Enhancer Stimulation Unmasks Latent Gene Transfer After Adenovirus-Mediated Gene Delivery Into Human Vascular Smooth Muscle Cells
Recombinant adenoviral vectors are being used increasingly for gene transfer studies in mammalian cells and gene therapy protocols in humans. High adenoviral titers are often required for successful transduction of vascular smooth muscle cells (VSMCs), defined as uptake and detectable expression of the foreign gene, but the relative contributions of efficiency of viral uptake and control of transcription are poorly understood. To explore the extent to which a lack of detectable gene expression may be due to inefficient transcription of a successfully transferred gene, we have used a replication-deficient adenovirus expressing β-galactosidase (RAd35β-Gal), under the control of the human cytomegalovirus major immediate-early promoter (CMV-IEP), which contains cAMP and nuclear factor-κB response elements, to investigate constitutive and inducible gene expression after gene transfer into human VSMCs. Histochemical staining with 5-bromo-4-chloro-3-indolyl β-d-galactopyranoside (X-gal), a quantitative spectrophotometric assay, SDS-PAGE, Western blotting, and Northern analysis were used to evaluate β−galactosidase expression in infected cells. After infection with RAd35β-Gal at 30, 100, and 1000 plaque-forming units per cell (pfu/cell), expression of β-galactosidase was augmented up to 17-, 19-, and 23-fold, respectively, in human VSMCs treated with forskolin and phorbol ester compared with unstimulated cells. After infection, the proportion of detectably transduced cells was increased by enhancer stimulation from 58% to 100% at 100 pfu/cell and from 9% to 62% at 10 pfu/cell, indicating quiescent viral DNA in unstimulated cells. At high adenoviral titers (1000 pfu/cell), the recombinant gene became the most abundant protein in cell extracts. These findings demonstrate that in human VSMCs, limited constitutive expression from the CMV-IEP, rather than failure of translocation of adenoviral DNA, may be responsible for the apparent failure of transduction at a low multiplicity of infection.
The expression of foreign genes in mammalian cells holds great promise for the investigation of gene function in complex models of disease and may lead to the development of novel therapeutic strategies. Direct gene transfer into VSMCs has aroused considerable interest because of their involvement in atherosclerosis and restenosis after therapeutic angioplasty. Transduction efficiency, defined as the proportion of exposed cells expressing the foreign gene, appears to be an important limiting factor in the application of gene transfer techniques to vascular biology, and in this respect, the use of cationic liposomes has proved particularly disappointing.1 Replication-deficient adenoviral vectors can effectively transduce VSMCs in culture, but high titers are required,2 3 with concentrations of ≈1010 pfu/mL needed for in vivo expression in arteries.2 3 4
Successful expression of a foreign gene is dependent not only on entry of DNA into the cell and translocation to the nucleus but also on efficient transcription. However, the relative contributions of each of these components to the requirement for high adenoviral titers to transduce VSMCs are unclear. The design of adenoviral and other vectors has involved the use of promoters thought to be constitutively active in target cells5 ; thus, a lack of recombinant protein production has previously been interpreted as a failure of cell entry and translocation to the nucleus.6 The human CMV-IEP and the RSV-LTR are able to use cellular transcription factors present in a number of cell types and are therefore said to exhibit constitutive activity. Because of its powerful enhancer activity,7 the CMV-IEP has been widely used in the field of gene transfer, including the first human therapeutic trial of arterial gene transfer,8 despite the fact that limited data exist on its activity in vascular cells.
The CMV-IEP contains a number of 18- and 19-bp repeats, which exhibit strong enhancer activity after binding the transcription factors, NF-κB9 and CREB,10 respectively. It is possible to stimulate these transcription factors pharmacologically by using phorbol ester for NF-κB9 11 and forskolin for CREB10 11 . We have used a replication-deficient adenovirus expressing β-galactosidase (RAd35β-Gal), driven by the CMV-IEP, to transduce human VSMCs in vitro. Stimulation of the cells with agents known to induce NF-κB and CREB was used to investigate constitutive and inducible expression from the CMV-IEP in human VSMCs and to test the hypothesis that gene expression after successful translocation of a foreign gene may remain undetectable until stimulated by activation of enhancer elements in the promoter.
