Intimal hyperplasia (IH) is the major cause of stenosis of vein grafts. Drugs such as statins prevent stenosis, but their systemic administration has limited effects. We developed a hyaluronic acid hydrogel matrix, which ensures a controlled release of atorvastatin (ATV) at the site of injury. The release kinetics demonstrated that 100% of ATV was released over 10 hours, independent of the loading concentration of the hydrogel. We investigated the effects of such a delivery on primary vascular smooth muscle cells isolated from human veins. ATV decreased the proliferation, migration, and passage of human smooth muscle cells (HSMCs) across a matrix barrier in a similar dose-dependent (5–10 µM) and time-dependent manner (24–72 hours), whether the drug was directly added to the culture medium or released from the hydrogel. Expression analysis of genes known to be involved in the development of IH demonstrated that the transcripts of both the gap junction protein connexin43 (Cx43) and plasminogen activator inhibitor-1 (PAI-1) were decreased after a 24–48-hour exposure to the hydrogel loaded with ATV, whereas the transcripts of the heme oxygenase (HO-1) and the inhibitor of tissue plasminogen activator were increased. At the protein level, Cx43, PAI-1, and metalloproteinase-9 expression were decreased, whereas HO-1 was upregulated in the presence of ATV. The data demonstrate that ATV released from a hydrogel has effects on HSMCs similar to the drug being freely dissolved in the environment.
Open surgical revascularization is still frequently the best option to treat coronary, lower limb, or cerebrovascular occlusive disease. Nevertheless, restenosis is a major reason of failure in 20–50% of these grafts, leading to the partial or complete occlusion of the anastomosis sites (Hwang et al., 2011). Postvascular intervention stenosis results mainly from intimal hyperplasia (IH) (Shah et al., 2003; Hwang et al., 2012), i.e., the thickening of the tunica intima due to proliferation of vascular smooth muscle cells (VSMCs), and from arterial remodeling, i.e., the rapid alteration of vein grafts implanted into arterial circulation (Ward et al., 2000). IH is an adaptive process that occurs in response to hemodynamic stress and injuries and following bypass graft interventions on arteries, veins, and artificial prosthesis (Zalewski et al., 2002; Sugimoto et al., 2009). It is characterized by the hyperproliferation and migration of VSMCs into the subintimal region and by an increase in matrix proteins, which together thicken the tunica intima (Newby, 1997) at the site of injury. This biological cascade is the main trigger of the dedifferentiation of poorly proliferating, contractile VSMCs into fast proliferating cells secreting extracellular matrix (Alexander and Owens, 2012; Nguyen et al., 2013).
Drugs preventing intraluminal vessel narrowing have been previously identified using an endovascular platform, such as a stent or a plain balloon (Schachner et al., 2006; Wiedemann et al., 2012). However, when open revascularization is mandatory, no platform is available for the local delivery of a drug. Thus, current treatments involve the repeat systemic administration of the active compound, which markedly increases its side effects (Wiedemann et al., 2012). These problems may be decreased by local application of a drug depot at the site of the surgery. Ideally, such a biocompatible depot should sustain a controlled and stable release of the active drug for a time sufficient to revert the phenotype of altered VSMCs, and should be biodegradable to avoid the need of a second surgery for its elimination. Here, we have investigated a hyaluronic acid hydrogel matrix, which can be loaded with a variety of drugs known to inhibit the proliferation and migration of vascular smooth muscle cells, as well as inflammation of the vessel wall (Baek et al., 2012a,b). Among these drugs, the 3-hydroxy-3-methylglutaryl–CoA reductase inhibitors, statins, have been shown to be fairly effective in preventing postsurgery stenosis (Qiang et al., 2012). However, the systemic administration of statins has limited effects on this prevention (Stettler et al., 2007), and undesirable side effects have been reported (Taylor et al., 2013). Therefore, we tested the effect of atorvastatin (ATV) after loading in a hyaluronic acid hydrogel on the proliferation, migration, and invasiveness of primary smooth muscle cells derived from human saphenous veins. The data document that the hydrogel suitably supports the local delivery of drugs, inasmuch as it allowed for a gradual release of ATV over 6 h, which resulted in several modifications of the VSMC phenotype, similar to those induced by the drug freely dissolved in the cell environment.
Material and Methods
Preparation of the Atorvastatin-Loaded Hydrogel.
