Proliferation and Differentiation of Smooth Muscle CellPrecursors Occurs Simultaneously During theDevelopment of the Vessel WallSANG HOON LEE,1 JILL E. HUNGERFORD,2,3 CHARLES D. LITTLE,3 AND M. LUISA IRUELA-ARISPE1*1Department of Pathology, Harvard Medical School and Beth Israel Deaconess Medical Center, Boston, Massachusetts2Department of Cell Biology, University of Virginia, Charlottesville, Virginia3Department of Cell Biology and Cardiovascular Developmental Biology Center, Medical University of South Carolina,Charleston, South Carolina

ABSTRACT Formation of the blood vesselwall depends on the recruitment, proliferation,and differentiation of smooth muscle cell (SMC)precursors. The temporal events associated withthe onset of expression of several SMC proteinshave been well characterized in mouse and avianspecies. However, the timing of cell proliferationduring this process has not been explored. Moreimportantly, it has not been clear whether com-mitment to the smooth muscle pathway pre-cludes proliferation during development. In thepresent study, we have determined the kinetics ofreplication in developing chick aortae betweendays 2.5 and 19 and have correlated these datawith the expression of various SMC differentia-tion markers. We found that proliferation of aor-tic SMC precursors occurs in two waves; an earlyphase of rapid proliferation (15–17%; betweendays 4 and 12), and a second phase, when replica-tion was reduced to less than 5% (days 16 tohatching). Proliferation of SMC during the firstwave occurred concomitantly with the progres-sive accumulation of SMC contractile proteins,such as SMa-actin, calponin,myosin heavy chain,and the 1E12 antigen. We also found that therelative proliferation capacity within each com-partment of the vessel wall, ie., intima, media,and adventitia varies throughout development.Approximately, 55–63%of all replicating cellswerefound in the tunica adventitia from days 6 to 12,whereas 35% were found in the tunica media(tunica media:adventitia 5 1:2). This ratio wasinverted after day 12, when most of the replicat-ing cells were located in the tunica media (tunicamedia:adventitia 5 2:1). In addition, we observeda ventral-to-dorsal gradient in the proliferationof SMC precursors between days 2.5 and 5. Theventral-to-dorsal proliferation gradientwas simi-lar to thepreviously describeddifferential expres-sion of two early SMC markers: a-actin and the1E12 antigen. These data support the conceptthat a polarity exists either in the pool of SMCprecursors or, in expression of factors that regu-late recruitment of presumptive SMC. Dev. Dyn.209:342–352, 1997. r 1997 Wiley-Liss, Inc.

Key words: aorta; chick; histone 3; proliferation;smoothmuscle cell

INTRODUCTION

Morphogenesis of the dorsal aorta is initiated by thefusion of two lateral channels of mesodermal cells at12–24 hr in the chicken embryo. Later, these mesoder-mal cells differentiate into endothelium and constitutethe tunica intima (Coffin and Poole, 1988; Poole andCoffin, 1989; Noden, 1993). By day 2, smooth musclecell (SMC) precursors are recruited from the localmesenchyme to form one to two layer-cells (Gonzalez-Crussi, 1971; Hirakow and Hiruma, 1981; Hungerfordet al., 1996). Presumptive SMC proliferate and differen-tiate to form the tunica media. The tunica media issurrounded by a layer of loose mesenchymal cells andextracellularmatrix, that constitute the tunica adventi-tia. In contrast to skeletal and cardiac muscle, SMCderive from diverse subpopulations of mesodermal andneural crest cells (Le Lievre and Le Douarin, 1975;Rosenquist and Beall, 1990; Hood and Rosenquist,1992; Miano et al., 1995; Kirby and Waldo, 1995;Topouzis and Majesky, 1996). To date, the molecularmechanisms involved in the embryonic SMC recruit-ment, proliferation, and differentiation have not beenelucidated.The expression of several contractile proteins has

provided a means to identify presumptive SMC duringdevelopment. These include SM a-actin, smooth musclemyosin heavy chain (SM-MHC), along with calciumregulatory proteins such as calponin. Smooth musclea-actin is the earliest marker expressed by presump-tive SMC that surround the dorsal aortic sac at embry-onic day 2 in chicken and quail (Duband et al., 1993;Hungerford et al., 1996).At this time, SM a-actin is alsodetected in cardiomyocytes and skeletal muscle (Skalli

