RESEARCH ARTICLE

Ezh2 restricts the smooth muscle lineage during mouse lungmesothelial developmentMelinda Snitow1,2,3,4, MinMin Lu1,4, Lan Cheng1,4, Su Zhou1,4 and Edward E. Morrisey1,2,3,4,5,*

ABSTRACTDuring development, the lung mesoderm generates a variety of celllineages, including airway and vascular smooth muscle. Epigeneticchanges in adult lung mesodermal lineages are thought to contributetowards diseases such as idiopathic pulmonary fibrosis and chronicobstructive pulmonary disease, although the factors that regulateearly lung mesoderm development are unknown. We show in mousethat the PRC2 component Ezh2 is required to restrict smooth muscledifferentiation in the developing lung mesothelium. Mesodermal lossof Ezh2 leads to the formation of ectopic smooth muscle in thesubmesothelial region of the developing lung mesoderm. Loss ofEzh2 specifically in the developing mesothelium reveals amesothelial cell-autonomous role for Ezh2 in repression of thesmooth muscle differentiation program. Loss of Ezh2 derepressesexpression of myocardin and Tbx18, which are important regulatorsof smooth muscle differentiation from the mesothelium and relatedcell lineages. Together, these findings uncover an Ezh2-dependentmechanism to restrict the smoothmuscle gene expression program inthe developing mesothelium and allow appropriate cell fate decisionsto occur in this multipotent mesoderm lineage.

KEY WORDS: Ezh2, Polycomb repressive complex 2, Lungdevelopment, Smooth muscle, Mesoderm, Mesothelium

INTRODUCTIONLung mesoderm-derived lineages provide a physical scaffold andinductive signaling for the conducting airway and gas-exchangingalveolar epithelium. Lung mesoderm develops into multiple tissuetypes, such as vascular smooth muscle, airway smooth muscle, andendothelium, in addition to poorly characterized parenchymal andinterstitial mesenchymal cells. Lung mesodermal lineages arehighly plastic in development and during injury repair, anddestabilization of cell identity and quiescence are associated withdiseases such as idiopathic pulmonary fibrosis (IPF) and chronicobstructive pulmonary disease (COPD) (Morrisey and Hogan,2010; Herriges and Morrisey, 2014; Hogan et al., 2014).The lung mesothelium is a multipotent mesoderm-derived cell

lineage that forms amonolayer surrounding the lung and contributesto lung development through both direct contribution ofmesodermally derived lineages as well as paracrine signaling. The

lung mesothelium is contiguous with other pleural cell types,including the epicardium of the heart. Mesothelial cells migrate intothe lung and contribute to the mesenchyme and generate asubstantial proportion of vascular smooth muscle (Que et al.,2008; Dixit et al., 2013). Mesothelial cells may contribute to IPF, adisease characterized by inappropriate fibroblast differentiation andexcessive proliferation (Mutsaers et al., 2015). Mesothelial cellsdifferentiate into fibroblasts upon exposure to Tgfβ, a pro-fibroticsignaling molecule, which is upregulated in IPF lungs (Batra andAntony, 2015). Additionally, a mouse model of fibrosis byintratracheal Tgfβ1 instillation induces expression of themesothelial-specific transcription factor Wilms tumor 1 (WT1) inthe lung parenchyma (Karki et al., 2014). These data suggest that themesothelium and mesothelial-derived cells might reactivate theirdevelopmental plasticity during lung injury and repair processes,and contribute to inappropriate fibrosis.

Ezh2 is the histone methyltransferase component of Polycombrepressive complex 2 (PRC2), which trimethylates lysine 27 ofhistone H3 (H3K27me3) and epigenetically represses transcriptionof developmentally regulated genes (Su et al., 2003; Boyer et al.,2006; Bracken et al., 2006; Lee et al., 2006). Ezh2 is highlyexpressed in the developing lung mesoderm, and its expressiondecreases as development proceeds (Galvis et al., 2015; Snitowet al., 2015). In addition to suppressing ectopic lineagedifferentiation during lung endoderm development (Galvis et al.,2015; Snitow et al., 2015), Ezh2 is required for normal developmentof many mesodermal tissues, such as the heart where it is essentialfor suppressing skeletal muscle genes in the myocardium (Delgado-Olguín et al., 2012). The crucial roles of Ezh2 in multiplemesodermal lineages led us to investigate its requirement in lungmesoderm development and lineage specification.

Here, we show that Ezh2 is required in the developing lungmesothelium to suppress smooth muscle differentiation. Ezh2directly suppresses expression of the smooth muscle mastertranscription factor myocardin (Myocd) as well as thetranscription factor Tbx18, which is known to promote smoothmuscle development from the epicardium (Wu et al., 2013).Together, these data indicate that Ezh2/PRC2 is required to suppressectopic smooth muscle development from the multipotent lungmesothelium by repressing the key transcriptional regulators Myocdand Tbx18.

RESULTSLoss of Ezh2 in mesoderm inhibits lung growth andrespiratory functionEzh2 was previously shown to be highly expressed in thedeveloping lung mesoderm (Galvis et al., 2015; Snitow et al.,2015). To assess the role of Ezh2 in lung mesoderm development,we generated a mesoderm-specific loss-of-function mutant bycrossing Ezh2flox/flox mice with the pan-mesodermal Cre-expressingline Dermo1cre that has robust activity in the lung mesoderm byReceived 7 January 2016; Accepted 17 August 2016

1Department of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA.2Department of Cell and Developmental Biology, University of Pennsylvania,Philadelphia, PA 19104, USA. 3Penn Center for Pulmonary Biology, University ofPennsylvania, Philadelphia, PA 19104, USA. 4Penn Cardiovascular Institute,University of Pennsylvania, Philadelphia, PA 19104, USA. 5Penn Institute forRegenerative Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA.

