FPR2 Agonism Attenuates Restenosis by Mitigating Neointimal Hyperplasia via ELOVL6

ABSTRACT

Restenosis following endovascular intervention in lower extremity arterial disease contributes to significant morbidity and mortality. This study investigates the role of formylpeptide receptor 2 (FPR2) in neointimal hyperplasia and evaluates the therapeutic potential of the selective FPR2 agonist BMS‐986235 in mitigating restenosis. FPR2 expression was significantly reduced in the popliteal and anterior tibial arteries of male amputees with restenosis compared to healthy controls. Whole‐body and myeloid‐specific FPR2 knockout mice consistently displayed exaggerated neointimal hyperplasia, accompanied by a marked reduction in vessel lumen diameter, following endothelial injury. Treatment with BMS‐986235 effectively slowed the progression of restenosis. Mechanistically, FPR2 activation maintained the differentiated state of vascular smooth muscle cells (VSMCs) and limited excessive M2 macrophages accumulation, thereby limiting neointimal remodeling. Transcriptomic analysis additionally identified ELOVL fatty acid elongase 6 (ELOVL6) as a novel downstream target of FPR2 activation, which was upregulated in restenosis models. Notably, BMS‐986235 reduced ELOVL6 expression in both macrophages and VSMCs, inhibiting VSMC proliferation and mitigating neointimal hyperplasia. FPR2 activation mitigates restenosis progression by preserving VSMC differentiation through the FPR2/ELOVL6 axis, highlighting its potential as a novel therapeutic target for prevention of restenosis.

Restenosis after revascularization in lower extremity arterial disease remains a major unmet need. We demonstrate reduced formylpeptide receptor 2 (FPR2) expression in human restenotic arteries and demonstrate that FPR2 deficiency exacerbates neointimal hyperplasia. Mechanistically, FPR2 activation suppressed vascular smooth muscle cell (VSMC) de‐differentiation and limited M2 macrophage accumulation. Transcriptomics profiling identified ELOVL6 as a pivotal downstream mediator, with its inhibition preventing VSMC de‐differentiation. The FPR2/ELOVL6 axis emerges as a promising therapeutic strategy for restenosis.

Abbreviations

Ad

adenovirus

ALT

alanine aminotransferase

ARG‐1

arginase‐1

AST

aspartate aminotransferase

CM

conditioned media

CNN1

calponin 1

DAPI

4′,6‐diamidino‐2‐phenylindole

EdU

5‐ethynyl‐2′‐deoxyuridine

ELOVL6

ELOVL fatty acid elongase 6

FPR2

formylpeptide receptor 2

GPCR

G‐protein coupled receptor

HASMC

human aortic smooth muscle cell

HDL‐C

high‐density lipoprotein cholesterol

H&E

hematoxylin and eosin

IL

interleukin

LDL‐C

low‐density lipoprotein cholesterol

LEAD

lower extremity arterial disease

LSCI

laser speckle contrast imaging

NC

normal control

PCNA

proliferating cell nuclear antigen

PDGF‐BB

platelet‐derived growth factor‐BB

siRNA

short interfering RNA

TAGLN

transgelin

TC

total cholesterol

TG

thyroglobulin

VSMC

vascular smooth muscle cell

α‐SMA

alpha smooth muscle actin

1. Introduction

Restenosis, resulting from neointimal hyperplasia that involves uncontrolled proliferation and migration of vascular smooth muscle cells (VSMCs), poses a significant challenge to the sustained efficacy of endovascular interventions such as angioplasty [1, 2]. Occurring in up to 60% of patients treated for lower extremity arterial disease (LEAD) within 1 year of the procedure, restenosis severely limits both short‐ and long‐term outcomes [3]. Despite considerable efforts to improve treatment outcomes, including the use of drug‐eluting and drug‐coated balloons and stents, the incidence of restenosis remains high [1]. Moreover, drugs such as paclitaxel and sirolimus, which are currently employed to mitigate restenosis, are not only costly but also exhibit suboptimal efficacy [4, 5]. Therefore, a clear, unmet clinical need exists to develop novel preventive strategies for restenosis.

Endovascular procedures induce mechanical injury, triggering substantial inflammation that stimulates VSMC proliferation and extracellular matrix deposition, ultimately leading to neointimal thickening and restenosis [6, 7, 8]. Under physiological conditions, VSMCs are highly differentiated cells in the tunica media, which is crucial for maintaining normal structure and blood vessel function. However, after endovascular procedures, VSMCs are prompted to de‐differentiate, characterized by excessive proliferation and migration [6, 7, 8]. This results in significant migration from the tunica media to the tunica intima, acting as the primary driver of restenosis [6, 7, 8]. Type 2 macrophages (M2 macrophages), also known as alternatively activated macrophages, are characterized by their production of anti‐inflammatory cytokines, such as IL‐10 and TGF‐β [9, 10, 11]. Functionally, they exhibit strong phagocytic capacity, are involved in the clearance of apoptotic cells and debris, facilitate tissue repair and wound healing, and possess both pro‐angiogenic and pro‐fibrotic properties [9]. Following the inflammatory cascade initiated by endovascular procedures, M2 macrophages infiltrate the vascular wall and secrete cytokines, chemokines, and growth factors that exacerbate VSMC proliferation and migration [12, 13, 14].

Restenosis is often associated with chronic, unresolved inflammation; thus, promoting the resolution of inflammation represents an innovative approach to minimizing restenosis. While macrophages contribute to hyperplasia, they also produce specialized pro‐resolving lipid mediators [15]. Moreover, administering pro‐resolving lipid mediators, such as resolvin E1 and aspirin‐triggered 15‐epi‐lipoxin A, limits hyperplasia by regulating M2 macrophages in various vascular injury models [16, 17]. Formylpeptide receptor 2 (FPR2), a G protein‐coupled receptor (GPCR) highly expressed in macrophages [18, 19], has been identified as the “master switch” for resolving inflammation [20, 21, 22]. Other functions of FPR2 have also been reported, including inhibiting adipocyte thermogenesis and regulation of aging, underscoring its broader role in inflammation, metabolism, and tissue homeostasis [23, 24]. This foundational understanding of FPR2's biology has inspired the development of FPR2 agonists. The small‐molecule FPR2‐agonist BMS‐986235 is more metabolically stable than endogenous specialized pro‐resolving lipid mediators and other FPR2‐agonists, making it an attractive candidate for the treatment of disorders involving inflammation and tissue remodeling, such as myocardial infarction [18, 25]. However, the therapeutic potential of BMS‐986235 in vascular disorders, e.g., restenosis, has not been investigated. Based on its role in regulating inflammation and tissue repair, we hypothesized that FPR2 agonism could mitigate VSMC de‐differentiation and thereby limit neointimal hyperplasia. In this study, we aim to investigate the therapeutic potential of FPR2 agonism, specifically through BMS‐986235, to modulate VSMC de‐differentiation and reduce neointimal hyperplasia, offering a promising therapeutic approach for restenosis.

2. Materials and Methods

All human and animal studies have received approval from the relevant ethics committee and have been conducted in accordance with the ethical standards defined in the 1964 Declaration of Helsinki and its subsequent amendments. The study aimed to create groups of equal size through randomization and blinded analysis. The experimental approach is summarized in Figure S1.

2.1. Patients and Sample Collection

Arteries were obtained from amputated lower limbs of patients who met the following inclusion criteria: (i) a history of lower limb atherosclerosis; (ii) previous endovascular treatment; (iii) restenosis at the site of endovascular surgery, accompanied by blockage of the vascular lumen. The exclusion criteria were: (i) absence of restenosis at the site of endovascular surgery during amputation; (ii) a thrombus as the main component of the vascular obstruction; (iii) no history of endovascular surgery; (iv) the presence of apparent infection or tissue erosion at the site of restenosis. Vascular segments were selected based on surgical history. Normal vessels were obtained from individuals with traumatic amputations who had not previously received an endovascular intervention, anatomically as close as possible to the vascular segments that displayed restenosis following percutaneous transluminal angioplasty. All patients were male (n = 10 for restenosis, n = 10 for normal; for detailed information, see Table S1) at Qilu Hospital of Shandong University and Beijing Jishuitan Hospital. Procedures during tissue collection strictly adhered to relevant ethical regulations for clinical specimen research, with protocols approved by the Ethics Committee of Qilu Hospital of Shandong University (Approval No. KYLL‐2021 (KS)‐769) and Beijing Jishuitan Hospital (Approval K2023‐021). Written informed consent was obtained from all participants after explaining the purpose, procedure, and possible consequences of this study. Vascular segments were isolated and washed in sterile normal saline. Some of these tissues were snap‐frozen in liquid nitrogen and stored at −80°C, while others were fixed in 4% paraformaldehyde (4% PFA, BL539A, Biosharp, Anhui, China) and subsequently paraffin‐embedded. Hematoxylin and eosin staining (H&E, G1121, Solarbio Science&Technology, Beijing, China) was employed to confirm the presence of restenosis prior to immunofluorescence, immunohistochemical staining, and western blot analyses.

2.2. Animal Model of Induced Restenosis

All animals were bred and maintained in pathogen‐free conditions. Mice were randomly assigned to different groups and kept on a Clinton‐Cybulsky high‐fat diet (40% fat, 1.25% cholesterol, TP28521, Trophic Animal Feed High‐Tech Co. Ltd., China).

Fpr2 knockout (Fpr2^−/−) mice with a C57BL/6J background were generated (Cyagen Bioscience Inc., S‐KO‐02106, Guangzhou, China) as described previously [26]. Exon 2 was selected as the target site, with the ATG start codon in exon 2 and the TGA stop codon in exon 2 (Transcript Fpr2‐201: ENSMUST00000064068).

