Development and pathologies of the arterial wall

Abstract

Arteries consist of an inner single layer of endothelial cells surrounded by layers of smooth muscle and an outer adventitia. The majority of vascular developmental studies focus on the construction of endothelial networks through the process of angiogenesis. Although many devastating vascular diseases involve abnormalities in components of the smooth muscle and adventitia (i.e., the vascular wall), the morphogenesis of these layers has received relatively less attention. Here, we briefly review key elements underlying endothelial layer formation and then focus on vascular wall development, specifically on smooth muscle cell origins and differentiation, patterning of the vascular wall, and the role of extracellular matrix and adventitial progenitor cells. Finally, we discuss select human diseases characterized by marked vascular wall abnormalities. We propose that continuing to apply approaches from developmental biology to the study of vascular disease will stimulate important advancements in elucidating disease mechanism and devising novel therapeutic strategies.

Introduction

Vascular development and disease

The networks and walls of blood vessels are constructed in a strictly regulated manner. In this review, we will first discuss the development of the intricate structure of blood vessels and then highlight select human vascular pathologies in which this structure is deficient.

Basic components of the cardiovascular system

As a testament to its importance, the heart and vasculature is the first system to function during morphogenesis. The early embryo must be capable of delivering nutrients to and removing waste products from all of its tissues in order to facilitate proper development of its essential structures. The heart starts beating early in development and generates the force needed to deliver blood through the large-bore arteries to the medium-sized arterioles and eventually to small capillaries. The constituents of blood, including oxygen and nutrients, traverse the thin walls of capillaries to reach the extravascular tissues of end organs and contribute to their physiological processes. In addition to generating substances required for key physiological functions, these processes also generate waste products which must be eliminated. This waste travels from the extravascular space into capillaries and then through the vasculature into venules and veins, and is either metabolized by the liver, excreted by the kidney, or exhaled by the lung.

Structure of the blood vessel

Blood vessels are composed of three tissue layers: the tunica intima, the tunica media, and the tunica adventitia. The intima, or inner layer, consists of a single layer of elongated endothelial cells (ECs) that line the vessel lumen, and the basement membrane is interposed between the ECs and the media (i.e., the middle layer). The media regulates vascular tone, and, depending on the overall thickness of specific vessels, the media contains a different number of layers. In larger vessels, these layers consist of circumferentially oriented smooth muscle cells (SMCs) and extracellular matrix (ECM) components, including elastin and collagen. In contrast, the mural cells of capillaries consist of pericytes instead of SMCs. The tunica adventitia, or outer vessel layer, is an assembly of loose ECM, fibroblasts, nerves, and small arteries known as the vasa vasorum that supply nutrients to the adventitia and outer medial layers of large arteries. Recently, the adventitia has been recognized as a dynamic environment, instrumental in the growth, disease, and repair of the artery. Although the majority of the vascular developmental literature has concentrated on the morphogenesis of EC networks, here, we will briefly discuss the intima and instead focus our attention on the development of the media and adventitia (i.e., the vascular wall) and on diseases affecting those layers.

Model organisms

A number of organisms have been widely utilized to model processes that occur in human development and/or disease. Model organisms that have been especially revealing of human vascular development include Drosophila, chick, quail, zebrafish, mice, and rats. Although it may seem that investigations of Drosophila have little relevance to mammalian vascular biology, there are important molecular and cellular similarities between, for instance, branching and tube morphogenesis of the Drosophila trachea and mammalian vascular systems [1–3]. In addition, the development of chick and quail embryos outside the mother permits physical manipulation as well as cross-species transplantation experiments with molecularly distinct host and donor tissues, to facilitate studies of the origin and fate of vascular tissues. Due to their transparency, short gestational period, widely available transgenics, and rapidly expanding means of facile genetic manipulations, the zebrafish is a particularly useful vascular model [4]. For instance, the study of endothelial tip and stalk cell formation and patterning has been investigated intensely in the zebrafish intersegmental vessels (ISVs), which are apparent as soon as 24 h post-fertilization.

Rodents are the most commonly used mammals for studying vascular development and disease. Due to the wealth of transgenic animals, powerful genetic engineering tools and similarities to human biology, mice have proven especially useful in elucidating the roles of gene products in human vascular development. For instance, the mouse retina, which is vascularized postnatally, is widely used to investigate sprouting angiogenesis. Although lacking the genetic tools of mice, the rat is a significantly larger animal and thus a useful complement, especially for surgical procedures, some disease models and studies of the microvasculature.

Lineage tracing

Lineage tracing and clonal analysis studies are central to the field of developmental biology for generating cell fate maps. Some of the first fate mapping studies used vital dyes to label cells in amphibians in a non-lethal manner [5]. Later studies improved on this technique to incorporate radioactive and fluorescent labels from donor embryos and species-specific differences in heterochromatin to identify and track groups of cells in host embryos [5]. More recently, sophisticated genetic tools have evolved which facilitate the non-invasive marking of specific cell populations with temporospatial precision during development. For instance, these tools include the Gal4/UAS system in Drosophila as well as the Cre-lox system and the tetracycline-controlled transcriptional activation system in mice. Such approaches have also been adapted to discern the origin of cell types in normal maintenance of adult tissues and in models of disease and regeneration. These studies raise many intriguing concepts linking developmental biology to disease, such as the idea that insults can induce the reprogramming of differentiated cell types to other cell fates through the process of cell dedifferentiation or transdifferentiation (Fig. 1) [6]. An important example in vascular biology is the potential reprogramming of ECs to other cell fates, such as SMCs, in development and disease [7–12].

Fig. 1.

Schematic diagram of development and lineage reprogramming of cells in the epigenetic landscape. In normal development, a pluripotent cell (green ball) rolls down bifurcating valleys, which represent all possible developmental paths. The cell makes a series of “choices” and differentiates into a mature cell (blue ball) at the bottom of the valley. Lineage reprogramming includes dedifferentiation and transdifferentiation, where a mature cell takes a step backward to a progenitor stage (cyan ball) or converts directly to another mature cell (yellow ball). Adapted from reference [6]

Pathologies of the arterial wall

A number of important human pathologies characterized by arterial wall dysfunction are discussed in this review. For instance, diseases associated with mutations in the gene encoding the elastin protein and affecting the arterial wall include supravalvular aortic stenosis (SVAS), Williams’ syndrome (WS), and autosomal dominant cutis laxa (ADCL). In addition, aneurysms or abnormal dilations of the aorta predispose to aortic rupture, thrombus, and distal embolization. The Marfan syndrome (MFS) due to fibrillin mutations is an important cause of thoracic aortic ascending aneurysm and follows an autosomal dominant pattern of inheritance. A clinically variable disorder that also includes aortic aneurysms and is associated with mutations in components of the transforming growth factor beta (TGFβ) signaling pathway is the Loeys–Dietz syndrome (LDS). Finally, we discuss atherosclerosis, which is the most common cause of mortality in developed nations and pulmonary hypertension (PH), a devastating disease defined by an elevated pulmonary artery (PA) blood pressure.

Vascular development

Tunica intima: vascular endothelium

ECs derive from the mesodermal germ layer and, thus, the development of blood vessels can be considered to initiate with gastrulation, a process which establishes the three primary embryonic germ layers, the endoderm, mesoderm, and ectoderm. Prior to the onset of gastrulation at embryonic day (E) 6.0–6.5, the mouse embryo is comprised of the hypoblast, which contributes to the extraembryonic endoderm, and the epiblast, the source of both the extraembryonic mesoderm and of all fetal tissue. Gastrulation initiates with the formation of the cranial-caudal primitive streak at the future posterior pole of the embryo. Many cells of the epiblast ingress through the primitive streak to become endoderm and mesoderm, whereas the ectoderm derives from the remaining epiblast.

Studies in which mouse epiblasts have been isolated (and often manipulated) and then allowed to develop in culture suggest that the position of cells in the epiblast as well as their timing and location of migration through the primitive streak are important factors determining mesoderm fate and patterning [13–15]. Although lateral and posterior epiblast cells predominantly become mesoderm [15], a clonal analysis in which E6.7 epiblasts were cultured for ~24 h after single cell labeling indicates that the progeny of an individual epiblast cell can give rise to cells that are not restricted to a single germ layer or only to the extraembryonic mesoderm [14]. The mesodermal fate is apparently established as cells of the epiblast migrate through the primitive streak and undergo an epithelial-to-mesenchymal (EMT) transition [13, 16]. In the early streak stage, migrating cells become the extraembryonic mesoderm, while at the mid-streak stage, progenitors of the cranial and cardiac mesoderm and lateral plate mesoderm (LPM) of the upper body ingress through the primitive streak. Finally, at the late streak stage, the epiblast cells destined to become the presomitic mesoderm (PSM) and trunk LPM ingress through the primitive streak. In addition to the timing of ingress, the location is also apparently important as progenitors for the extra-extraembryonic mesoderm, LPM, and PSM which migrate through the posterior, middle, and anterior segments of the primitive streak, respectively.

