Nitric Oxide and Nanotechnology: a Novel Approach to Inhibit Neointimal Hyperplasia

CORE Metadata, citation and similar papers at core.ac.uk

Provided by Elsevier - Publisher Connector From the Society for Vascular Surgery

Nitric oxide and nanotechnology: A novel approach to inhibit neointimal hyperplasia

Muneera R. Kapadia, MD,ab Lesley W. Chow, BS,bc Nick D. Tsihlis, PhD,ab Sadaf S. Ahanchi, MD,ab Jason W. Eng,ab Jozef Murar, BA,ab Janet Martinez, AAS,ab Daniel A. Popowich, MD,ab Qun Jiang, MD,ab Joseph A. Hrabie, PhD,d Joseph E. Saavedra, PhD,d Larry K. Keefer, PhD,e James F. Hulvat, PhD,bc Samuel I. Stupp, PhD,bc and Melina R. Kibbe, MD,ab Chicago, Ill; and Frederick, Md

Objective: Nitric oxide (NO) has been shown to inhibit neointimal hyperplasia after arterial interventions in several animal models. To date, however, NO-based therapies have not been used in the clinical arena. Our objective was to combine nanofiber delivery vehicles with NO chemistry to create a novel, more potent NO-releasing therapy that can be used clinically. Thus, the aim of this study was to evaluate the perivascular application of spontaneously self-assembling NO-releasing nanofiber gels. Our hypothesis was that this application would prevent neointimal hyperplasia. Methods: Gels consisted of a peptide amphiphile, heparin, and a diazeniumdiolate NO donor (1-[N-(3-Aminopropyl)-N- (3-ammoniopropyl)]diazen-1-ium-1,2-diolate [DPTA/NO] or disodium 1-[(2-Carboxylato)pyrrolidin-1-yl]diazen-1- ium-1,2-diolate [PROLI/NO]). Nitric oxide release from the gels was evaluated by the Griess reaction, and scanning electron microscopy confirmed nanofiber formation. Vascular smooth muscle cell (VSMC) proliferation and cell death were assessed in vitro by 3H-thymidine incorporation and Personal Cell Analysis (PCA) system (Guava Technologies, Hayward, Calif). For the in vivo work, gels were modified by reducing the free-water content. Neointimal hyperplasia .(per group 6 ؍ after periadventitial gel application was evaluated using the rat carotid artery injury model at 14 days (n Inflammation and proliferation were examined in vivo with immunofluorescent staining against CD45, ED1, and Ki67 per group), and graded by blinded observers. Endothelialization was assessed by Evans blue injection at 2 ؍ at 3 days (n .(per group 3 ؍ days (n 7 Results: Both DPTA/NO and PROLI/NO, combined with the peptide amphiphile and heparin, formed nanofiber gels and released NO for 4 days. In vitro, DPTA/NO inhibited VSMC proliferation and induced cell death to a greater extent than PROLI/NO. However, the DPTA/NO nanofiber gel only reduced neointimal hyperplasia by 45% (intima/media [I/M] area ؎ ratio, 0.45 ؎ 0.07), whereas the PROLI/NO nanofiber gel reduced neointimal hyperplasia by 77% (I/M area ratio, 0.19 P < .05) vs control (injury alone I/M area ratio, 0.83 ؎ 0.07; P < .05). Both DPTA/NO and PROLI/NO nanofiber ,0.03 gels significantly inhibited proliferation in vivo (1.06 ؎ 0.30 and 0.19 ؎ 0.11 vs injury alone, 2.02 ؎ 0.20, P < .05), yet had minimal effect on apoptosis. Only the PROLI/NO nanofiber gel inhibited inflammation (monocytes and leukocytes). Both NO-releasing nanofiber gels stimulated re-endothelialization. Conclusions: Perivascular application of NO-releasing self-assembling nanofiber gels is an effective and simple therapy to prevent neointimal hyperplasia after arterial injury. Our study demonstrates that the PROLI/NO nanofiber gel most effectively prevented neointimal hyperplasia and resulted in less inflammation than the DPTA/NO nanofiber gel. This therapy has great clinical potential to prevent neointimal hyperplasia after open vascular interventions in patients. (J Vasc Surg 2008;47:173-82.)

Clinical relevance. Atherosclerosis affects >79 million Americans, many of whom require arterial intervention; unfortu- nately, treatment modalities often fail secondary to the development of neointimal hyperplasia, necessitating reinterven- tion. It is well established that nitric oxide (NO) inhibits neointimal hyperplasia, but no NO-based therapies have been clinically applied owing to various concerns. In this study, we demonstrated inhibition of neointimal hyperplasia in the rat carotid artery balloon-injury model using the local application of a gel made from NO and self-assembling nanofibers. By inhibiting neointimal hyperplasia and subsequent restenosis after arterial interventions, we aim to improve long-term patency rates and reduce the number of repeat interventions.

From the Division of Vascular Surgerya, Institute for BioNanotechnology in (W81XWH-05-1-0381), and by the generosity of Mrs Hilda Rosen- Medicineb, and Department of Material Science and Engineeringc, bloom. In addition, part of this research was supported with federal funds Northwestern University; Basic Research Program, SAIC-Frederick Incd; from the National Cancer Institute, NIH, under Contract N01-CO- and Laboratory for Comparative Carcinogenesis/Center for Cancer Re- 12400 with SAIC-Frederick Inc, and by the Intramural Research Program search, National Cancer Institute-Frederick.e of the NIH, National Cancer Institute, Center for Cancer Research. Competition of interest: Dr Hrabie, Dr Saavedra, and Dr Keefer are listed as Presented at the Society for Vascular Surgery Meeting, Baltimore, Md, June inventors on the patents filed by the US Government for the nitric oxide 7-10, 2007. Dr Kapadia is the recipient of the 2007 Lifeline Resident donors used in this article. Dr Hulvat is an employee of and has a financial Research Prize. interest in Nanotope Inc, which is commercializing compounds related to Reprint requests: Melina R. Kibbe, MD, Division of Vascular Surgery, those used in this study. Dr Stupp is a consultant for and has a financial Northwestern University, 201 E Huron St, Galter 10-105, Chicago, IL interest in Nanotope Inc. 60611 (e-mail: [email protected]). This work was supported in part by funding from the National Institutes of 0741-5214/$34.00 Health (NIH, 1K08HL084203 and 5 R01 EB003806-02), the Depart- Copyright © 2008 by The Society for Vascular Surgery. ment of Veterans Affairs, VA Merit Review Grant, the U.S. Army TATRC doi:10.1016/j.jvs.2007.09.005 173 JOURNAL OF VASCULAR SURGERY 174 Kapadia et al January 2008

