The mesoangioblast is a vascular-associated progenitor cell, typically residing in artery wall, that can give rise to vascular cells as well as other mesodermal lineages including skeletal and cardiac muscle. The name “mesoangioblast” distinguishes this cell type from the “hemangioblast,” another progenitor cell that can give rise to vascular as well as hematopoetic lineages. Minasi and colleagues first defined the mesoangioblast by transplanting cells derived from embryonic mouse or quail dorsal aorta into chick embryos (Minasi et al. 2002). Donor cells were found in numerous mesodermal tissues, including bone, smooth, and cardiac muscle. These cells could be cultured and differentiated into osteoblasts, adipocytes, and skeletal myotubes.

Although mesoangioblasts possess a number of cell markers similar to mesenchymal stem cells (MSCs), they cannot be classified as MSCs. Embryonic mesoangioblasts express CD34, while MSCs and pericytes do not (Wang et al. 2012). Mesoangioblasts may also express Flk1, MEF2D, Sca1, and Thy1, with varying reports of cKit (Passman et al. 2008; Campagnolo et al. 2010; Zengin et al. 2006). Transcriptome analysis of mesoangioblasts also determined that they express receptors for FGF, PDGF, TGFβ, and Wnt proteins (Tagliafico et al. 2004).

While mesoangioblasts were first isolated from dorsal aorta, the developmental origin of mesoangioblasts remains unknown. Flk1 lineage tracing determined that Flk1+ positive cells can give rise to skeletal and cardiac muscle, similar to Minasi’s chimeric studies (Motoike et al. 2003). However, Flk1 is not a unique marker for mesoangioblasts. A recent study by Vodyanik and colleagues determined that hESC differentiated to an MSC lineage express genes specific to lateral plate mesoderm during differentiation, indicating that MSC and MSC-like cells have a lateral plate mesodermal origin (Vodyanik et al. 2010). However, the MSC subset Vodyanik studied appeared to possess a more restricted differentiation potential than mesoangioblasts, making it inaccurate to assume that mesoangioblasts arise from lateral plate mesoderm as well.

Mesoangioblasts have been studied as a potential cell therapy source for muscular regeneration, but have not been used to treat pulmonary diseases (Galli et al. 2005; Galvez et al. 2006; Sampaolesi et al. 2003, 2006). Mesoangioblasts were used to treat mice with X-chromosome-linked muscular dystrophy (mdx), alpha-SG-null mice, and golden retrievers with heredity muscular dystrophy. Animals that are administered mesoangioblasts showed some improvement in muscular function. Interestingly, when GFP-labeled mesoangioblasts were injected into the femoral arteries of mice with cardiotoxin-induced muscular injury, 5 % of the injected mesoangioblast population migrated to the lung (Galvez et al. 2006). Whether mesoangioblast treatment could alleviate the pathology of pulmonary injury remains unknown. The role of these cells in remodeling during lung disease has not been evaluated and a precise understanding of their biology is confounded by the lack of a specific marker to define their origin in vivo.

Smooth muscle proliferation and hypertrophy are characteristic of vascular remodeling in the pulmonary artery and arterioles. A population of smooth muscle cell (SMC) progenitors is responsible for SMC expansion. These progenitors are hypothesized to reside in tissue interstitium, arterial adventitia as well as the mesothelium. The characterization of smooth muscle progenitors to date includes their identification by cell marker analysis including, but not limited to, smooth muscle alpha actin (acta2), SM22a (tagln), Tie-1, and PDGFRα/β (Majesky et al. 2011b; Minasi et al. 2002). The first smooth muscle progenitors described were defined by their expression of PDGFR α and β during lung development and terminal differentiation and their recruitment to newly forming endothelial tubes by PDGF ligands (Lindahl et al. 1997; Hellstrom et al. 1999).

The mesothelium is an extra-pulmonary tissue that may provide a source of SMC progenitors in the adult lung during tissue homeostasis as well as disease-associated remodeling. The mesothelium is a component of the pleura that encases the adult lung. Mesothelial expression of the Wilms tumor 1 gene (WT1) was exploited to perform lineage tracing of mesothelium to both mesenchymal and smooth muscle lineages within the lung. However, WT-1 expression is not strictly limited to the mesothelial cells in the adult lung, complicating lineage-tracing studies to determine the role these cells play during adult tissue homeostasis and disease.

