Introduction
Vascular smooth muscle cells (VSMCs)are highly specialized cells that play crucial roles in the cardiovascularsystem under normal and numerous pathologic conditions. In arteries,the main function of VSMCs is to provide structural support to thevasculature and regulate blood vessel tone and diameter, blood pressure,and blood flow distribution, under normal conditions. VSMCs are quiescentand typically express a unique range of contractile proteins, includingsmooth muscle cell actin (ACTA2), smoothelin, SM22α, and smoothmuscle cell myosin heavy chain (MYH11) (Owens, 1995).
The origin of VSMCs is far more diversethan originally thought. They develop from a wide range of embryonictissues. Lineage mapping studies have shown that the vascular smoothmuscle in developing vertebrate embryos is a mosaic tissue. Differentvessels, or even different segments of the same vessel, are composedof subtypes of VSMCs arising from different embryonic tissues (
Majesky, 2007). For example, neuralcrest-derived VSMCs constitute the ascending aorta and the aorticarch, while VSMCs in the descending thoracic aorta, the abdominalaorta, and distal portions of the internal carotid arteries are mesoderm-derived(
Majesky, 2007). Theorigin of coronary vascular smooth muscle is thought to be the pro-epicardium(
Mikawa and Gourdie, 1996). Thus, it is essential to emphasize that the underlying mechanismscontrolling the differentiation of VSMCs from different progenitorsare likely to be different (
Shi andChen, 2016). VSMCs contribute significantly to formationof the neointima in vascular diseases; in this classical view intimalcells are derived from medial VSMCs (
Baumgartner and Studer, 1963;
Stemerman and Ross, 1972;
Schwartz et al., 1975;
Clowes et al., 1983;
Regan et al., 2000). However, recentstudies have suggested that adventitial VSMC progenitor cells alsocontribute to formation of the neointima. Several cellular modelshave been established to address VSMC origin in development and pathogenesis(see below). In this review, we provide a brief summary of the currentunderstanding of the regulatory mechanisms in VSMC phenotypic switchingand their role in vascular diseases.
VSMCs in vascular diseases
The blood vessel wall is composedof three layers: a single layer of the endothelium which constitutesthe intima, the media which is composed of VSMCs, and the ECM synthesizedby these VSMCs, and the adventitia which is a complex layer consistingof fibroblasts, collagen fibers, immune cells, and nerves. Blood vesselscontinuously generate mechanical signals and biochemical factors,and in response to these, VSMCs are involved in regulating physiologicfunctions as well as pathological changes taking place in the vascularwall (
Alexander and Owens, 2012). Unlike skeletal and cardiac muscle cells which are terminallydifferentiated, VSMCs maintain a high degree of plasticity both
in vivo and
in vitro. Following an injury such as angioplasty, the insertion of a stent,or in vascular diseases, VSMCs dedifferentiate, reduce the expressionlevels of MYH11, ACTA2, and cause a series of other conventional ‘VSMCmarkers’ in normal blood vessels, as has been previously described(Owens,1995;
Wang and Olson, 2004). During this so-called ‘phenotypic switching’ process,VSMCs dedifferentiate from a ‘contractile’ state to ahighly proliferative ‘synthetic’ state. VSMCs that undergophenotypic switching show increased rates of proliferation, migration,and synthesis of extracellular matrix (ECM) components, while at thesame time also acquiring macrophage markers and properties (
Rzucidlo et al., 2007;
Shi and Chen, 2016). This dedifferentiatedphenotype plays an important role in a large number of major diseasesin humans, including atherosclerosis, restenosis, calcification, andtumor development. Hence, unraveling the molecular regulatory pathwayinvolved in VSMC phenotypic switching should provide useful insightsinto the pathogenesis of cardiovascular diseases.
Atherosclerosis
Atherosclerosis is a chronic progressiveinflammatory disease characterized by the formation of plaque in theintima of medium-sized arteries. It is the leading cause of morbidityand mortality worldwide (
Virmaniet al., 2000;
Libbyet al., 2011). Atherosclerotic lesions most often gothrough a partial resolution process characterized by the formationof a fibrous cap. This fibrous cap serves as a ‘protective’barrier to separate platelets in the blood stream from pro-thromboticmaterials in the plaque. The thickness of the fibrous cap, and theextent of cap inflammation determine the stability of the atheroscleroticplaque. Acute rupture of unstable or ‘vulnerable plaques’frequently results in acute thrombotic vascular diseases, like myocardialinfarction and stroke (
Legein et al.,2013;
Vilahur and Badimon,2013). The historical view of the role played VSMCsin atherosclerosis is that VSMCs in advanced plaques are completelybeneficial in preventing rupture of the fibrous cap. However, recentstudies have shown that the role of VSMCs within the atheroscleroticplaque is determined by the balance between cell proliferation andmigration and between cell senescence and cell death (
Lacolley et al., 2012). It has beensuggested that VSMC proliferation may be predominantly protective,not just in advanced lesions but throughout the entire process ofatherosclerosis.
