Colchicine reduces neointima formation and VSMC phenotype transition by modulating SRF-MYOCD activation and autophagy

Abstract

Vascular smooth muscle cell (VSMC) phenotype transformation significantly contributes to vascular intimal hyperplasia. However, effective preventive and therapeutic measures are lacking. Colchicine, a binary alkaloid derived from Colchicum autumnale, is traditionally used for treating inflammatory diseases. Its role in neointima formation is not fully understood. Here, we investigated the role of colchicine in vascular intimal hyperplasia. We found that colchicine significantly reduced vascular intimal hyperplasia in an animal model at 7, 14, and 28 days post carotid artery ligation and increased the number of contractile-phenotype VSMCs (SMA-positive cells) in the neointimal areas. In vitro experiments demonstrated that colchicine facilitated the transition of VSMCs from a proliferative phenotype to a contractile phenotype. Additionally, colchicine attenuated PDGF-BB-induced phenotypic conversion and upregulated the expression of serum response factor (SRF) and myocardin (MYOCD). Further molecular mechanistic studies revealed that colchicine inhibited the expression of forkhead box protein O3A (FOXO3A) to increase the activation of the SRF‒MYOCD complex. FOXO3A can bind to MSX1/2, thereby inhibiting the expression of SRF–MYOCD and contractile genes. Moreover, colchicine maintains vascular homeostasis and stabilizes the contractile phenotype by affecting the expression of autophagy-related genes (LC3II, p62, and Beclin-1) induced by FOXO3A. Additionally, colchicine inhibited monocyte/macrophage infiltration and inflammatory cytokine expression. In summary, this study suggests that colchicine inhibits vascular intimal hyperplasia by modulating FOXO3A-mediated SRF-MYOCD activation and autophagy, providing new insights for future therapeutic approaches targeting occlusive vascular diseases.

Introduction

Approximately 48.6% of Americans ≥20 years of age are estimated to have cardiovascular disease (CVD) [1]. In China, the incidence of atherosclerotic cardiovascular disease (ASCVD) has rapidly increased. In 2016, approximately 2.4 million deaths were attributed to ASCVD, representing 25% of all-cause mortality and 61% of CVD-related deaths, a significant increase from one million deaths in 1990 [2]. The pathological progression of atherosclerosis leads to plaque deposition in the vessel wall, ultimately causing vascular obstruction or plaque rupture and subsequent thrombogenesis [3, 4]. Advancements in percutaneous coronary intervention (PCI) and coronary artery bypass grafting have improved survival rates among patients with coronary heart disease [5]. However, complications such as in-stent restenosis and vein graft disease due to intimal hyperplasia are still prevalent [6, 7]. Hence, there is an urgent need for new and effective strategies for the prevention and treatment of these complications.

Vascular smooth muscle cells (VSMCs) constitute one of the primary cell types comprising the vascular wall and play crucial roles in maintaining vascular structure and function [8]. Mature VSMCs regulate blood flow and pressure by modulating vessel diameter. Under normal physiological conditions, VSMCs exhibit a differentiated, quiescent, and contractile state with low proliferative rates, characterized by specific contractile proteins. VSMCs demonstrate significant cellular plasticity and are capable of markedly altering their phenotype in response to environmental stimuli and extracellular signals. The mature, differentiated phenotype of VSMCs undergoes physiological transcriptional profiling through the regulation of “contractile” protein-encoding gene expression, such as smooth muscle alpha-actin (SMA), smooth muscle myosin heavy chain (MYH11), and smooth muscle protein 22 alpha (SM22α, also known as transgelin) [9, 10]. When exposed to various pathological conditions (e.g., risk factors for atherosclerosis and hypertension), VSMCs undergo phenotypic switching, a biological process whereby VSMCs dedifferentiate or transdifferentiate into other phenotypes.

Colchicine, a bioactive alkaloid isolated and purified from the ancient medicinal plant Colchicum autumnale, has demonstrated efficacy in treating cardiovascular diseases in recent years. Clinical trials, such as LoDoCo, LoDoCo2, and CoLCoT, have demonstrated its effectiveness in treating conditions such as atherosclerosis, heart failure, myocardial infarction, atrial fibrillation, and COVID-19-induced cardiovascular complications [11–14]. Research suggests that colchicine can induce the transformation of pathogenic smooth muscle cell-derived macrophage-like cells and osteoblast-like cells into protective myofibroblast-like cells. This transformation thickens and stabilizes the fibrous cap, facilitating plaque regression, and relies on Notch3 signal transduction in plaque cells derived from smooth muscle cells [15]. Additionally, colchicine inhibits the release of inflammatory cytokines from activated endothelial cells and macrophages, thereby inhibiting the development of atherosclerosis [16]. However, the role and underlying molecular mechanisms of colchicine in neointima formation have yet to be elucidated. Therefore, this study aimed to explore the potential mechanisms underlying colchicine activity in intimal hyperplasia.

Materials and methods

Reagents

Colchicine (Cat#: HY-18982), MG-132 (Cat#: HY-13259), SC79 (Cat#: HY-18749), LY294002 (Cat#: HY-10108), and magnetic stand (Cat#: HY-K0200) were purchased from MedChemExpress (Monmouth Junction, NJ, USA). Platelet-derived growth factor-BB (PDGF-BB) (Cat#: 10572-H07Y) was purchased from Sino Biological (Beijing, China). siFOXO3A (Cat#: siB170217014430-1-5), siMSX1 (Cat#: siB13322204623-1-5), and siMSX2 (Cat#: siB160307104015-1-5) were obtained from RiboBio (Guangzhou, China). 3-MA (Cat#: 5142-23-4) was purchased from Aladdin (Shanghai, China). Protein A/G magnetic beads were purchased from NuoyiBio (Shanghai, China). The antibodies used are listed in Table S1.

