Fas Ligand and Nitric Oxide Combination to Control Smooth Muscle Growth while Sparing Endothelium

Abstract

Metallic stents cause vascular wall damage with subsequent smooth muscle cell (SMC) proliferation, neointimal hyperplasia, and treatment failure. To combat in-stent restenosis, drug-eluting stents (DES) delivering mTOR inhibitors such as sirolimus or everolimus have become standard for coronary stenting. However, the relatively non-specific action of mTOR inhibitors prevents efficient endothelium recovery and mandates dual antiplatelet therapy to prevent thrombosis. Unfortunately, long-term dual antiplatelet therapy leads to increased risk of bleeding/stroke and, paradoxically, myocardial infarction. Here, we took advantage of the fact that nitric oxide (NO) increases Fas receptors on the SMC surface. Fas forms a death-inducing complex upon binding to Fas ligand (FasL), while endothelial cells (ECs) are relatively resistant to this pathway. Selected doses of FasL and NO donor synergistically increased SMC apoptosis and inhibited SMC growth more potently than did everolimus or sirolimus, while having no significant effect on EC viability and proliferation. This differential effect was corroborated in ex vivo pig coronaries, where the neointimal formation was inhibited by the drug combination, but endothelial viability was retained. We also deployed FasL-NO donor-releasing ethylene-vinyl acetate copolymer (EVAc)-coated stents into pig coronary arteries, and cultured them in perfusion bioreactors for one week. FasL and NO donor, released from the stent coating, killed SMCs close to the stent struts, even in the presence of flow rates mimicking those of native arteries. Thus, the FasL-NO donor-combination has a potential to prevent intimal hyperplasia and in-stent restenosis, without harming endothelial restoration, and hence may be a superior drug delivery strategy for DES.

1. INTRODUCTION

Coronary artery disease (CAD) is the major cause of morbidity and mortality in western countries [1]. CAD is mainly caused by atherosclerosis, which is narrowing and hardening of arteries due to excessive buildup of plaque on the vessel wall. Invasive percutaneous coronary intervention (PCI) procedures such as atherectomy, balloon angioplasty, and stent deployment restore the blood flow in diseased coronary arteries. However, in-stent restenosis, which is the re-narrowing of the vessel as a response to wall injury and endothelial denudation, is one of the major drawbacks of this procedure [2–4]. Following stent deployment, endothelial denudation and exposure of the vessel wall to blood flow results in immediate platelet and fibrinogen adherence to the vessel wall, and the adhesion of leukocytes. In addition to circulating cells adhering to the lumen, vessel wall injury due to the high-pressure distension causes medial smooth muscle cells (SMC) and adventitial cell damage, followed by their proliferation and migration to the lumen surface [5–7].

In an effort to prevent the resulting intimal hyperplasia, drug-eluting stents (DES) were developed roughly 15 years ago. Currently marketed DES deliver antiproliferative drugs, mammalian target for rapamycin (mTOR) inhibitors (sirolimus, everolimus, biolimus A9, or zotarolimus), microtubule inhibitors (paclitaxel), or calcineurin blockers (tacrolimus or pimecrolimus) to the vessel injury site, and reduce in-stent restenosis while avoiding systemic toxicity [8, 9]. However, the non-specific antiproliferative effect of eluted drugs affects not only SMCs but also endothelial cells (EC), which results in the need for prolonged antiplatelet therapy following stent deployment [10, 11]. Although re-endothelialization between the stent struts of some of the newer DESs, such as everolimus-eluting stents, is comparable with that of bare metal stents, eNOS expression in the repopulated ECs of vessels with DES are significantly lower than those with bare metal stents (BMS) [12, 13]. Moreover, treatment failures with first- and second-generation DES is still too common: the COMPARE trial reported that 11.4% of patients required target-vessel revascularization at 5 years for paclitaxel-eluting stents (1st generation DES), and 7.4% revascularization for everolimus-eluting stents (2nd generation DES) [14]. Similarly, the RESOLUTE trial showed total cardiac event accrual of 15% at 5 years, including a 10% target vessel revascularization rate, due mainly to in-stent restenosis but also due to acute thrombosis. Patients with early in-stent restenosis - occurring before one year - had higher incidences of myocardial infarction (MI) and death as compared to other patients [15]. Thus, while DES certainly function better than bare-metal stents – which have a 30% restenosis rate – there are still major limitations to currently available DES for cardiac patients [16, 17]. Given the substantial use of these devices in countries like the US, there is an urgent need for a novel stent technology which prevents SMC proliferation more potently than currently available DES, and which differentiates its inhibitory effect between SMCs and ECs, thereby sparing endothelial function.