Myeloid‐specific Fpr2 knockout (Fpr2^MKO) mice were generated by crossing Fpr2^flox/flox mice with LysM‐Cre mice [27] (Cyagen Bioscience Inc., S‐CKO‐02479, Guangzhou, China). Exon 2 was selected as the conditional knockout region (cKO region), which contains a 1056 bp coding sequence. Deleting this region is expected to result in the loss of function of the mouse Fpr2 gene.

Twelve‐week‐old male mice (Fpr2^−/−, WT, Fpr2^MKO, Fpr2^flox/flox) were subjected to transluminal mechanical injury of the left femoral artery by introducing a guide wire [28]. The sample size was determined using a power calculation [29]. Only male animals were used in experiments to eliminate potential confounding factors from the estrogen cycle on vascular remodeling. The animal surgeon was blinded regarding the identity of gene type or treatment. In brief, mice were anesthetized with sodium pentobarbital (50 mg/kg, i.p.; Ganes Chemicals, Pennsville, NJ, USA). The left femoral artery was exposed by blunt dissection and was sutured proximal and distal with 6–0 silk loops to control blood flow during the operation temporarily. A wire, 0.38 mm in diameter (C‐SF‐15‐15, Cook, Bloomington, IN), was inserted into the femoral artery towards the iliac artery and left in place for 1 min to denude the intima and dilate the artery. Blood flow in the femoral artery was then restored. Animals were euthanized by exsanguination through cardiac puncture under anesthesia (sodium pentobarbital, 50 mg/kg, i.p.) 28 days after wire injury. Mouse femoral arteries were isolated and dissected for further ex vivo studies. The animal fate flow chart summarizes animal use and analysis in preclinical studies of the Fpr2^−/− and Fpr2^MKO mice (Figure S2).

All experimental procedures were approved by the Animal Ethics Committee of Shandong University (Approval No. 21010). All animal studies were conducted according to the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals and Animal Research and followed Reporting of In Vivo Experiments (ARRIVE) guidelines [30].

2.3. Verification of Mouse Genotype

Fpr2^−/− mice on a C57BL/6 background were generated as described previously [26]. Approximately 0.3 cm of the tail was cut off from each mouse, and the specimen obtained was digested in a water bath at 55°C using proteinase K for 20 min. DNA was extracted using the Mouse Genotyping Kit (PD101, Vazyme, Nanjing, China). The genomic DNA of each mouse was amplified using gene primers specific for Fpr2 to determine if the gene was knocked out. PCR products were separated by 2% agarose gel electrophoresis. The Fpr2 knockout gene has a product size of 756 bp, while the normal wild‐type gene produces a 623 bp product. If only a single band of 756 bp is present, the mouse genotype is considered Fpr2^−/−, indicating Fpr2 gene knockout mice. If only one band of 623 bp is present, the mouse genotype is Fpr2^+/+, indicating wild‐type (WT) mice. Fpr2 deletion was confirmed by PCR (Figure S4A).

To identify Fpr2 gene containing loxP sites, Fpr2^MKO mice were developed by crossing Fpr2^flox/flox mice with LysM‐Cre mice [27]. The procedure used to extract DNA was consistent with the previously described method. The product size for the Fpr2 gene containing the loxP site was 202 bp, while the product size for the normal Fpr2 gene was 143 bp. Mouse tail genomic DNA was also amplified to confirm the presence of the Cre recombinase gene (LysM‐Cre), resulting in an amplification product size of 700 bp. Fpr2^flox/flox mice exhibit only a single band (202 bp), whereas LysM‐Cre^+/− (Fpr2^MKO) mice, which conditionally knockout the Fpr2 gene, show both a 143 bp band and a 700 bp band (genotype confirmed by PCR, Figure S4B).

2.4. Blood Perfusion Monitored With Laser Speckle Contrast Imaging (LSCI)

Mice were anesthetized (sodium pentobarbital, 50 mg/kg, i.p.) and placed on a heating blanket at 37°C. The LSCI system [31] (Perimed Instruments AB, Sweden) was used on day 28 after the operation and prior to euthanasia. The blood perfusion ratio = injured side (left hind limb)/normal side (right hind limb) × 100%.

2.5. Receptor Agonists and Dosing Regimen

BMS‐986235 (1‐((3S,4R)‐4‐(2,6‐difluoro‐4‐methoxyphenyl)‐2‐oxopyrrolidin‐3‐yl)‐3‐phenylurea) was synthesized by WuXi AppTec Inc. (Shanghai, China) as previously described [32]. Twelve‐week‐old C57BL/6 mice were randomly assigned to one of three groups: sham (surgery without wire injury, receiving an equivalent volume of vehicle (0.8% Tween‐80 in saline i.p.)), restenosis (wire injury, receiving an equivalent volume of vehicle), or restenosis (wire injury) plus BMS‐986235 (3 mg/kg, i.p.) [32, 33]. All interventions were administered daily for 28 days, from the surgery day until the experimental endpoint. The experimental approach for animal use and analysis to evaluate the effect of BMS‐986235 can be found in Figure S3. The femoral arteries of these animals were collected at the end of the study for histological, quantitative real‐time polymerase chain reaction (qRT‐PCR) analyses, and transcriptomic profiling.

2.6. Analysis of Plasma Lipid Levels

As described in the manufacturer's specifications [34], mouse plasma was obtained by centrifugation (3000 rpm for 20 min) from whole blood containing EDTA. Total plasma cholesterol (TC, P/N: 105–000448‐00), triglycerides (TG, P/N: 105–000449‐00), high‐density lipoprotein cholesterol (HDL‐C, P/N: 105–000463‐00), and low‐density lipoprotein cholesterol (LDL‐C, P/N: 105–000464‐00) levels were assayed using kits from Mindary (Shenzhen, China).

2.7. Tissue Harvesting and Histology

Mouse femoral artery and patients' artery samples were fixed in 4% PFA for 24 h and then paraffin‐embedded. The 5 μm sections were initially stained with H&E to assess general appearance. Collagen Fiber and Elastic Fiber Staining (EVG staining, G1597, Solarbio Science & Technology, Beijing, China) stains both elastic fibers and collagen fibers in sections, making it suitable for observing these fibers within tissues. Masson's trichrome (G136, Solarbio Science & Technology, Beijing, China) was used to differentiate collagen fibers in the vessels, and images were captured using an optical microscope (Olympus, Tokyo, Japan), as previously described.

The quantification of neointima formation was performed in a blinded manner by experienced investigators. The areas of neointimal hyperplasia and the arterial lumen in the cross‐section of H&E‐stained artery segments were automatically measured in square pixels (pixel²) using a computerized image analysis system (Image‐Pro Plus 6.0 software). Arterial narrowness was defined as the ratio of the areas of neointimal hyperplasia to the artery lumen. In contrast, the percentage area of smooth muscle fiber was defined as the ratio of smooth muscle fiber in neointimal hyperplasia to the total area of neointimal hyperplasia, as previously described [35].

2.8. Immunofluorescent Staining

Detailed procedures for immunofluorescent staining have been described previously [36]. Expression levels and locations identified through double‐immunofluorescence staining include proliferating cell nuclear antigen (PCNA) and alpha‐smooth muscle actin (α‐SMA), ELOVL6 and α‐SMA, F4/80 and α‐SMA, and ELOVL6 and arginase‐1 (ARG‐1) in mouse femoral arteries subjected to wire injury or sham operations, as well as CD163 and α‐SMA in human vessel samples. The antibodies used were mouse anti‐α‐SMA (1:200, 67 735–1, Proteintech, Wuhan, China), rabbit anti‐PCNA (1:200, 13 110, Cell Signaling, Boston, USA), rabbit anti‐ELOVL6 (1:200, PAB13032, Abnove, Taiwan, China), mouse anti‐ARG‐1 (1:200, 66 129–1, Proteintech, Wuhan, China), rabbit anti‐CD163 (1:100, ab182422, Abcam, Cambridge, UK), rabbit anti‐iNOS (1:200, 13120, Cell Signaling, Boston, USA) and rabbit anti‐F4/80 (1:200, 30 325, Cell Signaling, Boston, USA). Levels of ARG‐1 were detected using rabbit anti‐ARG‐1 (1:200, 93 668, Cell Signaling, Boston, USA) in murine restenosed vessels. Tissue sections were incubated with the respective antibodies overnight at 4°C. Alexa Fluor 594 AffiniPure Goat Anti‐Mouse IgG (H + L) (1:500, JAC‐115‐585‐003, Jackson ImmunoResearch, PA, USA) was used for binding with primary antibodies derived from mice. Alexa Fluor 594 AffiniPure Goat Anti‐Rabbit IgG (H + L) (1:500, JAC‐115‐545‐003, Jackson ImmunoResearch, PA, USA) and Alexa Fluor 488 AffiniPure Goat Anti‐Rabbit IgG (H + L) (1:500, JAC‐111‐585‐003, Jackson ImmunoResearch, PA, USA) were used for binding with primary antibodies from rabbits. 4′,6‐Diamidino‐2‐phenylindole (DAPI, ab104139, Abcam, Cambridge, UK) was utilized to visualize the nuclei. Specimens were examined under a fluorescence microscope (Olympus, Tokyo, Japan).

2.9. Immunohistochemical Analysis

Immunohistochemical analysis was performed on 5 μm thick sections. After antigen retrieval using citrate buffer, the sections were incubated overnight at 4°C with rabbit anti‐FPR2 antibodies (1:200, ab203129, Abcam, Cambridge, UK), ELOVL6 (1:200, PAB13032, Abnove, Taiwan, China), mouse anti‐α‐SMA (1:200, 67 735–1, Proteintech, Wuhan, China), and rabbit anti‐CD80 (1:200, 15 416, Cell Signaling, Boston, USA). Following a rinse with PBS, the sections were incubated with the SABC‐POD (Rabbit/Mouse IgG) kit (SA1027/SA1028, BOSTER, Wuhan, China) for 30 min at 37°C. The sections were counterstained with hematoxylin using a DAB detection system (ZLI‐9017, Zhongshan, China), as previously described [15].