*Author for correspondence (emorrise@mail.med.upenn.edu)

E.E.M., 0000-0001-5785-1939

3733

© 2016. Published by The Company of Biologists Ltd | Development (2016) 143, 3733-3741 doi:10.1242/dev.134932

DEVELO

PM

ENT

E10.5 (Su et al., 2003; Yu et al., 2003; Yin et al., 2008). We alsocrossed the R26RmTmG reporter line into these mutants to track Crerecombination activity (Muzumdar et al., 2007). Dermo1cre:R26RmTmG:Ezh2flox/flox mutant mice (hereafter referred to asEzh2mesoderm-KO) die at P0 due to respiratory distress. They arecyanotic shortly after birth and attempt to breath, but are unableto inflate their lungs with air as assessed by buoyancy in PBS(Fig. 1A,B). Ezh2mesoderm-KO lungs are small (Fig. 1C,D) and havepoorly developed alveoli with reduced mesenchymal development,as noted by reduced vimentin expression (Fig. S1A-C). Ourlaboratory recently showed that Tgfβ is crucial for sacculation andearly alveologenesis in the lung (Wang et al., 2016). However, lossof Ezh2 did not alter the expression of Tgfβ pathway members in thelung mesoderm (Fig. S1D). Moreover, although the epithelium ofEzh2mesoderm-KO lungs is immature and exhibits a slight reductionin expression of Tgfβ pathway components (Fig. S1E-G), theseeffects are indirect and likely to result from reduced mesenchyme inEzh2mesoderm-KO lungs.Ezh2mesoderm-KO mutant lung mesenchyme is less proliferative

than that of sibling controls, with a 50% reduction in BrdUincorporation (Fig. 1E,F). Cell cycle inhibitors such as Cdkn2a (also

known as p16) are common targets of Ezh2-mediated epigeneticrepression (Bracken et al., 2007), and Cdkn2a expression is readilyobserved at E14.5 and E18.5 in the Ezh2mesoderm-KO lungmesenchyme, but not in controls (Fig. 1G,H). To examine theCdkn2a promoter for occupancy by the Ezh2/PRC2 histone markH3K27me3, we performed ChIP-qPCR in isolated lung mesoderm.We found that the Cdkn2a promoter is enriched for the H3K27me3mark, in contrast to a nearby gene desert and the constitutivelyactive gene β-actin (Actb) (Fig. 1I). Thus, Ezh2/PRC2 is responsiblefor repressing the cell cycle inhibitor Cdkn2a in order to allowproliferation and appropriate growth of the developing lungmesenchyme.

Lossof Ezh2 leads toectopic smoothmuscle formation in thelungExamination of Ezh2mesoderm-KO lungs revealed co-expression ofSM22α (also known as Tagln) and smooth muscle actin (SMA)in regions along the periphery of the lung adjacent to themesothelium (Fig. 2A-D). The co-expression of both SM22α andSMA suggests that these ectopic structures were composed ofsmooth muscle cells. 3D imaging by optical projection

Fig. 1. Ezh2 is required in mesoderm for lungdevelopment. (A) Ezh2mesoderm-KO mutant miceare cyanotic shortly after birth. (B) Lungs fromEzh2mesoderm-KOmice at P0 are not buoyant in PBS(n=3), whereas control lungs float (n=4).(C,D) Ventral (C) and dorsal (D) views ofEzh2mesoderm-KO show reduced lung size at E18.5.(E) IHC for BrdU at E18.5 reveals decreased BrdUincorporation in the GFP+ mesoderm of mutants.(F) BrdU incorporation quantified by litter, showingan average 50% reduction in Ezh2mesoderm-KO

mutants relative to their siblings (P=0.0018).(G) IHC for Cdkn2a (green) reveals increasedexpression in mutants at E18.5; the red channeldistinguishes autofluorescent blood cells (BC)from specific signal. (H) qPCR for Cdkn2aexpression shows ectopic expression in themutants at E14.5 and E18.5, whereas signal wasnot detected in controls. (I) ChIP-qPCR forH3K27me3 in isolated lung mesoderm at E18.5(n=4). The Cdkn2a promoter is enriched relative tonon-repressed control loci. Error bars indicates.e.m. Scale bars: 50 µm.

3734

RESEARCH ARTICLE Development (2016) 143, 3733-3741 doi:10.1242/dev.134932

DEVELO

PM

ENT

tomography (OPT) of whole-mount immunohistochemistry(IHC) for SM22α shows the ectopic smooth muscle formingdisorganized branch-like structures and sheets, as well as smallerdistal nodes (Fig. 2E,F, Movies 1,2). The phenotype is 100%penetrant, although the amount of surface area covered by ectopicsmooth muscle varies between lobes of the lung (Fig. S2A,B).To determine when this phenotype arises, we performed IHC forSM22α and SMA on lungs from E12.5, E14.5 and E16.5(Fig. 2G-L). There is no evidence for ectopic smooth muscle atE12.5 (Fig. 2J), but by E14.5 the Ezh2mesoderm-KO lungs havesporadic regions of ectopic smooth muscle in the very proximalportions of the lobes (Fig. 2K). By E16.5, the phenotype hasexpanded and extends distally through the lung, where itbecomes more noticeable by E18.5 (Fig. S2C). These datashow that the phenotype appears to develop in a proximal-to-distal manner over time. We did not observe other significantchanges in lung mesoderm lineages, including airway or vascularsmooth muscle (Fig. S2D). Thus, Ezh2 is required in lungmesoderm to prevent ectopic smooth muscle in the submesothelialmesenchyme.Since the lung mesoderm generates smooth muscle during

development, we examined the expression of pathways known toregulate this process. These data did not reveal any significantdysregulation of these pathways in the Ezh2mesoderm-KO lungs(Fig. S3A-D) (Morrisey and Hogan, 2010). This is supported bynormal vascular and airway smooth muscle development inEzh2mesoderm-KO lungs (Fig. S2D). Thus, the ectopic smoothmuscle in Ezh2mesoderm-KO lungs develops in close proximity tothe mesothelium, but does not appear to be induced by alteration inparacrine signaling factors that promote smooth muscledevelopment.

The smooth muscle master transcription factor Myocd isderepressed in ectopic smooth muscleTo explore the underlying cause of the ectopic smooth muscle inEzh2mesoderm-KO lungs, we examined expression of Myocd, a crucialtranscription factor of the smooth muscle lineage. Myocd isrecruited to the promoters of SM22α and SMA, and directlyinduces their transcription at high levels (Du et al., 2003; Wanget al., 2003). Enforced Myocd expression can reprogram non-smooth muscle cells into smooth muscle, making it a potentialcandidate for promoting ectopic smooth muscle formation inEzh2mesoderm-KO lungs (Du et al., 2003; Parmacek, 2007). In situhybridization for Myocd shows expression in the forming ectopicsmoothmuscle nodules at E14.5 (Fig. 3A) and in the ectopic smoothmuscle nodules at E18.5 (Fig. 3B).

