The origin and mechanisms of smooth muscle cell development in vertebrates

Smooth muscle cells (SMCs) are specialized cells found mainly in the walls or in proximity of hollow organs, including the circulatory, respiratory and digestive systems (Halayko et al., 1996; McHugh, 1995; Owens et al., 2004; Sparrow and Lamb, 2003). Although not discussed in detail in this Review, SMCs are also present in reproductive system (oviduct and epididymis), urinary tracts, skin and eyes (Jahoda et al., 1991)

Depending on the organ of interest, SMCs can be found as single cells or they can be organized into layer-forming sheet of intertwined and elongated cells (Burnstock, 1970; Gabella, 1981). Unlike skeletal muscle, smooth muscle (SM) tissue does not have sarcomeres or striations, but does have actin and myosin proteins that both form and control their contractile apparatus. The contraction of SM is initiated by calcium ions (Ca²⁺), which are released from sarcoplasmic reticulum and Ca²⁺-calmodulin-myosin systems (Hill-Eubanks et al., 2011). SM tissues are not controlled voluntarily, and the regulation of contraction is dependent on hormones, on parasympathetic nerves through the autonomic nervous system (ANS) and on locally released signals, such as nitric oxide (NO). In addition, the epithelium and endothelium have been linked to the regulation of SM contraction (Box 1). Often, the stretching-relaxation control response (stretching the muscle to induce its contraction) can regulate SM contraction in certain visceral organs (Chevalier, 2018). It is this contraction and relaxation of SMCs that regulates the function of the organs in which they are found (Huycke et al., 2019).

SMCs are identified through the expression of specific markers, such as α smooth muscle actin (αSMA/ACTA2), transgelin (TAGLN/SM22), smooth muscle-myosin heavy chain (SM-MHC/MYH11), caldesmon, serum response factor (SRF), myocardin and other signaling molecules (Table 1). Most of these markers are shared among SMCs; however, in principle, different combinations of them can be used to distinguish the different types of SM tissue present in circulatory (vascular SMCs and pericytes), respiratory (airway SMCs) and digestive (visceral/intestinal SMCs) systems (Table 1).

SMCs play crucial roles during organ development and morphogenesis, and provide different functions in adulthood depending on their location and organ distribution. In order to understand how a complete organ or tissue develops, it is crucial to reveal the mechanisms of SMC specification and differentiation (Badri et al., 2008; Majesky, 2007; McHugh, 1996; Nakano et al., 2011; Wang et al., 2015). Here, we describe the origin of SMC populations and the steps leading to SMC differentiation during the development of the cardiovascular system, airways and intestine. We describe how SMCs shape organogenesis and discuss how heterogeneities in the origins of SMCs could be used as an innovative therapeutic approach to treat SMC-related diseases.

Although SMCs share similar characteristics, in vertebrates SMCs arise from a wide range of embryonic structures that not only vary between different organs, but also within the same tissue. In this section, we describe the organization and function of SMCs in different organs. We review some of the key fate-mapping experiments that have identified the prominent populations that give rise to SMCs in the cardiovascular system, the respiratory system and intestinal tract in zebrafish, chicken and mice (Jiang et al., 2000; Le Lièvre and Le Douarin, 1975; Santoro et al., 2009; Wasteson et al., 2008) (Fig. 1). Some of the specific genetic and molecular tools that have been used to investigate the developmental origin of SMCs are summarized in Table 2.

SMCs that surround the circulatory system are called vascular mural cells and can be divided between vascular smooth muscle cells (vSMCs) and pericytes (Gaengel et al., 2009).

vSMCs are the most well-studied population among SMCs because their function is essential for arterial physiology and pathology (Basatemur et al., 2019). vSMCs are organized into concentric layers within the walls of arteries and veins (Carmeliet, 2000; Herbert and Stainier, 2011), where they play an essential role in establishing and stabilizing blood vessels during embryonic development (Fig. 2). Furthermore, vSMCs act as crucial regulators of vessel wall assembly and vessel maturation, because they are involved in the production and maintenance of the basement membrane of blood vessels, as well as promoting endothelial cell (EC) quiescence and abolishing responsiveness to vascular endothelial growth factor (VEGF) (Korff et al., 2001; Wagenseil and Mecham, 2009). In the adult, they also provide structural and functional support to vessels to maintain correct blood pressure and tissue perfusion.

It is worth mentioning that the embryonic origins of vSMC progenitors are highly heterogeneous compared with other types of SMC. As we discuss below, vSMC progenitors derive from several lineages depending on the vascular region (Fig. 1). The first insights into the embryonic origin of vascular SM have been derived from a lineage map study for neural crest cells (NCCs) using chick-quail chimeras (Le Lièvre and Le Douarin, 1975). Through this lineage-tracing strategy, it has been possible to visualize that cranial and cardiac NCC migrate toward the pharyngeal arch complex and differentiate into vSMCs associated with the branchial arch arteries. These NCCs also provide SMC progenitors at the level of the right and left common carotid arteries, right subclavian artery (but not the left one), the ducts arteriosus, as well as the ascending aorta and aortic arch (Jiang et al., 2000; Kulesa and Fraser, 2000; Lumsden et al., 1991). Interestingly, substituting the cranial neural crest population with cells from trunk neural crest does not produce vSMCs in those anatomical regions, indicating that environmental cues alone are not sufficient to induce SMC differentiation and that intrinsic factors are necessary (Bookman et al., 1987; Kirby, 1989).

