Hypercholesterolemia exacerbates in-stent restenosis in rabbits: Studies of the mitigating effect of stent surface modification with a CD47-derived peptide

Abstract

Background and aims:

Hypercholesterolemia (HC) has previously been shown to augment the restenotic response in animal models and humans. However, the mechanistic aspects of in-stent restenosis (ISR) on a hypercholesterolemic background, including potential augmentation of systemic and local inflammation precipitated by HC are not completely understood. CD47 is a transmembrane protein known to abort crucial inflammatory pathways. Our studies have examined the interrelation between HC, inflammation, and ISR and investigated the therapeutic potential of stents coated with a CD47-derived peptide (pepCD47) in the hypercholesterolemic rabbit model.

Methods:

PepCD47 was immobilized on metal foils and stents using polybisphosphonate coordination chemistry and pyridyldithio/thiol conjugation. Cytokine expression in buffy coat-derived cells cultured over the BM and pepCD47-derivatized foils demonstrated an M2/M1 macrophage shift with pepCD47 coating. HC and normocholesterolemic (NC) rabbit cohorts underwent bilateral implantation of BM and pepCD47 stents (HC) or BM stents only (NC) in the iliac location.

Results:

A 40% inhibition of cell attachment to pepCD47-modified compared to bare metal (BM) surfaces was observed. HC increased neointimal growth at 4 weeks post BM stenting. These untoward outcomes were mitigated in hypercholesterolemic rabbits treated with pepCD47-derivatized stents. Compared to NC animals, inflammatory cytokine immunopositivity and macrophage infiltration of peri-strut areas increased in HC animals and were attenuated in HC rabbits treated with pepCD47 stents.

Conclusions:

Augmented inflammatory responses underlie severe ISR morphology in hypercholesterolemic rabbits. Blockage of initial platelet and leukocyte attachment to stent struts through CD47 functionalization of stents mitigates the pro-restenotic effects of hypercholesterolemia.

1. Introduction

Development of ISR in stented atherosclerotic coronary and peripheral arteries is a common complication of stent angioplasty¹. Even in the era of advanced drug-eluting stents (DES), ISR complicates up to 10% of the coronary interventions², with the numbers being even higher for patients with certain comorbidities³. Given the decisive role of dyslipidemia in the progression of atherosclerosis, the significance of hypercholesterolemia for ISR pathogenesis has been scrutinized for a long time, yet the clinical data are controversial^4–6. The most plausible link between the elevated total and the low-density lipoprotein-associated cholesterol (LDL-C) and ISR is increased systemic and vascular inflammation associated with hypercholesterolemia^{7, 8}. Increased uptake of cholesterol by endothelial cells promotes monocyte attachment and transmigration into the subendothelial space via upregulated expression of cell adhesion molecules^{9, 10}. Monocytes differentiate into macrophages and produce multiple cytokines that, by paracrine action, amplify the initial damage to healthy endothelium triggered by hypercholesterolemia¹¹. Furthermore, the compensatory anti-inflammatory mechanisms mediated by nitric oxide (NO) become depressed because of reduced NO production by the dysfunctional endothelium and its decreased bioavailability due to NO sequestration by reactive oxygen species¹². In addition, hypercholesterolemic states induce thrombocythemia and trigger platelets to adhere to activated endothelium, thus priming it for neutrophil and monocyte ingress^{13, 14}.

CD47 is a “marker of self” ubiquitously expressed on the cell surface to protect the host from the attack by its own immune system^{15, 16}. We have previously shown that CD47^{17, 18}, and CD47-derived peptides^{19, 20} bestow protection from platelet and inflammatory cell attachment to the surface of biomaterials. We also demonstrated that the stainless-steel surfaces of stents functionalized with a 22-aminoacid peptide derived from the extracellular Ig domain of rat CD47 are partially protected from platelet and leukocyte association and reduce ISR after carotid artery implantation in a rat model of stent angioplasty²⁰. Since this prior research used healthy normocholesterolemic rats, the results do not entirely reflect the complex relationship between hypercholesterolemia, vascular and systemic inflammation, and the therapeutic effectiveness of CD47-modified stents. The current study aimed to elaborate on the role of inflammation as a link between hypercholesterolemia and the exacerbated response to vascular injury and to investigate the anti-inflammatory and anti-restenotic effects of CD47-derivatized stents in a clinically relevant animal model.