Little is known about the epigenetic regulation of the Myocdgene. We found by ChIP-qPCR that theMyocd promoter is enrichedfor the H3K27me3mark in isolated lungmesoderm (Fig. 3C). Thus,loss of Ezh2 leads to increased Myocd expression, possibly due todirect regulation by PRC2.

The ectopic smooth muscle structures in Ezh2mesoderm-KO

lungs do not contain myocardium or vascular endotheliumOne known regulator of Myocd expression is the myocardialtranscription factor Nkx2.5, which promotes expression of acardiac-specific isoform of Myocd (Ueyama et al., 2003). Nkx2.5is expressed in cardiac myocardium and the pulmonary veinmyocardium (Lien et al., 1999; Mommersteeg et al., 2007). Recentwork has demonstrated the existence of a common cardiopulmonaryprogenitor that generates cardiac myocardium as well as pulmonaryvenous myocardium and pulmonary vascular smooth muscle in thedeveloping lung (Peng et al., 2013). Therefore, one possible

Fig. 2. Ectopic smoothmuscle developsat the peripheryof the lung. (A-D) IHC forsmooth muscle markers SM22α and SMAindicates smooth muscle developmentaround the periphery of the Ezh2mesoderm-KO

lung at E18.5 (B, arrowheads). Insetshows morphological differences betweencontrol blood vessel (C, outlined by thedashed line) and Ezh2mesoderm-KO ectopicsmooth muscle (D). A, airway; V, bloodvessel. (E,F) OPT on whole-mount IHC forSM22α with E-cadherin counterstain tooutline lung endoderm. Insets of right andleft lung lobes reveal the patterning ofectopic smooth muscle in Ezh2mesoderm-KO

lungs. (G-L) IHC for SM22α and SMAthroughout lung development shows thatectopic smooth muscle developmentinitiates by E14.5 (H,K) and expandsbetweenE14.5 andE16.5 (I,L, arrowheadsin inset). Scale bars: 20 µm in A,C; 500 µmin E,F; 100 µm in G-L.

3735

RESEARCH ARTICLE Development (2016) 143, 3733-3741 doi:10.1242/dev.134932

DEVELO

PM

ENT

hypothesis to explain the Ezh2mesoderm-KO lung phenotype is thatloss of Ezh2 in cardiopulmonary progenitors disrupts theirnormal migration into the lung and drives the ectopic expressionof smooth muscle genes. We examined expression of themyocardial markers Nkx2.5 and myosin II heavy chain (MF20antibody) in Ezh2mesoderm-KO lungs. These experiments revealedthat the ectopic smooth muscle nodules in Ezh2mesoderm-KO lungsare not of myocardial origin (Fig. 4A,B). Additionally, despitehaving a branch-like structure, the nodules of ectopic smoothmuscle do not contain vascular endothelium and they lack anobvious lumen (Fig. 4C). Thus, the ectopic smooth musclein Ezh2mesoderm-KO lungs does not contain myocardium or adisorganized vasculature.

Ezh2 is required to repress smooth muscle lineagedifferentiation in the lung mesotheliumThe mesothelium is a mesodermally derived epithelial monolayerthat envelopes the lung and reduces friction against the chest cavityand other organs. During development, a population of mesothelialcells delaminate and migrate into the lung, contributing to vascularsmooth muscle (Que et al., 2008; Dixit et al., 2013). To determinewhether the mesothelium in Ezh2mesoderm-KO lungs contributes toectopic smooth muscle, we investigated the association of thesetissues. Confocal microscopy on sectioned tissue revealed that theSMA+ ectopic smooth muscle is immediately adjacent to, or formspart of, the exterior cell layer of the lung, and is thus intimatelyassociated with the mesothelium (Fig. 5A). Co-IHC for themesothelial marker WT1 with SM22α revealed that the SM22α+

ectopic smooth muscle is in close contact with WT1+ mesothelialcells (Fig. 5B).The localization of the ectopic smooth muscle nodules suggests

that the mesothelium could provide either a niche for othermesodermal lineages to develop into smooth muscle or that itdirectly generates the ectopic smooth muscle (Fig. 5A,B). Toaddress whether loss of Ezh2 specifically in the mesothelium wouldlead to ectopic smooth muscle formation, we crossed Ezh2flox/flox

mice with Wt1cre mice and RosamTmG mice to delete Ezh2specifically in the mesothelium and lineage trace the targetedcells (Muzumdar et al., 2007; Zhou et al., 2008). The Wt1cre:Ezh2flox/flox:RosamTmG mutant embryos (hereafter Ezh2Wt1-KO)exhibit multiple heart and lung phenotypes (data not shown).

Importantly, they develop ectopic smooth muscle that expressesSM22α (Fig. 5C). The ectopic smooth muscle is closely associatedwith WT1+ cells at the periphery of the lung, similar to theEzh2mesoderm-KO lungs (Fig. 5D, Fig. S4), and is lineage traced withtheWt1cre line (Fig. 5E). Other organs, including the heart, intestineand kidney, did not exhibit formation of ectopic smooth musclenodules (data not shown). Thus, Ezh2 is required autonomously inthe pulmonary mesothelium to suppress differentiation into ectopicsmooth muscle.

Aberrant expression of Tbx15/18 in Ezh2mesoderm-KO lungsThe formation of ectopic smooth muscle at the mesothelial layerof the lung superficially resembles coronary vessel smooth musclethat develops at the surface of the heart from the epicardium.Since the lungmesothelium and epicardium are both generated fromWT1+ cells during development, we examined Ezh2mesoderm-KO

lungs for expression of the key epicardial transcription factorTbx18 (Kraus et al., 2001; Wu et al., 2013). Tbx18 null embryosexhibit reduced development of coronary smooth muscle in theheart (Wu et al., 2013). We found that Tbx18 expression isincreased in Ezh2mesoderm-KO lungs at E14.5 and E18.5 (Fig. 6A).Moreover, expression of the highly related Tbx15 gene is alsoincreased in Ezh2mesoderm-KO lungs at E14.5 and E18.5 (Fig. 6B).Consistent with previous observations that Tbx genes are regulatedby PRC2 (Boyer et al., 2006), we found that the Tbx18 andTbx15 promoters are decorated by H3K27me3 in isolated lungmesoderm (Fig. 6C,D). Immunostaining using an antibody thatrecognizes both Tbx15 and Tbx18 shows Tbx15/18 expressionin rare mesothelial and submesothelial cells in control E18.5lungs, whereas Tbx15/18+ cells are found expanded and intermixedamong ectopic smooth muscle cells in Ezh2mesoderm-KO lungs(Fig. 6E,F). Thus, Ezh2 is required to repress the transcriptionfactors Tbx15 and Tbx18. The derepression of these twofactors along with Myocd is likely to drive ectopic smoothmuscle differentiation in Ezh2mesoderm-KO and Ezh2Wt1-KO lungs(Fig. 7).

