Abstract

   In-stent restenosis (ISR), caused by aggressive vascular smooth muscle
   cell (VSMC) proliferation, is a serious complication of stenting.
   Therefore, developing therapeutic approaches that target VSMC
   inhibition is imperative. Our previous study showed that VSMC
   hyperplasia was attenuated after iron stent degradation, and VSMC
   proliferation around the stented section was arrested. The corrosion
   products of the iron stents were primarily Fe[3]O[4] particles.
   Therefore, we hypothesized that Fe[3]O[4] particles generated by iron
   stents would prevent neointimal hyperplasia by inhibiting VSMC
   proliferation. To test this hypothesis, culture assays and flow
   cytometry were performed to investigate the proliferation of VSMC.
   Global gene sequencing and Kyoto Encyclopedia of Genes and Genomes
   enrichment analyses were performed to investigate the underlying
   mechanisms. Fe[3]O[4]-coated stents were implanted into rabbit carotid
   arteries to evaluate the inhibitory effects of Fe[3]O[4] on neointimal
   hyperplasia. The major findings of the study were as follows: 1)
   Fe[3]O[4] attenuated neointimal hyperplasia by preventing VSMC
   proliferation after stenting; 2) Fe[3]O[4] exerted inhibitory effects
   on VSMCs by downregulating proliferative genes such as SOX9, EGR4, and
   TGFB1, but upregulated inhibitory genes such as DNMT1, TIMP3, and PCNA;
   3) Fe[3]O[4] inhibited VSMCs by preventing phenotypic transformation
   from the contractile to the synthetic phase; and 4) Fe[3]O[4]-coated
   stents achieved satisfactory hemocompatibility in a rabbit model. Our
   study highlights the additional benefits of Fe[3]O[4] particles in
   inhibiting VSMC proliferation, indicating that Fe[3]O[4] coated stent
   potentially served as an attractive therapeutic approach for ISR
   prevention.

   Keywords: Fe[3]O[4] particle, Neointimal hyperplasia, Vascular smooth
   muscle cell, Cell proliferation

Graphical abstract

   Graphic abstract: Fe[3]O[4] generated from iron stents prevents
   neointimal hyperplasia by inhibiting VSMC proliferation.

   [31]Image 1
   [32]Open in a new tab

1. Introduction

   Until recently, in-stent restenosis (ISR), caused by aggressive
   vascular smooth muscle cell (VSMC) proliferation, was considered a
   severe complication of endovascular interventional surgery
   [[33]1,[34]2]. In addition, ISR continues to be one of the most common
   reasons for stent failure, requiring repeat revascularization
   [[35]3,[36]4]. Therefore, suppressing neointimal formation by
   inhibiting VSMC proliferation may be a novel therapeutic approach for
   preventing ISR. Although drug-eluting stents (DES) are frequently used
   worldwide to prevent intimal hyperplasia, concerns remain regarding the
   occurrence of ISR in the chronic stage [[37]5]. Limitations such as
   permanent metallic platforms, persistent local inflammation, and
   restricted vasomotion are the major causes of the aggressive
   proliferation of VSMCs and the generation of neointimal hyperplasia
   after stenting [[38][5], [39][6], [40][7]]. To address these drawbacks,
   bioresorbable stents (BRS) have been designed to address the long-term
   risk of complications associated with DES. The main advantage of BRS is
   the use of bioresorbable stent materials, which could be gradually
   degraded in vivo [[41]8,[42]9]. In addition, these degradation products
   are nontoxic, are completely cleaved into small molecules, and are
   ultimately discharged into body fluids [[43]10]. Moreover, iron and its
   alloys display long-term biocompatibility, adequate mechanical support
   performance, and reasonable degradation rates [[44]11,[45]12].
   Consequently, iron stents are considered one of the best candidates for
   biodegradable metal stents, and have received particular attention in
   the past few decades. In addition, our previous study showed that
   neointimal hyperplasia is inhibited after iron stent degradation.
   Moreover, VSMC proliferation in the stented sections was attenuated
   [[46]13]. These findings reveal the inhibitory effects of iron stent
   degradation on neointimal hyperplasia and VSMC proliferation. The
   products degraded by the iron stents were primarily composed of
   Fe[3]O[4]. Therefore, the mechanism underlying the effects of Fe[3]O[4]
   particles on VSMC proliferation needs to be determined.

   Owing to their high thermodynamic stability and magnetic and catalytic
   properties, Fe[3]O[4] particles are considered one of the most
   promising biomaterials for various applications such as biomolecule
   detection, magnetic resonance imaging, enzyme-like biocatalysts, and
   targeted gene delivery [[47][14], [48][15], [49][16]]. In addition,
   multifunctional applications of organized Fe[3]O[4]-based particles
   have a considerable impact on cancer therapy. Xie et al. reported that
   Fe[3]O[4]-solamargine induces apoptosis in pancreatic cancer cells but
   inhibits cell metastasis. In addition, the malignant growth of
   pancreatic cancer cells is attenuated by Fe[3]O[4] [[50]15]. Shen et
   al. have reported that Fe3+ participates in the Fenton reaction, which
   promotes the generation of reactive oxygen species (ROS), thereby
   increasing cancer cell apoptosis [[51]17]. Wei et al. reported that
   Fe[3]O[4] could easily be assembled into cancer cells [[52]18].
   Moreover, Fe[3]O[4] has been widely used for biomedical, diagnostic,
   and therapeutic purposes, such as microbial infection therapy,
   hyperthermia, and dermatologic diseases [[53]19]. In-stent restenosis
   is considered the “Achille's heel” of PCI, and the aggressive
   proliferation of VSMCs is the key mechanism for the generation of ISR
   after stenting. Therefore, there is an urgent need for therapeutic
   approaches that target VSMCs. Herein, we investigated the mechanisms
   underlying the effects of Fe[3]O[4] on VSMC proliferation, both in vivo
   and in vitro. The obtained results indicated that Fe[3]O[4] generated
   from iron stents displayed inhibitory effects by preventing VSMC
   proliferation and, consequently, attenuating neointimal hyperplasia.
   Moreover, Fe[3]O[4] inhibited VSMCs by impeding their phenotypic
   transformation from a contractile to a synthetic phenotype. Based on
   these findings, a crucial role may be ascribed to Fe[3]O[4] in the
   development of neointimal hyperplasia, which could serve as an
   attractive therapeutic approach for preventing ISR after stenting.

