Targeting the phenotypic switch of vascular smooth muscle cells to tackle atherosclerosis

Vascular smooth muscle cells (VSMCs) play fundamental roles in the pathophysiology of atherosclerosis. A crucial step in VSMC biology is the switch from contractile (differentiated, quiescent, non-migratory) to synthetic (dedifferentiated, proliferative, migratory) phenotypes (Fig. 1). The exact mechanisms underlying VSMC phenotypic switch are not fully known and various pieces of information have been recently added to this puzzle [1–5]. For instance, in this issue of Atherosclerosis, Bente Halvorsen’s team elegantly demonstrates that Neil3, known to participate in DNA repair, can mediate VSMC phenotypic switch via non-canonical mechanisms [6].

Fig. 1.

VSMC phenotypic switch.

The cartoon summarizes the main mechanisms underlying the phenotypic switch of vascular smooth muscle cells (VSMCs) from contractile to synthetic. ECM: extracellular matrix; IL-1: interleukin-1; KLF4: Krüppel-like factor 4; MYOCD: myocardin; PDGF: platelet-derived growth factor; TGF-β1: transforming growth factor beta 1; TNF-α: Tumor necrosis factor alpha; SRF: serum response factor.

Intriguingly, Neil3 deficiency was shown for the first time to trigger a shift in VSMC phenotype towards a proliferating, lipid-accumulating, and secretory macrophage-like cell phenotype, without major changes in DNA damage [6].

We report herein a concise overview of the main mechanisms regulating VSMC phenotypic switch, mainly relying on transcription factors and epigenetic mechanisms.

1. Regulation of transcriptional activity in VSMC phenotypic switch

The main transcription factors involved in VSMC phenotypic switch are myocardin (MYOCD), Krüppel-like factor 4 (KLF4), and octamer binding transcription factor (Oct4).

MYOCD, a potent cardiac and smooth muscle tissue-specific transcriptional coactivator of serum response factor (SRF), plays a vital role in inducing smooth muscle differentiation. Numerous downstream muscle-specific genes including Sm22α, Calponin, and Acta-2 are activated following the binding of MYOCD to the SRF on the CArG DNA motifs [7]. MYOCD-related transcriptional factors A and B (MRTFs) activate SRF-driven transcription. Several studies have shown that the overexpression of MYOCD and MRTFs induces anti-proliferative effects [8] while their inactivation promotes cell growth and oncogenesis [9]. Equally important, reduction of MRTFs inhibits normal cell cycle progression resulting in an anti-proliferative effect [10]. As a result of this effect, changes such as shortening of G₁ phase and extension of S and G₂ phase occur under normal growth conditions, whereas under serum-starved conditions there is aberrant entry into the S and G₂ phases without any subsequent cell division [11]. MYOCD, along with DNA repair factors like thymine-DNA glycosylase (TDG), recognizes DNA damage in actively expressed areas of the genome and repairs the region-specific DNA [12,13]. Proliferating cells have high expression of TDG and thus instructs the cells to repair DNA mismatches. Moreover, proliferating smooth muscle cells express increased levels of TDG that inhibits both the antiproliferative and promyogenic activities of MYOCD [14].

KLF4 plays an essential role in VMSC proliferation and differentiation [15]. Among the numerous transcriptional regulators that mediate responses to inflammation, proliferation, and differentiation in vascular tissues, KLF family of zinc finger transcription factors have been gaining attention, especially KLF4. There are three different domains of KLF4: transcriptional repression, transcriptional activation, and DNA-binding domains. There are speculations that KLF4 might have anti-proliferative properties because increased levels of KLF4 in numerous cell types have shown growth arrest, reduced cell cycle progression, and inhibition of DNA synthesis [16]. KLF4 regulates gene expression by recognizing and binding to GC-rich DNA sequences or CACCC (GT-box) through its DNA-binding domain. However, studies have demonstrated that KLF4 may transition from pro-cell survival to pro-cell death in certain circumstances [17]. After gamma-irradiation-induced DNA damage [18], KLF4 expression has been shown to be activated in a p53 dependent fashion. The role of p53 is to induce an arrest in the G [1] and S phase of the cell cycle following DNA damage, thereby preventing the irradiated cell to undergo apoptosis [18].