Cell culture

Human aortic smooth muscle cells (HASMCs) and 293 T and RAW264.7 cells were purchased from ATCC (Rockville, MD, USA) and cultured in high-glucose DMEM containing 10% fetal bovine serum and 50 μg/mL penicillin/streptomycin. HASMCs were selected for experiments within 10 passages and fasted for 12 h before appropriate treatment in serum-free medium.

In vivo mouse studies

The mouse in vivo study protocol was approved by the Ethics Committee of Hefei University of Technology (HFUT20220506003) and was in line with the Guidelines for the Care and Use of Laboratory Animals published by the National Institutes of Health. The animal studies were conducted in compliance with the ARRIVE guidelines. C57BL/6J wild-type mice were purchased from GemPharmatech (Nanjing, China) and raised at the Animal Center of Hefei University of Technology (Hefei, China).

To study the effect of colchicine on vascular neointima formation, carotid artery ligation was performed on mice to establish a vascular injury model [17]. Briefly, male C57BL/6J mice were randomly divided into 2 groups (5 mice/group), anesthetized by inhalation of isoflurane, and subcutaneously injected with buprenorphine (0.1 mg/kg) for analgesia. The left carotid artery was then ligated with a 6–0 silk suture to perform carotid artery injury in all the mice. Sham operation without ligation was performed on the right carotid artery. After the operation, the mice were fed a normal diet. One group of mice was intraperitoneally injected with colchicine solution (0.05 mg/kg, dissolved in PBS) for 7–28 consecutive days, and the other group of mice was administered the same amount of PBS.

At 7 d, 14 d, and 28 d, the mice were euthanized via the intraperitoneal injection of an overdose of pentobarbital (500 mg/kg). Then, the left and right carotid arteries of the mice were collected and used to prepare 5-μm frozen sections. The sections were stained with hematoxylin and eosin (HE) to assess the morphological changes in the vascular tissue and intimal hyperplasia at the site of injury. The livers, hearts and kidneys of the mice were collected to determine their histochemical characteristics via HE staining. Serum samples prepared from blood were subjected to biochemical assays via an automated biochemical analyzer to determine the levels or activities of alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), serum urea (UREA), creatinine (CREA), uric acid (UA), creatine kinase (CK), lactate dehydrogenase (LDH), and hydroxybutyrate dehydrogenase (HBDH).

Immunofluorescence staining

The expression of FOXO3A, SMA, OPN, SRF, and MYOCD in carotid arteries and that of LC3 and Beclin-1 in HASMCs were determined via immunofluorescence staining [18]. The mean fluorescence intensity (MFI) of the target protein was analyzed by ImageJ, and the MFI of the control group was defined as 1.

Determination of cell viability, cell cycling, and cell migration

The viability of HASMCs was determined via a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Briefly, cells in 96-well plates were treated with different concentrations of colchicine (0, 2.5, 5, 10, 20, or 40 nM) in serum-free medium for 24 h. After treatment and aspiration of the treated medium, 100 μL of MTT solution (5 mg/mL in serum-free medium) was added to the cells in each well and incubated for 4 h. After removal of the medium, 150 μL of DMSO was added to each well and incubated on a shaker at low speed with shaking for 10 min to fully dissolve the blue‒purple crystals formed within the cells. The absorbance value of each well at a wavelength of 570 nm was subsequently measured using a microplate reader (BioTek, USA). The results were normalized to the average of the control group and are expressed as a fold change relative to the control group.

The cell cycle distribution was determined by FACS [19]. Briefly, after colchicine treatment, HASMCs were fixed in 70% ethanol for 24 h at 4 °C and washed twice with PBS, RNase (1 mg/mL) was added, and the mixture was incubated at 37 °C for 30 min. The cells were then incubated with propidium iodide (PI) solution (2 mg/mL) for 30 min in the dark. Finally, the cell cycle distribution was determined via a FACS assay with Cytoflex (Beckman, USA).

To determine the effect of colchicine on HASMC migration, HASMCs (85% ~ 95% confluence) in a 12-well plate were wounded with a yellow pipette tip. After the cells were washed twice with PBS to remove suspended cells and cell debris, DMEM containing 2% FBS was added to the cells. Images were taken under a light microscope (Leica, Wetzlar, Germany), and the scratch width was recorded as W₀. The cells were then subjected to colchicine treatment and cultured for 24 h, after which photographs were taken, and the width was recorded as W₂₄. The mean migration distance (W₀ – W₂₄) in the control group was defined as 1. The relative migration distance was calculated as follows: Treatment _(W0–W24)/Ctrl _(W0–W24).

Preparation of the FOXO3A expression vector or siRNA transfection

The human FOXO3A overexpression plasmid (pCMV-HA-FOXO3A, also referred to as HA-FOXO3A) was purchased from Tsingke Biotechnology Co., Ltd. (Beijing, China), and its sequence and protein expression were confirmed. HASMCs in 6-well plates of approximately 90% density were transfected with the pCMV-HA empty vector or HA-FOXO3A using Lipofectamine® RNAiMAX Transfection Reagent (Invitrogen) for 12 h, after which the cells were incubated with complete DMEM and cultured for an additional 12 h and subsequently subjected to the desired treatment.

FOXO3A siRNA and control siRNA were purchased from RiboBio (Guangzhou, China). HASMCs in 6-well plates were transfected with control siRNA or FOXO3A siRNA via Lipofectamine® RNAiMAX Transfection Reagent (Invitrogen). After 24 h of transfection, the cells were incubated with complete DMEM, cultured for an additional 24 h, and then subjected to the indicated treatments.