The molecular regulation of VSMCphenotypic switching in atherosclerosis has been discussed extensively,and there has been significant progress in illuminating the underlyingmechanisms. One of the most classic discoveries was of serum responsefactor (SRF), which is a muscle-specific transcription factor knownto drive VSMC-specific gene expression (Owens et al., 2004). In addition,myocardin (MYOCD), a powerful myogenic coactivator that associatesfirmly with SRF, and stabilizes the binding of SRF at the degenerateCC(A/T-rich)6GG(CArG) cis-elements of all known CArG-dependent VSMCmarker genes, is also involved (
Wanget al., 2003b). The MYOCD-SRF regulatory module is acentral component of the regulatory mechanism that facilitates combinatorialinteractions between activating and repressing signals/co-factorsthat act on most VSMC contractile genes (
Bennett et al., 2016). Studies have shown that ApoE
−/− mice having a MYOCD-specific conditionalknockout exhibit increased atherosclerosis associated with VSMC phenotypicswitching, compared to ApoE
−/− mice lacking the MYOCD-specific conditional knockout. Loss of myocardintherefore represents a crucial permissive step in the process of phenotypictransition and inflammatory activation, including increased macrophagerecruitment, at the onset of vascular disease (
Ackers-Johnson et al., 2015).
In addition, Krüppel-like factors(KLFs), a group of transcription factors principally involved in regulatingcell growth, differentiation, proliferation, and apoptosis, also playimportant roles in the progression of atherosclerosis (
McConnell and Yang, 2010). Investigatorshave found that loss of KLF4 in VSMCs
in vivo results in a transient delay in phenotypic switching following vascularinjury (
Yoshida et al., 2008). Recent studies using ApoE
−/− mice have shown that a VSMC specific conditional knockout of KLF4results in significant reductions in plaque size with increased fibrouscap thickness, and increases multiple indices of plaque stability.In addition, the plaques in the these knockout mice contained reducednumbers of VSMC-derived mesenchymal stem cell-like cells, as wellas macrophage-like cells, suggesting that KLF4 promotes the transitionto a ‘macrophage’ phenotype (
Shankman et al., 2015). Similarly, studies have foundthat another KLF family member, KLF5, induces the phenotypic conversionof VSMCs and facilitates loss of contractile function in smooth muscle(
Kim et al., 2015).Dramatically, recent studies have shown that the transcriptional activitiesof myocardin, KLF4, KLF8, and KLF5 orchestrate the mechanism of theVSMC phenotype switch (see below) (
Ha et al., 2017). More extensive studies have shownthat KLF4 also has a critical role in the generation of vascular progenitorcells in the adventitia. A group of vascular progenitor cells in theadventitia of ApoE
−/− miceand human vessels that express the progenitor markers Sca1 and CD34(AdvSca1 progenitors) can differentiate into VSMCs
in vitro (
Hu et al., 2004;
Zenginet al., 2006;
Campagnoloet al., 2010). Interestingly, further studies by Majesky’sgroup have established that AdvSca1 cells are generated from VSMCsand that generation of AdvSca1 cells is KLF4-dependent both
in vivo and
in vitro (
Majesky et al., 2017). VSMC-derived AdvSca1 cells can differentiate into macrophage-likecells and endothelial-like cells
in vivo and play important roles in arterial homeostasis and disease. Insummary, these studies all indicate that KLF4 plays a key role inatherosclerosis and other vascular-related diseases.
Besides the aforementioned transcriptionfactors, several other stimulating cues that modulate VSMC plasticityin atherosclerosis have been reported, including growth factors andinflammatory mediators (
Beamish etal., 2010). For example, platelet-derived growth factor(PDGF), tumor necrosis factor α (TNFα), and interleukin-1β(IL1β) stimulate the phenotypic switching from the contractiletype to the synthetic type (
Ha etal., 2015). Conversely, the synthetic VSMC phenotypeis significantly converted into the contractile VSMC phenotype bystimulation of the insulin and insulin-like growth factor-1 (IGF-1)signaling pathways (
Wang et al.,2003a;
Hayashi et al.,2004). Therefore, the inflammatory response is closelyassociated with these VSMC phenotypic conversions in disease states.Previously, investigators have reported that that TNFα playsa critical role in vascular remodeling. For example, a deficiencyin TNFα reduces the thickness of vascular walls and the sizeof atherosclerotic lesions in TNFα/ApoE double knockout mice(
Ohta et al., 2005).However, the mechanism underlying the TNFα-induced phenotypicconversion of VSMCs is still unclear. Recent data from a study conductedby Ha et al. (2017) suggest that KLF8 stimulates the differentiationof VSMCs by enhancing myocardin, KLF4, and NFkB, while suppressing KLF5; In the presence of TNFα,the expression of KLF8 is downregulated, thereby relieving KLF5 repression,inducing VSMC dedifferentiation and promoting vascular disease (Fig.1).
Although the majority of studieshave focused on the intracellular signals within VSMCs that regulatephenotypic switching, VSMCs synthesize ECM that separates VSMCs fromeach other. The conventional view is that the ECM maintains VSMCsin the ‘contractile’ state and suppresses phenotypic switching.However, ECM, collagen, and elastin can be broken down by matrix metalloproteinases(MMPs) released from macrophages and VSMCs. This process will boostphenotypic switching and promote cell proliferation and migration(
Koyama et al., 1996;
Li et al., 1998;
Fukumoto et al., 2004). However,the mechanisms by which the ECM affects VSMCs may be more complexthan previously thought. For example, fibronectin (FN) is one of theearliest ECM proteins deposited at atherosclerosis-prone sites, andit is thought to promote atherosclerotic lesion formation. A studyhas demonstrated that while FN worsens the process of atherosclerosisby increasing the atherogenic plaque area, it also facilitates theformation of the protective fibrous cap (Rohwedder et al., 2012).Moreover, although VSMCs produce and respond to collagens in vitro,there is no direct evidence
in vivo that VSMCs are a crucial source of collagens and impact lesion developmentor fibrous cap formation. A recent study has demonstrated that a VSMC-derivedcollagen, collagen type XV (COL15A1), is critical for atheroscleroticlesion development. Interestingly, previous studies have also reportedthat VSMC
Col15a1 knockout resultsin a dramatic attenuation rather than an exacerbation of atheroscleroticlesion formation (
Durgin et al., 2017). This study is the first to examine how knocking out a given ECMgene exclusively in VSMCs impacts lesion pathogenesis
in vivo.
