Gene Therapy (1999) 6, 1184–1191  1999 Stockton Press All rights reserved 0969-7128/99 $12.00 http://www.stockton-press.co.uk/gt

BRIEF COMMUNICATION

Tissue-selective expression of dominant-negative proteins for the regulation of vascular smooth muscle cell proliferation JF Schmitt1, M-C Keogh1, U Dennehy1, D Chen1, F Lupu1, K Weston2, D Taylor2, VV Kakkar1 and NR Lemoine1,3 1

Thrombosis Research Institute, 3Imperial Cancer Research Fund, Molecular Oncology Unit, Imperial College School of Medicine; and 2CRC Centre for Cell and Molecular Biology, Institute of Cancer Research, London, UK

The transcription factors c-myb and c-myc are essential for vascular smooth muscle cell (VSMC) replication and are rapidly induced following mitogenic stimulation of quiescent VSMCs in vitro and in vivo following balloon catheter injury. Consequently, interference with c-myb and c-myc function provides a possible avenue for the prevention of VSMC proliferation associated with intimal hyperplasia. We have carried out studies focused on the inhibition of VSMC proliferation using dominant-negative gene constructs incorporating the DNA-binding domains of the c-myb or

c-myc genes fused to the repressor domain of the Drosophila engrailed gene. Transient transfection of rat, rabbit and human vascular SMCs results in a dramatic inhibition of proliferation for at least 72 h after transfection. Furthermore, this inhibition of cellular proliferation was found to be due, at least in part, to the induction of apoptosis. Coupling expression of the chimeric dominant-negative proteins to transcriptional regulatory elements of the human vascular smooth muscle ␣-actin gene allows specific targeting of vascular smooth muscle cells.

Keywords: 4-hydroxytamoxifen; dominant-negative; engrailed; human VSMC ␣-actin; myb; myc; tissue-specific expression

Excellent short-term results have been achieved by the use of percutaneous transluminal coronary angioplasty (PTCA) for the treatment of arterial narrowing (stenosis) resulting from arterial disease such as acute atherosclerosis. However, repeat PTCA is often necessary as reocclusion of the damaged area or restenosis occurs in 30– 40% of cases.1,2 Exactly how restenosis occurs is yet to be clearly defined although aberrant smooth muscle cell proliferation has been implicated. It has been proposed that exposure of the procoagulant intimal and medial smooth muscle cell layers to the bloodstream by the removal of the endothelial cell layer during PTCA results in the initiation of the coagulation cascade and the accumulation of mitogenic products in the damaged area. These stimulate the migration and proliferation of vascular smooth muscle cells (VSMC) and the synthesis of extracellular matrix, potentially leading to re-occlusion of the injured artery.3,4 In accordance with the above model, recent studies aimed at preventing restenosis have focused on gene therapy approaches for the transient control of smooth muscle cell proliferation, although the total contribution of VSMC proliferation to the developing restenotic lesion is far from accepted.5–7 Correspondence: M-C Keogh, Department of Biological Chemistry and Molecular Pharmacology, Bldg C1, Rm 207, Harvard Medical School, 240 Longwood Avenue, Boston MA 02115, USA The first two authors contributed equally to this work Received 5 November 1998; accepted 5 February 1999

The c-myb and c-myc transcription factors play a central role in the cell cycle and have been found to be essential for VSMC replication. Both are rapidly induced following mitogenic stimulation of quiescent VSMCs in vitro and balloon catheter injury in vivo.8–11 It is now recognized that c-myc and c-myb are required to protect cells from entering the programmed cell death pathway (apoptosis) when stimulated by growth factors and blocking their expression can cause the cellular fail-safe pathways to be circumvented.12–15 Consequently, down-regulation of cmyc or c-myb provides a possible avenue for interference with VSMC proliferation associated with restenosis. A number of gene therapy approaches for the treatment of restenosis have employed antisense strategies to inhibit VSMC proliferation. Antisense oligonucleotides targeted against c-myc and c-myb have been reported to result in reduced c-myc and c-myb protein levels and inhibition of proliferation of human and rat VSMCs in vitro.16–19 Similarly, in vivo transfection studies utilizing antisense c-myc and c-myb oligonucleotides have reported reduced c-myc or c-myb expression and reduced neointimal thickening following vascular injury in a number of animal models.16,20–22 However, therapeutic application of antisense oligonucleotides is hampered by cellular uptake and stability, as well as by their lack of cell- and possibly target mRNA-specificity.23–26 Consequently, antisense gene expression vectors containing the c-myc cDNA have been created and shown to reduce cmyc expression and cellular proliferation in a number of cell types.27–29