2. Materials and methods

2.1. pepCD47 functionalization of stainless steel.

316-grade stainless steel foils (Goodfellow; Coraopolis, PA, USA), tubes (Microgroup; Medway, MA, USA), and stents (C2D Medical; Ahmedabad, India) were washed in isopropanol and chloroform at 55°C to remove impurities and heated in an air atmosphere at 220°C for 30 min. The specimens were then reacted with 0.5% aqueous solution of polyallylamine bisphosphonate with latent thiol groups (PABT)²¹ at 60°C with shaking for 2 hours, followed by washing in double distilled water (DDW) and exposure to tris (2-carboxyethyl) phosphine hydrochloride (TCEP, 10 mg/ml) in 0.1M acetic buffer at 28°C for 10 min to deprotect thiols. The samples were then washed with degassed DDW and reacted in an argon atmosphere with a degassed 1% aqueous solution of polyethylenimine with installed pyridyldithio groups (PEI-PDT) ²¹ at 37°C for 40 min and thoroughly washed with DDW. A peptide derived from the extracellular Ig domain of rabbit CD47 (Ac-Gly-Asn-Tyr-ThrCys-Glu-Val-Thr-Glu-Leu-Ser-Arg-Glu-Gly-Lys-Thr-Val-Ile-Glu-Leu-Lys-AEEAc-AEEAc-Cys-OH) was custom synthesized by Bachem (Torrance, CA, USA). The CD47-derived peptide (0.5 mg in 1 ml of 50% trichloroacetic acid) bearing free thiols was further diluted with degassed dimethylformamide/PBS mixture and reacted in argon atmosphere with thiol-reactive pyridyldithio (PDT) groups on the surface of metal samples at 37°C with shaking for 30 min. The metal specimens were then washed in sterile PBS and stored at 4°C prior to use.

2.2. Chandler loop experiments.

PepCD47 surface-modified or unmodified stainless steel tube inserts (10 mm length, 6.3 mm diameter) were placed in succession with 1 cm intervals into the 40-cm sections of 1/4" PVC tubing (Terumo Cardiovascular Systems, Ann Arbor, MI). Ten ml of NC and HC rabbit blood anticoagulated with sodium citrate (0.6% final concentration) was perfused through the tubing for 3 hours at a shear rate of 20 dyn/cm². At the end of the perfusion period, the steel inserts were removed, gently washed with PBS, and processed as detailed below.

To visualize the platelets and inflammatory leukocytes attached to the adluminal surface of the inserts, non-fixed specimens were exposed to 10 μM Vybrant ^™ Carboxyfluorescein diacetate succinimidyl ester (CFDA SE) (Thermo Fisher) at 37°C for 15 mins. The samples were then thoroughly washed with PBS and fixed in 4% paraformaldehyde at room temperature for 1 hour. The inserts were then cut longwise, inverted, and imaged using a fluorescence microscope (Nikon TE300). The samples were also analyzed fluorimetrically (ex 485 nm, em 538 nm, cut-off 530 nm) using Spectra Max Gemini EM (Molecular Devices, Sunnyvale, CA).

Additionally, after removing from the Chandler loop apparatus, pepCD47-functionalized and non-modified stainless-steel inserts exposed to HC blood were gently washed with PBS and fixed in 2% glutaraldehyde in sodium cacodylate buffer with 0.1M sodium chloride (Electron Microscopy Sciences; Hatfield, PA) for 24 hours. The inserts were then cut longwise, inverted, rinsed with DDW, dehydrated serially with 70–100% ethanol and hexamethyldisilazane (Sigma-Aldrich; St. Louis, MO, USA), and sputter-coated with gold-palladium. Ten digital micrographs (magnification range 500–10000) at random areas were acquired for analysis using scanning electron microscopy (SEM) (Quanta250, FEI, Hillsboro, OR). Platelet and leukocyte attachment in 4 random fields was determined by manual counting.

