been ostensibly attributed to an inhibition of either mitochondrial respiration or mono-ADP-ribosylation [35,36]. In this study, we have shown that MIBG blocks an early step leading to expression of the c-fos and c-myc genes (Fig 7A and B). This novel observation is clearly inconsistent with a general effect on mitochondrial activity. Strong support for this view is also provided by the fact that growth inhibition is dissociated from cell death, as MIBG-induced changes in SMC morphology are not evident at concentrations below 500 l M (Fig. 2). The most significant outcome of this study was identify- ing a link between MIBG and c-fos gene expression, and therefore implies ADP-ribosylation regulates induction of the c-fos gene. Although this finding contrasts with the observation of Thyberg et al. [20] who reported that MIBG, at a concentration of 100 l M , did not inhibit induction of either c-fos or c-myc gene expression, this discrepancy may be correlated with the limited potency of MIBG (e.g. DNA synthesis was inhibited by only 50%) under their experi- mental conditions. Furthermore, the distinct methods used to prepare the cell cultures (collagenase digestion vs. explant culture), as well as the time in reduced serum media (48 h vs. 5 days) intended to produce a quiescent state, may contribute to differences in their response to mitogenic stimuli. In any case, the data presented in Fig. 1 indicate the concentration of MIBG required to inhibit SMC prolifer- ation is in agreement with the earlier studies that first reported MIBG inhibits cell proliferation [35,36]. Two potential routes by which MIBG could influence c-fos gene expression were examined. Their selection was based on information that each pathway is known to activate one of the key transcription factors that regulates c-fos promoter activity. MAP kinase and PI3-kinase were investigated because it has been established they couple immediate early gene induction to mitogenic signals gener- ated by both AngII and PGE 2 [26]. MAP kinase operates through Elk1, one component of the transcriptional machinery required for c-fos gene activity [37]. The lack of inhibition of MAP kinase activation (Fig. 7C) and Elk1 phosphorylation (Fig. 7E) by MIBG, however, suggests this pathway is independent of ADP-ribosylation. Expression of the c-fos gene also requires SRF, which activates c-fos gene transcription via the serum response element located at ) 300 relative to the site of transcription initiation [38]. Inhibition of PRK1/2 phosphorylation by MIBG could imply that SRF activation is ADP-ribosylation-dependent. This connection is based on published evidence linking PRK1/2 with SRF [31]. Furthermore, the fact that PRK1/2 is a downstream effector of Rho suggests ADP-ribosylation may contribute to the regulation of Rho-dependent signal- ling [39]. Accordingly, an association between Rho and ADP-ribosylation has been established previously [15]. In addition, inhibition of Rho by MIBG could explain the diverse effects of this compound on both cell proliferation and migration, as Rho is necessary for both processes [40]. Nevertheless, the exact target that mediates the presumed mono-ADP-ribosyltransferase function, however, remains to be identified. However, given the ability of ADP- ribosylation to regulate GTP/GDP-binding status of numerous proteins [41], it is possible to speculate that one of the GTP-binding proteins responsible for Rho activation (e.g. G 12/13 -dependent activation of p115 RhoGEF [42]), or possibly Rho itself, is the most likely target. It is also noteworthy that inhibition of Rho activation influences cytoskeletal organization and thus cell motility. Indeed, the GTP-binding