Conversely, it has been shown that vSMCs located at the level of the descending aorta (DA) derive from both splanchnic mesoderm and the somitic compartment, called the sclerotome (Jiang et al., 2000; Pouget et al., 2006; Wasteson et al., 2008; Wiegreffe et al., 2007) (Fig. 1). Lineage-tracing experiments have revealed two different populations of vSMCs in the DA during chick and mouse development (Wasteson et al., 2008; Wiegreffe et al., 2007, 2009): during early development, upon aortic fusion, splanchnic mesoderm-derived cells colonize the ventral regions of the DA and start to express transgelin, a marker of SMC differentiation. These cells are referred to as ‘primary SMCs’ (Wiegreffe et al., 2007). Soon after, sclerotome-derived cells invade and colonize the dorsal region of the DA, and begin also to express transgelin. This population, referred to as ‘secondary SMCs’, migrates along the ventral region of the DA, replacing the primary SMCs located in the ventral region of the vessel. The reason for two different populations is unknown; however, it has been suggested that a dual origin for cells that support the aortic wall could be advantageous during animal evolution, because failure of tissue to develop from one origin could be compensated for by the second origin (Wiegreffe et al., 2009).

Although NCCs give rise to vSMC progenitors at the level of the ascending aorta, SMCs in the heart vasculature have different developmental origins. Using quail-to-chick chimeras and human embryos, it has been demonstrated that, after the formation of the septation of the aortic sac, cells derived from the secondary heart field (SHF) surround the base region of the aorta and the pulmonary trunk and begin to express ACTA2 (Waldo et al., 2005; Yang et al., 2013) (Fig. 1). Using ROSA26:LacZ reporter mouse lines crossed with Wnt1-Cre and Mef2c-Cre animals (for cardiac NCC and SHF tracing, respectively), the SHF has been shown to contribute to the outer medial cells of the ascending aorta, while the inner medial cells are derived from cardiac NCCs (Sawada et al., 2017).

A further source of vSMC for the coronary vasculature in mammals is the pro-epicardium, a transient embryonic structure located dorsally in the developing heart tube. After the formation of the epicardial layer in the surface of the developing heart, epicardial cells undergo epithelial-to-mesenchymal transition (EMT), migrate into the heart and provide vSMC progenitors (Gittenberger-de Groot et al., 1998; Mellgren et al., 2008; Mikawa and Gourdie, 1996; Smith et al., 2011; Vrancken Peeters et al., 1999) (Fig. 1).

In addition to chick and mouse models, the zebrafish has become another important system with which to study the origin and function of vascular mural cells during cardiovascular development (Santoro et al., 2009). ACTA2 and transgelin expression can be seen as early as 60 hpf (hours post-fertilization) around the trunk vasculature and transgenic zebrafish reporter lines can mark vSMC throughout adulthood (Santoro et al., 2009; Whitesell et al., 2014). Using mural cell-specific elements [such as the acta2 promoter/enhancer element, sm22α-b (transgelin1) enhancer element, the abcc9 promoter and the pdgfrb promoter], different groups have been able to trace vSMC origin and differentiation in living zebrafish (Ando et al., 2016, 2019; Chen et al., 2017b; Whitesell et al., 2019, 2014). Similar to mammals, zebrafish vSMC derive from the lateral plate mesoderm and, more precisely, from a population of cells adjacent to the sclerotome region (Kelley et al., 2019; Santoro et al., 2009; Stratman et al., 2017). The advantages of the zebrafish system for live imaging have also shown the contribution of hemodynamics and NOTCH signaling for the recruitment of vSMC to the dorsal aorta (Ando et al., 2019; Chen et al., 2017b).

A final source of vSMCs is the mesenchyme, which also contributes to airway and intestinal SM lineages (discussed below). Using Cre reporter mouse lines, it has been shown that WT1⁺ and Tbx4⁺ mesenchymal lineages contribute to distinct subpopulations of bronchial SMCs and vSMCs, which are dependent on active Hh signaling (Dixit et al., 2013; Zhang et al., 2013; Moiseenko et al., 2017). Parallel studies have demonstrated that WT1⁺ serosal mesothelial cells can also give rise to SM of all major blood vessels in the mesenteries and gut (Wilm et al., 2005).

Although pericytes share similarities with vSMCs in terms of gene expression markers, functionally they are very different. The role of pericytes is to control the formation and permeability of the blood-brain barrier (BBB) (Armulik et al., 2010; Daneman et al., 2010), regulate capillary blood flow (Hall et al., 2014) and, finally, control vessel stabilization (Lindahl et al., 1997a).