Oct4 is a transcription factor primarily expressed in germline cells that is required to maintain embryonic stem cell (ESC) pluripotency. Oct4 transcription plays a critical role for inner cell mass of blastocysts formation, and loss of expression of Oct4 is associated with spontaneous ESC differentiation; moreover, Oct4 regulates VSMC phenotypic transition and plays an important role in the pathogenesis of atherosclerosis [5,19]. Downregulation of Oct4 contributes to cell cycle regulation of ESCs by blocking G₀/G₁ cycle and preventing proliferation. Further, Oct4 downregulation induces p21 in ESC; inhibition of CDK2 induces G₁ phase arrest followed by apoptosis or differentiation of impaired ESC and causes DNA damage response [20].

2. Role of epigenetics and non-coding RNAs in VSMC phenotypic switch

Mounting evidence has demonstrated that VSMC phenotype switching is regulated by non-coding RNAs, including miRNAs [21], small (approximately 22 nucleotides) single-stranded non-coding RNA molecules that repress protein expression via binding to the 3’ untranslated region of specific mRNAs and either blocking the transcription or promoting mRNA degradation [22].

The contractile phenotype of VSMCs is sustained by miR-143/145 and miR-1; miR-1 expression is triggered by the master regulator of contractile phenotype, MYOCD [23,24]; miR-1 targets the 3′UTR of Kruppel-like factor 4 (Klf 4) – a transcriptional factor that inhibits MYOCD and induces the expression of genes switching VSMCs to a macrophage-like phenotype [24,25]. Moreover, miR-1 limits VSMC proliferation targeting Proto-oncogene serine/threonine-protein kinase Pim-1 and Histone deacetylase 4 (HDAC4) [23]. Some data suggest that an augmented miR-1 expression may also play a role as a negative feedback loop [26].

Similarly, miR-143/145 expression mirrors VSMC differentiation [27]. Their expression is driven by serum response factor (SRF) – an indispensable coactivator of MYOCD [28,29]. In vivo experiments demonstrated that miR-143/145 deficiency results in VSMCs being locked in the synthetic phenotype [27]. miR-143/145 can target and downregulate Klf4, ETS Like-1 protein (Elk-1), and angiotensin converting enzyme (Ace) [27,28]. Moreover, miR-143/145 downregulation by cholesterol was demonstrated in vitro, and VSMCs with a low miR-143/145 expression were found in atherosclerotic plaques in vivo [28,30].

miR-221 promotes cell proliferation via repression of p27^Kip1 and inhibits the contractile phenotype via targeting Proto-oncogene c-kit and [31]. VSMCs proliferation is also induced by miR-222 which targets the 3′UTR of p27^Kip1 and p57^Kip2 [32]. When both miR-221 and miR-222 are upregulated, they can trigger a transformation of VSMCs in an osteogenic phenotype, characterized by high calcification activity [33]; miR-21 and miR-124 were also found to cause dedifferentiation of VSMCs targeting Specificity protein 1 (Sp1) expression [34].

Several miRNAs were shown to play an opposite role, favoring the contractile phenotype. One of the most potent factors in this sense is Platelet derived growth factor BB (PDGF-BB) [35–38]: PDGF-BB induces expression of miR-24 and miR-26a and the inhibition of these miRNAs mitigated PDGF-BB induced proliferation and dedifferentiation of VSMCs [31,39,40]. For instance, miR-24 was shown to suppress the contractile phenotype of VSMCs via repression of Tribbles-like protein-3 (Trb3), a negative regulator of NFκB, which in turn downregulates MYOCD transcription [39,41].

Another class of non-coding RNAs – long-non-coding RNA (lncRNAs) – was recently identified as a regulator of VSMCs phenotype. Among them are SMILR and MALAT1 [42]; the mechanism of their action is not fully understood, however, most likely it involves “sponging” or acting as a decoy for miRNAs [43]. Other lncRNAs that have been shown to modulate VSMC biology are CARMN, PEBP1P2, PVT1, NEAT1, AK098656, and ANRIL [2,44–48].

Epigenetic mechanisms also play significant roles in establishing the VSMCs phenotype. The expression of key genes defining the contractile phenotype of VSMCs – Acta2 and Myh11 – are controlled via histone acetylation in SRF binding regions [49]. MYOCD and KLF4 activities heavily rely on HDACs recruitment [12,13]. Methylation of histones H3 and H4 also increases SRF binding to its target regions [12]. Interestingly, di-methylation of lysine 4 of histone H3 (H3K4me2) in Myh11 could be used as a highly specific marker of non-differentiated VSMCs both in human and mice [50].