The feasibility of using the Fas/Fas ligand (FasL) system to target vascular progenitor cells contributing to intimal hyperplasia has been shown in many studies [18–24]. Fas receptor, a member of tissue necrosis factor family, is a death receptor that initiates apoptosis upon activation by its ligand, FasL [18, 19]. Ectopic and increased expression of FasL may be effective in inducing apoptosis in Fas-bearing vascular SMCs and macrophages [20, 23, 24]. ECs, however, are comparably resistant to Fas-mediated apoptosis, since they endogenously express surface FasL. One possible mechanism that protects ECs is that the Fas-mediated death signal is blocked by FLICE (Fas-associated death domain–like interleukin-1β–converting enzyme)-inhibitory proteins (FLIPs) [25–27]. Hence, the local delivery of FasL to sites of the injured vessel wall has the potential to decrease the number of SMCs without affecting re-endothelialization [20, 25]. Interestingly, over-expression of FasL in ECs via adenovirus-mediated gene transfer decreases balloon injury-related intimal hyperplasia, by inducing apoptosis in SMCs or infiltrating host inflammatory cells, without self-destruction of the ECs expressing the FasL [21–23, 28].

Intriguingly, nitric oxide (NO) is known to increase surface Fas receptors on vascular SMCs [29, 30]. NO from overexpression of inducible nitric oxide synthase (iNOS) results in pro-apoptotic p53 protein accumulation in human fibroblasts [31]. Furthermore, p53 induces surface Fas expression in VSMCs by translocation of a preformed pool from the Golgi apparatus [30, 32]. Therefore, we hypothesize that local delivery of NO can be used to increase Fas receptor expression on the cell surface and enhance the potential anti-SMC therapeutic effect of FasL delivered to the same region (Fig. 1).

Figure 1.

A-B: NO released from the stent moves Fas receptors to the cell surface. C: FasL oligomerizes the receptors and initiates apoptosis via caspase-dependent mechanisms.

Here, for the first time, we studied the effect of FasL and NO combination on SMCs and ECs. In addition, to demonstrate proof of concept, we showed the feasibility of releasing the FasL and NO combinations from a stent surface to control SMC proliferation/apoptosis and intimal thickening in an ex vivo pig coronary artery model.

2. MATERIALS AND METHODS

2.1. Cells and Materials

Human and porcine aortic SMCs (HAoSMCs and PAoSMCs) were isolated from aortas as previously described [33], and were cultured in DMEM with 10% FBS (Hyclone, Marlborough, MA, USA) at 37° C and 5% CO₂ until sub-confluence. Human aortic ECs (HAoECs) were purchased from Promocell (Heidelberg, Germany). All EC types were cultured in EGM-2 (Lonza, Basel, Switzerland). Cells at passage 3–6 were used in all experiments in this study. Recombinant human Fas ligand was purchased from BioLegend (San Diego, CA, USA). The NO donor DetaNONOate ((Z)-1-[N-(2-aminoethyl)-N-(2-ammonioethyl)amino]diazen-1-ium-1,2-diolate) was purchased from Cayman Chemical (Ann Arbor, MI, USA). Sirolimus and everolimus was purchased from LC Laboratories (Boston, MA, USA). Ethylene-vinyl acetate copolymer (EVAc), dichloromethane (DCM) and Ficoll 400 were purchased from Sigma (St. Louis, MO, USA).

2.2. Cell culture with FasL- and NO donor-combination, sirolimus and everolimus

Varying combinations of recombinant human FasL at 0 and 400 ng/mL and NO donor, DetaNONOate, at 0, 0.1, and 0.2 mM concentrations were added to culture medium of HAoSMCs, PAoSMCs and HAoECs in 96 well-plates for 48 hours to determine the effect of this combination on different cell types. Alternatively, HAoECs and HAoSMCs were cultured in 96-well plates with either 10 nM sirolimus, 10 nM everolimus or the combination of 400 ng/mL FasL and 0.1 mM DetaNONOate. The drugs were added into the culture medium when cells were 80% confluent. Cells were fixed with 4% PFA at 48 hours after adding the drugs, and ECs and SMCs viability and proliferation were assessed by immunofluorescent staining and live/dead staining.

Subsequently, HAoECs and HAoSMCs were cultured in 6-well plates with varying combinations of FasL and DetaNONOate to evaluate Caspase-8 activation, which is a specific marker of Fas-mediated or TNFα-mediated apoptosis. Cell lysates were collected at the end of 24-hours of drug treatment for Western Blot analysis.

Additionally, gene expression levels for the following inflammatory markers were analyzed: intracellular cell adhesion molecule (ICAM-1); vascular cell adhesion molecule-1 (VCAM-1); interleukin-6 (IL-6) and interleukin-8 (IL-8). These were quantified in HAoSMCs and HAoECs using qRT-PCR. In addition, the effect of the drugs on EC phenotype was assessed by quantifying the eNOS, VE-Cadherin, ZO-1 (Zonula occludens-1), and thrombomodulin expressions in HAoECs.

