Overexpression of Transforming Growth Factor-;1 Stabilizes Already-Formed Aortic Aneurysms
http://www.100md.com
《循环学杂志》
CNRS UMR 7054, Centre de Recherches Chirurgicales
Université Paris XII, UFR de Médecine (J.D., F.L., A.-M.G., C.P., P.D., J.-P.B., E.A.) and Service de Chirurgie Vasculaire et Endocrinienne, Assistance Publique des H;pitaux de Paris (J.D., P.D., J.-P.B., E.A.)
H;pital H. Mondor, Créteil
INSERM U643, Nantes (I.A.), France.
Abstract
Background— The cell response to transforming growth factor-;1 (TGF-;1), a multipotent cytokine with healing potential, varies according to tissue context. We have evaluated the ability of TGF-;1 overexpression by endovascular gene therapy to stabilize abdominal aortic aneurysms (AAAs) already injured by inflammation and proteolysis.
Methods and Results— Active TGF-;1 overexpression was obtained in already-developed experimental AAAs in rats after endovascular delivery of an adenoviral construct encoding for a mutated form of active simian TGF-;1 and in an explant model using human atherosclerotic AAA fragments incubated with recombinant active TGF-;1. Transient exogenous TGF-;1 overexpression by endovascular gene delivery was followed by induction of endogenous rat TGF-;1. Overexpression of active TGF-;1 in experimental AAAs was associated with diameter stabilization, preservation of medial elastin, decreased infiltration of monocyte-macrophages and T lymphocytes, and a decrease in matrix metalloproteinase-2 and -9, which was also observed in the explant model, in both thrombus and wall. In parallel with downregulation of the destructive process, active TGF-;1 overexpression triggered endoluminal reconstruction, replacing the thrombus by a vascular smooth muscle cell–, collagen-, and elastin-rich intima.
Conclusions— Local TGF-;1 self-induction after transient exogenous overexpression reprograms dilated aortas altered by inflammation and proteolysis and restores their ability to withstand arterial pressure without further dilation. This first demonstration of stabilization of expanding AAAs by delivery of a single multipotent self-promoting gene supports the view that endovascular gene therapy should be considered for treatment of aneurysms.
Key Words: aneurysm ; gene therapy ; transforming growth factor beta
Introduction
Transforming growth factor-;1 (TGF-;1), an important molecular determinant of wall strengthening during vascular development,1,2 mediates responses of vessels submitted to injury by promoting intimal tissue accumulation.3 Abrogation of TGF-; signaling shifts experimental atherosclerotic lesions toward an unstable, inflammatory plaque phenotype depleted in extracellular matrix.4–6
See p 939
Despite these observations and previous attempts,7 no report has documented the potential of TGF-;1 in reprogramming a dilated artery wall toward healing and stability. Abdominal aortic aneurysms (AAAs) show an inexorable tendency to expand under radial hemodynamic stress after artery wall extracellular matrix destruction by inflammation and proteolysis.8 AAAs are lesions in which one can test the ability of TGF-;1 to help inflammation-injured vessels return to normal function, eg, ability to withstand stress without expanding, with clinically relevant consequences because death risk by rupture is proportional to diameter.9
Because the response to TGF-;1 is modulated by other stimuli applied to cells, the question of whether TGF-;1 overexpression might help to control AAA expansion10 should be addressed in aortic tissues altered by ongoing inflammation and proteolysis. Accordingly, we have used already-developed experimental and human lesions to provide evidence that TGF-;1 overexpression induced by endovascular gene transfer, or incubation with recombinant protein, decreases proteolytic burden, slows destruction of the damaged wall, promotes wall reconstruction on the luminal side, and ultimately stabilizes AAA diameter.
Methods
Animals
Male Lewis rats and Hartley guinea pigs (Iffa Credo, Lyon, France) were housed and taken care of according to the European Union Standards. Intraperitoneal injections of 5 mg/100 g body weight of pentobarbital were used for anesthesia. Animals were fed a standard diet.
Gene Vectors
Vectors were derived from a serotype 5 human adenovirus, with E1- and E3-encoding region deletions, containing the cDNA of either Escherichia coli nls ;-galactosidase (Ad-LacZ) or a mutated active form of simian TGF-;1 (Ad-sTGF-;1),11 generously provided by Généthon, France. The constructs were obtained by homologous recombination between the plasmid pAdEasy-1 and a shuttle plasmid pTrack-CMV containing the genes of interest under the control of the cytomegalovirus long-terminal-repeat promoter. Single stocks of viruses from the 2 constructs were placed in aliquots and kept at –80°C until use.
Generation of Expanding Aortic Aneurysms
Guinea pig infrarenal aortas were decellularized using sodium dodecyl sulfate to obtain tubes of intact extracellular matrix12 that were sewn orthotopically into the aorta of 200-g male Lewis rats with 10-0 nylon interrupted sutures (xenograft).13 Fourteen days after implantation, a chimeric aneurysm (>50% diameter increase) had developed with a degraded extracellular matrix of guinea pig origin and cells and thrombus of rat origin.
Endovascular Vector Infusion in Already-Formed Aneurysms
Fourteen days after xenograft implantation, the experimental AAAs were isolated from blood flow by clamps during a second surgical procedure as described previously.14 The lumen was gently rinsed with 0.9% NaCl through a PE10 catheter introduced by means of an aortotomy performed downstream in the native aorta. The catheter was connected to a reservoir containing the viral suspension (0.8x109 IP per AAA) or the vehicle (0.9% NaCl) alone. The virus stock was delivered to the aneurysmal wall during a 20-minute infusion at a pressure of 80 mm Hg generated by a mercury manometer. After infusion, aortotomy and laparotomy were closed.
Euthanasia and Tissue Harvest
Aneurysms were harvested 3, 14, or 28 days after endovascular infusion. The aneurysm diameter was measured in the anesthetized animal using a grid in the eyepiece of the operative microscope. Rats were then euthanized by an overdose of intravenous pentobarbital. The aneurysmal lumen was gently rinsed with PBS, and the middle part was fixed in 70% ethanol, embedded in paraffin, and sectioned transversally at 5 μm. The 2 ends were snap-frozen in liquid nitrogen and kept at –80°C.
Histological and Immunohistochemical Analysis
Cross sections were stained with orcein for elastic fibers and sirius red for collagen. Percentages of stained surface were calculated with Perfect Image software. Polyclonal rabbit anti-synthetic TGF-;1 antibody (a generous gift from Dr Saez, INSERM, Lyon, France),15 goat anti–rat-TIMP-1 (RandD Sysytems), anti–Ser465-467–phosphorylated Smad 2 (Cell Signaling Technology), and 1F4 antibody for rat CD3 (T lymphocytes) (Oxford Biotechnology Ltd) were used with immunoperoxidase staining (Vectastain Elite ABC Kit, Vector Laboratories) counterstained with hematoxylin. Other primary antibodies were mouse anti-rat monoclonals: ED1 clone for monocyte-macrophages (Dakopatts) and 1A4 clone for -actin (smooth muscle cells) (Dakopatts) detected by alkaline phosphatase anti–alkaline phosphatase staining (Dakopatts). Negative controls were generated by omission of the primary antibody or with normal IgG or IgM (for anti–CD-3) (Sigma-Aldrich). Cells were counted using a grid in the microscope eyepiece.
Semiquantification of mRNA by RT-PCR
RNA was extracted from 6 aneurysms after separation of the luminal thrombus from the aneurysmal wall (media/adventitia). Reverse-transcription polymerase chain reaction (RT-PCR), comparative to the domestic gene 18s (QuantumRNA 18s Internal Standards Kit, Ambion), was used to determine mRNA levels. Two sets of primers, one for the gene of interest and one for 18s, were added to the same tube with cDNA and PCR mix. Because of the small amount of material available from each rat aneurysm, tissues were pooled by layers and groups. Total RNA was extracted with TRIzol (Life Technologies) and treated with grade I DNAse (Roche Molecular Biomedicals). RT was done with random primers, Superscript II (Life technologies), dNTP, dithiothreitol, and ribonuclease inhibitor (Roche Molecular Biochemicals). Then, 2 μg RT product was diluted to a final volume of 50 μL. Primers were as follows: matrix metalloproteinase-2 (MMP-2): 5'-CTATTCTGTCAGCACTTTGG-3'/5'-CAGACTTTGGTTCTCCAACTT-3'; MMP-9, 5'-CTGCGT-ATTTCCATTCATCTT-3'/5'-ATGCCTTTTATGTCGTCTTCA-3'; tissue inhibitor of metalloproteinase-1 (TIMP-1), 5'-CCCCAGA-AATCAACGAGAGACCA-3'/5'-ACACCCCACAGCCAGCACTAT-3'; collagen III, 5'-TGCCACCCTGAACTCAAGAG-3'/5'-GCCTTGC-GTGTTTGATATTC-3'; rat TGF-;1, 5'-CGGACTACTACGCCAAAG-AA-3'/5'-TCAAAAGACAGCCACTCAGG-3'; and primate TGF-;1, 5'-AACACATCAGAGCTCCGAGAA-3'/5'-GTCAATGTACAGCTGCCGCAC-3' (Genset Oligos SA). Final PCR conditions were chosen to avoid interference between the 2 sets of primers and to obtain a signal within a linear range of amplification with Taq polymerase (EurobioTaq, Eurobio). Negative controls were done without Superscript II. PCR products (10 μL) were run in a 2% agarose gel with 5 μg/mL ethidium bromide visualized under UV light by a video camera. Bands of amplified sequences corresponding to the gene of interest and to 18s were quantified with Gel Analyst. Results were expressed as a ratio between signals corresponding to the gene of interest and 18s.