There are multiple cell types involvedin the process of atherosclerosis, including macrophages, lymphocytes,endothelial cells (ECs), and VSMCs. In particular, macrophages andVSMCs are key contributors (
Tabaset al., 2015). The conventional viewpoint is that plaquesare composed primarily of macrophages, and that these macrophage-derivedfoam cells are more vulnerable to rupture compared to VSMCs (
Virmani et al., 2000;
Libby et al., 2011). In other wordsVSMCs in advanced plaque lesions are usually considered to be athero-protective,whereas macrophages are considered to be athero-promoting with respectto plaque stabilization. However, due to limitations, this viewpointmay be incorrect. In advanced plaques, there is an over-dependenceon markers that are either not specific or are upregulated in othercell types. For example, a study has shown that approximately 50%of foam cells in advanced human coronary artery lesions express theVSMC-specific marker, SM α-actin. Nevertheless, most of thesecells also express the macrophage marker CD68, and thus, it is uncertainwhether these cells represent VSMC-derived cells expressing macrophagemarkers or are macrophages expressing VSMC markers (
Allahverdian et al., 2014). Lineagetracing studies in advanced plaques in ApoE
−/− mouse have also indicated that medial VSMCs can convert to macrophage-likecells and constitute a major component in advanced atheroscleroticlesions. These macrophage-like cells express various macrophage markersincluding Mac-2/LGLS3 and CD68, and simultaneously they reduce theexpression of classic VSMC markers, such as ACTA2 (
Feil et al., 2014). Of note, an editorialon this latter paper emphasized that the labeling efficiency of VSMCsin these studies was too low to illustrate what proportion of macrophage-likecells were derived from VSMCs, and how these cells contribute to lesionpathogenesis in advanced lesions (Swirski and Nahrendorf, 2014). Similarstudies have found that cholesterol loading of VSMCs converts themto a macrophage-like state by downregulating the miR-143/145-myocardinaxis, which positively regulates the master VSMC differentiation transcriptionfactor myocardin. Although these cells would be classified as macrophagesby immuno-histochemistry analysis, their transcriptome and functionalproperties imply that their contribution to atherogenesis is not thatof classical macrophages (
Vengrenyuket al., 2015). In summary, the results from all thesestudies support the conclusion that most previous studies have misidentifiedVSMCs and macrophages in atherosclerotic lesions and that VSMCs exhibitan even greater degree of plasticity than previously recognized. Nevertheless,VSMCs and macrophages are inextricably linked in atherosclerotic lesions.Consistent with this view, recent experiments using a trans-well chambermodel have shown that VSMC-macrophage communication can induce changesin the protein composition of the ECM by significantly decreasingthe expression of ECM proteins (collagen I, III, elastin) in VSMCs,while increasing the expression and activity of metalloprotease MMP-9and expression of the collagenase MMP-1 in both macrophages and VSMCs.As discussed above, these changes result in the progression of theatheroma toward a ‘vulnerable plaque’ (
Butoi et al., 2016), suggesting thattargeting the smooth muscle-to-macrophage transdifferentiation orcellular cross-talk in the atherosclerotic plaque may be a novel therapeuticstrategy to slow-down or retard plaque progression.
Calcification
Vascular calcification (VC) is anactive and complicated process regulated by multiple factors thatinvolves several cytokines and their associated signaling pathways.VC is widely seen in numerous vascular-related diseases, includingatherosclerosis, chronic kidney disease (CKD), hypertension, and diabetes(
Bessueille and Magne, 2015). The conventional viewpoint suggests that vascular calcificationmainly occurs in the arterial intima and media. However,
in vitro studies have observed calcifiedlesions in the aortic adventitia in humans, suggesting that adventitialcalcification may arise from fibroblasts which had been transformedinto myofibroblasts or VSMCs (
Li etal., 2015). There are numerous vascular cells, includingVSMCs, myofibroblasts, vascular mesenchymal progenitors, and ECs thathave been shown to be involved in vascular calcification (
Steitz et al., 2001;
Bostrom et al., 2011). VSMCs aremajor participants in vascular calcification and undergo an osteogenicphenotype switch characterized by the loss of VSMC-specific markersand acquisition of osteogenic markers, including Runx1/Cbfa1, osteopontin,osteocalcin, and alkaline phosphatase (
Steitz et al., 2001;
Speer et al., 2009). Apoptosis of VSMCs also promotesthis deposition. Osteoblast-like cells promote calcification of thecollagen in the ECM, ultimately leading to increased stiffness anddecreased pliability of the arterial wall (
Shanahan et al., 2011).