2.3. FasL- and NO donor-releasing polymer matrices and stent coatings

EVAc, a material utilized in several FDA-approved devices, is a non-absorbable polymer used in variety of medical implants [29]. EVAc can be doped with an inert powder composed of an excipient like and the desired agent, which results in a porous matrix that facilitates the release of the doped agent. In aqueous solutions, the ficoll and the drug agent are then released in a sustained way. FasL- and NO donor-releasing materials were produced in a manner similar to a previously described method [34]. Recombinant human FasL, NO donor DetaNONOate and sufficient amount of ficoll were mixed in PBS with 0.01 M NaOH to obtain a mass ratio of 1:50:1700 FasL:DetaNONOate:Ficoll at pH≈12. The mixture was vortexed, quickly frozen and lyophilized for 48 hours to obtain a homogeneous powder. The resulting powder was crushed using a mortar and pestle to form small particles (<300 μm in diameter). EVAc was dissolved in DCM with a mass/volume ratio of 10% for matrices of drug-eluting polymer, and 1% for stent coating. Sufficient FasL/DetaNONOate/Ficoll powder was added to the polymer solution to obtain 40% mass loading (mass percent solids in solids plus polymer). The solution was vortexed to produce a homogeneous suspension, and the suspension was poured into a level glass mold that was prechilled on a block of dry ice. The matrix was removed from the glass mold, and DCM was evaporated at −20° C for 48 hours and then by lyophilizing for 24 hours. Control polymers were made by the same technique but with ficoll only. The resulting dry matrices, approximately 1.5 mm thick, were cut into small rectangular pieces weighing approximately 60 mg. These 60-mg EVAc matrices were incubated in 1 mL of PBS at 37 °C, and the PBS was completely replaced at periodic intervals. The DetaNONOate and FasL amount released into the PBS was detected by measuring peak UV absorption at 252 nm with NanoDrop 2000 spectrophotometer (Thermo Scientific, Waltham, MA, USA), and NanoOrange protein quantitation kit (Thermo Scientific, Waltham, MA, USA), respectively.

Balloon-expandable 2.75 mm-diameter CoCr alloy stents (Kossel Medtech, Suzhou, China) were pre-chilled on dry ice and dipped into an EVAc-drug suspension for 5 seconds. The stents were removed from the suspension, and DMC was evaporated at −20° C for 48 hours and then lyophilized for 24 hours. Control stents were coated with ficoll-only releasing EVAc. The mass ratios of FasL:DETANO:Ficoll:EVAc were 1:50:1700:2625 in all polymer matrices and stent coatings.

2.4. Ex-vivo pig coronary artery culture with FasL- and NO donor-releasing polymer matrices

Porcine hearts were procured from 20–50 kg outbred Yorkshire pigs at a local abattoir (J Latella and Sons Piggery, West Haven, CT) and pericardial coronary arteries were isolated. Artery segments 5–6 mm in length were cultured in 12-well tissue culture plates with 4 mL of DMEM:Vasculife (1:1) medium containing 20% FBS for up to 7 days. EVAc matrices with a mass of 60 mg and releasing FasL and DetaNONOate were suspended in the culture medium. Ficoll-only releasing matrices were used as control. The medium was changed every two days. Artery samples were fixed with 10% neutral buffered formalin at indicated times and processed for immunofluorescence staining analysis.

2.5. Ex vivo pig coronary artery culture with FasL- and NO donor-releasing stents in perfusion bioreactors

CoCr alloy stents were coated either with Ficoll-only-releasing or FasL- and DetaNONOate-releasing EVAc. The coated stents were sterilized with 13 kGy of gamma irradiation with a Mark I-68A 137Cs irradiator (JL Shepherd and Associates, San Fernando, CA) and deployed into freshly isolated pig coronary arteries, ex vivo. The stented arteries were mounted into perfusion bioreactors, as previously described [35, 36]. PharMed silicone tubing was attached to glass pipettes onto which the stented artery was tied, and a reservoir inlet and outlet completed the perfusion loop. Culture medium was pumped through the stented artery and then drained back to the reservoir via a Masterflex L/S roller pump. The vessels were cultured in DMEM:Vasculife (1:1) with 20% FBS for 7 days, and the culture medium was refreshed 3 times a week. The ratio of lumen area to vessel area (the area enclosed by tracing the media-adventitia interface), SMC apoptosis and proliferation was evaluated histologically, after one week of culture. Subsequently, pig coronary arteries with FasL- and DetaNONOate-releasing EVAc coated stents were cultured in the perfusion bioreactors for 14 days to evaluate the reestablishment of the EC coverage on the stent struts. Stented lumens were imaged by imaged by FE-SEM.