Incubation of Human Aneurysmal Tissues With Recombinant Active TGF-;1
Fragments from the maximum dilation zone of 5 human atherosclerotic AAAs (asymptomatic; diameter >50 mm) were collected during elective surgery in the Department of Vascular Surgery of the Henri Mondor hospital in agreement with the local ethics committee. The thrombus was immediately separated from the aneurysmal wall; both were kept in M199 with antibiotics but without serum at 4°C and were sterilely cut into small fragments (<2 mm3) as described previously.16 Explants of thrombus and wall were separately cultured in serum-free culture medium (50% M199, 50% RPMI) with increasing doses of recombinant human active TGF-;1 (rhTGF-;1) (Peprotech) for 24 hours. Quantification of human TIMP-1 protein in explant-conditioned media was done with an ELISA kit (RandD Systems, Inc) according to the manufacturer’s instructions.
Quantification of Gelatinases and TIMP-1
MMP-2 and MMP-9 activities in experimental AAAs, after separation of thrombus/intima from media/adventitia and extraction with a guanidine buffer, and in explant-conditioned media from 5 human atherosclerotic AAAs were quantified on 1% gelatin zymograms.13 MMP-2– and MMP-9–related bands were quantified with ImageMaster software (Pharmacia Biotech). Results were plotted against a standard curve generated with normal rat arteries for intergel normalization and with recombinant MMP-2 and MMP-9 (Medgene Science). Quantification of rat TIMP-1 protein was done with a DuoSet ELISA kit (RandD Systems Europe).
Statistical Analysis
Results are expressed as mean±SD. The percentage of diameter increase was calculated as follows: (diameter at harvest–diameter at infusion)x100/diameter at infusion. Comparisons between 2 groups were made by use of the nonparametric Mann-Whitney U test and between 3 groups with the nonparametric Kruskall-Wallis test; correlations between continuous variables were made with the Z correlation test (Statview, version 4.5). Values of P<0.05 was considered significant.
Results
TGF-;1 Overexpression Stops Expansion of Already-Formed AAAs
Anti–TGF-;1 immunostaining on cross sections showed a stable, albeit small, number of positive cells in AAAs after Ad-LacZ infusion (Figure 1A and 1D). As early as 3 days after Ad-sTGF-;1 infusion, the number of TGF-;1–positive cells increased 4-fold in the thrombus and media/adventitia (thrombus, from 0.25±0.16 to 1.0±0.56x103 cells/mm2 for Ad-LacZ and Ad-sTGF-;1, respectively, P<0.05; media/adventitia, from 0.16±0.17 to 0.59±0.21x103 cells/mm2, respectively, P<0.05) (Figure 1A). In the media/adventitia, ie, in the aneurysmal wall itself, overexpression of TGF-;1 was detected up to 28 days after Ad-sTGF-;1 infusion (Figure 1A) despite the decline in sTGF-;1 mRNA content observed by RT-PCR using primers specific to primate TGF-;1 (Figure 1B). RT-PCR with rat-specific primers showed a 6-fold increase in rat TGF-;1 mRNA content 28 days after Ad-TGF-;1 infusion, suggesting an induction of endogenous rat TGF-;1 after endovascular delivery of the sTGF-;1 gene (Figure 1C). Immunostaining with the antibody to phosphorylated Smad2, a marker of high-affinity serine-threonine TGF-; receptor activation, was markedly increased in cell nuclei 3 days after Ad-sTGF-;1 delivery, suggesting effective TGF-;1 signaling in this group (Figure 1D). These results demonstrate an early, significant, and sustained increase in TGF-;1 with efficient signal transduction on endovascular active sTGF-;1 gene transfer.
Infusion of Ad-LacZ instead of vehicle (0.9% NaCl) had no significant impact on AAA diameter expansion (28-day diameter increase: Ad-LacZ [n=8], 72.6±39.6%; vehicle [n=7], 57.1±26.8%; P=0.42). Mean external diameter of AAAs infused with Ad-LacZ had increased significantly 28 days later, from 2.8±0.6 to 4.8±1.6 mm (n=8; P=0.04), demonstrating disease progression in this control group and in vehicle-infused AAAs (data not shown). In contrast, mean AAA diameters remained stable 28 days after endovascular infusion of Ad-sTGF-;1 (from 3.1±0.5 to 3.3±0.6 mm; n=8; P=NS) (Figure 1E). The mean percentage of diameter increase after 28 days was 4.7±11.8 in Ad-sTGF-;1–infused vessels (comparisons of 28-day percent diameter increase: between Ad-LacZ and Ad-sTGF-;1, P=0.0011; between vehicle and Ad-sTGF-;1, P=0.0018; between the 3 groups [Kruskall-Wallis test], P=0.0008). These data demonstrate that overexpression of active TGF-;1 by the adenoviral vector prevented AAA expansion.
TGF-;1 Overexpression Decreases Medial Injury
Extracellular Matrix and Inflammation
Overexpression of TGF-;1 led to preservation of media elastic fiber network compared with controls (elastic fiber surface, 9.55±0.80% versus 19.71±1.28%, 28 days after Ad-LacZ and Ad-sTGF-;1 endovascular infusion, respectively; P=0.034) (Figure 2A). Overexpression of TGF-;1 was also followed by a 2-fold decrease in monocyte-macrophage density in the media/adventitia as early as 3 days after endovascular gene delivery (2.58±0.12 versus 1.33±0.35x 103 ED1-positive cells per 1 mm2 in Ad-LacZ– and Ad-sTGF-;1–infused AAAs, respectively; P=0.02) and up to 28 days after infusion (Figure 2B and 2C). TGF-;1 overexpression was also associated with a marked decrease in T lymphocyte infiltration 28 days after infusion (Figure 2D).
MMP-Dependent Proteolytic Balance in Experimental and Human Atherosclerotic AAAs
In experimental AAAs, overexpression of active TGF-;1 was associated with a 2- to 4-fold decrease in total and active MMP-9 at day 3 in both the thrombus and media/adventitia (Figure 3A). Because MMP-9 activity detected in AAAs may not originate entirely from synthesis in the artery wall itself,17 we used RT-PCR to examine MMP-9 mRNA sources in the thrombus and aneurysmal wall (ie, media/adventitia) after gene delivery. The most important effect of active TGF-;1 overexpression was a marked decrease in MMP-9 mRNA content (100-fold) in the thrombus of AAAs 3 days after Ad-sTGF-;1 infusion (Figure 3A). MMP-2 activity and mRNA content were decreased 3 days after TGF-;1 overexpression in the media/adventitia, but mRNA had increased slightly in the thrombus (Figure 3B). Because MMP activity in AAAs is also regulated by the inhibitors TIMPs,13 we examined TIMP-1 mRNA content and observed a massive increase (50-fold) in the thrombus 3 days after TGF-;1 overexpression and a 3-fold decrease in the media/adventitia (Figure 3C). TIMP-1 protein was also induced after TGF-;1 expression, as shown by immunohistochemistry at day 3 (Figure 3C) and ELISA on AAA extracts 28 days after gene transfer (26.1±3.8 and 32.7±3.2 ng/mg in Ad-LacZ and Ad-sTGF-;1 groups, respectively; P=0.049).
To demonstrate the relevance of the data obtained in the animal model, we have evaluated the impact of TGF-;1 on MMP and TIMP-1 production by human atherosclerotic AAA explants in culture. Addition of rhTGF-;1 decreased MMP-9 and MMP-2 activity after a 24-hour culture of explants (Figure 4) in conditioned medium from both aneurysmal thrombi (n=5) and walls (n=5). This decrease was observed for all forms of MMP-2 and MMP-9 (data not shown), activated MMP-9 (20 ng/mL rhTGF-;1: thrombus, –34.2±17.8%, P=0.001; wall, –37.4±17.7%, P=0.003), and activated MMP-2 (20 ng/mL rhTGF-;1: thrombus, –50.4±48.1%, P=0.007; wall, –75.0±22.7%, P=0.003) with reference to the media without rhTGF-;1. The decrease in MMP-9 and MMP-2 activity was TGF-;1 dose dependent (Z correlation test; active MMP-9: thrombus, P=0.001; wall, P=0.003; active MMP-2: thrombus, P=0.072; wall, P=0.003). Incubation with rhTGF-;1 had no impact on human TIMP-1 concentration in the conditioned medium (data not shown).