While vSMCs wrap around large endothelial tubes, producing basement membranes located between the SMC and EC, pericytes are solitary cells located sporadically along the arterioles, venules and capillaries (Armulik et al., 2011). They do not completely cover the smaller vessels and instead extend primary and secondary cytoplasmic processes along the surface of the endothelium (Hartmann et al., 2015) (Fig. 1). Furthermore, pericytes are embedded within the vascular basement membrane with a peculiar rounded cell body, long dendrite-like process and attached to the longitudinal axis of capillaries (Sims, 1986; Attwell et al., 2016). Pericytes interact with ECs both physically, by means of gap junctions, and through paracrine signaling (Armulik et al., 2005; Gaengel et al., 2009).

Although pericytes and vSMCs share some markers, specific markers can be used to distinguish between them (Table 1). In mammals, these markers include platelet-derived growth factor receptor β (PDGFRβ), which is involved in the proliferation and recruitment of pericytes (Hellstrom et al., 1999; Olson and Soriano, 2011), the nerve-glial antigen 2 (NG2), which is involved in the recruitment of pericytes to the BBB and tumor vasculature (Ozerdem et al., 2001; Stallcup, 2018), the GTPase activating protein regulator of G-protein signaling 5 (RGS5) (Mitchell et al., 2008), and CD146, a transmembrane glycoprotein that functions as an adhesion molecule (Chen et al., 2017a).

Although pericytes play a crucial role in the vascular development, maturation, stabilization and remodeling, their developmental origin is heterogenous. Pericytes in the mouse liver (Asahina et al., 2011), lung (Que et al., 2008) and gut (Wilm et al., 2005) have been traced from the mesothelium, an epithelial monolayer that lines the lung. Pericytes in most other organs originate from the ectoderm. Chick-quail chimera analyses indicate that pericytes in the central nervous system originate from NCCs (Etchevers et al., 2001). Supporting this, genetic lineage-tracing experiments in zebrafish and mice using neural crest-specific Cre recombinase transgenic lines (such as Wnt-1-Cre and Sox10-Cre in combination with loxP-mediated fluorescent reporter line) have demonstrated that brain pericytes are neural crest derived (Ando et al., 2016; Foster et al., 2008). Trans-differentiation from ECs into pericytes has also been suggested to occur in birds (DeRuiter et al., 1997), although this is not a major route of pericyte formation during embryonic development in other vertebrates (Santoro et al., 2009). Recent work in the zebrafish has confirmed that the pericytes that interact with vessels in the brain are derived from NCCs, except for those in the hindbrain, where pericytes are instead derived from the mesoderm (Ando et al., 2016) (Fig. 1). Interestingly, genetic fate-mapping analysis in embryonic mice has revealed that tissue myeloid progenitors give rise to pericytes in the skin, by showing that the number of pericytes is reduced in Pu.1 (Spi1^−/−) homozygous knockout mice, which lack myeloid progenitors (Yamazaki et al., 2017).

Overall, the ontogeny of pericytes and vSMCs is very distinct from that of other specialized SMC types and recent genome-wide quantitative transcriptomes studies have highlighted the complexity of this mural cell population in adult organs (Muhl et al., 2020; Vanlandewijck et al., 2018). Interestingly, single-cell transcriptomic analyses of the adult mouse brain have led to the identification of a population of perivascular fibroblast-like cells that are present on all vessel types, except capillaries (Vanlandewijck et al., 2018).

SMCs that surround the respiratory system are called airway smooth muscle cells (aSMCs). aSMCs encircle the bronchial tree, and play a vital role in airway structure and function, such as in the regulation of bronchomotor tone and in the control of the airway caliber (Stephens, 2001). They form layers that circumferentially surround the bronchial tree, with differences in organization between the upper and lower airways (Amrani and Panettieri, 2003) (Fig. 1). Moreover, aSMCs contribute to the support and homeostasis of the lung function during embryonic development (Jesudason, 2009). aSMC peristalsis (involuntary wave-like contractions) has been observed in a number of species ranging from chicks to humans (McCray, 1993; Parvez et al., 2006), and early impairment in differentiation of this cell lineage can lead to respiratory dysfunction at birth (Lindahl et al., 1997b). aSMC express similar markers to vSMCs, such as ACTA2, SM-MHC, desmin, SRF and myocardin (Table 1).