proteins Rho, Rac and cdc42, which regulate the cytoskeletal rearrangements necessary for migration, are known targets of toxin-mediated mono- ADP-ribosylation [41]. On the other hand, other possible targets should also be considered, including actin, desmin and tubulin, as mono-ADP-ribosylation has been shown to influence their assembly/disassembly [43–46]. Further investigation will be necessary to identify the mechanism by which MIBG inhibits cell migration. This investigation provides the first evidence to directly link mono-ADP-ribosylation with SMC proliferation and migration. Furthermore, expression of the c-fos gene was the earliest proliferative event exhibiting sensitivity to MIBG treatment, and represents a novel mechanism by which mono-ADP-ribosylation can influence cellular processes. As it is known that FOS mediates c-myc gene activation, the inhibition of c-myc expression by MIBG supports the premise that SRF-dependent c-fos gene expression is the target of this inhibitor. Thus, c-fos gene expression and PRK1/2 phosphorylation can now be employed as endpoints for identification of those 108 L. Yau et al. (Eur. J. Biochem. 270) Ó FEBS 2003 cellular events that are susceptible to inhibition by MIBG and should permit further examination of the role of mono-ADP-ribosylation in cell cycle control and cell proliferation. Acknowledgements This study was supported by grants from the Natural Sciences and Engineering Research Council and by funding to the CIHR Group in Experimental Cardiology. L. Y. was supported by a studentship from the St Boniface Research Foundation. References 1. Ueda, K. & Hayaishi, O. (1985) ADP-ribosylation. Annual Rev. Biochem. 54, 73–100. 2. Zahradka, P. & Ebisuzaki, K. (1982) A shuttle mechanism for DNA–protein interactions. The regulation of poly (ADP-ribose) polymerase. Eur. J. Biochem. 127, 579–585. 3. Lindahl, T., Satoh, M.S., Poirier, G.G. & Klungland, A. (1995) Post-translational modification of poly (ADP-ribose) polymerase induced by DNA strand breaks. Trends Biochem. Sci. 20, 405–411. 4. Duriez, P.J. & Shah, G.M. (1997) Cleavage of poly(ADP-ribose) polymerase: a sensitive parameter to study cell death. Biochem. Cell Biol. 75, 337–349. 5. Bhatia, M., Kirkland, J.B. & Meckling-Gill, K.A. (1996) Over- expression of poly (ADP-ribose) polymerase promotes cell cycle arrest and inhibits neutrophilic differentiation of NB4 acute pro- myelocytic leukemia cells. Cell Growth Differ. 7, 91–100. 6. Poirier, G.G., de Murcia, G., Jongstra-Bilen, J., Niedergang, C. & Mandel, P. (1982) Poly (ADP-ribosyl) ation of polynucleosomes causes relaxation of chromatin structure. Proc. Natl Acad. Sci. USA 79, 3423–3427. 7. Saxena, A., Saffery, R., Wong, L.H., Kalitsis, P. & Choo, K.H. (2002) Centromere proteins Cenpa, Cenpb, and Bub3 interact with PARP-1 protein and are poly(ADP-ribosyl)ated. J. Biol. Chem. 277, 26921–26926. 8. Meisterernst, M., Stelzer, G. & Roeder, R.G. (1997) Poly (ADP- ribose) polymerase enhances activator-dependent transcription in vitro. Proc. Natl Acad. Sci. USA 94, 2261–2265. 9. Soldatenkov, V.A., Chasovskikh, S., Potaman, V.N., Trofimova, I.,Smulson,M.E.&Dritschilo,A.(2002)Transcriptional repression by binding of poly(ADP-ribose) polymerase to pro- moter sequences. J. Biol. Chem. 277, 665–670. 10. Okazaki, I.J. & Moss, J. (1996) Mono-ADP-ribosylation: a reversible posttranslational modification of proteins. Adv. Phar- macol. 35, 247–280. 11. Okazaki, I.J. & Moss, J. (1999) Characterization of glycosylpho- sphatidylinositiol-anchored, secreted, and intracellular vertebrate mono-ADP-ribosyltransferases. Annual Rev. Nutr. 19, 485–509. 