Atherosclerosis is prevalent in all developed nations and is the leading cause of death and disability in the United States. Nearly 2400 deaths per day are caused by cardiovas- cular disease, and an estimated $432 billion was spent last year alone in the United States on cardiovascular disease, with a significant portion being attributed to the cost of repeat interventions.1 Although current treatment modal- ities improve blood flow, their long-term success is limited because of restenosis secondary to neointimal hyperplasia. Nitric oxide (NO) has many different vasoprotective properties.2 These include inhibition of platelet - aggrega tion,3 leukocyte chemotaxis,4 vascular smooth muscle cell (VSMC) proliferation and migration,5,6 and endothelial cell apoptosis.7 In addition, NO stimulates endothelial cell proliferation8 and is a potent vasodilator.9 These properties have led investigators to study the efficacy of NO-based therapies to inhibit the development of neointimal hyper- plasia in small- and large-animal models of arterial injury, vein bypass grafting, and prosthetic bypass 10-30grafting. Each of these therapies has demonstrated varying degrees of success, but none have been introduced to the clinical arena. Our goal is to combine nanotechnology with NO chemistry to create a potent NO-releasing therapy that can be used clinically to inhibit neointimal hyperplasia. The Fig 1. Our approach to inhibiting neointimal hyperplasia uses therapy should be simple to use, safe, and free from any side two components: (A) diazeniumdiolate nitric oxide donors, effects. Diazeniumdiolates are a class of NO donors that PROLI/NO, PAPA/NO, DPTA/NO, and DETA/NO (half- release NO spontaneously when placed in an aqueous lives listed are for physiologic conditions, 37°C and pH 7.4) and environment.31 Each mole of these compounds releases(B) heparin-binding peptide amphiphile. C, The peptide amphi- two moles of NO at varying rates, depending on the specific philes spontaneously form nanofibers in aqueous solutions when diazeniumdiolate.31 For our delivery vehicle, we usedexposed a to heparin, as depicted in this drawing. PROLI/NO, peptide amphiphile molecule that consists of a hydrophobic Disodium 1-[(2-carboxylato)pyrrolidin-1-yl]diazen-1-ium-1,2- fatty acid segment and a hydrophilic peptide segment con- diolate; PAPA/NO, 1-[N-(3-ammoniopropyl)-N-(n-propyl)amino] diazen-1-ium-1,2-diolate; DPTA/NO, 1-[N-(3-aminopropyl)-N- taining a sequence that binds to negatively charged (3-ammoniopropyl)]diazen-1-ium-1,2-diolate; DETA/NO, 1-[N- biopolymers, such as glycosaminoglycan heparin sulfate (2-aminoethyl)-N-(2-ammonioethyl)amino]diazen-1-ium-1,2- 32,33 (Fig 1B, ). When placed in an aqueous environment,diolate. aggregation of the hydrophobic segments induces sponta- neous assembly of the molecules into long nanofibers. The addition of heparin or other negatively charged molecules safety profiles.31,34 Gels were made by mixing -equal vol promotes nanofiber growth and networking, forming a umes of the following in order: peptide amphiphile (syn- macroscopic gel [FigC ].1 , The nanofiber gels are biocom-thesized as previously described32 and purified by high- patible and serve as an elegant delivery vehicle for the performance liquid chromatography) solution in ultrapure adventitial application of NO. Therefore, the aim of this water (3-wt %), 1-mol/L diazeniumdiolate (prepared as study was to examine the effect of NO-releasing nanofiber previously described35,36) in phosphate-buffered saline gels on the development of neointimal hyperplasia after (PBS), and heparin in PBS (2 wt %, heparin sodium salt, arterial injury, and our hypothesis was that our NO-based Sigma, St Louis, Mo). therapy would prevent neointimal hyperplasia. To determine NO production from each of the gels, 100 ␮L of gel was submerged in 500 ␮L of fresh PBS and METHODS maintained at 37°C. Every 24 hours, the supernatant was Nitric oxide-releasing nanofiber gels. The diazeni- removed to measure nitrite release (an indirect determinant umdiolates evaluated in this study include disodium 1-[(2- of NO production) by the Griess37 reactionAnother. 500 carboxylato)pyrrolidin-1-yl]diazen-1-ium-1,2-diolate (PROLI/ ␮L of fresh PBS was then added each day, and the cycle was NO), 1-[N-(3-ammoniopropyl)-N-(n-propyl)amino]dia- repeated until the gels no longer released NO. zen-1-ium-1,2-diolate (PAPA/NO), 1-[N-(3-aminopro- For in vivo experiments, the gel recipe was modified to pyl)-N-(3-ammoniopropyl)diazen-1-ium-1,2-diolate (DPTA/ reduce free-water content. PROLI/NO or DPTA/NO NO), and 1-[N-(2-aminoethyl)-N-(2-ammonioethyl)amino] (10 mg) was dissolved directly in 100 ␮L of 3-wt % heparin diazen-1-ium-1,2-diolate (DETA/NO; Fig A).1 , These (in PBS, pH 12.1 and 10.4, respectively, room tempera- were chosen based on their NO release rates as well as their ture). This solution was added to 100 ␮L of 4.5-wt % JOURNAL OF VASCULAR SURGERY Volume 47, Number 1 Kapadia et al 175