2.6. Histology and immunostaining

Cultured cells were fixed with 4% paraformaldehyde for ten minutes. Artery segments were fixed with 10% neutral buffered formalin, embedded in paraffin, and sectioned at 5 μm thickness. Tissue slides were stained for hematoxylin and eosin (H&E), as described previously [37]. Cell death was assessed by labeling DNA strand breaks (TUNEL stain, Sigma, St. Louis, MO, USA) according to manufacturer’s instructions. Additionally, fluorescein diacetate and propidium iodide (PI) (Sigma, St. Louis, MO, USA) were used to stain live and dead cells, respectively. For immunofluorescence, cultured cells were permeabilized and blocked with PBS containing 5% BSA and 0.2% Triton X-100 for 1 h, and subsequently incubated in primary antibodies against ki67 (Abcam, ab16667, rabbit monoclonal, 0.5 μg/mL, Cambridge, MA, USA). After washing cells with 0.2% Triton x-100 in PBS, secondary antibody (Alexafluor 555) was applied at 1:500 dilution for 1 h.

Tissue sections were incubated in sodium citrate buffer (pH 6) at 75° C for 20 minutes for antigen retrieval, permeabilized and blocked with PBS containing 5% BSA and 0.2% Triton X-100 for 1 hour, and subsequently incubated in primary antibodies against von Willebrand factor (vWF) (Dako, M081, rabbit monoclonal, 70 μg/mL, Agilent Technologies, Santa Clara, CA, USA) at 4° C. For a negative control, secondary antibodies were applied without incubating samples with primary antibodies. The cells and tissue slides were visualized using a Zeiss Axiovert 200m fluorescence microscope. Cell number was quantified by counting DAPI stained nuclei. Proliferating and apoptotic cells were quantified by counting nuclei that co-localized with ki67 or TUNEL, respectively. Cells were counted in 10 randomly taken 20X magnification images of each well or each vessel section. The number of viable cells were determined by subtracting the number of TUNEL⁺ or PI⁺ nuclei from the total number of nuclei (stained by DAPI or Hoechst). To quantify the vessel narrowing in H&E stained tissue sections, the intimal thickness, ratio of the intima to media, and intimal area was calculated for each vessel section, by ImageJ (National Institutes of Health, Bethesda, MD, USA).

To evaluate the EC recovery in the stented lumens, tissue segments were rinsed gently with (PBS) three times, and fixed with 2.5% glutaraldehyde solution for 30 min. Next, the samples were dehydrated sequentially in 50%, 75%, 90% and 100% ethanol solution. Finally, the luminal surfaces of the stented pig arteries were sputter-coated with gold, and imaged by FE-SEM (Model S-4700, Hitachi, Canada).

2.7. Western Blot

HAoECs and HAoSMCs, cultured on tissue culture plates in standard culture medium with varying combinations of DetaNONOate and FasL, were lysed in cold RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% [v/v] Triton X-100, 0.5% [w/v] SDC, and 0.1% [w/v] sodium dodecyl sulphate (SDS)) containing 1% protease inhibitor (Sigma Aldrich, St.Louis, MO). Protein lysate was mixed with laemmli sample buffer, reduced (DTT at 95 °C 5 min), and loaded on 4–20% gradient polyacrylamide gels (Bio-Rad, Hercules, CA). Blots were run for 60 min at 120 V and the protein was transferred to PVDF membrane and blocked for 1 h at room temperature in 5% nonfat dry milk. Anti-Caspase 8 primary antibody (Biolegend, San Diego, CA, USA) 1.6 μg/mL was applied overnight at 4 °C in 1% nonfat dry milk in TBST buffer (0.1% Tween-20 in Tris-buffered saline). After being rinsed with TBST, goat anti rat secondary antibody (Novus Biologicals, Centennial, CO, USA) were applied at a dilution of 1:1000 for 1 h at RT. Protein was detected using enhanced chemiluminescence (Thermo Scientific, Waltham, MA, USA).

2.8. Gene quantification with real-time qRT-PCR

The change in eNOS, VE-Cadherin, ZO-1, thrombomodulin, ICAM-1, VCAM-1, IL-6, and IL-8 gene expressions were evaluated with qRT-PCR. Total cellular RNA was isolated using the RNeasy Mini Kit (QIAGEN) following the manufacturer’s instructions. 500 ng of RNA from each sample was used to synthesize single-stranded cDNA using the iScript cDNA Synthesis Kit (Bio-Rad), according to the manufacturer’s protocol. All PCR reactions were run in triplicate using 1 μL of cDNA in a 25 μL final volume with iQ SYBR Green Supermix (Bio-Rad) and 200 nM each of forward and reverse eNOS, VE-Cadherin, ZO-1, thrombomodulin, ICAM-1, VCAM-1, IL-6, and IL-8 primers (Table S1). RT-qPCR was performed on the CFX96 Real-Time System (Bio-Rad) for 40 cycles. Average threshold cycle values (Ct) from the triplicate PCR reactions were normalized to β-actin (ΔCt) and reported as fold change using the 2-ΔΔCt method, with control cells assigned a fold change of 1. The experiment was performed in triplicate.

