Inhibition of lysine‐specific demethylase 1A suppresses neointimal hyperplasia by targeting bone morphogenetic protein 2 and mediating vascular smooth muscle cell phenotype

Abstract Objectives Vascular disorders are associated with phenotypical switching of vascular smooth muscle cells (VSMCs). We investigated the effect of bone morphogenetic protein (BMP)‐2 in controlling VSMC phenotype and vascular disorder progression. Lysine (K)‐specific demethylase 1A (KDM1A) has been identified to target BMP‐2 and is employed as a therapeutic means of regulating BMP‐2 expression in VSMCs. Materials and methods VSMCs were stimulated with angiotensin II, and the expression of KDM1A and BMP‐2 was detected. VSMC proliferation, apoptosis, and phenotype were evaluated. An in vivo aortic injury model was established, and VSMC behaviour was evaluated by the expression of key markers. The activation of BMP‐2–associated signalling pathways was examined. Results We confirmed the inhibitory effect of KDM1A on BMP‐2 activity and demonstrated that KDM1A inhibition prevented VSMC transformation from a contractile to synthetic phenotype. In angiotensin II‐treated VSMCs, KDM1A inhibition triggered a decrease in cell proliferation and inflammatory response. In vivo, KDM1A inhibition alleviated post‐surgery neointimal formation and collagen deposition, preventing VSMCs from switching into a synthetic phenotype and suppressing disease onset. These processes were mediated by BMP‐2 through canonical small mothers against decapentaplegic signalling, which was associated with the activation of BMP receptors 1A and 1B. Conclusions The regulatory correlation between KDM1A and BMP‐2 offers insights into vascular remodelling and VSMC phenotypic modulation. The reported findings contribute to the development of innovative strategies against vascular disorders.


| INTRODUC TI ON
Neointimal hyperplasia is a pathological process that often occurs after surgical intervention in vascular diseases such as pulmonary arterial hypertension and is the main contributor to post-intervention restenosis. 1 It is characterized by the excess proliferation and migration of vascular smooth muscle cells (VSMCs), promoting the development of cardiovascular diseases with cytopathological manifestations such as vascular stenosis and calcification. [2][3][4] VSMCs exist in and can interconvert between two phenotypic states, namely the contractile (differentiation) and synthetic (dedifferentiation) states. 4,5 The transformation from a contractile state to a synthetic one leads to enhanced VSMC proliferation and migration, secretion and synthesis of extracellular matrices, and formation of neointimal membranes, which is the key step in the initiation of severe vascular proliferative diseases. 6,7 Precise control of the phenotypic transformation of VSMCs is thus an important step in preventing the occurrence and development of the above-mentioned diseases at an early stage.
The phenotypic transformation of VSMCs is mainly induced by growth factors (eg, angiotensin II), mechanical stimulation, and molecular signalling. The stimulation signal is transmitted to the nucleus and eventually regulates the expression of smooth muscle cell differentiation marker genes. [8][9][10] Among the members of the bone morphogenetic protein (BMP) multifunctional cytokine family, BMP-2 is a well-known osteogenic factor 11,12 that also plays critical roles in embryonic development, 13,14 nerve growth, 15,16 and cell behaviour regulation. 17 BMP-2 signalling can occur via either canonical or non-canonical routes. Canonical BMP-2 signalling is closely linked to downstream small mothers against decapentaplegic (SMAD) signalling, involving BMP-transduced phosphorylation of receptor-regulated (R)-SMADs 1, 5, and 8 via the activation of BMP-2 receptors (BMPRs). 18 On the other hand, non-canonical BMP signalling triggers non-SMAD activity including mitogen-activated protein kinase and phosphoinositide 3 kinase pathways. 18,19 In tumour-related studies, BMP-2 has demonstrated an inhibitory effect on colon cancer cell growth and proliferation while promoting apoptosis. 20,21 In addition, BMP-2 signalling is involved in the maintenance of the contractile phenotype and has been shown to inhibit VSMC proliferation and neointimal hyperplasia. 22,23 Whether BMP-2 signalling is mediated by canonical or non-canonical routes in vascular remodelling remains to be investigated.
We hypothesized that BMP-2 signalling is directly or indirectly involved in the phenotypic transformation of VSMCs and that the occurrence of vascular diseases may be prevented by regulating BMP-2 expression. Transcriptome sequencing has identified lysinespecific histone demethylase 1 (or lysine (K)-specific demethylase 1A, KDM1A) as a compound that potentially targets BMP-2, 24 suggesting that its function in cardiovascular health may be non-negligible. Studies on KDM1A have revealed that histone methylation occurs through the interaction of histone methyltransferase and demethylase, dynamically regulating biological processes such as the activation and inhibition of gene transcription. 25 Presently, research on KDM1A has focused on tumours, and it has been found to be highly expressed in tumour tissues and cells, participating in multiple signalling pathways that affect tumorigenesis, tumour invasion and metastasis, and drug resistance. 26 For example, high levels of KDM1A are often closely associated with the clinical progression of highly aggressive medulloblastoma, whereas inhibition of KDM1A using drugs reduced the tumorigenicity of human neuroblastoma cell lines in nude mice. 24 Yet, the involvement of KDM1A in the phenotypic transformation of VSMCs, the occurrence of neointimal hyperplasia and the development of vascular diseases have not been investigated.
Herein, we induced phenotypic switching of VSMCs using angiotensin II and investigated the resulting changes in KDM1A and BMP-2 expression. Cell behaviour including proliferation, migration, and apoptosis and the secretion of a variety of related factors were examined to validate the conversion of VSMCs from a contractile to a synthetic state. A rat model of aortic endothelial balloon injury was established to further investigate the relationship between KDM1A and BMP-2 at the in vivo level and to elucidate the possible involvement of BMPRs and SMADs in the phenotypic remodelling of VSMCs. The findings of our study aim to clarify the mechanism of KDM1A in regulating the biological behaviour of VSMCs and neointimal formation, revealing new targets and insights for the prevention and treatment of vascular proliferative diseases.