Several studies have reported that aSMCs are derived from mesenchymal progenitors that relocate around the developing lung epithelium (El Agha et al., 2014; Yang et al., 1999). Early reports have suggested that during mammalian embryogenesis, aSMCs originate from NCCs and mesenchymal cells before vSMCs (Low and White, 1998). Furthermore, studies in mice using a LacZ-Fgf10 transgenic reporter line, which specifically labels the distal lung mesenchyme, have revealed that Fgf10⁺ cells include progenitors of the parabronchial SMCs (Mailleux et al., 2005). Another study has indicated that a subset of aSMCs cells might originate from proximal lung mesenchyme (Shan et al., 2008). By using the Wt1-Cre^ERT2 transgenic line to lineage trace and visualize mesothelial cells during murine development and differentiation, Dixit and colleagues have been able to show that Wt1⁺ mesothelial cells migrate and enter the fetal lung to provide different lung progenitors, such as aSMCs, vSMCs and fibroblasts (Cano et al., 2013; Dixit et al., 2013). More recently, however, an in vivo lineage-tracing study using reporter lines that mark different progenitors has shown that Fgf10⁺ (distal lung mesenchyme) and Wt1⁺ (mesothelium) provide only a minor contribution to the SMC lineage, whereas Axin2⁺ and Gli1⁺ cells, which are enriched in sub-epithelial mesenchyme, produce most of the aSMCs during embryonic development, but with low specificity (Moiseenko et al., 2017).

Organ culture and single-cell studies have described that such mesenchymal progenitor niches are in close proximity to the budding and bifurcating airway branches (Kim et al., 2015; Kumar et al., 2014). Furthermore, recent studies coupling genetic lineage tracing, single-cell RNA sequencing and organoid culture have indicated that, in the lung, a subset of mesenchymal cells express LGR5 and LGR6, known markers of epithelial stem cells. A specific subset of these mesenchymal cells (LGR6⁺) are progenitors for aSMCs and promote airway differentiation of epithelial progenitors via WNT-FGF10 cross talk (Lee et al., 2017). During development, the lung mesoderm also generates vSMCs. Recent findings have revealed that EZH2 is a crucial molecular determinant required to restrict SM differentiation in the developing lung mesothelium (Snitow et al., 2016).

Overall, compared with vSMCs, the embryonic origins of aSMCs have been less intensively studied. Considering the different experimental approaches and transgenic lines that have been used, the extent to which mesenchymal and mesothelial progenitors contribute to the aSMC lineage is far from understood. Further studies are still needed to unveil and elucidate the complete process of airway myogenesis, as well as to decode how the integration of mechanical and biological signaling shape this tissue.

SMCs embedding the walls of visceral organs are commonly called intestinal (or visceral) smooth muscle cells (iSMCs). They regulate peristalsis: the involuntary movement of the longitudinal and circular muscles that occur in progressive wavelike contractions in esophagus, stomach and intestines allowing movement of food through the digestive system. Studies in chicken and zebrafish embryos have described the emergence, propagation and physiological-molecular development of SM-dependent gut peristalsis in the lower digestive tract (Abrams et al., 2012; Chevalier et al., 2020, 2017; Gays et al., 2017). During murine embryonic development, it has been shown that iSMCs confer shape and mobility of the intestine through peristaltic movements (Gabella, 2002), facilitated by the presence of circumferential and longitudinal layers of iSMCs that wrap around the gut tube (Kedinger et al., 1990). The events leading to this final anatomical structure of SM around the gut have been clearly described in mammals: an inner layer of circumferentially smooth muscle is established and, later, an outer longitudinal layer is formed and aligned (McHugh, 1995, 1996).

In vertebrates, endodermal epithelium is surrounded by layers of SMCs derived from splanchnic mesoderm. After the closure and maturation of the epithelium, the splanchnic mesoderm gives rise to mesenchymal cells that surround the primitive gut and form two different layers of SM with the circumferential and longitudinal orientations. In the mouse, circular iSMCs differentiate at E13, while longitudinal layers differentiate at E16 (Walton et al., 2016). In addition, in chicken, circular and longitudinal iSMCs do not differentiate simultaneously, with circular iSMCs developing at E6 and longitudinal iSMCs at E13 (Le Guen et al., 2015). Recent studies have helped to understand how the muscle layers are specified at the correct time and with the correct orientation (Huycke et al., 2019), and the development of iSMCs in the human gastrointestinal tract (Wallace and Burns, 2005).

The development and patterning of iSMCs have also been studied and characterized in the zebrafish model, which has many features that can be considered homologous to amniotes. The zebrafish intestine has a single layer of epithelial cells facing the lumen attached to mesenchyme and surrounded by circular and longitudinal SM. Here, circular layers of iSMCs develop and differentiate at around 72 hpf, while longitudinal layers develop at 120 hpf (Georgijevic et al., 2007; Wallace et al., 2005). Recently, genetic lineage tracing of zebrafish embryos has shown that visceral iSMCs originate from the lateral plate mesoderm (Prummel et al., 2020). Studies in zebrafish embryos have also shed light on early embryonic events leading to the formation of the first circular layer of iSMCs (Horne-Badovinac et al., 2003). It has been shown that the lateral plate mesoderm undergoes a partial EMT; unlike canonical EMT, lateral plate mesoderm migrates as a cohesive layer of mesenchymal cells, retaining features of both epithelial (such as cell-cell adhesions) and mesenchymal (such as the ability to migrate) cells, surrounding first the dorsal and then the ventral region of the primitive gut (Gays et al., 2017). Although the precise signals driving such cellular interactions remain to be determined, a variety of extracellular matrix molecules (e.g. laminin) and growth factors (e.g. transforming growth factor β TGFβ), have been shown to be involved during this process during zebrafish embryonic development (Gays et al., 2017; Yin et al., 2010).