It is generally accepted that serumphosphate and calcium levels, magnesium, bone morphogenetic proteins(BMPs), leptin, and inflammatory cytokines drive this phenotypic transitionby inducing oxidative stress in VSMCs. In addition, other vasculartissue factors such as adrenomedullin, C-type natriuretic peptide,angiotensin, and aldosterone are also involved in regulating vascularcalcification (
McCarty and DiNicolantonio,2014). It is worth mentioning that expression of theosteogenic transcription factor Runx2 plays a critical role in VSMCcalcification
in vitro (
Chen et al., 2006;
Byon et al., 2008). Runx2 belongsto the runt-related transcription factor family, and its expressionis significantly increased in calcified vascular tissue compared withnormal vascular tissue, suggesting that Runx2 may be important invascular calcification (
Steitz etal., 2001;
Tyson etal., 2003). Runx2 either directly or indirectly promotesthe expression of many proteins found in osteoblasts that regulatethe extracellular deposition of hydroxyapatite (
Liu and Lee, 2013). Studies usingsmall interfering RNA, as well as VSMC-specific Runx2 deficient mousemodels, have demonstrated that Runx2 is essential for oxidative stress-inducedVSMC calcification, and is sufficient to induce VSMC calcificationby itself, both
in vitro and
in vivo (
Byon et al., 2008;
Sun et al., 2012). In addition, elevated serum phosphateand calcium, two other vital mediators of mediators of VC, have beenindependently correlated with inflammation. Using incubation of humanaortic smooth muscle cells (HASMCs) in a high phosphate medium, arecent study has demonstrated that elevated calcium and phosphatelevels have direct effects on vascular calcification, accompaniedby the activation of NF-kB signaling,increased expression of the pro-inflammatory mediators, and increasedproduction of reactive oxygen/nitrogen species (ROS/RNS). Therefore,these studies provide evidences for novel mechanisms whereby highphosphate can directly trigger vascular calcification (
Martinez-Moreno et al., 2017).
Restenosis
Restenosis refers to an abnormal(>50%) narrowing of the vessel diameter compared with the normal vessel,and often occurs following percutaneous angioplasty procedures. VSMCaccumulation in the arterial intima is an important event in the pathogenesisof post-angioplasty restenosis (
Glassand Witztum, 2001;
Marxet al., 2011). Restenosis is mainly caused by formationof the neointima. Injury to the arterial wall induces the dedifferentiation,migration, and proliferation of medial-derived VSMCs, and initiatesan inflammatory response, all of which contribute to restenosis (
Mitra and Agrawal, 2006). Thus, themolecular mechanisms underlying the pathological restenosis responseand the formation of the neointima have been widely investigated,and consequently have remarkably improved the treatment of patientswith coronary artery disease.
During the development of restenosis,several processes similar to those occurring during atherosclerosispromote a dedifferentiated VSMC phenotype (
Newby and Zaltsman, 2000;
Schober, 2008. Both are characterizedby activation of VSMCs, resulting in an inflammatory environment,neointimal hyperplasia, and vessel occlusion (
Libby et al., 2011;
Alexander and Owens, 2012). Phosphataseand tensin homolog (PTEN), which is a dual-specificity protein andlipid phosphatase, has been implicated as a negative regulator ofVSMC proliferation and injury-induced vascular remodeling. PTEN functionsas a cytoplasmic lipid phosphatase to regulate both basal and growthfactor-stimulated PI3-kinase/Akt-mediated signaling (
Dahia, 2000;
Vazquez et al., 2000;
Tamguney and Stokoe, 2007). Thus,changes in VSMC PTEN/Akt signaling following vascular injury are relatedto increased VSMC proliferation and neointima formation. Furgesonet al. have shown that carotid ligation in mice having a VSMC-specificheterozygous PTEN deletion resulted in enhanced neointima formation,increased VSMC hyperplasia, reduced expression of VSMC markers (α-SMAand calponin), increased PI3-kinase/Akt/mTOR signaling, and NF-kB activity compared with wild-type mice(
Furgeson et al., 2010). These data indicate that inactivation of PTEN exclusively in VSMCspromotes a pro-inflammatory phenotype and enhances neointima formation.Consistent with this viewpoint, a further study, using both mousegenetic models and in vitro approaches, demonstrated that PTEN isan essential regulator of the transcriptional activity of SRF whichplays an important role in the mechanism that dynamically regulatesthe expression of VSMC contractile genes (
Horita et al., 2016). Thus, loss of the PTEN-SRF axispromotes reprogramming of VSMCs into a proliferative, and inflammatoryphenotype. Therefore, targeting PTEN-SRF nuclear interactions hasthe potential to produce novel therapeutics that preserve the maturedifferentiated VSMC phenotype in order to inhibit in-stent restenosis.Collectively, these data support the conclusion that an alterationin VSMC PTEN signaling acts as a critical initiating determinant drivingpathological vascular remodeling.
Abdominal aortic aneurysms (AAAs)
AAA is a progressive degenerativedisease with no available pharmacological treatment which may resultin significant morbidity and mortality, occurring mainly in olderadults (
Baxter et al., 2008). AAA is defined as an acquired focal dilation (aneurysm) greaterthan 1.5 times the normal size of abdominal aorta.