2. Methods and materials

2.1. Cell culture

   The vascular smooth muscle cells were harvested from Sprague–Dawley
   rats using enzymatic dispersion. A detailed procedure for isolating and
   culturing VSMCs has been described previously [[54]10]. The third to
   fifth passages of VSMCs were used in subsequent experiments. VSMCs were
   cultured in Dulbecco's modified Eagle's medium (DMEM) containing
   glucose (Gibco) and 20 % fetal bovine serum (FBS; Marlborough, MA, USA,
   Hycone). In subsequent experiments, VSMCs were deprived of serum for
   24 h. The starved cells were then co-cultured with Fe[3]O[4] and 316 L
   stainless-steel particles. The concentrations of Fe[3]O[4] and 316 L
   stainless-steel particles were 20 mg/L. Concentration test were as
   following: 5 different concentrations were set before the assay
   conducted (5 mg/L, 10 mg/L, 20 mg/L, 40 mg/L and 80 mg/L). Based on
   results obtained from those concentration test, the optimal
   concentrations for the co-cultural experiments was 20 mg/L (if the
   concentration too high(e.g. 40 mg/L or 80 mg/L), cells tended to be
   deactivate and low viability. On the other hand, little differences
   were observed if the concentration is too low (e.g. 5 mg/L or 10 mg/L).
   In consistent with the concentration in-vitro, the concentration of
   Fe[3]O[4] particles in vivo experiment was also set as the 20 mg/L. To
   achieve this concentration, the electrohydrodynamic jetting, and
   air-brush spray coating technique was introduced to controlling the
   Fe[3]O[4] releasing from the stent surface, thus performed the
   long-term therapeutic effects. This study was approved by the Ethics
   Committee of Xiangya Hospital of Central South University and was
   conducted following the Guidelines for the Care and Use of Laboratory
   Animals published by the US National Institutes of Health.

2.2. Cell wound-healing assay

   A wound-healing assay was performed to measure the migratory activity
   of VSMCs in vitro. VSMCs were seeded at 1 × 10^5 cells per milliliter
   into 6-well plates and incubated in the medium. The growth medium
   solution was discarded and replaced with DMEM containing 0.4 % FBS
   after the VSMCs reached approximately 90 % confluence. The monolayer of
   the VSMCs was scraped off using a sterile pipette tip (200 μL) after
   12-h incubation. The resultant floating cells were removed by scraping
   them off using warm phosphate-buffered saline (PBS). The VSMCs were
   co-cultured with Fe[3]O[4] and 316 L stainless-steel particles. A
   phase-contrast microscope equipped with a 50 objective and digital
   camera was used to record the dynamic changes in the scratched areas at
   0 and 72 h after scraping. Track width was measured at these time
   points.

2.3. Immunofluorescence staining

   Differentially treated VSMCs were seeded onto coverslips (24-well
   plates) at a density of approximately 1 × 10^4. The cells were then
   washed two times with warm PBS and fixed using paraformaldehyde (4 %)
   for 15 min at room temperature. After fixation, the cells were washed
   with PBS two times and then permeabilized with 0.3 % Triton X-100 in
   PBS for 10 min. Bovine serum albumin (BSA, 5 %) in PBS was added for
   60 min at room temperature to block nonspecific sites on the fixed
   cells. 4′,6-diamidino-2-phenylindole (DAPI) was used to stain the cell
   nuclei (room temperature for 15 min). Stained cells were mounted using
   an anti-fluorescence quenching agent. An Olympus IX73 imaging system
   was used to visualize the stained cells. Image-Pro Plus software
   (version 6.0) was used to measure the mean fluorescence intensity after
   staining.

2.4. Cell-cycle analysis

   Flow-cytometric DNA analysis was performed to measure the distribution
   of VSMCs at different experimental stages. Briefly, cold PBS was used
   to wash treated VSMCs twice. Thereafter, the cells were centrifuged for
   5 min at 4 °C. The pellet was resuspended in propidium iodide (PI)
   solution (50 μg/mL) containing RNase A (0.1 mg/mL) at 4 °C. A
   FACSCalibur cytometer (BD Biosciences, San Jose, CA, USA) was used to
   analyze each sample. A minimum of 1 × 10^4 cells was counted in each
   sample. CellQuest software (BD Biosciences, San Jose, CA, USA) was used
   to calculate the percentage of cells in each cycle phase.

2.5. RNA isolation and real-time PCR

   The RNeasy Mini Kit (Qiagen) was used to collect total cellular RNA
   from the treated VSMCs, following the manufacturer's instructions. The
   collected RNA was subjected to reverse transcription using a TaqMan
   Reverse Transcription Kit (Applied Biosystems) according to the
   manufacturer's instructions. Real-time PCR amplifications were
   performed using the iQTM SYBR Green supermix (Bio-Rad). The comparative
   Ct method was used to obtain relative quantities of the collected
   mRNAs, which were then normalized to glyceraldehydes-3-phosphate
   dehydrogenase (GAPDH).

2.6. RNA extraction and library construction

   A TRIzol Total RNA Extraction Kit was used to isolate RNA from the
   samples. RNA quality was determined using spectrophotometry and agarose
   gel electrophoresis (0.8 %). The 260/280 absorbance ratio of 1.8–2.2
   was set to detect the RNA quality for library construction and
   sequencing. cDNA (APExBIO, Cat. No. K1159) was obtained using oligo-dT
   primers for transverse mRNA. cDNA products were purified using the
   AMPure XP system. Library fragments were amplified using PCR after
   library construction; fragment sizes ranged from 350 to 550 bp. The
   Agilent 2100 Bioanalyzer (Agilent, USA) was used to quality-assess the
   library. The Illumina NovaSeq 6000 sequencing platform (Paired end150)
   was used to sequence the library and generate raw reads.