Once SMC progenitors are established, they then differentiate and cover the tissue of interest in a process called myogenesis. Myogenesis not only involves canonical cell signaling, but also mechanical forces that drive SMC differentiation and recruitment (Table 3).

Studies in vertebrates have shown that different signaling pathways, such as PDGF, TGFβ, NOTCH, Sonic hedgehog (Shh), WNT and chemokine/NF-kB signaling, are involved in mural cell (vSMC and pericyte) differentiation, recruitment and stabilization. As most of these pathways have been reviewed in detail previously (Armulik et al., 2011; Badimon and Borrell-Pages, 2017; Döring et al., 2014; Gaengel et al., 2009; Goumans and Ten Dijke, 2018), we focus on the most exciting ones.

Disruption of PDGFRβ signaling strongly affects mural cells recruitment and causes abnormal aortic expansion and elasticity in vertebrates (Hellstrom et al., 1999; Olson and Soriano, 2011; Shimada et al., 1998; Stratman et al., 2017). An interesting signaling crosstalk between PDGF, the CXCR4 chemokine receptor and its associated ligand CXCL12 (also known as SDF1α) has been proposed, because SDF1α expression has been shown to be stimulated by PDGFβ (Song et al., 2009). Knockout of Cxcr4 in mice results in defective arterial patterning with associated defects in mural cell coverage of the vasculature (Li et al., 2013). Additional evidence for an involvement of the SDF1α/CXCR4 axis in vSMC recruitment has recently been provided in zebrafish (Fig. 2). Here, it has been shown that sdf1a/cxcr4 zebrafish mutants have fewer vSMCs associated with the dorsal aorta, suggesting that SDF1α/CXCR4 autocrine activity leads to increased production of PDGF by ECs (Stratman et al., 2020). Overall, chemokine signaling in vSMC recruitment and vascular maturation is evolutionarily conserved (Stratman et al., 2020; Li et al., 2013; Song et al., 2009).

Hemodynamic forces (also called shear- or mechano-stress) are essential for the vascular maturation and the induction of vascular myogenesis in vertebrates (Chen et al., 2017b; Wu et al., 2008). The gene encoding the transcription factor KLF2 is one of the best known flow-response genes (Huddleson et al., 2004). Studies using Klf2-null mice indicate that KLF2 is required for vSMC migration, and have elucidated a novel mechanism involving communication between PDGF and CXCR4 in vascular maturation (Wu et al., 2008). Studies in zebrafish further support the role of shear stress on vascular myogenesis (Chen et al., 2017b). After arterialization, blood flow causes hemodynamic forces that stimulate endothelial primary cilia to activate Notch signaling, as shown by previous studies that have reported the ciliary localization of Notch1 and Notch3 receptors (Ezratty et al., 2011; Leitch et al., 2014). Endothelial Notch activation leads then to Foxc1b expression, which promotes vSMC recruitment and differentiation (Chen et al., 2017b). Our knowledge of shear stress in driving mural cell recruitment is limited and requires further characterization.

Interestingly, expression of two master regulators of mural cell differentiation, SRF and microRNAs (miRNAs or miRs), are also mechanosensitive (Kumar et al., 2014; Turczyńska et al., 2013). For example, ex vivo experiments in mouse portal veins have shown a role for stretching of the vascular wall in driving the expression of vSMC markers, mediated by Rho activity and actin polymerization through a SRF-dependent mechanism (Albinsson et al., 2004) (Fig. 2). Indeed, during vSMC differentiation, many signaling pathways converge on the regulation of SRF and its co-factor myocardin and myocardin-related transcription factors (MRTFs) (Olson and Nordheim, 2010; Owens et al., 2004). SRF and myocardin form a key regulatory complex for vSMC differentiation by binding to CArG box DNA elements, thereby regulating vSMC-specific markers (Majesky, 2007) (Fig. 2). A lack of, or reduction in, SRF has been directly linked to a SMC phenotype characterized by impairment of SM differentiation, proliferation and migration in both the cardiovascular and gastrointestinal systems (Browning et al., 1998). Similarly, myocardin loss of function is lethal in mouse embryos due to the lack of vSMCs in the dorsal aorta, which leads to severe vascular abnormalities (Amrani and Panettieri, 2003). Seminal work has suggested that actin polymerization triggers the nuclear localization of MRTF to stimulate SRF-dependent transcription of the master regulators involved in SMC differentiation (Miralles et al., 2003).