Extensive studies have suggestedthat inflammation in the aortic wall causes the production of localinflammatory mediators, with macrophages and VSMCs releasing proteasessuch as MMPs in response to cytokine production. These events leadto the breakdown of the ECM proteins, collagen and elastin, whichis followed by VSMC apoptosis in the later stages of the aneurysm(
Ailawadi et al., 2009). Thus, loss of VSMCs and the resulting degradation of collagenand elastin are now viewed as critical steps in aneurysm developmentand progression (
Ailawadi et al., 2003;
Owens, 2007). However,as an important contributor to aneurismal change, less attention hasbeen paid to the role of VSMCs compared to the role of inflammation.There is growing evidence that aortic VSMCs have the potential todirectly participate in the degenerative process. As an example, recentstudies have demonstrated that AAA-SMCs have a unique gene expressionprofile and a so called pro-elastolytic phenotype, which is able todegrade significantly more insoluble elastin than cells derived fromnormal aorta, that directly contributes to the pathogenesis of theaneurysm (
Airhart et al., 2014).
Other cell signaling mechanisms involvedin the formation of these aortic aneurysms remain incompletely understood.TGF-β signaling is known to regulate VSMC growth and apoptosis,MMP-dependent proteolysis, and vascular inflammation. TGF-β signalingis widely recognized as being an important signaling pathway in theinitiation of aneurysms and their progression, although, interestingly,the role of TGF-β signaling is still controversial (
Wang et al., 2013). Early studieshave shown that increased TGF-β signaling leads to aneurismaldilatation in AAAs (
Fukui et al.,2003). In contrast, Wang et al. have shown that TGF-βacts as a protector against inflammatory AAA progression and complicationsin an angiotensin II-infusion mouse model (
Wang et al., 2010). Yet, similarstudies using a smooth muscle-specific disruption of the TGF-βtype II receptor (Tgfbr2) animal model has suggested that TGF-βsignaling in VSMCs contributes to the pathogenesis of elastase-inducedAAA, and that disruption can prevent its formation (
Gao et al., 2014). How can we explainthese contradictory findings? A variety of molecules that may playopposing functions exist in the TGF-β pathway. According to thelocations of the different aneurysms (e.g. thoracic vs. abdominal)and the major types of cells (SMCs vs. inflammatory cells) that areregulated by TGF-β signaling, TGF-β signaling can act asa master upstream modulator that regulates both pro- and anti-inflammatorypathways (
Li et al., 2014). This may explain why TGF-β1 has been reported to promotethoracic aortic aneurysm development, while at the same time playinga protective role in AAA development (
Wang et al., 2013). This protective role of TGF-β1in AAA development was addressed in a recent study that showed thatthe expression of TGF-β1 is markedly elevated in aneurysmal tissuegroup compared to nonaneurysmal tissue (
Doyle et al., 2015).
Resident vascular progenitor cells and VSMCs
It is widely accepted that VSMCscontribute significantly to neointimal formation following vascularinjury. The origin of these neointimal VSMCs, however, is still underdebate. The fact that most intimal cells express VSMC marker genes,including α-SMA and SM22α, after vascular injury led tothe hypothesis that intimal cells are derived from medial VSMCs (
Baumgartner and Studer, 1963;
Stemerman and Ross, 1972;
Schwartz et al., 1975;
Clowes et al., 1983;
Regan et al., 2000). However, thishypothesis was challenged by the finding that adventitial cells initiallyrespond to injury by increasing cell proliferation and then latermigrating into the neointima (
Holifieldet al., 1996;
Scottet al., 1996;
Shi etal., 1996;
Mason etal., 1999;
Oparil etal., 1999). An early report showed that cells from theadventitial side of uninjured canine carotid arteries proliferate
in vitro and express α-SMA rather thanSM-MHC. In contrast, VSMCs from adult carotid media did not proliferate
in vitro, and maintained expression of VSMCmarker proteins (
Holifield et al.,1996). Direct evidence for the ability of adventitialcells to migrate through the media into the neointima was obtainedby seeding these adventitial cells onto the adventitial side of anartery and detecting the movement of these cells following arterialendothelial injury (
Mason et al.,1999;
Li et al., 2000;
Hu et al., 2004;
Rodriguezmenocal et al., 2009). Inthis experiment, β-galactosidase was used to label adventitialcells prior to endothelial injury of rat common carotid arteries.As a result, β-galactosidase-positive cells were observed withinthe medial layer after 3 days, and in the neointima 7 and 14 daysafter endothelial injury (
Holifieldet al., 1996).
In recent years, numerous reportshave demonstrated the existence of resident cardiovascular progenitorcells (
Kovacic and Boehm, 2009;
Campagnolo et al., 2010;
Orlandi and Bennett, 2010;
Hu and Xu, 2011;
Majesky et al., 2011a, 2011b;
Psaltis et al., 2011;
Torsney and Xu, 2011;
Plass et al., 2012;
Li and Izpisua Belmonte, 2016). Emerging data has suggestedthat several distinct progenitor populations have the capacity todifferentiate into VSMCs. The recruitment of stem/progenitor cellspresent into the vessel wall are largely responsible for VSMC accumulationin the intima during vascular remodeling that occurs following suchevents as neointimal hyperplasia and arteriosclerosis (
Scott et al., 1996; Li, 2000;
Sartore et al., 2001;
Abedin et al., 2004;
Hirschi and Majesky, 2004;
Aicher et al., 2005;
Xu, 2007;
Du et al., 2012). Among all the residentvascular progenitor cells, adventitial Sca1
+ progenitor cells are among the most important (Hu, 2004;
Passman et al., 2008). Hu et al.first identified a Sca1
+ cell residingin the adventitia as being a VSMC progenitor cell (
Hu et al., 2004). When Sca-1
+ cells expressing the
LacZ gene were transferred to the adventitial side of vein grafts inApoE-deficient mice,
b-gal
+ cells were found in the adventitialmedia at 2 weeks, and in the adventitial intima at 4 weeks, wherethey no longer expressed the Sca1 antigen and became immunopositivefor VSMC marker proteins.