The LPM is implicated as the major source of ECs. However, quail chick orthotopic transplantation experiments suggest that the presomitic/somitic mesoderm eventually gives rise to all the ECs of the developing avian descending aorta [17]. Interestingly, the data for the descending aorta in the mouse differ, suggesting that ECs of this vessel derive from the LPM [18] with perhaps a minor contribution from the PSM [19]. Through the process of vasculogenesis, mesodermal cells differentiate into primitive ECs and coalesce into the initial blood vessel tubes. Subsequently, the initial vascular tubes ramify through angiogenesis, and the EC network is refined through pruning and regression (reviewed in [20]).

To form a functional circulatory system that delivers oxygen to developing target tissues, hematopoietic cells develop in parallel with blood vessels. In the early twentieth century, Florence Sabin reported that, in the developing chick, ECs give rise to red blood cells that are released into circulation [21]. Indeed, subsequent studies of human and mouse development indicate that the initial primitive blood cells are predominantly nucleated erythroid progenitors which emerge at a similar time and location as ECs from the extra-embryonic yolk sac [22]. In the murine embryonic yolk sac, this primitive erythropoiesis initiates at ~E7–7.75 followed by definitive hematopoiesis of multi-lineage progenitor cells at ~8.25 [22]. Definitive hematopoiesis initiates in the embryo proper in the the aorta-gonad-mesonephros at ~E10 and subsequently shifts to the fetal liver at ~E11 and finally, prior to birth, to the bone marrow which remains the primary site of postnatal hematopoiesis [22].

In addition to segregating from other mesodermal-derived fates, arterial and venous ECs differentiate to distinct fates during embryogenesis. Historically, the prevailing theory was that the identity of arteries and veins was determined by oxygen levels and hemodynamic factors. More recently, it has been shown that markers specific to arteries or veins are detected in their respective vessels early in the process of vascular morphogenesis. For example, prior to the start of angiogenesis, ephrinB2, a transmembrane ligand, is expressed in the mouse embryo in an arterial-specific manner while EphB4 (a receptor for ephrinB2) is expressed in a relatively venous-specific pattern [23–25]. In the zebrafish embryo, prior to the first heartbeat or initial expression of ephrinB2 and EphB4, presumptive ECs express Notch-related genes such as gridlock (grl), a transcription factor downstream of the Notch receptor [26–28]. Before vessel formation, grl is expressed as two bilateral stripes in the LPM, which contains artery and vein precursors [28, 29], and thereafter, grl is expressed within the artery (dorsal aorta) but not the vein (cardinal vein) of the trunk [28]. In a pivotal study using laser-induced dye uncaging, Zhong et al. [27] showed that, while EC precursors for the zebrafish dorsal aorta and cardinal vein are spatially mixed in the LPM, the progeny of each individual EC precursor is delimited to one of the vessels in an arterial-venous decision guided by grl.

In addition to marking early arterial ECs, Notch pathway members also control arterial EC fate as integral components of a signaling cascade that also involves vascular endothelial growth factor (VEGF) and sonic hedgehog (Shh). Downregulating the Notch pathway in zebrafish results in reduced ephrinB2 expression in the dorsal aorta and defects in blood vessel formation, while contiguous regions of the cardinal vein expand, and EphB4 expression increases within the vein [27, 30]. In contrast, activating the Notch pathway reduces expression of Flt, a marker of venous identity, without changing arterial marker expression or dorsal aorta size [27, 30]. In addition, loss of VEGF or sonic hedgehog (shh) in zebrafish results in loss of arterial identity, while exogenous expression of VEGF and Shh results in expression of arterial markers [31]. Taken together, these studies show that VEGF acts upstream of the Notch pathway and downstream of Shh to control arterial EC fate [31].

The above studies raise interesting questions regarding the timing of specification of arterial and venous EC identity and if and when these identities are absolutely committed. Moyon and colleagues attempted to address these questions by dissecting the dorsal aorta, carotid artery, jugular vein, or cardinal vein from the E2–15 quail and grafting the vessel into an E2 chick coelom [32]. The host embryos were stained with arterial-specific antibodies and the quail-specific anti-EC antibody QH1 to determine whether the grafted vessels generated ECs that colonized host arteries, veins, or neither vessel type [32]. If the quail vessels were harvested until approximately E7 and subsequently grafted into the chick, ECs colonized both host arteries and veins. However, if harvested after E7, plasticity of the grafted vessels decreased; after E10, ECs derived from quail arteries or veins and grafted into the chick almost exclusively contributed to host arteries or veins, respectively [32]. Significantly, when ECs were isolated from the E11 quail dorsal aorta wall by collagenase treatment and subsequently grafted, the plasticity of the ECs returned to that of an E5 vessel, suggesting to the authors that an unknown vascular wall-derived signal modulates EC fate [32]. In a related investigation in embryonic mice, Red-Horse and colleagues showed that EC sprouts from the sinus venosus, the structure that returns blood from the embryonic heart, dedifferentiate as they migrate over and through the myocardium [33]. They suggest that ECs that invade the myocardium differentiate into the ECs of the coronary artery and capillaries, while those that remain superficial redifferentiate into coronary vein ECs [33].

In addition to EC specification and formation of arterial or venous tubular cords, these EC tubes must lumenize to facilitate the delivery of nutrients and oxygen to target tissues as well as to remove waste products. There has been significant controversy regarding two proposed models for the mechanism underlying blood vessel lumenization that may, in fact, simply differ depending on vessel size and type [34]. We will first discuss the recently proposed and more widely accepted extracellular lumenization model in which the vessel lumen forms between two or more ECs [35]. For instance, in the developing aorta, apical surfaces of adjacent ECs face each other and, upon increased apical expression of negatively charged CD34-sialomucins, ECs de-adhere and initiate inter-EC lumen formation perhaps as a result of electrostatic repulsion [35]. EC polarity and shape also play important roles in extracellular lumen formation. EC-specific knockout of β1-integrin causes the mislocalization of polarity proteins [36]. Furthermore, β1-integrin-deficient ECs are cuboidal-shaped and become stratified, occluding the vessel lumen [36]. In contrast, the second model, known as intracellular lumenization, proposes that the vascular lumen forms from a coalescence of vacuoles within a single EC followed by intercellular vacuole fusion. This process has been observed in cultured ECs [37–39] and in zebrafish ISVs [40]; however, recent studies suggest that at least the primary lumen of zebrafish ISVs forms extracellularly at junctions between ECs [34, 41]. Subsequently, expansion of the apical membrane and lumen formation are likely to be driven at least partly by vacuole contribution to the apical domain [34].

Branching of the EC tubes establishes a ramified vascular network and is largely regulated through signaling by VEGF, Notch, and TGFβ. This process is initiated by endothelial tip cells, located at the end of vascular sprouts, and stalk cells that comprise the stem of the budding vasculature. Hypoxia stimulates expression of the ligand VEGF-A, and, in addition to promoting EC survival, VEGF-A induces the initial extension of tip and stalk cells through vascular endothelial growth factor receptor (VEGFR)2-mediated signaling [42]. Tip cells divide rarely if at all and express the platelet-derived growth factor (PDGF)-B ligand, VEGFR2, and the Notch ligand delta-like ligand 4, while stalk cells are proliferative and are enriched with Notch and the homeobox protein HLX1 [43–46]. In comparison to branching and sprouting, much less is known about how angiogenesis terminates; however, the sphingosine 1 phosphate receptor 1 (S1PR1) was recently demonstrated to play an important role in curtailing angiogenesis by serving as an EC-autonomous inhibitor of sprouting [47, 48]. S1PR1 ^(−/−) mice and S1PR1 zebrafish morphants display an increase in EC filopodia formation and angiogenic hypersprouting, and dysregulated VEGF-induced signaling has been implicated as an important underlying mechanism [47, 48]. In addition, signaling through the TGFβ type 1 receptor activin receptor-like kinase 1 (ALK1) has been shown to synergize with activated Notch in stalk cells to attenuate VEGF signaling, tip cell formation, and EC sprouting [49].