Western blot and quantitative real-time PCR (qRT‒PCR)

After the indicated treatments, protein expression in total protein from tissue samples or cells was determined by Western blotting. Briefly, total protein was extracted from HASMCs or tissue samples via RIPA lysis buffer. The protein samples were then separated by SDS‒PAGE and transferred to NC membranes. The membrane was blocked with 5% skim milk in PBST for 1 h at room temperature. Then, depending on the molecular weight of the target protein, the membrane is cut into smaller pieces before hybridization. The membrane was then incubated with a specific primary antibody overnight at 4 °C. After three washes with PBST, the membrane was incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. The images were then captured with a chemiluminescence imaging system (Qinxiang, ChemiScope 3300 Mini, China). The density values of the bands were measured via ImageJ. The density of the target band was normalized to that of GAPDH in the respective samples to reduce the variance.

After treatment, total RNA was extracted from the cells via TRIzol reagent (Invitrogen, Carlsbad, CA, USA) and reverse transcribed into cDNA. The levels of OPN, CNN1, SMA, SM22α, SRF, MYOCD, IL-1β, IL-6, and TNF-α in cells were analyzed via qRT‒PCR with the primers listed in Supplementary Table S2. For qRT‒PCR, the target gene level was normalized to that of GAPDH mRNA in the corresponding sample.

Coimmunoprecipitation (Co-IP) assay

The endogenous interaction between FOXO3A and MSX1/2 in HASMCs was determined by Co-IP. Briefly, after the HASMC proteins were extracted, MSX1/2 (FOXO3A) or IgG antibodies were incubated with the protein extract overnight. Protein A/G magnetic beads were then added and incubated on a shaker for 1 h at room temperature, followed by 4–5 washes with PBST on a magnetic stand. The bead-bound protein was then eluted with 2× SDS loading buffer. The corresponding protein concentration was determined by Western blotting.

Luciferase reporter assay

To determine the effect of FOXO3A on the transcriptional activity of SRF, we transfected HASMCs with a luciferase reporter plasmid containing SRF response element (SRE) (pHHSRF-Luc) and pCMV-HA or HA-FOXO3A in serum-free medium for 12 h, followed by stabilization in complete medium for 12 h. The cells were then incubated with serum-free medium with or without colchicine (COL, 10 nM) for 24 h. Luciferase activity was analyzed via a dual luciferase reporter assay.

Ubiquitination assay

HASMCs at approximately 90% confluence were incubated with serum-free medium with or without colchicine (COL, 10 nM) for 24 h. Then, MG132 (10 nM) was added to DMEM and incubated for 8 h. Cell lysates were immunoprecipitated (IP) with an anti-FOXO3A antibody at 4 °C overnight. Finally, the expression of ubiquitin (Ub) was determined via Western blotting.

Analysis of FOXO3A stability

To determine the stability of the FOXO3A protein, HASMCs (~90% confluence) were incubated with serum-free medium with or without colchicine (COL, 10 nM) for 24 h. The cells were then treated with MG132 at the indicated time points (0 h, 4 h, 8 h, and 12 h). Total cellular proteins were harvested and used to determine FOXO3A expression via Western blotting.

Data analysis

All experiments were repeated at least 3 times, and representative results are given. All values are expressed as the means ± SDs. The normally distributed data were analyzed via the one-sample K‒S nonparametric test of SPSS 22 software, and parametric statistics were used (Student’s t test was used for two groups, and one-way ANOVA was used for more than two groups). The difference was statistically significant at P < 0.05.

Results

Colchicine inhibits carotid artery ligation-induced intimal hyperplasia

To investigate the impact of colchicine on intimal hyperplasia in mice, we established a model of neointima formation by ligating the left common carotid artery in mice. As determined by HE staining, colchicine did not affect histochemistry characteristics of liver, heart and kidney (Fig. S1a). Moreover, colchicine did not affect the serum functional parameters of the liver (ALT, AST, and ALP), kidney (UREA, CREA, and UA) or heart (CK, LDH, and HBDH) (Fig. S1b), suggesting that colchicine has little toxicity in vivo. Stenosis and intimal hyperplasia in the left carotid artery gradually increased after ligation. Interestingly, 0.05 mg/kg colchicine significantly reduced the neointimal area at 7, 14, and 28 days after carotid artery ligation (Fig. 1a). Furthermore, the mRNA expression of SMA and SM22α was markedly reduced, whereas that of osteopontin (OPN, a marker for proliferative VSMCs) was increased in the carotid arteries after ligation. Moreover, colchicine reversed these effects (Fig. 1b). Additionally, the immunofluorescence staining results demonstrated that colchicine decreased OPN expression but increased SMA expression in the neointimal region (Fig. 1c, d). The mRNA expression of SMA and SM22α was more remarkable than the protein expression, implying that the contractile proteins were more stable and that the mRNA expression was sensitive in response to the stimulus. These results suggest that colchicine can inhibit neointima formation by promoting the expression of contractile genes.

Fig. 1. Colchicine reduces neointima formation.

After the carotid artery ligation model was established in C57BL/6J mice, the mice were randomly divided into control and experimental groups. The control group received an intraperitoneal injection of physiological saline (PBS), whereas the experimental group received an intraperitoneal injection of colchicine (Col, 0.05 mg/kg). Carotid artery tissues were collected on Days 7, 14, and 28 after carotid artery ligation. a Hematoxylin and eosin (HE) staining was performed to assess neointimal formation, and the neointimal area was quantitatively analyzed. *P < 0.05 (n = 5). b mRNA levels of SM22α, SMA, and OPN in the carotid artery on Day 28 after ligation with or without colchicine treatment were determined via qRT‒PCR. *P < 0.05 (n = 5). Immunofluorescence staining was used to evaluate the expression of SMA (c) and OPN (d) in the carotid artery. *P < 0.05 (n = 5).