Overexpression of TGF-;1 Promotes Aneurysmal Wall Reconstruction
TGF-;1 overexpression was followed by the development of an intima rich in vascular smooth muscle cells (VSMCs) as demonstrated by anti–-actin immunostaining starting 14 days after Ad-sTGF-;1 delivery (Figure 5A and 5D). VSMCs replaced the preexisting endoluminal thrombus in Ad-sTGF-;1–infused AAAs, whereas thrombus remained the main structure in contact with circulating blood in AAAs infused with Ad-LacZ (Figure 5D). Overexpression of TGF-;1 was followed by an accumulation of collagen in the intima replacing the thrombus (collagen surface at day 28, 8.2±5.4% versus 30.9±17.5% in Ad-LacZ– versus Ad-sTGF-;1–infused AAAs, respectively; P=0.02) (Figure 5B and 5D). Type III collagen mRNA content was increased in the neointima 14 and 28 days after TGF-;1 overexpression (Figure 5C) but not in the media/adventitia (data not shown). In some areas at day 28 after Ad-sTGF-;1 infusion, elastic fibers were detected by orcein staining, mostly on the luminal side of the intima, whereas virtually none were detected in Ad-LacZ–infused vessels (Figure 5D). Overall, the development of a tissue made of VSMCs surrounded by a dense extracellular matrix network on the luminal side of the AAA could be interpreted as wall reconstruction.
Discussion
TGF-;1 is a multipotent cytokine that induces expression of fibrillar collagen18 and elastin19 genes. Under defined circumstances, TGF-;1 also downregulates inflammation, promotes VSMC growth20,21 and differentiation,22 and inhibits MMP-dependent proteolysis.23 We evaluated the impact of active TGF-;1 overexpression on the remodeling process of already-developed AAAs. Because the response to TGF-; is conditioned by costimulatory factors and tissue context,24,25 we overexpressed TGF-;1 in inflammatory and proteolytically injured aortas, ie, in already-formed experimental AAAs in vivo and in human AAAs in vitro. Aneurysmal degeneration of aortic xenografts is reproducible with a low variability, allowing assessment of the impact of TGF-;1 overexpression on diameter, the remodeling criterion used to predict rupture in the human disease.9 Here, we document the efficiency of the endovascular route for gene delivery to overexpress active TGF-;1 in experimental AAAs. At the interface between blood and aorta injured by inflammation and proteolysis, both experimental14 and atherosclerotic AAAs17 display a luminal thrombus that has recently been shown to be associated with wall injury.26 Accordingly, we analyzed separately these 2 components of AAAs with respect to their responsiveness to TGF-;1. We provide evidence that overexpression of a mutated active form of TGF-;1 decreases the proteolytic burden and the inflammatory and destructive process in the aneurysmal wall in experimental and human AAAs; promotes wall reconstruction by inducing the formation of a VSMC-, collagen-, and elastin-rich intima replacing the luminal thrombus; and ultimately stabilizes the diameter of expanding, already-formed AAAs in vivo. The stabilization of AAA diameter after Ad-sTGF-;1 infusion was observed compared with both Ad-LacZ infusion (vector without the gene of interest) and infusion of vehicle alone. Our results support the idea that overexpression of active TGF-;1 adapts the structure of expanding aortic walls injured by inflammation and proteolysis and reestablishes their ability to resist wall stress without further dilation.
Inhibition of Wall Destruction
Injury of the medial layer by inflammation and proteolysis is the hallmark of human atherosclerotic AAAs8 and is reproduced in the xenograft aneurysm model.12–14,27 In both conditions, injury of the elastic fiber network is observed with upregulation of proteolytic enzymes capable of degrading elastin, particularly the gelatinases A (MMP-2) and B (MMP-9), which are 2 powerful elastases. The suppression of aneurysmal degeneration after TIMP-1 overexpression in this experimental model13 led to the concept that MMPs are terminal effectors of aneurysm formation/expansion in this setting. The mechanisms driving diameter expansion of already-formed AAAs are poorly documented,28 largely because it is difficult to reproduce further diameter increase in animal models of already-developed AAAs. As a working hypothesis, it can be accepted that the interplay between infiltrating T lymphocytes29,30 and macrophages31,32 results in an aggravation of matrix injury by proteases and ultimately in diameter increase. Neoepitopes exposed by ongoing matrix degradation33 and xenogenic proteins34 are proposed as stimuli perpetuating wall inflammation and destruction. In the xenograft model, xenoepitopes elicit wall infiltration by monocyte-macrophages (80%) and T lymphocytes (10%) in a ratio similar to that observed in atherosclerotic AAAs.29,35 In this work, gene-mediated overexpression of active TGF-;1 decreased T-lymphocyte and monocyte-macrophage infiltration, as well as MMP-2 and MMP-9. The preservation of elastic fibers in the media is a marker of the protective effect of TGF-;1 toward proteolytic injury. Data from explants of human origin show that the wall of atherosclerotic AAAs responds similarly to active TGF-;1 stimulation by a decrease in MMP-2 and MMP-9 release. These results, after TGF-;1 overexpression in experimental and atherosclerotic AAAs, are concordant with those obtained by opposite TGF-;1–blocking strategies in atherosclerotic mice,4,6 which have been shown to promote the development of plaques with a rupture-prone phenotype, ie, reduced extracellular matrix and increased inflammation.
The luminal thrombus responded to active TGF-;1 in experimental AAAs by decreased MMP-9 activity and mRNA and increased TIMP-1 mRNA, with a simultaneous marked decrease in T lymphocytes. Thrombi from atherosclerotic AAAs also responded to TGF-;1 stimulation by decreasing MMP-2 and MMP-9 released into conditioned medium, suggesting the participation of cells in this structure classically regarded as noncellular or harboring polymorphonuclear leukocytes.17 We observed that the surface of the thrombi from human AAAs harbors mesenchymal cells (E. Allaire, unpublished observation), which may respond to cytokine/growth factor stimulation.
Taken together, our data demonstrate that experimental and atherosclerotic AAAs respond to TGF-;1 by a downregulation of proteolysis and wall destruction.
Reconstruction of the Artery Wall
TGF-;1 has been described as a healing factor in various tissues by promoting cell growth and matrix accumulation.36 Atherosclerotic and experimental AAAs exhibit a decreased number of VSMCs in the media.37 We have shown previously that endovascular seeding of VSMCs stabilized the diameter of expanding experimental AAAs, supporting the concept that correction of the quantitative VSMC defect allows functional healing of AAAs.14,38 Accumulation of VSMCs is elicited in normal and injured vessel wall by TGF-;1 overexpression.20,21 In the present study, overexpression of active TGF-;1 resulted in the development of a VSMC-rich intimal thickening, which tended to replace the luminal thrombus and likely participated in wall strengthening because of its dense collagen and elastic fiber networks. Platelet-derived growth factor-BB, which is localized in the thrombus of experimental AAAs (E. Allaire, unpublished observation), has been shown to facilitate VSMC proliferation39 and collagen synthesis40 on TGF-;1 stimulation. Interaction between these 2 growth factors is currently being investigated in our laboratory.
Time- and Layer-Specific Impact of TGF-;1 Overexpression
Our data in experimental AAAs show that TGF-;1 expression remains high after endovascular gene delivery, even when transgene expression is exhausted at day 28. Endogenous rat TGF-;1 mRNA content was high in the aneurysmal wall itself at day 28. These data support the idea that TGF-;1 acts as a self-promoting cytokine in the aneurysmal wall and that endogenous synthesis relays delivered gene expression after its exhaustion. Active TGF-;1 overexpression appears to reprogram the aneurysmal vessel, resulting in a new wall structure able to withstand hemodynamic stress. This observation is relevant for future gene therapy approaches because it suggests that a time-limited overexpression of TGF-;1 by a short-living vector may be sufficient for long-term AAA stabilization.
Determination of which cells in AAAs are targeted by gene transfer requires further studies. Using a double-labeling approach, we have observed that both -actin– and ED1-positive cells demonstrated ;-galactosidase activity (J. Dai, E. Allaire, unpublished data). Although aneurysmal wall and thrombus respond in different ways to active TGF-;1, overexpression finally establishes a new interplay between these 2 layers, which allows recovery of stability in response to arterial hemodynamic stress.
Conclusions
We provide support for the view that TGF-;1 triggers multiple responses in injured wall and thrombus of AAAs. In a model reproducing the complexity and heterogeneity of aortic walls submitted to inflammation and proteolysis, active TGF-;1 overexpression is shown to promote stability of expanding AAAs. The clinical application of this first approach to stabilize expanding AAAs by endovascular delivery of a single multipotent gene requires further studies, particularly with other gene vectors applied to human atherosclerotic lesions.
Acknowledgments
This work was supported by the CNRS, l’Assistance Publique des H;pitaux de Paris, La Fondation de l’Avenir pour la Recherche Médicale Appliquée (grant ET9-210), by la Fondation de France, and by grant PROGRES from INSERM. We thank Dr Saez for providing us with anti–TGF-;1 antibody; the Vector Core of Nantes University Hospital, which is supported by the Association Fran;aise contre les Myopathies, for producing the adenoviral vectors used in this study; and Mary Osborne-Pellegrin for correction of the manuscript.
References
Shih SC, Ju M, Liu N, Mo JR, Ney JJ, Smith LE. Transforming growth factor ;1 induction of vascular endothelial growth factor receptor 1: mechanism of pericyte-induced vascular survival in vivo. Proc Natl Acad Sci U S A. 2003; 100: 15859–15864.
Carmeliet P. Angiogenesis in health and disease. Nat Med. 2003; 9: 653–660.
Wolf YG, Rasmussen LM, Ruoslahti E. Antibodies against transforming growth factor-;1 suppress intimal hyperplasia in a rat model. J Clin Invest. 1994; 93: 1172–1178.