peptide amphiphile (pH 7.4, room temperature) in deion- Morphometric analysis. Carotid arteries harvested at ized water. The final solution was placed on a glass slide to 14 days were examined histologically for evidence of neo- gel ex vivo for 10 minutes before application. The final pH intimal hyperplasia using routine hematoxylin and eosin for the PROLI/NO and DPTA/NO nanofiber gels was staining. A modified Verhoeff von Gieson stain was used to 10.87 and 10.27, respectively. The control gel was made up evaluate elastin and collagen. Digital images were collected of 100 ␮L of 3-wt% heparin and 100 ␮L of 4.5-wt% peptide with light microscopy using an Olympus BHT microscope amphiphile; it did not contain PROLI/NO or DPTA/NO. (Melville, NY) with ϫ4, ϫ10, and ϫ40 objectives. Six Scanning electron microscopy. Gels were dehy- evenly spaced sections through each injured carotid artery drated in a graded ethanol series and dried by the critical were morphometrically analyzed. Lumen area, intimal area point method in a Samdri 790A critical point drying appa- (I), medial area (M), and external elastic lamina circumfer- ratus (Tousimis Research Corporation, Rockville, Md). ence were measured (arbitrary units) using ImageJ software The dried samples were sputter-coated with gold/palla- (National Institutes of Health, Bethesda, Md). dium (3nM) using a Cressington Sputter Coater (Cressing- Immunohistochemistry. Carotid arteries harvested ton Scientific Instruments, Watford, UK) before imaging at 3 days were examined for evidence of inflammation and with a Hitachi S-4800 scanning electron microscope (Hi- proliferation using immunofluorescent staining. Sections tachi Kokusai Electric America, Woodbury, NY). fixed with 2% paraformaldehyde or acetone were perme- Cell culture. Vascular smooth muscle cells were iso- abilized with Triton-X100 in PBS. Sections were then lated and cultured from the abdominal aorta of Sprague- blocked with goat serum (Sigma, St Louis, Mo) in 0.5% Dawley rats (Harlan, Indianapolis, Ind) using the explant bovine serum albumin (Vector, Burlingame, Calif). Pri- method38 and maintained as previously described.37 mary antibody in bovine serum albumin was applied for 1 Proliferation assay. Vascular smooth muscle cells hour: anti-ED1 (macrophage, 1:500; Serotec, Raleigh, plated in 12-well plates (4 ϫ 104 cells/well) were growth- NC), anti-CD45 (lymphocyte, 1:500; BD Pharmingen, arrested for 24 hours with starvation media that contained San Diego, Calif) or anti-Ki67 (1:100; BD Biosciences, San Jose, Calif). Secondary antibody in bovine serum albumin no fetal bovine serum. Cells were then exposed to media was applied for 1 hour (goat antimouse Alexa Fluor 555, containing one of the two NO donors (PROLI/NO or 1:3000; Invitrogen, Carlsbad, Calif). Coverslips were af- DPTA/NO powder) in the presence of tritiated 3H-thymi- fixed with gelvatol. dine (5 ␮Ci/mL, PerkinElmer, Wellesley, Mass) for an From each animal, six sections from the area of injury additional 24 hours. The incorporation of 3H-thymidine were stained. Eight independent blinded observers qualita- into trichloroacetic acid–precipitated DNA was quantified tively graded the inflammation and proliferation in each by scintillation counting. treatment group on a scale of 0 to 3. Digital images were Cell death assay. Vascular smooth muscle cells plated acquired using a Zeiss LSM-510 microscope (Hallberg- in 6-well plates (1 ϫ 105 cells/well) were growth-arrested moos, Germany) at ϫ40. for 24 hours with starvation media, after which they were Apoptosis was evaluated in carotid arteries harvested at exposed to media with one of the two NO donors 3 and 14 days by terminal deoxynucleotidyl transferase- (PROLI/NO or DPTA/NO powder) for 24 hours. Cells mediated deoxy uridine triphosphate biotin nick-end label- were trypsinized, collected, pelleted, and resuspended in ing (TUNEL) using a commercial system according to the ␮ ␮ PBS (250 L). This suspension (40 L) was added to 160 manufacturer’s instructions (DeadEnd Colorimetric ␮ L of Guava ViaCount Reagent, and cell death was assessed TUNEL system; Promega, Madison, Wis). using the Personal Cell Analysis (PCA) system (Guava Endothelialization. Rat carotid arteries were har- Technologies, Hayward, Calif). vested at 7 days. Thirty minutes before euthanasia, rats Animal surgery. All animal procedures were per- received an intravenous injection of Evans blue dye (0.5 mL formed in accordance with the Guide for the Care and Use of 0.5%, Sigma). Carotid arteries were procured after in-situ of Laboratory Animals (National Institutes of Health Pub- perfusion with 500 mL of PBS and photographed. Blue lication 85-23, 1996) and approved by the Northwestern staining indicated area of increased endothelial permeabil- University Animal Care and Use Committee. Adult male ity. Images were qualitatively analyzed by two independent Sprague-Dawley rats weighing 350 to 400 g underwent observers. 20,37 carotid artery balloon injury as previously described. Statistical analysis. Results are expressed as mean Ϯ After injury and restoration of flow, one of the NO-eluting standard error of the mean. Differences between multiple nanofiber gels or control therapies (200 ␮L) was applied to groups were analyzed using one-way analysis of variance the external surface of the injured common carotid artery with the Student-Newman-Keuls post hoc test for all pair- and the neck incision was closed. Rats were euthanized at 3 wise comparisons using SigmaStat software (SPSS, Chi- days (n ϭ 2 per group), 7 days (n ϭ 3 per group), and 14 cago, Ill). Statistical significance was assumed when P Ͻ days (n ϭ 6 per group). .05. Tissue processing. Carotid arteries were harvested af- ter in-situ perfusion-fixation with PBS (250 mL) and 2% RESULTS paraformaldehyde (500 mL). Tissue was processed as pre- Nitric oxide release from nanofiber gels. Initially, viously described.20 the diazeniumdiolate nanofiber gels tested included JOURNAL OF VASCULAR SURGERY 176 Kapadia et al January 2008