2.3. Cell culture experiments assessing macrophage polarization.

Fresh anticoagulated blood harvested from NC and HC rabbits was mixed 1:1 with PBS and carefully layered on top of Ficoll-Paque Plus (Cytiva, Marlborough, MA, USA) in the 50-ml tubes. The tubes were then centrifugated (550 G, 20 min, breaks-off) at room temperature, and the formed “buffy coats” were collected and depleted of platelets by centrifugation at 200 G (10 min × 3). The resulting pellets were resuspended in RPMI-1640 medium and were seeded in the wells of a 12-well plate containing 15 mm × 15 mm stainless steel foil coupons, either non-modified or pepCD47-functionalized (n=4 for each group). The medium was changed after 4 hours for complete RPMI-1640 supplemented with 100 ng/ml of human macrophage colony-stimulating factor (M-CSF; Peprotech, Cranbury, NJ, USA) that was shown to support the growth and differentiation of rabbit macrophages²². The medium was changed every other day. After 6 days of the culture, cells growing on the cell culture plastic (control) and those growing on the different steel substrates were lysed. RNA was isolated using an RNeasy kit (Qiagen, Germantown, MD, USA) and reverse transcribed in cDNA with TaqMan reverse transcription reagents (Applied Biosystems, Waltham, MA, USA). cDNA was amplified in RT-PCR reaction with Sybr Green and rabbit inflammatory cytokine primers (IDT, Coralville, IA, USA), reflecting the M1 and M2 polarization phenotypes.

2.4. In vivo studies.

All animal experiments were pre-approved by the University of Pennsylvania IACUC and conformed to all relevant regulations. Fifteen rabbits of both genders were used in these studies. Nine rabbits were fed an HC diet²³ (a standard rabbit chow supplemented with 1% cholesterol and 3% peanut oil) (Bioserv, Flemington, NJ, USA) for 3 weeks prior to blood harvest or before stent surgery and continued until the sacrifice. Total cholesterol and malondialdehyde (MDA) levels in plasma were measured using a cholesterol fluorimetric assay and lipid peroxidase assay kits from Cayman Chemicals and Eagle Biosciences, respectively. Blood for monocyte isolation (0.75% of body weight) was harvested from the central ear artery. Blood donors were used for surgery at least 3 weeks after blood harvesting. For stent surgery, in isoflurane-anesthetized rabbits of either gender (3–4 kg) vascular access was surgically obtained with arteriotomy in the right carotid artery. Under fluoroscopic guidance, a 4-Fr sheath was advanced over the 0.035" guidewire to the distal aorta. The guidewire was exchanged for a 0.014" guide wire, over which a 3-Fr Fogarty catheter was inserted in the left common iliac artery. A balloon of the Fogarty catheter was inflated and passed across the artery three times to denude the endothelium, after which the Fogarty catheter was exchanged for a 3.0 mm angioplasty catheter loaded with a stent. The stent was deployed in the segment of the iliac artery devoid of the side branches at 12–14 atm to achieve a stent/artery diameter ratio of 1.3. The guidewire was then relocated into the right common iliac artery, and the sequence of Fogarty balloon denudation/stent deployment was repeated. In six NC and three HC rabbits, both stents were BMS, while in the remaining six HC animals, one of the stents was BMS, and the other was a pepCD47 stent. All animals were euthanized 28 days after stent placement. Venous blood was harvested, plasma was separated, and the total cholesterol and malondialdehyde (MDA) levels in plasma were measured using a cholesterol fluorimetric assay and lipid peroxidase assay kits from Cayman Chemicals and Eagle Biosciences, respectively.

2.5. Post-mortem analysis.

Immediately after euthanasia, the animals were consecutively perfused with saline and 10% formalin. The stented iliac segments were excised and post-fixed in 10% formalin for 72 hours. The samples were then extensively washed in the tap water and exposed to the mixture of hydrochloric, hydrofluoric, and nitric acid as described by us before²⁴ to dissolve the struts. “Destented” arteries were then washed in DDW for 30 min and re-fixed in 10% formalin for 3–5 days, prior to routine paraffin embedding and sectioning. The final numbers of processed stented arteries of each type were NC, BMS – 12; HC, BMS – 14; and HC, pepCD47 – 6. The paraffin sections were stained with an elastic stain kit (Sigma-Aldrich, HT-25A) for morphometry according to the Verhoef-van Gieson method. The arterial micrographs were captured as digital images under 20x magnification, and the areas of lumen, neointima, and media were calculated using Image J (v 1.53)-generated tracings of the respective anatomic arterial compartments. The extent of ISR expressed as neointima area, neointima thickness, the percent of cross-sectional narrowing, and the neointima/media ratio was determined for each section using computerized morphometry. The data from individual sections was averaged for each artery, and the latter was used for statistical comparison between groups.