Mallat Z, Gojova A, Marchiol-Fournigault C, Esposito B, Kamate C, Merval R, Fradelizi D, Tedgui A. Inhibition of transforming growth factor-; signaling accelerates atherosclerosis and induces an unstable plaque phenotype in mice. Circ Res. 2001; 89: 930–934.
Lutgens E, Gijbels M, Smook M, Heeringa P, Gotwals P, Koteliansky VE, Daemen MJ. Transforming growth factor-; mediates balance between inflammation and fibrosis during plaque progression. Arterioscler Thromb Vasc Biol. 2002; 22: 975–982.
Robertson AK, Rudling M, Zhou X, Gorelik L, Flavell RA, Hansson GK. Disruption of TGF-; signaling in T cells accelerates atherosclerosis. J Clin Invest. 2003; 112: 1342–1350.
Ribourtout E, Desfaits AC, Salazkin I, Raymond J. Ex vivo gene therapy with adenovirus-mediated transforming growth factor ;1 expression for endovascular treatment of aneurysm. J Vasc Surg. 2003; 38: 576–583.
Shah PK. Inflammation, metalloproteinases, and increased proteolysis: an emerging pathophysiological paradigm in aortic aneurysm. Circulation. 1997; 96: 2115–2117.
Powell JT, Greenhalgh RM. Clinical practice: small abdominal aortic aneurysms. N Engl J Med. 2003; 348: 1895–1901.
Fukui D, Miyagawa S, Soeda J, Tanaka K, Urayama H, Kawasaki S. Overexpression of transforming growth factor ;1 in smooth muscle cells of human abdominal aortic aneurysm. Eur J Vasc Endovasc Surg. 2003; 25: 540–545.
Josien R, Douillard P, Guillot C, Muschen M, Anegon I, Chetritt J, Menoret S, Vignes C, Soulillou JP, Cuturi MA. A critical role for transforming growth factor-beta in donor transfusion-induced allograft tolerance. J Clin Invest. 1998; 102: 1920–1926.
Allaire E, Guettier C, Bruneval P, Plissonnier D, Michel JB. Cell-free arterial grafts: morphologic characteristics of aortic isografts, allografts, and xenografts in rats. J Vasc Surg. 1994; 19: 446–456.
Allaire E, Forough R, Clowes MM, Starcher B, Clowes AW. Local overexpression of TIMP-1 prevents aortic aneurysm degeneration and rupture in a rat model. J Clin Invest. 1998; 102: 1413–1420.
Allaire E, Muscatelli-Groux B, Pagès C, Guinault A-M, Goussard A, Mandet C, Bruneval P, Melliere D, Becquemin JP. Vascular smooth muscle cell endovascular therapy stabilizes already developed aneurysms in a model of aortic injury elicited by inflammation and proteolysis. Ann Surg. 2004; 239: 417–427.
Avallet O, Gomez E, Vigier M, J;gou B, Saez JM. Sertoli cell-germ interactions and TGFb1 expression and secretion in vitro. Biochem Biophys Res Commun. 1997; 238: 905–909.
Walton LJ, Franklin IJ, Bayston T, Brown LC, Greenhalgh RM, Taylor GW, Powell JT. Inhibition of prostaglandin E2 synthesis in abdominal aortic aneurysms: implications for smooth muscle cell viability, inflammatory processes, and the expansion of abdominal aortic aneurysms. Circulation. 1999; 100: 48–54.
Fontaine V, Jacob MP, Houard X, Rossignol P, Plissonnier D, Angles-Cano E, Michel JB. Involvement of the mural thrombus as a site of protease release and activation in human aortic aneurysms. Am J Pathol. 2002; 161: 1701–1710.
Roberts AB, Sporn MB, Assoian RK, Smith JM, Roche NS, Wakefield LM, Heine UI, Liotta LA, Falanga V, Kehrl JH. Transforming growth factor type ;: rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen formation in vitro. Proc Natl Acad Sci U S A. 1986; 83: 4167–4171.
Sauvage M, Hinglais N, Mandet C, Badier C, Deslandes F, Michel JB, Jacob MP. Localization of elastin mRNA and TGF-;1 in rat aorta and caudal artery as a function of age. Cell Tissue Res. 1998; 291: 305–314.
Nabel EG, Shum L, Pompili VJ, Yang ZY, San H, Shu HB, Liptay S, Gold L, Gordon D, Derynck R. Direct transfer of transforming growth factor ;1 gene into arteries stimulates fibrocellular hyperplasia. Proc Natl Acad Sci U S A. 1993; 90: 10759–10763.
Majesky MW, Lindner V, Twardzik DR, Schwartz SM, Reidy MA. Production of transforming growth factor ;-1 during repair of arterial injury. J Clin Invest. 1991; 88: 904–910.
Hautmann MB, Adam PJ, Owens GK. Similarities and differences in smooth muscle alpha- induction by TGF-; in smooth muscle versus non-smooth muscle cells. Arterioscler Thromb Vasc Biol. 1999; 19: 2049–2058.
Vaday GG, Schor H, Rahat MA, Lahat N, Lider O. Transforming growth factor-; suppresses tumor necrosis factor -induced matrix metalloproteinase-9 expression in monocytes. J Leukoc Biol. 2001; 69: 613–621.
McCaffrey TA, Consigli S, Du B, Falcone DJ, Sanborn TA, Spokojny AM, Bush HL Jr. Decreased type II/type I TGF-; receptor ratio in cells derived from human atherosclerotic lesions: conversion from an antiproliferative to profibrotic response to TGF-;1. J Clin Invest. 1995; 96: 2667–2675.
Battegay EJ, Raines EW, Seifert RA, Bowen-Pope DF, Ross R. TGF-; induces bimodal proliferation of connective tissue cells via complex control of an autocrine PDGF loop. Cell. 1990; 63: 515–524.
Kazi M, Thyberg J, Religa P, Roy J, Hedin U, Swedenborg J. Influence of intraluminal thrombus on structural and cellular composition of abdominal aortic aneurysm wall. J Vasc Surg. 2003; 38: 1283–1292.
Allaire E, Hasenstab D, Clowes MM, Starcher B, Clowes AW. Overexpression of plasminogen activator-inhibitor prevents arterial extracellular matrix degradation by blocking matrix metalloproteinase activity. Circulation. 1998; 98: 249–255.
Juvonen J, Surcel HM, Satta J, Teppo AM, Bloigu A, Syrjala H, Airaksinen J, Leinonen M, Saikku P, Juvonen T. Elevated circulating levels of inflammatory cytokines in patients with abdominal aortic aneurysm. Arterioscler Thromb Vasc Biol. 1997; 17: 2843–2847.
Ocana E, Bohorquez JC, Perez-Requena J, Brieva JA, Rodriguez C. Characterisation of T and B lymphocytes infiltrating abdominal aortic aneurysms. Atherosclerosis. 2003; 170: 39–48.
Curci JA, Thompson RW. Adaptive cellular immunity in aortic aneurysms: cause, consequence, or context; J Clin Invest. 2004; 114: 168–171.
Bayston T, Ramessur S, Reise J, Jones KG, Powell JT. Prostaglandin E2 receptors in abdominal aortic aneurysm and human aortic smooth muscle cells. J Vasc Surg. 2003; 38: 354–359.
Curci JA, Liao S, Huffman MD, Shapiro SD, Thompson RW. Expression and localization of macrophage elastase (matrix metalloproteinase-12) in abdominal aortic aneurysms. J Clin Invest. 1998; 102: 1900–1910.
Hirose H, Ozsvath KJ, Xia S, Tilson MD. Molecular cloning of the complementary DNA for an additional member of the family of aortic aneurysm antigenic proteins. J Vasc Surg. 1997; 26: 313–318.
Halme S, Juvonen T, Laurila A, Saikku P, Surcel HM. Chlamydia pneumoniae reactive T lymphocytes in the walls of abdominal aortic aneurysms. Eur J Clin Invest. 1999; 29: 546–552.
Pearce WH, Koch AE. Cellular components and features of immune response in abdominal aortic aneurysms. Ann N Y Acad Sci. 1996; 800: 175–185.
Shah M, Revis D, Herrick S, Baillie R, Thorgeirson S, Ferguson M, Roberts A. Role of elevated plasma transforming growth factor-;1 levels in wound healing. Am J Pathol. 1999; 154: 1115–1124.
Henderson EL, Geng YJ, Sukhova GK, Whittemore AD, Knox J, Libby P. Death of smooth muscle cells and expression of mediators of apoptosis by T lymphocytes in human abdominal aortic aneurysms. Circulation. 1999; 99: 96–104.
Losy F, Dai J, Pages C, Ginat M, Muscatelli-Groux B, Guinault A-M, Rousselle E, Smedile G, Loisance D, Becquemin JP, Allaire E. Paracrine secretion of transforming growth factor-;1 in aneurysm healing and stabilization with endovascular smooth muscle cell therapy. J Vasc Surg. 2003; 37: 1301–1309.
Majack RA, Majesky MW, Goodman LV. Role of PDGF-A expression in the control of vascular smooth muscle cell growth by transforming growth factor-;. J Cell Biol. 1990; 111: 239–247.