Transcriptionally targeted dominant-negative gene constructs JF Schmitt et al

An alternative approach to antisense strategies is to block normal protein function through the employment of competitive inhibitors or dominant negative proteins such as H-ras30,31 or constitutively active retinoblastoma (b)32 mutants and indeed both have demonstrated their ability to inhibit neointimal formation in animal models. Dominant-negative c-myb expression constructs incorporating the c-myb DNA binding domain alone or coupled to the Drosophila engrailed repressor gene have allowed examination of c-myb function in T cell development.33 Similar deletion or chimeric c-myb constructs have also been used to examine the role of c-myb in cell-cycle progression and the regulation of cytoplasmic Ca2+ concentration in rat VSMC cell lines.33,34 The dominant-negative chimeric myb-engrailed (MEnT, MybEn, MERT) and myc-engrailed (MycEn) constructs used in this study were designed on the basis that the activity of c-myb and c-myc proteins is dependent on the integrity of their trans-activation and DNA-binding domains. As described previously,33 the dominantnegative chimeric myb-engrailed gene constructs (MybEn

and MEnT) incorporate the DNA-binding domain of the murine c-myb gene (aa 84–213) linked to the repressor domain of the Drosophilia engrailed protein (aa 2–298).35 Design of the myc-engrailed construct was slightly different as the functional activity of c-myc requires dimerisation with Max.36,37 Dimerisation is mediated by the basic helix–loop–helix leucine zipper (bHLH-LZ) domain at the carboxy-terminus of the protein. Hence the dominantnegative chimeric myc-engrailed construct includes the bHLH-LZ domain of human c-myc (aa 262–439) linked to the repressor domain of Drosophilia engrailed (aa 2–298) (Figure 1a). Deletion derivatives of myc which retain the bHLH-LZ, but not the amino-terminal transactivating domain, form biologically inactive DNA-binding dimers with Max and function as dominant negative suppressers of Myc transforming activity.37,38 It was postulated that the additional presence of the engrailed repressor domain would have a more pronounced effect than simple deletion mutants.33 The two main expression plasmids utilised in this study are the pCI/h999 expression vector (Figure 1b) and

Figure 1 Schematic of dominant-negative chimeric proteins and expression vectors. (a) Schematic depicting the component elements of the chimeric proteins utilized in these experiments and their origin. MEnT consists of the c-myb DNA binding domains fused to the negative regulatory domain of the Drosophila homeobox engrailed gene as described.33 The MERT fusion protein was obtained by coupling of a mutant estrogen receptor gene sequence to the myb-engrailed segment of MEnT as described.13 For this study MEnT was subcloned into the pCINeo expression vector (Promega, Southampton, UK). The MybEn chimera was amplified from MEnT with PCR primers to remove the 9E10 myc-epitope tag. The MycEn chimera was assembled as follows: the engrailed gene sequence from the MybEn chimera was PCR amplified with an ATG start codon incorporated into the forward primer. This was cloned 5⬘ of the c-myc DNA binding/dimerization domain (amino acids 262–440) in the BsIIKS+ vector (Stratagene, Amsterdam, The Netherlands) before subcloning into pCI or pCI/h999. (b) Graphic representation of the muscle-specific expression vector pCI/h999 showing the location of functional elements and restriction enzyme sites available for cloning in the plasmid polylinker. The vector backbone is based on the commercially available vector pCI (Promega). An in-depth description of the vector has been published.39 Cloning of genes in cis into the polylinker of this vector confers a musclespecific expression pattern. Amp, ampicillinr; ER, estrogen receptor; HBD, hormone binding domain; MCS, multiple cloning site.

1185

Transcriptionally targeted dominant-negative gene constructs JF Schmitt et al

1186

its parent pCI (Promega), both of which have been described previously.39 The inclusion of human VSMC ␣actin regulatory elements in the pCI/h999 expression vector (as opposed to the CMV-containing pCI) allows the specific-targeting of VSMCs in vitro and we predict in vivo. The pCI/h999 vector has been tested in vitro in cells from a number of lineages including muscle (skeletal, cardiac and vascular smooth), endothelium (HUVEC, BAEC, PAEC), hepatocytes, fibroblasts and T cells. The vector is active in all muscle types tested, although relatively more active in VSMCs. Since the vector is intended for delivery and use in a specific anatomical location (a site of intimal hyperplasia) rather than systemically, general expression in muscle does not detract from its potential usefulness. Of the cells comprising this lesion type (VSMCs, ECs, fibroblasts, T cells and macrophages), only VSMCs will be expected to support expression. This cellular-specificity is not possible with