Abbreviations used: Dig, digoxigenin; FBS, fetal bovine serum;MoAb, monoclonal antibody; PBS, phosphate buffered saline; SMC,smooth muscle cell; SM-MHC, smooth muscle myosin heavy chain.Grant sponsor: American Heart Association; Grant number: 96-

1218.*Correspondence to: Dr. M. Luisa Iruela-Arispe, Department of

Pathology, Beth Israel Deaconess Medical Center, 330 Brookline Ave,Boston, MA02215. E-mail: liruela@bidmc.harvard.eduReceived 16 January 1997; Accepted 14April 1997

DEVELOPMENTALDYNAMICS 209:342–352 (1997)

r 1997 WILEY-LISS, INC.

et al., 1986; Glukhova et al., 1988; Ruzicka andSchwartz, 1988; Sawtell and Lessard, 1989; Sugi andLough, 1992). Later, expression is lost in striatedmuscle and becomes specifically associated with vascu-lar and visceral SMC. Recently, a novel SMC marker,the 1E12 antigen, had been identified (Hungerford etal., 1996). The 1E12 antigen is detected first in presump-tive SMC of dorsal aorta by embryonic day 3 in quailand continues to be expressed in the aorta throughdevelopment and in the adult. This marker has beenshown to identify SMC in both vascular and visceralorgans (Hungerford et al., 1996). Recent studies sug-gest that 1E12 antibody recognizes an isotype of SMa-actinin (Hungerford et al., 1997). Calponin is anactin- and tropomyosin-binding protein implicated inSMC contraction (Takahashi and Nadal-Ginard, 1991;Applegate et al., 1994). Calponin is first detected in theSMC of dorsal aortae by day 6 of avian development(Duband et al., 1993), however the protein is expressedby multiple myogenic cell lineages during embryogen-esis (Miano and Olson, 1996). The mRNA of smoothmusclemyosin heavy chain (SM-MHC), themost distinc-tive marker of SMC, is first detected in the developingaorta at 10.5 days postcoitum duringmouse embryogen-esis (Miano et al., 1995).The temporal relationship between SMC prolifera-

tion and differentiation in vivo is not well understood.Studies performed in vitro have demonstrated that thehighly proliferative and synthetic SMC phenotype isassociated with the loss of contractile proteins such asmyosin and tropomyosin, and with some loss of SMa-actin (Chamley-Campbell et al., 1979; Thyberg et al.,1990; Owens, 1995). Recently, Belknap and colleagues(1996) showed that a rapid increase of tropoelastinexpression, in the late fetal and early postnatal period,was associated with a decrease in SMC proliferationduring rat aortic development. A major conclusion ofthat study was that the synthetic phenotype, as evalu-ated by tropoelastin mRNA, was not compatible withproliferation. The findings were similar to those seen invascular disease. However, the study was performedrather late in development, when most of the eventsassociated with SMC differentiation had already oc-curred. Therefore, these data do not clearly differenti-ate between proliferation and commitment ofmesenchy-mal cells to the SMC differentiation pathway duringvascular development. Thus, it remains to be seenwhether proliferation is compatible with SMC differen-tiation during vessel development or whether theseprocesses are mutually exclusive.In the present study, we have noted that sequential

expression of several SMC differentiation markers oc-curs in association with relatively high periods ofproliferation in the tunica media. Two distinct waves ofproliferation were observed during aortic development;one from days 4–12 (17–15%) and a second, from day 12to hatching (5–2%). The first wave of proliferationcoincides with the onset of expression of several SMCmarkers: SM a-actin, 1E12 antigen, and calponin. The

second wave of proliferation occurs in cells that expressSM-MHC in addition to the earlier markers. Theseresults provide the first detailed analysis of the relation-ship between SMC proliferation and the temporalexpression of SMC differentiation markers during vas-cular development. We also found evidence for a ventralto dorsal gradient in the rate of proliferation andrecruitment of SMC precursors. These data providesupport for the hypothesis that environmental signalsdifferentially regulate SMC recruitment to the ventralsurface of the developing aorta.