DISCUSSIONPrevious studies have demonstrated key roles for epigeneticregulation of patterning and lineage decisions in the lungendoderm (Zacharek et al., 2011; Wang et al., 2013; Galvis et al.,

Fig. 3. Myocardin is derepressed in ectopic smoothmuscle. (A,B) In situ hybridization for myocardin (Myocd) reveals overlapping expression with SM22α andSMA in ectopic smooth muscle in Ezh2mesoderm-KO lungs (arrowheads, dashed lines) at E14.5 (A) and E18.5 (B). (C) ChIP-qPCR for H3K27me3 in isolatedmesoderm at E18.5. The Myocd promoter is enriched relative to the non-repressed Actb locus and a nearby gene desert. Error bars indicate s.e.m. Scale bars:50 µm in A; 100 µm in B.

3736

RESEARCH ARTICLE Development (2016) 143, 3733-3741 doi:10.1242/dev.134932

DEVELO

PM

ENT

2015; Snitow et al., 2015). Studies on human lung diseases havedemonstrated loss of epigenetic regulation in adult lung mesodermalcell lineages (Ito et al., 2002, 2005; Huang et al., 2013). However,little is known about epigenetic regulation of the developingmesoderm of the lung. In this study, we have shown an essential rolefor Ezh2/PRC2 in restricting the differentiation of the smoothmuscle lineage from the mesothelium of the lung.Ezh2 is known to regulate the development of a variety of

mesodermal tissues, including B cell development (Su et al., 2003),skeletal muscle satellite stem cells (Juan et al., 2011; Woodhouseet al., 2013), limb bud development (Wyngaarden et al., 2011),heart development (Chen et al., 2012; Delgado-Olguín et al., 2012;He et al., 2012), fetal hematopoiesis (Mochizuki-Kashio et al.,2011) and endothelial development (Delgado-Olguín et al., 2014).In many of these tissues, and as found in our study presented here,there is a requirement for Ezh2-mediated suppression of the cellcycle inhibitor Cdkn2a, to allow progenitor cell proliferation. Moreimportantly, our studies have revealed a previously unappreciatedrole for Ezh2 in suppressing ectopic expression of the smoothmuscle transcription factor Myocd in the developing lungmesothelium. The direct regulation of Myocd by Ezh2 issupported by our finding that the Myocd promoter region isdecorated by the Ezh2 repressive mark H3K27me3.The mesothelium is a multipotent mesodermally derived

epithelial layer that surrounds the lung, and migrates into the lungto generate multiple lung mesoderm lineages including lungalveolar mesenchyme and vascular smooth muscle (Que et al.,2008; Dixit et al., 2013). Mesothelia surrounding the gut (serosalmesothelium) and heart (epicardium) similarly generate vascularsmooth muscle in these organs (Mikawa and Gourdie, 1996;

Wilm et al., 2005; Wu et al., 2013). The origins of smooth musclelineages in the lung remain poorly understood. Recent evidencedemonstrates that a multipotent cardiopulmonary progenitor(CPP) generates both vascular and airway smooth muscle in thelung (Peng et al., 2013). Despite these findings and those presentedin the present report, the overall quantitative contribution ofmesothelial, CPP, or other sources of smooth muscle in the lungremains unclear.

The restricted pattern of the ectopic smooth muscle nodules in ourEzh2 lung mesoderm mutants suggests that lack of Ezh2 creates apermissive rather than an instructive environment for Myocdexpression. Such instructive cues could come from niche effectsexerted by the WT1+ mutant mesothelium, which closely associateswith the ectopic smooth muscle. However, we did not observe anysignificant gene expression changes in signaling pathways known toregulate smooth muscle development from the mesothelium (Whiteet al., 2006; Dixit et al., 2013). We did observe derepression of bothTbx15 and Tbx18, which are transcription factors implicated inregulating muscle development (Agulnik et al., 1998; Kraus et al.,2001; Singh et al., 2005; Airik et al., 2006). In particular, Tbx18plays a crucial role in the development of epicardial smooth muscle(Wu et al., 2013). Tbx18 may function to create a niche for smoothmuscle development through the regulation of signaling factorssuch as such as Tgfβ or Notch that promote smooth muscledifferentiation in the epicardium, in lung vascular smooth muscle,and in mesothelial delamination and migration into the developinglung (Farin et al., 2007; Morimoto et al., 2010; Greulich et al.,2012). Although we have not observed alterations in these pathwaysin our studies, they could be disrupted in small subsets of cells thatare difficult to detect and whose altered differentiation leads to the

Fig. 4. Ectopic smooth muscle is neithermyocardium nor vasculature. (A) Co-IHC for SMAand Nkx2.5 and (B) SM22α and myosin II heavychain (with MF20) in E18.5 lung parenchyma andpulmonary myocardium reveals that ectopic smoothmuscle in Ezh2mesoderm-KO lungs does not expressthese myocardial markers, and that pulmonarymyocardium forms normally in mutant lungs. (C) Co-IHC for SMA and CD31 shows that ectopic smoothmuscle does not contain a vascular lumen withCD31+ endothelium. Ectopic smooth muscle isoutlined. Scale bars: 20 µm.

3737

RESEARCH ARTICLE Development (2016) 143, 3733-3741 doi:10.1242/dev.134932

DEVELO

PM

ENT

observed phenotype. Since our lineage tracing analysis shows thatEzh2Wt1-KOmice also develop ectopic smooth muscle nodules, thesedata support the concept that Ezh2 is required in mesothelial cells todirectly suppress smooth muscle lineage differentiation through theinhibition of multiple transcriptional regulators of the smoothmuscle lineage, including Myocd and Tbx18.