SMC progenitors in homeostasis, development, and disease are controlled by a variety of cell signaling pathways, including FGF, Shh, Wnt, PDGF, and Notch. FGF10-FGFR2b signaling during lung development prevents SMC differentiation and limits matrix deposition (De Langhe et al. 2006). Studies by Passman et al. demonstrated that a gradient of Shh in the adventitia regulates the number of Sca1 SMC progenitors lacking differentiated markers, but expressing transcription factors of SMC lineage potential, including serum response factor, Klf-4, Msx1, and fox04 (Passman et al. 2008). Regulation of proliferation by the Wnt and PDGFR signaling pathways in SM22-labeled SMC progenitors was also reported by Cohen et al. (2009). A temporal increase in Notch signaling was required for Ve-cad^neg/CD45^neg/Tie1^pos/CD31^dim terminal differentiation to SMC, but not the maintenance of a stable vessel (Chang et al. 2012). These complex interactions during development involve coordinated signaling between the epithelium, endothelium, and mesenchyme to regulate appropriate tissue patterning and development.

A pericyte is a perivascular-associated progenitor cell found within the basement membrane of the vasculature. Pericytes were first described by Charles-Marie Benjamen Rouget (Armulik et al. 2011) and named “Rouget cells,” until being renamed for their proximity to endothelial cells by Zimmerman in a 1923 publication. While pericytes are thought to derive from progenitors found in the neurocrest in the forebrain or mesenchyme, lung-resident pericytes also derive from the mesothelium (Que et al. 2008; Bergers and Song 2005). Pericytes, like many mesenchymal progenitor cells, exist in a heterogeneous population with no single cell marker. Also, because many other stromal populations exist in the periendothelial space, they are often confused with other cell types, such as MSC, vascular SMC (vSMC), Fb, and macrophages. Thus, the “gold standard” for pericyte identification is a combination of analysis using multiple markers as well as ultrastructural analysis.

Pericytes have been found to express SMA, Desmin, RGS5, PDGFβ, and NG2 among other markers (Armulik et al. 2011; Crisan et al. 2008; Feng et al. 2010; Rock et al. 2011; Bergers and Song 2005). These makers are not specific to pericytes. For example, all of the markers mentioned are also found on vSMC (Armulik et al. 2011). To correctly identify a pericyte, high-resolution histological examination is needed. Pericytes are vascular associated cells with a large nucleus, small cytoplasm, and long processes encircling the capillary wall (Bergers and Song 2005). Pericytes communicate with endothelial cells through paracrine signaling as well as direct contact in the form of gap and adherens junctions as well as “peg-and-socket” contacts in which pericyte processes are inserted into endothelial invaginations (Armulik et al. 2011; Feng et al. 2010).

Pericytes have been isolated from multiple human organs, including skeletal muscle, placenta, skin, pancreas, and bone marrow (Crisan et al. 2008). Crisan isolated pericytes using flow cytometry based on the expression of CD146 and the exclusion of CD34-, CD45-, and CD56-positive cells. These CD146-positive cells could give rise to adipocytes, chondrocytes, and osteoblasts both in vitro as well as express myofibers when administered to cardioxin-treated mouse skeletal muscle and become ectopic bone when implanted into mouse hind limb (Crisan et al. 2008). Based on marker expression and cell lineage potential, pericytes are similar to MSCs and to date have typically been grouped as the same population due to their perivascular localization.

Another potential source of multipotent cells in the lung are the resident MSCs. The first MSCs defined were bone marrow-derived mesenchymal stem cells (BM-MSC) identified by Friedenstein (Friedenstein et al. 1968; Lee et al. 2011). He established several foundational concepts of MSC biology, determining that isolated BM-MSCs grew in tissue culture plates in formations termed colony-forming unit-fibroblast (CFU-F) and that BM-MSCs could differentiate into a host of different mesenchymal lineages. Subsequent work discovered MSC populations outside the BM-MSC, prompting the International Society of Cellular Therapy to define an MSC by three conditions: (1) MSCs be adherent to plastic (2) MSCs must express specific cell markers (CD73, CD90, CD105) but not express other markers (CD45, CD34, CD14, CD11b), and (3) MSCs must differentiate into mesenchymal lineages including osteoblasts, adipocytes, and chondroblasts (Dominici et al. 2006). While it is common to differentiate potential MSC populations to multiple mesenchymal lineages in vitro, such an experiment is difficult to recapitulate in vivo. MSC are distinct from fibroblasts in their ability to form colonies in a CFU-F assay and their previously mentioned multilineage differentiation potential.