the utilization of antisense oligonucleotides or viral promoters. This work describes a dominant-negative gene approach to the regulation of VSMC proliferation using chimeric myb-engrailed (MEnT, MybEn, MERT) and mycengrailed (MycEn) gene constructs (Figure 1a). Two of the myb-engrailed constructs utilised have been described previously: the tamoxifen-responsive, 9E10 epitopetagged MERT and the 9E10 epitope-tagged MEnT.13,33 An additional myb-engrailed chimera (MybEn) and a novel myc-engrailed chimera (MycEn) were created and cloned in the expression vectors pCI and pCI/h999 (Figure 1b).39 The sequence of the MybEn and MycEn chimeras was confirmed by DNA sequencing (data not shown). Following transfection into mammalian cells it is expected that the pCI vector will support expression of genes linked in cis in all cell types, while the genes will be expressed only in muscle cells when the pCI/h999 vector is used.39 To

Figure 2 Expression of the chimeric proteins MycEn and MybEn. (a) Empty pCI or luciferase, MycEn and MybEn in the pCI or pCI/h999 expression vectors were transfected into rabbit VSMCs or the HUVEC cell line ECV304 as previously described39,40 and total RNA harvested 36 h later from approximately 2 × 106 cells for Northern analysis using the RNeasy mini kit (Qiagen, Surrey, UK) according to the manufacturer’s instructions. The RNA eluent was precipitated, resuspended in 10 ␮l formamide and electrophoresed on a 1.1% agarose denaturing formaldehyde gel. Following visualization of marker tracks, the gel was pressure blotted on to Hybond-N membrane (Amersham Pharmacia Biotech, Buckinghamshire, UK) with 20 × SSC, baked for 2 h at 80°C to fix and probed at 65°C in 20 ml Rapid-Hyb buffer (Amersham Pharmacia Biotech) for 4 h with the random primed DNA-probe indicated (50 ng of random primed S26 or engrailed (En)). Following probing the membrane was sequentially washed at 65°C for 20 min with (2 × SSC, 0.1% SDS), (1 × SSC, 0.1% SDS), (0.5 × SSC, 0.1% SDS) and (0.1 × SSC, 0.1% SDS) as indicated and exposed to Kodak Biomax MS X-ray film (Amersham Pharmacia Biotech). S26, final wash 65°C for 20 min with (0.1 × SSC, 0.1% SDS), overnight exposure −70°C; En, final wash 65°C for 20 min with (0.5 × SSC, 0.1% SDS), 96 h exposure −70°C. (b) The luciferase, MycEn and MybEn genes in the pCI or pCI/h999 expression vectors were subjected to in vitro transcription/translation as described. In vitro transcription of the sense-strand RNA was performed using the Ribomax T7 kit (Promega) according to the manufacturer’s instructions. RNA transcripts derived were translated using the Flexi-rabbit reticulocyte lysate system (Promega) according to the manufacturer’s instructions. The inclusion of 35S-methionine (1200 Ci/mmol at 10 mCi/ml, ICN Pharmaceuticals, Hampshire, UK) allowed the visualization of the translated products resolved by reducing SDS-PAGE and autoradiography. In each case, a major protein species of the expected size as deduced from amino acid sequence is visible (luciferase, MycEn and MybEn proteins are 63.8, 52.5 and 50.2 kDa, respectively). Negative control empty vector pCI is also shown. The position of each species of molecular weight markers is shown in the left margin.

Transcriptionally targeted dominant-negative gene constructs JF Schmitt et al

confirm this expression pattern, the MycEn and MybEn constructs in the pCI or pCI/h999 vectors were transfected into rabbit VSMCs or the HUVEC cell line ECV304 as described39,40 and total RNA harvested 36 h later for Northern analysis. The resulting membranes were probed for engrailed or the endogenous ribosomal pro-

tein S2641 with the expected product sizes of S26, MycEn and MybEn mRNAs being 0.44, 1.43 and 1.35 kb, respectively. The expression pattern observed is as expected: transcription of MycEn and MybEn in pCI is supported in both cell types, but only in rabbit VSMCs when the pCI/h999 expression vector is used (Figure 2a).