Hyaluronic acid gels (Fortélis Extra, kindly provided by Anteis, Geneva, Switzerland), consisting of 25.5 mg/ml cross-linked hyaluronic acid obtained from biofermentation and suspended in phosphate buffer, were provided by Anteis (Geneva, Switzerland) and used as received. Calcium ATV, obtained from Chemos GmbH (Regenstauf, Germany), was dissolved in 33% ethanol aqueous solution and incorporated by gentle stirring in the hydrogel at the desired concentration. The gel was freeze-dried in a Telstar LyoBeta 15 (Telstar, Terrassa, Spain) using primary drying at −40°C under 0.2 mbar for 1 hour, followed by 10-hour secondary drying at 20°C to eliminate both alcohol and water. It was then reconstituted to the initial volume with sterile ultrapure water (MilliQ academic; Millipore, Billerica, MA) over a period of 24 hours. This rehydration restored the macroscopic transparency and viscosity of the unloaded gel, which could be easily disposed with a syringe. For cell culture experiments, the preparation was made under sterile conditions in a laminar flow (Steag LFH07.15; Luftechnik+Metallbau AG, Wettingen, Switzerland) using autoclaved materials.
In Vitro Release of ATV.
The in vitro release of ATV was determined under sink conditions. To this end, 100 µl of ATV-loaded hydrogel was placed in the 24 wells of a cell culture plate and covered with a dialysis membrane with a cutoff of 14 kDa (Merck, Darmstadt, Germany), which was maintained in place using a silicon O-ring. Five hundred microliters of RPMI 1640 (Invitrogen, Carlsbad, CA) culture medium supplemented with 10% fetal calf serum (FCS), and 1% penicillin/streptomycin was added to each well. At defined time intervals, 200 μl of medium was sampled and replaced with the same volume of fresh medium. Samples were extracted with 300 μl of acetonitrile and analyzed by high-performance liquid chromatography–UV to determine the ATV concentration. The high-performance liquid chromatography system consisted of a Waters LC Module 1 (Waters Corporation, Milford, MA) and a Nucleosil, 125/4, 100-5 C18 column (Macherey-Nagel, Oensingen, Switzerland). The mobile phase (acetonitrile/10 mM, pH 3, acetate buffer: 55/45) was delivered at a flow rate of 1 ml/min. The method has been fully validated, and a limit of quantification of 500 ng/ml and limit of detection of 50 ng/ml were obtained. A trueness of 98–102% was determined, and the intermediate precision was <2%; moreover, the three replicates injected at three different days demonstrate the repeatability of the method. The injection volume was 20 µl, and the drug was detected at 245 nm. A standard plot of ATV concentrations ranging from 6.25 to 50 μg/ml was prepared under identical conditions. Release profiles were compared using the similarity factor f2 and difference factor f1. (U.S. Food and Drug Administration, Guidance for industry CMC5-1995: immediate release solid oral dosage forms; U.S. Food and Drug Administration, 1995) (Shah et al., 1998). Equivalent profiles show f1 values close to 0 (generally less than 15) and f2 values close to 100 (generally greater than 50). Three independent experiments were run.
Samples of human saphenous veins were obtained from patients undergoing peripheral artery bypass surgery and prepared for explant culture, as previously described (Corpataux et al., 2005ab). Primary smooth muscle cells were cultured from human saphenous veins from 23 different patients, predominantly male (82%), with a mean age of 68.2 ± 12.1 years. The Ethics Committee of the University of Lausanne approved the experiments, which conform with the principles outlined in the Declaration of Helsinki for use of human tissue. In brief, vein segments discarded at surgery were placed in RPMI 1640 medium. Adhering fat and connective tissue were discarded. The adventitia was carefully removed, and the vessel was opened longitudinally. The inner surface was scraped to remove endothelial cells. Vein explants of 1–2 mm were plated, with luminal side down on the dry surface of a 12-well culture plate. Explants were gently covered with one drop of RPMI 1640 medium and placed overnight in a 37°C incubator gassed with air/5% CO2. The next day, culture medium was carefully added to the wells, taking care not to detach the explants. Medium was changed every 2–3 days for 1–2 weeks, until cells started to migrate out from the explants. Human smooth muscle cells (HSMCs) were identified by immunostaining using antibodies to α-smooth muscle actin (ab5694; Abcam, Cambridge, UK) and desmin (M 0760; DAKO Schweiz AG, Baar, Switzerland). Primary proliferating HSMCs, which featured a doubling time of 72–96 hours, were grown to confluence, passed once per week, and cultured until passage 3.