RESULTSExpression of SMMarkers During Developmentof the ChickenAorta

To evaluate the relationship between proliferationand differentiation in the vascular wall, we first con-firmed the developmental expression pattern of fourpreviously characterizedmarkers of presumptive SMCs:SM a-actin, 1E12 antigen, calponin, and SM-myosin(Kuro-o et al., 1991; Drew et al., 1991; Aikawa et al.,1993; Duband et al., 1993; Hungerford et al., 1996).Although expression of these markers has been exam-ined previously, such an analysis has not been compiledin a single species. Furthermore, these data wereessential to establish a framework for the subsequentproliferation studies.In avian development, formation of the dorsal aorta

is initiated by fusion of two lateral aortae that cometogether following gastrulation. Fusion of the two lat-eral aortae occurs at stage 16 (day 2). The newly formeddescending aorta is then comprised of a single layer ofendothelial cells. Shortly after (stage 17), one or twolayers of mesenchymal cells loosely associate with theendothelium of the developing vessel. It is critical in astudy of this nature to have consistency at sampling theaorta specimens, since origin, distribution, and organi-zation of smooth muscle cell precursors varies in thethoracic and abdominal regions. Figure 1 illustrates theplane of sectioning use in all transversal sampling.Expression of SM a-actin was first detected at this

time in the presumptive SMCs associated with theaortic endothelium at embryonic stage 17 (data notshown). A longitudinal section of embryonic stage 19(day 3) immunostained with SM a-actin showed posi-tive cells located adjacent to endothelium (Fig. 2A). Asthe vessel matures, several layers of cells becomejuxtaposed, and this is coincident with an increase inthe levels of SM a-actin (Fig. 2B). Expression of SMa-actin continued to increase throughout developmentand clearly defines the tunica media (Fig. 2C). A secondearly marker of presumptive SMCs, 1E12 antigen, wasfirst detected at stage 18 (days 2.5–3); this represents a8–12 hr delay from the onset of SM a-actin expression.The 1E12 antigen was first detected in the most proxi-mal layers ofmesenchymal cells adjacent to the endothe-lium (day 3; Fig. 2D). This marker was later expressedin the entire media of the vessel wall, as well as in theSMC layers of the vasa vasorum present in the adventi-

343PROLIFERATION AND DIFFERENTIATION OF THE VASCULAR WALL

tia of the aorta in day 11 embryos (Fig. 2E). Atembryonic day 18, the tunica media is comprised ofseven to eight layers of SMCs, the 1E12 antigen wasexpressed throughout this layer (Fig. 2F). No signalwas detected in the adventitia.Calponin is another SMC-specific contractile protein

found exclusively in smooth muscle cells, both duringdevelopment and in the adult (Duband et al., 1993;Samaha et al., 1996). Calponin was not detected in thedescending aorta by embryonic day 4 (Fig. 2G), but wasobserved in aortic smooth muscle cells at day 6 (Fig.2H). Expression was initially confined to the innerlayers of tunica media at day 10 (data not shown) andthen progressively accumulated throughout the tunicamedia (Fig. 2I), similar to 1E12 and SM a-actin.SM-MHC was not detected in the vessel wall at day 8(Fig. 2J), but was first observed in smooth muscle cellsadjacent to the tunica intima at day 10 (data notshown). Progressive expansion of SM-MHC expressionfrom inner to outer SMC layers in the tunica media wasevident by embryonic day 11 (Fig. 2K). A strong signalfor SM-MHC was observed throughout the tunica me-dia by day 18 (Fig. 2L). Counterstaining of nuclei withHoechst (blue) is presented in all panels to provide asense for the architecture of the vessel wall and toemphasize the contrast between ‘‘differentiated’’ and‘‘undifferentiated’’ mesenchyme.In summary, our results are consistent with previous