MSCs are also prominent in remodeling due to injury. Remodeling in arteries during disease or in response to injury is often coupled with the formation of ectopic tissue structures such as bone, fat, or cartilage as well as microvascularization of the outer most layers (Majesky et al. 2011a). Primitive mesenchymal cells associated with arteries have been termed calcifying vascular cells (CVC) and vascular adventitial fibroblasts (VAFs) and represent an additional group of multipotent mesenchymal stromal cells (MSC) (Tintut et al. 2003; Hoshino et al. 2008). These cells were similar in surface phenotype to bone marrow MSC including the expression of CD44, CD105, and CD29 and lack of all hematopoietic markers as well as CD34, distinguishing these populations from previously described mesoangioblasts (Tintut et al. 2003; Hoshino et al. 2008).

Resident MSC populations identified in the neonate and adult lung described to date differ both in their origin and function (Foronjy and Majka 2012). Lung MSCs have been identified in bronchoalveolar lavage fluid from patient allograft tissue or tracheal aspirates from ventilated neonates, respectively, through their ability to adhere to plastic (Lama et al. 2007; Hennrick et al. 2007). Concurrently, MSC have been identified and isolated by flow cytometry via visualization of a side population phenotype by Hoechst 33342 vital dye staining in combination with the absence of the hematopoietic marker, CD45 (Summer et al. 2007; Martin et al. 2008; Irwin et al. 2007). Lama et al. showed conclusively that MSCs reside in an adult lung in humans by demonstrating donor origin of BAL MSC up to 11 years after transplant (Lama et al. 2007). Jun et al. also showed that MSCs reside in the adult lung by transplanting genetically labeled bone marrow and observing the composition of the MSC following injury in murine models (Jun et al. 2011). Sca1 has also been reported to enrich for a population of mesenchymal/fibroblast progenitor cells resident in lung tissue (McQualter et al. 2009). Multipotent MSC have also been isolated from endarterectomized tissues from patients with chronic thromboembolic pulmonary hypertension (CTEPH) (Firth et al. 2010).

These MSC populations, independent of their origin, demonstrate multilineage mesenchymal differentiation potential to osteocyte, adipocyte, and chondrocyte lineages, and in varying combinations, express the characteristic mesenchymal cell surface determinants ABCG2, CD90, CD105, CD106, CD73, CD44, Stro-1, and Sca1. In addition, they lack the hematopoietic markers c-kit and CD34 (Jun et al. 2011; Martin et al. 2008; Chateauvieux et al. 2007; Summer et al. 2007; McQualter et al. 2009). Both MSC isolated from BAL and via ABCG2 expression are capable of suppressing T-cell proliferation (Jarvinen et al. 2008; Jun et al. 2011; Lee et al. 2009). Analysis of gene expression of lung MSCs compared to BM-MSCs revealed that only lung MSCs express FOXF1, HOXA5, and HOXB5, transcription factors unique to the developing lung (Walker et al. 2011). In addition to delineating BM-MSC from organ-resident MSCs, this finding may indicate that MSCs are inherently organ-specific.

Pericytes and MSC have very similar cell surface characteristics and differentiation potential. Due to these similarities, pericytes have been hypothesized to be MSC in adult tissues. However, the limitations to address this hypothesis included the lack of a specific marker that could be used to distinguish MSC. Pericytes function to stabilize blood vessels, whereas the function of MSC during adult tissue homeostasis, other than as a supporting or stromal cell, is unknown. Recently, studies by Jun and Chow et al. reported the identification of ABCG2 as a marker to label lung MSC, which lack expression of SMA and NG2 protein (Chow et al. 2013; Jun et al. 2011). Further global gene profiling analyses demonstrated that these ABCG2^pos lung MSC were distinct from NG2 pericytes as well as lung fibroblasts. The differences between NG2^pos pericytes and ABCG2^pos MSC likely represent pericyte heterogeneity within the lung. Interestingly, these ABCG2^pos MSC exhibit vascular endothelial, smooth muscle, and myofibroblast differentiation potential, additionally qualifying them as endothelial and smooth muscle progenitors (Chow et al. 2013).