PROLI/NO (125 to 500␮M), there was no statistically significant effect vs control. In contrast, the longer-acting NO donor, DPTA/NO, induced a concentration-depen- dent inhibition of VSMC proliferation (125 to 1000␮M; P Ͻ .05). Furthermore, the antiproliferative effect of DPTA/NO was more potent than PROLI/NO: DPTA/NO inhibited VSMC proliferation by 51% to 78% (P Ͻ .05). PROLI/NO and DPTA/NO induce minimal vas- cular smooth muscle cell death in vitro. To determine if PROLI/NO and DPTA/NO induce VSMC death in vitro, the Guava PCA system was implementedC (andFig ). D3, Although not statistically significant, there was a trend towards increased cell death with increasing concentrations of PROLI/NO (125 to 1000␮M). At lower concentra- tions, DPTA/NO (125 to 500␮M) did not induce statis- tically significant VSMC death. However, at the highest concentration evaluated, DPTA/NO (1000␮M) induced 21.4% Ϯ 0.9% cell death in VSMCs (P Ͻ .05). Nitric oxide-releasing nanofibers inhibit neointi- mal hyperplasia in the rat carotid artery injury model. Fig 2. A, Nitrite release from nanofiber, PROLI/NO nanofiber, Because of the multiple-day NO-release and the ability of and DPTA/NO nanofiber gels in vitro, determined using the PROLI/NO and DPTA/NO to form gels with peptide Griess reaction (n ϭ 2 per group). Data are representative of three separate experiments. Scanning electron microscopy images amphiphile and heparin, we examined the in vivo effect of (ϫ4000 original magnification) of the (B) nanofiber gel, (C) gel application on neointimal hyperplasia after rat carotid PROLI/NO nanofiber gel, and (D) DPTA/NO nanofiber gel. artery balloon injury (Fig 4). The gel was applied to the NO, Nitric oxide. See Fig 1 for the expansion of the abbreviations adventitial surface of the carotid artery only. Balloon injury PROLI/NO and DPTA/NO. produced reproducible neointimal hyperplasia at 14 days (Table; Fig A4)., With the application of the nanofiber control gel, there was no statistically significant difference PROLI/NO, PAPA/NO, DPTA/NO, and DETA/NO. vs injury alone for either the intimal area (3.20 Ϯ 0.34 vs When combined with peptide amphiphile and heparin, 3.63 Ϯ 0.35, respectively) or the I/M area ratio (0.69 Ϯ DETA/NO and PAPA/NO did not form a consistent gel; 0.07 vs 0.83 Ϯ 0.07, respectively). However, with applica- instead, most of the product remained liquid. We therefore tion of PROLI/NO and DPTA/NO nanofiber gels, neo- used only the PROLI/NO and DPTA/NO nanofiber gels. intimal hyperplasia was significantly inhibited. The To determine the amount and duration of NO release PROLI/NO nanofiber gel showed an 80% reduction of from the gels (100 ␮L), Griess reactions were conducted. intimal area (0.71 Ϯ 0.14 vs injury alone, P Ͻ .05) and a Nitrite release was observed for 4 days from both the 77% reduction of I/M area ratio (0.19 Ϯ 0.03 vs injury PROLI/NO and DPTA/NO nanofiber gels, with most of alone, P Ͻ .05). Although not as impressive, the the nitrite release occurring in the first 2 A).days DPTA/NO(Fig 2, nanofiber gel demonstrated modest inhibition The PROLI/NO nanofiber gel appeared to have steady of neointimal hyperplasia: 40% reduction of intimal area nitrite release for the initial 2 days, whereas the DPTA/NO (2.19 Ϯ 0.34 vs injury alone, P Ͻ .05) and 45% reduction nanofiber gel had similar release on day 1 compared with of I/M area ratio (0.45 Ϯ 0.07 vs injury alone, P Ͻ .05). the PROLI/NO nanofiber gel, but on day 2, the nitrite There was a statistically significant difference in inhibition release doubled. Both PROLI/NO and DPTA/NO nano- of neointimal hyperplasia between PROLI/NO and fiber gels had a significant decline in NO release on day 3 DPTA/NO nanofiber gels (P Ͻ .05). No statistically sig- and 4, and by day 5, there was no difference compared with nificant difference was found between the medial areas for control gels. Scanning electron microscopy of the nanofi- injury alone, nanofiber control gel, and DPTA/NO nano- ber, PROLI/NO nanofiber, and DPTA/NO nanofiber fiber gel groups; however, the PROLI/NO nanofiber gel gels revealed distinct nanofiber formation B-D(Figs). 2, treatment resulted in a 15% medial area reduction vs injury PROLI/NO and DPTA/NO inhibit vascular alone (P Ͻ .05). There was no difference in the circumfer- smooth muscle cell proliferation in vitro. In an effort to ence of the external elastic lamina between the groups. It characterize the effects of PROLI/NO and DPTA/NO on therefore appears that our NO-nanofiber gels had very little proliferation of VSMC, an in vitro 3H-thymidine prolifer- effect on adaptive vascular remodeling but had a large effect ation assay was conducted (AFig and 3B,). The highest on inhibiting the development of neointimal hyperplasia. concentration of PROLI/NO (1000␮M) inhibited VSMC Nitric oxide-releasing nanofiber gels inhibit inflam- proliferation by 32% (P Ͻ .05). Although there was a trend mation and proliferation in vivo. To characterize the toward less proliferation with lower concentrations of effect of the PROLI/NO and DPTA/NO nanofiber gels JOURNAL OF VASCULAR SURGERY Volume 47, Number 1 Kapadia et al 177

Fig 3. A, B, Effect of PROLI/NO and DPTA/NO on vascular smooth muscle cell proliferation and (C, D) cell death in vitro. Proliferation and cell death were quantified using tritiated thymidine and the Guava Personal Cell Analysis (PCA) system (Hayward, Calif), respectively, (n ϭ 3 per group). Data are representative of 3 separate experiments. See Fig 1 for expansion of the abbreviationsPROLI/NO andDPTA /NO. on neointimal hyperplasia at a cellular level, we conductedNitric oxide-releasing nanofiber gels induce little immunohistochemical staining for markers of inflamma- apoptosis in vivo. We sought to determine whether the tion and proliferation at 3 days (Fig 5). Similar NO-releasingnumbers of nanofiber gels limited neointimal hyperplasia by monocytes (anti-ED1) were seen in the injury alone, nano- inducing apoptosis in vivo. Sections stained using TUNEL fiber, and DPTA/NO groups. In the PROLI/NO group, did not reveal any distinct patterns and showed little apoptosis however, there was a reduction in the observed mononu- among the different treatment groups (data not shown). clear infiltrate vs injury alone. When the CD45-positive These data are fairly consistent with the in vitro data. leukocytes were examined, the injury alone and Nitric oxide-releasing nanofiber gels stimulate re- DPTA/NO groups appeared to have similar leukocyte establishment of the endothelial barrier after arterial infiltration; however, nanofiber and PROLI/NO nanofiber injury. To determine the effect of the NO-releasing nano- groups had fewer leukocytes compared with injury alone. fiber gels on the endothelium after balloon injury, animals To summarize, the DPTA/NO nanofiber gel appeared to were injected with Evans blue dye premortem; arterial have no effect on inflammation compared with injury segments with increased endothelial permeability stained alone, whereas the PROLI/NO nanofiber gel appeared to blue (Fig 6). By qualitative analysis, untreated carotid ar- inhibit inflammation after arterial injury. teries and arteries treated with nanofiber gel exhibited the Next we determined the effect of the nanofiber gels on greatest amount of blue staining. Arteries treated with cellular proliferation (anti-Ki67) in vivo. Injury alone ex- PROLI/NO nanofiber gel demonstrated a very small hibited the most proliferation, and similar levels were ob- amount of blue staining, and those treated with served with the nanofiber gel; most of the proliferating cells DPTA/NO nanofiber gel had little to no staining, suggest- appeared to be in the adventitia. Both PROLI/NO and ing little endothelial permeability by 7 days. These data DPTA/NO nanofiber gels demonstrated an inhibition of suggest that both NO-releasing nanofiber gels stimulate cellular proliferation in vivo, and the effect appeared to be the re-establishment of an intact endothelial layer. greatest on the adventitial cells; however, the PROLI/NO nanofiber gel seemed to inhibit proliferation dramatically, DISCUSSION whereas the DPTA/NO nanofiber gel inhibited prolifera- Neointimal hyperplasia develops after arterial injury and tion modestly. This pattern mirrored the effect on neointi- endothelial denudation through a series of events that in- mal hyperplasia. cludes platelet adherence and aggregation, leukocyte chemo- JOURNAL OF VASCULAR SURGERY 178 Kapadia et al January 2008

Fig 4. Rat carotid artery sections from uninjured, injury alone, nanofiber gel, PROLI/NO nanofiber gel, and DPTA/NO nanofiber gel animals euthanized at 14 days (n ϭ 6 per group). A, Representative sections (ϫ100 original magnification) from each group using routine staining with hematoxylin and eosin (H&E) and Verhoff-van Gieson (VvG). B, Graphic representation of intima/media (I/M) area ratio presented with the standard error of the mean. Morphometric analysis conducted on 6 sections per rat. Units are arbitrary. IA, Injury alone, NG, nanofiber gel, PNG, PROLI/NO nanofiber gel; DNG, DPTA/NO nanofiber gel. NO, Nitric oxide. See Fig 1 for the complete expansion of the abbreviations PROLI/NO and DPTA/NO.