A total of 12,500 HSMCs were seeded per well in a 24-well plate and incubated in RPMI 1640 containing 0.4% FCS (growth arrest medium) for 24 hours. The next day, the growth arrest medium was removed and replaced with 600 μl of RPMI 1640 containing 10% FCS supplemented with ATV (5 or 10 μM). To evaluate the effect of ATV release by hydrogel, only 500 μl of RPMI 1640 medium was added, and 100 μl of ATV-loaded hydrogel (5 or 10 μM) or unloaded hydrogel was placed within a transwell insert of 8-μm pore polycarbonate membrane (Falcon; BD Biosciences Discovery Labware, Bedford, MA). Cell viability was assessed by the MTT test (Chen et al., 2011; Loo et al., 2011; Wang et al., 2011) prior to the addition of ATV or ATV-loaded gel (time 0) and after 24, 48, and 72 hours of culture. To this end, the cells were incubated with 10 μl of MTT labeling solution (5 mg/ml) in 200 µl of medium at 37°C for 4 hours and then solubilized in 200 μl of dimethylsulfoxide. Absorbance at 570 nm was measured with a microtiter plate reader with a reference wavelength of 630 nm, the reaction solvent being used as a blank.
The chemotactic-induced transmigration of HSMCs across a matrix barrier was investigated using a Boyden chamber (Back et al., 2005; Corpataux et al., 2005a,b; Erices et al., 2011) made of a polycarbonate membrane insert with 8-μm pores (Falcon; BD Biosciences Discovery Labware) placed in 24-well tissue culture plates. Confluent HSMCs were trypsinized and suspended in RPMI 1640 medium containing 10% serum supplemented with 0.25% bovine serum albumin and 50 ng/ml platelet-derived growth factor (migration medium). A total of 105 HSMCs in 500 μl of migration medium supplemented with various concentrations of ATV (5 or 10 μM) were loaded into the upper well of each chamber. In the experiments testing the delivering system, 100 μl of hydrogel (5 or 10 μM ATV) deposited on the bottom of wells and covered with a dialysate membrane (D9527, molecular mass 12 kDa; Sigma-Aldrich, St. Louis, MO) fixed by a rubber band were preincubated for 24 hours in the presence of 500 μl of RPMI 1640 medium, which was supplemented with 10% FCS, 0.25% bovine serum albumin, and 50 ng/ml platelet-derived growth factor prior to being used for the transmigration experiments. After a 24-hour culture at 37°C, the cell suspension was removed from the top of the inserts, which were washed with phosphate-buffered saline (PBS) and fixed in 100% ethanol at −20°C for 30 minutes. The upper side of the membranes was then rubbed with a moist cotton swab and a spatula to remove the cells, which did not transmigrate, and the membranes were stained for 10 minutes with hematoxylin, washed in PBS, and examined under a ×400 light microscope to score the nuclei of migrating HSMCs. Transmigration was expressed as the mean number of migrated cells per high-power field.
Cell migration was studied by a wound-healing assay (Chen et al., 2011; Erices et al., 2011), using silicon culture inserts (Ibidi, Martinsried, Germany) which define a cell-free gap of 500 µm. A total of 15,000 HSMCs in 70 μl of migration medium were seeded on both sides of the insert. After 24 hours, when cells reached a 90% confluence, the inserts were removed and the HSMCs were overlaid with 600 μl of culture medium supplemented with different concentrations of ATV. For the experiments involving the delivering system, 100 μl of hydrogel was overlaid in a transwell insert made of a polycarbonate membrane with an 8-μm pore size (Falcon; BD Biosciences Discovery Labware), which was placed in the wells of a 24-well plate, containing 500 μl of migration medium. Cultures were photographed at time 0, just after the silicone insert was removed, and thereafter every 12 hours for a period of 48 hours. Cells that migrated in the 500-µm area initially defined by the silicone insert were counted under a light microscope at a magnification of ×350.