Vascular branching is also mediated by pathways that share similarity with branching programs observed in other ramified structures such as neurons, the mammalian lung, and the Drosophila trachea. Indeed, a number of families of axon guidance cues and their receptors play important roles in patterning of EC networks [50]. These ligand–receptor pairs include Slit-Robo, netrin-UNC5B, semaphorin-plexin/neuropilin, and ephrin-EphrinB4 [50]. Airway branching morphogenesis in the mouse and Drosophila is induced by fibroblast growth factor (FGF)-mediated signaling pathways, and these pathways are inhibited by Sprouty (SPRY) [51–53]. In the vasculature, the ligands FGF1 and FGF2 are known to induce angiogenesis in vivo [54], and the SPRY proteins inhibit angiogenesis and EC proliferation [2, 55, 56]. In human umbilical vein ECs, overexpression of Spry1 upregulates cell cycle inhibitors p21 and p27 [56], and Spry4 overexpression has also been shown to increase p21 levels [2]. Furthermore, mice that are null for Spry4 or have undergone knockdown of Spry2 and Spry4 are resistant to ischemic insults due, at least partially, to accelerated neovascularization [55].

In addition to ECs, SMCs have also been described within the intima of blood vessels: in diffuse intimal thickening (DIT) of human vessels as well as in the diseased vasculature. Interestingly, DIT is classified as normal human intima [57] but is also implicated as an atherosclerotic precursor as it is consistently present in regions of arteries that are especially prone to atherosclerosis and absent from atherosclerosis-resistant arteries [58, 59]. In addition, animals that do not spontaneously develop atherosclerosis lack DIT [59]. In contrast to the circumferential orientation of SMCs in the tunica media, DIT smooth muscle is elongated longitudinally [58]. Depending on the setting (e.g., normal development or maturation, injury due to balloon, wire or vessel ligation, vascular graft, or atherosclerosis), intimal SMCs are likely to have diverse cellular sources, including circulating bone marrow-derived stem cells, macrophages, or pre-existing SMCs, ECs, or adventitial cells of the vascular wall [12, 60–64].

The vascular wall: tunica media and tunica adventitia

Media: mural cells and extracellular matrix

Cellular and extracellular matrix components

The tunica media is the middle layer of arteries, stabilizing the intimal endothelium and providing the vessel with structural integrity and dynamic tone. In medium- and large-sized arteries, the media consists of mural cells or SMCs and ECM components including elastin and collagen. In order to maintain blood pressure and perfusion, SMCs are able to dynamically contract or relax, adjusting the caliber of blood vessels. The strength of the vessel wall is largely determined by collagen, whereas elastin allows the vessel to stretch when receiving cardiac output during systole and recoil during diastole to maintain blood pressure. Facilitating the transfer of substances to and from the vascular compartment, the capillary wall is much thinner than that of large vessels, and the mural cells of capillaries are pericytes. While vascular smooth muscle cells (VSMCs), pericytes, and the ECM play pivotal roles in various vascular diseases, there are relatively few studies delineating the development of these components in comparison to the vast literature investigating the morphogenesis of EC networks and tubes.

Differences exist between pericytes and VSMCs, but these cell types are generally considered to exist along a continuum and lack firm distinctions (reviewed in [65]). Pericytes are located in the intima, imbedded in the basement membrane of capillary ECs, while first-layer VSMCs share a basement membrane with ECs but are located in the media, separated from the intima by the internal elastic lamellae. Additionally, pericytes tend to have an irregular orientation, and VSMCs are oriented circumferentially around the EC tube. Pericyte functions include microvessel structural support, phagocytosis and communication between multiple ECs. VSMCs are critical regulators of vascular tone. Many of the molecular markers of these two cell types overlap, but the most frequently utilized markers of pericytes include platelet-derived growth factor receptor beta (PDGFR-β), neuron glial 2, and regulator of G-protein signaling 5. Common markers of SMCs include alpha-smooth muscle actin (SMA), transgelin (also known as SM22-alpha), smooth muscle myosin heavy chain (SMMHC), and smoothelin.

Vascular smooth muscle cell origins

The developmental origins of VSMCs differ depending on the specific blood vessel and are even different within specific regions of individual blood vessels, such as the aorta (reviewed in [66]). The borders between VSMCs from different lineages are sharply demarcated with rare mixing among cells of separate origins. Interestingly, aortic dissection occurs preferentially at boundaries where SMCs of different embryological origin intersect, possibly due to regional variance in mechanical properties [67]. Aortic root SMCs derive from the LPM, and neural crest cells of the ectoderm give rise to SMCs of the aorticopulmonary septum, aortic arch, and cranial vessels [66]. Within the adult descending aorta, SMCs and ECs derive from the PSM and LPM, respectively [18]. Wasteson and colleagues used HoxB6-Cre to mark LPM-derived cells, and identified that these cells are the source of descending aortic ECs and also temporarily of cells on the ventral aspect of the descending aorta wall at ~E9.5–10.51 [8]. Meox1-Cre marks cells derived from the PSM, and, subsequently, cells labeled by this Cre replace the LPM-derived aortic SMCs [18]. A clonal analysis conducted by Buckingham and colleagues provides further evidence for the PSM origin of descending aortic SMCs [19]. They used a mouse in which the nlaacZ gene, consisting of a duplication of the lacZ coding sequence that encodes a truncated and inactive β-galactosidase (β-gal), is knocked-into the α-cardiac actin locus [19]. This nlaacZ requires a rare intragenic recombination event to yield a functional nlacZ sequence which generates active β–gal. There was a low frequency of embryos with labeled cells in either the dorsal aorta or the PSM-derived myotome at ~E10; however, a strikingly high percentage of embryos with β-gal⁺ dorsal aortic cells also had labeled myotome cells [19], suggesting a common origin. Topouzis and Majesky indicated that the origin of VSMCs has significant functional consequences because the effect of TGFβ stimulation on DNA synthesis differs between ectodermally-derived aortic arch SMCs and mesodermally-derived abdominal aortic SMCs in the chick embryo [68].

In a recent study, Cheung et al. [67] generated subtypes of aortic SMC-like cells from human embryonic stem cells (hESCs) to explore the embryological origin of these SMCs in relation to their disease susceptibility. These hESCs were initially induced to the LPM, PSM, or neuroectoderm lineage, and, subsequently, differentiated into SMCs. Interestingly, in response to interleukin (IL)-1β, lineage-specific SMCs differentially activated matrix metalloproteases (MMPs) and tissue inhibitors of metalloproteases (TIMPs), and the authors hypothesized that varying proteolytic activation may result in regional differences in vessel wall mechanical properties, and thus aortic dissection at the boundaries between SMCs of different origins (Fig. 2) [67].

Fig. 2.

The different embryological origins of aortic SMCs may contribute to the site of aortic dissection. Aortic SMCs originate from three distinct developmental lineages. The aortic root is derived from LPM (blue solid arrow), while the ascending aorta and arch are neural crest derived (red solid arrow). The descending aortic SMCs originate from PSM (green solid arrow). In vitro hESC-derived SMC subtypes predicted the differential MMP and TIMP activation in aortic SMCs of corresponding origins (dotted arrows) in response to IL-1β. The authors propose that the origin-specific SMCs display differential proteolytic ability in disease settings, which may result in differential loss of the structural integrity in different regions along the aortic wall. This difference in mechanical properties may predispose aortic dissection to occur preferentially at the boundaries between different SMC lineages (indicated by black jagged bolts). Figure from reference [67]

Due to their prominent role in atherosclerotic heart disease, a number of studies have examined the origin of coronary artery SMCs from the proepicardium (reviewed in [69]). The proepicardium is a transitory structure that forms as an outgrowth of coelomic mesothelium at the caudal base of the murine developing heart at E9.5 and is the source of epicardial cells, which migrate as a mesothelial sheet over the myocardium and become the pericardium. Cardiomyocyte-derived signals provoke an EMT by which some epicardial cells lose their cell–cell adhesion and invade the underlying myocardium. In addition, studies have shown that the proepicardium and epicardium contribute to the coronary artery SMCs using lineage analysis with dyes and viral vectors as well as genetic approaches with the Wilms tumor1 (Wt1)-CreER [69, 70]. Extending this paradigm beyond the coronary arteries, additional investigations suggest that the outer layer of a variety of organs may be considered an important source of VSMCs. For example, in the developing gut, expression of the Wt1 protein is limited to the serosal mesothelium, and a Wt1-Cre yeast artificial chromosome (YAC) transgene marks a lineage of cells that includes the SMCs of the gut vasculature [71].