Lung mesothelium has been implicated in the pathogenesis ofmultiple diseases including IPF, where fibrotic lesions often initiate inthe submesothelial mesenchyme and expand internally into the lung.Expression of the lungmesothelial markerWT1 has been observed inlung fibrotic lesions, suggesting that mesothelial cells or theirdifferentiated progeny may contribute to lesion formation (Mubarak

Fig. 5. Loss of Ezh2 specifically in the lung mesotheliumleads to the formation of ectopic smooth musclenodules. (A) IHC for SMA shows that ectopic smooth muscleforms along the exterior cell layer of the lung at E18.5. (B) Co-IHC for SM22α and WT1 shows that ectopic smooth muscleforms next to the WT1+ mesothelium at E18.5. (C) IHC forSM22α and GFP lineage tracing reveals ectopic smoothmuscle (arrowheads) forming in Ezh2Wt1-KO lungs along theGFP+ mesothelium at E18.5. (D) IHC for WT1 and SM22αconfirms the close association of ectopic smooth muscle withmesothelium at E18.5. (E) IHC for SM22α and GFP lineagetracing showing colocalization of these markers by confocalmicroscopy confirms that ectopic smooth muscle arises fromthe WT1+ lineage in Ezh2Wt1-KO mutants. Scale bars: 20 µmin A,B,D,E; 100 µm in C.

Fig. 6. Repression of the T-box factors Tbx18 andTbx15 by Ezh2 in the lung mesoderm. (A,B) qPCR forthe epicardial transcription factor Tbx18 (A) and its closeparalog Tbx15 (B) shows ectopic expression of thesetranscription factors in Ezh2mesoderm-KO lungs at E14.5 andE18.5. **P<0.01, ***P<0.001. (C,D) ChIP-qPCR forH3K27me3 in isolated mesoderm at E18.5. The Tbx18 (C)and Tbx15 (D) promoters are enriched relative to non-repressed control loci. Error bars indicate s.e.m. (E,F) IHCfor Tbx15/18 reveals their expression (arrowheads) in thedeveloping lung mesothelium (E), and that expression islocalized in the ectopic smooth muscle (SM) nodules inEzh2mesoderm-KO lungs at E18.5 (F). Scale bars: 50 µm.

3738

RESEARCH ARTICLE Development (2016) 143, 3733-3741 doi:10.1242/dev.134932

DEVELO

PM

ENT

et al., 2012; Karki et al., 2014). Mesothelial contribution to fibrosishas also been demonstrated in the liver, where the lineage-tracedmesothelium invades into the submesothelial region anddifferentiates into fibroblasts following chemical injury (Asahinaet al., 2011; Li et al., 2013). The etiology of IPF is unknown, and ourresults suggest that Polycomb-mediated transcriptional repressionmight have a role in regulating mesothelial fate in the adult to preventthis disease. Our findings, along with the studies mentioned above,suggest that mesothelial cells may contribute significantly towardsfibrotic lung disease through their predisposition to differentiate intosmooth muscle or myofibroblast cells. Future studies to assesswhether pathological smooth muscle remodeling in pulmonaryhypertension, COPD, bronchiolitis obliterans and asthma involvesPolycomb-mediated regulation might provide important informationregarding the etiology of these diseases.

MATERIALS AND METHODSAnimalsDermo1cre, Wt1cre, Ezh2flox/flox and RosamTmG mice and their genotypinghave been described previously (Su et al., 2003; Yu et al., 2003; Muzumdaret al., 2007; Zhou et al., 2008). BrdU was administered intraperitoneally(60 mg/kg body weight) 90 min prior to dissection. All animal procedureswere performed in accordance with the Institute for Animal Care and UseCommittee at the University of Pennsylvania.

Float testLungs were collected from seven neonatal pups shortly after birth. Lungsfrom three cyanotic pups and four of their non-cyanotic siblings weresequentially placed in PBS to determine buoyancy. Pups were subsequentlygenotyped and matched with lung sample data.

Quantitative RT-PCRQuantitative PCR (qPCR) was performed as previously described (Snitowet al., 2015) using the primers listed in Table S1. Statistical significance wascalculated with GraphPad Prism 5 software using a two-tailed unpaired t-test.

ChIP-qPCRThe epithelial cell depletion protocol was modified fromWang et al. (2016).Lung mesoderm was isolated by digesting E18.5 lungs in 1× dispase (BDBiosciences), 480 U/ml collagenase type I (Life Technologies) and 0.33 U/mlDNase I (Roche) for 20 min at 37°C, pipetting briefly every 5 min. 5 mMEDTA was then added to prevent epithelial cells from clumping. Digestedlung cells were gently pelleted, and washed twice with PBS containing5 mM EDTA and 1% BSA. Epithelial cells were removed by incubationwith 5 µg rat anti-EpCAM/CD326 (eBioscience, 14-5791-85) in 1%BSA inPBS containing 5 mM EDTA for 30 min at 4°C, washed three times with1% BSA in PBS containing 5 mM EDTA, then incubated with 50 µl sheepanti-rat IgG Dynabeads (4.5 µm in diameter; Life Technologies) for 30 minat 4°C. Dynabeads linked to epithelial cells were removed by magnetic pull-down, and the mesoderm-containing supernatant was washed three times inPBS prior to cross-linking.

Lung mesoderm was cross-linked with 1% formaldehyde for 10 min andsonicated to an average length of 150-500 bp using a Diagenode Bioruptor

on high amplitude for 10 cycles of 30 s on/off. Immunoprecipitation wasperformed as previously described (Snitow et al., 2015), and analyzed byreal-time qPCR using the primers listed in Table S1.

HistologyTissues were fixed overnight in fresh 2% or 4% paraformaldehyde,dehydrated, and embedded in paraffin wax and sectioned at a thickness of6-8 μm. Hematoxylin and Eosin (H&E) staining was performed usingstandard procedures. The Myocd in situ hybridization probe was describedpreviously (Du et al., 2003). In situ hybridization and IHC were performedas described (Wang and Morrisey, 2010; Tian et al., 2011; Li et al., 2012;Wang et al., 2013). IHC used the following antibodies: p16/Cdkn2a (SantaCruz, sc-1661; 1:100), GFP (Aves, GFP-1020; 1:1000), BrdU (Abcam,ab6326; 1:100), SM22α (Abcam, ab10135; 1:100), SMA (Sigma, A5228;1:200), Nkx2.5 (Santa Cruz, sc-8697; 1:50), MF20 (Developmental StudiesHybridoma Bank; 1:20), PECAM (CD31; R&D Systems, MAB3628;1:500), WT1 (Santa Cruz, sc-192; 1:50), Tbx15/18 (ThermoFisher, PA5-38563; 1:50), SP-C (Chemicon, AB3786; 1:500) and Pdpn (DevelopmentalStudies Hybridoma Bank, T1 alpha 8.1.1; 1:50).