Table. Morphometric analysis of carotid arteries 14 days after balloon injury

Treatment group* Circumference Lumen area Intimal area Medial area I/M area ratio I/(IϩM)

Injury alone 1.64 Ϯ 0.03 8.53 Ϯ 0.77 3.63 Ϯ 0.35 4.37 Ϯ 0.17 0.83 Ϯ 0.07 0.42 Ϯ 0.02 Nanofiber gel 1.64 Ϯ 0.02 12.00 Ϯ 0.31† 3.20 Ϯ 0.34 4.56 Ϯ 0.16 0.69 Ϯ 0.07 0.38 Ϯ 0.02 PROLI/NO 1.68 Ϯ 0.02 15.57 Ϯ 0.43† 0.71 Ϯ 0.14†‡ 3.70 Ϯ 0.17†‡ 0.19 Ϯ 0.03†‡ 0.14 Ϯ 0.02† nanofiber gel DPTA/NO 1.68 Ϯ 0.03 12.75 Ϯ 0.66† 2.19 Ϯ 0.34† 4.59 Ϯ 0.17 0.45 Ϯ 0.07† 0.26 Ϯ 0.03† nanofiber gel

I, Intimal; M, Medial; PROLI/NO, disodium 1-[(2-carboxylato)pyrrolidin-1-yl]diazen-1-ium-1,2-diolate; DPTA/NO, 1-[N-(3-aminopropyl)-N-(3-ammo- niopropyl)]diazen-1-ium-1,2-diolate *All groups had 6 rats, 6 sections per rat. Data are presented as mean Ϯ SE. †P Ͻ .05 vs injury alone. ‡P Ͻ .05 vs DPTA/NO nanofiber gel. taxis, and VSMC and fibroblast proliferation and migra- period, indicating either (1) an increased half-life for the tion.39-42 This cascade often results in luminal narrowingNO donorand in the nanofiber gel environment by shielding it restenosis. We examined the effects of two NO-releasing from hydrogen ions or (2) prolonged retention of NO itself nanofiber gels on neointimal hyperplasia in an attempt to within the gel. NO retention within the hydrophobic cores develop a simple-to-use, potent, NO-based clinical therapy. of the nanofiber would be expected given the low dipole An unexpected result from this research is that in vitro, moment of NO. Furthermore, Moller43 demonstrated et al PROLI/NO had little effect, whereas DPTA/NO demon- NO partitioning to the hydrophobic phase of biologic strated inhibition of VSMC proliferation and induction of membranes. Although the exact disposition of NO and its cell death. The NO-eluting nanofiber gels created using diazeniumdiolate precursor in the nanofiber gels will re- these two diazeniumdiolates exhibited prolonged nitrite quire further study, an enhanced therapeutic effect of NO release (a decomposition product of NO) during a 4-day in this novel delivery vehicle is evident. JOURNAL OF VASCULAR SURGERY Volume 47, Number 1 Kapadia et al 179

Fig 5. Immunofluorescence staining for markers of inflammation and proliferation. A, Representative sections (ϫ400 original magnification) from uninjured, injury alone, nanofiber gel, PROLI/NO nanofiber gel, and DPTA/NO nanofiber gel treatment groups (n ϭ 2 per group, 3 sections per animal). Positive staining is indicated by red and elastic lamina is green. Graphic representation of scoring by 8 blinded observers (scale, 0-3) for (B) monocytes (anti-ED1), (C) leukocytes (anti-CD45), and (D) proliferation (anti-Ki67). Each data point represents scoring for a particular animal. U, Uninjured; IA, injury alone; NG, nanofiber gel; NO, nitric oxide; PNG, PROLI/NO nanofiber gel; and DNG, DPTA/NO nanofiber gel. See Fig 1 for the complete expansion of PROLIthe /NOabbreviations and DPTA/NO.

Several investigators have successfully demonstrated Kown et 30 aldemonstrated that L-arginine polymer- inhibition of neointimal hyperplasia using NO-based ap- treated vein grafts exhibited 43% less neointimal hyperpla- proaches in animal models.10-30 These approaches include sia in a rabbit vein bypass30 Kaulmodel. et 26 alapplied a inhalational NO therapy,10 L-arginine supplementa- biodegradable polymer (polylactic-polyglycolic acid) con- tion,11-13 systemic administration of NO donors,14 nitric taining spermine/NO (2.5% w/w) to the periadventitial oxide synthase gene therapy,15-21 and local delivery ofsurface of balloon-injured rat ileofemoral arteries, which intraluminal and perivascular NO donors.22-27,30 How- resulted in a 69% reduction in intimal area.24 Chaux et al ever, none of these therapies have been implemented clin- used the same biodegradable polymer with spermine/NO ically because of various concerns, including lack of signif- in a rabbit jugular vein grafting model. At 28 days there was icant and durable effect, systemic side effects, safety a 41% reduction in intimal area 24vs control. concerns, and complicated delivery schemes. Two studies from Dr West’s group utilized polyethyl- Our novel therapy uses periadventitial delivery. Be- ene glycol (PEG)-based hydrogels covalently modified with cause NO is freely diffusible, it penetrates all layers of the S-nitrosocysteine groups.44,45 After rat carotid artery- bal vascular wall. Thus, adventitial delivery effectively reduces loon injury, the hydrogels were applied to the periadventi- neointimal hyperplasia by inhibiting both adventitial fibro- tial arterial surface. Photopolymerization of the hydrogels blasts and medial VSMCs. Another advantage of localized was conducted with an ultraviolet light, and NO release was therapy is that the NO delivery is concentrated at the site of observed for approximately 24 hours. A 75% inhibition of injury. This avoids systemic side effects, such as hypoten- neointimal hyperplasia was seen in arteries harvested at 14 sion, coagulopathy, and headaches. days. These studies thus demonstrate, similar to our study, JOURNAL OF VASCULAR SURGERY 180 Kapadia et al January 2008