2.7. RNA-sequence

   Following treatment with Fe[3]O[4]/S316L particles, VSMCs were
   harvested. Total RNA was extracted from each sample using TRIzol
   reagent (Invitrogen). A NanoPhotomer® spectrophotometer (IMPLEN, CA,
   USA) was used to check the purity of RNA. The RNA Nano 6000 Assay Kit
   of the Bioanalyzer 2100 system (Agilent Technologies) was used to
   assess RNA integrity. The Qubit® RNA Assay Kit in the Qubit®2.0
   Fluorometer (Life Technologies, CA, USA) was used to measure the RNA
   concentration. NEBNext® UltraTM RNA Library Prep Kit for Illumina®
   (NEB, USA) was introduced to generate the sequencing libraries
   according to the manufacturer's recommendations. Index codes were then
   added to each sample in the attribute sequences. The cBot Cluster
   Generation System equipped with the TruSeq PE Cluster Kit v3-cBot-HS
   (Illumina) was used to measure the clustering of index-coded samples
   following the manufacturer's instructions. Read numbers mapped to each
   gene were counted using Feature Counts v1.5.0-p3. The FPKM of each gene
   was calculated based on the read count and gene length mapped to the
   gene. The DESeq2 R package was used to identify the differentially
   expressed genes (DEGs) between the Fe[3]O[4-] and S316L
   particle-treated groups. The threshold for significant differential
   expression was determined by an absolute log2(fold change) > 1 and a
   corrected P-value of <0.05. DAVID [53] and GSEA [54] annotations were
   used to perform gene ontology (GO) and Kyoto Encyclopedia of Genes and
   Genomes (KEGG) pathway enrichment analyses. A threshold of padj <0.05
   was set for DEGs that were significantly enriched in the GO terms and
   KEGG pathway analyses.

2.8. RNA-seq data analysis

   TrimGaloreto was performed to filter the raw paired-end FASTQ reads.
   Following quantification of gene expression using StringTie and
   reference genome-guided transcriptome assembly, clean reads were
   aligned to the mm10/hg19 mouse/human genome. DEseq2 was used to
   identify the DEGs among those samples. Functional enrichment analysis
   of significant DEGs was performed using the Cluster Profiler. KEGG
   pathway categories and GO were used to identify the potential genes in
   the identified modules. Statistical significance was set at P < 0.05.

2.9. Stent device and animal surgical procedure

   Iron stents are manufactured from iron tubes by tube drawing, laser
   cutting, chemical etching, and electropolishing. Fe[3]O[4] was
   deposited onto the surface of the stent platform using an
   electroplating process. The Fe[3]O[4] coatings designed to be lasting
   over 3 months in stent surface. The stent was crimped onto
   rapid-exchange balloon catheters (F3.0 15 mm; F2.75, 15 mm). Ethylene
   oxide was used to sterilize the device after fabrication. All rabbits
   received a loading dose of 10 mg aspirin and 10 mg clopidogrel a day
   before stent implantation surgery. For anesthesia via intubation and
   mechanical ventilation, ketamine (15 mg/kg) and midazolam (3 mg/kg)
   were administered via intramuscular (IM) injection. Ten minutes after
   the IM injection, tracheal intubation was performed. Isoflurane (range
   0–3.6 %) in 100 % oxygen was delivered via the Jackson Rees
   modification of Ayre's T-piece non-rebreathing system. Carprofen
   (3 mg/kg) and dextrose saline (5 mL/kg per h) were intravenously
   administered (IV). The heart rate (HR), respiratory rate (RR), and
   arterial oxygen saturation of hemoglobin (SpO[2]) were monitored during
   surgery. The right femoral artery was surgically exposed after
   successful anesthesia, and a 4F guide catheter was introduced over a
   0.356 mm guidewire. The contrast agent used for arterial angiography
   was iohexol. The stent device system was delivered to and placed in the
   right carotid artery through the guidewire. The stent system was
   dilated using a balloon under suitable pressure for at least 10 s after
   deployment at the target vascular site. To maintain an activated
   coagulation time of >300 s during implantation surgery, heparin (200
   IU/kg) was continuously administered through the catheter during
   surgery. All rabbits were administered amoxicillin (1 g/day) for 5 d
   after surgery. To prevent thrombosis in the stented position, all
   rabbits received clopidogrel (10 mg) and aspirin (10 mg) daily for
   three months. Subsequently, aspirin (10 mg/day) was administered for
   three months. Carotid artery tissue was harvested three months after
   stent implantation. All animal experiments were approved by the
   Institutional Animal Ethics Committee of Xiangya Hospital (Animal
   Protocol Number: 2019103012).

2.10. The follow-up of carotid angiography

   All rabbits received 1000 U of heparin (intravenous injection) before
   carotid angiography. The carotid artery angiography images were
   captured using the CGO-2100 Cath-Lab system (Wandong, China). After the
   heparin injection, an angiographic catheter was inserted through the
   proximal right femoral artery. A contrast agent was then injected into
   the right femoral artery through an angiographic catheter to highlight
   the carotid artery. After the injection of the contrast agent, X-ray
   images of the same location were collected before and after the
   contrast agent injection using the CGO-2100 Cath-Lab system, and the
   images were subtracted to obtain the target artery.

2.11. Morphological investigation

   All rabbits were euthanized by administering a lethal dose of
   pentobarbital (intravenous injection). The carotid artery was carefully
   exposed, and the stented artery tissue was collected. To prepare for
   histopathological investigation, the harvested arterial tissue was
   placed in 3.5 % neutral buffered formalin. The stented arterial tissue
   was sectioned longitudinally, and the stented arterial lumen was
   dehydrated and embedded in resin. The samples were sectioned into
   slides, and Masson's trichrome and immunofluorescent staining were
   performed. The stent structure and neointimal tissue thickness were
   measured using a biological microscope (DM2500, Leica, Germany). The in
   vitro hemolysis assay in the present study evaluated hemoglobin release
   in the plasma (as an indicator of red blood cell lysis) following the
   Fe[3]O[4]-coated stent and S316L stents. The specimen used in this
   assay was anticoagulant-treated whole blood collected from healthy
   individuals. Triton X-100 was used as a hemolytic agent and added to
   the positive control.

2.12. Scanning transmission electron microscopy

   For scanning transmission electron microscopy (STEM) observations, a
   Leica EM UC6 ultramicrotome was used to prepare sliced polished
   cross-sections (approximately 50 nm thick). An FEI Titan G2 60–300
   Chemi STEM, equipped with a Super-X EDS with four windowless silicon
   drift detectors and a Cs probe corrector, was used to obtain
   high-resolution, high-angle annular dark-field scanning transmission
   electron microscopy (HAADF-STEM).

2.13. Statistical analysis

   Quantitative data are represented as means ± SEM from at least three
   independent assays. Before statistical analysis, the data were tested
   for normal distribution using the Shapiro–Wilk test. An unpaired
   two-sided Student's t-test was used to compare continuous and normally
   distributed variables. Non-normally distributed variables were compared
   using the Mann–Whitney U test. Statistical significance was set at
   P < 0.05.