Figure 3 Nuclear translocation of the MERT chimeric protein reduces rat VSMC proliferation and induces apoptosis. (a) Rat VSMCs were transfected with the MERT construct and a stable integrated population (MERT+ Rat VSMC) selected by the addition of G418 (0.5 mg/ml) 48 h after transfection. MERT+ and non-transfected rat VSMCs were plated in triplicate, grown to approximately 50% confluence and 0–100 nm 4-OH-TAM added. Proliferation was then analyzed by tritiated thymidine ({methyl-3H}TdR) (Amersham Pharmacia Biotech) incorporation at the indicated intervals over the next 72 h. At the time-points indicated, each well was pulsed with 2 mCi/ml tritiated thymidine for 4 h and the cells then rinsed twice with PBS and three times with 5% trichloroacetic acid (TCA) in PBS. Following this the wells were rinsed with 70% ethanol, air dried and stored at −20°C until analysis. For scintillation counting, the cell monolayers were suspended by incubation in 400 ␮l 0.25 m NaOH at 37°C for 1 h with gentle shaking. Samples were then transferred to scintillation vials, 4 ml of Optiphase Hisafe scintillation cocktail (Wallac, Milton Keynes, UK) was added and the vials assayed. Results are expressed as mean counts per minute (c.p.m.) of the triplicate samples with the standard error of the mean. (b) The number of apoptotic nuclei was measured by TUNEL at the time-points indicated (in hours). Non-transfected and MERT+ rat VSMCs were seeded on glass coverslips overnight, 4-OH-TAM added to 100 nm and coverslips removed from culture at the indicated time-points. Coverslips were washed in PBS, fixed for 10 min in 4% PFA in PBS and stored at −20°C under methanol until assayed. For the TUNEL procedure, monolayers were washed with PBS and labeled for 1 h at 37°C using terminal deoxytransferase (TdT, Boehringer Mannheim, East Sussex, UK) in TdT buffer with 6 mm FITC-12-dUTP, 60 mm dATP and 1 mm CoCI2. Unincorporated label was removed by PBS washing and the cells mounted in Vecta shield (Vector Laboratories, Peterborough, UK) containing DAPI for fluorescence microscopy. Microscopic determination of apoptotic and normal nuclei was determined by counting under excitation with a fluorescein filter (␭288 nm). Normal nuclei stain blue only, while apoptotic nuclei show co-localization of blue DAPI and green FITC. The mean percentage of apoptotic nuclei relative to normal nuclei and standard error of the mean were determined after counting cells in a minimum of eight microscopic fields. Paired Student’s t tests were applied where indicated with the aid of the Sigmaplot program (SPSS UK Ltd, Surrey, UK) and significance (*) was determined as a probability that two means indicated were different at the P ⬍ 0.05 level. 4-OH-TAM, 4hydroxytamoxifen; CPM, counts per minute; hrs, hours; VSMC, vascular smooth muscle cell.

1187

Transcriptionally targeted dominant-negative gene constructs JF Schmitt et al

1188

To demonstrate frame fidelity of each clone, the novel MybEn and MycEn chimeras in the pCI and pCI/h999 expression vectors were subjected to in vitro transcription/translation. The enzymatic activity of the luciferase gene transcribed/translated from pCI Luc or pCI/h999 Luc was used as a positive control for the method (data not shown). The translation products observed are shown (Figure 2b) and their respective sizes correspond to the production of the appropriate fulllength proteins. The construction of the myb-engrailed-tamoxifen responsive chimera (MERT) has been described previously.13,33 A stably transfected rat VSMC MERT+ line was created for initial studies of the effect of mybengrailed chimeric proteins on a relatively homogeneous population. The addition of 4-hydroxytamoxifen (4-OHTAM) to these stable integrants is assumed to cause translocation of the MERT+ protein from the cytoplasm to the nucleus of expressing cells to exert its effects.42 Stable MERT+ rat VSMC and untransfected rat VSMC were cultured in the presence of 0–100 nm 4-OH-TAM for 0–72 h. Analysis of cell proliferation by incorporation of tritiated thymidine revealed that the MERT+ cells showed dosedependent 4-OH-TAM inhibition of cell proliferation within 48 h of addition. After 4 days culture in the presence of 100 nm 4-OH-TAM the rate of thymidine incorporation of the MERT+ transfectants was only 30% of that observed for MERT+ cells grown in the absence of 4-OHTAM. In contrast, proliferation of untransfected rat smooth muscle cells was unaffected by 4-OH-TAM addition (Figure 3a). The TUNEL procedure was used to determine whether the inhibition of proliferation observed in rat VSMC MERT+ transfectants treated with 4-OH TAM was a consequence of apoptosis. Apoptotic nuclei could be observed as soon as 4 h after the addition of 100 nm 4OH-TAM to the MERT+ population and their proportion relative to normal nuclei increased steadily (and significantly at each time-point (P ⬍ 0.05)) to represent 20% of the population by 16 h (Figure 3b). In contrast, little change in the level of apoptosis was evident over the 16 h when untransfected cells were grown in the presence of 100 nm 4-OH-TAM. Although stable integrants such as the MERT+ rat VSMC line are useful for functional studies, a gene therapy protocol for a transient condition such as intimal hyperplasia might more rationally exploit transient transfection. Thus the activity of the novel myb- and mycengrailed chimeras we assembled were examined in vitro under these conditions. Primary human VSMCs were transiently transfected with pCI, the chimeric proteins MybEn or MycEN in pCI or pCINeo MEnT as indicated. The proliferation of each population was determined by tritiated thymidine incorporation after 48 h and the level of apoptosis by TUNEL at 24 h. A reduction in the proliferation level is seen with MybEn-, MycEn- and MEnT-transfected populations versus pCI-transfected (Figure 4a). This reduction in proliferation is accompanied by the induction of apoptosis (Figure 4b). The induction of apoptosis by myb-engrailed chimeras was not unexpected as previous studies have shown them to mediate apoptosis by a pathway which can be inhibited by bcl-2 in T cells.13 It should be understood, however, that lymphocytes are widely acknowledged to be on a ‘hair-trigger’ to apoptosis and thus fin-