studies done in chicken and mouse; SM a-actin is thefirst marker to be detected at embryonic day 2.5,followed by 1E12 at day 3. Calponin is then evident atday 6, followed by SM-MHC at days 10–11. Oncepresent, expression of these markers is constant in themedia through the adult. Interestingly, a gradient ofexpression within the media (from the lumen to the

adventitia) was evident with all markers and at differ-ent times during the development of the aorta; i.e., atday 2.5 with a-actin, at day 4 with 1E12, at day 6 withcalponin, and at day 12 withmyosin. The data suggest arecurrent radial wave of differential gene expression inthe vessel wall.

Analysis of Cell Proliferation During VesselDevelopment In Vivo

Temporal maturation of the vessel wall is the directresult of two processes: (1) an increase in cell number,and (2) cellular differentiation, as indicated by progres-sive expression of SMC markers. It has been postu-lated, although never entirely proven, that these twoevents are mutually exclusive; i.e., that full differentia-tion of SMC is not consistent with a proliferativephenotype. To correlate the temporal expression ofSMC differentiation markers with growth rates in theaorta, in situ hybridization for histone 3 was performedon sections from embryonic days 4, 6, 8, 12, 16, 19, andadult. Histone 3 provided a means to identify cells inthe S-phase of the cell cycle and committed to mitosis.To examine the proliferative rate during vascular devel-opment, histone 3 positive cells and total cell number,as assessed by toluidine blue, were counted. Percentproliferation was obtained for all the time points indi-cated. Figure 3 shows a time-course analysis of theproliferation rate in the developing dorsal aorta. Thehighest proliferation was observed at day 8, neverthe-less elevated proliferative activity of at least 15–17%was maintained from day 4 to day 12. Growth rate wassubsequently reduced to 5% at day 16, and 2.5% at day19 prior to hatching. No proliferation was detected inthe adult aorta.To determine whether replicating embryonic SMC of

the tunica media expressed SMC differentiation mark-ers, we examined the distribution of histone 3 positivecells in the early aortic wall and compared it to theexpression of several SMC markers. By embryonic day5, SM a-actin was expressed in every cell of the tunicamedia (Fig. 4A,C). Interestingly, replicating cells werealso distributed throughout the tunica media (Fig.4B,D). The 1E12 antigen was expressed in the entiretunicamedia by day 8 (Fig. 4E), at which time prolifera-tion of SMCs was also evident in the tunica media (Fig.4F). By day 12, the expression of most SM differentia-tion markers including SM-MHC (Fig. 4G) occurredcoincidently with proliferation (Fig. 4H). Note thatproliferating cells were found at all levels, i.e., nearlumen and near the tunica adventitia. These resultsindicate that the proliferation of SMCs within thevessel wall occurs simultaneously with expression ofSMC differentiation markers.To evaluate the relative changes in total cell number

within each compartment of the vessel wall, we countedtoluidine blue stained cells in the intima, media, andadventitia from day 4 to adult. The ratio of cell number(tunica media:adventitia) was approximately 1:2 fromday 6 to day 8 and decreased to 2:3 at day 12. By day 16,

Fig. 1. Schematic representation of a day 3 chicken embryo. The lineA–B indicates the plane of sectioning used in sampling aortae for thisstudy. A single transversal section was made caudal to the cardiac regionand at the level of the anterior limb.