The levels of human connexin43 (Cx43), heme oxygenase-1 (HO-1), tissue plasminogen activator (tPA), and plasminogen activator inhibitor-1(PAI-1) mRNA were determined by quantitative reverse transcription polymerase chain reaction, using the Fast SYBR Green Master Mix (Applied Biosystems, Foster City, CA) in a ViiA7 Instrument (Applied Biosystems). Briefly, RNA extracted from HSMCs using TriPure isolation reagent (Roche Applied Science, Indianapolis, IN) was treated for 30 minutes in the presence of DNase I (DNA-free kit; Ambion, Cambridge, UK). One microgram of total RNA was used for reverse transcription (Promega, Madison, WI). Equivalent amounts of cDNA from each reaction were processed for reverse-transcription polymerase chain reaction analysis. Negative controls included amplification of distilled water and RNA samples which had not been reverse transcribed. The primers used to amplify specific cDNAs are given in Table 1, and were designed using the free online software Primer3 (http://frodo.wi.mit.edu/primer3/) (Alonso et al., 2010a,b).
HSMCs were washed once with cold PBS and immediately collected and homogenized in lysis buffer containing SDS, as previously published (Alonso et al., 2010a,b). Protein content was measured using a detergent-compatible protein assay kit (Bio-Rad Laboratories, Reinach, Switzerland). Samples of total cell extracts (15 µg) were resolved by SDS-PAGE (10%) and transferred to a polyvinylidene fluoride membrane (Immobilon P; Millipore). Membranes were incubated for 1 hour in PBS containing 5% milk and 0.1% Tween 20 (blocking buffer). Saturated membranes were incubated overnight at 4°C in blocking buffer containing monoclonal anti–heme oxygenase-1 antibodies (ab13248 diluted 1/500; Abcam), rabbit polyclonal anti-Cx43 antibodies (AB1728 diluted 1/1000; Millipore), rabbit polyclonal anti–matrix metalloproteinase-9 (MMP9) antibodies (ab38898 diluted 1/500; Abcam), rabbit polyclonal anti–PAI-1 antibodies (NBP1-19773 diluted 1/1000; Novus Biologicals, Littleton, CO), or monoclonal antibodies anti–α-tubulin (T5168 diluted 1:3000; Sigma-Aldrich). After incubation at room temperature for 1 hour with a relevant secondary antibody conjugated to horseradish peroxidase (Fluka Chemie, Gmbh, Buchs, Switzerland; diluted 1:20,000), membranes were revealed by enhanced chemiluminescence according to the manufacturer’s instructions (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK). Densitometric analyses of immunolabeled proteins were performed using the ImageQuant Software (GE Healthcare).
HSMCs grown to confluence on glass slides were fixed for 10 minutes in acetone at −20°C, air dried, rinsed in PBS, and permeabilized for 1 hour in PBS supplemented with 1.5% bovine serum albumin and 0.1% Triton X-100. The cells were incubated overnight at 4°C in the presence of either a rabbit antibody against Cx43 (1/100; Cell Signaling, Danvers, MA), a monoclonal antibody against heme oxygenase-1 (ab13248, 1/50; Abcam), or a rabbit antibody against PAI-1 (NBP1-19773, 1/50; Novus Biologicals). Cells were washed and further exposed for 1 hour at room temperature to appropriate Alexa Fluor 488– or 594–conjugated antibodies (1/1000; Invitrogen). Cells were then washed, mounted in PBS containing 50% glycerol and 0.4 μg/ml 4,6-diamidino-2-phenylindole, and photographed under fluorescence microscopy (Leica Camera AG, Nidau, Switzerland).
Data are presented as mean ± S.E.M. One-way analysis of variance was performed to compare the mean values between groups, using the post-hoc Bonferroni test, as provided by the Statistical Package for the Social Sciences (SPSS 17.0; SPSS, Inc., Chicago, IL). A P value <0.05 was considered significant.
Atorvastatin Is Released from the Hydrogel in a Controlled Manner
The hydrogel released the entire amount of loaded ATV in the cell culture medium over 10 hours (Fig. 1A), whatever the initial concentration, as indicated by the similarity factor (f2 = 62) and difference factor (f1 = 5.9). Therefore, the shape of the release curve did not depend on ATV concentration within the concentration range (2.5–10 µM) studied. In contrast, the release of ATV per hour increased with the initial concentration of the drug (Fig. 1B), resulting in significantly different release profiles (f1 < 15 and f2 > 50 for all pair comparisons). The data show that the release of ATV from the hydrogel is sustained and controlled.