3. Results

3.1. VSMC proliferation and viability were inhibited by Fe[3]O[4] particles
in vitro

   A co-culture medium containing Fe[3]O[4] was used to investigate the
   effect of Fe[3]O[4] on VSMC proliferation. As shown in [55]Fig. 1a–b,
   the number of BrdU-positive cells decreased in the Fe[3]O[4] particle
   group, indicating an inhibitory effect of Fe[3]O[4] on VSMCs. In
   addition to VSMC proliferation, migration plays an essential role in
   the development of neointimal hyperplasia. Therefore, scratch assays
   were performed to investigate VSMC migration after treatment with
   Fe[3]O[4]. The scratch closure areas and gap distances were measured at
   different time points. As shown in [56]Fig. 1c–e, as the culture time
   increased, the closure area in the control group narrowed, and the gap
   distance was reduced. However, the closure area in the
   Fe[3]O[4]-treated group was larger, with a wider gap distance,
   indicating the inhibition of migration by the Fe[3]O[4] particles. To
   investigate whether the effects of Fe[3]O[4] on VSMCs are
   concentration-dependent, we compared VSMC proliferation and viability
   at typical low (20 μM), medium (60 μM), and typical high (100 μM)
   concentrations. With the increase in the concentration of Fe[3]O[4],
   VSMC proliferation gradually declined ([57]Fig. 1f). Moreover, VSMC
   viability was attenuated with increasing Fe[3]O[4] concentration
   ([58]Fig. 1g). These results suggested that Fe[3]O[4] inhibits VSMC
   proliferation in a concentration-dependent manner. The pH value can
   change after co-culture with Fe[3]O[4], which influences the
   proliferation and viability of VSMCs. The pH of the co-culture medium
   was monitored over time. As shown in [59]Fig. 1h, although the
   Fe[3]O[4]concentration differed in each medium, no statistical
   differences in pH changes were detected in each medium or at different
   intervals.

Fig. 1.

   [60]Fig. 1
   [61]Open in a new tab

   VSMC proliferation was inhibited by Fe[3]O[4] in vitro. a-b) The
   proliferation of VSMCs was attenuated, indicating an inhibitory effect
   of Fe[3]O[4] on VSMCs. Representative images of BrdU-positive cells
   after Fe[3]O[4] particle pretreatment. BrdU (red) immunolabelling with
   DAPI staining (green) of VSMCs. The white arrow indicates the
   BrdU-positive cells; c-e) The closure area in the Fe[3]O[4]-treated
   group was larger and with a wider gap distance than that of the
   control; the red dashes denote the scratch wound closure area; f-g)
   increasing concentrations of Fe[3]O[4] led to inhibition of VSMC
   proliferation and viability; h) no statistically significant changes in
   pH were observed in the different intervals. Similar results were
   obtained from three independent experiments. *P < 0.05.

   VSMC: vascular smooth muscle cell; BrdU: bromodeoxyuridine; DAPI:
   4′,6-diamidino-2-phenylindole.

3.2. Differential gene expression analysis after exposure to Fe[3]O[4]
particles

   To explore the potential gene regulation affected by Fe[3]O[4], global
   VSMC genes were identified using transcriptome sequencing after
   pretreatment with Fe[3]O[4]. VSMCs co-cultured with S316L
   stainless-steel particles were used as controls. Total RNA was
   extracted, labeled, and hybridized to an oligonucleotide microarray.
   Adjusted p-values (false discovery rate (FDR) < 0.05) and fold change
   (FC) ratios (|log2FC| ≥1) were used to identify DEGs between the two
   groups. The per-base content of transcriptome sequencing is shown in
   [62]Fig. S1a. Raw reads were cleaned using Fastq, and the following
   analysis was based on clean reads; the percentage of clean reads was
   98.79 % ([63]Fig. S1b). The expression levels of transcripts per
   million during transcriptome sequencing are shown in [64]Fig. S1c. A
   total of 53,929 reference-based transcripts were assembled from the
   mapped data using the String Tie Software. After removing low-quality
   and redundant sequences, 53,391 transcripts were obtained,
   corresponding to 26,740 reference transcripts. Simultaneously, 26,075
   genes were assembled from mapped reads. In total, 11089 DEGs were
   identified at the transcriptomic level. Of these DEGs, 6863 were
   upregulated and 4226 were downregulated. The top 30 upregulated and
   downregulated DEGs are discussed as heat maps in [65]Fig. 2a–b. A
   volcano plot was used to show the overall distribution of the DEGs. In
   the volcano plots, the downregulated (blue dots) and upregulated genes
   (red dots) were sorted by statistical significance (p-value) ([66]Fig.
   2c).

Fig. 2.

   [67]Fig. 2
   [68]Open in a new tab

   Differentially expressed genes after pretreatment with Fe[3]O[4]
   particles. a-b) Top 30 upregulated and downregulated DEGs are
   represented as heatmaps; c) overall distribution of DEGs is displayed
   as a volcano plot, in which the upregulated (red) and downregulated
   (blue) genes are sorted by statistical significance; d) GO enrichment
   analysis showed that pathways such as smooth muscle cell
   differentiation, smooth muscle cell proliferation, and cell-cycle
   division were downregulated in the Fe[3]O[4]-pretreated group (n = 3);
   e) differentially expressed genes related to VSMC proliferation after
   Fe[3]O[4] treatment corresponded to eight downregulated genes, namely
   DNMT1, TIMP3, TIMP1, PCNA, FGF1, AIFM2, ARF5, and BID. Yet, SOX9, EGR4,
   TGFB1, VEGFA, and CCND1 were upregulated; f) PPI network analyses
   identified 211 interactions among the 75 genes within the DEGs from the
   comparison group. SOX9, TIMP1, CCND1, TGFB1, and EGR1 played essential
   roles in maintaining the tight connections of the whole network.

   DEG: differentially expressed gene; GO: Gene Ontology; VSMC: vascular
   smooth muscle cell; PPI: protein–protein interaction.