2.10. Image Analysis for Double Immunofluorescent Staining and Immunohistochemical Staining

ImageJ (National Institutes of Health, USA) was used to analyze immunohistochemical staining, immunofluorescent staining, and double immunofluorescent staining. Images were thresholded for each channel, and the positive cell count, fluorescence intensity, and integrated density were measured. The percentage of the area that was positive was calculated as (positive staining area/total area) × 100 [8].

2.11. RNA‐Seq Analysis and Differential Gene Identification

2.11.1. RNA extraction, RNA‐Seq, and data analysis

Relevant RNA from all samples was isolated and purified according to the manufacturer's protocol, utilizing TRIzol (Thermo Fisher, CA, USA) [37]. The amount and purity of total RNA were verified using the NanoDrop ND‐1000 (NanoDrop, Wilmington, DE, USA), and the integrity of the RNA was assessed with the Bioanalyzer 2100 (Agilent, CA, USA). The concentration, RIN value, and total RNA were maintained at > 50 ng/μL, > 7.0, and > 1 μg, respectively, to meet the requirements for subsequent experiments. Oligo (dT) beads (Dynabeads Oligo (dT), Thermo Fisher, CA, USA) captured PolyA‐containing mRNA through two rounds of purification. The captured mRNA was disrupted at high temperatures using the NEBNextR Magnesium RNA Fragmentation Module for fragmentation (94°C, 5–7 min). Fragmented RNA was synthesized by reverse transcriptase (Invitrogen SuperScript II Reverse Transcriptase, CA, USA). E. coli DNA polymerase I (NEB, MA, USA) and RNase H (NEB, MA, USA) facilitated double‐stranded synthesis to convert these DNA and RNA composite duplexes into DNA duplexes. dUTP Solution (Thermo Fisher, CA, USA) complements the ends of double‐stranded DNA to form blunt ends. An “A” base is added at each end to facilitate connection with a “T” base on the opposite end, and the fragment size is screened and purified using magnetic beads. The second strand was digested by UDGase (NEB, MA, USA) and subsequently underwent PCR (predenaturation at 95°C for 3 min; denaturation at 98°C for 15 s, annealing at 60°C for 15 s, extension at 72°C for 30 s for a total of 8 cycles; and final extension at 72°C for 5 min) to create a strand‐specific library with fragment sizes of 300 ± 50 bp. Finally, we used Illumina NovaSeq 6000 (LC Bio‐Technology CO. Ltd. Hangzhou, China) to sequence it bilaterally according to standard practice in PE150.

2.11.2. Data processing of differential expression genes (DEGs)

Genes that satisfied the inclusion criteria of p < 0.05 and |logFC| ≥ 2 were identified as DEGs and included in the study. Bioinformatics and Evolutionary Genomics, an online website tool, was used to identify common DEGs and to plot Venn diagrams (http://bioinformatics.psb.ugent.be/webtools/Venn/). Genes affected by disease were collected by comparing the restenosis group and the sham group. Meanwhile, genes affected by treatment were obtained by comparing the restenosis group (RS) and the restenosis + BMS‐986235 group (RS + BMS‐986235). Each circle represented a dataset (genes changed in disease and genes altered by treatment), and the overlap between the circles amounted to the overlap between the datasets. The heatmap was plotted by ggplot2 (version 3.3.3) [38].

2.12. Cell Culture

Primary human aortic smooth muscle cells (HASMCs) were purchased from ScienCell (6110, San Diego, California, USA) and cultured according to the manufacturer's instructions. Human monocytic leukemia (THP‐1) cells were purchased from the Chinese Academy of Sciences (SCSP‐567, Shanghai, China). The THP‐1 cell line was maintained in RPMI 1640 media (11 875 093, Gibco, California, USA) supplemented with 10% fetal bovine serum (FBS, A5256701), penicillin G (100 IU/mL), and streptomycin (100 μg/mL) and incubated at 37°C. THP‐1 macrophages were obtained by treating THP‐1 monocytes with 100 ng/mL phorbol myristate acetate (PMA, P8103, Sigma‐Aldrich, St Louis, Missouri, USA) for 48 h in 6‐well culture plates with 3 × 10⁶ cells in each well. M2‐macrophage polarization was induced by incubation with 20 ng/mL of interleukin‐4 (IL‐4, 200‐04, Peprotech, Suzhou, China) and 20 ng/mL of interleukin‐13 (IL‐13, 200‐13, Peprotech, Suzhou, China) [39].

THP‐1 was pre‐treated with BMS‐986235 (10 μM) for 30 min before inducing M2 macrophage polarization. Supernatant from the cell culture was collected 24 h after BMS‐986235 treatment and then centrifuged and purified. These supernatants, hereafter referred to as conditioned media (CM), were frozen at −80°C for further use. Collected CM includes M2 macrophage CM (M2‐CM) and BMS‐986235 conditioned M2‐macrophage media (M2 + BMS‐CM). Unpolarized macrophage CM (M0‐CM) was used as the control.

Macrophage CM generation is described above and previously [40]. After thawing, supernatants were diluted 1:2 with a complete medium and added to HASMCs. Further experiments were carried out after 24 h of incubation.

2.13. ELOVL6 and FPR2 Transfection With Adenovirus and siRNA

HASMCs and THP‐1 were transfected with an adenovirus containing ELOVL6 (Vigene Bioscience, Jinan, China), while the negative control was an empty adenovirus. siRNA (Vigene Bioscience, Jinan, China) downregulates ELOVL6 expression in both cell types, with siRNA‐NC as the negative control. Protein was extracted 72 h after the siRNA/adenovirus treatment.

After transfection for 72 h, the obtained macrophages were polarized into M2 macrophages using 20 ng/mL of IL‐4 and 20 ng/mL of IL‐13. The collected conditioned medium was frozen at −80°C until further use and is referred to as adenovirus‐ELOVL6 conditioned media (Ad‐ELOVL6‐CM) or siRNA‐ELOVL6 conditioned media (siRNA‐ELOVL6‐CM). Normal macrophage‐conditioned medium (Ad‐NC‐CM and siRNA‐NC‐CM) was used as a control. Additionally, conditioned medium from macrophages (M0 and M2) was collected and frozen for later use, both with and without BMS‐986235 treatment.

FPR2 was knocked down by siRNA (Vigene Bioscience, Jinan, China) in both M2 macrophages and HASMCs, along with negative controls (siRNA‐NC).

2.14. Western Blotting

Western blotting has been performed to comply with guidelines [41]. Cells and human vessel tissue samples were lysed in RIPA lysis buffer (P0013B, Beyotime Biotechnology, Shanghai, China). Equal amounts of protein were loaded onto SDS‐PAGE gels. After electrophoresis, the separated proteins were transferred to the PVDF membrane (Millipore, Billerica, MA, USA) [42]. Next, the membrane was blocked with QuickBlock blocking buffer (P0252, Beyotime Biotechnology, Shanghai, China) for 15 min and incubated with primary antibodies for mouse anti‐TAGLN (1:1000, 60 213–1, Proteintech, Wuhan, China), rabbit anti‐CNN1 (1:1000, BM4088, BOSTER, Wuhan, China), rabbit anti‐FPR2 (1:1000, NLS1878, Novus Biologicals, Colorado, USA), rabbit anti‐ELOVL6 (1:1000, PAB13032, Abnova, Taiwan, China), rabbit anti‐β‐Tubulin (1:1000, 2146, Cell Signaling Technology, Danvers, MA, USA), rabbit anti‐ARG‐1 (1:1000, 93 668, Cell Signaling Technology, Danvers, MA, USA), and rabbit anti‐β‐actin (1:1000, 970S, Cell Signaling Technology, Danvers, MA, USA) at 4°C overnight. After washing with PBS, the membranes were incubated with HRP‐conjugated secondary antibodies (JAC‐115‐035‐003 for mouse; JAC‐305‐035‐003 for rabbit) for 1 h at room temperature. Blots were visualized using the enhanced chemiluminescence detection kit (WBKLS0500, Millipore, Massachusetts, USA), with Amersham ImageQuant 800. The concentration‐response curves for phosphorylated proteins were normalized to total protein levels, as calculated from the band intensities in the western blot.

2.15. Quantitative Real‐Time Polymerase Chain Reaction (qRT‐PCR)

RNAs from mouse artery samples and cells were isolated using TRIzol reagent (15 596 018, Invitrogen, California, USA), following the manufacturer's specifications [43]. Target RNA expression was normalized to β‐actin expression in vessel tissue samples. Fluorescent signals were normalized against an internal reference, and the threshold cycle (Ct) was set within the exponential phase of the PCR. The relative gene expression was calculated by comparing each PCR target's cycle numbers (Ct). The target PCR Ct values were normalized by subtracting the β‐actin Ct value to generate the ΔCt value. The relative expression level between treatments was then calculated using the following equation: relative gene expression = 2^{−(ΔCtsample−ΔCtcontrol)}.

The primer sequences for all genes mentioned are provided in Table S2.

2.16. EdU Incorporation Assay

Cell proliferation was assessed using the 5‐ethynyl‐2′‐deoxyuridine (EdU) assay kit (C10310‐1, RiboBio, Wuhan, China), as detailed previously [44]. HASMCs were seeded in 96‐well plates at a density of 1 × 10⁴ cells per well. Following serum starvation, the cells were exposed to various media for 24 h. All cells were then incubated with 50 μM EdU for 2 h at 37°C. After fixation with 4% formaldehyde in PBS and permeabilization with 0.5% Triton X‐100, the cells were treated with 100 μL of the 1 × apollo reaction cocktail for 30 min and washed with 0.5% Triton X‐100. DNA staining was performed using Hoechst 33342 for 30 min. The staining results were captured using fluorescence microscopy (OLYMPUS, Tokyo, Japan), and the number of EdU‐positive cells in five random fields was counted under a fluorescence microscope.