The Atorvastatin-Loaded Gel Decreases the Viability of HSMCs
The viability of HSMCs increased with time in medium devoid of ATV or containing a drug-free hydrogel, as revealed by the MTT test (Fig. 2). A similar pattern was seen when HSMCs were cultured in the presence of 2.5 µM ATV (not shown). In contrast, the addition of 5–10 µM ATV to the culture medium significantly decreased this viability (Fig. 2). A similar decrease was observed with HSMCs cultured in the presence of an ATV-loaded hydrogel (Fig. 2). This decrease reached significance after 24-hour and 72-hour culture in the presence of 10 and 5 µM ATV, respectively. From these observations, we chose to test these two concentrations in further experiments. The results indicate that ATV interferes with the viability of HSMCs, and that this effect is not altered when the drug is released by a hydrogel.
The Atorvastatin-Loaded Hydrogel Decreases the Transmigration of HSMCs
Control HSMCs rapidly transmigrate across an artificial membrane (Fig. 3). A significant decrease in this ability was observed after exposure to 5–10 µM ATV, whether the drug was directly added to medium or released by a hydrogel (Fig. 3). At the same concentrations, ATV also reduced the migration of HSMCs, as assessed in a wound-healing assay (Fig. 4). A significant, faster change was observed in the presence of 10 µM (24 hours) compared with 5 µM ATV (36 hours) (Fig. 4). The data document that the drug released by the hydrogel significantly affected the in vitro function of HSMCs.
The Atorvastatin-Loaded Hydrogel Selectively Modulates the Expression of Markers of HSMC Differentiation
ATV decreased HSMC expression of Cx43, MMP9, and PAI-1, three markers of IH (Deglise et al., 2005; Berard et al., 2013), at both the transcript (Fig. 5) and protein levels (Figs. 6 and 7). In contrast, expression of HO-1 (Lee et al., 2004) and tPA (Saucy et al., 2010; Berard et al., 2011) was upregulated by ATV (Figs. 5–7).These changes were observed whether ATV was directly added to the medium or was released by a hydrogel, but in the latter condition, became significant with a delay of about 24 hours compared with the former condition (Figs. 5–7). Immunofluorescence showed that, despite these quantitative changes, ATV did not alter the intracellular distribution of the Cx43, HO-1, and PAI-1 proteins (Fig. 7).
The implantation of a vein graft into arterial circulation often results in the development of IH, leading to vessel stenosis (Owens et al., 2009; Owens, 2010). This process is associated with the dedifferentiation of HSMCs, which turn from a contractile to a secretory phenotype, characterized by increased proliferation and migration (Owens et al., 2004; Mitra et al., 2006; Alexander and Owens, 2012; Nguyen et al., 2013). Given that IH involves various biological mechanisms, a therapy combining different drugs may help prevent interference with the vessel stenosis. Candidates, such as statins, have limited effects after systemic administration due to liver catabolism, digestive clearance, and frequent side effects of the high dosage required for systemic efficiency. It would be beneficial to selectively provide the drugs at the site of the stenosis (Wiedemann et al., 2012). As yet, however, no method can achieve such a locally targeted therapy.
As a first approach toward such development, we investigated a hydrogel platform that could ensure a local and controlled delivery of statins in a surgical field. Various thermosensitive gels have been tested for drug administration that, however, usually have a rather short residence time in vivo (Le Renard et al., 2010). We have selected a hydrogel made of hyaluronic acid, a key component of extracellular matrix in most native tissues, for two reasons. First, the cross-linking of hyaluronic acid increases the gel viscosity, extending its in vivo persistence (Elder et al., 2011). Second, hyaluronic acid could help reduce IH formation (Ferns et al., 1995; Chajara et al., 2003). Here, we have tested such a gel for the delivery of ATV, a statin which inhibits HSMC proliferation (Corpataux et al., 2005a,b) and IH (Qiang et al., 2012).
By use of primary HSMCs from human saphenous veins, we demonstrated that ATV decreases the viability, proliferation, and transmigration of HSMCs and that these effects were similarly observed whether the drug was directly added to the culture medium or was loaded on the hyaluronic acid hydrogel. The difference between the two conditions only related to the time course. Thus, the effects of the ATV hydrogel were somewhat delayed compared with those of the free ATV, and were sustained for the 48 hours, which were investigated here. This time frame should provide a sufficient therapeutic window to interfere with the early stages of IH development, which is launched by the endothelium disruption. The administration of 80 mg of ATV results in a maximum plasma concentration of the active molecule of ~ 50–200 ng/ml after 1–2hours (Bahrami et al., 2005), i.e., concentrations that are significantly lower than those we tested (2.5–5 µM), and could partially explain the limited effect of atorvastatin taken orally. The latter concentrations, which were intended to reach a range of those that are usually tested in vitro (Saijonmaa et al., 2004; Suski et al., 2013), are essential to reach a high local concentration of the drug. The need for elevated concentrations to detect functional effects of ATV in vitro is also likely due to the higher proliferation of SMCs in culture, which contrasts with their usually quiescent state in vivo. Moreover, the continuous stimulation of SMC with the serum, growth factors, and nutrients of culture media could contribute to the desensitization of cells to lower concentrations of ATV. At any rate, these concentration differences do not undervalue the interest of our experimental observations, which were all made under rigorously similar conditions for control SMC and cells exposed to ATV.