Similarly, numerous studies have identified that miRs are potent regulators of vSMC biology. Among these, miR-143 and miR-145 control vSMC differentiation. Indeed, Boettger and colleagues have reported that the mir-143/145 gene cluster promotes a contractile phenotype of murine vSMCs (Boettger et al., 2009). Furthermore, it has been shown by in vivo and in vitro studies that miR-145 and miR-143 regulate vSMC fate and plasticity by directly targeting a network of transcription factors, including KLF4 (Kruppel-like factor 4), myocardin and ELK1 (member of ETS oncogene family), and dysregulating the TGFβ signaling cascade, including SMAD2, SMAD3 and TGFβ ligands (Cordes et al., 2009; Zhao et al., 2015) (Fig. 2). A role for miR-143 and miR-145 in mediating the communication between vSMCs and ECs to modulate angiogenesis and vessel stabilization has also been proposed (Climent et al., 2015). Moreover, it has been shown that miR-143 and miR-145 are among those mechano-sensitive miRNAs, and the mir-143/145 cluster is regulated through KLF2 (Hergenreider et al., 2012). Here, the authors have proposed a mechanism through which flow promotes the transfer of endothelial-derived miR-143/145 to adjacent mural cells via extracellular vesicles. Another miRNA, miR26a, has been proposed to regulate mural cell differentiation and function, both in zebrafish and in mice (Leeper et al., 2011). Although the mechanism and the cell-autonomous function of this miRNA is still controversial, miR-26a acts at the intersection between PDGF and the BMP/TGFβ family during vSMC maturation (Watterston et al., 2019; Yang et al., 2017).

The molecular mechanisms that regulate epigenetic control of vSMC differentiation are also currently under intense study, but many pathways and factors still need to be identified. Innovative new tools and approaches, such as single-cell RNA sequencing and epigenetic assays, are allowing the identification of mural cell heterogeneity (Vanlandewijck et al., 2018), specific intermediate (progenitor) cell subpopulations of mural cells (Liu et al., 2019), as well as novel molecular determinants of vSMC during development and disease (Muhl et al., 2020; van Kuijk et al., 2019).

For aSMCs, the mesenchymal progenitor niches are located at the branch tip of the developing lung, and different studies have demonstrated a role for epithelial-derived morphogens in their recruitment and differentiation (Kumar et al., 2014; Mailleux et al., 2005). Although the precise mechanisms are not yet completely defined, different studies have demonstrated the involvement of FGF9, FGF10, SHH, BMP and WNT signaling in this process (Fig. 3). Expression of these molecules at the level of the pulmonary mesothelium, airway stalks and mesenchymal progenitors leads to aSMC recruitment, differentiation and envelopment of the epithelium. Blocking FGF9 signaling by knocking out its receptors in the airway mesenchyme results in ectopic SM differentiation and impaired epithelial morphogenesis (Yi et al., 2009). Mutant mice lacking Shh, which is normally expressed in the embryonic lung epithelium at the distal bud tips, show an absence of aSMCs (Miller et al., 2004), while exogenous recombinant Shh induces aSMC differentiation (Weaver et al., 2003). Fgf10 hypomorphic embryos have diminished aSMCs recruitment, although this may be an indirect effect due to impaired Shh epithelial expression (Ramasamy et al., 2007). FGF10 is expressed by early SMC progenitors and is involved in BMP4 expression at the level of the epithelium, which is required for entry into the aSMC lineage (Mailleux et al., 2005). The interplay between FGF10 and SHH signaling during mesenchymal differentiation and lung morphogenesis has been investigated, but remains unclear (Herriges et al., 2015) (Fig. 3).

Although these signaling cues are necessary for SMC differentiation, they are not sufficient. Yang and colleagues have demonstrated that, in absence of mechanical tension, mesenchymal progenitors do not differentiate in the SM lineage, while mechanical stretch (in the form of cell spreading, elongation or stretching) induces the expression of SM markers (Yang et al., 2000); αSMA, SM22α and desmin expression levels increase when uniaxial continuous stretching is induced through the use of a silastic membrane, and in ex vivo experiments, the numbers of αSMA-positive cells surrounding the bronchial tree increases. The authors have identified that SRF and its antagonist SRFΔ5 are key players involved during this tension-induced myogenesis, which has been confirmed by analyzing two different human hypoplastic lung pathologies (Yang et al., 2000). The same group have also identified a set of stretch-responsive factors, designated as tension-induced/inhibited proteins (TIPs), that are able to promote myogenesis possibly through chromatin remodeling and histone acetylation (Jakkaraju et al., 2005).

Numerous signals are implicated in differentiation of SMCs in developing gut (Fig. 4). In particular, endodermal-derived SHH and Indian hedgehog (IHH) signaling act on the surrounding mesenchyme to control its differentiation into iSMCs (Kim and Choi, 2009; Mao et al., 2010). However, opposing results have been obtained from chicken versus mouse. Tissue-grafting experiments and pharmacological modulation of the Hedgehog (Hh) pathway in chick explant culture suggests that this pathway inhibits SM differentiation (Sukegawa et al., 2000). Conversely, studies in mouse reveal that Hh signaling promotes SM formation via the direct activation of myocardin, a master regulator of SM differentiation (Ramalho-Santos et al., 2000; Zacharias et al., 2011). Another secreted factor that is implicated in gut SM formation is BMP signaling (de Santa Barbara et al., 2005). It has been proposed that BMP4 inhibits gut myogenesis: BMP4 overexpression results in reduction of mesenchymal layers and decreased SM in developing chick gut. BMP4 is activated by and dependent on Hh secretion by the adjacent endoderm, suggesting a level of integration between the two pathways in this context (Roberts et al., 1998).