| MATERIAL S AND ME THODS
Detailed Materials and Methods can be found in the online Materials S1.

| VSMC isolation, culture and treatment
Vascular smooth muscle cells were extracted from the thoracic and abdominal aortic tissues of Sprague Dawley rats following standard procedures for isolating primary cells. Cells were cultured with smooth muscle culture medium until they reached 80%-90% confluence. Morphological characterization via bright-field and fluorescence microscopy was carried out to confirm the successful extraction of VSMCs, which typically show peak-valley structures.
To induce phenotypic switching of VSMCs, the cells were treated with angiotensin II (Ang-II) with or without pre-treatment using ORY-1001, a specific inhibitor of KDM1A (denoted as KDM-inh).
Cells that were treated with neither Ang-II nor KDM-inh were used as controls.

| Evaluation of VSMC proliferation and apoptosis
Vascular smooth muscle cells were treated as described previously and subjected to characterization of proliferation and apoptosis. The viability of Ang-II-induced VSMCs was first assessed using a 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay based on the instructions provided in the assay kit. Then, VSMC proliferation was evaluated using a scratch assay. Images of scratches in the cell monolayer were acquired immediately or 24 hours after scratching, and the wound-healing rate was calculated using ImageJ.
EdU staining was also carried out to visualize VSMC proliferation following the instructions provided in the EdU staining kit. The percentage of EdU-positive cells was calculated using ImageJ. To examine the migratory ability of treated or non-treated VSMCs, a Transwell migration assay was performed following the manufacturer's instructions. Cell cycle progression and apoptosis were analysed using flow cytometry.

| Identification of VSMC phenotype
The phenotype of VSMCs was identified by immunofluorescence staining of α-smooth muscle actin (α-SMA) and osteopontin (OPN), which are markers of the contractile and synthetic phenotypes, respectively. After treatment as described previously, the VSMCs were fixed, permeabilized, blocked, and incubated with respective primary antibodies overnight at 4°C. After incubation with the corresponding secondary antibodies for 1 hour at 37°C, the cells were observed using a fluorescence microscope. The production of growth factors and inflammatory cytokines, which also function as markers of VSMC phenotype, was assessed using enzyme-linked immunosorbent assay kits following the manufacturer's instructions.