12. Terashima, M., Mishima, K., Yamada, K., Tsuchiya, M., Wakutani, T. & Shimoyama, M. (1992) ADP-ribosylation of actins by arginine-specific ADP-ribosyltransferase purified from chicken heterophils. Eur. J. Biochem. 204, 305–311. 13. Li, P.L., Chen, C.L., Bortell, R. & Campbell, W.B. (1999) 11,12- Epoxyeicosatrienoic acid stimulates endogenous mono-ADP- ribosylation in bovine coronary arterial smooth muscle. Circ. Res. 85, 349–356. 14. Tanuma, S., Kawashima, K. & Endo, H. (1988) Eukaryotic mono (ADP-ribosyl) transferase that ADP-ribosylates GTP-binding regulatory G i protein. J. Biol. Chem. 263, 5485–5489. 15. Maehama, T., Sekine, N., Nishina, H., Takahashi, K. & Katada, T. (1994) Characterization of botulinum C3-catalyzed ADP-ribosylation of rho proteins and identification of mammalian C3-like ADP-ribosyltransferase. Mol. Cell. Biochem. 138, 135–140. 16. Schwartz, S.M., Campbell, G.R. & Campbell, J.H. (1986) Replication of smooth muscle cells in vascular disease. Circ. Res. 58, 427–444. 17. Owens, G.K. (1995) Regulation of differentiation of vascular smooth muscle cells. Physiol. Rev. 75, 487–517. 18. Kocher, O., Gabbiani, F., Gabbiani, G., Reidy, M.A., Cokay, M.S., Peters, H. & Huttner, I. (1991) Phenotypic features of smooth muscle cells during the evolution of experimental carotid artery intimal thickening. Biochemical and morphologic studies. Laboratory Invest. 65, 459–470. 19. Thyberg, J., Blomgren, K., Roy, J., Tran, P.K. & Hedin, U. (1997) Phenotypic modulation of smooth muscle cells after arterial injury is associated with changes in the distribution of laminin and fibronectin. J. Histochem. Cytochem. 45, 837–846. 20. Thyberg, J., Hultgardh-Nilsson, A. & Kallin, B. (1995) Inhibitors of ADP-ribosylation suppress phenotypic modulation and pro- liferation of smooth muscle cells cultured from rat aorta. Differ- entiation 59, 243–252. 21. Grainger, D.J., Hesketh, T.R., Weissberg, P.L. & Metcalfe, J.C. (1992) Hexamethylenebisacetamide selectively inhibits the pro- liferation of human and rat vascular smooth-muscle cells. Bio- chem. J. 283, 403–408. 22. Saward, L. & Zahradka, P. (1997) Coronary artery smooth muscle in culture: migration of heterogeneous cell populations from vessel wall. Mol. Cell. Biochem. 176, 53–59. 23. Zahradka, P., Elliot, T., Hovland, K., Larson, D.E. & Saward, L. (1993) Repression of histone gene transcription in quiescent 3T6 fibroblasts. Eur. J. Biochem. 217, 683–690. 24. Yau, L., Elliot, T., Lalonde, C. & Zahradka, P. (1998) Repression of phosphoenolpyruvate carboxykinase gene activity by insulin is blocked by 3-aminobenzamide but not by PD128763, a selective inhibitor of poly (ADP-ribose) polymerase. Eur. J. Biochem. 253, 91–100. 25. Yau, L. & Zahradka, P. (1997) Immunodetection of activated mitogen-activated protein kinase in vascular tissues. Mol. Cell. Biochem. 172, 59–66. 26. Saward, L. & Zahradka, P. (1997) Angiotensin II activates phosphatidylinositol 3-kinase in vascular smooth muscle cells. Circ. Res. 81, 249–257. 27. Inageda, K., Nishina, H. & Tanuma, S. (1991) Mono-ADP- ribosylation of G s by an eukaryotic arginine-specific ADP-ribo- syltransferase stimulates the adenylate cyclase system. Biochem. Biophys. Res. Commun. 176, 1014–1019. 28. Zahradka, P., Harris, K.D., Triggs-Raine, B., Lamontagne, G. & Leblanc, N. (1995) PCR-based analysis of voltage-gated K + channels in vascular smooth muscle. Mol. Cell. Biochem. 145, 39–44. 29. Smets, L.A., Loesberg, C., Janssen, M. & van Rooij, H. (1990) Intracellular inhibition of mono (ADP-ribosylation) by meta- iodobenzylguanidine: specificity, intracellular concentration and effects on glucocorticoid-mediated cell lysis. Biochim. Biophys. Acta 1054, 49–55. 