344 LEE ET AL.

the increasing population of the cells in the tunicamedia caused themedia to adventitia ratios to switch to3:2. In the adult aorta, the tunica media to adventitiaratio is 2:1 (Fig. 5A).To examine the proliferation rate in the vessel, we

counted histone 3 positive cells in the tunica intima,media, and adventitia. As shown in Figure 5B, the

growth rate of endothelial cells in the tunica intimadecreased rapidly during development. In contrast, theproliferation rate of mesenchymal cells in the tunicaadventitia was consistently high (approximately 60%)during day 6 to day 12, compared to approximately 35%in tunica media. Thus, the ratio of proliferating cells inthe tunica media:tunica adventitia was 1:2. After day

Fig. 2. Expression of SMC differentiation markers in the developingchick aorta. Cross sections of dorsal aortae at embryonic days 3 (A, D), 4(G), 6 (H), 8 (J), 11 (B, E, K ), and 18 (C, F, I, L) were immunolabeled withSM a-actin MoAb (A–C), 1E12 MoAb (D–F), calponin MoAb (G–I), orSM-myosin polyclonal Ab (J–L; in green). Hoechst 33258 was used as acounterstain to reveal nuclei (in blue). SM a-actin and 1E12 antibodiesidentified presumptive SMCs associated with the endothelium of the aortaat embryonic day 3 (A, D). High expression of SM a-actin (B, C) and 1E12(E, F) was detected in the tunica media at day 11 (B, E) and 18 (C, F). Note

that 1E12 positive cells were also found in a small vessel (branch or vasavasorum) present in the adventitia (E; arrow). SMC precursors associatedwith the endothelium were negative for calponin at embryonic day 4 (G;arrows). Expression of calponin in the vessel wall was first detected atdays 5–6 (H) and increased calponin expression was observed in SMCsby day 18 (I). SM-myosin was not detected in the vessel wall at day 8 (J).Several layers of SMCs were immunostained with SM-myosin antibody byday 11 (K). SM-myosin was highly expressed in themedia at day 18 (L). D,Dorsal surface; L, lumen; V, ventral surface. Scale bar 5 50 µm.

345PROLIFERATION AND DIFFERENTIATION OF THE VASCULAR WALL

16, more than 60% of proliferating cells were found intunica media and the ratio (tunica media:adventitia)was switched to 3:2. This transition mirrored thechange in total cell number shown in Figure 4A.

AVentral to Dorsal Gradient of SMCDifferentiation is Evident During EarlyStages of Development

Morphological analysis of dorsal aortae from embry-onic days 2–6 showed a ventral to dorsal assymmetry inthe condensation of mesenchymal cells during thedevelopment of the tunica media. In Figure 6A (aorta atday 4), three to four layers of condensed mesenchymeare evident on the ventral side of the vessel; while onlyone to two layers can be observed on the dorsal surfaceof the descending aorta. To evaluate whether thisgradient was also observedwith regard to SMCdifferen-tiation, we performed SM a-actin immunohistochemis-try. In a cross section of descending aortae at embryonicday 3, presumptive SMCs appeared assymmetricallydistributed in the vessel wall. Specifically two to threelayers of positive cells were consistently associatedwith the ventral surface (Fig. 6B; V), while a singlelayer of cells was detected at the dorsal surface (Fig. 6B;D). The majority of SM a-actin positive cells were foundon the ventral side of the descending aorta at day 4 (Fig.6C,D). In addition, SM a-actin was expressed by a

group of mesenchymal cells which were distal to theintima and appeared to bemigrating towards the vesselwall (Fig. 6C). In order to determine whether theventral to dorsal gradient was exclusive to SM a-actin,we performed similar experiments with 1E12 antibody.Figure 6E shows a longitudinal section of aorta from aday 3 embryo. Staining was evident in a region whereseveral layers of mesoderm surrounded the ventralsurface of the vessel, while only fragmented stainingwas evident on the dorsal side.A ventral to dorsal pattern of presumptive SMC

recruitment and/or differentiation was also observed inmouse descending aorta at 10.5 days postcoitum, asindicated by expression of SM a-actin (Fig. 6F). There-fore, the ventral to dorsal gradient of differentiationappears to be a general phenomena conserved in bothavian and mammalian development.