Materials and Methods
Human VSMCs were cultured from segments of aorta discarded at transplantation using an explant method. Briefly, the endothelial layer was removed by abrasion with a sterile cotton bud, and 2-mm cubes of underlying media were placed in 12-well tissue culture flasks (Nunc) at a density of ≈5 pieces/cm2. VSMCs grew out onto the plastic after 5 to 7 days and were passaged for the first time after ≈3 weeks. Cells were grown in medium 199 with 20% FCS, 100 U/mL penicillin, 100 mg/mL streptomycin, and 0.25 mg/mL amphotericin B. Secondary cultures that were grown in the same medium were passaged just before confluence and were split 1:2 using trypsin/EDTA. For these studies, cells were used between passages 6 and 11. Positive staining with anti–α-smooth muscle actin staining (Sigma Chemical Co) confirmed that the cells were VSMCs.
Replication-Deficient Adenovirus Vector
A replication-deficient (E1 deletion) adenovirus vector (RAd35β-Gal) expressing β-galactosidase using the CMV-IEP (−299 to +65) has been described before.11 The replication-deficient adenovirus vector Av1LacZ4 (provided by Genetic Therapy Inc) contains nuclear-targeted β-galactosidase under the control of the RSV-LTR and has been described before.2 The virus was grown on 293 cells, purified by cesium density gradient centrifugation, and dialyzed before storage at −70°C. Plaque assay on 293 cells was used to quantify the viral stocks.
In Vitro Gene Transfer
Human VSMCs were maintained in medium 199 containing 0.5% FCS for 48 hours to induce quiescence before exposure to the adenovirus. The cells were washed with EBSS (GIBCO) before being incubated with 1 mL of adenoviral suspension at various concentrations in Optimem (GIBCO) for 90 minutes at 37°C. After removal of the viral suspensions, the cells were washed three times with EBSS and then maintained in medium 199 with 0.5% FCS for 48 hours before staining or harvesting. For enhancer stimulation experiments, cells were exposed to viral suspensions as described above and then incubated for 48 hours with 10 mmol/L forskolin or 50 ng/mL PMA with 4 mg/mL PHA in medium 199 containing 0.5% FCS. At this 48-hour time point, β-galactosidase expression was determined by histochemical staining, biochemical assay, SDS-PAGE, Western analysis, and Northern blotting, as detailed below. Cell density was such that incubation with 1 mL of 2×106 pfu/mL resulted in an MOI of 10 pfu/cell, 1 mL of 2×107 pfu/mL led to 100 pfu/cell, and 1 mL of 2×108 resulted in 1000 pfu/cell. Our experiments were carried out on a constant VSMC density and a constant infecting volume. Therefore, MOI was determined by changes in viral titers.
Analysis of β-Galactosidase Expression
After exposure to the adenovirus, cells were maintained in medium with 0.5% serum for 48 hours (with or without enhancer-stimulating agents). Cells were washed with PBS, fixed with 0.5% gluteraldehyde, and incubated with the chromogenic substrate X-gal for 2 hours at 37°C. Blue cytoplasmic staining was used to identify successfully transduced cells. At least 500 cells were counted in each sample for estimation of transduction efficiency.
Total cytoplasmic RNA was prepared from cultured cells using a Nonidet P-40 lysis method.12 RNA samples were run on 1.5% agarose gels containing 2.25 mol/L formaldehyde before being transferred to Hybond-N (Amersham International). Filters were hybridized with LacZ, adenoviral DBP, or GAPDH 32P probes generated from purified plasmid inserts using an oligolabeling kit (Pharmacia). Filters were washed with 0.1× SSC with 0.1% SDS for 10, 30, and 60 minutes before exposure to Fuji RX x-ray film. The plasmid pBL 1393-DBP, from which the full-length cDNA for DBP was excised, was a kind gift from Prof W. Russell, University of St. Andrews, UK.