| mRNA and protein expression of key markers
Quantitative reverse transcription-polymerase chain reaction was performed to detect the mRNA expression of KDM1A, and BMP-2 after VSMCs was treated as described previously, using GAPDH as a housekeeping gene. The primer sequences were

| Histological and immunohistochemical evaluation of injured tissues
The collected aortic tissues were cut into pieces, fixed, and dehydrated by conventional means. Tissue specimens were embedded in paraffin blocks and sliced at a thickness of approximately 5 μm.
Then, the tissue sections were subjected to haematoxylin and eosin (H&E) staining to observe general tissue morphology and Masson's trichrome staining to evaluate collagen deposition, using well-established histological protocols. The ratio of intimal to medial (I/M) thickness was calculated using ImageJ, and collagen deposition was quantified using Image-Pro Plus. The tissues were also subjected to immunohistochemical staining for KDM1A, BMP-2, α-SMA, OPN, and proliferating cell nuclear antigen (PCNA). After nuclear staining, dehydration, and mounting, the stained tissue samples were viewed and analysed using a microscope.

| Involvement of BMPR and SMAD signalling in neointimal hyperplasia
Proteins were extracted from aortic tissues of rats subjected to aortic endothelial balloon injury and/or treated with KDM-inh and/or BMP-2. To investigate the role of BMPR signalling, western blot was performed to quantify the protein expression of BMPR-1A, BMPR-1B, and BMPR-2. Protein expression was normalized to that of GAPDH. To explore the involvement of SMAD signalling pathways, western blot was carried out to examine the phosphorylated levels of individual SMADs (1, 5, and 8) relative to the respective levels of total SMADs. The intensity of the protein bands was acquired using Tanon software.

| Statistical analysis
All experiments were performed in triplicates (n = 3), and the data are presented as the mean ± standard deviation (SD). Statistical analysis was carried out by one-way analysis of variance with Tukey's test for multiple comparisons using orIgInPro 8. ImageJ ver. 1.8 and Image-Pro Plus 6.0 were used for image analysis. P < .05 is considered statistically significant.

| Ang-II enhanced VSMC proliferation and revealed inverse correlation between KDM1A and BMP-2
Isolated VSMCs ( Figure S1 in Materials S1) were stimulated with Ang-II and/or KDM-inh to induce phenotypical switching from the contractile to synthetic (or proliferative) state. We observed a notable increase in the proliferation of Ang-II-stimulated VSMCs compared to that of control cells, as revealed by MTT assay (Figure 1A (P < .05; Figure 1D,E) and BMP-2 (P < .05; Figure 1D,F) showed the same trends as their respective mRNA counterparts, with KDMinh playing an antagonistic role against Ang-II in VSMCs.
We further examined the impact of Ang-II on VSMC proliferation ( Figure 2A) and migration ( Figure S2 in Materials S1) using a scratch assay and Transwell assay, respectively. We noted that 24 hours after scratching, cells stimulated with Ang-II proliferated with greater capacity than did control cells, closing the scratch gap more than three times more efficiently. However, in the presence of KDM-inh, the proliferative capacity of the VSMCs was hindered, and wound closure did not proceed as effectively. EdU staining confirmed these results, showing that Ang-II promoted whereas KDM-inh suppressed the proliferation of VSMCs ( Figure 2B). Additionally, examination of cell cycle progression showed that Ang-II promoted the entry of VSMCs into the G0/G1 phase, whereas KDM-inh reduced the proportion of cells in the G0/G1 phase to a level similar to that of control cells ( Figure 2C). VSMC apoptosis was correspondingly suppressed by Ang-II ( Figure 2D), though not to a drastic extent. Again, KDM-inh counteracted the effects of Ang-II as anticipated.

| Ang-II-induced contractile-to-synthetic transition of VSMCs requires KDM1A
To In all cases, the presence of a KDM-inh counteracted the action of Ang-II (P < .05).

| KDM1A and BMP-2 showed inverse correlation in in vivo aortic injury model
In  Figure 4C).

| KDM1A inhibition and BMP-2 stimulation attenuated neointimal hyperplasia and affected VSMC behaviour in vivo
Histological examination of aortic endothelial tissues ( Figure 5) revealed a remarkable increase in the intimal thickness in rats   The data are presented as the mean ± SD of three independent replicates, n.s., not significant at P < .05 (comparisons between unmarked groups are all significant at P < .05)

| Regulation of neointimal hyperplasia via BMP-2 involves SMAD and BMPR signalling
With regard to BMP-2 signalling, we were interested in the involvement of SMADs during neointimal formation. We proceeded to evaluate the expression of various SMADs in response to aortic endothelial balloon injury ( Figure 7A,B). In particular, SMADs 1, 5, and 8 have been widely implicated in BMP-2 signalling. Herein, we noticed that balloon-induced injury led to a drastic decrease in the phosphorylation of all three SMADs (P < .05), whereas KDM-inh and BMP-2 individually recovered their expression to various degrees (not all differences were significant). However, the combination of KDM-inh and BMP-2 effectively triggered the phosphorylation of SMADs 1, 5, and 8 after injury (P < .05), and in the case of SMAD5 and SMAD8, phosphorylation was restored to levels close to those of control rats that were not subjected to injury.
Knowing that downstream BMP/BMPR signalling is critically im- The opposite was observed for BMPR-2, whereby rats treated with both KDM-inh and BMP-2 exhibited a large decline in BMPR-2 expression compared to that in non-treated injured rats (P < .05).