3.3. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway
enrichment analysis after pretreatment of VSMCs with Fe[3]O[4]

   To further confirm the potential functions of the target genes, GO and
   KEGG functional enrichment analyses were performed using the 11089 DEGs
   identified. GO enrichment analysis showed that pathways such as smooth
   muscle cell differentiation, smooth muscle cell proliferation, and
   cell-cycle division were downregulated in the Fe[3]O[4]-pretreated
   group (n = 3), suggesting inhibition of VSMC proliferation by Fe[3]O[4]
   particles ([69]Fig. 2d). Consistent with the GO enrichment analysis,
   the KEGG enrichment analysis showed that Fe[3]O[4] affected growth
   factor activity, DNA binding, and growth receptor binding ([70]Figs.
   S1d–e). Moreover, it was confirmed through GSEA that the pathways
   involved in VSMC apoptosis and cellular senescence were downregulated
   ([71]Fig. S1f). Among these pathways, VSMC proliferation and viability
   after Fe[3]O[4] treatment mainly involved eight downregulated genes:
   DNMT1, TIMP3, TIMP1, PCNA, FGF1, AIFM2, ARF5, and BID. However, SOX9,
   EGR4, TGFB1, VEGFA, and CCND1 were upregulated in this group ([72]Fig.
   2e). Protein–protein interaction (PPI) network analyses identified 211
   interactions among 75 genes within the DEGs in the comparison group
   ([73]Fig. 2f). Based on the PPI network analysis, SOX9, TIMP1, CCND1,
   TGFB1, and EGR1 play essential roles in maintaining the tight
   connections of the entire network. To validate the reproducibility and
   repeatability of the DEGs identified using transcriptome sequencing, we
   selected three genes involved in EC proliferation, PCNA, DNMT1, and
   TIMP3, for qRT-PCR analysis. The results showed that these genes were
   significantly differentially expressed and were consistently
   downregulated ([74]Fig. 3a). In contrast, SOX9, EGR4, and TGFB1 were
   upregulated in VSMCs after exposure to Fe[3]O[4] ([75]Fig. 3b).

Fig. 3.

   [76]Fig. 3
   [77]Open in a new tab

   Fe[3]O[4] degraded from the stent inhibits VSMC proliferation. a-b)
   RT-qPCR analysis showed that PCNA, DNMT1, and TIMP3 were significantly
   downregulated in the Fe[3]O[4]-pretreated group. In contrast, SOX9,
   EGR4, and TGFB1 were upregulated; c-d) fewer VSMCs were observed around
   the Fe[3]O[4]-coated stent surface. In contrast, the surface of the
   S316L stent was almost surrounded by VSMCs; e-f) gene expression
   involved in VSMC proliferation, including DNMT1, TIMP3, and PCNA, was
   downregulated in the Fe[3]O[4]-coated stent group. Conversely,
   representative gene expression of proliferative inhibition markers
   (SOX9, EGR4, and TGFB1) was upregulated; g) compared with the viability
   of VSMCs attached to the S316L stent, VSMC viability around the
   Fe[3]O[4]-coated stent was significantly attenuated. Similar results
   were obtained from three independent experiments. *P < 0.05,
   **P < 0.01.

   VSMC: vascular smooth muscle cell.

3.4. Fe[3]O[4] degraded from the stent inhibited VSMC proliferation

   Fe[3]O[4] is generated on the surface of the iron stent after
   implantation. To investigate the effects of Fe[3]O[4] degradation by
   iron stents on VSMCs, a stent coated with Fe[3]O[4] particles
   (Fe[3]O[4]-coated stent) was introduced and co-cultured with VSMCs. An
   S316L stent without a Fe[3]O[4] coating was used as a control. As the
   co-culture time increased, numerous Fe[3]O[4] particles were generated
   on the Fe[3]O[4]-coated stent surface. The morphology of the stent
   surface became rough and uneven, with extensive corrosion pits
   generated around the stent. However, the S316L stent surface was smooth
   and intact. Consistent with the Fe[3]O[4] generation, the proliferation
   rate of VSMCs around the stent was arrested. As shown in [78]Fig. 3c–d,
   fewer VSMCs were observed around the stent surface. By contrast, the
   surface of the S316L stent was surrounded by VSMCs. In addition, the
   expression of genes involved in VSMC proliferation, such as DNMT1,
   TIMP3, and PCNA, was downregulated, indicating growth arrest caused by
   Fe[3]O[4] particles in vitro ([79]Fig. 3e). In contrast to
   proliferation inhibition, the representative gene expression of
   proliferative inhibition markers (SOX9, EFR4, and TGFB1) was
   upregulated in the VSMCs harvested from the iron stent group ([80]Fig.
   3f). Numerous studies have confirmed that OPN, MMP-9, and epiregulin
   (EREG) are highly expressed in contractile VSMCs. To confirm the
   results of the RNA-seq assay, markers of the contractile phenotype,
   including OPN, MMP-9, and EREG, were measured. Consistent with previous
   findings, these markers were upregulated in the Fe[3]O[4]-treated
   group; the results are described in [81]Figs. S2d–f. In addition to
   VSMC proliferation, VSMC viability was measured using an MTT assay. As
   shown in [82]Fig. 3g, compared with the viability of VSMCs attached to
   the S316L stent, VSMC viability around the Fe[3]O[4]-coated stent was
   significantly attenuated, suggesting the inhibition of viability after
   the generation of Fe[3]O[4] particles.

3.5. Fe[3]O[4] prevented VSMC differentiation from the contractile to the
synthetic phase

   Differentiation of VSMCs from a physiologically contractile phenotype
   to an adverse proliferative or synthetic state is one of the major
   events leading to restenosis after stenting. As the proliferation of
   VSMCs was inhibited by Fe[3]O[4], we investigated whether Fe[3]O[4]
   inhibited VSMC differentiation by preventing the switch from a
   contractile to a proliferative phenotype. To address this issue, a
   co-culture medium was used, and VSMCs were stimulated with LPS. The
   concentration of LPS used was 100 ng/mL (Sigma-Aldrich L5293-2 ML,
   Burlington, MA, USA). Phenotypic transformation after treatment with
   Fe[3]O[4] was characterized by measuring the mRNA expression of a
   repertoire of contractile marker genes, including Myh11, α-SMA, and
   SM22α. As shown in the RT-qPCR results in [83]Fig. 4a, the RNA
   transcripts of these representative contractile markers significantly
   increased after treatment with Fe[3]O[4]. However, gene expression
   involved in VSMC proliferation, such as TGFB1, PDGF, and FGF-2, was
   downregulated ([84]Fig. 4b). As α-SMA is considered one of the specific
   biomarkers of the contractile phenotype for VSMCs, its expression was
   measured both using IF staining and WB assay. As presented in [85]Figs.
   S2a–d, the results obtained from IF staining and WB were consistent
   with the findings obtained from RT-qPCR, indicating the upregulation of
   the contractile phenotype of VSMCs after exposure to Fe[3]O[4.] To
   confirm the effect of Fe[3]O[4] on VSMC proliferation and
   differentiation, deferoxamine-conjugated nanoparticles (DFO-NPs) were
   added to eliminate the influence of Fe[3]O[4] on VSMCs. As expected,
   the proliferation of VSMCs increased after the administration of
   DFO-NPs ([86]Fig. 4c). In addition, the viability of VSMC was measured
   using MTT assays. Compared with the cell viability seen in the
   Fe[3]O[4]-pretreated group, [87]Fig. 4d shows the promotion of cell
   viability after treatment with DFO-NPs. Moreover, the contractile
   marker genes, including Myh11, α-SMA, and SM22α, were downregulated
   following exposure to DFO-NPs ([88]Fig. 4e). However, the expression of
   genes involved in the synthetic phase transformation, such as TGFB1 and
   PDGF, was upregulated after the administration of DFO-NPs ([89]Fig.
   4f). These results indicated the inhibitory effect of Fe[3]O[4] on VSMC
   differentiation into a proliferative phenotype.