Assay of β-Galactosidase Activity in Cell Extracts
Extracts of harvested cells were prepared by freezing/thawing (between dry ice with ethanol and 37°C) three times, centrifugation, and recovery of the supernatant. After incubation with ONPG, absorbance at 414 nm, using a 96-well plate reader (Titertek Multiskan), was used to quantify β-galactosidase activity (expressed as absorbance units at 414 nm/μg protein per hour). Absorbance at 414 nm was measured at multiple time points after incubation with ONPG, and β-galactosidase activity for each sample was determined from the linear part of the curve.
SDS-PAGE and Western Blotting
Cell extracts were prepared as described above, and protein concentration was measured using a standard assay kit (Bio-Rad). Protein (15 μg) from each sample was run on 8% SDS-PAGE gels. Protein bands were assessed after Coomassie blue staining. For Western analysis, proteins were transferred onto Immobilon-P, and specific protein expression was detected using a monoclonal antibody to β-galactosidase (Sigma), anti-mouse IgG alkaline phosphatase conjugate (Sigma), and 5-bromo-4-chloro-3-indolyl phosphate/nitro blue tetrazolium (Sigma).
Detection of Adenoviral DNA by PCR
Isolation of DNA From Nuclei
Human VSMCs were infected with RAd35β-Gal as described previously and harvested 48 hours later by trypsinization. The cells were washed once in PBS and then resuspended in 400 μL of Nonidet P-40 lysis buffer (0.5% Nonidet P-40, 10 mmol/L Tris, pH 7.4, 1 mmol/L MgCl2, and 150 mmol/L NaCl) at 4°C. The cells were maintained on ice for 3 minutes and then centrifuged at 8000 rpm for 5 minutes in a microfuge. The supernatant was removed, and the nuclear pellet was resuspended in 50 μL of proteinase K solution (50 μg/mL proteinase K, 10 mmol/L Tris, pH 7.8, 5 mmol/L EDTA, and 0.5% SDS) and incubated for 6 hours at 37°C. The DNA was extracted using phenol/chloroform, precipitated with ethanol, and resuspended in 50 μL Tris EDTA buffer.
PCR of Adenoviral DNA
PCR primers spanning the adenoviral hexon gene were used to give a 308-bp product.13 The primer sequences were 5′-GCCGCAGTGGTCTTACATGCACATC-3′ (sense) and 5′-CAGCACGCCGCGGATGTCAAAGT-3′ (antisense). For each reaction, 2 μL of nuclear DNA preparation was mixed with 3 μL of 10× reaction buffer (Promega), with 80 μmol/L dNTPs, 1 mmol/L of each primer, 1 U Taq polymerase (Promega), 6 mmol/L MgCl2, and nuclease-free water in a total volume of 30 μL. After an initial denaturation at 95°C for 5 minutes, samples were subjected to 35 rounds at 95°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds, with a final extension step of 72°C for 10 minutes. Samples were run on 1% agarose gels containing 0.5 μg/mL of ethidium bromide, and amplification products were visualized with UV light.
Constitutive Expression From the CMV-IEP in Human VSMCs
Initial experiments were performed to assess the constitutive activity of the CMV-IEP after VSMCs were infected with RAd35β-Gal. At an MOI of 1000 pfu/cell, virtually all cells were transduced, whereas at 100 pfu/cell, ≈40% of cells were transduced (Fig 1⇓). A similar dose response was observed when β-galactosidase enzyme activity was analyzed quantitatively in cell extracts using an ONPG assay (Fig 2⇓). Experiments were performed on a constant VSMC density such that infection with 1 mL of 2×108 pfu/mL resulted in an MOI of 1000 pfu/cell, 1 mL of 2×107 led to 100 pfu/cell, and 2×106 produced 10 pfu/cell. No blue staining or β-galactosidase activity was observed in uninfected cells.