Halloran BG, So BJ, Baxter BT. Platelet-derived growth factor is a cofactor in the induction of 1 (I) procollagen expression by transforming growth factor-; 1 in smooth muscle cells. J Vasc Surg. 1996; 23: 767–773.(Jianping Dai, MD, PhD; Fr)
Université Paris XII, UFR de Médecine (J.D., F.L., A.-M.G., C.P., P.D., J.-P.B., E.A.) and Service de Chirurgie Vasculaire et Endocrinienne, Assistance Publique des H;pitaux de Paris (J.D., P.D., J.-P.B., E.A.)
H;pital H. Mondor, Créteil
INSERM U643, Nantes (I.A.), France.
Abstract
Background— The cell response to transforming growth factor-;1 (TGF-;1), a multipotent cytokine with healing potential, varies according to tissue context. We have evaluated the ability of TGF-;1 overexpression by endovascular gene therapy to stabilize abdominal aortic aneurysms (AAAs) already injured by inflammation and proteolysis.
Methods and Results— Active TGF-;1 overexpression was obtained in already-developed experimental AAAs in rats after endovascular delivery of an adenoviral construct encoding for a mutated form of active simian TGF-;1 and in an explant model using human atherosclerotic AAA fragments incubated with recombinant active TGF-;1. Transient exogenous TGF-;1 overexpression by endovascular gene delivery was followed by induction of endogenous rat TGF-;1. Overexpression of active TGF-;1 in experimental AAAs was associated with diameter stabilization, preservation of medial elastin, decreased infiltration of monocyte-macrophages and T lymphocytes, and a decrease in matrix metalloproteinase-2 and -9, which was also observed in the explant model, in both thrombus and wall. In parallel with downregulation of the destructive process, active TGF-;1 overexpression triggered endoluminal reconstruction, replacing the thrombus by a vascular smooth muscle cell–, collagen-, and elastin-rich intima.
Conclusions— Local TGF-;1 self-induction after transient exogenous overexpression reprograms dilated aortas altered by inflammation and proteolysis and restores their ability to withstand arterial pressure without further dilation. This first demonstration of stabilization of expanding AAAs by delivery of a single multipotent self-promoting gene supports the view that endovascular gene therapy should be considered for treatment of aneurysms.
Key Words: aneurysm ; gene therapy ; transforming growth factor beta
Introduction
Transforming growth factor-;1 (TGF-;1), an important molecular determinant of wall strengthening during vascular development,1,2 mediates responses of vessels submitted to injury by promoting intimal tissue accumulation.3 Abrogation of TGF-; signaling shifts experimental atherosclerotic lesions toward an unstable, inflammatory plaque phenotype depleted in extracellular matrix.4–6
See p 939
Despite these observations and previous attempts,7 no report has documented the potential of TGF-;1 in reprogramming a dilated artery wall toward healing and stability. Abdominal aortic aneurysms (AAAs) show an inexorable tendency to expand under radial hemodynamic stress after artery wall extracellular matrix destruction by inflammation and proteolysis.8 AAAs are lesions in which one can test the ability of TGF-;1 to help inflammation-injured vessels return to normal function, eg, ability to withstand stress without expanding, with clinically relevant consequences because death risk by rupture is proportional to diameter.9
Because the response to TGF-;1 is modulated by other stimuli applied to cells, the question of whether TGF-;1 overexpression might help to control AAA expansion10 should be addressed in aortic tissues altered by ongoing inflammation and proteolysis. Accordingly, we have used already-developed experimental and human lesions to provide evidence that TGF-;1 overexpression induced by endovascular gene transfer, or incubation with recombinant protein, decreases proteolytic burden, slows destruction of the damaged wall, promotes wall reconstruction on the luminal side, and ultimately stabilizes AAA diameter.
Methods
Animals
Male Lewis rats and Hartley guinea pigs (Iffa Credo, Lyon, France) were housed and taken care of according to the European Union Standards. Intraperitoneal injections of 5 mg/100 g body weight of pentobarbital were used for anesthesia. Animals were fed a standard diet.
Gene Vectors
Vectors were derived from a serotype 5 human adenovirus, with E1- and E3-encoding region deletions, containing the cDNA of either Escherichia coli nls ;-galactosidase (Ad-LacZ) or a mutated active form of simian TGF-;1 (Ad-sTGF-;1),11 generously provided by Généthon, France. The constructs were obtained by homologous recombination between the plasmid pAdEasy-1 and a shuttle plasmid pTrack-CMV containing the genes of interest under the control of the cytomegalovirus long-terminal-repeat promoter. Single stocks of viruses from the 2 constructs were placed in aliquots and kept at –80°C until use.
Generation of Expanding Aortic Aneurysms
Guinea pig infrarenal aortas were decellularized using sodium dodecyl sulfate to obtain tubes of intact extracellular matrix12 that were sewn orthotopically into the aorta of 200-g male Lewis rats with 10-0 nylon interrupted sutures (xenograft).13 Fourteen days after implantation, a chimeric aneurysm (>50% diameter increase) had developed with a degraded extracellular matrix of guinea pig origin and cells and thrombus of rat origin.
Endovascular Vector Infusion in Already-Formed Aneurysms
Fourteen days after xenograft implantation, the experimental AAAs were isolated from blood flow by clamps during a second surgical procedure as described previously.14 The lumen was gently rinsed with 0.9% NaCl through a PE10 catheter introduced by means of an aortotomy performed downstream in the native aorta. The catheter was connected to a reservoir containing the viral suspension (0.8x109 IP per AAA) or the vehicle (0.9% NaCl) alone. The virus stock was delivered to the aneurysmal wall during a 20-minute infusion at a pressure of 80 mm Hg generated by a mercury manometer. After infusion, aortotomy and laparotomy were closed.
Euthanasia and Tissue Harvest
Aneurysms were harvested 3, 14, or 28 days after endovascular infusion. The aneurysm diameter was measured in the anesthetized animal using a grid in the eyepiece of the operative microscope. Rats were then euthanized by an overdose of intravenous pentobarbital. The aneurysmal lumen was gently rinsed with PBS, and the middle part was fixed in 70% ethanol, embedded in paraffin, and sectioned transversally at 5 μm. The 2 ends were snap-frozen in liquid nitrogen and kept at –80°C.
Histological and Immunohistochemical Analysis
Cross sections were stained with orcein for elastic fibers and sirius red for collagen. Percentages of stained surface were calculated with Perfect Image software. Polyclonal rabbit anti-synthetic TGF-;1 antibody (a generous gift from Dr Saez, INSERM, Lyon, France),15 goat anti–rat-TIMP-1 (RandD Sysytems), anti–Ser465-467–phosphorylated Smad 2 (Cell Signaling Technology), and 1F4 antibody for rat CD3 (T lymphocytes) (Oxford Biotechnology Ltd) were used with immunoperoxidase staining (Vectastain Elite ABC Kit, Vector Laboratories) counterstained with hematoxylin. Other primary antibodies were mouse anti-rat monoclonals: ED1 clone for monocyte-macrophages (Dakopatts) and 1A4 clone for -actin (smooth muscle cells) (Dakopatts) detected by alkaline phosphatase anti–alkaline phosphatase staining (Dakopatts). Negative controls were generated by omission of the primary antibody or with normal IgG or IgM (for anti–CD-3) (Sigma-Aldrich). Cells were counted using a grid in the microscope eyepiece.
Semiquantification of mRNA by RT-PCR
RNA was extracted from 6 aneurysms after separation of the luminal thrombus from the aneurysmal wall (media/adventitia). Reverse-transcription polymerase chain reaction (RT-PCR), comparative to the domestic gene 18s (QuantumRNA 18s Internal Standards Kit, Ambion), was used to determine mRNA levels. Two sets of primers, one for the gene of interest and one for 18s, were added to the same tube with cDNA and PCR mix. Because of the small amount of material available from each rat aneurysm, tissues were pooled by layers and groups. Total RNA was extracted with TRIzol (Life Technologies) and treated with grade I DNAse (Roche Molecular Biomedicals). RT was done with random primers, Superscript II (Life technologies), dNTP, dithiothreitol, and ribonuclease inhibitor (Roche Molecular Biochemicals). Then, 2 μg RT product was diluted to a final volume of 50 μL. Primers were as follows: matrix metalloproteinase-2 (MMP-2): 5'-CTATTCTGTCAGCACTTTGG-3'/5'-CAGACTTTGGTTCTCCAACTT-3'; MMP-9, 5'-CTGCGT-ATTTCCATTCATCTT-3'/5'-ATGCCTTTTATGTCGTCTTCA-3'; tissue inhibitor of metalloproteinase-1 (TIMP-1), 5'-CCCCAGA-AATCAACGAGAGACCA-3'/5'-ACACCCCACAGCCAGCACTAT-3'; collagen III, 5'-TGCCACCCTGAACTCAAGAG-3'/5'-GCCTTGC-GTGTTTGATATTC-3'; rat TGF-;1, 5'-CGGACTACTACGCCAAAG-AA-3'/5'-TCAAAAGACAGCCACTCAGG-3'; and primate TGF-;1, 5'-AACACATCAGAGCTCCGAGAA-3'/5'-GTCAATGTACAGCTGCCGCAC-3' (Genset Oligos SA). Final PCR conditions were chosen to avoid interference between the 2 sets of primers and to obtain a signal within a linear range of amplification with Taq polymerase (EurobioTaq, Eurobio). Negative controls were done without Superscript II. PCR products (10 μL) were run in a 2% agarose gel with 5 μg/mL ethidium bromide visualized under UV light by a video camera. Bands of amplified sequences corresponding to the gene of interest and to 18s were quantified with Gel Analyst. Results were expressed as a ratio between signals corresponding to the gene of interest and 18s.