The sections were also immunostained using the primary antibodies to the following rabbit antigens: CD68 (ProSci, 34–191, 1:100), TNFα (MyBioSource, MBS438365, 1:100), IL1β (Bioss, bs-6319R-biotin (1:100), and IL6 (Abbexa, abx131988, 1:200). The goat anti-mouse IgG antibody conjugated to HRP-polymer (Abcam, ab214879) was used as the secondary antibody for all primary antibodies except biotinylated anti-IL1β antibody. The latter was complemented by the Vectastain Elite ABC HRP kit (Vector Lab; PK-6100). In all cases, the color was developed using the eBioscience DAB Advanced chromogenic kit (Invitrogen; 8801–4965-72).

2.6. Statistics.

Data are presented as means ± SD unless specified otherwise. The normality of the data was confirmed with a Kolmogorov-Smirnov test. Differences between the groups were analyzed by ANOVA followed by a post-hoc Tukey's test. A t-test was used to analyze the differences in the experiments that included a single experimental group and a control. A correlation analysis was conducted to investigate the relationship between hypercholesterolemia and the extent of ISR. Statistical significance was assigned at p < 0.05.

3. Results

To study the attachment of bloodborne cells to the native and pepCD47-modified metal substrate, the relative number of cells adhered to the stainless-steel samples was quantified using CFDA SE staining and fluorimetry after 3-hour exposure to NC, and HC rabbit blood recirculated in a Chandler loop apparatus. While no significant differences between the non-modified and pepCD47-modified samples was noted with NC blood (Fig. 1 A, B, and E), the number of cells on the metal surface exposed to HC blood was reduced by 35% as a result of CD47 functionalization (Fig. 1 C, D and E). These results were further confirmed by scanning electron microscopy that demonstrated a 3.2-fold decrease in the number of platelets from HC blood deposited on CD47-modified steel surface (Fig. 1 F-H). The presence of pepCD47 fostered M2 polarization of the macrophages derived from the buffy coats that originated from NC and HC rabbit blood as evidenced by increased M2/M1 cytokine gene expression ratio in macrophages cultured on the bare metal and CD47-functionalized surfaces (Fig 2). The expression of signature M1 cytokines, IL1β and IL12, was suppressed in macrophages growing on the CD47-modified surface compared to macrophages cultured on bare metal steel (Fig. 2 A and B). In contrast, the M2 cytokines, IL10, and TGFβ demonstrated opposite kinetics (Fig. 2 C and D).

Fig. 1. Blood interaction with pepCD47-modified and non-modified stainless steel surface in the Chandler loop flow experiments.

(A-D) Fluorescence microscopy images (FITC filter set; original magnification - 200x) of the adluminal surface of non-modified (A, B) and pepCD47-modified (C, D) stainless steel inserts exposed to NC (A, C) or HC (B, D) rabbit blood recirculating in the Chandler loop apparatus (n=4 per sample type, blood type). E. Fluorimetry measurement of the relative fluorescence emitted by CFDA SE stained cells associated with the adluminal surface. (F-G) Scanning electron microscopy images (original magnification – 500x; insets – 5,000x) of the adluminal surface of non-modified (F) and pepCD47-modified (G) stainless steel inserts exposed to HC rabbit blood recirculating in the Chandler loop apparatus (n=3 per sample type). (H) The surface density of attached platelets.

Fig. 2. Relative expression of M1 (IL1β and IL12) and M2 (IL10 and TGFβ) macrophage polarization markers by macrophages growing on the bare metal and pepCD47-modified surface.

(A-D) Relative quantification (RQ) by 2^−δδct method of IL1β (A), IL12 (B), IL10 (C) and TGFβ (D) expression by rabbit NC and HC blood-derived macrophages grown on the cell culture plastic (control), bare metal surface (BMS) and pepCD47-modified steel (n=3 per condition).

To investigate the impact of hypercholesterolemia on ISR development and the antirestenotic effectiveness of CD47 stent functionalization, BMS and stents modified with pepCD47 were implanted in rabbits fed normal chow and HC diet. Exposure to a hypercholesterolemic diet (4 weeks before the intervention and 4 weeks after stent deployment) considerably increased neointimal formation in the stented iliac arteries (Fig. 3 A and B). Neointimal area (Fig. 3 D), neointimal thickness (Fig. 3 E), percent of luminal stenosis (Fig. 3 F), and neointima-to-media ratio (Fig. 3 G) increased 2.3–5.2 fold in the hypercholesterolemic animals. These hyperplastic responses to stent implantation were reversed 29–52% (p<0.01 for all comparisons) in the arteries treated with the pepCD47-derivatized stents (Fig. 3 A-G).