Enhancer Stimulation Augments Gene Expression in Human VSMCs
Extracts from cells treated with forskolin or PMA/PHA after infection with RAd35β-Gal had significantly greater β-galactosidase activity than extracts from infected but unstimulated cells (Fig 3⇓). Augmentation was demonstrated in VSMCs infected at 30, 100, and 1000 pfu/cell. Individually, forskolin and PMA/PHA increased β-galactosidase activity, but their combined use led to the greatest levels of activity, over 20 times that seen in unstimulated VSMCs after infection. No β-galactosidase activity was detected in uninfected VSMCs treated with forskolin or PMA/PHA.
Northern analysis demonstrated an increase in β-galactosidase mRNA after enhancer stimulation (Fig 4⇓, top). In keeping with the enzyme activity data, the increased expression after treatment with PMA/PHA was greater than that with forskolin. To exclude the unlikely possibility that enhancer stimulation might allow adenoviral replication, Northern analysis was performed to examine the expression of the DNA binding protein from the E2a region. Expression of DBP was observed in 293 cells (in which E1-deleted adenoviruses can replicate) but not in VSMCs stimulated with PMA/PHA (Fig 4⇓, bottom). Our results are thus consistent with increased transcription of β-galactosidase mRNA after enhancer stimulation.
Enhancer Stimulation Unmasks Latent Gene Transfer in Human VSMCs
The augmented β-galactosidase activity in cell extracts after infection at 100 pfu/cell may have resulted from increased expression in VSMCs already constitutively transduced or from VSMCs in which constitutive transduction was undetectable. To address this issue, we investigated whether enhancer stimulation would increase the percentage of cells showing histochemical evidence of β-galactosidase activity after infection at 100 and 10 pfu/cell. The use of both enhancer-stimulating agents together increased the proportion of transduced cells at 100 pfu/cell from 40% to 98% and at 10 pfu/cell from 5% to 55% (Fig 5⇓, top left and top right). No transduction was observed in uninfected cells treated with forskolin and PMA/PHA. To maximize our ability to detect successful transduction, the incubation time with X-gal was increased from 2 hours to 8 and 24 hours. These increased staining times led to only a small increment in the proportion of transduced VSMCs in both stimulated and unstimulated cells but did not influence the ability of enhancer stimulation to unmask β-galactosidase expression that was previously undetectable (Fig 5⇓, middle left and bottom left). These results suggest that failure to detect gene expression after infection at 10 or 100 pfu/cell is, to a large extent, due to insufficient gene expression rather than failure of translocation of the viral genome to the nucleus, in that successful translocation of the exogenous gene was unmasked by enhancer stimulation.
To test whether the effect seen with enhancer stimulation was specific to the CMV-IEP, stimulation experiments were performed on human VSMCs infected with an adenovirus vector expressing β-galactosidase driven by an RSV promoter (Av1LacZ4), which does not contain NF-κB or CREB binding sites. Stimulation with forskolin or phorbol ester had no effect on the proportion of transduced cells (Fig 6⇓).
To determine whether viral DNA was present in the nuclear preparations of cells infected at a low MOI, human VSMCs in culture were infected with RAd35β-Gal at various MOIs. After 48 hours, DNA was isolated from nuclear extracts, and PCR to a region of the hexon gene was used to identify the presence of adenoviral DNA. The expected 308-bp fragment was amplified from cells infected at titers down to 0.1 pfu/cell but not from uninfected cells (Fig 7⇓). Thus, adenoviral DNA was identified after infection at MOIs when no transduction was observed.