30. Hill, C.S., Wynne, J. & Treisman, R. (1995) The Rho family GTPases RhoA, Rac1, and CDC42Hs regulate transcriptional activation by SRF. Cell 81, 1159–1170. 31. Morissette, M.R., Sah, V.P., Glembotski, C.C. & Brown, J.H. (2000) The Rho effector, PKN, regulates ANF gene transcription in cardiomyocytes through a serum response element. Am. J. Physiol. Heart Circ. Physiol. 278, H1769–H1774. 32. Marks, P.A., Richon, V.M., Kiyokawa, H. & Rifkind, R.A. (1994) Inducing differentiation of transformed cells with hybrid polar compounds: a cell cycle-dependent process. Proc. Natl. Acad. Sci. USA 91, 10251–10254. Ó FEBS 2003 ADP-ribosylation and SMC proliferation (Eur. J. Biochem. 270) 109 33. Weissberg, P.L., Grainger, D.J., Shanahan, C.M. & Metcalfe, J.C. (1993) Approaches to the development of selective inhibitors of vascular smooth muscle cell proliferation. Cardiovasc. Res. 27, 1191–1198. 34. Beckmann, J.D., Illig, M., Romberger, D. & Rennard, S.I. (1992) Induction of fibronectin gene expression by transforming growth factor beta-1 is attenuated in bronchial epithelial cells by ADP- ribosyltransferase inhibitors. J. Cell Physiol. 152, 274–280. 35. Smets, L.A., Bout, B. & Wisse, J. (1988) Cytotoxic and antitumor effects of the norepinephrine analogue meta-iodo-benzylguanidine (MIBG). Cancer Chemother. Pharmacol. 21, 9–13. 36. Loesberg, C., van Rooij, H., Nooijen, W.J., Meijer, A.J. & Smets, L.A. (1990) Impaired mitochondrial respiration and stimulated glycolysis by m-iodobenzylguanidine (MIBG). Int. J. Cancer 46, 276–281. 37. Treisman, R. (1996) Regulation of transcription by MAP kinase cascades. Curr. Opin. Cell Biol. 8, 205–215. 38. Treisman, R., Marais, R. & Wynne, J. (1992) Spatial flexibility in ternary complexes between SRF and its accessory proteins. EMBO J. 11, 4631–4640. 39. Amano, M., Mukai, H., Ono, Y., Chihara, K., Matsui, T., Hamajima, Y., Okawa, K., Iwamatsu, A. & Kaibuchi, K. (1996) Identification of a putative target for Rho as the serine-threonine kinase protein kinase N. Science 271, 648–650. 40. Seasholtz, T.M., Majumdar, M., Kaplan, D.D. & Brown, J.H. (1999) Rho and Rho kinase mediate thrombin-stimulated vascular smooth muscle cell DNA synthesis and migration. Circ. Res. 84, 1186–1193. 41. Moss, J. & Vaughan, M. (1988) ADP-ribosylation of guanyl nucleotide-binding regulatory proteins by bacterial toxins. Adv. Enzymol. Relat. Areas Mol. Biol. 61, 303–379. 42. Gohla, A., Offermanns, S., Wilkie, T.M. & Schultz, G. (1999) Differential involvement of Galpha12 and Galpha13 in receptor- mediated stress fiber formation. J. Biol. Chem. 274, 17901–17907. 43. Clancy, R., Leszczynska, J., Amin, A., Levartovsky, D. & Abramson, S.B. (1995) Nitric oxide stimulates ADP ribosylation of actin in association with the inhibition of actin polymerization in human neutrophils. J. Leukoc. Biol. 58, 196–202. 44. Terashima, M., Yamamori, C. & Shimoyama, M. (1995) ADP- ribosylation of Arg28 and Arg206 on the actin molecule by chicken arginine-specific ADP-ribosyltransferase. Eur. J. Biochem. 231, 242–249. 45. Yuan, J., Huiatt, T.W., Liao, C.X., Robson, R.M. & Graves, D.J. (1999) The effects of mono-ADP-ribosylation on desmin assem- bly-disassembly. Arch. Biochem. Biophys. 363, 314–322. 46. Scaife, R.M., Wilson, L. & Purich, D.L. (1992) Microtubule protein ADP-ribosylation in vitro leads to assembly inhibition and rapid depolymerization. Biochemistry 31, 310–316. 110 L. Yau et al. (Eur. J. Biochem. 270) Ó FEBS 2003 . Endogenous mono-ADP-ribosylation mediates smooth muscle cell proliferation and migration via protein kinase N-dependent induction of c- fos expression Lorraine. of these proteins. MIBG also blocked the stimulation of DNA and RNA synthesis, prevented smooth muscle cell migration and suppressed the induction of c-fos and