Fig. 4.

   [90]Fig. 4
   [91]Open in a new tab

   Fe[3]O[4] prevents VSMC differentiation from the contractile to the
   synthetic phase. a) the representative contractile markers Myh11,
   α-SMA, and SM22α were significantly upregulated after treatment with
   Fe[3]O[4]; b) genes involved in VSMC proliferation, such as TGFB1,
   PDGF, and FGF-2, were downregulated in the Fe[3]O[4]-pretreated group;
   c) VSMC proliferation increased after administration of DFO-NPs; d)
   cell viability was improved after treatment with DFO-NPs; e-f)
   contractile marker genes were downregulated but genes involved in the
   synthetic phase transformation were upregulated after administration of
   DFO-NPs. Similar results were obtained from three independent
   experiments. *P < 0.05, **P < 0.01.

   VSMC: vascular smooth muscle cell; DFO-NP: deferoxamine-conjugated
   nanoparticles.

3.6. Fe[3]O[4]-coated stents inhibited VSMC hyperplasia by impeding
proliferation

   To further determine the effect of Fe[3]O[4] on VSMCs in vivo, a
   Fe[3]O[4]-coated stent was deployed into the carotid arteries of
   rabbits via transfemoral interventional surgery. Eight male rabbits
   (mean weight: 4.8 kg ± 0.4 kg) were included and implanted with the
   Fe[3]O[4]-coated stent. Accordingly, another seven male rabbits (mean
   weight 4.6 kg ± 0.3 kg) that received S316L stent implantation were set
   as the control. The intimal layer of the stented arterial section was
   removed using a scraper. Therefore, the Fe[3]O[4]-coated stent was in
   contact with the medium layer (mainly consisting of VSMCs), and the
   VSMCs in the medium layer were stimulated after scraping the intimal
   tissue layer ([92]Fig. 5a and b). All the rabbits remained healthy
   until the final follow-up period. Optical coherence tomography (OCT)
   detection indicated that the struts of the Fe[3]O[4]-coated stent were
   completely embedded in the medial tissue layer ([93]Fig. 5c).
   Meanwhile, as the implantation time increased, the Fe[3]O[4] particles
   degraded from the stent surface, and dark black particles infiltrated
   extensively into the stented media layer tissue ([94]Fig. 5d–e). Upon
   exposure to Fe[3]O[4], the thickness of the neointimal layer in the
   Fe[3]O[4]-coated stent group was lower than that in the S316L stent
   group ([95]Fig. 5f–g), indicating that the inhibition of neointimal
   hyperplasia occurred after implantation of the Fe[3]O[4]-coated stent.
   In addition, the thicknesses of the intima, media, and lumen were
   measured. No differences were identified in the thicknesses of the
   intima and media layers. However, the lumen area was larger in the
   Fe[3]O[4]-coated stent group ([96]Figs. S2g–i). To further evaluate the
   proliferation profile of VSMCs, the medium-layer tissue was harvested,
   and the VSMC cell cycle was qualitatively assessed using flow
   cytometry. Consistent with the HE staining findings, cell-cycle
   measurements showed that the percentage of VSMCs in the S phase was
   reduced in the Fe[3]O[4]-coated stent group ([97]Fig. 6a–b). To
   identify the phenotypic switch of VSMCs after Fe[3]O[4]-coated stent
   deployment, the level of α-SMA, a representative contractile marker,
   was measured further. As shown in [98]Fig. 6c–d, compared with the
   α-SMA expression level in the S316L stent group, the expression of
   α-SMA was increased in the Fe[3]O[4]-coated stent group. In addition,
   the relative mRNA expression of SM22α was upregulated after receiving
   the Fe[3]O[4]-coated stent implantation ([99]Fig. 6e).

Fig. 5.

   [100]Fig. 5
   [101]Open in a new tab

   Fe[3]O[4]-coated stent inhibits neointimal hyperplasia in vivo. a-b) An
   Fe[3]O[4]-coated stent was deployed in the carotid artery via
   transfemoral interventional surgery. Red arrows indicate the scraper
   device used for removing the intima layer of the stented artery
   section, and the yellow arrow indicates the position of the stent
   implantation after the balloon catheter was discharged; c) OCT revealed
   that the struts of the Fe[3]O[4]-coated stent were embedded in the
   media layer tissue; d-e) The carotid artery tissue harvested after the
   stent was implanted for three months. Fe[3]O[4] particles were degraded
   from the stent surface, and dark black particles infiltrated into the
   stented media layer tissue; f-g) with exposure to Fe[3]O[4], neointimal
   layer thickness in the Fe[3]O[4]-coated stent group was decreased
   compared with that seen in the S316L stent group. Yellow dashed lines
   demarcate the neointimal layer after stenting. The results were
   obtained from three independent measurements at different sites per
   sample. *P < 0.05.

   OCT: optical coherence tomography.

Fig. 6.

   [102]Fig. 6
   [103]Open in a new tab

   Fe[3]O[4]-coated stent inhibits VSMC hyperplasia by preventing VSMC
   proliferation. a-b) cell-cycle measurement indicated that the
   percentage of VSMCs in the S phase was decreased in the
   Fe[3]O[4]-coated stent group; c-d) compared with α-SMA expression in
   the S316L stent group, the expression of α-SMA in the media layer
   increased after receiving the Fe[3]O[4]-coated stent implantation; e)
   the relative mRNA expression of SM22α was upregulated after
   Fe[3]O[4]-coated stent implantation. *P < 0.05.