Slides were mounted with Vectashield mounting medium containingDAPI (Vector Laboratories). BrdU+ nuclei and DAPI-stained nuclei werecounted manually, assisted by the multipoint counting tool in Fiji software(Schindelin et al., 2012); epithelial cells were excluded from this analysis.Statistical analysis was performed with GraphPad Prism software using aone-tailed paired t-test. P<0.05 was considered significant.

Whole-mount immunohistochemistryThe whole-mount IHC protocol was modified from Metzger et al. (2008).Lungs were dissected and fixed overnight in a 1:4 ratio of dimethylsulfoxide:methanol, bleached for 5-10 h in a 1:1:4 ratio of 30% H2O2:DMSO:methanol, washed and stored in 100% methanol. Lungs wererehydrated in a gradient series of methanol in PBST (PBS with 0.1% Tween20): 75, 50, 25, 0%. Lungs were blocked overnight in blocking buffer (2.5%Triton X-100, 10% donkey serum and 0.05% sodium azide in PBS). Lungswere incubated for 48 h with primary antibodies SM22α (Abcam, ab10135;1:200) and E-cadherin (Sigma, U3254; 1:200) diluted in blocking buffer,washed overnight in PBST, incubated for 48 h in secondary antibody,washed overnight in PBST, then washed for several hours in PBS. Lungswere then embedded in 1% low-melting agarose, dehydrated overnight in100% methanol, then cleared overnight in a 1:2 benzoyl alcohol:benzoylbenzoate solution (BABB). OPT was imaged on a Bioptonics OPT Scanner3001 M and reconstructed with the included software packages. Lungsimaged for surface area analysis were not optically cleared with BABB.Surface area was measured using Fiji software.

AcknowledgementsWe gratefully acknowledge Yi Wang and David Frank for advice on lung digestionand epithelial cell depletion; Rachel Kadzik and Kurt Engleka for advice on whole-mount IHC; Jessica Grindheim and Shanru Li for helpful discussion and advice onchromatin immunoprecipitation; and the Histology Core at the University ofPennsylvania Cardiovascular Institute for histology services on sectioned tissue.

Competing interestsThe authors declare no competing or financial interests.

Author contributionsM.S. and E.E.M. designed the experiments, analyzed the data and wrote the paper.M.S. performed experiments. M.M.L., L.C. and S.Z. performed histologicalsectioning, in situ hybridization and IHC of sectioned tissue.

FundingThis work was supported by funding from the National Institutes of Health[HL087825, HL110942 and HL100405]. Deposited in PMC for release after 12months.

Supplementary informationSupplementary information available online athttp://dev.biologists.org/lookup/doi/10.1242/dev.134932.supplemental

Fig. 7. Model of Ezh2 repression of the smooth muscle lineage duringlung development. Ezh2 represses Myocd, the master transcription factor ofthe smooth muscle lineage. Ezh2 also represses Tbx18, which mightcontribute to a specialized epicardial-like niche that promotes smooth muscledevelopment.

3739

RESEARCH ARTICLE Development (2016) 143, 3733-3741 doi:10.1242/dev.134932

DEVELO

PM

ENT

ReferencesAgulnik, S. I., Papaioannou, V. E. and Silver, L. M. (1998). Cloning, mapping, andexpression analysis of TBX15, a newmember of the T-Box gene family.Genomics51, 68-75.

Airik, R., Bussen, M., Singh, M. K., Petry, M. and Kispert, A. (2006). Tbx18regulates the development of the ureteral mesenchyme. J. Clin. Invest. 116,663-674.

Asahina, K., Zhou, B., Pu, W. T. and Tsukamoto, H. (2011). Septum transversum-derived mesothelium gives rise to hepatic stellate cells and perivascularmesenchymal cells in developing mouse liver. Hepatology 53, 983-995.

Batra, H. and Antony, V. B. (2015). Pleural mesothelial cells in pleural and lungdiseases. J. Thorac. Dis. 7, 964-980.

Boyer, L. A., Plath, K., Zeitlinger, J., Brambrink, T., Medeiros, L. A., Lee, T. I.,Levine, S. S., Wernig, M., Tajonar, A., Ray, M. K. et al. (2006). Polycombcomplexes repress developmental regulators in murine embryonic stem cells.Nature 441, 349-353.

Bracken, A. P., Dietrich, N., Pasini, D., Hansen, K. H. and Helin, K. (2006).Genome-wide mapping of Polycomb target genes unravels their roles in cell fatetransitions. Genes Dev. 20, 1123-1136.

Bracken, A. P., Kleine-Kohlbrecher, D., Dietrich, N., Pasini, D., Gargiulo, G.,Beekman, C., Theilgaard-Monch, K., Minucci, S., Porse, B. T.,Marine, J.-C. et al. (2007). The Polycomb group proteins bind throughout theINK4A-ARF locus and are disassociated in senescent cells. Genes Dev. 21,525-530.

Chen, L., Ma, Y., Kim, E. Y., Yu,W., Schwartz, R. J., Qian, L. andWang, J. (2012).Conditional ablation of Ezh2 in murine hearts reveals its essential roles inendocardial cushion formation, cardiomyocyte proliferation and survival. PLoSONE 7, e31005.

Delgado-Olguın, P., Huang, Y., Li, X., Christodoulou, D., Seidman, C. E.,Seidman, J. G., Tarakhovsky, A. and Bruneau, B. G. (2012). Epigeneticrepression of cardiac progenitor gene expression by Ezh2 is required for postnatalcardiac homeostasis. Nat. Genet. 44, 343-347.

Delgado-Olguın, P., Dang, L. T., He, D., Thomas, S., Chi, L., Sukonnik, T.,Khyzha, N., Dobenecker, M.-W., Fish, J. E. and Bruneau, B. G. (2014).Ezh2-mediated repression of a transcriptional pathway upstream ofMmp9 maintains integrity of the developing vasculature. Development 141,4610-4617.

Dixit, R., Ai, X. and Fine, A. (2013). Derivation of lung mesenchymal lineages fromthe fetal mesothelium requires hedgehog signaling for mesothelial cell entry.Development 140, 4398-4406.

Du, K. L., Ip, H. S., Li, J., Chen, M., Dandre, F., Yu, W., Lu, M. M., Owens, G. K.and Parmacek, M. S. (2003). Myocardin is a critical serum response factorcofactor in the transcriptional program regulating smooth muscle celldifferentiation. Mol. Cell. Biol. 23, 2425-2437.