Figure 4 The MybEn and MycEn chimeras induce apoptosis. (a) Primary human VSMCs (5 × 104 cells per well in sextuplet) were transiently transfected with empty vector pCI, the chimeric proteins MybEn or MycEN in pCI, or pCINeo MEnT as indicated (transfection efficiency determined as approximately 40%). The proliferation of each population was assayed by a 4-h tritiated thymidine pulse (as per Figure 3a) after 48 h. For clarity, data are expressed as the observed counts per minute from each group relative to the pCI transfected population which is given a value of one. (b) Primary human VSMCs (5 × 104 cells per well in triplicate on glass coverslips) were transiently transfected with empty vector pCI, the chimeric proteins MybEn or MycEN in pCI, or pCINeo MEnT as described. The level of apoptosis was examined by TUNEL after 24 h. The mean percentage of apoptotic nuclei relative to normal nuclei and standard error of the mean were determined after counting cells in a minimum of 20 microscopic fields. Paired Student’s t tests were applied with the aid of the Sigmaplot program to compare the level of apoptosis induced by each construct relative to pCI. Significance (*) was determined as a probability that the two means were different at the P ⬍ 0.05 level. VSMC, vascular smooth muscle cell.

dings in these cells may not translate well to other tissues. Other studies utilizing c-myb cDNA lacking the DNAbinding domain as well as a myb-engrailed fusion demonstrated that in addition to its effect on the cell cycle, cmyb has a role in regulating intracellular stores of calcium.34 For c-myc to induce apoptosis, however, it requires association with its dimerisation partner Max.43 In addition, Myc-induced apoptosis is usually associated with c-myc overexpression,44 although the reverse appears to hold for some B-lymphocyte lines.45 It is possible that apoptosis induced by chimeric MycEn is due to

Transcriptionally targeted dominant-negative gene constructs JF Schmitt et al

1189

Figure 5 Coupling of the human VSMC a-actin transcriptional enhancer in cis to MybEn and MycEn chimeras confers VSMC-specific cytopathicity. (a) Transfection of the luciferase reporter gene in pCI or pCI/h999 was performed to demonstrate efficient transfection of the lineages and the specificity of the pCI/h999 vector. The pCI, pCI-Luc or pCI/h999-Luc constructs were transfected in triplicate (5 × 104 cells per well) into rat and rabbit VSMCs, the HUVEC cell line ECV304 or bovine aortic endothelial cells (BAECs) as described.39,40 The cells were lysed 48 h later and luciferase activity measured according to the manufacturer’s instructions (Promega) as previously described.40 Data are expressed as the mean of triplicate measurements of luciferase enzymatic activity (in light units) with the standard error of the mean also shown. (b) Primary rat and rabbit VSMCs, the HUVEC cell line ECV304 or bovine aortic endothelial cells (BAECs) (5 × 104 cells per well in sextuplet) were transiently transfected with empty vector pCI, the chimeric proteins MybEn or MycEN in pCI or pCI/h999, or pCINeo MEnT as indicated. The proliferation of each population was assayed by a 4-h tritiated thymidine pulse (as per Figure 3a) after 48 h. For clarity, data are expressed as the observed counts per minute from each group relative to the pCI transfected population which is given a value of one. (c) Primary rabbit VSMCs and bovine aortic endothelial cells (BAECs) (5 × 104 cells per well in triplicate on glass coverslips) were transiently transfected with empty vector pCI, the chimeric proteins MybEn or MycEN in pCI or pCI/h999, or pCINeo MEnT as described (5 × 104 cells per well). The level of apoptosis was examined by TUNEL after 24 h. The mean percentage of apoptotic nuclei relative to normal nuclei and standard error of the mean were determined after counting cells in a minimum of 20 microscopic fields. Paired Student’s t tests were applied with the aid of the Sigmaplot program to compare the level of apoptosis induced by each construct relative to pCI within each lineage. Significance (*) was determined as a probability that the two means were different at the P ⬍ 0.05 level. BAEC, bovine aortic endothelial cells; Luc, luciferase; RLU, relative light units; VSMC, vascular smooth muscle cells.