Incubation of Human Aneurysmal Tissues With Recombinant Active TGF-;1
Fragments from the maximum dilation zone of 5 human atherosclerotic AAAs (asymptomatic; diameter >50 mm) were collected during elective surgery in the Department of Vascular Surgery of the Henri Mondor hospital in agreement with the local ethics committee. The thrombus was immediately separated from the aneurysmal wall; both were kept in M199 with antibiotics but without serum at 4°C and were sterilely cut into small fragments (<2 mm3) as described previously.16 Explants of thrombus and wall were separately cultured in serum-free culture medium (50% M199, 50% RPMI) with increasing doses of recombinant human active TGF-;1 (rhTGF-;1) (Peprotech) for 24 hours. Quantification of human TIMP-1 protein in explant-conditioned media was done with an ELISA kit (RandD Systems, Inc) according to the manufacturer’s instructions.
Quantification of Gelatinases and TIMP-1
MMP-2 and MMP-9 activities in experimental AAAs, after separation of thrombus/intima from media/adventitia and extraction with a guanidine buffer, and in explant-conditioned media from 5 human atherosclerotic AAAs were quantified on 1% gelatin zymograms.13 MMP-2– and MMP-9–related bands were quantified with ImageMaster software (Pharmacia Biotech). Results were plotted against a standard curve generated with normal rat arteries for intergel normalization and with recombinant MMP-2 and MMP-9 (Medgene Science). Quantification of rat TIMP-1 protein was done with a DuoSet ELISA kit (RandD Systems Europe).
Statistical Analysis
Results are expressed as mean±SD. The percentage of diameter increase was calculated as follows: (diameter at harvest–diameter at infusion)x100/diameter at infusion. Comparisons between 2 groups were made by use of the nonparametric Mann-Whitney U test and between 3 groups with the nonparametric Kruskall-Wallis test; correlations between continuous variables were made with the Z correlation test (Statview, version 4.5). Values of P<0.05 was considered significant.
Results
TGF-;1 Overexpression Stops Expansion of Already-Formed AAAs
Anti–TGF-;1 immunostaining on cross sections showed a stable, albeit small, number of positive cells in AAAs after Ad-LacZ infusion (Figure 1A and 1D). As early as 3 days after Ad-sTGF-;1 infusion, the number of TGF-;1–positive cells increased 4-fold in the thrombus and media/adventitia (thrombus, from 0.25±0.16 to 1.0±0.56x103 cells/mm2 for Ad-LacZ and Ad-sTGF-;1, respectively, P<0.05; media/adventitia, from 0.16±0.17 to 0.59±0.21x103 cells/mm2, respectively, P<0.05) (Figure 1A). In the media/adventitia, ie, in the aneurysmal wall itself, overexpression of TGF-;1 was detected up to 28 days after Ad-sTGF-;1 infusion (Figure 1A) despite the decline in sTGF-;1 mRNA content observed by RT-PCR using primers specific to primate TGF-;1 (Figure 1B). RT-PCR with rat-specific primers showed a 6-fold increase in rat TGF-;1 mRNA content 28 days after Ad-TGF-;1 infusion, suggesting an induction of endogenous rat TGF-;1 after endovascular delivery of the sTGF-;1 gene (Figure 1C). Immunostaining with the antibody to phosphorylated Smad2, a marker of high-affinity serine-threonine TGF-; receptor activation, was markedly increased in cell nuclei 3 days after Ad-sTGF-;1 delivery, suggesting effective TGF-;1 signaling in this group (Figure 1D). These results demonstrate an early, significant, and sustained increase in TGF-;1 with efficient signal transduction on endovascular active sTGF-;1 gene transfer.
Infusion of Ad-LacZ instead of vehicle (0.9% NaCl) had no significant impact on AAA diameter expansion (28-day diameter increase: Ad-LacZ [n=8], 72.6±39.6%; vehicle [n=7], 57.1±26.8%; P=0.42). Mean external diameter of AAAs infused with Ad-LacZ had increased significantly 28 days later, from 2.8±0.6 to 4.8±1.6 mm (n=8; P=0.04), demonstrating disease progression in this control group and in vehicle-infused AAAs (data not shown). In contrast, mean AAA diameters remained stable 28 days after endovascular infusion of Ad-sTGF-;1 (from 3.1±0.5 to 3.3±0.6 mm; n=8; P=NS) (Figure 1E). The mean percentage of diameter increase after 28 days was 4.7±11.8 in Ad-sTGF-;1–infused vessels (comparisons of 28-day percent diameter increase: between Ad-LacZ and Ad-sTGF-;1, P=0.0011; between vehicle and Ad-sTGF-;1, P=0.0018; between the 3 groups [Kruskall-Wallis test], P=0.0008). These data demonstrate that overexpression of active TGF-;1 by the adenoviral vector prevented AAA expansion.
TGF-;1 Overexpression Decreases Medial Injury
Extracellular Matrix and Inflammation
Overexpression of TGF-;1 led to preservation of media elastic fiber network compared with controls (elastic fiber surface, 9.55±0.80% versus 19.71±1.28%, 28 days after Ad-LacZ and Ad-sTGF-;1 endovascular infusion, respectively; P=0.034) (Figure 2A). Overexpression of TGF-;1 was also followed by a 2-fold decrease in monocyte-macrophage density in the media/adventitia as early as 3 days after endovascular gene delivery (2.58±0.12 versus 1.33±0.35x 103 ED1-positive cells per 1 mm2 in Ad-LacZ– and Ad-sTGF-;1–infused AAAs, respectively; P=0.02) and up to 28 days after infusion (Figure 2B and 2C). TGF-;1 overexpression was also associated with a marked decrease in T lymphocyte infiltration 28 days after infusion (Figure 2D).
MMP-Dependent Proteolytic Balance in Experimental and Human Atherosclerotic AAAs
In experimental AAAs, overexpression of active TGF-;1 was associated with a 2- to 4-fold decrease in total and active MMP-9 at day 3 in both the thrombus and media/adventitia (Figure 3A). Because MMP-9 activity detected in AAAs may not originate entirely from synthesis in the artery wall itself,17 we used RT-PCR to examine MMP-9 mRNA sources in the thrombus and aneurysmal wall (ie, media/adventitia) after gene delivery. The most important effect of active TGF-;1 overexpression was a marked decrease in MMP-9 mRNA content (100-fold) in the thrombus of AAAs 3 days after Ad-sTGF-;1 infusion (Figure 3A). MMP-2 activity and mRNA content were decreased 3 days after TGF-;1 overexpression in the media/adventitia, but mRNA had increased slightly in the thrombus (Figure 3B). Because MMP activity in AAAs is also regulated by the inhibitors TIMPs,13 we examined TIMP-1 mRNA content and observed a massive increase (50-fold) in the thrombus 3 days after TGF-;1 overexpression and a 3-fold decrease in the media/adventitia (Figure 3C). TIMP-1 protein was also induced after TGF-;1 expression, as shown by immunohistochemistry at day 3 (Figure 3C) and ELISA on AAA extracts 28 days after gene transfer (26.1±3.8 and 32.7±3.2 ng/mg in Ad-LacZ and Ad-sTGF-;1 groups, respectively; P=0.049).
To demonstrate the relevance of the data obtained in the animal model, we have evaluated the impact of TGF-;1 on MMP and TIMP-1 production by human atherosclerotic AAA explants in culture. Addition of rhTGF-;1 decreased MMP-9 and MMP-2 activity after a 24-hour culture of explants (Figure 4) in conditioned medium from both aneurysmal thrombi (n=5) and walls (n=5). This decrease was observed for all forms of MMP-2 and MMP-9 (data not shown), activated MMP-9 (20 ng/mL rhTGF-;1: thrombus, –34.2±17.8%, P=0.001; wall, –37.4±17.7%, P=0.003), and activated MMP-2 (20 ng/mL rhTGF-;1: thrombus, –50.4±48.1%, P=0.007; wall, –75.0±22.7%, P=0.003) with reference to the media without rhTGF-;1. The decrease in MMP-9 and MMP-2 activity was TGF-;1 dose dependent (Z correlation test; active MMP-9: thrombus, P=0.001; wall, P=0.003; active MMP-2: thrombus, P=0.072; wall, P=0.003). Incubation with rhTGF-;1 had no impact on human TIMP-1 concentration in the conditioned medium (data not shown).
Overexpression of TGF-;1 Promotes Aneurysmal Wall Reconstruction
TGF-;1 overexpression was followed by the development of an intima rich in vascular smooth muscle cells (VSMCs) as demonstrated by anti–-actin immunostaining starting 14 days after Ad-sTGF-;1 delivery (Figure 5A and 5D). VSMCs replaced the preexisting endoluminal thrombus in Ad-sTGF-;1–infused AAAs, whereas thrombus remained the main structure in contact with circulating blood in AAAs infused with Ad-LacZ (Figure 5D). Overexpression of TGF-;1 was followed by an accumulation of collagen in the intima replacing the thrombus (collagen surface at day 28, 8.2±5.4% versus 30.9±17.5% in Ad-LacZ– versus Ad-sTGF-;1–infused AAAs, respectively; P=0.02) (Figure 5B and 5D). Type III collagen mRNA content was increased in the neointima 14 and 28 days after TGF-;1 overexpression (Figure 5C) but not in the media/adventitia (data not shown). In some areas at day 28 after Ad-sTGF-;1 infusion, elastic fibers were detected by orcein staining, mostly on the luminal side of the intima, whereas virtually none were detected in Ad-LacZ–infused vessels (Figure 5D). Overall, the development of a tissue made of VSMCs surrounded by a dense extracellular matrix network on the luminal side of the AAA could be interpreted as wall reconstruction.