Although one study similarly identified the lung mesothelium as the source of ~1/3 of all pulmonary vascular SMA⁺ cells through the use of the Wt1-Cre YAC transgene and a panel of Cre reporters [72], the origin of pulmonary artery (PA) SMCs has recently become controversial. Morimoto and colleagues [73] reported that embryos with the Wt1-Cre YAC transgene and a ROSA26R-YFP Cre reporter rarely have YFP⁺ lung VSMCs. These authors further suggest that most SMCs of the proximal PAs arise from ECs, based on experiments using the Tie1-Cre [73]. Several previous studies have also been interpreted as suggesting that transdifferentiation of ECs into VSMCs is important in both disease and developmental contexts [7–12]. However, our recent results with the vascular endothelial cadherin-Cre [74] and mTomato/mGFP Cre reporter [75] indicate that ECs are not a significant source of the SMCs of the large PAs at E18.5 [76]. Our additional experiments indicate that, instead, these cells largely derive from local mesenchyme [76].

Smooth muscle cell differentiation

A critical early step in characterizing the morphogenesis of any tissue is to define the morphology and molecular markers that represent the differentiated phenotype of specific cell types of the tissue of interest. Prominent endoplasmic reticulum and Golgi, a euchromatic nucleus, and lack of a distinctly filamentous cytoplasm characterize early, undifferentiated cells that are destined to the VSMC fate [77]. On the other hand, a heterochromatic nucleus, myofilaments, and decreased synthetic organelles characterize mature VMSCs [77]. Additionally, expression of a number of contractile and cytoskeletal proteins occurs during SMC differentiation. The most abundant smooth muscle protein is SMA encompassing 40 % of the total protein in differentiated SMCs [78]. SMA is an early but nonspecific SMC marker, as it is expressed in skeletal muscle and a variety of other cell types, including embryonic cardiac muscle [78, 79]. SM22α binds actin and tropomyosin and is an early SMC marker and a more specific marker of adult SMCs; however, it is also expressed in other muscle types during development [79]. During embryogenesis, SMMHC is expressed slightly later than SMA and SM22α, and, in contrast to these markers, SMMHC expression is specific to the SMC lineage [80]. Similarly, smoothelin is a SMC-specific cytoskeletal protein, but its expression is first noted late in differentiation when SMCs are part of a contractile tissue [81].

Depending on their environment, VSMCs can assume a variety of phenotypes, which complicates the study of their development [78]. During initial blood vessel development, VSMCs are often proliferative and migratory and synthesize large amounts of ECM components, whereas mature VSMCs are predominantly sedentary and quiescent and express contractile proteins but not ECM. Yet, these differing characteristics between synthetic and contractile states are certainly not absolute [82]. Adult VSMCs may appear phenotypically stable, but they are not terminally differentiated, and, thus, in vascular disease, extracellular cues can stimulate VSMCs to dedifferentiate via a process referred to as phenotypic modulation [78].

These VSMC phenotypes result from gene expression profiles that switch between a differentiated contractile set of genes and a distinct undifferentiated synthetic set of genes [83]. Serum response factor (SRF) is a ubiquitous transcription factor that modulates the expression of almost all SMC contractile and cytoskeletal genes. SRF binds as a homodimer to a 10-base-pair DNA consensus sequence CC(A/T)₆GG known as the CArG box (i.e., C, AT rich, G box), and at least one CArG box is found in the regulatory region of almost all SMC genes. Interestingly, early growth response genes also contain the CArG box sequence within the serum response enhancer element [83], and, thus, a higher order of control is required to regulate the expression of distinct growth or differentiation gene sets in each cell at specific time points. Indeed, the regulation of expression of smooth muscle contractile and growth genes is coordinated via competition for docking sites on SRF between a smooth muscle- and cardiac muscle-specific transcriptional co-activator of SRF known as myocardin and ternary complex factors, such as Elk-1 [84]. Myocardin is considered a master regulator of SMC differentiation as ectopic expression of myocardin in nonmuscle cells is sufficient to activate expression of SMC contractile genes [85], and murine embryos lacking myocardin die at E10.5 due to a lack of VSMCs [86].

More recently, the role of chromatin in the dynamic control of SMC-specific gene transcription has been appreciated (reviewed in [87, 88]). The fundamental unit of chromatin is the nucleosome, consisting of 146 base pairs of genomic DNA wrapped around histone proteins, and individual nucleosomes are linked by DNA. In addition, many non-histone proteins complex with nucleosomes and linker DNA. SMC-specific histone post-translational modifications (e.g., acetylation and methylation) within the CArG box chromatin have been shown to influence the binding of SRF and its co-activators to chromatin [87, 89]. Overall, the differentiation of SMCs is intricately controlled by the dynamic relationship between SRF, myocardin, the CArG box, and modified histones.

Patterning of developing vascular smooth muscle cell layer

In comparison to our understanding of the mechanisms underlying SMC gene expression, there is relatively little known with regard to the recruitment of SMCs and/or their precursors to the vascular wall, the distribution of these cells around the nascent EC tube, or the pattern of differentiation of VSMC precursors within or in proximity to the vascular wall (reviewed in [77]). Historically, studies of vascular wall patterning have in general focused on the aorta and found that loose, undifferentiated mesenchymal cells surround the early EC tube. The location of initial mesenchymal cell consolidation and SMA expression within the descending aorta depends on the cranial-caudal position of the tissue: proximally, these processes initiate on the dorsal aspect of the aorta versus distally where they begin on the ventral side [77, 78]. Classical studies in the chick aorta suggest that outer medial layers mature initially with early presumptive SMC condensation and elongation and elastic tissue accumulation [90, 91]. On the other hand, investigations of the rodent or quail aorta indicate that the cell layer immediately adjacent to the ECs is the first to consolidate and express SMC markers, and, thereafter, additional SMC layers form [92–95].

We recently showed that the PA wall in mice is constructed radially, from the inside out, via sequential induction of successive cell layers (Fig. 3) [76]. The process of generating a relatively mature inner smooth muscle layer encompasses a series of morphological and molecular transitions, including alignment of longitudinally elongated PDGFR-β⁺ cells, upregulation of SMC markers, dowregulation of PDGFR-β, longitudinal shortening, and reorientation to a circumferentially elongated morphology [76]. After commencing in the first layer, a similar process is initiated and then completed in the subsequent layer. This sequential developmental program ceases during assembly of the outermost layer, resulting in relatively “undifferentiated” adventitial cells.

Fig. 3.

Radial development of the pulmonary artery (PA) wall. Longitudinal sections through the left PA wall at the indicated ages in the embryonic mouse, stained for SMA (red), PDGFR-β (green), and nuclei (DAPI, blue) as indicated. Arrowheads, nuclei elongated longitudinally. The inset in E shows a transverse section through the left PA wall. Schematics at the bottom summarize SMA (red) and PDGFR-β (green) expression (orange, co-expression of SMA and PDGFR-β) and cell shape and orientation changes in the forming layers of the developing PA wall. E endothelial cell (EC) layer; 1 first (inner) smooth muscle cell (SMC) layer; 2 second SMC layer; A adventitial cell layer; Lu PA lumen. Square cells circumferentially-oriented cells. Scale bars 10 μm. Figure from reference [76]

A number of mechanisms have been implicated as contributing to this inside–outside radial patterning of vessel wall morphogenesis. In the morphogen gradient model [83], depending on discrete concentration thresholds, an EC-derived signal diffuses through the developing media and adventitia, inducing responses in the cells that reside in these compartments (i.e., changes in morphology, gene expression, and/or proliferation) [76, 96]. In contrast, in the relay mechanism [97], a short-range or surface-bound EC signal induces neighboring cells that in turn propagate the signal either by secreting a morphogen or activating adjacent cells, thereby perpetuating the signal (i.e., the “bucket brigade” or lateral induction model [98–100]). We recently reported a third mechanism in which some of the progeny of inner-layer SMCs migrate radially outward to contribute to, and perhaps seed, subsequent layer(s) of SMCs (Fig. 4) [76].

Fig. 4.