β-Galactosidase Protein Production
To assess the proportion of β-galactosidase protein relative to the total amount of endogenous protein in the infected human VSMCs, SDS-PAGE electrophoresis was performed. SDS-PAGE gels of extracts of VSMCs exposed to 30 pfu/cell showed no detectable band, with Coomassie staining, at the level of the 116-kD size marker (data not shown). However, at 1000 pfu/cell, a new 116-kD band (the appropriate size for β-galactosidase) was seen in unstimulated cells. This band became the predominant protein species in the cell when infected cells were stimulated with a combination of forskolin and PMA/PHA. No 116-kD band was seen when uninfected cells were treated with both agents (Fig 8⇓, left). The identity of the 116-kD band and its upregulation after enhancer stimulation were confirmed using Western blotting with an antibody to β-galactosidase (Fig 8⇓, right).
In the present study, we have explored the ability of a replication-deficient adenovirus carrying the gene for β-galactosidase, driven by the CMV-IEP, to transduce human VSMCs in culture. We have shown that at low adenoviral titers, when viral DNA has been successfully delivered to the nucleus, pharmacological stimulation of enhancer elements in the CMV-IEP is required to produce detectable gene expression. By stimulating enhancer elements within the CMV-IEP, we have been able to augment expression and substantially increase the proportion of detectably transduced cells after infection at 10 and 100 pfu/cell. Therefore, our data suggest that failure to produce detectable VSMC transduction at low adenoviral titers may be due more to limited activity of the CMV-IEP in human VSMCs than to failure of adenoviral delivery. These are important observations, because previously it has been assumed that successful translocation of adenoviral DNA would inevitably lead to detectable gene expression if driven by promoters such as the CMV-IEP and the RSV-LTR.6 8
Successful transduction depends on a number of factors, including DNA delivery, translocation to the nucleus, and transcription. Thereafter, detection of transduction depends on the sensitivity of the assay to detect the protein product. However, it should be emphasized that in gene transfer studies, both in vitro and in vivo, the desirable quantity of recombinant product will depend much more on the biological activity of the protein than the sensitivity of an assay to detect it. Therefore, optimal promoter activity will be dictated by the potency of the gene product. Thus, β-galactosidase, which has been widely used to assess transduction after gene delivery, can be used only as a guide to illustrate the principle of enhancer stimulation to maximize gene expression. In our experiments, we were able to demonstrate viral DNA in nuclear extracts of cells in which no protein product could be detected by histochemical staining. Although this may reflect, to some extent, sensitivity of the PCR to detect viral DNA and insensitivity of the assay for β-galactosidase, our findings are consistent with the view that low constitutive activity from the CMV-IEP in human VSMCs may lead to misleadingly low estimates of successful gene delivery. In addition, enhancer stimulation considerably increased the amount of gene product produced by overtly transduced cells. Thus, a 20-fold increase in the β-galactosidase activity of cell extracts was seen with enhancer stimulation at 1000 pfu/cell (when 100% of cells were demonstrably transduced), such that β-galactosidase became the most abundant protein species. However, the production of such large amounts of recombinant protein by all cells is probably undesirable and likely to be toxic.
The identification of viral promoter/enhancers (such as the CMV-IEP, RSV-LTR, and simian virus 40), which are active in a number of cell types, has allowed the development of versatile expression vectors. The CMV-IEP has emerged as possibly the most potent of these promoters.7 14 However, our findings demonstrate that the CMV-IEP has limited activity in human VSMCs. HCMV is a common human pathogen that is capable of lifelong latency after primary infection but may cause disease after reactivation, particularly in immunocompromised hosts.15 Activation of IEP is required to initiate the regulated cascade of HCMV gene expression in permissive cells.16 Therefore, it is not surprising that the activity of the CMV-IEP in human cells is highly regulated, and both positive9 10 and negative17 regulatory sequences have been described. The region of the CMV-IEP used in RAd35β-Gal (−299 to +65) is known to contain the CREB and NF-κB binding sites responsible for enhancer activity, and our data demonstrate that in human VSMCs, expression from RAd35β-Gal can be augmented using forskolin and phorbol ester. Although basal NF-κB activity has been documented in VSMCs,18 it would appear from our results that this activity is insufficient to drive maximal CMV-IEP activity in these cells. The RSV-LTR exhibits constitutive activity in some cell types and has also been widely used in adenoviral gene transfer vectors. Our experiments with a replication-deficient adenovirus expressing β-galactosidase driven by the RSV-LTR showed lower levels of transduction at 10 and 100 pfu/cell than those achieved with RAd35β-Gal before enhancer stimulation. This suggests that the RSV-LTR also has limited activity in human VSMCs, since it is unlikely that the uptake and translocation of the two type 5 adenoviruses into human VSMCs are influenced by the promoter sequence in the genome.