Discussion
TGF-;1 is a multipotent cytokine that induces expression of fibrillar collagen18 and elastin19 genes. Under defined circumstances, TGF-;1 also downregulates inflammation, promotes VSMC growth20,21 and differentiation,22 and inhibits MMP-dependent proteolysis.23 We evaluated the impact of active TGF-;1 overexpression on the remodeling process of already-developed AAAs. Because the response to TGF-; is conditioned by costimulatory factors and tissue context,24,25 we overexpressed TGF-;1 in inflammatory and proteolytically injured aortas, ie, in already-formed experimental AAAs in vivo and in human AAAs in vitro. Aneurysmal degeneration of aortic xenografts is reproducible with a low variability, allowing assessment of the impact of TGF-;1 overexpression on diameter, the remodeling criterion used to predict rupture in the human disease.9 Here, we document the efficiency of the endovascular route for gene delivery to overexpress active TGF-;1 in experimental AAAs. At the interface between blood and aorta injured by inflammation and proteolysis, both experimental14 and atherosclerotic AAAs17 display a luminal thrombus that has recently been shown to be associated with wall injury.26 Accordingly, we analyzed separately these 2 components of AAAs with respect to their responsiveness to TGF-;1. We provide evidence that overexpression of a mutated active form of TGF-;1 decreases the proteolytic burden and the inflammatory and destructive process in the aneurysmal wall in experimental and human AAAs; promotes wall reconstruction by inducing the formation of a VSMC-, collagen-, and elastin-rich intima replacing the luminal thrombus; and ultimately stabilizes the diameter of expanding, already-formed AAAs in vivo. The stabilization of AAA diameter after Ad-sTGF-;1 infusion was observed compared with both Ad-LacZ infusion (vector without the gene of interest) and infusion of vehicle alone. Our results support the idea that overexpression of active TGF-;1 adapts the structure of expanding aortic walls injured by inflammation and proteolysis and reestablishes their ability to resist wall stress without further dilation.
Inhibition of Wall Destruction
Injury of the medial layer by inflammation and proteolysis is the hallmark of human atherosclerotic AAAs8 and is reproduced in the xenograft aneurysm model.12–14,27 In both conditions, injury of the elastic fiber network is observed with upregulation of proteolytic enzymes capable of degrading elastin, particularly the gelatinases A (MMP-2) and B (MMP-9), which are 2 powerful elastases. The suppression of aneurysmal degeneration after TIMP-1 overexpression in this experimental model13 led to the concept that MMPs are terminal effectors of aneurysm formation/expansion in this setting. The mechanisms driving diameter expansion of already-formed AAAs are poorly documented,28 largely because it is difficult to reproduce further diameter increase in animal models of already-developed AAAs. As a working hypothesis, it can be accepted that the interplay between infiltrating T lymphocytes29,30 and macrophages31,32 results in an aggravation of matrix injury by proteases and ultimately in diameter increase. Neoepitopes exposed by ongoing matrix degradation33 and xenogenic proteins34 are proposed as stimuli perpetuating wall inflammation and destruction. In the xenograft model, xenoepitopes elicit wall infiltration by monocyte-macrophages (80%) and T lymphocytes (10%) in a ratio similar to that observed in atherosclerotic AAAs.29,35 In this work, gene-mediated overexpression of active TGF-;1 decreased T-lymphocyte and monocyte-macrophage infiltration, as well as MMP-2 and MMP-9. The preservation of elastic fibers in the media is a marker of the protective effect of TGF-;1 toward proteolytic injury. Data from explants of human origin show that the wall of atherosclerotic AAAs responds similarly to active TGF-;1 stimulation by a decrease in MMP-2 and MMP-9 release. These results, after TGF-;1 overexpression in experimental and atherosclerotic AAAs, are concordant with those obtained by opposite TGF-;1–blocking strategies in atherosclerotic mice,4,6 which have been shown to promote the development of plaques with a rupture-prone phenotype, ie, reduced extracellular matrix and increased inflammation.
The luminal thrombus responded to active TGF-;1 in experimental AAAs by decreased MMP-9 activity and mRNA and increased TIMP-1 mRNA, with a simultaneous marked decrease in T lymphocytes. Thrombi from atherosclerotic AAAs also responded to TGF-;1 stimulation by decreasing MMP-2 and MMP-9 released into conditioned medium, suggesting the participation of cells in this structure classically regarded as noncellular or harboring polymorphonuclear leukocytes.17 We observed that the surface of the thrombi from human AAAs harbors mesenchymal cells (E. Allaire, unpublished observation), which may respond to cytokine/growth factor stimulation.
Taken together, our data demonstrate that experimental and atherosclerotic AAAs respond to TGF-;1 by a downregulation of proteolysis and wall destruction.
Reconstruction of the Artery Wall
TGF-;1 has been described as a healing factor in various tissues by promoting cell growth and matrix accumulation.36 Atherosclerotic and experimental AAAs exhibit a decreased number of VSMCs in the media.37 We have shown previously that endovascular seeding of VSMCs stabilized the diameter of expanding experimental AAAs, supporting the concept that correction of the quantitative VSMC defect allows functional healing of AAAs.14,38 Accumulation of VSMCs is elicited in normal and injured vessel wall by TGF-;1 overexpression.20,21 In the present study, overexpression of active TGF-;1 resulted in the development of a VSMC-rich intimal thickening, which tended to replace the luminal thrombus and likely participated in wall strengthening because of its dense collagen and elastic fiber networks. Platelet-derived growth factor-BB, which is localized in the thrombus of experimental AAAs (E. Allaire, unpublished observation), has been shown to facilitate VSMC proliferation39 and collagen synthesis40 on TGF-;1 stimulation. Interaction between these 2 growth factors is currently being investigated in our laboratory.
Time- and Layer-Specific Impact of TGF-;1 Overexpression
Our data in experimental AAAs show that TGF-;1 expression remains high after endovascular gene delivery, even when transgene expression is exhausted at day 28. Endogenous rat TGF-;1 mRNA content was high in the aneurysmal wall itself at day 28. These data support the idea that TGF-;1 acts as a self-promoting cytokine in the aneurysmal wall and that endogenous synthesis relays delivered gene expression after its exhaustion. Active TGF-;1 overexpression appears to reprogram the aneurysmal vessel, resulting in a new wall structure able to withstand hemodynamic stress. This observation is relevant for future gene therapy approaches because it suggests that a time-limited overexpression of TGF-;1 by a short-living vector may be sufficient for long-term AAA stabilization.
Determination of which cells in AAAs are targeted by gene transfer requires further studies. Using a double-labeling approach, we have observed that both -actin– and ED1-positive cells demonstrated ;-galactosidase activity (J. Dai, E. Allaire, unpublished data). Although aneurysmal wall and thrombus respond in different ways to active TGF-;1, overexpression finally establishes a new interplay between these 2 layers, which allows recovery of stability in response to arterial hemodynamic stress.
Conclusions
We provide support for the view that TGF-;1 triggers multiple responses in injured wall and thrombus of AAAs. In a model reproducing the complexity and heterogeneity of aortic walls submitted to inflammation and proteolysis, active TGF-;1 overexpression is shown to promote stability of expanding AAAs. The clinical application of this first approach to stabilize expanding AAAs by endovascular delivery of a single multipotent gene requires further studies, particularly with other gene vectors applied to human atherosclerotic lesions.
Acknowledgments
This work was supported by the CNRS, l’Assistance Publique des H;pitaux de Paris, La Fondation de l’Avenir pour la Recherche Médicale Appliquée (grant ET9-210), by la Fondation de France, and by grant PROGRES from INSERM. We thank Dr Saez for providing us with anti–TGF-;1 antibody; the Vector Core of Nantes University Hospital, which is supported by the Association Fran;aise contre les Myopathies, for producing the adenoviral vectors used in this study; and Mary Osborne-Pellegrin for correction of the manuscript.
References
Shih SC, Ju M, Liu N, Mo JR, Ney JJ, Smith LE. Transforming growth factor ;1 induction of vascular endothelial growth factor receptor 1: mechanism of pericyte-induced vascular survival in vivo. Proc Natl Acad Sci U S A. 2003; 100: 15859–15864.
Carmeliet P. Angiogenesis in health and disease. Nat Med. 2003; 9: 653–660.
Wolf YG, Rasmussen LM, Ruoslahti E. Antibodies against transforming growth factor-;1 suppress intimal hyperplasia in a rat model. J Clin Invest. 1994; 93: 1172–1178.