Clonal analysis of inner layer cells of the PA wall. a Clonal analysis scheme showing early (E11.5) marking of an inner layer cell and four possible patterns of its proliferation and migration: longitudinal (L), circumferential (C), and radial (R) expansion and longitudinal with mixing with unlabeled cells (L, M). E endothelial cell (EC) layer; 1, 2 first and second SMC layers; A adventitial cell layer. b A GFP-marked left PA clone in a SMMHC-CreER, ROSA26R ^mTmG/+ embryo, induced by a limiting dose of tamoxifen at E11.5 and analyzed at E13.5 after staining for clone marker (GFP, green), SMA (red), and PECAM (white). An individual coronal confocal section of the four-cell clone is shown (left panel) along with a maximal projection (center panel). 1–4 cells of clone; Lu PA lumen. In the clone schematic (right panel), the positions of marked cells are indicated by circles color-coded to highlight the layer in which the cell resides: green (layer 1), red (layer 2), and blue (adventitia). For cells located superficial (white circles) or deep (gray circles) to the lumen, we were unable to determine which layer they reside in (nd not determined). This clone expanded longitudinally (L) and circumferentially (C), with mixing (M). c Sixteen-cell clone, induced and analyzed as in b, except clone marker was multicolor (rainbow, Rb) ROSA26R^Rb Cre reporter and the clone was analyzed at E18.5. Left panel Cerulean channel of section: bright cells are Cerulean+ (numbered). Center panel Cerulean, mOrange and mCherry channels of the same section. All labeled cells in left PA express Cerulean marker, confirming clonality. Clone expanded in all three axes (L, C, and R), with some cells (red in schematic) having invaded layer 2. For clarity, only every other cell in the schematic is numbered. Figure adapted from reference [76]

Several signaling pathways involving an EC-derived signal that is received by mesenchymal cell receptors have been identified to play important roles in the morphogenesis of the vascular wall (Fig. 5; reviewed in [101, 102]). The PDGF pathway, with the PDGF-B ligand secreted by ECs and its receptors expressed in undifferentiated mesenchyme (PDGFR-α and PDGFR-β) and pericytes (PDGFR-β), is perhaps the most well-studied pathway in vascular wall development. Mice with a targeted deletion of PDGF-B or PDGFR-β develop microvascular aneurysms, hemorrhage, and perinatal lethality, which are considered a consequence of reduced smooth muscle coating of medium-sized arteries and a dearth of pericytes [103–106]. On the other hand, gain-of-function strategies with conditional knock-in of activating PDGFR-β mutations at the PDGFR-β locus in mice enhances cell proliferation and reduces differentiation gene expression in aortic SMCs, thus resulting in increased aortic wall thickness [76, 107]; interestingly, we recently demonstrated that this response is vessel-specific as the PA of these mutants is indistinguishable from that of controls [76]. However, PDGF-B-loaded agarose beads implanted in E12 embryonic lung cultures induces ectopic SMCs in the lung mesenchyme [76]. Another secreted ligand, TGFβ, has also been widely implicated in vascular wall differentiation. When co-cultured with ECs, undifferentiated embryonic mesenchymal 10T1/2 cells are induced to express SMC markers and elongate in a TGFβ-dependent manner [108]. Furthermore, direct treatment of the 10T1/2 cells with TGFβ1 results in similar changes [108].

Fig. 5.

Signaling pathways mediating mural cell recruitment, differentiation, and vascular stabilization. Multiple ligand–receptor complexes have been implicated in pericyte recruitment to the endothelium, including PDGF-B/PDGFR-β, stromal-derived factor (SDF)-1α/C-X-C chemokine receptor type 4 (CXCR4), heparin binding epidermal growth factor (HB-EGF)/EGF receptors (ErbBs), Shh/patched (Ptc), and angiopoietin (Ang)1/Tie-2. The cellular response to TGFβ/TGFβ receptor signaling is dependent on receptor composition and relative ligand level. A ligand–receptor pair is indicated by the same color. N-cadherin and Notch-mediated vessel stabilization requires direct contact between a pericyte and an EC. Figure from reference [102]

In addition to signaling pathways with secreted ligands, the Notch pathway, which utilizes plasma membrane bound ligands such as jagged 1 (JAG1), plays pivotal roles in arterial SMC differentiation in vivo (reviewed in [109]). EC-derived JAG1 is necessary for normal aortic and yolk sac vessel SMC differentiation [110], and the regulation of ductus arteriosus closure and ascending aortic wall differentiation was recently attributed to a process of lateral induction mediated by JAG1 on SMCs [99, 100]. The receptor Notch3 is specifically expressed in arterial SMCs of adult humans [111], and, although blood vessels of newborn mice null for Notch3 lack a phenotype at birth [111, 112], Notch3 is required for postnatal maturation of the tunica media of small vessels [96, 97]. Moreover, CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy; OMIM ID 125310) syndrome in humans is caused by NOTCH3 mutations, and is characterized pathologically by degeneration and eventual loss of VSMCs [61, 62]. Interestingly, CADASIL patients display an abnormal vascular accumulation of the extracellular domain of NOTCH3 [113]. On a final note, we point out that other signaling pathways, such as those mediated by angiopoietin-Tie and S1P ligand–receptor pairs, do not involve an EC-derived ligand and/or mesenchymal receptors but play important roles in vascular wall morphogenesis [101].

Extracellular matrix: collagen and elastic fibers

In addition to smooth muscle development, proper maturation of the ECM in the tunica media is also critical for normal vascular function. Important structural matrix proteins include collagen, elastin, and fibrillins, and expression of structural matrix proteins is dynamic during the development of the mouse aorta. Gene expression profiling of the aorta shows that many of these proteins have an initial large increase of expression at E14, sometimes followed by a brief decrease at birth and then a steady rise for approximately 2 weeks, and lastly a decline to low levels at 2–3 months that persists into adulthood [114, 115].

Circumferential collagen fibers have high tensile strength, bearing most of the stress forces in the aorta, whereas elastic fibers have high recoil and low tensile strength and thus distribute stress throughout the vascular wall, including onto collagen fibers [114, 116]. Seventeen types of collagen are expressed in the developing murine aortic wall [114, 116], and elastin comprises up to 50 % of the dry weight of the aorta [116]. Elastin is secreted by VSMCs as the soluble monomeric gene product tropoelastin and undergoes post-translational modifications and cross-linking. Ultimately, elastic fibers are organized into circumferential elastic lamellae in the tunica media that alternate with rings of VSMCs to form lamellar units.

In addition to the elastin core, elastic fibers are also comprised of fibrous structures known as microfibrils. Fibrillin molecules are linked head to tail and are the structural building blocks of microfibrils. The human genome includes three fibrillin genes, and fibrillin1 is a major component of the microfibril. The temporal pattern of fibrillin1 expression during aortic development is similar to that of most structural proteins, such as elastin, except that its peak expression occurs at postnatal day (P) 0 [115]. Fibrillins are implicated in modulating TGFβ bioavailability by directly interacting with latent TGFβ binding proteins (LTBPs; reviewed in [117]). TGFβ is sequestered through binding to LTBP, and, in response to microfibril proteolytic injury, active TGFβ is released from the microfibril and initiates tissue repair by stimulating the secretion of ECM proteins, anti-proteases, and other factors [117, 118].

Adventitia: connective tissue and progenitor cells

Due to a scarcity of scientific investigation and limited specific markers and genetic tools, little is known about the development of the tunica adventitia, the outer layer of blood vessels. The adventitia is a collagen-rich connective tissue containing perivascular nerves and small arteries, known as the vasa vasorum. Until recently, the adventitia was considered relatively static and loosely organized; however, many investigators now view it as a dynamic tissue containing many interacting cell types (e.g., progenitor cells, fibroblasts, leukocytes, ECs, mural cells, and adipocytes) which are instrumental in the growth, inflammation, disease, and repair of the artery [119]. Based on experiments with quail chick transplants, the coronary vascular adventitia is believed to originate from the epicardium [120]. Quail epicardial cells that were grafted into the pericardial space of the E2 chick undergo EMT and give rise to both coronary vascular SMCs (consistent with findings discussed above regarding SMC origins in the tunica media) as well as coronary perivascular fibroblasts [120].