2.9. Statistical analysis

Data were expressed as mean +/− SD and analyzed using one-way ANOVA test with Tukey’s multiple comparison using GraphPad Prism-7.01 (GraphPad Software, Inc., La Jolla, CA, USA). A p value of <0.05 was considered statistically significant.

3. RESULTS

3.1. Effect of FasL-NO donor combination, Sirolimus, and Everolimus on SMC and EC viability and proliferation.

To evaluate the extent to which the FasL-NO donor combination had a differential effect on SMCs and ECs, we added varying dosage combinations of an NO donor, DetaNONOate (0, 0.1, and 0.2 mM), and FasL (0 and 400 ng/mL) into SMC and EC culture medium in 96-well plates. We found that 0.1 mM DetaNONOate with 400 ng/mL FasL resulted in potent SMC inhibition (Fig. 2 and 3) (>40% inhibition for both pig and human SMCs, n=6) while sparing arterial endothelium (Fig. 2 and 4, Fig. S2 and S3). DetaNONOate alone at 0.1 mM, with or without FasL, did not affect cultured EC cell number or rates of apoptosis (Fig. 2), but concentrations of DetaNONOate above 0.2 mM negatively affected EC proliferation (Fig. S4). These results demonstrate that tuning of the concentrations in the FasL and NO donor allows the identification of doses having no significant effect on cultured EC viability and proliferation, while increasing cultured SMC apoptosis in a synergistic way (Fig. S1).

Figure 2.

The number of live SMCs, normalized to untreated group, in human (A) and porcine (B) aortic SMCs cultured with FasL and NO donor. 48 hours of treatment with varying concentrations of human FasL and DetaNONOate (NO donor) decreased the number of viable cells in human and pig aortic SMCs synergistically. Human aortic ECs resist treatment with DetaNONOate and FasL. 48 hours of treatment with certain doses of human FasL and DetaNONOate did not cause apoptosis (C) or affect proliferation (D) in ECs. NS: Not significant, *: p<0.05, n = 8. Protein expression of full length Caspase-8, cleaved Caspase-8 and β-actin (E). FasL did not cause cleavage of Caspase-8 even with 0.2 mM DetaNONOate. Cleaved Caspase-8 expression relative to full length Caspase-8 for each sample (F) (Experiments repeated in triplicates, *: p<0.05).

Figure 3.

Human aortic SMCs cultured for 48 hours in standard culture medium either with 10 nM sirolimus (B), 10 nM everolimus (C) or 0.1 mM DetaNONOate+400 ng/mL FasL (D). Blue: DAPI, Red: ki67. FasL with NO donor decreases the total number of cells and number of proliferating cell significantly more than both sirolimus and everolimus at end of 48 hours of culture. Total cell numbers were normalized to that of in control group (Scale bar: 100 μm, n=8–17, experiments were repeated four times. One-way ANOVA, *: p<0.0001 and †: p<0.05). qRT-PCR analysis of IL-6 and IL-8 gene change in human aortic SMCs. Changes in gene expression levels, relative to the housekeeping gene β-actin, were normalized to the untreated group (experiments were repeated in triplicate).

Figure 4.

Human aortic ECs cultured for 48 hours in standard culture medium either with 10 nM sirolimus, 10 nM everolimus or 0.1 mM DetaNONOate + 400 ng/mL FasL. Total cell number was decreased significantly in Sirolimus and Everolimus groups (n=8, p<0.01) but not in FasL+NO donor group. In FasL+NO donor group ki67+ cell number was significantly higher than in sirolimus and everolimus groups. Total cell numbers were normalized to that of in control group (n=8, * indicates p < 0.001).

Immunoblotting analysis showed that Caspase-8 cleavage is observed only in SMCs treated with combination of FasL and DetaNONOate (both 0.05 or 0.1 mM). In ECs, however, FasL did not cause cleavage of Caspase-8 ,even with 0.2 mM concentration of DetaNONOate (Fig. 2 E and F).

Next, we cultured human aortic SMCs and ECs with 10 nM sirolimus, 10nM everolimus and the FasL-DetaNONOate combination (400 ng/mL and 0.1 mM, respectively) to compare the effects of these drugs on cultured cell viability and proliferation. Figures 3 shows representative panels of SMC nuclei and ki67⁺ nuclei images of each group. In the cell culture wells treated with FasL and NO donor, there are only a few living SMCs in any of the 10 X-magnified images, and in most of the images there are no ki67⁺ nuclei (proliferating cells) (Fig. 3 D). Both total SMC numbers (n=8–17, One-way ANOVA, p<0.0001, between FasL-detaNONOate and any other group) and the percentage of the proliferating SMCs (p<0.05) were significantly less in the FasL-DetaNONOate treated group than in any other group (Fig. 3). Compared to the control group with no drug treatment, sirolimus, everolimus and the FasL-DetaNONOate combination significantly decreased total SMC number by 36±9.8%, 53±13.8%, and 93±2.1%, respectively (p<0.0001). The percentage of proliferating (ki67⁺) cells were 5.5±3.7%, 10.6±4.5% and 1.4±4.8% in sirolimus, everolimus and FasL-DetaNONOate treated SMCs, while 10±2.5% of the SMCs were ki67⁺ in the control group (Fig. 3E).