Transcriptionally targeted dominant-negative gene constructs JF Schmitt et al

1190

its overexpression and dimerisation with Max, rather than by directly interfering with endogenous c-myc function. To increase the usefulness of these constructs further for the treatment of a condition such as intimal hyperplasia, the ability to target VSMCs specifically is desirable. For this reason the chimeric genes were cloned into the pCI/h999 expression vector which specifically supports transcription in muscle lineages only (Figure 2a).39 Primary rat and rabbit VSMCs, the HUVEC cell line ECV304 or bovine aortic endothelial cells (BAECs) were then transiently transfected with pCI, pCINeo MEnT, or the chimeric proteins MybEn or MycEN in pCI or pCI/h999 as indicated. The proliferation of each population was determined by tritiated thymidine incorporation after 48 h and the level of apoptosis by TUNEL at 24 h. Transfection of the luciferase reporter gene in pCI or pCI/h999 was performed to demonstrate efficient transfection of the lineages and the specificity of the pCI/h999 vector in each case (Figure 5a). Coupling of the human VSMC ␣-actin transcriptional enhancer in cis to MybEn and MycEn confers VSMC-specific cytopathicity. A reduction in the proliferation level is seen with both MybEn and MycEn in all populations transfected with the pCI form, but only in VSMCs if the pCI/h999 vector is used (Figure 5b). This specific reduction in proliferation is accompanied by the specific induction of apoptosis (Figure 5c). A significant increase in the apoptosis level is observed for all constructs (pCI or pCI/h999 forms) relative to the empty vector pCI (P ⬍ 0.05) in the rabbit VSMC populations while BAECs are only affected when transfected with constitutively expressed cassettes (pCI form only). This reflects the fact that transcription is unsupported by the pCI/h999 vector in non-VSMCs (Figures 2a and 5a). In summary, our findings demonstrate that expression of the chimeric dominant-negative myb- and mycengrailed proteins described can dramatically decrease the proliferation level of an exponentially growing population, due at least in part to the induction of apoptosis. Experiments have shown that even in the case of VSMCs which stably express the inducible myb-engrailed protein MERT, only a percentage of the population is targeted for apoptosis following induction. Furthermore, by the coupling of these constructs to the human VSMC ␣-actin smooth muscle promoter/enhancer we were able to demonstrate VSMC-specific expression and action of the dominant-negative constructs. Together, these findings have a potential application in gene therapy-mediated regulation of VSMC proliferation/extracellular matrix production in the developing restenotic lesion. The advantages of the constructs we describe here over those previously utilized are many-fold: the inclusion of the VSMC ␣-actin regulatory elements in the pCI/h999 expression vectors allows the specific targeting of VSMCs in vitro and we predict in vivo. The use of the dominantnegative approach also contributes to specific cell targeting, as opposed to the by stander effect observed with gene-prodrug therapies such as the HSVtk gene and gancyclovir. Another advantage of the myb- and mycengrailed chimeras is their apparent inability to block cell proliferation completely, even in a stably transformed VSMC population. In the treatment of restenosis, this is a desirable outcome as the aim is to reduce the proliferation of the VSMC pool, not destroy it completely. With

the advent of more efficient methods of gene delivery, such that a greater percentage of cells in the vessel wall could be transfected, it could be important to utilize an inducible system. The studies we describe show mybengrailed chimeras to be amenable to the tamoxifeninducible system, while previous studies have shown the activity of a c-myc-ER fusion.46 Finally, the chimeric c-myb and c-myc constructs were found to be very effective at inhibiting VSMC proliferation in rat, rabbit and human VSMCs by a mechanism which involves apoptosis. A percentage of VSMCs undergo apoptosis in the vascular remodeling which occurs during the development of restenosis.47–49 Thus the controlled induction of apoptosis in the remodeling vessel attempts to improve on the endogenous situation.

Acknowledgements These studies were supported by funding from the Garfield Weston Foundation. We would like to thank the Liver Transplant Unit, Kings College Hospital and the Royal Brompton Hospital for supplying human arteries. We would also like to thank the staff of the immunohistochemistry laboratory, TRI for their aid and technical expertise.