   VSMC: vascular smooth muscle cell.

3.7. Fe[3]O[4]-coated stents displayed satisfactory hemocompatibility

   The Fe[3]O[4] coated on the stent surface inhibited VSMC proliferation,
   revealing its promising clinical application. Therefore, the
   biocompatibility of Fe[3]O[4]-coated stents was further evaluated in
   vivo. The concentration of iron ions can be increased by the release of
   Fe[3]O[4] particles into the blood, which may cause heavy metal
   toxicity in vital organs. Therefore, serum iron ions were monitored
   over time. As shown in [104]Fig. S3a, the concentration of soluble iron
   ions was higher after the implantation of the Fe[3]O[4]-coated stent.
   However, the iron ion levels remained normal (ISO standard). The
   hemolysis rate was measured, and the results are shown in [105]Figs.
   S3b and c. The supernatants exhibited a negligible hemolysis effect in
   both the Fe[3]O[4]-coated and S316L stents, indicating that the
   Fe[3]O[4]-coated stent did not cause severe hemolysis after
   implantation. PT assays were performed to further examine the
   activation of coagulation factors after Fe[3]O[4]-coated stent
   implantation. As shown in [106]Fig. S3d, the partial thromboplastin
   time in the Fe[3]O[4]-coated stent group was higher than that in the
   original blood plasma, suggesting a low activation of coagulation by
   the Fe[3]O[4]-coated stent. However, the values returned to normal
   after heparin administration. To evaluate the cytotoxic effects of
   Fe[3]O[4] on vital organs, pathological staining of the kidneys, lungs,
   and spleen was performed after the implantation of the Fe[3]O[4]-coated
   stent. [107]Fig. S3e shows no histomorphological abnormalities in the
   vital organs.

4. Discussion

   Although DES is frequently used worldwide, permanent metallic
   platforms, persistent inflammatory reactions, and late stent neointimal
   hyperplasia have limited its clinical application in revascularization
   [[108]5,[109]6]. In-stent restenosis, caused by VSMC proliferation
   after stent implantation, remains a severe complication of endovascular
   interventional surgery [[110]5,[111]20,[112]21]. Therefore,
   understanding the proliferation, migration, viability, and phenotypic
   modulation of VSMCs after stenting may reveal the mechanisms critical
   for discovering novel therapeutic candidates. Our previous findings
   showed that neointimal hyperplasia is attenuated after iron stent
   degradation. Moreover, VSMC proliferation around the stented area is
   inhibited by iron stent degradation [[113]13]. These observations
   indicate the inhibitory effects of iron stent degradation on VSMCs. The
   degradation product released from the iron stent is primarily
   Fe[3]O[4]. Therefore, we determined the mechanism underlying the effect
   of Fe[3]O[4] on VSMC proliferation, both in vivo and in vitro.

   Fe[3]O[4] particles have received increasing attention in the past few
   decades and are considered one of the most promising biomaterials for
   various applications, such as targeted delivery of genes, molecular
   imaging, stem cell tracking, enzyme-like biocatalysts, and tissue
   repair [[114]22,[115]23]. Moreover, owing to their effective
   biocompatibility and unique magnetic behavior, Fe[3]O[4] particles can
   be conjugated with targeting moieties, such as nucleic acids,
   antibodies, enzymes, and cell lines [[116]16,[117]19,[118][24],
   [119][25], [120][26]]. The multifunctional application of organized
   Fe[3]O[4]-based particles has had a substantial influence on cancer
   therapy; Fe[3]O[4] can easily be assembled into cancer cells. In
   addition, owing to their cancer-targeting properties, Fe[3]O[4]
   particles have potential applications in cancer diagnosis and therapy.
   Shen et al. reported that Fe^3+ participates in the Fenton reaction,
   which generates ROS, consequently inducing cancer cell apoptosis
   [[121]15,[122]17,[123]27,[124]28]. In addition to preventing cancer
   cell proliferation, Fe[3]O[4] is widely used for biomedical,
   diagnostic, and therapeutic purposes. Owing to these attractive
   advantages, Fe[3]O[4] particles are considered one of the most
   promising candidates for biomedical and clinical applications.
   Consistent with the previous findings, we investigated the role of
   Fe[3]O[4] particles generated from iron stents in VSMCs. Through in
   vitro co-culture experiments, we observed that Fe[3]O[4] prevented the
   proliferation of VSMCs. Moreover, by comparing VSMC proliferation and
   viability at typical low (20 μM), medium (60 μM), and typical high
   (100 μM) concentrations, we observed that Fe[3]O[4] exposure reduced
   VSMC proliferation and viability in a concentration-dependent manner.
   These findings revealed the beneficial application of VSMC-targeted
   approaches to preventing ISR.

   TIMP3 levels dramatically decrease in various cardiovascular diseases.
   As replenishment ameliorates cardiovascular diseases, TIMP3 is a
   promising therapeutic agent for cardiovascular diseases such as
   neointimal hyperplasia after stenting [[125]29]. In addition to playing
   an important role in cardiovascular disease, TIMP3 exerts an
   anti-angiogenic effect by directly interacting with the VEGF receptor-2
   to inhibit endothelial cell proliferation and tube formation. TIMP3
   induces apoptosis in VSMCs and certain cancer cells [[126][30],
   [127][31], [128][32]]. These findings indicated the critical role of
   TIMP3 in VSMC inhibition and its application as a therapy for
   cardiovascular diseases. This is consistent with the previous findings.
   By examining cell cycle-associated genes using RNA-seq, we observed
   that TIMP3 expression in VSMCs decreased dramatically after exposure to
   Fe[3]O[4], revealing the inhibitory effects of Fe[3]O[4] on VSMC
   proliferation. Our findings will be valuable in compiling advances and
   knowledge regarding TIMP3 biological function and will form the basis
   for further investigations to verify its clinical application in
   inhibiting VSMC proliferation after stenting.