2.17. Scratch Assay

Scratch assays were employed to assess VSMC migration, as previously described [45]. Confluent cell monolayers in a 12‐well plate were wounded by manually scraping them with a 10 μL pipette tip. Before scraping, the cells were incubated in various media for 24 h. Following scraping, all cells were cultured in a serum‐free medium. The cell migration area was observed using microscopy at 0 and 24 h post‐scraping. ImageJ was used to delineate the wound area through edge enhancement and thresholding methods; the area was calculated at 0 h and 24 h, with the differences in wound area expressed as a percentage of the original wound area to indicate the migration profile. The cell migration rate was calculated using the formula: (scratch area at 0 h—scratch area at 24 h)/scratch area at 0 h × 100 (%).

2.18. Data Collection

We conducted a comprehensive search of the PubMed database (https://pubmed.ncbi.nlm.nih.gov/) for studies publicly available up to August 30, 2024. The search was performed using the following keywords: “Restenosis” (study keyword), “ Homo sapiens ” (organism), and “Expression profiling by array” (study type). Studies were included based on the following criteria: (i) samples included both restenosis and normal tissues or cells; (ii) gene expression profiling of mRNA was conducted; and (iii) sufficient information was provided to perform the analysis. Following a systematic review, we found that only Arencibia et al. [46] met the inclusion criteria, and a further analysis of the gene expression profiles was conducted. This gene expression profile comprised three normal and three restenosis cell samples. Differentially expressed genes (DEGs) were identified based on the inclusion criteria of p < 0.05 and |logFC| ≥ 1, and subsequent correlation analyses were performed.

2.19. Statistical Analysis

All data were analyzed for normality and equal variance. Data are presented as mean ± SEM. Comparisons were performed using independent‐measures one‐way ANOVA with Bonferroni's correction for multiple comparisons, or unpaired Student's t‐test where applicable (α = 0.05). Post hoc tests were conducted only if the F value of the one‐way ANOVA reached the necessary statistical significance level. For data sets that did not follow a normal distribution or when sample size was smaller than 5, the Mann–Whitney U test nonparametric analysis (2‐group analysis) or Kruskal‐Wallis nonparametric test, followed by Dunn multiple comparison, was used. Group sizes were all n ≥ 3, where n equaled the number of independent values, and all data were included in statistical analyses. Where indicated, multiple fields of view from each well, considered as technical replicates, were combined and presented as the mean. Values from separate wells were then treated as biological replicates for analysis. Correlation coefficients were calculated using Pearson's correlation test. p < 0.05 was deemed statistically significant. All data were analyzed using GraphPad Prism 9.0.1 software (GraphPad, La Jolla, CA).

2.20. Materials

BMS‐986235 (Purity: 99.90%) was prepared at WuXi AppTec Inc. (Shanghai, China). For the in vitro model study, the stock solution (10 mM) was dissolved in dimethyl sulfoxide (DMSO), which was purchased from Sigma‐Aldrich (D4540, St Louis, Missouri, USA). The stock solution was diluted with PBS (BL302A, Biosharp, Anhui, China) to 200 μM, and the final concentration of 10 μM was added to the cell culture medium. The same volume of DMSO and PBS was used as a control. For the animal model studies, the stock solution (3 mg/mL) was dissolved in DMSO, diluted with vehicle (0.8% Tween‐80 in saline) to 3 mg/kg, and administered based on mouse body weight. The sham and restenosis groups received the same volume of vehicle (0.8% Tween‐80 in saline i.p.), while the restenosis plus BMS‐986235 group received 3 mg/kg (i.p.) [32, 33]. Details of other materials and suppliers are provided in the specific sections.

3. Results

3.1. Reduced Vascular FPR2 Expression in Patients With Restenosis

The expression of FPR2 was compared in artery samples collected from patients after restenosis and normal (amputees due to trauma) (n = 10 per group). Male samples were used to eliminate potential hormonal influences (e.g., varying oestradiol levels). The mean age of the restenosis group was 50.4 years compared to 46.5 years in the normal group (Table S1). Vessels from the restenosis group were obtained from amputees who had undergone endovascular treatment, selecting segments with visible intimal hyperplasia. Normal artery segments were sourced from patients with traumatic amputations, free of atherosclerosis or endovascular treatment, from a location matched as closely as possible to the restenosis segments (Figure 1A).

FIGURE 1.

Lower vascular FPR2 expression in patients with restenosis. (A) Schematic protocol for collecting human samples from normal vessels and restenosed arteries. (B) Representative cross‐sections of human arteries showing restenosis compared to normal vessels, stained with H&E and EVG (n = 5). The yellow arrow in the EVG‐stained image points to the internal elastic lamina. Scale bar = 500 μm. (C) Representative double immunofluorescence staining alongside quantitative analysis of normal and restenosed arteries using CD163 (green, M2 macrophage marker) and α‐SMA (red, VSMC marker) (n = 5). Scale bar = 500 μm; scale bar (top right corner) = 50 μm. (D) Representative immunohistochemical images together with quantitative analysis of FPR2 expression in human normal and restenosed arteries (n = 5). Red arrows indicate positive staining for FPR2. Scale bars = 500 μm and 50 μm (right top corner). (E) Representative western blot images and quantitative analysis of FPR2 expression in normal and restenosed arteries (n = 8). β‐actin was used as a loading control. The yellow dashed line marks the dividing line between the intima and media. Data represent the mean ± SEM from n independent experiments. Independent measures used an unpaired Student's t‐test for comparison of two groups. ** p < 0.01, ***p < 0.001. FPR2, formylpeptide receptor 2; DAPI, 4′,6‐diamidino‐2‐phenylindole; α‐SMA, alpha‐smooth muscle actin; RS, restenosis; H&E, hematoxylin and eosin; EVG, verhoeff's van gieson staining.

Neointimal hyperplasia (Figure 1B), primarily composed of VSMCs (Figure S5A), was evident in human arteries of restenosis sections compared to the normal group. In addition, human restenosis exhibited higher VSMC (α‐SMA positive) and M2‐macrophages (CD163 positive) in the area of neointimal hyperplasia (Figure 1C). At the same time, there was no such increase in M1‐macrophages (CD80 positive, Figure S5B). FPR2 protein expression was markedly lower in vessels with restenosis using immunohistochemistry (Figure 1D) and western blot (Figure 1E).

3.2. Global and Myeloid‐Specific Fpr2 Knockout Exacerbates Wire Injury‐Induced Neointimal Hyperplasia

To clarify the role of FPR2 in neointimal hyperplasia, we compared the effects of wire injury in mice with global Fpr2‐knockout (Fpr2^−/−) and C57BL/6 (WT) mice. Global Fpr2 knockout did not influence body weight or plasma lipid levels (Table 1). Compared to WT mice, Fpr2^−/− mice reduced blood perfusion of the injured limb (Figure 2A,B), exhibited more severe neointimal hyperplasia, indicated by greater arterial narrowing (the ratio of the area of neointimal hyperplasia to arterial lumen) (Figure 2A,B), and increased smooth muscle fiber (Masson staining) in the neointimal hyperplasia area (Figure 2A,B). Lower levels of VSMC contractile markers CNN1 and TAGLN accompanied these changes (Figure 2C; Figure S6A) consistent with a de‐differentiated phenotype. Furthermore, vessels from Fpr2^−/− mice showed higher cell proliferation markers (PCNA) in VSMCs (expressing the VSMC marker α‐SMA) compared to WT mice (Figure 2D).

TABLE 1.

Characteristics of wire injured global and macrophage‐specific Fpr2 knock‐out mice.

Group	WT n = 6	Fpr2−/− n = 6	P	Fpr2flox/flox n = 5	Fpr2MKO n = 5	P
Weight (g)	26.8 ± 0.6	25.0 ± 0.4	NS	26.1 ± 1.2	25.4 ± 1.1	NS
ALT (U/L)	24.9 ± 4.2	27.8 ± 2.5	NS	26.8 ± 2.3	22.4 ± 4.6	NS
AST (U/L)	90.8 ± 7.0	88.0 ± 5.4	NS	81.9 ± 4.2	103.4 ± 4.3	NS
TG (mmol/L)	0.5 ± 0.1	0.5 ± 0.1	NS	0.4 ± 0.3	0.3 ± 0.3	NS
TC (mmol/L)	3.1 ± 0.2	3.2 ± 0.2	NS	3.7 ± 0.4	3.5 ± 0.5	NS
LDL‐C (mmol/L)	0.9 ± 0.1	0.9 ± 0.2	NS	0.8 ± 0.3	0.8 ± 0.2	NS
HDL‐C (mmol/L)	2.5 ± 0.1	2.5 ± 0.2	NS	3.5 ± 0.5	3.2 ± 0.5	NS

Note: All values are expressed as the mean ± SEM and analyzed using Kruskal‐Wallis nonparametric test.

Abbreviations: ALT, alanine aminotransferase; AST, glutamic‐oxalacetic transaminase; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; NS, not significant; TC, total cholesterol; TG, triglycerides.

FIGURE 2.