gel adheres, ensuring that the treatment reaches the tar- geted location as accurately as possible. An interesting finding is the opposing results from the in vitro and in vivo studies. DPTA/NO more effectively inhibited VSMC proliferation and induced VSMC cell death in vitro, yet the PROLI/NO nanofiber gel more successfully inhibited neointimal hyperplasia in vivo. This may be occurring for several reasons. The PROLI/NO powder has a very short half-life; therefore, its effect in vitro may not be sustained enough to exert an effect. When administered as part of a gel, the PROLI/NO nanofiber gel was shown to release NO for 4 days, and this prolonged Fig 6. Re-endothelialization of injured rat carotid arteries. NO release may partially increase the effect seen. The Groups include injury alone and injury with nanofiber, PROLI/NO nanofiber gel also inhibited inflammation, as PROLI/NO nanofiber, or DPTA/NO nanofiber gels (n ϭ 3 per evidenced by less monocyte and leukocyte infiltration, group). Arteries were procured at 7 days postinjury. Arterial areas whereas the DPTA/NO nanofiber gel did not. Inhibition lacking endothelium stain (arrow)blue . See Fig 1 for the completeof leukocyte infiltration would impact the arterial injury expansion of the abbreviations PROLI/NO and DPTA/NO. cascade by limiting growth factor and cytokine secretion and subsequent VSMC proliferation and migration. Thus, these properties of the PROLI/NO nanofiber gel may that perivascular NO application is effective in reducing contribute to its heightened efficacy. neointimal hyperplasia. Another fascinating aspect of this research is the effec- Our NO-eluting nanofiber therapy has several addi- tiveness of the NO donors with use of the nanofiber deliv- tional attractive features. The first is the use of customizable ery vehicle. Prior work in our laboratory examined the nanostructures formed from peptide amphiphiles as a de- effect of 20 mg of PROLI/NO powder applied to the livery vehicle. These molecules spontaneously assemble adventitial surface of the rat carotid artery after balloon into nanofiber gels under physiologic conditions, and sim- injury and demonstrated 86% inhibition of neointimal hy- ilar nanofibers support growth and differentiation of vari- perplasia at 14 days, as measured by the I/M area ratio ous cell types in 46,47vitro. The spontaneous formation is(unpublished data). In this study, half as much important because it requires no additional activators, such PROLI/NO (10 mg) resulted in a similar degree of inhi- as ultraviolet light. Thus, our approach is simple to use. bition (77%), and this may be because the gel application is Peptide amphiphile nanofibers have also been used as more controlled and allows for more even distribution of bioactive coatings for tissue engineering 48implants, and the drug on the entire external surface of the artery. In preliminary testing in murine models has revealed no large- addition, the peptide amphiphile is known to specifically scale immune response to the nanofibers.49 Rajangam et bind heparin, an inherent property that may increase the al32 recently reported a peptide amphiphile moleculelocal that heparin concentration at the site of injury and poten- was designed to bind heparin. Heparin-binding nanofiber tiate the antiproliferative actions of NO. There is prece- gels were shown in vitro and in vivo to bind and control the dence for combining diazeniumdiolate NO donors with bioavailability of various growth factors (vascular endothe- heparin: Saavedra et50 synthesizedal a heparin-methoxym- lial growth factor and fibroblast growth 32factor-2), pre- ethyl-PIPERAZI/NO conjugate, which released NO on senting another potential therapeutic axis for inhibiting hydrolysis and inhibited adenosine 5=-diphosphate-in- neointimal hyperplasia. In vivo testing of this molecule in duced platelet aggregation but also retained significant other animal models is underway, and thus far has revealed heparin-like anticoagulant properties in vitro. The two no significant biocompatibility issues (unpublished data). complementary antithrombotic mechanisms are likely oc- The other component of our approach uses diazenium- curring with application of our treatment gels as well. diolates. Diazeniumdiolates are advantageous because of Regardless, we found that use of this nanofiber gel delivery their predictable, spontaneous NO release in an aqueous vehicle provided advantages over the PROLI/NO powder environment, and they can be tailored to a wide range of alone and will enable transition of this therapy to the NO release rates.31 Thus, our unique approach to marryclinical arena. these two entities—peptide amphiphiles and the diazeni- Although our NO-releasing nanofiber gel is an exciting umdiolates—has several beneficial therapeutic qualities. therapeutic candidate, one potential limiting factor is the Another important physical characteristic of our therapy is potentiated antithrombotic actions of heparin and NO. that the peptide amphiphile and heparin solutions are stable Because both of these products inhibit platelet adherence when refrigerated. When combined with the NO-heparin and aggregation, their combination may lead to increased solution, the self-assembling nature of the peptide amphi- bleeding if proper hemostasis is not achieved. In five ani- phile system allows the nanofibers to form a gel nearly mals treated with the DPTA/NO nanofiber gel and three instantly and thus requires very little preparation time. animals treated with PROLI/NO nanofiber gel, a superfi- Finally, when the therapy is applied to the injured area, the cial hematoma was observed at euthanasia (14 days) vs JOURNAL OF VASCULAR SURGERY Volume 47, Number 1 Kapadia et al 181