qRT-PCR analysis showed that inflammatory markers IL-6 and IL-8 were dramatically upregulated with the FasL and DetaNONOate combination (Fig. 3 F and G). However, the relative change in VCAM-1 and ICAM-1 genes were undetectable in all of the SMC groups.

For human aortic ECs, sirolimus and everolimus decreased total EC number more than 19% and 15%, respectively, compared to the control group (n=7–9, one-way ANOVA, p<0.01), while the total number of FasL-DetaNONOate treated ECs was not significantly different from the control (Fig. 4). Additionally, the number of ki67⁺ ECs in the FasL-DetaNONOate treated group was significantly higher than in the sirolimus- and everolimus-treated groups. These results show that the FasL-NO donor combination has a potent ability to prevent SMC proliferation while having little effect on EC proliferation. This property differentiates the effects of FasL-NO donor combination from the effects of sirolimus and everolimus.

3.2. Effect of FasL-NO donor combination, Sirolimus, and Everolimus on EC quality

We also evaluated the effect of sirolimus, everolimus or FasL-DetaNONOate treatment on eNOS, VE-Cadherin, ZO-1 and thrombomodulin expressions in cultured human aortic ECs. VE-Cadherin, ZO-1 and thrombomodulin expressions did not change significantly after 48 hours of treatment with any of the drugs. However, both 10 nM sirolimus and 10 nM everolimus decreased eNOS expression significantly (more than 67% and 72%, respectively) (Fig. 5) while the combination of FasL (400 ng/mL) and DetaNONOate (0.1 mM) decreased eNOS expression by 27% (Fig. 5). Thus, all three drug treatments decreased eNOS expression, but the effect of FasL-DetaNONOate combination on eNOs expression was significantly less than the others. All three drug treatments decreased the relative expression of VCAM-1, ICAM-1, and IL-6, while IL-8 expression increased with all of the drugs (Fig. S5).

Figure 5.

qRT-PCR analysis of eNOS (A), VE-Cadherine (B), ZO1 (C) and Thrombomodulin (D) gene change in Human aortic ECs cultured for 48 hours in standard culture medium either with 10 nM sirolimus, 10 nM everolimus or 400 ng/mL FasL + 0.1 mM DetaNONOate. Changes in gene expression levels, relative to the house-keeping gene β-actin, were normalized to the untreated group. The reduction in eNOS gene with FasL and NO donor combination is significantly less than that of with sirolimus or everolimus. There was no significant change in VE-Cadherine, ZO1 and Thrombomodulin gene expression (n = 3; **: p < 0.0001, †: p<0.05).

3.3. FasL and NO donor release kinetics of polymer matrices

To create a stent coating capable of simultaneous release of FasL and DetaNONOate in a sustained way, we doped EVAc polymer with an inert powder of recombinant FasL and DetaNONOate, using ficoll as an excipient [34]. DetaNONOate has a 20-hour half-life in aqueous solutions at 37 °C at pH≈7.4. When suspended in PBS at 37° C, EVAc matrices released both FasL and DetaNONOate in a sustained manner for two weeks (Fig. 6).

Figure 6.

EVAc matrices suspended in PBS at 37° C released DetaNONOate and FasL for at least two weeks. Average cumulative release is given in terms of the percentage of the total loading in the matrices. n=3.

3.4. Effect of sustained release of FasL- and NO donor- combination on pig SMCs and coronary arteries

To delineate the effects of the co-release of NO and FasL in native arteries, freshly harvested porcine coronary arteries were cultured in dishes under standard culture conditions in DMEM:Vasculife (50:50) with 20% FBS in the presence of FasL-DetaNONOate-loaded EVAc matrices. Histological outcomes were assessed after 7 days. Suspension of FasL- and DetaNONOate-releasing EVAc matrices in culture medium was compared to control culture conditions, containing EVAc matrices that were not pre-loaded with the drug. In coronary arteries that were cultured with excipient-only control matrices for 7 days, the intima region became noticeably thicker as compared to freshly excised coronary arteries (Fig. 7 A, E). However, the FasL- and DetaNONOate-releasing EVAc matrices largely prevented intimal thickening (Fig. 7). The intimal thickness, area, and intima/media ratio was significantly less in the presence of FasL- and DeatNONOate-releasing EVAc matrices as compared to control EVAc matrices (Fig. 7 M–O). TUNEL staining showed that drug-releasing EVAc matrices were associated with increased SMC apoptosis in the intimal region (green, Fig. 7 B, F, J). However, vWF staining and absence of the TUNEL staining on the luminal cell layer showed that EC remained viable on the lumen, implying a lack of toxicity to this cell type (yellow stain, Fig 7 C, G, K). These results show that effective doses of FasL-NO donor combination can be released from EVAc polymer matrices for at least one week, and that the drug combination has a differential effect on SMC and EC viability.