Global Fpr2 deficiency exacerbates injury‐induced neointima formation in vivo, which is associated with adverse VSMC de‐differentiation. (A) Representative LSCI images and representative images of morphological staining (H&E and Masson staining) for arterial sections from WT and Fpr2^−/− mice (n = 10). Scale bar = 50 μm. For LSCI, the red box highlights the limb post‐surgery, while the blue box indicates the normal limb. In Masson staining, blue indicates collagen, and red indicates muscle. (B) Quantitative analysis of blood perfusion, arterial narrowness, and the percentage of muscle fiber area in neointimal hyperplasia in WT and Fpr2^−/− mice. (C) Relative protein expression of VSMC contractile markers (CNN1 and TAGLN) in injury‐induced neointima formation in WT and Fpr2^−/− mice (n = 3). β‐actin was used as a loading control. (D) Representative immunofluorescent images and quantitative analysis of injury‐induced neointima formation samples from WT and Fpr2^−/− mice stained with PCNA (green) and α‐SMA (red). Scale bar = 75 μm and 25 μm (right top corner). (E‐G) Representative immunofluorescence images of injury‐induced neointima formation in WT and Fpr2^−/− mice, stained for ARG‐1, with quantitative analysis of ARG‐1‐positive area and ARG‐1‐positive cell number (n = 5). Scale bar = 75 μm and 25 μm (right top corner). (H) Relative protein expression of M2 macrophage marker (ARG‐1) in WT and Fpr2^−/− mice (n = 3). β‐actin was used as a loading control. The yellow dashed line indicates the boundary between the intima and media. Data represent the mean ± SEM from n independent experiments. Independent measures used an unpaired Student's t‐test or Mann–Whitney U test nonparametric analysis for two comparisons. *p < 0.05, **p < 0.01, ****p < 0.0001. Fpr2, formylpeptide receptor 2; WT, wild‐type; LSCI, laser speckle contrast imaging; PCNA, proliferating cell nuclear antigen; α‐SMA, alpha smooth muscle actin; CNN1, calponin 1, basic; TAGLN, transgelin; ARG‐1, arginase‐1; DAPI, 4′,6‐diamidino‐2‐phenylindole; H&E, hematoxylin and eosin.

Given the increased M2 macrophage content in human restenosis samples, we assessed the impact of FPR2 on macrophage content at the injured sites in mice. Consistent with the human samples, the content of M2 macrophages (as indicated by ARG‐1 expression) was higher in Fpr2^−/− mice than in WT mice (Figure 2E–H; Figure S6A–C) within the area of neointimal hyperplasia of Fpr2 deficient mice.

To explore the role of FPR2 in macrophage during restenosis, we generated myeloid‐selective Fpr2‐deficient (Fpr2^MKO) mice. As occurred with the global Fpr2 knockout, myeloid‐specific Fpr2 deletion did not affect body weight or plasma lipid levels (Table 1). A similar VSMC phenotype switch in response to injury was observed in global Fpr2^−/− and Fpr2^MKO animals. This was indicated by greater intimal thickness, lower blood perfusion in the limb subjected to wire injury, and higher levels of VSMCs in neointimal hyperplasia in Fpr2^MKO mice (Figure 3A,B). Fpr2^MKO mice showed lower VSMC contractile markers (CNN1 and TAGLN, Figure 3C; Figure S6D), increased PCNA and α‐SMA levels (Figure 3D), and elevated M2 macrophage markers (ARG‐1) in Fpr2^MKO (Figure 3E–H; Figure S6D).

FIGURE 3.

Myeloid‐specific deficiency of Fpr2 exacerbates injury‐induced neointima formation in vivo and is associated with detrimental de‐differentiation of VSMCs. (A) Representative LSCI images and representative cross‐sections of Fpr2^flox/flox and Fpr2^MKO mice stained with H&E and Masson (n = 6). Scale bar = 50 μm. (B) Quantitative analysis of blood perfusion analysis and arterial narrowing, along with the percentage of muscle fiber area in neointimal hyperplasia. (C) Relative protein expression of VSMC contractile markers (CNN1 and TAGLN) in neointima formation of Fpr2^flox/flox and Fpr2^MKO mice (n = 3). β‐actin was used as a loading control. (D) Representative immunofluorescent images and quantitative analysis of restenosis samples from Fpr2^flox/flox and Fpr2^MKO mice stained with PCNA (green) and α‐SMA (red). Scale bar = 75 μm and 25 μm (right top corner). (E‐G) Representative immunofluorescence images of injury‐induced neointima formation in Fpr2^flox/flox and Fpr2^MKO mice, stained for ARG‐1, with quantitative analysis of ARG‐1‐positive area and ARG‐1‐positive cell number (n = 5). Scale bar = 75 μm and 25 μm (right top corner). (H) Relative protein expression of M2 macrophage marker (ARG‐1) in Fpr2^flox/flox and Fpr2^MKO mice (n = 3). β‐actin was used as a loading control. The yellow dashed line indicates the boundary between the intima and media. Data represent the mean ± SEM from n independent experiments. Independent measures used unpaired Student's t‐test or Mann–Whitney U test nonparametric analysis for two comparisons. *p < 0.05, **p < 0.01, ***p < 0.001. Fpr2, formylpeptide receptor 2; WT, wild‐type; LSCI, laser speckle contrast imaging; PCNA, proliferating cell nuclear antigen; α‐SMA, alpha smooth muscle actin; CNN1, calponin 1, basic; TAGLN, transgelin; ARG‐1, arginase‐1; DAPI, 4′,6‐diamidino‐2‐phenylindole; H&E, hematoxylin and eosin.

Given that myeloid‐specific Fpr2 knockout mice exhibited similar neointimal hyperplasia to global Fpr2 knockout mice in the restenosis model, we hypothesized that FPR2 inhibits VSMC de‐differentiation in restenosis by a macrophage‐dependent mechanism.

3.3. FPR2 Agonist BMS‐986235 Prevents De‐differentiation of VSMCs

Employing the selective FPR2‐agonist BMS‐986235, we investigated whether FPR2 activation regulates VSMC de‐differentiation to mitigate neointimal hyperplasia. Given that elevated M2‐macrophage content drives neointimal hyperplasia, we first examined whether FPR2 signaling in M2 macrophages influences the VSMC phenotype. BMS‐986235 (10 μM) treatment maximally inhibited IL‐10 production (Figure S7), and was therefore used in subsequent experiments. To further elucidate the link between macrophage FPR2 activation and VSMC proliferation and migration, conditioned media from macrophages treated with vehicle or BMS‐986235 (M0‐CM, M2‐CM, and M2 + BMS‐986235‐CM) were collected and used to culture HASMCs. To differentiate between the direct and indirect effects of BMS‐986235, HASMCs cultured in M2‐CM were additionally treated with BMS‐986235 (M2‐CM + BMS‐986235) (Figure 4A). BMS‐986235 supplementation (M2 + BMS‐986235‐CM) significantly reduced HASMC proliferation (EdU incorporation assay, Figure 4B), migration (scratch assay, Figure 4C; Figure S11A), and upregulated contractile markers (CNN1 and TAGLN, Figure 4D). Furthermore, a significant inhibitory effect was observed in the M2‐CM + BMS‐986235 group compared to the M2‐CM group. However, this effect was still less pronounced than in the M2 + BMS‐986235‐CM group (Figure 4B–D), indicating that FPR2 agonism may reduce macrophage‐mediated VSMC de‐differentiation.

FIGURE 4.

The FPR2 small‐molecule agonist BMS‐986235 indirectly reduces M2 macrophage‐induced HASMC de‐differentiation and directly maintains the differentiated phenotype of HASMCs. (A) The schematic outlining the protocol for obtaining macrophage‐conditioned medium (CM) for HASMCs culture is as follows: Group 1: M0 macrophages were pre‐treated with the vehicle. Group 2: M0 macrophages were pre‐treated with the vehicle and then polarized to M2 macrophages. Group 3: M0 macrophages were pre‐treated with BMS‐985235 and subsequently polarized to M2 macrophages. In all three groups (1–3), the supernatant was collected and added to the HASMCs. Further experiments were conducted 24 h after the CM was added. Group 4: M0 macrophages were pre‐treated with the vehicle, polarized to M2 macrophages, and their supernatant was collected to culture HASMCs simultaneously treated with BMS‐986235. (B) Representative images and quantitative analysis of the EdU assay demonstrating the effect on HASMC proliferation in CM, along with quantification of EdU‐positive cells (n = 5). Scale bar = 75 μm. (C) Representative images and quantitative analysis of the scratch assay illustrating the effect on HASMC migration in CM (n = 5). Images were captured at 0 and 24 h post‐scratch (enhanced edge image). Scale bar = 200 μm. (D) Representative western blot and quantitative analysis of the VSMC contractile markers (CNN1 and TAGLN) protein expression in HASMCs treated with CM (n = 3). β‐Tubulin was utilized as a loading control. (E) The schematic outlining the protocol for HASMC culture is as follows: Group 1: HASMCs were treated with PBS and vehicle. Group 2: HASMCs were pre‐treated with the vehicle and then added with PDGF‐BB. Group 3: HASMCs were pre‐treated with BMS‐985235 and then added with PDGF‐BB. (F) Representative images and the quantification of HASMCs assessed by the EdU assay (n = 5). Scale bar = 75 μm. (G) Representative images and quantification of HASMCs evaluated by the scratch assay (n = 5). Images were acquired at 0 and 24 h postscratch (enhanced edge image). Scale bar = 200 μm. (H) Representative western blot and quantitative analysis of VSMC contractile markers (CNN1 and TAGLN) protein expression in HASMCs pre‐treated with BMS‐986235 followed by PDGF‐BB stimulation (n = 3). β‐Tubulin was used as a loading control. Data are expressed as mean ± SEM from n independent experiments. Independent measures one‐way ANOVA with Bonferroni's correction or Kruskal‐Wallis nonparametric test for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. HASMC, human aortic smooth muscle cell; EdU, 5‐ethynyl‐2′‐deoxyuridine; PDGF, platelet‐derived growth factor; CNN1, calponin 1, basic; TAGLN, transgelin; CM, conditioned medium.

To better simulate the microenvironment of injured arteries, equal consideration should be given to the direct effects of BMS‐986235 on VSMC. For a focused investigation into FPR2's role in VSMC phenotypic switching, HASMCs were treated with recombinant human platelet‐derived growth factor‐BB (PDGF‐BB) to induce a de‐differentiated state. We then examined the direct effects of FPR2 agonism on this process. Notably, BMS‐986235 significantly attenuated PDGF‐BB‐induced VSMC proliferation and migration (Figure 4E–H, Figure S11B), indicating that FPR2 activation in HASMCs directly inhibits de‐differentiation.