References 1 CABRI Trial participants. First-year results of CABRI (Coronary angioplasty versus bypass revascularisation investigation). Lancet 1995; 346: 1179–1183. 2 Kadel C et al. Long-term follow up after PTCA in patients with single-vessel disease. Am Heart J 1992; 124: 1159–1169. 3 Libby P. Lesion versus lumen. Nature Med 1995; 1: 17–18. 4 Casscells W, Engler D, Willerson JT. Mechanisms of restenosis. Texas Heart Inst J 1994; 21: 68–77. 5 Schwartz RS, Holmes DRJ, Topol EJ. The restenosis paradigm revisited: an alternative proposal for cellular mechanisms. Am Coll Cardiol 1992; 20: 1284–1293. 6 Andersen HR et al. Remodeling rather than neointimal formation explains luminal narrowing after deep vessel wall injury. Insights from a porcine coronary (re)stenosis model. Circulation 1996; 93: 1716–1724. 7 Andres V. Control of vascular smooth muscle cell growth and its implication in atherosclerosis and restenosis. Int J Mol Med 1998; 2: 81–89. 8 Bauters C et al. Proto-oncogene expression in rabbit aorta after wall injury. First marker of the cellular process leading to restenosis after angioplasty? Eur Heart J 1992; 13: 556–559. 9 Brown KE, Kindy MS, Sonenshein GE. Expression of the c-myb proto-oncogene in bovine vascular smooth muscle cells. J Biol Chem 1992; 267: 4625–4630. 10 Gadeau AP, Campan M, Desgranges C. Induction of cell-cycle dependent genes during cell cycle progression of arterial smooth muscle cells in culture. J Cell Physiol 1991; 146: 356–361. 11 Miano JM et al. Early proto-oncogene expression in rat aortic smooth muscle cells following endothelial removal. Am J Pathol 1992; 137: 761–765. 12 Bennet M, Evan G, Newby A. Deregulated expression of the cmyc oncogene abolishes inhibition of proliferation of rat vascular smooth muscle cells by serum reduction, interferon-gamma, heparin and cyclic nucleotide analogues and induces apoptosis. Circ Res 1993; 74: 525–538. 13 Taylor D, Badiani P, Weston K. A dominant negative interfering myb mutant causes apoptosis in T cells. Genes Dev 1996; 10: 2732–2744. 14 Wu M et al. Inhibition of c-myc expression induces apoptosis of WEHI 231 murine B cells. Mol Cell Biol 1996; 16: 5015–5025.

Transcriptionally targeted dominant-negative gene constructs JF Schmitt et al

15 Salomoni P et al. Resistance to apoptosis in CTLL-2 cells constitutively expressing c-Myb is associated with induction of BCL2 expression and Myb-dependent regulation of bcl-2 promoter activity. Proc Natl Acad Sci USA 1997; 94: 3296–3301. 16 Bennett RM et al. Inhibition of vascular smooth muscle cell proliferation in vitro and in vivo by c-myc antisense oligodeoxynucleotides. J Clin Invest 1994; 93: 820–828. 17 Biro S et al. Inhibitory effects of antisense oligodeoxynucleotides targeting c-myc mRNA on smooth muscle cell proliferation and migration. Proc Natl Acad Sci USA 1993; 90: 654–658. 18 Ebbecke M et al. Antiproliferative effects of a c-myc antisense oligonucleotide on human arterial smooth muscle cells. Basic Res Cardiol 1992; 87: 585–591. 19 Shi Y et al. Downregulation of c-myc expression by antisense oligonucleotides inhibits proliferation of human smooth muscle cells. Circulation 1993; 88: 1190–1195. 20 Simons M et al. Antisense c-myb oligonucleotides inhibit intimal arterial smooth muscle cells accumulation in vivo. Nature 1992; 359: 67–70. 21 Gunn J et al. The effect of oligonucleotides to c-myb on vascular smooth muscle cell proliferation and neointima formation after porcine coronary angioplasty. Circ Res 1997; 80: 520–531. 22 Shi Y et al. Cellular, molecular, biological and immunological research: transcatheter delivery of c-myb antisense oligomers reduces neointimal formation in a porcine model of coronary artery balloon injury. Circulation 1994; 90: 944–951. 23 Burgess TL et al. The antiproliferative activity of c-myb and cmyc antisense oligonucleotides in smooth muscle cells is caused by a nonantisense mechanism. Proc Natl Acad Sci USA 1995; 92: 4051–4055. 24 Villa AE et al. Effects of antisense c-myb oligonucleotides on vascular smooth muscle cell proliferation and response to vessel wall injury. Circ Res 1995; 76: 505–513. 25 Ho PT, Parkinson DR. Antisense oligonucleotides as therapeutics for malignant diseases. Semin Oncol 1997; 24: 187–202. 26 Castier Y et al. The activity of c-myb antisense oligonucleotide to prevent intimal hyperplasia is nonspecific. J Cardiovasc Surg 1998; 39: 1–7. 27 Xin Y, Min W. Retrovirus mediated transfer of antisense human c-myc gene into human esophageal cancer cells suppressed cell proliferation and malignancy. Sci China 1992; 35: 76–83. 28 Sklar MD et al. Depletion of c-myc with specific antisense sequences reverses the transformed phenotype in ras oncogenetransformed NIH3T3 cells. Mol Cell Biol 1991; 11: 3699–3710. 29 Fujita M, Shiku H. A human T lymphotropic virus type 1 (HTLV-1) long terminal repeat-directed antisense c-myc construct with an Epstein–Barr virus replicon vector inhibits cell growth in a HTLV-1-transformed human T cell line. FEBS Letts 1991; 322: 15–20. 30 Indolfi C et al. Inhibition of cellular ras prevents smooth muscle cell proliferation after vascular injury in vivo. Nature Med 1995; 1: 541–545. 31 Ueno H et al. Adenovirus-mediated transfer of a dominant-