Cellular Proliferation is Consistent With theVentral to Dorsal Gradient of Differentiation

In order to examine the proliferative rates of presump-tive SMCs on the ventral and dorsal surface of thedeveloping aorta, the expression of histone 3 mRNAwas determined by in situ hybridization and prolifera-tion was evaluated on both sides. A marked differencein the number of histone 3 positive cells was evidentbetween the ventral and dorsal surface of the vesselfrom day 3 to day 6 (Fig. 7). Cell replication was clearlyassymetric favoring the ventral surface of the aorta.This gradient of cell division was less evident but stillpresent on subsequent days 7–8 (Fig. 7). By day 10, theproliferation rate on both ventral and dorsal sides wasessentially identical (Fig. 7). The ratio of proliferationon the ventral and dorsal surfaces was 3:1 at day 4, and2:1 on days 6 and 8.

DISCUSSION

The regulation of SMC proliferation and differentia-tion during vascular development is poorly understood.In the present study, we have used a combination ofimmunofluorescence and in situ hybridization tech-niques to establish the temporal pattern of proliferationin the developing aorta and its relationship to SMCdifferentiation. Main conclusions from this study in-clude: (1) proliferation of SMC during embryogenesisoccurs concomitantly with differentiation, as indicatedby expression of four well recognized SMC markersduring periods of consistently high proliferation rates;(2) morphogenesis of the tunica media occurs, at leastin part, by mitosis of cells within the media, althoughthe contribution of cells from the adventitia has notbeen excluded by this study; (3) there is a ventral-to-dorsal gradient of proliferation and differentiationthroughout the length of the aorta. This last pointmight suggest the existence of localized signals (growthfactors, extracellular matrix, etc.) that regulate mesen-chymal commitment to SMC phenotype.It has been proposed that mechanisms that regulate

proliferation during vascular development are also

Fig. 3. Quantification of histone 3 positive cells during the develop-ment of the dorsal aorta. In situ hybridization with histone 3 was used toidentify proliferating cells in the chick aorta at embryonic days 4, 8, 12, 16,19, and 30-week-old adult. Note the decrease in proliferating cells afterday 16. Data are expressed as the percentage of histone 3 positive cells,relative to total nuclei. Difference in distribution of proliferating cells inpooled specimens in early (4–12) and late (16–210 days) stages ofdevelopment is significant (Contingenct X2 for 1 degree of freedom withYates’ correction 5 3.85: P , 0.05).

346 LEE ET AL.

Fig. 4. Localization of replicating SMCs in the tunica media. Crosssections of dorsal aortae at embryonic days 5 (A, B, C, D ), 8 (E, F), and 12(G, H) were immunolabeled with SM a-actin MoAb (A, C), 1E12 MoAb (E),and SM-myosin polyclonal antibody (G) and hybridized with a histone 3riboprobe (B, D, F, H). Serial sections were used inAand B to ask whether

expression of SM a-actin was coincident with progression through the cellcycle. C and D correspond to the areas indicated by squares in A and Brespectively. Arrows in C and D indicate proliferating cells (D) that expressSM a-actin (C). Proliferation coincided with onset of 1E12 antigenexpression (E, F) and SM-myosin (G, H). L, Lumen. Scale bar 5 50 µm.