3.4. The FPR2 Agonist BMS‐986235 Inhibits Wire Injury‐Induced Neointimal Hyperplasia

To evaluate the efficacy of the FPR2 agonist against wire injury‐induced neointimal hyperplasia, wild‐type (WT) mice were randomized to receive either BMS‐986235 or vehicle. Treatment with BMS‐986235 did not affect the body weight or plasma lipid levels of the mice (Table 2). BMS‐986235 treatment improved blood perfusion in the injured limb (LSCI, Figure 5A) and reduced arterial narrowing caused by wire injury, as illustrated in Figure 5B,C. Additionally, it lowered the level of smooth muscle fibers in the neointimal hyperplasia area (Figure 5B,C), upregulated contractile markers (Cnn1 and Tagln, Figure 5D), diminished VSMC proliferation (PCNA‐positive VSMCs, Figure 5E), and decreased M2 macrophages (ARG‐1 levels, Figure 5F). Overall, BMS‐986235 treatment preserved a differentiated VSMC phenotype in preclinical models of restenosis.

TABLE 2.

Characteristics of wire‐injured mice after BMS‐986235.

Group	Sham n = 6	Restenosis n = 6	Restenosis+ BMS‐986235 n = 6	P a	P b
Weight (g)	26.4 ± 0.9	26.9 ± 0.3	27.6 ± 0.3	0.02	NS
ALT (U/L)	29.8 ± 7.2	23.0 ± 4.8	31.9 ± 6.8	NS	NS
AST (U/L)	106.4 ± 6.5	97.4 ± 6.6	92.4 ± 4.0	NS	NS
TG (mmol/L)	0.8 ± 0.2	0.6 ± 0.1	0.5 ± 0.1	NS	NS
TC (mmol/L)	3.3 ± 0.1	3.0 ± 0.1	2.9 ± 0.2	NS	NS
LDL‐C (mmol/L)	1.0 ± 0.2	0.7 ± 0.0	0.9 ± 0.1	NS	NS
HDL‐C (mmol/L)	2.7 ± 0.2	2.8 ± 0.2	2.5 ± 0.1	NS	NS

Note: All values are expressed as the mean ± SEM and analyzed using Kruskal‐Wallis nonparametric test.

Abbreviations: ALT, alanine aminotransferase; AST, glutamic‐oxalacetic transaminase; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; NS, not significant; TC, total cholesterol; TG, triglycerides.

^a Comparing restenosis vs. sham.

^b Comparing restenosis+BMS‐986235 vs. restenosis.

FIGURE 5.

The FPR2 agonist BMS‐986235 reduces injury‐induced neointima formation. (A) LSCI were conducted on mouse treated with or without BMS‐986235, and quantitative analysis of blood perfusion analysis (n = 5). In LSCI, the red box marks the limb after surgery, while the blue box highlights the normal limb. (B, C) Morphological staining (H&E and Masson staining) were conducted on mouse artery sections (n = 14). In Masson staining, blue represents collagen, and red indicates muscle. Quantitative analyses included arterial narrowing and the percentage of muscle fiber area in neointimal hyperplasia. (D) Relative mRNA expression of VSMC contractile markers (Cnn1 and Tagln). (E) Representative images of double immunofluorescence staining showing injury‐induced neointima formation with PCNA (green) and α‐SMA (red). Scale bar = 75 μm and 25 μm (right top corner). (F) Representative immunofluorescence staining revealing injury‐induced neointima formation with ARG‐1. Scale bar = 75 μm and 25 μm (right top corner). The yellow dashed line indicates the boundary between the intima and media. Data are presented as mean ± SEM from n independent experiments. Statistical analyses employed one‐way ANOVA with Bonferroni's correction for multiple comparisons. ND, not detect, **p < 0.01, ***p < 0.001, ****p < 0.0001. α‐SMA, alpha smooth muscle actin; Cnn1, calponin 1, basic; Tagln, transgelin; LSCI, laser speckle contrast imaging; PCNA, proliferating cell nuclear antigen; DAPI, 4′,6‐diamidino‐2‐phenylindole; ARG‐1, arginase‐1; H&E, hematoxylin & eosin; RS, restenosis.

3.5. FPR2‐Agonist BMS‐986235 Limits the Progression of Restenosis Partially via ELOVL6

To investigate the molecular pathways and proteins involved in the protective effect of FPR2 activation against restenosis, we performed RNA‐seq analysis on arterial samples from mice subjected to wire injury. By comparing the sham group with the restenosis group, we identified 295 genes with altered expression levels (Figure 6A red circle). Additionally, we identified 123 genes with different expression levels in BMS‐986235‐treated mice, representing genes affected by FPR2 agonism (Figure 6A blue circle). Ten genes were influenced by both restenosis and BMS‐986235 treatment, as illustrated by the heat map (Figure 6B; Dataset S1). Among these 10 genes, Elovl6 was the only gene that exhibited a consistent change in mRNA levels upon validation by qRT‐PCR (Figure 6C; Figure S8). Elovl6 mRNA expression was higher in restenotic arteries compared to sham counterparts. In contrast, BMS‐986235 treatment reduced Elovl6 expression compared with the restenosis group. Utilizing qRT‐PCR data from murine tissue samples, we found a significant correlation between increased Elovl6 mRNA levels and decreased levels of VSMC contractile markers, specifically Cnn1 (r ² = 0.8289) and Tagln (r ² = 0.6796) (Figure 6D). Consistent with this observation, BMS‐986235 treatment decreased ELOVL6 protein levels in M2 macrophages (Figure 6E) and VSMCs (Figure 6F) within neointima cross‐sections.

FIGURE 6.

Transcriptomics reveals that BMS‐986235 treatment inhibits neointima formation via ELOVL6. (A) The Venn diagram displays differentially expressed genes among the sham vs. restenosis group and the restenosis+BMS‐986235 vs. restenosis group from transcriptomics analysis. The red and blue circles overlap, indicating the potential differential expression of genes affected by BMS‐986235 treatment in injury‐induced restenosis. Common genes in the overlapping area are influenced by both restenosis and BMS‐986235 treatment. (B) Heat map of differentially expressed genes. Red represents relative upregulation, while blue indicates relative downregulation (after correction). (C) Relative mRNA expression of Elovl6 (n = 5) measured by qRT‐PCR. (D) Correlation analysis between Elovl6 mRNA expression and VSMC contractile marker (Cnn1 and Tagln) mRNA in sham, restenosis, and restenosis+BMS‐986235. (E) Representative images of double immunofluorescence staining and quantitative analysis showing ELOVL6 (green) and ARG‐1 (red) expression in neointima formation across sham, restenosis, and restenosis+BMS‐986235 groups. Scale bar = 75 μm and 25 μm (right top corner). (F) Representative images of double immunofluorescence staining and quantitative analysis showing ELOVL6 (green) and α‐SMA (red) expression in neointima formation across sham, restenosis, and restenosis+BMS‐986235 groups. Scale bar = 75 μm and 25 μm (right top corner). (G) Representative western blot of ELOVL6 expression in human normal and restenosis arteries (n = 8). β‐actin was used as a loading control. (H) Correlation between ELOVL6 and FPR2 protein expression in human normal and restenosis vessels. The yellow dashed line indicates the boundary between the intima and media. Data represent the mean ± SEM from n independent experiments. Independent measures unpaired Student's t‐test for two comparisons, along with one‐way ANOVA with Bonferroni's or Kruskal‐Wallis nonparametric test for multiple comparisons. ND, not detect, *p < 0.05, **p < 0.01, ****p < 0.0001. Pearson's test was used for correlation analysis. α‐SMA, alpha smooth muscle actin; Cnn1, calponin 1, basic; Tagln, transgelin; DAPI, 4′,6‐diamidino‐2‐phenylindole; ARG‐1, arginase‐1; RS, restenosis.

Expression of ELOVL6 was further investigated in tissue samples from patients suffering from restenosis. Arterial sections affected by restenosis exhibited significantly greater ELOVL6 protein expression compared to normal controls, and this increase in expression was particularly evident in the intima (Figure 6G, Figure S9A). ELOVL6 protein expression was negatively correlated with the expression of FPR2 in human normal and restenosis arteries (r ² = 0.4245, Figure 6H). We also mined a published gene expression profile from a study exposing VSMCs to stent fragments in vitro, resembling in‐stent restenosis [46]. After analyzing the correlation of differentially expressed genes, we revealed that elevated human Elovl6 expression (Figure S9B) corresponded with a significant reduction in VSMC contractile marker in human (Cnn1 and myosin heavy chain 10 (MYH10), Figure S9C).

3.6. FPR2 Regulates VSMC De‐differentiation Partially via ELOVL6

To examine the role of ELOVL6 in vitro, we employed transient overexpression (adenovirus) and knockdown (siRNA) strategies in both M2 macrophages and HASMCs. Successful manipulation of ELOVL6 levels in both cell types was confirmed by western blotting (Figure S10A,B). HASMCs treated with conditioned media from M2 macrophages overexpressing ELOVL6 (Figure 7A) demonstrated enhanced HASMC proliferation (Figure 7B), migration (Figure 7C; Figure S11C), and reduced contractility (Figure 7D). In contrast, M2 macrophages with ELOVL6 knockdown exhibited the opposite effects on VSMC proliferation, migration, and contractility (Figure 7B–D). These results suggest that ELOVL6 in macrophages plays a role in macrophage‐regulated VSMC de‐differentiation. In line with observations in M2 macrophages, direct upregulation of ELOVL6 in HASMCs led to increased proliferation and migration while inhibiting the expression of VSMC contractile markers (Figure 7E–H; Figure S11D). Conversely, downregulation of ELOVL6 in HASMCs suppresses proliferation and migration while promoting the expression of VSMC contractile markers (Figure 7E–H).