Mallat Z, Gojova A, Marchiol-Fournigault C, Esposito B, Kamate C, Merval R, Fradelizi D, Tedgui A. Inhibition of transforming growth factor-; signaling accelerates atherosclerosis and induces an unstable plaque phenotype in mice. Circ Res. 2001; 89: 930–934.
Lutgens E, Gijbels M, Smook M, Heeringa P, Gotwals P, Koteliansky VE, Daemen MJ. Transforming growth factor-; mediates balance between inflammation and fibrosis during plaque progression. Arterioscler Thromb Vasc Biol. 2002; 22: 975–982.
Robertson AK, Rudling M, Zhou X, Gorelik L, Flavell RA, Hansson GK. Disruption of TGF-; signaling in T cells accelerates atherosclerosis. J Clin Invest. 2003; 112: 1342–1350.
Ribourtout E, Desfaits AC, Salazkin I, Raymond J. Ex vivo gene therapy with adenovirus-mediated transforming growth factor ;1 expression for endovascular treatment of aneurysm. J Vasc Surg. 2003; 38: 576–583.
Shah PK. Inflammation, metalloproteinases, and increased proteolysis: an emerging pathophysiological paradigm in aortic aneurysm. Circulation. 1997; 96: 2115–2117.
Powell JT, Greenhalgh RM. Clinical practice: small abdominal aortic aneurysms. N Engl J Med. 2003; 348: 1895–1901.
Fukui D, Miyagawa S, Soeda J, Tanaka K, Urayama H, Kawasaki S. Overexpression of transforming growth factor ;1 in smooth muscle cells of human abdominal aortic aneurysm. Eur J Vasc Endovasc Surg. 2003; 25: 540–545.
Josien R, Douillard P, Guillot C, Muschen M, Anegon I, Chetritt J, Menoret S, Vignes C, Soulillou JP, Cuturi MA. A critical role for transforming growth factor-beta in donor transfusion-induced allograft tolerance. J Clin Invest. 1998; 102: 1920–1926.
Allaire E, Guettier C, Bruneval P, Plissonnier D, Michel JB. Cell-free arterial grafts: morphologic characteristics of aortic isografts, allografts, and xenografts in rats. J Vasc Surg. 1994; 19: 446–456.
Allaire E, Forough R, Clowes MM, Starcher B, Clowes AW. Local overexpression of TIMP-1 prevents aortic aneurysm degeneration and rupture in a rat model. J Clin Invest. 1998; 102: 1413–1420.
Allaire E, Muscatelli-Groux B, Pagès C, Guinault A-M, Goussard A, Mandet C, Bruneval P, Melliere D, Becquemin JP. Vascular smooth muscle cell endovascular therapy stabilizes already developed aneurysms in a model of aortic injury elicited by inflammation and proteolysis. Ann Surg. 2004; 239: 417–427.
Avallet O, Gomez E, Vigier M, J;gou B, Saez JM. Sertoli cell-germ interactions and TGFb1 expression and secretion in vitro. Biochem Biophys Res Commun. 1997; 238: 905–909.
Walton LJ, Franklin IJ, Bayston T, Brown LC, Greenhalgh RM, Taylor GW, Powell JT. Inhibition of prostaglandin E2 synthesis in abdominal aortic aneurysms: implications for smooth muscle cell viability, inflammatory processes, and the expansion of abdominal aortic aneurysms. Circulation. 1999; 100: 48–54.
Fontaine V, Jacob MP, Houard X, Rossignol P, Plissonnier D, Angles-Cano E, Michel JB. Involvement of the mural thrombus as a site of protease release and activation in human aortic aneurysms. Am J Pathol. 2002; 161: 1701–1710.
Roberts AB, Sporn MB, Assoian RK, Smith JM, Roche NS, Wakefield LM, Heine UI, Liotta LA, Falanga V, Kehrl JH. Transforming growth factor type ;: rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen formation in vitro. Proc Natl Acad Sci U S A. 1986; 83: 4167–4171.
Sauvage M, Hinglais N, Mandet C, Badier C, Deslandes F, Michel JB, Jacob MP. Localization of elastin mRNA and TGF-;1 in rat aorta and caudal artery as a function of age. Cell Tissue Res. 1998; 291: 305–314.
Nabel EG, Shum L, Pompili VJ, Yang ZY, San H, Shu HB, Liptay S, Gold L, Gordon D, Derynck R. Direct transfer of transforming growth factor ;1 gene into arteries stimulates fibrocellular hyperplasia. Proc Natl Acad Sci U S A. 1993; 90: 10759–10763.
Majesky MW, Lindner V, Twardzik DR, Schwartz SM, Reidy MA. Production of transforming growth factor ;-1 during repair of arterial injury. J Clin Invest. 1991; 88: 904–910.
Hautmann MB, Adam PJ, Owens GK. Similarities and differences in smooth muscle alpha- induction by TGF-; in smooth muscle versus non-smooth muscle cells. Arterioscler Thromb Vasc Biol. 1999; 19: 2049–2058.
Vaday GG, Schor H, Rahat MA, Lahat N, Lider O. Transforming growth factor-; suppresses tumor necrosis factor -induced matrix metalloproteinase-9 expression in monocytes. J Leukoc Biol. 2001; 69: 613–621.
McCaffrey TA, Consigli S, Du B, Falcone DJ, Sanborn TA, Spokojny AM, Bush HL Jr. Decreased type II/type I TGF-; receptor ratio in cells derived from human atherosclerotic lesions: conversion from an antiproliferative to profibrotic response to TGF-;1. J Clin Invest. 1995; 96: 2667–2675.
Battegay EJ, Raines EW, Seifert RA, Bowen-Pope DF, Ross R. TGF-; induces bimodal proliferation of connective tissue cells via complex control of an autocrine PDGF loop. Cell. 1990; 63: 515–524.
Kazi M, Thyberg J, Religa P, Roy J, Hedin U, Swedenborg J. Influence of intraluminal thrombus on structural and cellular composition of abdominal aortic aneurysm wall. J Vasc Surg. 2003; 38: 1283–1292.
Allaire E, Hasenstab D, Clowes MM, Starcher B, Clowes AW. Overexpression of plasminogen activator-inhibitor prevents arterial extracellular matrix degradation by blocking matrix metalloproteinase activity. Circulation. 1998; 98: 249–255.
Juvonen J, Surcel HM, Satta J, Teppo AM, Bloigu A, Syrjala H, Airaksinen J, Leinonen M, Saikku P, Juvonen T. Elevated circulating levels of inflammatory cytokines in patients with abdominal aortic aneurysm. Arterioscler Thromb Vasc Biol. 1997; 17: 2843–2847.
Ocana E, Bohorquez JC, Perez-Requena J, Brieva JA, Rodriguez C. Characterisation of T and B lymphocytes infiltrating abdominal aortic aneurysms. Atherosclerosis. 2003; 170: 39–48.
Curci JA, Thompson RW. Adaptive cellular immunity in aortic aneurysms: cause, consequence, or context; J Clin Invest. 2004; 114: 168–171.
Bayston T, Ramessur S, Reise J, Jones KG, Powell JT. Prostaglandin E2 receptors in abdominal aortic aneurysm and human aortic smooth muscle cells. J Vasc Surg. 2003; 38: 354–359.
Curci JA, Liao S, Huffman MD, Shapiro SD, Thompson RW. Expression and localization of macrophage elastase (matrix metalloproteinase-12) in abdominal aortic aneurysms. J Clin Invest. 1998; 102: 1900–1910.
Hirose H, Ozsvath KJ, Xia S, Tilson MD. Molecular cloning of the complementary DNA for an additional member of the family of aortic aneurysm antigenic proteins. J Vasc Surg. 1997; 26: 313–318.
Halme S, Juvonen T, Laurila A, Saikku P, Surcel HM. Chlamydia pneumoniae reactive T lymphocytes in the walls of abdominal aortic aneurysms. Eur J Clin Invest. 1999; 29: 546–552.
Pearce WH, Koch AE. Cellular components and features of immune response in abdominal aortic aneurysms. Ann N Y Acad Sci. 1996; 800: 175–185.
Shah M, Revis D, Herrick S, Baillie R, Thorgeirson S, Ferguson M, Roberts A. Role of elevated plasma transforming growth factor-;1 levels in wound healing. Am J Pathol. 1999; 154: 1115–1124.
Henderson EL, Geng YJ, Sukhova GK, Whittemore AD, Knox J, Libby P. Death of smooth muscle cells and expression of mediators of apoptosis by T lymphocytes in human abdominal aortic aneurysms. Circulation. 1999; 99: 96–104.
Losy F, Dai J, Pages C, Ginat M, Muscatelli-Groux B, Guinault A-M, Rousselle E, Smedile G, Loisance D, Becquemin JP, Allaire E. Paracrine secretion of transforming growth factor-;1 in aneurysm healing and stabilization with endovascular smooth muscle cell therapy. J Vasc Surg. 2003; 37: 1301–1309.
Majack RA, Majesky MW, Goodman LV. Role of PDGF-A expression in the control of vascular smooth muscle cell growth by transforming growth factor-;. J Cell Biol. 1990; 111: 239–247.
Halloran BG, So BJ, Baxter BT. Platelet-derived growth factor is a cofactor in the induction of 1 (I) procollagen expression by transforming growth factor-; 1 in smooth muscle cells. J Vasc Surg. 1996; 23: 767–773.(Jianping Dai, MD, PhD; Fr)