Recently, Huycke and colleagues have demonstrated that the correct orientation of the circumferential and longitudinal layers is driven by unique mechanical forces and Shh signaling at different stages of gut development in chick. The authors have demonstrated that the orientation of the first (circumferential) SM layer is driven by residual strain created by differential growth of the mesenchyme, with cells in the inner mesenchyme having shorter cell cycles than cells of the outer mesenchyme (Huycke et al., 2019). Conversely, the longitudinal orientation is driven by cyclic contraction of the inner layer that induces SMCs of the outer layer to align perpendicular to the axis of strain. During this process, cell signaling involving SHH, BMP and BMP antagonists (such as noggin and gremlin 1) orchestrate the timing and location of these SM layers along the gastrointestinal tract (Huycke et al., 2019) (Fig. 4).

Studies in zebrafish have identified a crucial role for miR-145 in iSMC differentiation (Zeng et al., 2009; Zeng and Childs, 2012). miR-145 governs expression of BMP4 morphogens that drive both epithelial and SMC differentiation in intestine through autocrine and paracrine mechanisms (Zeng and Childs, 2012) (Fig. 4). TGFβ has been identified as a crucial player in the migration of SMC progenitors from the lateral plate mesoderm in the direction of developing tube (Gays et al., 2017). TGFβ acts through two distinct genes, Zeb1 and Foxo1, that encode a zinc-finger and homeodomain transcription factor, and a Forkhead family transcription factor, respectively. During iSMC development and differentiation, ZEB1 (which is known to function in EMT) controls lateral plate mesoderm migration, while FOXO1A is involved in maintaining mesodermal progenitor stemness. TGFβ-induced miR-145 negatively regulates zeb1 and foxo1a expression and, therefore, blocks EMT and mesodermal stemness; these biological processes need to be terminated in order to allow LPM differentiate in mature iSMCs (Gays et al., 2017) (Fig. 4).

The cardiovascular, respiratory and gastrointestinal systems are constantly subjected to mechanical stimuli; blood, gas and food create two types of mechanical forces on the internal wall of these organs: shear stress and pressure. Both forces affect the endothelium and epithelium, but also SMCs. A process called mechano-transcription is currently the focus of intense study. For example, it has been shown that vSMCs possess a mechanosensitive cell cycle regulation through the transcriptional regulation of p27Kip1 (Sedding et al., 2003), emerin and lamin A/C, two important components of nuclear envelope proteins localized beneath the inner nuclear membrane (Qi et al., 2016). It has been shown that mechanical forces are crucial also for iSMC differentiation by regulating transcription directly in these cells (Shi, 2017).

Some studies have shown a role for the mechano-sensors and mechano-transducers YAP and TAZ in growth and differentiation of gut mesenchyme (Cotton et al., 2017). The Hippo and YAP and TAZ pathways (YAP/TAZ) are regulated by mechanical cues and are involved in the regulating of the organ size and tumor growth (Dupont et al., 2011; Moya and Halder, 2019). Genetic analyses using mice genetic models have highlighted the functional interplay between Hippo and SHH pathways in gastrointestinal development (Cotton et al., 2017). Specifically, it has been proposed that myocardin expression is driven by SHH signaling, but it is inhibited by YAP/TAZ, which spatially and temporally regulate iSMC differentiation (without affecting SHH transduction). When YAP/TAZ are downregulated in specific restricted zone, myocardin expression occurs, enabling iSMC lineage specification (Cotton et al., 2017) (Fig. 4).

Mechanical forces produced by the cellular environment have important roles in the differentiation and recruitment of SMCs. On the other hand, several studies have demonstrated that mechanical forces produced by SMCs instruct the morphogenesis of different organs during their development.

Although aSMCs are mainly known to play a crucial function postnatally, their role in lung morphogenesis has been studied in mammals (Kim et al., 2015). Indeed, localized SM differentiation contributes to shaping lung branches: the position and morphology of domain branches are highly stereotyped, as is the pattern of SM that differentiates around the base of each branch (Goodwin et al., 2019).

Shyer and colleagues have shown that differentiating SM layers play crucial roles in the formation of villi in the human and chick gut. The circular inner SM layer prevents expansion of the gut tube and causes compressive forces in the epithelium, which induce longitudinal ridges and intestinal folding. Meanwhile, the longitudinal outer SM layers fold the ridges into parallel zigzags, leading to villi formation (Shyer et al., 2013). These studies conclude that it is possible that the formation of human and chick villi occurs through similar steps. In addition to promoting villification, circular SM tone and contractions are also necessary for the anisotropic growth of the gut, which in turn is essential for high-aspect ratio of the organ, increased epithelial surface and compartmentalization of digestion (Khalipina et al., 2019).