   The medial layer of the arterial wall is mainly composed of VSMCs.
   Unlike skeletal muscle cells, VSMCs exhibit remarkable sensing,
   plasticity, and adapting properties [[129]33]. ISR is mainly determined
   by aggressive VSMC proliferation and therefore accelerates neointimal
   generation around the stent area. Despite advancements in DES
   technology, ISR and the requirement for repeat revascularization
   continue to be the most common reasons for stent failure, occurring at
   a rate of 1–2% per year on modern DES platforms [[130]5]. Among the
   pathogenic mechanisms underlying ISR, the increased motility and
   phenotypic transformation of VSMCs are the main causes of neointimal
   hyperplasia following stenting. VSMCs displayed low proliferation but
   high contractility under physiological conditions [[131]34,[132]35].
   Because of their contractility, VSMCs perform the primary functions of
   arterial dilatation and contraction, therefore playing an essential
   role in regulating blood pressure and flow. However, VSMCs are not
   terminally differentiated cells but display the capacity to
   differentiate into inflammatory, synthetic, or macrophage-like
   phenotypes upon stimulation [[133]36,[134]37]. The synthetic phenotype
   of VSMCs is characterized by increased motility, loss of contractility,
   and a high proliferation rate. We have previously shown that neointimal
   hyperplasia is attenuated after iron stent degradation. In contrast,
   previous studies have shown that the proliferation and viability of
   VSMCs are inhibited by Fe[3]O[4] particle exposure in vitro. Therefore,
   the physiological phenotype transformation was investigated, and the
   contractile properties of VSMCs were characterized by measuring the
   mRNA expression of a repertoire of contractile marker genes. As
   expected and consistent with the growth arrest of VSMCs, the mRNA
   levels in these representative contractile biomarkers, such as α-SMA,
   SM-22α, and Myh11, were decreased after treatment with Fe[3]O[4].
   Moreover, the extent of the reduction in expression was dependent on
   Fe[3]O[4] concentration. These findings indicate that Fe[3]O[4] may
   inhibit VSMC differentiation by preventing the phenotypic switch from
   the contractile to the proliferative phases. Collectively, these
   results support the inhibitory effects of Fe[3]O[4] on the phenotypic
   transformation of VSMCs from contractile to synthetic, suggesting a
   possible solution for preventing vascular remodeling and restenosis by
   targeting VSMCs after stenting. Moreover, these findings highlight the
   additional advantages of iron stents over other types of stents (such
   as DES and other biodegradable stents), potentially rendering them an
   efficient approach in clinical applications. However, the effects of
   Fe[3]O[4] on ECM production, inflammatory responses, and endothelial
   activation have not been determined owing to its in vitro experimental
   design. Hence, additional in vivo studies are required to further
   identify these factors.

   VSMC proliferation can be influenced by local hemodynamic factors, such
   as blood flow velocity and wall shear stress. In contrast, VSMCs
   undergo complex phenotypic changes in response to environmental stimuli
   in vivo. Therefore, the findings of the in vitro studies need to be
   confirmed in animal models. To address these concerns, a
   Fe[3]O[4]-coated stent was manufactured and deployed into the carotid
   arteries of rabbits via transfemoral interventional surgery. An S316L
   stent was used as the control. The intimal layer of the stented artery
   was removed; as a result, the Fe[3]O[4]-coated stent could come into
   direct contact with the VSMCs after deployment. VSMCs in the medium
   layer were stimulated by scraping the intimal tissue layer. Based on
   this animal model, as the implantation time increased, the Fe[3]O[4]
   particles degraded from the stent surface, and the dark yellow granules
   extensively infiltrated the stented media layer tissue. Upon exposure
   to Fe[3]O[4], the thickness of the medium layer in the Fe[3]O[4]-coated
   stent group was lower than that in the S316L stent group. In addition,
   the inhibition of the VSMC cycle in the Fe[3]O[4]-coated stent group
   was validated using flow cytometry. We observed a phenotypic switch in
   VSMCs after Fe[3]O[4]-coated stent deployment. Our results showed that
   the expression of α-SMA, a representative contractile marker, was
   significantly attenuated in the intima layer after receiving the
   Fe[3]O[4]-coated stent implantation. Beyond this, the expression levels
   of contractile markers, including α-SMA, SM-22α, and Myh11, were
   decreased in the Fe[3]O[4]-coated stent group. However, to obtain a
   good capacity for surgical trauma, only male rabbits were included in
   the experiment, which may bring the statistical bias to the results.
   Consistent with the in vitro results, data from the in vivo experiments
   provide valid evidence that Fe[3]O[4] reduces neointimal hyperplasia by
   preventing aggressive proliferation and phenotypic switching in VSMCs.
   Our findings are consistent with the diverse biomedical applications of
   Fe[3]O[4] particles proposed in a previous study, highlighting their
   advantages in the fields of medicine and healthcare. This highlights
   the beneficial effects of Fe[3]O[4] in preventing aggressive
   proliferation of VSMCs, indicating that Fe[3]O[4] particles generated
   by iron stents could potentially serve as an attractive therapeutic
   approach for preventing ISR. However, the detailed underlying
   mechanisms remain unclear and require further investigation.

5. Conclusions

   Neointimal hyperplasia caused by the aggressive proliferation of VSMCs
   after stenting is considered the “Achille's heel” of PCI and is the
   main cause of ISR. Therefore, suppression of neointima formation by
   impeding VSMC proliferation may be a novel therapeutic approach for
   preventing ISR. This study indicates that Fe[3]O[4] generated from iron
   stents inhibits VSMC proliferation and thus attenuates neointimal
   hyperplasia. Moreover, Fe[3]O[4] inhibited VSMCs by impeding their
   phenotypic transformation from a contractile to a synthetic phenotype.
   Consistent with previous findings, our study highlights the beneficial
   effects of iron stents in inhibiting VSMC proliferation, indicating
   that the Fe[3]O[4] particles generated by iron stents may serve as an
   attractive therapeutic approach for preventing ISR.

Ethics approval statement

   All experimental procedures were approved by the Institutional Animal
   Ethics Committee of.

   Xiangya Hospital.

CRediT authorship contribution statement

   Yalan Deng: Writing – review & editing, Project administration,
   Conceptualization. Jiabing Huang: Visualization, Formal analysis,
   Writing – original draft, Methodology, Investigation. Changqing Chen:
   Resources, Data curation. Yanbing Wen: Validation, Supervision. Dongxu
   Qiu: Writing – review & editing, Project administration, Funding
   acquisition, Conceptualization.

Declaration of competing interest

   The authors declare that they have no known competing financial
   interests or personal relationships that could have appeared to
   influence the work reported in this paper.

Acknowledgements