32

33 34

35 36

37 38

39

40

41

42

43

44 45

46

47

48

49

negative H-ras suppresses neointimal formation in ballooninjured arteries in vivo. Arterioscler Thromb Vasc Biol 1997; 17: 898–904. Chang MW et al. Cytostatic gene therapy for vascular proliferative disorders with a constitutively active form of the retinoblastoma gene product. Science 1995; 267: 518–522. Badiani P et al. Dominant interfering alleles define a role for cmyb in T-cell development. Genes Dev 1994; 8: 770–782. Husain M et al. c-myb-dependent cell cycle progression and Ca2+ storage in cultured vascular smooth muscle cells. Circ Res 1997; 80: 617–626. Jaynes JB, O’Farrell PH. Active repression of transcription by the engrailed homeodomain protein. EMBO J 1991; 10: 1427–1433. Amati B, Land H. Myc-Max-Mad: a transcription factor network controlling cell cycle progression, differentiation and death. Curr Opin Genet Dev 1994; 4: 102–108. Amati B et al. Oncogenic activity of the c-Myc protein requires dimerization with Max. Cell 1993; 72: 233–245. Mukherjee B, Morgenbesser SD, DePinho R. Myc family oncoproteins function through a common pathway to transform normal cells in culture: cross-interference by Max and transacting dominant mutants. Genes Dev 1992; 6: 1480–1492. Keogh M-C et al. Design of a muscle cell-specific expression vector utilising human VSMC ␣-actin regulatory elements. Gene Therapy 1999; 6: 616–628. Keogh M-C et al. High efficiency reporter gene transfection of vascular tissue in vitro and in vivo using a cationic lipid–DNA complex. Gene Therapy 1997; 4: 162–171. Vincent S, Marty L, Fort P. S26 ribosomal protein RNA: an invariant control for gene regulation experiments in eukaryotic cells and tissues. Nucleic Acids Res 1993; 21: 1498. Pavlik EJ et al. Antagonism to estradiol in the mouse: reduced entry of receptors complexed with 4-hydroxytamoxifen into a Mg2+-soluble chromatin fraction. Endocrinology 1986; 118: 1124–1134. Amati B et al. The c-Myc protein induces cell cycle progression and apoptosis through dimerization with Max. EMBO J 1993; 12: 5083–5087. Green DR. A Myc-induced apoptosis pathway surfaces. Science 1997; 278: 1246–1247. Sonenshein GE. Down-modulation of c-myc expression induces apoptosis of B lymphocyte models of tolerance via clonal deletion. J Immunol 1997; 158: 1994–1997. Littlewood TD et al. A modified oestrogen receptor ligand-binding domain as an improved switch for the regulation of heterologous proteins. Nucleic Acids Res 1995; 23: 1686–1690. Bochaton-Piallat M et al. Apoptosis participates in cellularity regulation during rat aortic intimal thickening. Am J Pathol 1995; 146: 1059–1064. Hamet P et al. The time window of apoptosis: a new component in the therapeutic strategy for cardiovascular remodeling. J Hypertens 1996; 14: S65–S70. Isner JM et al. Apoptosis in human atherosclerosis and restenosis. Circulation 1995; 91: 2703–2711.

1191