FIGURE 7.

ELOVL6 inhibits VSMC de‐differentiation through M2 macrophages and directly in HASMCs. (A) Schematic protocol for preparing macrophage CM following ELOVL6 transfection. M0 macrophages were infected with adenovirus (Ad‐NC or Ad‐ELOVL6) or transfected with siRNA (siRNA‐NC or siRNA‐ELOVL6) and then induced to polarize into M2 macrophages. The resulting supernatant was added to HASMCs. (B) Representative images and quantitative analysis of the EdU assay show the effect on HASMC proliferation cultured in CM (n = 5). Scale bar = 75 μm. (C) Representative images and quantitative analysis of the scratch assay showing the effect on the migration of HASMCs cultured in CM (n = 5). Images were taken at 0 and 24 h post‐scratch (enhanced edge image). Scale bar = 200 μm. (D) Representative western blot and quantitative analysis of VSMC contractile marker (CNN1 and TAGLN) protein expression in HASMCs cultured in CM (n = 3). β‐Tubulin was used as a loading control. (E) Schematic protocol for the HASMC culture assay following ELOVL6 transfection. HASMCs were infected with adenovirus (Ad‐NC or Ad‐ELOVL6) and transfected with siRNA (siRNA‐NC or siRNA‐ELOVL6), then incubated with PDGF‐BB to induce de‐differentiation. (F) Representative images and quantification of HASMCs assessed by the EdU assay (n = 5). Scale bar = 75 μm. (G) Representative images and quantification of HASMCs assessed by the scratch assay (n = 5). Images were taken at 0 and 24 h post‐scratch (enhanced edge image). Scale bar = 200 μm. (H) Representative western blot and quantitative analysis of VSMC contractile marker (CNN1 and TAGLN) protein expression in HASMCs (n = 3). β‐Tubulin was used as a loading control. Data represent the mean ± SEM from n independent experiments. Independent measures one‐way ANOVA with Bonferroni's correction or Kruskal‐Wallis nonparametric test for multiple comparisons. *p < 0.05, **p < 0.01, ****p < 0.0001. HASMC, human aortic smooth muscle cell; CM, conditioned media; EdU, 5‐ethynyl‐2′‐deoxyuridine; CNN1, calponin 1, basic; TAGLN, transgelin; DAPI, 4′,6‐diamidino‐2‐phenylindole; NC, normal control; Ad, adenovirus; siRNA, small interfering RNA.

To determine the FPR2‐dependency of the therapeutic effect of BMS‐986235, we administered BMS‐986235 to FPR2‐knockdown M2 macrophages (Figure S10C) and then treated HASMCs with the CM. Compared to HASMCs with FPR2 knockdown alone, those exposed to BMS‐986235 showed no significant changes in proliferation, contractile markers, or migration (Figure 8A–C; Figure S11E), indicating a lack of functional response without FPR2 signaling. In contrast, HASMCs treated with CM from control macrophages (with siRNA‐NC) exposed to BMS‐986235 showed a significant reduction in proliferation and migration, accompanied by a marked increase in the expression of contractile markers (Figure 8A–C; Figure S11E). Additionally, HASMCs transfected with siRNA targeting FPR2 (Figure S10D) neutralized its inhibitory effects on proliferation and migration and its ability to enhance contractile biomarker expression (Figure 9A–C; Figure S11F). Notably, FPR2 knockdown in both M2 macrophages and HASMCs eliminated the capacity of BMS‐986235 to suppress HASMC de‐differentiation. Furthermore, BMS‐986235 significantly reduced ELOVL6 levels in cells, but this reduction was no longer observed in FPR2‐knockdown cells (Figures 8C and 9C), suggesting that the downregulation of ELOVL6 by BMS‐986235 is critically FPR2‐dependent.

FIGURE 8.

BMS‐986235 attenuates VSMC de‐differentiation through the FPR2/ELOVL6 axis in macrophages. (A) Representative images and quantification of EdU‐positive cells in HASMCs treated with siRNA‐NC + Vehicle/BMS‐986235‐CM and siRNA‐FPR2 + Vehicle/BMS‐986235‐CM (n = 5). Scale bar = 75 μm. (B) Representative images and quantitative analysis of the scratch assay in HASMCs treated with siRNA‐NC + Vehicle/BMS‐986235‐CM and siRNA‐FPR2 + Vehicle/BMS‐986235‐CM (n = 5). Images were captured at 0 and 24 h post‐scratch (enhanced edge image). Scale bar = 200 μm. (C) Representative western blot and quantitative analysis of ELOVL6 expression and VSMC contractile markers (CNN1 and TAGLN) in HASMCs treated with siRNA‐NC + Vehicle/BMS‐986235‐CM or siRNA‐FPR2 + Vehicle/BMS 986235‐CM (n = 3). β‐Tubulin served as a loading control. Data represent the mean ± SEM from n independent experiments. Independent measures one‐way ANOVA with Bonferroni's correction or Kruskal‐Wallis nonparametric test for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. HASMC, human aortic smooth muscle cell; CM, conditioned media; EdU, 5‐ethynyl‐2′‐deoxyuridine; CNN1, calponin 1, basic; TAGLN, transgelin; DAPI, 4′,6‐diamidino‐2‐phenylindole; NC, normal control; Ad, adenovirus; siRNA, small interfering RNA.

FIGURE 9.

BMS‐986235 suppresses VSMC de‐differentiation by directly targeting the FPR2/ELOVL6 signaling axis in HASMCs. (A) Representative images and quantitative analysis of the EdU assay in HASMCs treated with BMS‐986235 after transfection with siRNA‐FPR2 or siRNA‐NC (n = 5). Scale bar = 75 μm. (B) Representative images and quantitative analysis of scratch assay in HASMCs treated with BMS‐986235 following transfection with siRNA‐FPR2 or siRNA‐NC (n = 5). Images were captured at 0 and 24 h post‐scratch (enhanced edge image). Scale bar = 200 μm. (C) Representative western blot and quantitative analysis of ELOVL6 expression and VSMC contractile markers (CNN1 and TAGLN) in HASMCs treated with BMS‐986235 post‐transfection with siRNA‐FPR2 or siRNA‐NC (n = 3). β‐Tubulin was used as a loading control. Data represent the mean ± SEM from n independent experiments. Independent measures one‐way ANOVA with Bonferroni's correction or Kruskal‐Wallis nonparametric test for multiple comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. HASMC, human aortic smooth muscle cell; CM, conditioned media; EdU, 5‐ethynyl‐2′‐deoxyuridine; CNN1, calponin 1, basic; TAGLN, transgelin; DAPI, 4′,6‐diamidino‐2‐phenylindole; NC, normal control; Ad, adenovirus; siRNA, small interfering RNA.

4. Discussion

We investigated the integral role of the immunomodulatory GPCR, FPR2, in neointimal hyperplasia, a key driver of restenosis following endovascular procedures such as stenting or angioplasty. Our integrative approach has established that FPR2 plays a protective role against the progression of restenosis. We present the first observation of significantly reduced FPR2 expression in artery sections from patients with restenosis post‐angioplasty. Consistent with these human observations, global and myeloid‐specific Fpr2‐knockout mice exhibited exacerbated neointima formation, greater vessel lumen narrowing, increased M2 macrophage content, and enhanced VSMC de‐differentiation in response to vascular injury. These findings highlight the essential role of FPR2 in regulating myeloid cell function, particularly in M2 macrophages, during restenosis [17]. Importantly, pharmacological activation of FPR2 with the selective agonist BMS‐986235 effectively inhibited VSMC de‐differentiation, as demonstrated by reduced proliferation and migration in human aortic smooth muscle cells (HASMCs) and attenuation of neointimal hyperplasia in a murine wire injury model. Transcriptomic analysis (RNA‐seq) further identified ELOVL6, a long‐chain fatty acid elongase, as a downstream effector of FPR2 signaling in both VSMCs and M2 macrophages, suggesting a mechanistic link between FPR2 activation and its protective effects in restenosis. Thus, our story supports further investigation of FPR2 agonists in preventing restenosis after angioplasty.

Our study is the first to demonstrate a significant reduction in FPR2 expression in human patients with restenosis and murine models of restenosis. The mechanism underlying the downregulation of FPR2 in restenosis patients is not yet fully understood. Previous observations suggest that this downregulation may be associated with alterations in immune response pathways. FPR2, a receptor expressed on immune cells such as macrophages and neutrophils, is known to be modulated in response to inflammatory signals [47]. In the context of restenosis, the persistent inflammatory environment could induce a compensatory downregulation of FPR2 [26, 48, 49]. This regulatory response may serve as a protective mechanism to limit excessive immune activation and prevent further damage to the vessel wall. Specifically, decreased FPR2 expression might mitigate the detrimental effects of sustained inflammation by reducing immune cell activation and vascular cell apoptosis. Further research is needed to elucidate this potential protective role and its implications for restenosis management. Our data suggest that global Fpr2 knockout mice exhibited an exaggeration of neointimal hyperplasia. This is consistent with the reported reduction in FPR2 expression in abdominal aortic aneurysms, which is associated with inflammatory cell infiltration and leads to vascular wall rupture [49]. Furthermore, neutrophil adhesion and extravasation were elevated following ischemia–reperfusion injury of the mesenteric artery in Fpr2‐null mice compared to their wild‐type counterparts [50]. Given that the M2 macrophage is one of the primary immune cell types driving the progression of restenosis [12, 51], we also evaluated the role of macrophage FPR2 in neointimal hyperplasia using myeloid‐specific Fpr2‐knockout mice. The targeted deletion of FPR2, specifically in macrophages, produced results similar to the global FPR2 knockout, further highlighting the importance of macrophages. Our study provides the first evidence supporting the possibility that FPR2, previously described as the master switch in the resolution of inflammation [20, 21, 22], maybe a druggable target for preventing restenosis.