Colchicine inhibits the proliferation and migration of VSMCs

The transformation of the VSMC phenotype to a proliferative phenotype, with increased proliferation and migration properties, is the main pathophysiological process during neointima formation. To further investigate the effects of colchicine on vascular intimal hyperplasia, we determined the effects of colchicine on HASMC proliferation and migration in vitro. Initially, the MTT assay was employed to assess changes in cell viability under colchicine treatment. The results revealed that colchicine had no significant cytotoxic effects on HASMC viability at 5 nM, moderately reduced cell viability at 10 nM, and significantly reduced HASMC viability at concentrations exceeding 20 nM (Fig. 2a). Furthermore, flow cytometry analysis indicated that 10 nM colchicine did not induce cell apoptosis (Fig. 2b). Thus, 10 nM colchicine was selected for subsequent experiments.

Fig. 2. Colchicine inhibits the proliferation and migration of HASMCs.

a HASMCs were treated with colchicine at the indicated concentrations for 24 h. An MTT assay was used to determine the optimal concentration of colchicine for HASMCs. ns: not significantly different, *P < 0.05 (n = 5). b HASMCs were treated with 10 nM colchicine for 24 h. Flow cytometry was used to detect apoptosis. ns: not significantly different (n = 3). c HASMCs were treated with colchicine for 24 h in serum-free medium, and the cell cycle distribution was analyzed via flow cytometry. *P < 0.05 (n = 3). d, e HASMC migration was assessed via a scratch assay after treatment with colchicine for 24 h in medium containing 2% FBS or with PDGF-BB alone or in combination with colchicine for 24 h. *P < 0.05 (n = 3). f, g HASMCs were treated with colchicine (10 nM), PDGF-BB (20 ng/mL) or a combination of colchicine and PDGF-BB for 24 h. p-AKT, AKT, PTEN, and PCNA protein expression was examined via Western blotting. *P < 0.05 (n = 3).

The G1/S transition in the cell cycle serves as a crucial checkpoint for cell proliferation. Our findings demonstrated that colchicine induced cell cycle arrest at the G0 phase (Fig. 2c), impeding the cell cycle transition to the S phase. Furthermore, the potential impact of colchicine on the migration of HASMCs was assessed through scratch assays. Colchicine significantly inhibited basal and PDGF-BB-induced HASMC migration (Fig. 2d, e). Previous research has indicated that PDGF-BB-induced cell proliferation and migration are regulated by the AKT signaling pathway. Colchicine inhibited the phosphorylation of AKT (p-AKT) while increasing the negative feedback regulator phosphatase and tensin homolog (PTEN) in the AKT signaling pathway. Moreover, colchicine suppressed the expression of the proliferation-related gene proliferating cell nuclear antigen (PCNA) (Fig. 2f). Furthermore, colchicine inhibited PDGF-BB-induced activation of p-AKT and PCNA (Fig. 2g). The inhibition of PTEN expression by PDGF-BB was also reversed by colchicine (Fig. 2g). These data suggest that colchicine inhibits HASMC proliferation and migration.

Colchicine inhibits VSMC phenotype transition

The phenotypic transformation of VSMC, characterized by abnormal proliferation and migration capabilities as well as dysregulated autophagy, is a significant factor in the onset and progression of vascular remodeling diseases. We subsequently explored the impact of colchicine on the phenotypic transition of VSMC. Colchicine increased the expression of contraction genes, including SMA, SM22α, MYH11, and calponin 1 (CNN1) (Fig. 3a, b). The expression of the VSMC proliferative marker OPN was reduced by colchicine (Fig. 3a, b).

Fig. 3. Colchicine activates the SRF-MYOCD complex to increase the expression of contractile genes in VSMC.

a, b, e, f HASMCs were treated with colchicine for 24 h. c, d, g HASMCs were treated with PDGF-BB alone or in combination with colchicine for 24 h. The expression of OPN, SMA, SM22α, CNN1, MYH11 (a, c), SRF and MYOCD (e) was assessed via qRT‒PCR. The protein expression of OPN, SMA, SM22α, CNN1 (b, d), SRF, and MYOCD (f, g) was analyzed via Western blotting. *P < 0.05 (n = 3 or 4). h Immunofluorescence staining was conducted to detect the expression of SRF and MYOCD in the neointimal region. *P < 0.05 (n = 5).

PDGF-BB effectively promotes the proliferation and migration of VSMCs and induces phenotypic transformation. We confirmed that PDGF-BB inhibits the expression of contractile genes such as SMA, CNN1, and SM22α (Fig. 3c, d). Colchicine restored the expression of SMA, SM22α, and CNN1 inhibited by PDGF-BB while inhibiting PDGF-BB-induced OPN expression (Fig. 3c, d), thus maintaining the contractile phenotype of HASMCs.

The CArG-SRF-MYOCD complex is a crucial regulatory factor that maintains VSMC differentiation and regulates the transcription of VSMC contractile-related genes. In vitro, colchicine increased the expression of SRF and MYOCD in HASMCs (Fig. 3e, f) and restored the expression of SRF and MYOCD that was inhibited by PDGF-BB (Fig. 3g). In vivo, the expression levels of SRF and MYOCD in the neointima region of the ligated carotid artery were significantly lower than those in the nonligated carotid artery (Fig. 3h). Additionally, colchicine restored the expression levels of SRF and MYOCD in the neointimal region following vascular injury (Fig. 3h).