Fig. 3. Effects of hypercholesterolemia and stent surface functionalization with pepCD47 on the extent of ISR.

(A-C) Representative microscopy images of stented iliac arteries of NC (A) and HC (B, C) rabbits treated with bare metal (A, B) and pepCD47 stents (original magnification is 20x). Stent struts were removed by acid treatment prior to paraffin embedding²⁴, sectioning, and Verhoef-van Gieson staining. N – neointima, M – media, A – adventitia, L – lumen, S – strut positions. (D-G) Morphometric values of neointimal area (D), neointima thickness above the struts (E), percent of luminal stenosis (F), and neointima to media area ratio in the treatment groups.

In the individual animals treated with BMS, the severity of ISR expressed as neointima to media ratio strongly correlated (Pearson coefficient of 0.826; p<0.0001) with total cholesterol levels in rabbit blood (Supplementary Fig. 1). BMS implantation in hypercholesterolemic animals was associated with increased plasma concentrations of inflammatory markers, TNFα (Supplementary Fig. 2 A) and IL1β ( Supplementary Fig. 2 B), and lipid peroxidation marker, malondialdehyde (MDA; Supplementary Fig. 2 C) compared to normocholesterolemic rabbits. Likewise, local arterial expression of inflammatory markers, TNFα (Fig. 4 A, B, and D), IL1β (Fig. 4 E, F, and H), and IL6 (Fig. 4 I, J, and L) was significantly upregulated in the peri-strut regions of hypercholesterolemic rabbits compared to normocholesterolemic counterparts. Furthermore, inflammatory macrophage infiltrates were significantly more predominant in the peri-strut regions of hypercholesterolemic rabbits as compared to normocholesterolemic animals (Fig. 4 M, N, and P). Local expression of inflammatory markers and the presence of macrophages around the struts was significantly reduced in the arteries of hypercholesterolemic animals treated with pepCD47-functionalized stents (compare Fig. 4 B and C, F and H, J and K, N and O). Macrophages accumulating in the peri-strut location appear to present the principal source of pro-inflammatory cytokines, as apparent from dual immunofluorescence staining of macrophages and IL1β-producing cells (Supplementary Fig. 3). Activation of quiescent cells of the vascular wall by the macrophage-released cytokines drives and sustains cell proliferation, migration, and ECM production, thus contributing to the neointimal formation and ISR²⁵. In BMS-treated hypercholesterolemic animals, proliferation in the neointimal, medial, and adventitial compartments increased compared to normocholesterolemic counterparts (Fig. 5 A, B, and D) and was significantly reduced in the arteries treated with pepCD47-modified stents (Fig. 5 B, C and D).

Fig. 4. Immunohistochemical assessment of the inflammatory cytokine expression in the peri-strut areas of NC and HC rabbits implanted bare metal and pepCD47-modified stents.

(A-P) Representative microscopic images (100x magnification) of iliac arteries stented with BMS (A, B, E, F, I, J, M, N) and pepCD47-functionalized stents (C, G, K, O) and immunostained for TNFα (A-C), IL1β (E-G), IL6 (I-K) and CD68-positive macrophages (M-O). Quantification of TNFα (D), IL1β (H), and IL6 (L) expression was based on a semi-quantitative (0–4) scale. Macrophage infiltration around the struts (P) was calculated as a percentage of a strut circumference occupied with the CD68-positive cells (n=4 for each marker/treatment group combination).

Fig. 5. Cell proliferation in the stented arteries.

(A-C) Ki67 immunohistochemistry was performed on the sections from the stented iliac arteries of NC (A) and HC (B, C) rabbits treated with BMS (A, B) and pepCD47-modified stents (C). Ki67 labeling was expressed as the number of Ki67-positive cells per mm² of neointimal, medial, and adventitial compartments (D).

n – neointima, m – media, a – adventitia, l – lumen, s – strut positions.