We also document that ATV differentially regulates the expression of specific genes involved in the development of IH, decreasing Cx43, MMP9, and PAI-1, and increasing tPA and HO-1. By immunofluorescence studies, we further demonstrated that the levels of Cx43 were time-dependently decreased in the presence of both free ATV and ATV released from hydrogel, whereas those of HO-1 and tPA were increased under the very same conditions. Vascular cells express four connexins (Cx37, Cx40, Cx43, and Cx45) in various amounts depending on species and vascular beds (Alonso et al., 2010b). These proteins appear to be involved in different aspects of IH. Thus, we previously demonstrated that Cx43 is expressed in HSMCs of human veins and upregulated in the media layer with the development of IH (Deglise et al., 2005). Cx43 participates in controlling the migration and proliferation of HSMCs (Song et al., 2009) and is increased with the synthetic state of these cells, which develops in early atherosclerotic lesions (Haefliger et al., 2004). Moreover, statins reduced Cx43 in the aortas of atherosclerotic rabbits (Wang et al., 2005). Heme oxygenases degrade heme to biliverdin, carbon monoxide, and free iron (Otterbein et al., 2000, 2003). The exogenous administration of HO-1 reduced the development of atherosclerosis and restenosis in balloon-injured rat models (Juan et al., 2001; Tulis et al., 2001). Simvastatin increased cytoprotective HO-1 in HSMCs (Lee et al., 2004), partially accounting for its anti-inflammatory effects (Lee et al., 2004). IH remodeling requires the integrated effects of the fibrinolytic system, the MMPs, and their inhibitors. We recently showed that PAI-1 (Ha et al., 2009) is induced by arterial shear stress and promotes IH (Berard et al., 2013), likely by enhancing the degradation of the extracellular matrix, which facilitates the migration of HSMCs from the media to the intima layer (Muto et al., 2012). It is remarkable that several of the factors which contribute to IH are simultaneously but differentially regulated by ATV, in a way that interferes with the pathological remodeling of the vascular wall.
If the drug and other candidate statins were made selectively available at the site of venous stenosis, it may be feasible to prevent or hinder IH if the effective drug levels can be maintained for the few early days, during which the phenotypic change in HSMCs is launched. Our experiments indicate that this can be achieved, at least in vitro, using a hydrogel of native hyaluronic acid, which releases active ATV in a sustained manner. The results pave the way toward the generation of further platforms that could release a combination of antistenosis drugs, in vivo and over extended periods of time.
Participated in research design: Dubuis, May, Alonso, Luca, Mylonaki, Delie, Jordan, Saucy, Haefliger.
Conducted Experiments: Delie, Jordan, Déglise, Corpataux, Saucy, Haefliger.
Performed data analysis: Dubuis, May, Alonso, Delie, Jordan, Saucy, Haefliger.
Wrote or contributed to the writing of the manuscript: Meda, Delie, Jordan, Saucy, Haefliger.
- Received August 13, 2013.
- Accepted September 25, 2013.
↵1 Current affiliation: Dr. E. Gaeub AG, Bern, Switzerland.
C.D., L.M., F.S., and J.-A.H. contributed equally to this work.
This work was supported by the Swiss National Science Foundation [Grants 31003A-138528 and 310030-141162]; the Octav and the Marcella Botnar Foundation; the Novartis Foundation; and the Emma Muschamp Foundation.
- similarity factor
- difference factor
- fetal calf serum
- heme oxygenase-1
- human smooth muscle cell
- intimal hyperplasia
- matrix metalloproteinase
- plasminogen activator inhibitor-1
- phosphate-buffered saline
- tissue plasminogen activator
- Copyright © 2013 by The American Society for Pharmacology and Experimental Therapeutics