Recent studies suggest that the adventitia provides a niche for stem/progenitor cells. Zengin et al. [121] identified a “vasculogenic zone” of cells expressing the stem cell marker CD34 (but not the EC marker CD31) at the interface between the media and adventitia of human internal thoracic arteries. This vasculogenic zone is able to form vascular structures in human internal thoracic arterial ring explant cultures and promote the development of microvessels in a tumor model in vivo [121]. Another group reported two distinct cell populations in this same anatomical location with putative stem cell capabilities, one expressing CD34 and the other expressing the hematopoietic progenitor c-kit [122]. Upon exposure to VEGF in culture, a subset of these progenitors attains an EC phenotype [122]. Cells that are derived from the adventitia of another vessel, the human PA, and that express markers of mesenchymal stem/progenitor cells but not of endothelial or hematopoietic cells, have been shown to differentiate into osteogenic, adipogenic, and smooth muscle cells in culture [123]. Campagnolo et al. [124] isolated CD34⁺CD31⁻ cells from human saphenous veins that express the stem cell marker Sox2 and pericyte/mesenchymal markers in culture, and exhibit clonogenic and multi-lineage differentiation capacities. When injected into murine ischemic limbs, these cells induce neovascularization [124].

Interestingly, the intensely studied growth factor Shh is expressed in the vasculogenic zone of medium- and large-sized arteries of the perinatal mouse [125]. Hedgehog proteins signal by binding patched-1 (Ptc1) or patched-2 (Ptc2) receptors, and β-gal activity in Shh reporter mice, Ptc1-lacZ or Ptc2-lacZ, indicate that Shh signaling is active in the adventitia during the late embryonic period and early postnatal period [125]. Cells expressing stem cell antigen1 (Sca1) are located in the adventitia of the mouse between the aortic and pulmonary trunks, initially in the late embryonic stages and persisting into adulthood. Shh appears to be critical for this population of cells, as the number of adventitial Sca1⁺ cells is markedly reduced in Shh ^(−/−) mice [125].

In addition to Shh, it is almost certain that other molecular signals as well as ECM components play principal roles in regulating the vasculogenic zone of arteries. In a recent review, Majesky and colleagues [119] propose Wnt, Notch, and bone morphogenetic protein (BMP) as candidate signaling pathways in this capacity, given their role in progenitor/stem cell niches in other tissues. Studies have also suggested that the local ECM milieu provides dynamic structural, adhesive, and biomechanical inputs that influence stem/progenitor cell proliferation, survival, and differentiation [126–129]. Arteries dampen the ventricular pulse pressure gradient by expanding during systole and recoiling during diastole, and the largest changes in diameter occur in the outermost layers of the vessel (i.e., the adventitia) [119, 130]. These biomechanical dynamics are likely to have profound effects on adventitial progenitor cells via matrix–receptor-mediated signaling [119, 129, 131].

Vascular diseases

Diseases associated with elastin mutations

Loss of function mutations in a single allele of the human elastin gene ELN leads to supravalvular aortic stenosis (SVAS; OMIM ID 185500), a congenital narrowing of the large arteries that occurs either sporadically or with autosomal dominant inheritance. In SVAS, there is breakdown of elastic lamellae, a change in ECM composition, and VSMC hypertrophy and hyperplasia with increased medial layers [132]. Similarly, between E18 and P1, Eln ^(+/−) mice develop an increased blood pressure and correspondingly accumulate more lamellar units on the outer aspect of the outside lamellar unit, thereby preserving the tension per lamellar unit [132–134]. The lifespan of Eln ^(+/−) mice is normal [134]; in contrast, Eln null embryos suffer a progressive obstructive arteriopathy, starting at the end of gestation, and die by P4.5 [135]. The cellular origins of excess SMCs in Eln ^(+/−) or Eln ^(−/−) mutants have not been rigorously assessed, but in the Eln null aorta, these cells are likely, to at least partly, derived from pre-existing SMCs, as the late gestation subendothelial aortic cells are hyperproliferative. We recently reported that construction of the multi-layered PA wall involves the induction of mesenchymal cells surrounding the EC tube and the invasion of inner layer SMC marker⁺ cells into outer layers [76]. We suggest that similar developmental programs are likely to be reactivated at the end of gestation in the Eln ^(+/−) aorta, to construct the additional outer smooth muscle layers from differentiation of adventitial cells and outward radial migration of SMC marker⁺ cells. In a recent study, Ge and colleagues [136] developed a human-induced pluripotent stem cell (iPSC) line from a patient with SVAS, and discovered that SVAS iPSC-derived SMCs exhibited fewer organized SMA filament bundles and proliferated and migrated at higher rates than control iPSC-derived SMCs.

Two additional human disorders, Williams’ syndrome (WS; OMIM ID 194050) and autosomal dominant cutis laxa (ADCL; OMIM ID 123700), have also been associated with mutations in ELN. WS is a multisystem disorder caused by heterozygous deletion of approximately 26–28 genes, including ELN, on chromosome 7 (reviewed in [137]). The disorder is characterized by SVAS, facial dysmorphisms, connective tissue abnormalities, a distinctive neurobehavioral phenotype, and endocrine manifestations. ADCL is characterized by loose, inelastic skin and often by internal manifestations such as aneurysms, PA stenosis, and hernias [138]. Additionally, the effects of elastin on vascular homeostasis is important in humans without known ELN mutations, as defective elastic lamellae are associated with atherosclerosis in such patients [139], and it has been proposed that injury-induced focal disruption of elastin in mature arteries results in SMC phenotypic modulation and neointima formation [140].

Aneurysmal disease

Arterial aneurysms, defined as enduring, localized dilations to at least one-and-a-half times the normal diameter, are susceptible to arterial rupture, thrombus, and/or distal embolization, which are often catastrophic events [141]. Although any artery can be affected, most aneurysms occur in the infrarenal abdominal aorta. The pathological processes involved in the development of abdominal aortic aneurysm (AAA) are thought to involve upregulation of proteolytic pathways, apoptosis, oxidative stress, inflammation, SMC apoptosis, and ECM degradation, as well as hemodynamic factors [141]. AAA is associated with atherosclerosis, advanced age, smoking, elevated cholesterol, hypertension, and a family history of AAA [142]. Although family history is a risk factor, no causative single gene mutations have been identified; instead, AAA is a complex trait, influenced by multiple genetic and environmental factors.

Thoracic aortic aneurysms (TAAs) are another relatively common type of aneurysm. Similar to AAAs, descending TAAs are associated with traditional cardiovascular risk factors, whereas ascending TAAs are not. Instead, ascending TAAs are often a result of cystic medial necrosis or due to one of a number of single gene mutations displaying classic Mendelian genetics (reviewed in [143]). Human disorders involving TAA are often the result of mutations in genes that encode structural proteins within the media (e.g., SMA [144], SMMHC [145]), or ECM (e.g., fibrillin114 [6, 147], collagen genes [148–154]) or signaling factors that function within the medial layer (e.g., TGFβ-receptor 1 or 21 [28, 129]). See Table 1 for a list of many gene mutations associated with TAAs in humans [143], and the section below for a discussion of fibrillin1 and TGFβ receptor mutants. Although not discussed in detail here, COL3A1 mutations in humans are well known to cause Ehlers–Danlos syndrome type IV (OMIM ID 130050) with vascular manifestations that include vessel fragility and large vessel aneurysm and rupture [114, 151, 152].

Table 1.