Colchicine suppresses FOXO3A expression to restore the expression of vascular contractile genes

Previous studies have suggested that FOXO3A contributes to the aberrant expression of contractile genes in HASMCs. FOXO4 has been shown to be involved in inhibiting SRF/MYOCD activity. We previously demonstrated that colchicine affects FOXO3A mRNA expression in endothelial cells [20]. Therefore, we explored the potential involvement of FOXO proteins in colchicine-mediated inhibition of intimal hyperplasia. Colchicine increased FOXO1 mRNA expression but had no effect on FOXO4 or FOXO6 mRNA expression (Fig. S2). Moreover, colchicine significantly inhibited the expression of FOXO3A mRNA and protein in HASMCs (Fig. 4a, b, S2), suggesting that colchicine maintains the contractile phenotype of VSMC by inhibiting FOXO3A expression. Immunofluorescence staining and qRT‒PCR also confirmed that the expression of FOXO3A was elevated in injured vessels, whereas colchicine significantly suppressed FOXO3A expression in mice (Fig. 4c, d). Therefore, we further determined the direct impact of FOXO3A on HASMC phenotypic transition. We first confirmed the efficiency of FOXO3A overexpression or silencing (Fig. 4e, g, h). FOXO3A siRNA transfection increased the protein expression of contractile genes (SMA, SM22α, and CNN1) but inhibited the protein expression of OPN (Fig. 4f). Conversely, FOXO3A overexpression inhibited the protein expression of contractile genes (SMA, SM22α, and CNN1) but increased OPN protein expression (Fig. 4i). Importantly, we found that colchicine significantly reversed the downregulation of contractile genes and the upregulation of proliferative genes caused by FOXO3A overexpression (Fig. 4j). These results indicate that FOXO3A participates in the colchicine-mediated inhibition of VSMC phenotype transition.

Fig. 4. Colchicine inhibits FOXO3A to increase VSMC contractile gene expression.

After treatment with colchicine (10 nM) for 24 h, FOXO3A protein (a) and mRNA (b) levels were assessed by Western blot and qRT‒PCR, respectively. *P < 0.05 (n = 3). c FOXO3A expression in the neointimal area was determined by immunofluorescence staining. *P < 0.05 (n = 5). d FOXO3A mRNA levels in the carotid artery on Day 28 post-vascular injury or after colchicine treatment were examined via qRT‒PCR. *P < 0.05 (n = 5). e, f HASMCs were transfected with control siRNA (siCtrl) or FOXO3A siRNA (siFOXO3A) for 24 h in serum-free medium, followed by treatment with complete medium for another 24 h. The protein expression of FOXO3A (e), SMA, CNN1, SM22α, and OPN (f) was detected via Western blotting. *P < 0.05 (n = 3). g‒i HASMCs were transfected with pCMV-HA or pCMV-HA-FOXO3A (HA-FOXO3A) for 24 h. FOXO3A mRNA expression was assessed via qRT‒PCR (h). The protein expression of FOXO3A (g), SMA, CNN1, SM22α, and OPN (i) was analyzed via Western blotting. *P < 0.05 (n = 3). j After HASMCs were transfected with pCMV-HA/HA-FOXO3A for 12 h, the cells were treated with colchicine (10 nM) for 24 h. The protein expression of p-FOXO3A, FOXO3A, OPN, SMA, CNN1, and SM22α was evaluated via Western blotting. *P < 0.05 (n = 3).

FOXO3A inhibits the expression and transcriptional activity of the SRF-MYOCD complex

It has been reported that FOXO3A (Ser253), which is an inactive form of FOXO3A with increasing instability, can be phosphorylated via the PI3K/AKT pathway [21]. We determined that colchicine reduced p-AKT in HASMCs (Fig. 2f). We wondered whether colchicine regulated FOXO3A and contractile gene expression via p-AKT. Colchicine reduced p-AKT, which was further reduced by the AKT inhibitor LY294002 (Fig. S3a). The AKT inhibitor LY294002 further enhanced the colchicine-induced increase in contractile gene expression (Fig. S3b). However, LY294002 did not affect FOXO3A expression (Fig. S3b). Conversely, the AKT activator affected the expression of contractile genes but not FOXO3A (Fig. S3c, d), suggesting that the colchicine-mediated regulation of FOXO3A does not depend on p-AKT.

Previous studies have indicated that FOXO3A inhibits MYOCD expression directly or via catalase [22, 23]. We confirmed that silencing FOXO3A in HASMCs significantly increased the protein levels of SRF and MYOCD (Fig. 5a). Conversely, overexpression of FOXO3A inhibited the expression of SRF and MYOCD (Fig. 5b). More importantly, colchicine reversed the inhibitory effect of FOXO3A overexpression on SRF and MYOCD expression (Fig. 5c). Additionally, we observed that the decrease in the expression of contractile genes (SM22α, SMA, CNN1, SRF, and MYOCD) induced by FOXO3A overexpression was reversed by SRF or MYOCD overexpression (Fig. 5d, e). The results of the luciferase assay revealed that FOXO3A significantly inhibited the transcriptional activity of the SRF response promoter, which was partially restored by colchicine (Fig. 5f).

Fig. 5. Colchicine restores FOXO3A-inhibited SRF‒MYOCD expression.