In vitro, adventitial Sca1
+ progenitor cells candifferentiate into smooth muscle cells expressing the mature VSMCmarkers α-smooth muscle actin (ACTA2), calponin (CNN1), andsmooth muscle myosin heavy chain (MYH11). Passman et al. have alsoreported that Sca1
+ , CD34
+ , and PDGFRβ
+ cellsresiding in an adventitial niche, characterized by sonic hedgehog(Shh) signaling, could be differentiated into smooth muscle-like cells
in vitro (
Passman et al., 2008). The differentiation of adventitialSca1
+ progenitors into VSMC can be regulatedby PDGF-BB. MMP-811 and stromal-cell derived factor-12 also seem toregulate adventitial Sca1
+ progenitor cellrecruitment during atherosclerosis or neointimal formation after arterialendothelial injury (
Shikatani etal., 2016).
However, the origin of the adventitialSca1
+ progenitor cells remains unknown.Little attention has been paid to the possibility that VSMCs may alsomove in the opposite direction, that is, into the adventitia. Recentreports have demonstrated that a high percentage of VSMCs in atheroscleroticlesions lack detectable expression of conventional VSMC markers, butexhibit a macrophage-like phenotype (
Feil et al., 2014;
Vengrenyuk et al., 2015). These findings suggest thatVSMCs exhibit an even greater degree of plasticity than that previouslyrecognized. Using fate-mapping and lineage-tracing approaches, Majeskyet al. have demonstrated that a distinct subpopulation of adventitialSca1
+ progenitor cells derived from differentiatedVSMCs can undergo a reprogramming-like process in situ to generatemultipotent progenitor cells (
Majeskyet al., 2017).
Other than Sca1
+ cells, Rafael Kramann have demonstrated that Gli1
+ progenitor cells are also an important adventitial cell source forVSMCs and contribute to neointimal formation after acute arterialinjury. Gli1
+ cells located in the arterialadventitia are progenitors of VSMCs and contribute to neointima formationand repair after acute injury to the femoral artery. Genetic fatetracing has indicated that adventitial Gli1
+ MSC-like cells migrate into the media and neointima during arteriosclerosisin ApoE
−/− mice with chronic kidney disease(
Kramann et al., 2016).
Furthermore, Tang et al. isolatedcells from the medial layer of the blood vessel wall and identifieda new type of multipotent vascular stem cell (MVSC) expressing markerswhich included Sox17, Sox10, S100β, and neural filament-mediumpolypeptide (NFM) (
Tang et al., 2012). MVSCs can differentiate into MSC-like cells and subsequently VSMCs.Importantly, lineage-tracing experiments have shown that MVSCs, aswell as proliferative or synthetic VSMCs, are not derived from matureVSMCs. Following vascular injury, MVSCs become proliferative and contributedsignificantly to vascular remodeling and neointima formation. Thesefindings define a novel MVSC-VSMC differentiation pathway, and supporta new viewpoint that the differentiation of MVSCs contributes to vascularremodeling and diseases instead of the de-differentiation of VSMCs
In summary, resident vascular progenitorcells significantly contribute to neointimal formation and vasculardiseases.
In vitro VSMC differentiationmodels
Given that VSMC differentiation playsa critical role in the process of vascular disease, we will next summarizethe in vitro models that have been established to define the molecularmechanisms involved in the regulation of VSMC differentiation andphenotypic modulation.
Mesoderm-derived models
C3H/10T1/2 Cells
The 10T1/2 cell line was generatedfrom primary cultures of 14- to 17-day whole C3H mouse embryos (
Reznikoff et al., 1973). It has beenreported that these multipotent 10T1/2 cells can differentiate intoVSMCs by coculture with ECs or treatment with transforming growthfactor-β1 (TGF-β1, 1 ng/mL) for 24 to 48 h, as evidencedby a phenotypic change from a polygonal to a spindled-shaped phenotype,and the expression of VSMC-specific markers including α-SMA,SM22α, and calponin (
Hirschiet al., 1998). This cell model is most commonly usedfor studying VSMC differentiation because of the ease of acquisitionof these cells from the American Type Culture Collection, undemandingculture conditions (DMEM+ 10% FBS), and rapid induction of differentiationby TGF-β1 (within 48 h).
However, it has been suggested that10T1/2 cells cannot fully differentiate into VSMCs upon TGF-β1treatment, but rather differentiate into myofibroblasts that do notexpress MYOCD and the mature VSMC marker MHY11 (
Yoshida and Owens, 2005;
Shi et al., 2012). These conflictingreports maybe due to the differences in culturing methods and manipulationof these cells in different laboratories. Regardless, the 10T1/2 cellsdifferentiation system has been a useful model for studying the molecularmechanisms underlying TGF-β1-dependent VSMC differentiation.