Studies using adenoviruses for gene transfer into arteries in vivo have demonstrated more efficient transduction of endothelial and neointimal VSMCs compared with medial VSMCs.3 19 The use of high adenoviral titers themselves has been shown to provoke marked inflammatory and morphological changes in rabbit arteries20 and toxic effects in rat arteries.21 Therefore, strategies aimed at achieving efficient gene expression using lower adenoviral loads (as we have demonstrated in cultured human VSMCs) by the identification of the most appropriate promoters for specific cell types represent a logical approach for the potential refinement of adenoviral vectors. Although any extrapolation from our in vitro work to the in vivo situation must be regarded as speculative, our data raise the possibility that preferential gene expression by intimal VSMCs and endothelial cells may be due to differences in promoter activity in these cells compared with medial VSMCs, in addition to better viral access, since the pattern of gene expression differs markedly between medial VSMCs, neointimal VSMCs, and endothelial cells.12 22 Improved medial transduction has been documented after mechanical disruption of the endothelium23 ; however, this injury may modulate the transcriptional activity of medial VSMCs as well as facilitate viral access.
Our findings illustrate a number of important principles that may help in the design of adenoviral vectors for gene transfer. First, the lowest adenoviral titers needed to translocate DNA successfully to the target cells should be established, since translocation of DNA to the nucleus is all that is required of the virus, and higher titers will simply induce confounding nonspecific responses. By stimulating the CMV-IEP, we were able to show that failure to detect gene expression in human VSMCs does not necessarily imply failure of gene transfer. Second, the activity of the promoters used in adenoviral constructs should be evaluated in the context of the particular cell type to be targeted and should ideally encompass potential phenotypic and species variations in target cells. Third, manipulation of transcription factors by pharmacological or other means24 is a powerful mechanism for the control of gene expression after gene transfer. Clearly, toxic compounds, such as phorbol esters, cannot be used in vivo, but they can be used experimentally to illustrate the principle of enhancer stimulation. A greater understanding of promoter activity in target cell types and the identification of more potent promoters for VSMCs should lead to a more rational and effective strategy for adenovirus-mediated gene transfer.
Selected Abbreviations and Acronyms
|CMV-IEP||=||cytomegalovirus immediate early promoter|
|CREB||=||cAMP-responsive binding protein|
|DBP||=||DNA binding protein|
|EBSS||=||Earle's balanced salt solution|
|FCS||=||fetal calf serum|
|MOI||=||multiplicity of infection|
|NF-κB||=||nuclear factor κB|
|PCR||=||polymerase chain reaction|
|PMA||=||phorbol 12-myristate 13-acetate|
|RSV-LTR||=||Rous sarcoma virus long terminal repeat|
|VSMC||=||vascular smooth muscle cell|
Dr Clesham is a Medical Research Council Training Fellow. Dr Browne is supported by the Wellcome Trust, and Professor Weissberg is supported by the British Heart Foundation.
- Received January 30, 1996.
- Accepted October 3, 1996.
Pickering JG, Jekanowski J, Weir L, Takeshita S, Losordo DW, Isner JM. Liposome-mediated gene transfer into human vascular smooth muscle cells. Circulation. 1993;89:13-21.