Another important aspect to keep in mind is that the SMCs interact with their environment. Different in vitro studies have demonstrated that the co-culture of ECs and SMCs inhibits EC growth and movement by inducing EC quiescence. These responses are mediated by TGFβ action and abrogated VEGF responsiveness (Antonelli-Orlidge et al., 1989; Korff et al., 2001; Sato and Rifkin, 1989). Furthermore, Fortuna and colleagues have shown that, in zebrafish, vSMC recruitment to the dorsal aorta is required for the acquisition of noradrenergic specification by sympathetic precursors (Fortuna et al., 2015). Overall, SMCs support and regulate organ development and differentiation throughout the vertebrate body.

In this Review, we have discussed how the SM of different organs share similarities, such as the expression of common markers and performing their function by constricting the hollow organs they ensheathe (e.g. circulatory, respiratory and digestive systems). However, specific differences do exist in the organization, origin, signaling between vSMC, aSMC and iSMC that reflect the specialization and functions of the organs and tissue with which they interact. The SMCs of different organs might be organized very differently and perform specialized functions, as well as expressing both specific and unique marker genes. Furthermore, the population of SMCs present within circulatory, respiratory and digestive systems are not homogenous; they have different origins and rely on different signals for specification. Overall, should we think of smooth muscle cells as a single cell type or do we need to consider them as heterogeneous populations? So far, we probably do not know enough about SMCs to fully address this question; we need to better understand their heterogeneity. It is these types of questions, however, that make SM a puzzling and exciting tissue to study.

The recent progress in understanding the origin, function and signaling of SMC development in a variety of organs can provide the basis for a better understanding, modeling and treatment of pathologies based on SMC abnormalities. For regenerative vascular medicine, there is a pressing need for new model systems to investigate vSMC development and heterogeneity (Sinha and Santoro, 2018). Induced-pluripotent stem cells (iPSCs) are an exciting, clinically relevant candidate for cell-based applied therapy. In particular, human IPSC-derived vSMCs are being used to regenerate or tissue-engineer both large vessels and the microvasculature (Cheung et al., 2012). Single-cell transcriptomics is emerging as an essential technique, enabling researchers to investigate different cell populations in more depth than ever before (Jakab and Augustin, 2020). As we have discussed, recent work has demonstrated that this new technology can be used to identify disease-relevant transcriptional signatures in vSMC-lineage cells in healthy blood vessels, with implications for disease susceptibility, diagnosis and prevention (Dobnikar et al., 2018).

In addition, aSMCs play an integral part in the pathogenesis of chronic airway diseases, such as asthma, where they contribute to airway remodeling and inflammation. Therefore, targeting the signaling pathways that regulate airway SM responses might provide new therapeutic approaches and treatments for chronic pulmonary disease.

The cellular plasticity of iSMCs is not only important for the normal differentiation and maturation of gastrointestinal SM, but also appears to play a significant role in a variety of intestinal diseases. Obstruction of the small intestine is a frequently encountered complication of many congenital and acquired gastrointestinal disorders (Gabella, 1990; Lin et al., 2012; MacDonald, 2008). Hypertrophy of the circular iSMCs in the constricted part of the intestine is the most common effect of the disease. Chronic partial obstruction of the small intestine can dramatically alter peristaltic contractile properties. We know relatively little about the molecular mechanisms that contribute to the phenotypic remodeling of intestinal SM in this obstructive disease (Chen et al., 2008; Shi, 2017). Mechanical stretch has been shown to alter gene transcription in iSMCs and this stretch-altered gene expression (e.g. mechano-transcription) could play a crucial role in pathogenesis of motility dysfunction and abdominal pain obstruction. It can be expected that future studies may be applicable to our understanding of the molecular events associated with motility dysfunction in chronic-intestinal pseudo-obstructions.

The anatomical organization and function of SMCs clearly demonstrates the complexity involved in understanding and decoding normal SM development in different organs. In recent years, important progress has been made in understanding the biology of SMCs, but many questions still remain unresolved: are there organ-specific and segment-specific patterns during smooth muscle development? How is the SMC molecular differentiation program integrated with the stemness program? Are there SMC stem/progenitor cells that reside in different organs, and can they be reactivated and eventually move through organs in embryos and adult? Does the SM of different organs develop in a manner similar to other SM tissues, i.e. vascular, respiratory and urogenital? How do the various SM layers of the mature vascular and gastrointestinal tract develop? Does organ metabolism influence their recruitment and differentiation? Such challenges need to be faced in the near future to support studies on SMC, as they represent a mysterious, but exciting, tissue in vertebrate bodies with great therapeutic potential.

We sincerely apologize to all our colleagues whose contributions to SMC biology we were unable to cite due to the complexity and extensiveness of the covered topics as well as space limitation. We thank the members of the Santoro lab for critical comments on the manuscript, as well as Ellen Jane Corcoran for editorial and language assistance.