HASMCs were transfected with siCtrl/siFOXO3A (a) or pCMV-HA/HA-FOXO3A (b) for 24 h in serum-free medium, followed by incubation in complete medium for another 24 h. The protein expression of SRF and MYOCD was assessed via Western blotting. *P < 0.05 (n = 3). c After HASMCs were transfected with pCMV-HA/HA-FOXO3A for 12 h, the cells were treated with colchicine (10 nM) for 24 h. The protein expression of SRF and MYOCD was evaluated via Western blotting. *P < 0.05 (n = 3). d, e HASMCs were transfected with pCMV-HA/HA-FOXO3A for 12 h, stabilized for 12 h, and then transfected with HA-SRF or HA-MYOCD for another 12 h. The protein expression of SRF, MYOCD, SM22α, SMA, and CNN1 was examined via Western blotting. *P < 0.05 (n = 3). f HASMCs were transfected with the SRF-responsive element (SRE) reporter plasmid (pHHSRF-Luc) and pCMV-HA or HA-FOXO3A for 24 h in serum-free medium. The cells were then transferred to complete medium with or without colchicine (10 nM) for 24 h. Luciferase activity was analyzed via a dual-luciferase reporter assay. *P < 0.05 (n = 5).

FOXO3A inhibits SRF‒MYOCD expression by binding MSX1/2

The transcriptional activity of the SRF-MYOCD complex is regulated by various factors, among which Msh homeobox 1/2 (MSX1/2) is important upstream negative regulatory factors. These proteins can prevent SRF or SRF‒MYOCD from binding to the CArG box on the contraction gene promoter of VSMC, thereby inhibiting the expression of contraction genes. We found that colchicine significantly inhibited the expression of MSX1/2 (Fig. 6a) and that MSX1/2 was upregulated in cells overexpressing FOXO3A (Fig. 6b). More importantly, colchicine significantly inhibited the expression of MSX1/2 induced by FOXO3A overexpression (Fig. 6c). Next, we determined that the inhibition of contractile genes (SMA, SM22α, CNN1, SRF, and MYOCD) by FOXO3A overexpression was blocked by MSX1 siRNA or MSX2 siRNA transfection (Fig. 6d, e). To further explore the potential mechanism by which FOXO3A inhibits SRF/MYOCD activity, we hypothesized that FOXO3A may directly interact with MSX1/2. Indeed, we found that endogenous FOXO3A interacts with MSX1/2 in HASMCs, as determined by a co-IP assay (Fig. 6f). Moreover, we determined that colchicine increased FOXO3A ubiquitination (Fig. 6g), facilitating the degradation of FOXO3A (Fig. 6h). In summary, FOXO3A can bind with MSX1/2 to inhibit SRF/MYOCD activity. By increasing ubiquitination-mediated degradation, colchicine reduces FOXO3A expression, thereby increasing the expression of contractile genes.

Fig. 6. FOXO3A interacts with MSX1/2 to regulate contractile gene expression.

a After treatment with colchicine (10 nM) for 24 h, MSX1/2 protein levels in HASMCs were assessed by Western blotting. *P < 0.05 (n = 3). b HASMCs were transfected with pCMV-HA/HA-FOXO3A for 12 h in serum-free medium, followed by incubation in complete medium for another 12 h. The expression of the FOXO3A and MSX1/2 proteins was detected by Western blotting. *P < 0.05 (n = 3). c HASMCs were transfected with pCMV-HA/HA-FOXO3A for 12 h in serum-free medium, followed by treatment with colchicine (10 nM) for 24 h. FOXO3A and MSX1/2 protein expression was examined via Western blotting. *P < 0.05 (n = 3). d, e HASMCs were transfected with pCMV-HA/HA-FOXO3A for 12 h in serum-free medium and then transfected with siMSX1 (d) or siMSX2 (e) for another 24 h. The expression of the FOXO3A, MSX1/2, SRF, MYOCD, SMA, and SM22α proteins was evaluated via Western blotting. *P < 0.05 (n = 3). f HASMC lysates were extracted and subjected to co-IP with anti-FOXO3A or anti-MSX1/2 antibodies. g HASMCs were treated with colchicine (10 nM) for 24 h in serum-free medium. Then, the cells were treated with MG132 (10 nM) for another 8 h. The cell lysate was subjected to an IP assay with an anti-FOXO3A antibody. The expression of ubiquitin was determined by Western blotting. *P < 0.05 (n = 3). h HASMCs were treated with colchicine (10 nM) for 24 h in serum-free medium. The cells were then treated with MG132 (10 nM) for the indicated times. The expression of FOXO3A was determined by Western blotting. *P < 0.05 (n = 3).

Colchicine inactivates FOXO3A-mediated autophagy

Intimal hyperplasia is closely related to abnormal autophagy. As an important intracellular degradation pathway, autophagy is involved in regulating key physiological processes of VSMCs, including survival, proliferation, migration, and matrix secretion [24]. FOXO3A regulates autophagy in skeletal muscle and the myocardium by activating genes involved in autophagosome formation. We attempted to determine whether colchicine reduces VSMC phenotype transition by inactivating autophagy. Colchicine significantly increased the expression of P62 but inhibited the expression of LC3-II and Beclin-1 (Fig. 7a, b). In contrast, overexpression of FOXO3A inhibited the expression of P62 and activated the expression of LC3-II and Beclin-1 (Fig. 7c, d). More importantly, colchicine can counteract the regulatory effects of FOXO3A overexpression on P62 and LC3-II levels (Fig. 7e). These findings indicate that the autophagy pathway mediated by FOXO3A is involved in the phenotypic transition of HASMCs, whereas colchicine can inhibit VSMC autophagy.

Fig. 7. Colchicine stabilizes vascular contraction by inhibiting autophagy.