Figure 7.

FasL- and DetaNONOate- releasing EVAc matrices were suspended in culture medium of porcine coronary arteries. A-D: Fresh native; E-H: EVAc control; I-L: FasL- and DetaNONOate- releasing EVAc matrices, all 7 days. Extensive TUNEL staining in SMCs (J), spares the vWF-expressing endothelium (L) (Scale bar: 100 μm). M-O: Intimal thickness, intima/media ratio, and intimal area (Experiments were repeated four times. *: p < 0.05. NS: Not significant).

To illustrate the application of this drug release technology in the context of vascular stents, balloon-expandable CoCr alloy stents were coated with EVAc films containing either FasL, DetaNONoate, and excipient or excipient only. These stents were deployed into freshly isolated pig coronary arteries. The stented arteries were cultured in perfusion bioreactors with 25 mL/min flow for 7, to partially mimic flow in the porcine coronary artery as described before [38]. At the end of 7 days, there were acellular regions in the vascular media region near the FasL- and DetaNONOate-releasing stent struts (Fig. 8 A). TUNEL stain (green) in Fig. 8 E and F shows that the stents that co-released FasL and NO donor resulted in significantly higher SMC apoptosis in the stent proximity than did the excipient only devices, indicating the local effect of the released drugs (Fig. 8G). These results show that stents can locally deliver functional amounts of FasL and NO donor to the vessel wall, even when there is physiological flow of fluid through the coronary vessel lumen.

Figure 8.

Pig coronary arteries cultured ex vivo in perfusion bioreactors with 25 mL/min flow either with control stent (A-C) or FasL- and DetaNONOate-releasing stent (D-F). FasL- and DetaNONOate-releasing stents created acellular regions in proximity to the stent struts (A). FasL- and NO donor-releasing stents (E) caused significantly higher apoptosis in SMCs of pig coronary arteries than did control stents (B, C and G) when cultured ex vivo with 25 mL/min flow (Experiments were repeated in triplicates, *: p < 0.05). SEM images showing the EC coverage on FasL- and NO donor-releasing stents after culture of pig coronary arteries in perfusion bioreactors for 14 days (H: 100x and I:200x magnification). Green: TUNEL (Scale bar: 100 μm).

In a subsequent experiment, pig coronary arteries with FasL and DetaNONoate-releasing stents were cultured in perfusion bioreactors for 14 days to evaluate the reestablishment of EC coverage on the stented lumens. SEM images show a confluent EC coverage, aligned in the direction of the flow, on and between the stent struts at the end of 14 days (Fig. 8 H and I). These results suggest that our FasL and DetaNONoate-releasing coating does not prevent EC migration and formation of a dense EC lining on the stent surface.

4. DISCUSSION

This is the first study to exploit the ability of NO to enhance Fas-mediated apoptosis to control smooth muscle growth, while also sparing endothelial cells. We showed the feasibility of a novel drug-eluting stent using a FasL-NO donor combination to potently inhibit smooth muscle cell proliferation with minimal damage to the endothelium.

Fas-mediated apoptosis is known to be effective in controlling intimal hyperplasia. Here, we studied the delivery of Fas ligand together with NO donor directly from a polymer coating to facilitate Fas-mediated apoptosis locally. We showed that concentrations of FasL and NO donors released from the EVAc matrix can be tuned to increase SMC apoptosis dramatically without harming endothelium (Fig.7 J–L). Moreover, the EVAc coating on the stents can deliver sufficient amounts of FasL and NO donor to trigger apoptosis only in the SMCs that are in the vicinity of the stent struts, despite arterial-like flow and shear applied during perfusion bioreactor culture (Fig. 8). Furthermore, a 14 day-culture experiment showed that ECs can migrate over the FasL and DetaNONoate-releasing stent coating and form a confluent layer on the stented lumen (Fig. 8 G and H). These results demonstrate the advantage of Fas-DetaNONoate combination over drugs delivered by currently available DES designs [12, 13, 39].