Genes associated with aneurysm conditions in humans

Gene (protein)	Human phenotype	OMIM for gene	References
ACTA2 (alpha smooth muscle actin)	Inherited ascending thoracic aortic aneurysms and dissections; Moyamoya disease 5; multi-systemic smooth muscle dysfunction syndrome	102620	[144, 209, 210]
ACVRL1 (activin receptor-like kinase I)	Hereditary hemorrhagic telangiectasia, type 2; aortic and medium-sized arterial aneurysms	601284	[211]
COL1A1 (collagen α-1(I))	Osteogenesis imperfecta; Ehlers–Danlos syndrome, type 7A; very rare aortic aneurysm; dissection of medium-sized arteries	120150	[148, 149]
COL1A2 (collagen α-2(I))	Osteogenesis imperfecta; Ehlers–Danlos syndrome, cardiac valvular form type 7B; very rare aortic aneurysm	120160	[150]
COL3A1 (collagen α-1(III))	Ehlers–Danlos syndrome, type 4; arterial dissection is common with rare aneurysm	120180	[151, 152]
COL4A1 (collagen α-1(IV))	Hereditary angiopathy, with nephropathy, aneurysms and muscle cramps	120130	[153]
COL4A5 (collagen α-5(IV))	Alport syndrome, X-linked; aortic aneurysm and dissection	303630	[154]
EFEMP2 (fibulin-4)	Cutis laxa type IB, autosomal recessive with aneurysm; ascending aortic aneurysm and tortuosity	604633	[212, 213]
ELN (elastin)	Cutis laxa autosomal dominant with aneurysm; aortic aneurysm and dissection	130160	[138]
ENG (endoglin)	Hereditary hemorrhagic telangiectasia, type I; aortic and arterial aneurysms	131195	[214]
FBN1 (fibrillin-1)	Marfan syndrome; ascending aortic aneurysm and dissection	134797	[117, 146, 147]
JAG1 (jagged 1)	Alagille syndrome, tetralogy of Fallot, aortic aneurysm	601920	[215]
MYH11 (smooth muscle myosin heavy-chain)	Inherited ascending thoracic aortic aneurysms and dissections	160745	[145]
NOTCH1 (NOTCH1)	Bicuspid valve, ascending aortic aneurysm	190198	[216]
PLOD1 (lysyl hydroxylase 1)	Ehlers–Danlos syndrome, type 6; rare aneurysm	153454	[143, 217, 218]
PLOD3 (lysyl hydroxylase 3)	Osteopenia, joint contractures and fractures, arterial rupture, deafness, arterial aneurysms	603066	[219, 220]
SLC2A10 (glucose transporter type 10)	Arterial tortuosity syndrome with tortuosity, elongation, stenosis and aneurysm formation in the major arteries	606145	[221]
SMAD3 (SMA- and MAD-related protein 3)	Loeys–Dietz syndrome type 3; inherited arterial aneurysms and dissections and early-onset osteoarthritis	603109	[222–224]
TGFBR1 (TGF-β receptor type 1)	Loeys–Dietz syndrome type 1A and 2A; aortic root and diffuse large and medium vessel aneurysms	190181	[158, 159]
TGFBR2 (TGF-β receptor type 2)	Loeys–Dietz syndrome type 1B and 2B; aortic root and diffuse large and medium vessel aneurysms	190182	[158, 159]

Fibrillin1- and TGFβ receptor-associated genetic diseases of the aorta

FBN1, the gene encoding fibrillin1 in humans, is comprised of 65 exons, and the resulting large protein product is distributed in the vasculature, skin, lung, kidney, cartilage, tendon, muscle, cornea, and ciliary zonules. Inherited human diseases associated with FBN1 include familial ectopia lentis, familial thoracic aneurysm, and, most commonly, the Marfan syndrome (MFS; OMIM ID 154700) [117]. The MFS has an autosomal dominant inheritance with an estimated frequency of 5/10,000, with 20 % of the affected patients having no affected parents [118]. Mutations in FBN1 are associated with an increased risk of aortic dilation, dissection, and rupture, and, similarly, targeted deletion of Fbn1 in mice is embryonic lethal due to aortic aneurysm and rupture [155]. Historically, the majority of patients with the MFS die from cardiovascular complications; however, over the last 30 years, significant progress in diagnosis and treatment of the disease has resulted in an increase in life expectancy by approximately 30 years [156, 157]. Given the broad tissue expression of fibrillin1, the syndrome displays a great deal of pleiotropy. In addition to the aorta, affected structures also include the mitral valve, myocardium, eyes, skeleton, skin, and lungs. The majority of clinical features of the MFS results from altered tissue resistance and, thus, manifest and progress with the “wear and tear” of advancing age.

Relatively recently, it became apparent that a group of patients, many of whom had been categorized as having MFS, were phenotypically discrete and have mutations in TGFβ receptors (TGFβ-R1 or TGFβ-R2) [128, 129]. Termed the Loeys–Dietz syndrome (LDS; OMIM IDs 609192 and 610168), this disorder is inherited in an autosomal dominant manner and displays significant clinical variability [158]. Unique features that distinguish it from the MFS include an absence of ectopia lentis, a propensity for aneurysms beyond the aortic root, occurrence of arterial tortuosity, a predisposition to dissect at smaller aortic diameter, hypertelorism, bifid uvula or cleft palate, several autoimmune disorders, and club foot [158, 159]. It is generally accepted that patients with the LDS require close monitoring and prophylactic aortic root surgery at smaller aortic root dimensions than in the MFS [157].

Atherosclerosis

Atherosclerosis of the coronary arteries is the underlying cause of myocardial infarction (MI) and the leading cause of mortality in developed nations. The earliest lesions of this disease are subintimal accumulations of cholesterol-engorged macrophages, and these “fatty streaks” are often noted in the first decade of life of persons living in Western cultures [160]. Fatty streaks are the precursors of more advanced lesions that typically have a fibrous cap comprised of SMCs and ECM enclosing a lipid-rich necrotic core. As plaques progress, they often become increasingly complex with calcification, ulceration, and hemorrhage from small vessels that grow into the lesion from the adjacent media. Progressive severe coronary artery atherosclerosis often results in exertional chest pain known as angina, that results from the demand of myocardial tissue for nutrients and oxygen not being met by the supply via the coronary vasculature because of obstruction to blood flow. More strikingly, acute rupture of a plaque in the coronary arteries exposes subendothelial tissue to the lumen resulting in thrombus formation and thus MI.

A number of studies have implicated ECs and SMCs in playing key roles in atherosclerosis pathogenesis. As with inflammatory reactions, EC interactions with leukocytes are critical in the initial stages of leukocyte recruitment to atherosclerotic lesions [161, 162], and involve a sequential multi-step process of leukocyte rolling, arrest, and eventual transmigration through the EC layer [163]. Historically, this inflammatory stimulus has been implicated in recruiting medial SMCs to the subendothelial space and inducing them to proliferate and produce ECM, and proteases, thereby contributing to vascular disease pathogenesis. Based on a cell lineage study, medial SMCs labeled with the SM22α-CreER are apparently a contributing source of cells in atherosclerotic lesions of ApoE ^(−/−) mice [164]; however, it is not clear whether these pre-existing SM22α⁺ cells are a cellular source of a sizable percent of intimal SMCs or if this is a minor contribution. Alternatively, intimal SMCs have been proposed to derive from ECs, the adventitia or circulating precursors such as monocytes and mesenchymal stem cells [12, 60–64]. Using mouse models of mechanical injury, chronic atherosclerosis, and graft vasculopathy, Nagai and colleagues recently found that some bone marrow-derived monocytes contribute to intimal SMA⁺ cells [165]. However, these cells did not express the SMC-differentiated marker SMMHC and instead express markers of inflammatory monocytes [165]. Studies utilizing rigorous lineage analysis and comprehensive SMC marker profiling are needed to firmly establish the origin and nature of intimal SMCs in atherosclerotic lesions.

Recent findings suggest that the adventitia is important in the pathogenesis of atherosclerosis (reviewed in [166–168]). Within the adventitia, the vasa vasorum and resident stem/progenitor cells have been suggested to contribute to atherosclerosis. Increased vasa vasorum capillary density is noted in regions of human coronary arteries with atherosclerosis [169], and is an early finding associated with EC dysfunction and plaque progression in hypercholesterolemic animal models [170, 171]. The vasa vasorum likely extends into the intima, providing nourishment as well as an easy route for leukocyte trafficking to the growing atheroma. Interestingly, inhibiting vasa vasorum-derived microvessel growth may be a viable strategy for attenuating plaque progression. Applying angiogenic stimulators to the arterial adventitia has been shown to exacerbate experimental injury-induced neointimal thickening [172], and, conversely, treatment of LDL-receptor ^(−/−), Apolipoprotein B-48-deficient mice with an inhibitor of angiogenesis decreases vasa vasorum density and aortic atherosclerosis plaque area and cholesterol deposition [173]. Finally, as described above, stem/progenitor cells are enriched at the interface between the adventitia and outer media of normal vessels [119, 125, 174], and this enrichment is also evident in the aortic adventitia of Apolipoprotein E ^(−/−) mice [175]. Recent evidence suggests that these stem/progenitor cells perform a variety of functions in atherosclerosis progression as well as injury repair [174].

Pulmonary hypertension

Pulmonary hypertension (PH) is a devastating disease defined by a mean pulmonary artery pressure (PAP) greater than or equal to 25 mmHg [176]. This increase in right ventricle (RV) afterload eventually culminates, in severe cases, in RV failure and death. The current classification of the World Health Organization classifies PH into five etiologic groups [176, 177]. The first group consists of pulmonary artery hypertension (PAH), which includes heritable and sporadic idiopathic pulmonary artery hypertension (IPAH). The other groups include PH resulting from left heart disease, parenchymal lung disease, hypoxemia, and chronic thromboembolism. The histopathology of PH is characterized by an increased SMC burden in the media of the large PAs, distal extension of SMCs to normally non-muscularized pulmonary arterioles, and small vessel obliterations known as plexiform lesions that are largely composed of hyper-proliferative ECs as well as SMA⁺ cells, inflammatory cells, and ECM. The cellular source(s) of the excessive SMCs is not established; although a number of origins have been proposed including ECs, pre-existing SMCs, adventitial fibroblasts, pericytes, airway epithelium, and circulating stem cells [10, 178–180], no formal lineage tracing experiments have been reported. In addition, hypoxia has been shown to induce adventitial remodeling with increased ECM and activation of fibroblasts [181, 182]. Taken together, these vascular pathologies markedly increase pulmonary vascular resistance and thus RV strain.