4. Discussion

We have previously demonstrated the anti-thrombotic and anti-inflammatory properties of immobilized CD47^{17–20, 26}. These previous studies were largely performed using in vitro and ex vivo model systems. With respect to using the immobilized CD47 strategy to address the biocompatibility issues of clinically relevant devices, we have demonstrated that CD47-modified stents significantly reduced restenosis markers in a rat carotid artery stenting model²⁰. These current studies have begun to expand our previous findings into a large animal preclinical model system. Specifically, we have confirmed the pathogenic link between diet-induced hypercholesterolemia and the augmented inflammatory response to vascular intervention that leads to heightened neointimal expansion and ISR in the rabbit model. In addition, we have demonstrated the therapeutic anti-restenotic effectiveness of stent surface modification with a CD47-derived peptide that curtails the hypercholesterolemia-driven inflammatory reactions triggered by the stent implantation.

4.1. Exacerbation of ISR by hypercholesterolemia is mediated by hypercholesterolemia-induced inflammation.

Augmented neointimal response after balloon angioplasty, endarterectomy, and stent implantation has been demonstrated in hypercholesterolemic subjects across multiple mammalian species and types of vascular injury^27–29. Systemic and local arterial inflammation preceding vascular intervention was identified as a pathogenetic link between hypercholesterolemia and restenosis^30–32. The pro-inflammatory effects of hyperlipidemia are multifaceted. Liver and adipose tissue damage by the excessive dietetic cholesterol triggers the production and secretion of pro-inflammatory cytokines, IL6, IL12, CRP, and TNFα, as well as of chemokines CCL2 and fractalkine (CX3CL1)⁸. In parallel, hypocholesterolemia-induced platelet activation leads to the formation of platelet/leukocyte aggregates that bind avidly to the balloon- or stent-injured arterial wall bridging between the systemic and local vascular inflammation⁸. Inflammation at the site of vascular intervention is further augmented by persisting hypercholesterolemia through the enhanced monocyte accumulation, conversion of macrophages into the foam cell, followed by their death, failed efferocytosis, and activation of NLRP3 inflammasome by the accumulated cholesterol crystals³³. In accordance with the previous reports^{23, 34, 35}, we observed increased concentrations of pro-inflammatory cytokines and MDA in the plasma of rabbits fed the HC diet for 8 weeks compared to NC littermates (Supplementary Fig. 2). Likewise, we confirmed the previously demonstrated exacerbated neointimal growth after stenting in the hypercholesterolemic animals (Fig. 3), and reported a direct correlation between the total cholesterol levels and ensuing ISR (Supplementary Fig. 1) that was not demonstrated previously. Plastic embedding techniques typically used for the histological preparation of stented arteries are not routinely compatible with immunohistochemical methods. Therefore, very little data about arterial tissue production and secretion of pro-inflammatory mediators are available. In situ acid dissolution of stent struts within the explanted arteries²⁴ enabled paraffin embedding of “destented” vasculature and subsequent immunohistochemical analysis. Our results (Fig. 4) revealed the peri-strut expression of TNFα, IL1β, and IL6, augmented by persistent hypercholesterolemia (Fig. 4 A, B, and D; E, F, and H, and I, J, and L, respectively). Likewise, peri-strut macrophage infiltrates detected with the anti-CD68 antibody were significantly more intense in the arteries of the HC compared to the NC animals (Fig. 4 M, N, and P). We have also validated macrophages as the principal source of pro-inflammatory cytokine expression in the stented arteries (Supplementary Fig. 3). Since the paracrine effect of cytokines secreted by inflammatory leukocytes was firmly established as the main trigger of smooth muscle cells proliferation in the iatrogenically manipulated arteries²⁵, we predictably observed increased proliferative activity in all layers of the stented arteries from the HC rabbits compared to their NC counterparts (Fig. 5).