Expanding upon previous observations that FPR2 agonism by BMS‐986235 promotes the resolution of inflammation in heart failure [18], we investigated its potential to limit the progression of restenosis. It has been well documented that VSMC de‐differentiation, characterized by VSMC migration from the media to the intima and subsequent proliferation of those cells, leads to intimal hyperplasia and restenosis following vascular surgery [6, 7, 8]. We demonstrated for the first time that the FPR2 agonist BMS‐986235 ameliorates restenosis by inhibiting VSMC de‐differentiation indirectly via macrophages, in addition to its direct actions on VSMCs. Unlike resolvin‐D1, an endogenous FPR2 agonist that only reduces PDGF‐induced VSMC migration [52], BMS‐986235 acts on both M2 macrophages and VSMCs to suppress VSMC proliferation and migration, potentially making it more effective. Moreover, small‐molecule FPR2 agonists, such as BMS‐986235 and Cmpd43, offer advantages over unstable endogenous ligands, for example by facilitating oral administration, potentially effectively addressing restenosis from neointima formation [19, 53]. Our research revealed that BMS‐986235 restricts M2 macrophage polarization during restenosis and suppresses VSMC de‐differentiation. Nevertheless, since FPR2 is abundantly expressed on macrophages and regulates key pro‐resolving functions, including cytokine release, efferocytosis, and trafficking, it is plausible that BMS‐986235 may also influence macrophage migration in the context of restenosis. While this was outside the scope of the present study, it represents an important future direction. An additional consideration is whether FPR2 primarily regulates monocyte migration into the vascular wall, with subsequent differentiation into macrophages exhibiting an M2‐like phenotype (including specific M2 subtypes), rather than directing the migration of mature M2 macrophages themselves. While this was not directly assessed in the present study, it represents an important mechanism that warrants future investigation to fully define the role of FPR2 signaling in immune cell recruitment and fate during restenosis.

In contrast, in the study by Cooray et al. [54], administering the anti‐inflammatory FPR2 agonist (AnxA1) to wild‐type and Fpr2/3 knockout mice increased IL‐10 expression in the peritoneum. Similarly, García et al. [33] investigated the immunomodulatory effects of BMS‐986235 in human whole‐blood samples. They observed a dose‐dependent increase in IL‐10 production when whole blood from healthy donors was exposed to BMS‐986235. These findings suggest that BMS‐986235 modulates IL‐10 levels in human whole blood, further validating its role in influencing cellular immune responses. As reported in the literature, the EC₅₀ of BMS‐986235 for human FPR2 was ~0.41 nM in HEK293 cells stably overexpressing FPR2 [32]. However, in our study, macrophages and VSMCs exhibited significantly lower basal FPR2 expression levels compared to those in the recombinant HEK293 model. This difference may account for the higher effective concentrations of BMS‐986235 required in our experimental system. Specifically, the variation between the concentration of BMS‐986235 employed in our study and the reported EC₅₀ of ~0.41 nM for human FPR2 in HEK293 cells reflects inherent methodological differences, including species‐specific factors, cellular context, and assay conditions. Collectively, our data indicate that FPR2 agonism exerts tissue‐specific and disease‐specific effects. Therefore, it is crucial to consider these factors when evaluating the role of FPR2 agonists in various physiological and pathological contexts, which may lead to a more comprehensive understanding of their therapeutic potential and underlying mechanisms.

To investigate the downstream molecular mechanism underlying FPR2 agonism in preventing restenosis, we employed RNA‐seq analysis. While RNA‐seq provides high‐throughput coverage and sensitivity for detecting low‐abundance transcripts, it can also generate false‐positive DEGs due to factors such as insufficient sequencing depth or background noise. To minimize this, we applied stringent thresholds (p < 0.05 and |logFC| ≥ 2). Because RNA‐seq quantification reflects relative expression levels and is influenced by gene length, GC content, and sequencing biases, validation by quantitative real‐time PCR (qRT‐PCR) is essential. ELOVL6 [55]–a rate‐limiting enzyme responsible for elongating saturated and monounsaturated fatty acids–emerged as a critical gene, being consistently validated at the mRNA level by qRT‐PCR with statistical significance. By contrast, Pm20d1 and Slc4a4 displayed expression trends concordant with RNA‐seq but without statistical significance, likely reflecting biological sample variability. Accordingly, these genes were excluded as major regulatory factors in our analysis. Although both FPR2 and ELOVL6 are involved in lipid metabolism, the mechanism behind the FPR2‐induced reduction in ELOVL6 remains to be explored. It has been reported that myeloid FPR2 plays a critical role in diet‐induced obesity and its related complications by modulating energy expenditure [48]. Additionally, ELOVL6 activity controls the length and de‐saturation of fatty acids to reduce body weight [56]. FPR2 agonists, such as BMS‐986235, trigger a Gi‐protein coupled signaling cascade that inhibits forskolin‐stimulated cyclic adenosine monophosphate (cAMP) accumulation [25], one of the key signaling pathways in lipid metabolism [25, 57, 58], leading to the activation of transcription factors such as CREB, CREM, and ATF‐1. Furthermore, it has been reported that cAMP upregulates ELOVL6 in bovine embryos [57], suggesting that FPR2 could regulate ELOVL6 through the classic FPR2/Gi/cAMP axis.

We then explored how ELOVL6 influences VSMC de‐differentiation. Given that elevated ELOVL6 in VSMCs accelerates the development of advanced atherosclerotic lesions, neointima, and foam cell formation by amplifying the mitotic response, this effect is likely mediated through ROS production and AMPK/KLF4 signaling‐induced phenotypic switching of VSMCs [59, 60]. Although there is limited literature on the role of ELOVL6 in M2 macrophages, it has been reported that ELOVL6 deficiency induces a reparative phagocytic phenotype that promotes myelin regeneration and repair of damaged myelin [61]. Our results suggest that FPR2 agonism could help mitigate restenosis, at least in part, by influencing ELOVL6.

Given that ELOVL6 may mediate FPR2‐regulated neointimal hyperplasia, we sought to confirm that restenosis was associated with altered ELOVL6 expression in human tissue. As anticipated, ELOVL6 levels were elevated and negatively correlated with FPR2 expression in tissue obtained from patients experiencing restenosis, aligning with the inhibitory effect of FPR2 agonism on neointimal hyperplasia and the subsequent reduction in ELOVL6 expression. Our findings from human tissue closely paralleled those obtained in our experimental models. Increased ELOVL6 expression negatively correlated with VSMC contractile markers such as CNN1 and TAGLN. Consistent with our data, an association between ELOVL6 and VSMC contractile markers was confirmed in an in‐stent injury gene expression profile, a distinct cell model for restenosis. Using the previously published gene expression profile [46], we analyzed the reported gene expression profile to show that the VSMC contractile markers Cnn1 and MYH10 negatively correlate with Elovl6 levels. The consistency observed in the association analyses between our RNA‐seq and the independent RNA‐seq dataset reinforces the negative association between VSMC contractile markers and Elovl6. The alignment of our findings with the observation from human tissue highlights the influence exerted by ELOVL6, mediated through FPR2 agonism, across different species.

We then demonstrated that regulating ELOVL6 expression in vascular injury depends on FPR2. The FPR2 agonist BMS‐986235 reduced ELOVL6 in both M2 macrophages and VSMCs, and this effect was abolished when FPR2 was knocked down. Our findings indicate an association between FPR2 and ELOVL6, a novel discovery with significant implications for understanding the pathogenesis and prevention of restenosis. In addition to the partial regulation attributed to the changes in ELOVL6 expression induced by FPR2 agonism, further exploration of the signaling pathway responsible for the phenotypic switch in VSMCs is necessary. While our proof‐of‐concept data demonstrated the role of FPR2 in restenosis, additional research is required to evaluate the clinical efficacy of FPR2 agonists in patients.

In conclusion, our findings establish FPR2 as a promising and novel target for limiting restenosis [19]. The observed reduction in FPR2 expression in both human and murine models of restenosis, coupled with the efficacy of the selective FPR2‐agonist BMS‐986235 in preserving VSMC differentiation and mitigating neointimal hyperplasia, underscores the critical role of FPR2 in preventing vascular hyperplasia and supports the development of FPR2‐based pharmacotherapy for restenosis. Furthermore, our study identifies the FPR2/ELOVL6 axis as a key regulatory pathway in M2 macrophages and VSMCs, providing mechanistic insight into FPR2's protective effects. Future investigations into the molecular mechanisms governing FPR2‐mediated VSMC phenotype modulation and the clinical translation of pro‐resolving FPR2 agonists will be essential for advancing GPCR‐targeted therapies in vascular disease management.

Author Contributions

Cheng Xue Qin and Xiaojun Zhou conceived the research concept and design and manuscript writing. Qian Zhang and Xiaoting Wang provided contributions to the conduction of the experiments and contributed to writing the article. Mi Zhou provided access to human samples. Qian Zhang, Yuqin Zha, Peishen Zhao, Owen L. Woodman, Mi Zhou, Yuguo Chen, Xiaojun Zhou, and Cheng Xue Qin discussed its content and reviewed/edited it before submission. All authors read and approved the final manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: fsb271020‐sup‐0001‐Supinfo.docx.

Dataset S1: fsb271020‐sup‐0002‐DatasetS1.xlsx.

Zhang Q., Zha Y., Wang X., et al., “ FPR2 Agonism Attenuates Restenosis by Mitigating Neointimal Hyperplasia via ELOVL6 ,” The FASEB Journal 39, no. 17 (2025): e71020, 10.1096/fj.202501823R.

Data Availability Statement

The authors declare that the data supporting the findings are available within the paper. Any remaining data that support the results of the study will be available from the corresponding authors upon reasonable request.