Our results show that FasL-NO donor combination increased relative IL-6 and IL-8 expression in SMCs, dramatically. This result is not surprising, since IL-6 and IL-8 upregulation in dying SMCs is reported in numerous studies [40, 41]. It is well known that IL-6 and IL-8 trigger SMC proliferation, decrease contractile markers, increase monocyte recruitment and upregulate monocyte-chemoattractant protein 1 (MCP-1) in vascular SMCs [40, 42]. However, it is also reported that delivery of Fas receptor or Fas ligand gene into the vessel wall reduced intimal hyperplasia, both by triggering vascular SMC apoptosis and by inhibiting T cell infiltration [20, 22, 28]. Conversely, Fas ligand deficiency was shown to increase T lymphocyte and macrophage infiltration into the vessel wall, and to enhance intimal hyperplasia in injured vessels [43]. Finally, and interestingly, patients with high levels of Fas and Fas ligand in their serum are more resistant to PCI-induced intimal hyperplasia [44]. Thus, our results, together with the findings of the previous studies, warrant further research testing FasL-NO donor releasing stents in animal models.

Sirolimus (rapamycin) and everolimus, both of which are mTOR inhibitors, are released by commonly used first- and second-generation DES. Considering the relatively high rates of the target vessel failure with currently available DES [12, 14, 16, 45, 46], there is a need for a novel drug strategy that prevents cell growth in and on the vessel wall to a much higher extent. In the present study, we compared the FasL-NO donor combination with two mTOR inhibitor drugs sirolimus and everolimus, in terms of their ability to prevent SMC growth and their effect on ECs. Sirolimus, at 10 nM concentration, has been previously shown to decrease the number of proliferating vascular SMC by 65% in 5 hours and total cell number by 55% in 24 hours, while decreasing proliferating ECs by 85% and total EC number by 55% [47]. Everolimus at 10 nM decreased SMC proliferation by 40% at 24 hours after application [48]. Therefore, we chose 10 nM as the working concentration for sirolimus and everolimus to compare their effects on SMCs with the combination of FasL and NO donor DetaNONOate, 400 ng/mL and 0.1mM, respectively. We showed that, at 48 hours after addition to the cell culture medium, the FasL-NO donor combination decreases total SMC number by 93±2.1% by triggering apoptosis, while sirolimus and everolimus decreased total cell number by 36±9.8% and 53±13.8%, respectively. At the end of 48 hours, the percentages of proliferating cells were 5.5%, 10.6% (not significantly different than the control group) and 1.4% for sirolimus, everolimus, and the FasL-NO donor combination.

Notably, the FasL-NO donor combination used here had minimal effect on EC viability, proliferation and function, unlike sirolimus and everolimus. mTOR inhibitors are known to reduce EC proliferation and migration [49] and eNOS expression [12, 39] in ECs. Non-complete or non-functional de novo endothelium creates the need for long-term dual antiplatelet therapy, which in turn leads to increased bleeding/stroke risk. Despite the low rates of definite thrombosis reported in long-term studies, the significant target vessel revascularization [46] and treatment failure may be related to non-functional endothelium. Here we showed that FasL-NO donor combination (400 ng/mL and 0.1 mM, respectively) resulted in more than 90% reduction in SMC numbers, and had no significant effect on arterial EC viability or proliferation. This same dose of FasL-NO reduced the eNOS gene expression by 27% in cultured ECs, while both sirolimus and everolimus caused more than a 67% decrease. Therefore, FasL-NO donor combination is an outstanding candidate drug combination to inhibit intimal hyperplasia and restenosis, while minimizing the damage to endothelial restoration.

While some prior studies have shown that NO donors promote EC survival by increasing resistance to external apoptotic stimuli or increasing vascular endothelial growth factor synthesis [50, 51], others report a negative effect on EC proliferation or adhesion protein expression at higher concentrations of NO donors [52–54]. Our results showed that FasL alone can inhibit SMC growth significantly, and even low concentration of NO donor (0.05mM DetaNONOate) can enhance this effect significantly, triggering Caspase 8 cleavage (Fig. S3 and Fig. 2). In contrast, even the combination of FasL and higher DetaNONOate concentration (0.2 mM) did not cause Fas-mediated apoptosis in ECs. However, 0.2 mM Detanonoate decreased EC proliferation significantly, independent of FasL (Fig. S4). The dip-coated stents were able to trigger significantly more SMC apoptosis in the stent proximity, while sparing a confluent EC layer in the lumen (Fig. 8). Yet, it is important to note that the doses that are delivered to the vessel wall must be determined precisely for subsequent translational in vivo studies. In the present study, we demonstrated the unique ability of the FasL-NO donor combination to differentially effect SMCs and ECs. However, further research is needed to model the effective release doses and the diffusion kinetics of NO donor- and FasL, with the goal of increasing SMC apoptosis significantly in the intima without affecting EC viability.

5. CONCLUSION

In the present study, we showed that combined delivery of FasL and NO donor can create dramatically different effects on vascular SMCs and ECs. This unique property of FasL and NO donor combination may result in a novel therapeutic application to control smooth muscle growth and cell infiltration into vessel walls following percutaneous interventions while sparing endothelial cells.