Neural crest stem cell–derived models
Monc-1 cells
Monc-1 is a neural crest cell linethat was immortalized by retroviral transfection of the
v-myc gene in primary cultures of mouse neuralcrest cells (
Rao and Anderson, 1997). Monc-1 cells can differentiate into neurons, gliacytes, chondrocytes,melanocytes, and VSMCs (
Rao and Anderson,1997;
Jain et al., 1998). It has been shown that Monc-1 cells can differentiate to a VSMClineage following stimulation with TGF-β1, resulting in the inductionof VSMC markers. In contrast, bone morphogenic protein 2 (BMP-2),another member of the TGF-β superfamily, is unable to inducethe expression of VSMC marker genes in Monc-1 cells (
Chen and Lechleider, 2004). Thisinconsistency is probably due to the antagonistic effect of the
v-myc gene on BMP-2 signaling. In additionto TGF-β1, fetal bovine serum (FBS) can also induce VSMC differentiationfrom Monc-1 cells (
Jain et al., 1998). FBS-induced Monc-1 cells exhibit a differentiation state similarto the proliferative synthetic phenotype of VSMCs, rather than thecontractile phenotype, because of lack of a spindle-shaped morphologyand a lack of contraction in response to carbachol (
Jain et al., 1998;
Chen and Lechleider, 2004). The limitationof this model is the requirement of a complicated medium for growthand maintenance of the undifferentiated state. On the other hand,Monc-1 cells expressing the
v-myc proto-oncogene in a constitutive manner may generate undefined effectssuch as abnormal proliferation during the differentiation process.
JoMa1 cells
JoMa1 is an immortalized multipotentneural crest cell line derived from neural crest primary culturesfrom a transgenic mouse line expressing a conditional tamoxifen-inducible
c-Myc oncogene (
Maurer et al., 2007). TGF-β1 treatment, combinedwith tamoxifen removal from the culture medium, results in a lossof
c-Myc and induces the differentiationof JoMa1 cells into VSMCs as evidenced by a change in cell morphology,a strong expression of α-SMA (over 90% of cells) and SM g-actin, and a weak expression of calponin.A clonally derived subline of JoMa1 cells, termed JoMa1.3, showeda purer VSMC population expressing higher levels of CNN1 than itsparental line following TGF-β1 treatment.
Compare to Monc-1 cells, JoMa1 andJoMa1.3 cells do not express
c-Myc during differentiation due to the absence of tamoxifen, which avoidsthe potential interference of this oncogene in the differentiationprogram. In the presence of tamoxifen, TGF-β1 induction wouldlead to incompatibility between the proliferation and differentiationsignals in these cells because the proliferative stimulus via
Myc, and the differentiation stimuli viaTGF-β1, delivered in parallel induces cell death (
Maurer et al., 2007). Although JoMa1is an attractive in vitro model for studying VSMC differentiationin a neural crest linage, it has limitations similar to the Monc-lline with respect to requiring complex culture conditions to maintainits undifferentiated state.
Stem/progenitor cell-derived models
P19 and A404 cells
The P19 cell line is derived froma teratocarcinoma that was formed following transplantation of a 7.5-daymouse C3H/He embryo into the testis (
McBurney and Rogers,1982;
McBurney, 1993). It appears to use mechanisms similarto those of normal embryonic stem cells (ESCs) to differentiate (
McBurney, 1993). Studies have shownthat P19 cells differentiate into fibroblast-like cells upon treatmentwith retinoic acid (RA; 10
−6 mol/L for 48 h) or dimethyl sulfoxide and 7.5% fetalbovine serum (after 5 to 7 days), following which they express smoothmuscle α-actin (ACTA2), exhibit calcium influx patterns, andacquire the ability to respond to phenylephrine, angiotensin II, andendothelin (
Rudnicki et al., 1990;
Blank et al., 1995). Although the P19 cell line can differentiate into functional VSMC-likecells, it still requires additional enrichment methods to increasethe yield of VSMCs.
The A404 cell line is a P19-derivedclonal cell line containing an ACTA2 promoter/intron-driven puromycinresistance gene (
Manabe and Owens,2001). All-
trans RA,combined with puromycin treatment, stimulates a rapid differentiationof multipotent A404 cells into VSMCs that express myocardin and multipleVSMC markers including α-SMA, calponin, and SM-MHC, with a differentiationefficiency of more than 90% (
Manabeand Owens, 2001;
Spinet al., 2004). Although the A404 cell line is a highlyefficient model system for VSMC differentiation, it is not a naturalVSMC progenitor that exists in vivo because it is established by introducingthe α-SMA promoter into the cells.
ESC differentiation system
Embryonic stem cells (ESCs) are pluripotentstem cells derived from the inner cell mass of blastocysts. ESCs candifferentiate into VSMCs in an adherent monolayer culture system (
Huang et al., 2006;
Xie et al., 2009). When treatedwith all-
trans RA, monolayers ofhESC cultures differentiate toward a VSMC lineage, and after up to30 days in culture, 40%–90% of the cells stably express α-SMAand SM-MHC, demonstrate morphological changes consistent with a contractileVSMC phenotype (
Huang et al., 2006;
Xie et al., 2009), and exhibit a smooth muscle-like contraction frequency (
Huang et al., 2006). Moreover, thesedifferentiated cells display functional calcium responses followingtreatment with the vasoconstrictor caffeine and the depolarizing agentKCl (
Xie et al., 2009). To improve the efficiency of VSMC differentiation from ESCs, theESC surface marker Sca-1 can be used to isolate VSMC progenitor subpopulations(
Xiao et al., 2006,
2007). Treatment of ESC-derivedSca-1
+ cells with PDGF has been reportedto induce VSMC differentiation at an efficiency of more than 95% afterlong-term culture on a surface containing collagen IV (
Xiao et al., 2007).