Farin, H. F., Bussen, M., Schmidt, M. K., Singh, M. K., Schuster-Gossler, K. andKispert, A. (2007). Transcriptional repression by the T-box proteins Tbx18 andTbx15 depends on Groucho corepressors. J. Biol. Chem. 282, 25748-25759.

Galvis, L. A., Holik, A. Z., Short, K. M., Pasquet, J., Lun, A. T. L., Blewitt, M. E.,Smyth, I. M., Ritchie, M. E. and Asselin-Labat, M.-L. (2015). Repression of Igf1expression by Ezh2 prevents basal cell differentiation in the developing lung.Development 142, 1458-1469.

Greulich, F., Farin, H. F., Schuster-Gossler, K. and Kispert, A. (2012). Tbx18function in epicardial development. Cardiovasc. Res. 96, 476-483.

He, A., Ma, Q., Cao, J., von Gise, A., Zhou, P., Xie, H., Zhang, B., Hsing, M.,Christodoulou, D. C., Cahan, P. et al. (2012). Polycomb repressive complex 2regulates normal development of the mouse heart. Circ. Res. 110, 406-415.

Herriges, M. and Morrisey, E. E. (2014). Lung development: orchestrating thegeneration and regeneration of a complex organ. Development 141, 502-513.

Hogan, B. L. M., Barkauskas, C. E., Chapman, H. A., Epstein, J. A., Jain, R.,Hsia, C. C. W., Niklason, L., Calle, E., Le, A., Randell, S. H. et al. (2014). Repairand regeneration of the respiratory system: complexity, plasticity, andmechanisms of lung stem cell function. Cell Stem Cell 15, 123-138.

Huang, S. K., Scruggs, A. M., Donaghy, J., Horowitz, J. C., Zaslona, Z.,Przybranowski, S., White, E. S. and Peters-Golden, M. (2013). Histonemodifications are responsible for decreased Fas expression and apoptosisresistance in fibrotic lung fibroblasts. Cell Death Dis. 4, e621.

Ito, K., Caramori, G., Lim, S., Oates, T., Chung, K. F., Barnes, P. J. and Adcock,I. M. (2002). Expression and activity of histone deacetylases in human asthmaticairways. Am. J. Respir. Crit. Care Med. 166, 392-396.

Ito, K., Ito, M., Elliott, W. M., Cosio, B., Caramori, G., Kon, O. M., Barczyk, A.,Hayashi, S., Adcock, I. M., Hogg, J. C. et al. (2005). Decreased histonedeacetylase activity in chronic obstructive pulmonary disease. N. Engl. J. Med.352, 1967-1976.

Juan, A. H., Derfoul, A., Feng, X., Ryall, J. G., Dell’Orso, S., Pasut, A., Zare, H.,Simone, J. M., Rudnicki, M. A. and Sartorelli, V. (2011). Polycomb EZH2controls self-renewal and safeguards the transcriptional identity of skeletal musclestem cells. Genes Dev. 25, 789-794.

Karki, S., Surolia, R., Hock, T. D., Guroji, P., Zolak, J. S., Duggal, R., Ye, T.,Thannickal, V. J. and Antony, V. B. (2014). Wilms’ tumor 1 (Wt1) regulates

pleural mesothelial cell plasticity and transition into myofibroblasts in idiopathicpulmonary fibrosis. FASEB J. 28, 1122-1131.

Kraus, F., Haenig, B. and Kispert, A. (2001). Cloning and expression analysis ofthe mouse T-box gene Tbx18. Mech. Dev. 100, 83-86.

Lee, T. I., Jenner, R. G., Boyer, L. A., Guenther, M. G., Levine, S. S., Kumar,R. M., Chevalier, B., Johnstone, S. E., Cole, M. F., Isono, K.-i. et al. (2006).Control of developmental regulators by Polycomb in human embryonic stem cells.Cell 125, 301-313.

Li, S., Wang, Y., Zhang, Y., Lu, M. M., DeMayo, F. J., Dekker, J. D., Tucker, P. W.and Morrisey, E. E. (2012). Foxp1/4 control epithelial cell fate during lungdevelopment and regeneration through regulation of anterior gradient 2.Development 139, 2500-2509.

Li, Y., Wang, J. and Asahina, K. (2013). Mesothelial cells give rise to hepaticstellate cells and myofibroblasts via mesothelial-mesenchymal transition in liverinjury. Proc. Natl. Acad. Sci. USA 110, 2324-2329.

Lien, C. L., Wu, C., Mercer, B., Webb, R., Richardson, J. A. and Olson, E. N.(1999). Control of early cardiac-specific transcription of Nkx2-5 by a GATA-dependent enhancer. Development 126, 75-84.

Metzger, R. J., Klein, O. D., Martin, G. R. and Krasnow, M. A. (2008). Thebranching programme of mouse lung development. Nature 453, 745-750.

Mikawa, T. and Gourdie, R. G. (1996). Pericardial mesoderm generates apopulation of coronary smooth muscle cells migrating into the heart along withingrowth of the epicardial organ. Dev. Biol. 174, 221-232.

Mochizuki-Kashio, M., Mishima, Y., Miyagi, S., Negishi, M., Saraya, A., Konuma,T., Shinga, J., Koseki, H. and Iwama, A. (2011). Dependency on the polycombgene Ezh2 distinguishes fetal from adult hematopoietic stem cells. Blood 118,6553-6561.

Mommersteeg, M. T., Brown, N. A., Prall, O. W. J., de Gier-de Vries, C., Harvey,R. P., Moorman, A. F. M. and Christoffels, V. M. (2007). Pitx2c and Nkx2-5 arerequired for the formation and identity of the pulmonary myocardium. Circ. Res.101, 902-909.

Morimoto, M., Liu, Z., Cheng, H.-T., Winters, N., Bader, D. and Kopan, R. (2010).Canonical Notch signaling in the developing lung is required for determination ofarterial smooth muscle cells and selection of Clara versus ciliated cell fate. J. CellSci. 123, 213-224.

Morrisey, E. E. and Hogan, B. L. M. (2010). Preparing for the first breath: geneticand cellular mechanisms in lung development. Dev. Cell 18, 8-23.

Mubarak, K. K., Montes-Worboys, A., Regev, D., Nasreen, N., Mohammed,K. A., Faruqi, I., Hensel, E., Baz, M. A., Akindipe, O. A., Fernandez-Bussy, S.et al. (2012). Parenchymal trafficking of pleural mesothelial cells in idiopathicpulmonary fibrosis. Eur. Respir. J. 39, 133-140.