Lee SW, Trapnell BC, Rade JJ, Virmani R, Dichek DA. In vivo adenoviral vector–mediated gene transfer into balloon-injured rat carotid arteries. Circ Res. 1993;73:797-807.
Lemarchand P, Jones M, Yamada I, Crystal RG. In vivo gene transfer and expression in normal uninjured blood vessels using replication-deficient recombinant adenovirus vectors. Circ Res. 1993;72:1132-1138.
Chang MW, Barr E, Seltzer J, Jiang YQ, Nabel GJ, Nabel EG, Parmacek MS, Leiden JM. Cytostatic gene therapy for vascularly proliferative disorders with a constitutively active form of the retinoblastoma gene product. Science. 1995;267:518-522.
Isner JM, Walsh K, Symes J, Pieczek A, Takeshita S, Lowry J, Rossow S, Rosenfeld K, Weir L, Brogi E, Schainfeld R. Arterial gene therapy for therapeutic angiogenesis in patients with peripheral artery disease. Circulation. 1995;91:2687-2692.
Stamminger T, Fickenscher H, Fleckenstein B. Cell type-specific induction of the major immediate early enhancer of human cytomegalovirus by cyclic AMP. J Gen Virol. 1990;71:105-113.
Wilkinson GWG, Akrigg A. Constitutive and enhanced expression from the CMV major IE promoter in a defective adenovirus vector. Nucleic Acids Res. 1992;20:2233-2239.
Shanahan CM, Weissberg PL, Metcalfe JC. Isolation of gene markers of differentiated and proliferating vascular smooth muscle cells. Circ Res. 1993;73:193-204.
Allard A, Girones R, Juto P, Wadell G. Polymerase chain reaction for detection of adenoviruses in stool samples. J Clin Microbiol. 1990;28:2659-2667.
Tobin JO. Cytomegalovirus infection. In: Weatherall DJ, Ledingham J, Warrell DA, eds. Oxford Textbook of Medicine. New York, NY: Oxford University Press; 1987:5.76-5.80.
Malone CL, Vesole DH, Stinski MF. Transactivation of a human cytomegalovirus early promoter by gene products from the immediate-early gene IE2 and augmentation by IE1-mutational analysis of the viral proteins. J Virol. 1990;64:1498-1506.
Liu R, Baillie J, Sissons JGP, Sinclair JH. The transcription factor YY1 binds to negative regulatory elements in the human cytomegalovirus major immediate early enhancer/promoter and mediates repression in non permissive cells. Nucleic Acids Res. 1994;22:2453-2459.
Lawrence R, Chang L, Siebenlist U, Bressler P, Sonenshein GE. Vascular smooth muscle cells express a constitutive NF-kB like activity. J Biol Chem. 1994;269:28913-28919.
Guzman RJ, Lemarchand P, Crystal RG, Epstein SE, Finkel T. Efficient and selective adenovirus-mediated gene transfer into vascular neointima. Circulation. 1993;88:2838-2848.
Newman KD, Dunn PF, Owens JW, Schulick AH, Virmani R, Sukhova G, Libby P, Dichek DA. Adenovirus-mediated gene transfer into normal rabbit arteries results in prolonged vascular cell activation, inflammation and neointimal hyperplasia. J Clin Invest. 1995;96:2955-2965.
Schulick AH, Newman KD, Virmani R, Dichek DA. In vivo gene transfer into injured carotid arteries: optimisation and evaluation of acute toxicity. Circulation. 1995;91:2407-2414.
Schwartz SM, deBlois D, O'Brien ERM. The intima: soil for atherosclerosis and restenosis. Circ Res. 1995;77:445-465.
Willard JE, Landau C, Glamann DB, Burns D, Jessen ME, Pirwitz MJ, Gerard RD, Meidell RS. Genetic modification of the vessel wall: comparison of surgical and catheter-based techniques for delivery of recombinant adenovirus. Circulation. 1994;89:2190-2197.