After treatment with colchicine (10 nM) for 24 h, P62, Beclin-1 and LC3 protein levels were assessed by Western blot (a) or immunofluorescence staining (b). *P < 0.05 (n = 3 or 5). c, d HASMCs were transfected with pCMV-HA/HA-FOXO3A for 12 h in serum-free medium, followed by incubation in complete medium for another 12 h. P62, Beclin-1 and LC3 protein expression was detected by Western blot (c) or immunofluorescence staining (d). *P < 0.05 (n = 3 or 5). e HASMCs were transfected with pCMV-HA/HA-FOXO3A for 12 h in serum-free medium, followed by colchicine (10 nM) treatment for 24 h. P62 and LC3 protein expression was examined via Western blotting. *P < 0.05 (n = 3). f, g After overexpressing FOXO3A in HASMCs for 24 h, the cells were treated with 3-MA (5 mM) for 12 h. The expression of FOXO3A, P62, LC3 (f), SMA, SM22α, OPN, SRF, and MYOCD (g) was evaluated via Western blotting. *P < 0.05 (n = 3).

To further determine the role of autophagy in the phenotypic transition of VSMC, we used the autophagy inhibitor 3-methyladenine (3-MA) to observe whether the expression of autophagy- and phenotype transition-related genes was affected by FOXO3A. We overexpressed FOXO3A in HASMCs and subsequently treated them with 3-MA. 3-MA significantly inhibited the protein expression of FOXO3A and the expression of autophagy-related genes (Fig. 7f). 3-MA also restored the expression of the SMA, SM22α, CNN1, SRF, and MYOCD contractile proteins in FOXO3A-overexpressing cells (Fig. 7g). These results indicate that inhibiting autophagy can maintain the contractile phenotype in FOXO3A-overexpressing HASMCs. In summary, these results indicate that FOXO3A promotes VSMC phenotype transition by activating autophagy, whereas colchicine inhibits VSMC phenotype transition by inhibiting FOXO3A-induced autophagy.

Colchicine reduces inflammation in the neointimal area

Colchicine has been demonstrated to inhibit inflammation in multiple models. Monocyte/macrophage infiltration represents the initial stage of vascular inflammation. Cytokines, chemokines, and adhesion molecules serve as crucial proinflammatory mediators that initiate the infiltration of inflammatory cells into the vascular wall and foster the development of inflammatory responses. Our study revealed a significant reduction in the areas positive for F4/80 (a monocyte/macrophage marker) and c-c motif chemokine ligand 2 (CCL2) in the hyperplastic region of mice treated with colchicine (Fig. S4a, b), suggesting that colchicine can suppress monocyte/macrophage infiltration by reducing the expression of the chemokine CCL2. In vitro cell experiments also demonstrated that colchicine reduced LPS-induced interleukin 1β (IL-1β), tumor necrosis factor α (TNF-α), interleukin 6 (IL-6) and NF-κB expression (Fig. S4c, d).

Discussion

Vascular intimal hyperplasia is a critical pathological process in atherosclerosis, in-stent restenosis, and aneurysm formation. VSMC phenotypic switching is considered an early and central step in the development of vascular intimal hyperplasia. In this context, colchicine has been reported to attenuate VSMC proliferation and collagen secretion [25, 26]. By inhibiting METTL14-mediated m⁶A modification, colchicine can increase global mRNA stability, thereby increasing sclerostin expression to inhibit the WNT/β-catenin signaling pathway [25]. Colchicine also limits abdominal aortic aneurysm formation by inhibiting immune cell infiltration [26], indirectly affecting the phenotypic switching of VSMC. March et al. reported that a biodegradable microsphere form of a colchicine analog inhibits DNA synthesis in VSMC [27]. Recent research has demonstrated that atherosclerosis is a type of cancer of VSMCs that is characterized by DNA damage and genomic instability [28]. However, the underlying mechanism by which colchicine affects VSMC phenotypic switching is not fully understood.

Our study demonstrated that colchicine significantly inhibited the neointimal area induced by carotid artery ligation injury in mice at days 7, 14, and 28 while also increasing the expression of contractile proteins in the neointimal region. In vitro experiments revealed that colchicine inhibits VSMC phenotypic switching and mitigates excessive intimal hyperplasia by inducing cell cycle arrest at the G0/G1 phase, thereby preventing cell transition to the S phase and inhibiting VSMC proliferation and migration. We also determined that colchicine increased the transcriptional activity of the SRF‒MYOCD complex to regulate the expression of VSMC contractile proteins. Colchicine also reduced monocyte/macrophage infiltration and inflammation in the neointimal areas, which was another mechanism by which colchicine reduced intimal hyperplasia.