The diagnosis and treatment of PH is extensively reviewed elsewhere [176, 183]. In general, treatment is directed at the underlying condition, and there are no cures for PH with the possible exception of high-risk pulmonary endarterectomy in combination with anti-coagulation for select chronic thromboembolic patients [183]. Much attention has been directed at treating PAH, yet the prognosis remains poor with a 3-year mortality rate of ~45 % in patients diagnosed with incident PAH [184]. The most commonly used drugs include calcium channel blockers, prostacyclins, endothelin receptor antagonists, and phosphodiesterase-5 inhibitors, which reduce PAP through vasodilation. In addition, prostacyclin and endothelin receptor antagonists inhibit SMC proliferation [185, 186]; however, all these treatments have limited ameliorative effects on SMC or EC differentiation, migration, or recruitment. This void in therapeutic options is striking, but, perhaps it is not surprising given that the pathways governing these phenomena in vivo are not well understood.

Although the principal mechanisms regulating PH in vivo mostly remain a mystery, human and mouse genetic studies have been informative. Approximately 70 % of patients with heritable PAH and 10–40 % of sporadic IPAH patients are heterozygous for hypomorphic mutations in the gene encoding bone morphogenetic protein receptor 2 (BMPR2), a TGFβ superfamily receptor [187]. Recent studies have noted similarities between PAH and cancer with enhanced and dysregulated cell proliferation and migration as well as BMPR2 mutations [188, 189]. Bmpr2 ^(−/−) embryonic mice fail to gastrulate [190], but Bmpr2 ^(+/−) mice are viable, and studies indicate that these Bmpr2 heterozygotes either do not spontaneously develop PH [191] or have only mild increases in PAP [192]. Interestingly, the penetrance of heterozygous BMPR2 mutations in humans is only ~25 %, suggesting that a “second hit”, such as an additional genetic, epigenetic, or environmental insult, is required for developing manifestations of PAH. Experiments with attenuated Bmpr2-mediated signaling in specific cell types have been utilized to explore the role of Bmpr2 in PH. Deletion of Bmpr2 in lung ECs with an Alk1-Cre predisposes to increased PAP, RV hypertrophy, distal pulmonary arteriole muscularization, and occlusions [193], changes that are characteristic of human PAH. In addition, SMC-specific expression of a dominant-negative Bmpr2 in mice results in increased PAP, distal muscularization, and, in some cases, small vessel occlusions [194, 195].

Similar to Bmpr2-mediated signaling, the Notch pathway is also implicated in dysregulated cell proliferation in cancer [196] and PAH [197, 198]. Consistent with the role of Notch in establishing SMC identity in development, Notch3 and its downstream target Hes5 are upregulated in PA SMCs isolated from IPAH patients, and Notch3 over-expression induces proliferation in cultured VSMCs [197, 199]. Additionally, Notch3 ^(−/−) mice are resistant to developing hypoxia-induced PH [197]. Notch-mediated stimulation of the contractile phenotype in human VSMCs requires the induction of microRNA 143/145 [200], while microRNAS have been implicated in PH [201]. We recently contributed to a study demonstrating that, in PAH, there is disruption of a novel microRNA 424/503-mediated link between signaling pathways regulated by the ligands FGF and apelin [202].

Inflammatory mediators are also integral to the pathogenesis of PAH. In PAH patients, levels of the cytokine IL-6 are increased in the serum and lung [203, 204]. IL-6 signals through the signal transducers and activators of transcription (STATs), and activation of STAT3 promotes the expression of the proto-oncogene Pim1 in PA SMCs from PAH patients [205]. Consistent with this observation, targeted deletion of IL-6 in mice confers a protective effect against the development of hypoxia-induced PH [206]. Conversely, as a gain-of-function strategy, overexpression of IL-6 by airway epithelium results in increased PAPs and extensive pulmonary vascular remodeling, including small vessel occlusions, distal arteriole muscularization, and thickened walls of the proximal vessels [207].

Although many of the key mechanisms underlying PAH remain to be defined, an increasing number of novel treatments are in phase II–IV clinical trials including peroxisome proliferator-activated receptor gamma-agonists, statins, tyrosine-kinase inhibitors, soluble guanylate cyclase activators, Rho-kinase inhibitors, vasoactive intestinal peptides, and serotonin transporter inhibitors [208]. These therapeutics target a wide array of cellular processes, yet they all reduce SMC proliferation and increase SMC apoptosis [208].

Summary

Taken together, this review highlights that both understanding the normal construction of vascular structures in detail and applying developmental biological techniques (e.g., lineage analysis) to disease contexts have the potential to dramatically enhance our understanding of vascular disease pathophysiology. The vascular system is complex, and its formation requires a meticulous coordination of diverse processes involving multiple cell layers including: (1) EC branching to generate the vascular network; (2) SMC coverage of the endothelial network to provide stability and tone; and (3) adventitial connective tissue development to serve as a niche for crucial inflammatory and progenitor cells. These events utilize specific and carefully regulated signals (e.g., Notch, angiopoietin/Tie, TGFβ, and PDGF) to strictly modulate the recruitment of progenitor cells from a number of sources as well as both the differentiation of these cells into vascular cells and their proliferation. Many recent advancements in vascular biology are instrumental in treatments that help manage vascular diseases, such as PAH and atherosclerosis; yet, the mortality rate of these diseases remains unacceptably high. We suggest that a key factor underlying PAH lethality is that current treatments are primarily directed at dilating blood vessels but do not focus on limiting or reversing the accumulation of excessive SMCs. This limitation largely stems from a lack of knowledge regarding the cellular source(s) of the excess smooth muscle. Similarly, the origin(s) of SMCs which contribute to DIT and atherosclerotic plaques are not well defined. Furthermore, increasing evidence implicates the adventitia as a stem cell niche important for these and other vascular diseases, but little is known regarding how to manipulate this niche as our understanding of even normal adventitial development is rudimentary. In sum, elucidating the key cell types in vascular disease, their cellular sources and the signaling pathways underlying their recruitment are instrumental initial steps in devising novel effective strategies for preventing or reversing vascular disease.

Abbreviations

AAA

Abdominal aortic aneurysm

ADCL

Autosomal dominant cutis laxa

ALK1

Activin receptor-like kinase 1

BMP

Bone morphogenetic protein

CADASIL

Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy

DIT

Diffuse intimal thickening

E

Embryonic day

EC

Endothelial cell

ECM

Extracellular matrix

EMT

Epithelial-to-mesenchymal transition

FGF

Fibroblast growth factor

hESC

Human embryonic stem cell

IL

Interleukin

IPAH

Idiopathic pulmonary arterial hypertension

iPSC

Induced pluripotent stem cell

ISV

Intersegmental vessel

JAG1

Jagged 1

LDS

Loeys–Dietz syndrome

LPM

Lateral plate mesoderm

LTBP

Latent TGFβ binding protein

MFS

Marfan syndrome

MI

Myocardial infarction

MMPs

Matrix metalloproteases

P

Postnatal day

PA

Pulmonary artery

PAH

Pulmonary arterial hypertension

PAP

Pulmonary arterial pressure

PDGF

Platelet-derived growth factor

PH

Pulmonary hypertension

PSM

Presomitic mesoderm

RV

Right ventricle

S1PR1

Sphingosine 1 phosphate receptor 1

SRF

Serum response factor

SMA

Alpha smooth muscle actin

SMC

Smooth muscle cell

SMMHC

Smooth muscle myosin heavy chain

STAT

Signal transducer and activator of transcription

SVAS

Supravalvular aortic stenosis

TAA

Thoracic aortic aneurysm

TGF

Transforming growth factor

TIMPs

Tissue inhibitors of metalloproteases

VEGF

Vascular endothelial growth factor

VSMC

Vascular smooth muscle cell

WS

Williams syndrome

Wt1

Wilms tumor 1

YAC

Yeast artificial chromosome

β-gal

β-Galactosidase