To distinguish differences between generic MSC, fibroblasts, side population MSC, pericytes, and ABCG2^posMSC, we have formulated a comparison of defined “mesenchymal” cell populations present in the distal lung (Fig. 12.2). Here we compare the initial lung MSC described by the efflux of Hoechst 33342 dye and the appearance of the side population phenotype (CD45^neg SP (Summer et al. 2007; Martin et al. 2008; Irwin et al. 2007; Jun et al. 2011)), to lung fibroblasts, pericytes, ABCG2^posMSC. All of the populations share common traditional MSC cell surface markers. However, only fibroblasts lack CFU-F ability. Although there is significant overlap, based on global gene expression analysis, the CD45^neg SP is a bit broader than the ABCG2^posMSC. This is likely due the expression of transporters and additional multidrug-resistant transporters on the cell surfaces of other lung cell populations which can efflux Hoechst dye. This was confirmed by Jun et al. who demonstrated that ABCG2 marked ~80 % of the CD45^neg SP (Jun et al. 2011). These findings illustrate that ABCG2 is a more specific marker than the Hoechst 33342 staining. There is significant similarity between the ABCG2^posMSC and NG2^pos pericytes. ABCG2^posMSC are precursors of NG2 pericytes. The differences in gene expression are likely due to function. However, it is clear that in their naïve states, they represent two distinct subpopulations of cells. All of these cell populations share characteristics of the “classically defined MSC.”

The endothelial progenitor cell (EPC) is a cell type that generates endothelial cells in response to injury. The first indications of the existence of EPCs came from the spontaneous generation of microvasculature on implanted prosthetics in animals (Scott et al. 1994). Asahara and colleagues were the first to isolate and characterize EPCs from human peripheral blood (Asahara et al. 1997). While Asahara’s initial isolation protocol has been adapted by other researchers (reviewed in Hirschi et al. 2008) (Yoder 2012), the most common method of EPC isolation is to plate peripheral blood on fibronectin-coated dishes, replate non-adherent cells onto separate fibronectin plates to deplete other hematopoietic lineages (such as macrophages and monocytes), and culture cells until clusters emerge at 7 days. Tissue-resident EPCs have also been isolated from artery walls in a location termed “the vasculogenic zone” between the smooth muscle and adventitial layers (Zengin et al. 2006).

While there is no individual accepted marker that defines an EPC, the cells that make up these clusters express CD34, CD133, and VEGFR2/KDR (Hirschi et al. 2008; Mao et al. 2013). Several functional assays define EPCs both in vitro and in vivo. In vitro, EPCs must have the capacity to uptake Dil-Ac-LDL, stain with lectin dye, as well as form tubes when plated in Matrigel (Firth and Yuan 2012). In vivo, EPCs must be able to form tubes when injected into animals in matrigel “plugs.” Several investigators have verified that EPCs contribute to neovascularization in vivo (Asahara et al. 1997; Lam et al. 2008, 2011), and therefore likely have the potential to contribute to pathological remodeling during disease. Studies of the origin, localization, and function of EPC during vascular homeostasis and remodeling are complicated by the lack of tissue and cell-specific markers of this cell type.

Endothelial cells (EC) also represent a potential progenitor population in the vasculature. EC can be induced to become migratory mesenchymal cells through the loss of endothelial markers and the progressive gain of mesenchymal characteristics (Kalluri and Weinberg 2009). Endothelial-to-mesenchymal transition (or EndoMT) has mainly been studied in the context of endocardial cushion and heart valve formation, but has become the focus of many investigators in research into sources of myofibroblasts during fibrosis. EndMT can be induced after bleomycin injury to the lungs (Hashimoto et al. 2010). Sixteen percent of lung fibroblasts isolated from transgenic animals expressing beta-galactosidase in TIE2⁺ cells treated with bleomycin were lacZ⁺, indicating these cells were likely derived from a vascular origin, the caveat of these studies being that Tie-2 expression is not restricted to EC.

Endothelial cells may also be induced to exhibit MSC characteristics. Medici and colleagues determined that a mouse model with constitutive activation of Alk2 had endothelially derived chondrocytes and osteoblasts (Medici et al. 2010). Injection of BMP4, an Alk2-activating ligand, could also generate endothelially derived chondrocytes and osteoblasts. Endothelial cell lines transfected with an Alk4-inducing adenovirus also began to show expression of the MSC markers STRO-1, CD10, CD44, CD71, and CD90. In addition, endothelial cells treated with TGFβ2 and BMP4 (both activators of Alk4) and cultured in osteogenic, chondrogenic, and adipogenic medium could differentiate into these three lineages. Endothelial cells treated with TGFβ2 and BMP4, placed in sponges and implanted into nude mice, could also differentiate into bone cartilage or fat. Taken together, these studies demonstrate that endothelial cells can acquire MSC phenotypes both in vivo and in vitro.