Furthermore, the absence of FOXO3A is crucial for the colchicine-maintained contractile phenotype of HASMCs. Colchicine significantly reduced the expression of FOXO3A in HASMCs. Additionally, we discovered that FOXO3A can bind with MSX1/2 to inhibit the activity of the SRF-MYOCD transcriptional complex. Colchicine disrupts the binding of FOXO3A with MSX1/2, thereby activating the SRF‒MYOCD complex to maintain vascular homeostasis. FOXO3A is crucial for regulating cellular differentiation, apoptosis, and immunity [29, 30]. Various external stimuli, such as growth factors and cytokines, can activate PI3K to activate AKT. Activated AKT can phosphorylate FOXO3A, causing its exclusion from the nucleus and translocation to the cytoplasm, thereby reducing the transcriptional activation of target genes [20]. 14-3-3 proteins have also been demonstrated to contribute to FOXO3A accumulation in the cytoplasm following phosphorylation by AKT [31]. JNK can phosphorylate FOXO3A at serine 574, which selectively binds to the promoters of proapoptotic genes but not to other well-described FOXO3A targets [32]. Therefore, the phosphorylation of FOXO3A and FOXO3A may target different genes. Our research revealed that colchicine can inhibit the phosphorylation of AKT and increase p-FOXO3A. Moreover, AKT activation or inhibition did not affect FOXO3A expression. Interestingly, the expression and activity of FOXO3A in cells were significantly suppressed by colchicine. These data suggest that the colchicine-mediated regulation of FOXO3A does not depend on AKT. We determined that the knockdown of MSX1/2 can reduce the exogenous FOXO3A protein level, implying that the binding of FOXO3A with MSX1/2 can increase FOXO3A stability. Considering that FOXO3A overexpression induces HASMC autophagy, MSX1/2 may increase FOXO3A stability via the proteasome pathway but not the autophagy pathway. Indeed, we determined that colchicine reduced FOXO3A stability by increasing its degree of ubiquitination. The underlying mechanism by which MSX1/2 increases FOXO3A stability requires further investigation.

The role of FOXO3A in VSMC phenotype switching is inconsistent. Yu et al. reported that FOXO3A activation increases VSMC apoptosis, proliferation, and neointima formation through MMP13 [33]. FOXO3A also accelerates abdominal aortic aneurysm formation by regulating VSMC phenotypic transition to an inflammatory phenotype [34]. Endothelial cell FOXO3A is a key molecule in BMP4-promoted intimal hyperplasia [35]. BMP4 upregulates FOXO3 to increase ROS production via SOD-1, thereby inhibiting endothelial cell migration [35]. In contrast, the overexpression of constitutively activated FOXO3A inhibits neointimal hyperplasia via the inhibition of cysteine-rich angiogenic protein 61 (CYR61, CCN1) [36]. These inconsistent data may be the result of FOXO3A expression and modification in response to different stimuli. For example, PI3K/AKT has been reported to phosphorylate FOXO3A at serine 253 to decrease FOXO3A stability [21]. JNK can phosphorylate FOXO3 at serine-574, resulting in selectively binding to the promoters of proapoptotic genes but not to other well-described FOXO3 targets [32]. However, the phosphorylation of FOXO3A by p38, macrophage stimulating 1 (MST1) and AMPK promotes its nuclear entry and increases its transcriptional activity [37–39]. Moreover, the activity of FOXO3A is also regulated by methylation, acetylation and interactions with other factors [40]. We demonstrated that colchicine inhibited FOXO3A expression to activate the SRF‒MYOCD complex and contractile gene expression, which may be related to its interaction with MSX1/2.

Intimal hyperplasia is closely associated with abnormalities in autophagy. Autophagy regulates various physiological functions of VSMCs, including survival, proliferation, migration, and extracellular matrix secretion [41]. Both excessive and insufficient autophagy can lead to intimal hyperplasia [42, 43]. FOXO3A can participate in the development of myocardial hypertrophy or cardiomyopathy by regulating autophagy levels [44, 45]. MYOCD was demonstrated to inhibit autophagy, thereby affecting the phenotypic transformation of VSMCs and alleviating neointimal hyperplasia in a mouse carotid artery injury model [46]. We determined that colchicine increased P62 but inhibited Beclin-1 and LC3-II, suggesting that colchicine may reduce autophagy to maintain the contractile phenotype of VSMC. Interestingly, we determined that FOXO3A overexpression can activate autophagy by affecting P62, Beclin-1 and LC3-II expression, which can be reversed by colchicine. Moreover, FOXO3A-mediated autophagy and VSMC phenotype transformation-related genes were attenuated by the autophagy inhibitor 3-MA. Therefore, our data suggest that colchicine may inhibit autophagy by reducing FOXO3A expression.

Our study has several limitations. First, we determined that colchicine improved neointima formation in vivo by inhibiting FOXO3A expression. However, we do not have FOXO3A VSMC-specific knockout mice to validate our data in vivo. In the future, we will construct FOXO3A VSMC-specific knockout mice to investigate the role of FOXO3A in the inhibition of neointima formation by colchicine. Second, the mechanisms by which MSX1/2 affects FOXO3A ubiquitination should be further investigated. Finally, drug-eluting stents containing rapamycin have been used to reduce in-stent restenosis. Despite the multiple roles of rapamycin in cell function, it can activate autophagy. Our data suggest that colchicine may inhibit neointima formation, at least partially, via the inhibition of autophagy. The interplay between colchicine and rapamycin needs further investigation in the future.

In conclusion, our study demonstrated that colchicine inhibits vascular intimal hyperplasia. Mechanistically, colchicine inhibited FOXO3A expression to 1) inhibit MSX1/2 expression, thereby releasing the SRF-MYOCD complex to induce contractile gene expression, and 2) inhibit autophagy by affecting P62 and LC3 expression. Additionally, colchicine reduced monocyte/macrophage infiltration and inflammation in neointimal areas. Therefore, colchicine has potential as a therapeutic agent for the treatment of vascular proliferative diseases.

Author contributions

Conceptualization and supervision: YLC, XXY, CRM and LKM; Methodology and experiments: BCZ, SNW and WYZ; Investigation: BCZ, SNW, WYZ, MMZ; Data analysis: SNW, HM, LD; Writing-original draft preparation: BCZ, SNW, WYZ; Writing-review and editing: YLC, XXY, CRM and LKM; Funding acquisition: YLC, BCZ and CRM. All authors have read and agreed to the final version of the manuscript.

Data availability

The data that support the findings of this study are available from the corresponding authors upon reasonable request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41401-024-01438-x.