| D ISCUSS I ON
The transition from a contractile to synthetic phenotype is a necessary step in the early developmental stages of vascular diseases such as atherosclerosis and restenosis. 34  In a well-established study using rat aortic VSMCs, Nakaoka et al suggested that BMP-2 inhibited neointimal hyperplasia caused by balloon injury, implicating the therapeutic potential of BMP-2 in the prevention of vascular proliferative diseases. 23 Our histological analysis of rat aortic tissues showed that KDM-inh and BMP-2 were able to attenuate neointimal formation and tissue fibrosis after balloon-induced injury ( Figure 5). We note also that BMP-2 (and consequently, KDM-inh) promoted the contractile phenotype in VSMCs and inhibited their proliferation, as signified by the increased expression of α-SMA and decreased expression of PCNA in injured aortic tissues treated by BMP-2 ( Figure 6). This is complementary to our in vitro observations (Figures 1-3) and is consistent with other reports demonstrating the importance of BMP-2 in the maintenance of contractile markers and suppression of proliferation in VSMCs. 40 notype. This interaction between BMPs and MRTFs was possibly due to non-SMAD pathways. 40 Herein, we revealed that KDM-inh suppressed neointimal hyperplasia in injured aortic tissues by mediating canonical SMAD-related pathways ( Figure 7A,B). The same phenomenon was observed when injured tissues were treated by BMP-2 ( Figure 5). The activation of R-SMADs (1, 5, and 8) upon administration of KDM-inh and BMP-2 was accompanied by enhanced expression of BMPR-1A and BMPR-1B, but BMPR-2 signalling was disrupted ( Figure 7C,D).
The premise and results of our investigation may seem to disagree with a number of studies reporting that BMP-2 contributes to vascular calcification, and thus atherosclerosis. We propose several explanations for the controversy. First, KDM1A signalling, which is the key to this study, could be much more potent than BMP-2 signalling. KDM1A itself may have unknown, unreported pro-inflammatory or pro-atherogenic effects, which may override those of BMP-2. While KDM1A  Figure 7C,D), which is normally downregulated in pro-atherogenic conditions. 47 Another possibility is that, as we have mentioned previously, BMP-2 signalling plays pleiotropic roles in regulating cellular processes. Depending on the specific growth conditions and state of proliferation, BMP-2 may have different effects on VSMCs in vitro or function as part of a continuum. 48 It was also suggested that while BMP-2 halts VSMC proliferation and causes cell cycle arrest, only further and continuous exposure to BMP-2 will result in vascular calcification. 48 Because of the prominent implications of BMP-2 in osteochondrogenic functions, in future studies, it will be interesting to explore whether these functions have an influence on VSMC phenotype and vascular remodelling in the context of neointimal hyperplasia. Specific molecular interactions between KDM1A and various BMPRs, especially BMPR-2, will also be investigated to provide further insights into the mechanism of the KDM1A/BMP signalling axis.
Collectively, the findings reported here demonstrated the regulatory correlation between KDM1A and BMP-2 signalling in vascular remodelling, specifically in terms of VSMC phenotypic modulation (in vitro and in vivo) and the control of neointimal hyperplasia. We speculate that an in-depth understanding of the functions and implications of KDM1A will offer insights into the development of innovative strategies for the treatment of vascular disorders.

ACK N OWLED G EM ENTS
This study was funded by the National Natural Science Foundation of China (Grant no. 81701166).

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

AUTH O R S ' CO NTR I B UTI O N S
XZ and HY designed the study; XZ, TH, and HZ performed cell culture and Western blot; XZ, TH, and WP carried out immunohistochemistry and microscopic experiments; XZ, TH, YZ, and QL were involved in in vivo experiments; XZ and TH performed flow cytometry; XZ, TH, WP, YZ, and QL analysed the data; XZ, TH, and HY drafted and revised the manuscript. All authors read and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.