Mutsaers, S. E., Birnie, K., Lansley, S., Herrick, S. E., Lim, C.-B. and Prêle, C. M.(2015). Mesothelial cells in tissue repair and fibrosis. Front. Pharmacol. 6, 113.

Muzumdar, M. D., Tasic, B., Miyamichi, K., Li, L. and Luo, L. (2007). A globaldouble-fluorescent Cre reporter mouse. Genesis 45, 593-605.

Parmacek, M. S. (2007). Myocardin-related transcription factors: criticalcoactivators regulating cardiovascular development and adaptation. Circ. Res.100, 633-644.

Peng, T., Tian, Y., Boogerd, C. J., Lu, M. M., Kadzik, R. S., Stewart, K. M., Evans,S. M. and Morrisey, E. E. (2013). Coordination of heart and lung co-developmentby a multipotent cardiopulmonary progenitor. Nature 500, 589-592.

Que, J., Wilm, B., Hasegawa, H., Wang, F., Bader, D. and Hogan, B. L. M. (2008).Mesothelium contributes to vascular smooth muscle and mesenchyme duringlung development. Proc. Natl. Acad. Sci. USA 105, 16626-16630.

Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch,T., Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B. et al. (2012). Fiji: anopen-source platform for biological-image analysis. Nat. Methods 9, 676-682.

Singh, M. K., Petry, M., Haenig, B., Lescher, B., Leitges, M. and Kispert, A.(2005). The T-box transcription factor Tbx15 is required for skeletal development.Mech. Dev. 122, 131-144.

Snitow, M. E., Li, S., Morley, M. P., Rathi, K., Lu, M. M., Kadzik, R. S., Stewart,K. M. and Morrisey, E. E. (2015). Ezh2 represses the basal cell lineage duringlung endoderm development. Development 142, 108-117.

Su, I.-H., Basavaraj, A., Krutchinsky, A. N., Hobert, O., Ullrich, A., Chait, B. T.and Tarakhovsky, A. (2003). Ezh2 controls B cell development through histoneH3 methylation and Igh rearrangement. Nat. Immunol. 4, 124-131.

Tian, Y., Zhang, Y., Hurd, L., Hannenhalli, S., Liu, F., Lu, M. M. and Morrisey,E. E. (2011). Regulation of lung endoderm progenitor cell behavior by miR302/367. Development 138, 1235-1245.

Ueyama, T., Kasahara, H., Ishiwata, T., Nie, Q. and Izumo, S. (2003). Myocardinexpression is regulated by Nkx2.5, and its function is required forcardiomyogenesis. Mol. Cell. Biol. 23, 9222-9232.

Wang, Y. and Morrisey, E. E. (2010). Regulation of cardiomyocyte proliferation byFoxp1. Cell Cycle 9, 4251-4252.

Wang, Z., Wang, D.-Z., Pipes, G. C. T. and Olson, E. N. (2003). Myocardin is amaster regulator of smooth muscle gene expression. Proc. Natl. Acad. Sci. USA100, 7129-7134.

Wang, Y., Tian, Y., Morley, M. P., Lu, M. M., Demayo, F. J., Olson, E. N. andMorrisey, E. E. (2013). Development and regeneration of Sox2+ endoderm

3740

RESEARCH ARTICLE Development (2016) 143, 3733-3741 doi:10.1242/dev.134932

DEVELO

PM

ENT

progenitors are regulated by a Hdac1/2-Bmp4/Rb1 regulatory pathway. Dev. Cell24, 345-358.

Wang, Y., Frank, D. B., Morley, M. P., Zhou, S., Wang, X., Lu, M. M., Lazar, M. A.and Morrisey, E. E. (2016). HDAC3-dependent epigenetic pathway controls lungalveolar epithelial cell remodeling and spreading via miR-17-92 and TGF-betasignaling regulation. Dev. Cell 36, 303-315.

White, A. C., Xu, J., Yin, Y., Smith, C., Schmid, G. and Ornitz, D. M. (2006). FGF9and SHH signaling coordinate lung growth and development through regulation ofdistinct mesenchymal domains. Development 133, 1507-1517.

Wilm, B., Ipenberg, A., Hastie, N. D., Burch, J. B. E. and Bader, D. M. (2005). Theserosal mesothelium is a major source of smooth muscle cells of the gutvasculature. Development 132, 5317-5328.

Woodhouse, S., Pugazhendhi, D., Brien, P. and Pell, J. M. (2013). Ezh2maintains a key phase of muscle satellite cell expansion but does not regulateterminal differentiation. J. Cell Sci. 126, 565-579.

Wu, S.-P., Dong, X.-R., Regan, J. N., Su, C. and Majesky, M. W. (2013). Tbx18regulates development of the epicardium and coronary vessels. Dev. Biol. 383,307-320.

Wyngaarden, L. A., Delgado-Olguin, P., Su, I.-h., Bruneau, B. G. and Hopyan, S.(2011). Ezh2 regulates anteroposterior axis specification and proximodistal axiselongation in the developing limb. Development 138, 3759-3767.

Yin, Y., White, A. C., Huh, S.-H., Hilton, M. J., Kanazawa, H., Long, F. and Ornitz,D. M. (2008). An FGF-WNT gene regulatory network controls lung mesenchymedevelopment. Dev. Biol. 319, 426-436.

Yu, K., Xu, J., Liu, Z., Sosic, D., Shao, J., Olson, E. N., Towler, D. A. and Ornitz,D. M. (2003). Conditional inactivation of FGF receptor 2 reveals an essential rolefor FGF signaling in the regulation of osteoblast function and bone growth.Development 130, 3063-3074.

Zacharek, S. J., Fillmore, C. M., Lau, A. N., Gludish, D.W., Chou, A., Ho, J. W. K.,Zamponi, R., Gazit, R., Bock, C., Jager, N. et al. (2011). Lung stem cell self-renewal relies on BMI1-dependent control of expression at imprinted loci. CellStem Cell 9, 272-281.

Zhou, B., Ma, Q., Rajagopal, S., Wu, S. M., Domian, I., Rivera-Feliciano, J.,Jiang, D., von Gise, A., Ikeda, S., Chien, K. R. et al. (2008). Epicardialprogenitors contribute to the cardiomyocyte lineage in the developing heart.Nature 454, 109-113.

3741

RESEARCH ARTICLE Development (2016) 143, 3733-3741 doi:10.1242/dev.134932

DEVELO

PM

ENT