Human embryonic stem cell-derivedmesenchymal cells (hES-MCs), derived from H9 human embryonic stemcells, are natural mesoderm-derived VSMC progenitors (
Boyd et al., 2009). hES-MCs havethe capacity to produce three mesenchymal lineages including osteocytes,chondrocytes, and VSMCs (
Boyd et al.,2009). It has been reported that hES-MCs can robustlydifferentiate into the VSMC lineage upon TGF-β1 induction, accompaniedby a high level of VSMC marker expression including α-SMA, SM22α,SM-MHC, and calponin, as well as contraction in response to the muscarinicagonist carbachol or KCl, suggesting a functional VSMC phenotype (
Guo et al., 2013).
hES-MCs appear to be a more attractive
in vitro model for VSMC differentiation than10T1/2 cells. Similar to 10T1/2 cells, hES-MCs exhibit such advantagesas simple culture methods (αMEM+ 10% MSC-qualified FBS) andrapid differentiation (within 72 h of TGF-β1 induction). Remarkably,unlike 10T1/2 cells, hES-MCs can stably express SM-MHC and myocardinupon TGF-β1 treatment, signifying a full differentiation towarda mature and functional VSMC lineage rather than immature VSMCs ormyofibroblasts (
Guo et al., 2013). Most importantly, hES-MCs may be used to study mechanisms underlyingVSMC differentiation in humans because of their human origin.
iPSC differentiation system
Recently, several studies have establishedmethods for generating VSMCs from induced pluripotent stem cells (iPSCs),such as lateral plate mesoderm-derived, coronary-like, VSMCs frompluripotent stem cells, neural crest-derived VSMCs from skin-derivedprecursors, and contractile or synthetic VSMCs from human iPSCs (
Steinbach and Husain, 2016;
Yang et al., 2016). These methodsmay be very useful for personalized medicine and regenerative medicineapproaches. They may also be used to establish models for hereditaryfamilial syndromes that affect VSMCs such as Marfan’s or Loeys-DietzSyndrome, or even systemic conditions such as progeria (Liu et al.,2011). Compared to ESCs, iPSCs are mainly used for SM tissue engineeringand regeneration to avoid both ethical problems and the recipient'simmune response that are encountered with ESCs. In spite of the potentialstrengths of iPSC-derived VSMCs in the treatment of vascular disease,there are still issues that need to be considered. It has been reportedby
Xie et al. (2009) that there are different VSMC marker expression patterns in iPSCsgenerated by different methods. This may be due in part to the randominsertion of the viral vectors encoding the four transcription factorsused in iPS generation.
In summary, a variety of in vitro VSMC differentiation models areavailable. Careful consideration should be given to their relativeadvantages and their intrinsic limitations to select the model thatis suitable for the experimental purpose.
Perspectives
Our understanding of the role ofVSMCs in cardiovascular diseases has evolved remarkably over the lastfew years. Recent studies have clarified that the development andprogression of numerous cardiovascular diseases are strongly associatedwith different VSMC phenotypes. Thus, further investigation of themechanism behind VSMC phenotype regulation is essential to illuminatethe pathophysiology of VSMC-related cardiovascular diseases to developtherapeutics to treat these diseases. However, many questions remainto be addressed: (1) How is VSMC differentiation from progenitorsof different embryonic origin controlled? In other words, in the invitro differentiation models, what are the mechanisms underlying progenitor-specificVSMC differentiation? (2) Many transcription factors and inflammatorymediators are responsible for stimulating the VSMC phenotype switch.While some evidence indicates that interactions exist between them,what is the mechanism underlying the regulation of transcription factorsby inflammatory cytokines during VSMC phenotypic switching? (3) Asmany of the factors that play important roles in the control of VSMCdifferentiation are also involved in regulating many other cellularprocesses, Is there a cross talk between the diverse cellular processesamong phenotypic switching? (4) What is the relation between VSMCsand other cell types, including ECs, macrophages, and circulatingprogenitor cells under pathological conditions, and how can they bedistinguished? (5) Finally, and most importantly, as direct interventionalstudies to prevent phenotypic switching have been lacking, is it possibleto identify the factors and mechanisms that promote beneficial changesin VSMC phenotype, and processes that can replace more conventionaltherapies in the near future?
Abbreviations
AAA: Abdominal aortic aneurysms;ACTA2: Smooth muscle alpha actin; CNN1: calponin; KLFs: Krüppel-likefactors; EC: Endothelial cells; ECM: Extracellular matrix; MVSC: Multipotentvascular stem cell; MYH11: Smooth muscle-myosin heavy chain; SRF:Serum response factor; MYOCD: Myocardin; VC: Vascular calcification;VSMC: Vascular smooth muscle cell.
Higher Education Press and Springer-Verlag GmbHGermany, part of Springer Nature
PDF
(487KB)