none in the injury alone group. To overcome this issue, REFERENCES heparin could be substituted with another sulfated glycos- 1. Rosamond W, Flegal K, Friday G, Furie K, Go A, Greenlund K, et al. aminoglycan without antithrombin-binding properties that Heart disease and stroke statistics - 2007 update - A report from the would also induce gel formation. One such potential sub- American Heart Association Statistics Committee and Stroke Statistics stitute is heparan sulfate; however, it is much more costly Subcommittee. Circulation 2007;115:E69-171. than heparin. 2. Bredt DS, Snyder SH. Nitric oxide: a physiologic messenger molecule. Annu Rev Biochem 1994;63:175-95. Another limitation is that this study only compares the 3. Radomski MW, Palmer RM, Moncada S. Endogenous nitric oxide outcome of one delivery vehicle, namely the self-assembled inhibits human platelet adhesion to vascular endothelium. Lancet 1987; nanofiber gel. A future study could analyze the efficacy of 2:1057-8. this gel delivery vehicle against other gel delivery vehicles 4. Kubes P, Suzuki M, Granger DN. Nitric oxide: an endogenous modu- using the same NO donor. This form of analysis would lator of leukocyte adhesion. Proc Natl Acad SciUSA1991;88:4651-5. 5. Garg UC, Hassid A. Nitric oxide-generating vasodilators and 8-bromo- provide data with respect to the gel delivery vehicle alone. cyclic guanosine monophosphate inhibit mitogenesis and proliferation Finally, we do not know if any of the hydrolyzed of cultured rat vascular smooth muscle cells. J Clin Invest 1989;83: breakdown products or if the peptide amphiphile or by- 1774-7. products have effects on the vasculature or, specifically, 6. Dubey RK, Jackson EK, Lüscher TF. Nitric oxide inhibits angiotensin neointimal hyperplasia. Some of these molecular byprod- II-induced migration of rat aortic smooth muscle cell. Role of cyclic- nucleotides and angiotensin1 receptors. J Clin Invest 1995;96:141-9. ucts include the amino acids alanine, glycine, lysine, 7. Tzeng E, Kim YM, Pitt BR, Lizonova A, Kovesdi I, Billiar TR. Adeno- leucine, and arginine; salts from sodium hydroxide, hydro- viral transfer of the inducible nitric oxide synthase gene blocks endothe- chloric acid, and possibly trifluoroacetic acid also may be lial cell apoptosis. Surgery 1997;122:255-63. present. The nanofiber gel alone serves as a control, and no 8. Ziche M, Morbidelli L, Masini E, Amerini S, Granger HJ, Maggi CA, et statistically significant reduction in neointimal hyperplasia al. Nitric oxide mediates angiogenesis in vivo and endothelial cell growth and migration in vitro promoted by substance P. J Clin Invest was observed with its application, suggesting that the pep- 1994;94:2036-44. tide amphiphile products do not affect results. This, how- 9. Ignarro LJ, Buga GM, Wood KS, Byrns RE, Chaudhuri G. Endotheli- ever, is something that may need to be examined more um-derived relaxing factor produced and released from artery and vein specifically in the future. is nitric oxide. Proc Natl Acad SciUSA1987;84:9265-9. 10. Lee JS, Adrie C, Jacob HJ, Roberts JDJ, Zapol WM, Bloch KD. Chronic inhalation of nitric oxide inhibits neointimal formation after CONCLUSION balloon-induced arterial injury. Circ Res 1996;78:337-42. By combining NO and nanotechnology, we created a 11. Davies MG, Kim JH, Dalen H, Makhoul RG, Svendsen E, Hagen PO. novel approach that successfully inhibited neointimal hy- Reduction of experimental vein graft intimal hyperplasia and preserva- tion of nitric oxide-mediated relaxation by the nitric oxide precursor perplasia in the rat carotid artery injury model. Although L-arginine. Surgery 1994;116:557-68. both NO-eluting nanofiber gels inhibited neointimal hy- 12. McNamara DB, Bedi B, Aurora H, Tena L, Ignarro LJ, Kadowitz PJ, et perplasia, the PROLI/NO nanofiber gel was clearly more al. L-arginine inhibits balloon catheter-induced intimal hyperplasia. effective. Before this therapy can be used in the clinical Biochem Biophys Res Commun 1993;193:291-6. arena, it must be evaluated in a large-animal model, and 13. Chen C, Mattar SG, Lumsden AB. Oral administration of L-arginine reduces intimal hyperplasia in balloon-injured rat carotid arteries. J Surg long-term studies will need to be conducted for toxicity as Res 1999;82:17-23. well as efficacy. Overall, this therapy has promising clinical 14. Groves PH, Banning AP, Penny WJ, Newby AC, Cheadle HA, Lewis potential as a novel NO-based therapy for open vascular MJ. The effects of exogenous nitric oxide on smooth muscle cell and cardiovascular surgical procedures for the prevention of proliferation following porcine carotid angioplasty. Cardiovasc Res restenosis and the associated patient morbidity. 1995;30:87-96. 15. von der Leyen HE, Mann MJ, Dzau VJ. Gene inhibition and gene We would like to express our thanks to the Northwest- augmentation for the treatment of vascular proliferative disorders. ern University Institute for BioNanotechnology in Medi- Seminars in Interventional Cardiology 1996;1:209-14. 16. Chen L, Daum G, Forough R, Clowes M, Walter U, Clowes AW. cine and to Lynnette Dangerfield for her administrative Overexpression of human endothelial nitric oxide synthase in rat vascu- support. lar smooth muscle cells and in balloon-injured carotid artery. Circ Res 1998;82:862-70. AUTHOR CONTRIBUTIONS 17. Janssens S, Flaherty D, Nong Z, Varenne O, Van Pelt N, Haustermans C, et al. Human endothelial nitric oxide synthase gene transfer inhibits Conception and design: MK, JFH, SS, MRK* vascular smooth muscle cell proliferation and neointima formation after Analysis and interpretation: MK, LC, JAH, JS, LK, JFH, balloon injury in rats. Circulation 1998;97:1274-81. MRK* 18. Varenne O, Pislaru S, Gillijns H, Van Pelt N, Gerard RD, Zoldhelyi P, Data collection: MK, LC, SA, JE, JM, NT, DP, QJ, et al. Local adenovirus-mediated transfer of human endothelial nitric oxide synthase reduces luminal narrowing after coronary angioplasty in Writing the article: MK, MRK* pigs. Circulation 1998;98:919-26. Critical revision of the article: MK, LC, NT, JAH, JS, LK, 19. Kibbe MR, Tzeng E, Gleixner SL, Watkins SC, Kovesdi I, Lizonova A, JFH, MRK* et al. Adenovirus-mediated gene transfer of human inducible nitric Final approval of the article: LK, MRK* oxide synthase in porcine vein grafts inhibits intimal hyperplasia. J Vasc Statistical analysis: MK, MRK* Surg 2001;34:156-65. 20. Shears LL, Kibbe MR, Murdock AD, Billiar TR, Lizonova A, Kovesdi I, Obtained funding: LK, SS, MRK* et al. Efficient inhibition of intimal hyperplasia by adenovirus-mediated Overall responsibility: MK, MRK* inducible nitric oxide synthase gene transfer to rats and pigs in vivo. *Dr Kibbe J Am Coll Surg 1998;187:295-306. JOURNAL OF VASCULAR SURGERY 182 Kapadia et al January 2008