4.2. CD47 functionalization of stents inhibits ISR by blocking early steps of inflammatory pathways.

The physiological responses elicited by CD47/SIRPα interaction have become a major topic of interest in cancer research³⁶ and transplant biology³⁷. Recently, the importance of defective efferocytosis induced by CD47 signaling has been realized as a crucial mechanism of vascular inflammation³⁸ and atherosclerotic plaque growth and destabilization^39–41. Furthermore, a blockage of CD47 signaling either with anti-CD47 antibodies⁴⁰ or a low-molecular-weight inhibitor of the CD47/SIRPs pathway⁴¹ was shown to be protective against atherosclerosis in the mouse models. Unresolved vascular inflammation is central to the pathogenesis of ISR⁴². To this end, inhibition of CD47/SIRPα signaling was demonstrated to counteract neointimal expansion in wire-injured mouse femoral arteries⁴³. Although our previous study²⁰ and the present experiments demonstrate the anti-restenotic efficacy of CD47-functionalized stents, our results do not contradict the reports that established the therapeutic effect of CD47 inhibition in comparable animal models of vascular disease. The controversy is explained by the immobilization of the SIRPα-interacting moiety (CD47 peptide⁴⁴) on the stent surface rather than being displayed on the cell membrane. In this presentation, CD47 aborts attachment and spreading of platelets²⁶ and cells of myeloid origin²⁶, expressing SIRPα. In confirmation of these data, our present studies demonstrated a reduced number of platelets and leukocytes from hypercholesterolemic rabbit blood attached to the pepCD47-modified vs. non-modified stainless steel substrate (Fig. 1). CD47 binds thrombospondin-1⁴⁵, which was implicated in the development of restenosis^{46, 47}. Thus, in addition to its direct anti-inflammatory effects, pepCD47 immobilized on the stent surface may hypothetically sequester thrombospondin-1, limiting its impact on ISR development. Recent studies have highlighted the importance of M1/M2 macrophage equilibrium in the progression of restenosis⁴⁸. Our findings contribute to this growing body of evidence by showing that pepCD47-functionalized surfaces foster the M2 polarization of attached macrophages (Fig. 2). This observation is in accordance with biased M1 polarization in mouse tumors upon treatment with anti-CD47 antibodies⁴⁹. The underlying mechanism responsible for the CD47-induced M2 polarization has been demonstrated to be mediated by the downstream SIRPα signaling through SHP-1 and Akt2⁵⁰.

The ISR severity in hypercholesterolemic rabbits in our study was reduced by 29–52% in the animals treated with pepCD47 stents (Fig. 3 and Fig. 6), partially offsetting the pro-restenotic consequences of hypercholesterolemia. In agreement with the hypothesis of inflammatory response mitigation by stent-immobilized pepCD47, the macrophage infiltration of peri-strut areas and local arterial expression of pro-inflammatory cytokines in HC animals was reduced to the values observed after stent deployment in the NC rabbit model. The counteraction to the damaging impact of hypercholesterolemia on cytokine expression in the stented vasculature was paralleled by the decrease of the proliferative activity in all layers of the stented arteries (Fig. 5).

Fig. 6. Graphical abstract.

Functionalizing the metal stent surface with CD47-derivatized peptide reduces in-stent restenosis in the rabbit iliac arteries by mitigating platelet and leukocyte deposition on the stent struts.

4.3. Limitations of the study.

Although the data presented herein shows that pepCD47 can significantly reduce the pathophysiology of stent implantation in a hypercholesterolemic rabbit, there are several elements of our investigations that need to be pursued in future studies. First, our present rabbit experiments did not address the immediate and early responses in the stented arteries of NC and HC animals, did not examine the therapeutic value of pepCD47 stents in the NC rabbit cohort, and did not validate the long-term persistence of the anti-restenotic effects achieved with pepCD47 stent functionalization in HC rabbits. Additionally, we used a single injury model that entails stent implantation in the arteries without pre-existing restenotic lesions. Finally, given recent reports linking CD47 expression with the modulating effects on phagocytotic behavior and efferocytosis of differently polarized macrophage populations as a contributing factor to restenosis^{43, 48}, future studies will need to be pursued in this important mechanistic area.

4.4. Conclusions.

In conclusion, the present studies support the view that through systemic and local mechanisms, hypercholesterolemia primes the arterial wall to a more vigorous inflammatory response to stenting-triggered injury and, consequently, a more robust neointimal formation than observed in normocholesterolemic controls. This untoward outcome is effectively mitigated by pepCD47 immobilized on the surface of stent struts that results in CD47-SIRPα signaling, inhibiting both platelet and leukocyte binding to the stent surface, and subsequent pro-inflammatory and thrombotic activity.

• CD47-derived peptide (pepCD47) immobilized on a metal substrate decreases cell attachment to the surface

• Immobilized pepCD47 promotes M2 polarization of macrophages

• Hypercholesterolemia increases the severity of ISR in rabbits

• PepCD47 stents reduce ISR severity in hypercholesterolemic rabbits

• PepCD47-mediated reduction of arterial inflammation mitigates ISR