21. Wang K, Kessler PD, Forudi F, Zhou Z, Zhou X, Tarakji K, et al. Local 35. Saavedra JE, Southan GJ, Davies KM, Lundell A, Markou C, Hanson adenoviral-mediated inducible nitric oxide synthase (iNOS) gene trans- SR, et al. Localizing antithrombotic and vasodilatory activity with a fer inhibits neointimal formation in the porcine coronary stented novel, ultrafast nitric oxide donor. J Med Chem 1996;39:4361-5. model. Am J Cardiol 2001;88:51G-2G. 36. Hrabie JA, Klose JR, Wink DA, Keefer LK. New nitric oxide-releasing 22. Marks DS, Vita JA, Folts JD, Keaney JFJ, Welch GN, Loscalzo J. zwitterions derived from polyamines. J Org Chem 1993;58:1472-6. Inhibition of neointimal proliferation in rabbits after vascular injury by 37. Kibbe MR, Li J, Nie S, Watkins SC, Lizonova A, Kovesdi I, et al. a single treatment with a protein adduct of nitric oxide. J Clin Invest Inducible nitric oxide synthase (iNOS) expression upregulates p21 and 1995;96:2630-8. inhibits vascular smooth muscle cell proliferation through p42/44 23. Schwarzacher SP, Lim TT, Wang BY, Kernoff RS, Niebauer J, Cooke mitogen-activated protein kinase activation and independent of p53 and cyclic guanosine monophosphate [In Process Citation]. J Vasc Surg JP, et al. Local intramural delivery of L-arginine enhances nitric oxide 2000;31:1214-28. generation and inhibits lesion formation after balloon angioplasty. 38. Yu SM, Hung LM, Lin CC. cGMP-elevating agents suppress prolifer- Circulation 1997;95:1863-9. ation of vascular smooth muscle cells by inhibiting the activation of 24. Chaux A, Ruan XM, Fishbein MC, Ouyang Y, Kaul S, Pass JA, et al. epidermal growth factor signaling pathway [see comments]. Circulation Perivascular delivery of a nitric oxide donor inhibits neointimal hyper- 1997;95:1269-77. plasia in vein grafts implanted in the arterial circulation. J Thorac 39. Clowes AW, Reidy MA, Clowes MM. Mechanisms of stenosis after Cardiovasc Surg 1998;115:604-12. arterial injury. Lab Invest 1983;49:208-15. 25. Fulton GJ, Davies MG, Barber L, Gray JL, Svendsen E, Hagen PO. 40. Clowes AW, Reidy MA, Clowes MM. Kinetics of cellular proliferation Local effects of nitric oxide supplementation and suppression in the after arterial injury. I. Smooth muscle growth in the absence of endo- development of intimal hyperplasia in experimental vein grafts. Eur J thelium. Lab Invest 1983;49:327-33. Vasc Endovasc Surg 1998;15:279-89. 41. Fingerle J, Johnson R, Clowes AW, Majesky MW, Reidy MA. Role of 26. Kaul S, Cercek B, Rengstrom J, Xu XP, Molloy MD, Dimayuga P, et al. platelets in smooth muscle cell proliferation and migration after vascular Polymeric-based perivascular delivery of a nitric oxide donor inhibits injury in rat carotid artery. Proc Natl Acad SciUSA1989;86:8412-6. intimal thickening after balloon denudation arterial injury: role of 42. Davies MG, Hagen PO. Pathobiology of intimal hyperplasia. Br J Surg nuclear factor-kappaB. J Am Coll Cardiol 2000;35:493-501. 1994;81:1254-69. 27. Harnek J, Zoucas E, Sjuve R, Arner A, Ekblad E, Schou H, et al. Local 43. Moller M, Botti H, Batthyany C, Rubbo H, Radi R, Denicola A. Direct infusion of the nitric oxide donor SIN-1 after angioplasty - Effects on measurement of nitric oxide and oxygen partitioning into liposomes intimal hyperplasia in porcine coronary arteries. Acta Radiologica 2003; and low density lipoprotein. J Biol Chem 2005;280:8850-4. 44:395-402. 44. Masters KS, Lipke EA, Rice EE, Liel MS, Myler HA, Zygourakis C, et 28. Yoon J, Wu CJ, Homme J, Tuch RJ, Wolff RG, Topol EJ, et al. Local al. Nitric oxide-generating hydrogels inhibit neointima formation. delivery of nitric oxide from an eluting stent to inhibit neointimal J Biomater Sci Polym Ed 2005;16:659-72. thickening in a porcine coronary injury model. Yonsei Medi J 2002;43: 45. Lipke EA, West JL. Localized delivery of nitric oxide from hydrogels 242-51. inhibits neointima formation in a rat carotid balloon injury model. Acta Biomater 2005;1:597-606. 29. Hou DM, Narciso H, Kamdar K, Zhang P, Barclay B, March KL. 46. Silva GA, Czeisler C, Niece KL, Beniash E, Harrington DA, Kessler JA, Stent-based nitric oxide delivery reducing neointimal proliferation in a et al. Selective differentiation of neural progenitor cells by high-epitope porcine carotid overstretch injury model. Cardiovasc Interventional density nanofibers. Science 2004;303:1352-5. Radiology 2005;28:60-5. 47. Beniash E, Hartgerink JD, Storrie H, Stendahl JC, Stupp SI. Self- 30. Kown MH, Yamaguchi A, Jahncke CL, Miniati D, Murata S, Grunen- assembling peptide amphiphile nanofiber matrices for cell entrapment. felder J, et al. L-arginine polymers inhibit the development of vein graft Acta Biomaterialia 2005;1:387-97. neointimal hyperplasia. J Thorac Cardiovasc Surg 2001;121:971-80. 48. Harrington DA, Cheng EY, Guler MO, Lee LK, Donovan JL, Claussen 31. Keefer LK. Progress toward clinical application of the nitric oxide- RC, et al. Branched peptide-amphiphiles as self-assembling coatings for releasing diazeniumdiolates. Annu Rev Pharmacol Toxicol 2003;43: tissue engineering scaffolds. J Biomed Mater Res Part A 2006;78a:157- 585-607. 67. 32. Rajangam K, Behanna HA, Hui MJ, Han XQ, Hulvat JF, Lomasney 49. Guler MO, Hsu L, Soukasene S, Harrington DA, Hulvat JF, Stupp SI. JW, et al. Heparin binding nanostructures to promote growth of blood Presentation of RGDS epitopes on self-assembled nanofibers of vessels. Nano Lett 2006;6:2086-90. branched peptide amphiphiles. Biomacromolecules 2006;7:1855-63. 33. Behanna HA, Rajangam K, Stupp SI. Modulation of fluorescence 50. Saavedra JE, Mooradian DL, Mowery KA, Schoenfisch MH, Citro ML, through coassembly of molecules in organic nanostructures. J Am Davies KM, et al. Conversion of a polysaccharide to nitric oxide- Chem Soc 2007;129:321-7. releasing form. Dual mechanism anticoagulant activity of diazenium- 34. Hrabie JA, Keefer LK. Chemistry of the nitric oxide-releasing diazeni- diolated heparin. Bioorg Med Chem Lett 2000;10:751-3. umdiolate (“nitrosohydroxylamine”) functional group and its oxygen- substituted derivatives. Chem Rev 2002;102:1135-54. Submitted Sep 3, 2007