involved in response to vascular injury in adult vessels.Vascular SMC replication is an important event in theformation of atherosclerotic lesions, hypertension, andrestenosis of arteries (Campbell and Campbell, 1985;Schwartz and Reidy, 1987; Ross, 1993). Atherosclerotic

and restenotic injury are caused by extensive pheno-typic modulation from the ‘‘contractile’’ to ‘‘synthetic’’phenotype that, among several features, is character-ized by an increase in growth capacity (Schwartz et al.,1986; Ross, 1993; Owens, 1995). The mechanisms thatregulate SMCmodulation to the synthetic and prolifera-tive state has been the focus of several laboratories.Research in this area revealed that a number of cyto-kines are involved in this transition, including PDGFand TGF-b (Collins et al., 1987; Hannan et al., 1988;Sato et al., 1990; Ross et al., 1993). Nevertheless, itappears certain that multiplicity of molecular path-ways are involved in pathological SMC replicationassociated with arterial lesions. Therefore, a moreconcrete understanding of the processes that contributeto SMC growth and recruitment during developmentmight provide insights on the proliferation of SMCduring vascular disease.It has been suggested that there is a cause-effect

relationship between SMC differentiation and cellularproliferation. Several groups have shown that in pri-mary culture, proliferation of adult smooth muscle cellsis followed by the loss of myosin and actin containingfilaments (Chamley-Campbell et al., 1979; Thyberg etal., 1990). However, other experiments have indicatedthat the differentiation of SMCs in vitro is not totallydependent on a withdrawal from a proliferative state(Owens et al., 1986; De Mey et al., 1989; Owens, 1995).Our data supports this later concept, since proliferationof SMCs during vessel development, especially fromembryonic days 6 to 16, occurs concomitantly withexpression of contractile proteins.We also evaluated changes in the proliferative rate of

cells in the tunica intima, media, and adventitia, andits relationship with the development and maturationof the vessel wall. The ratio of proliferating cellsbetween tunica media and adventitia was 1:2 from day6 to day 12, when more than 60% of proliferating cellswere localized in the tunica media. The ratio betweentunica media:adventitia was switched to 2:3 after day16. The kinetics of growth rate in both tunicamedia andadventitia appear to be essential to the increase invessel size and maturation. Nevertheless, the growth oftunica media appears to occur, at least in part, byproliferation within themedia throughout developmentand not only by addition of cells from the adventitia, asoriginally proposed.Our studies were performed exclusively in the de-

scending aorta. It is not clear whether the rate/patternof proliferation in the ascending aorta or other ‘‘morecranial’’ vessels will be similar to the one described inthis study. Unlike in the abdominal aorta, the tunicamedia of the aortic arch arteries, for example, hassignificant contribution from neural crest cell popula-tions (Le Lievre and Le Douarin, 1975; Kirby andWaldo, 1990; Topouzis and Majesky, 1996). Rosenquistand Beall (1990) have reported that smooth musclea-actin is first expressed from the medial region of theaortic arch arteries nearest the heart in chick embryos

Fig. 5. Quantification of proliferating and total cell number in the threetunicas of the vessel wall. A: Distribution of cells in the tunica intima,media, and adventitia. Total cell number was assessed by countingtoluidine blue labeled cells within three vascular layers. Data are ex-pressed as the percentage of cells per compartment. B: Distribution ofproliferating cells in the vessel wall during development. Time courseanalysis of proliferation rates in each compartment was determined bycounting histone 3 positive cells in each compartment. (Statistical analy-sis of the three compartments differ significantly from random and havecontingenct X2 with Yates’ correction 5 13.85: P , 0.001 [for A] and 19.3:P , 0.001 [for B].)

348 LEE ET AL.

Fig. 6. A ventral to dorsal pattern in SMC recruitment. A: Toluidineblue staining of a cross section of embryonic day 4 aorta. Note themultilayered condensation of mesenchymal cells on the ventral surface(V) of the vessel wall, versus the poor condensation of mesenchymal cellson the dorsal surface (D). B: Transverse section of a day 3 dorsal aortaimmunostained with SM a-actin MoAb and counterstained with Hoechst33258 (blue nuclear label). C: Day 4 aorta was immunolabeled with SMa-actin. Positive cells were observed predominantly on the ventral surface(V) of the aorta. A group of mesenchymal cells on ventral side, which were

not juxtaposed to the vessel wall, were also positive for SM a-actin(arrows). In addition, note SM a-actin positive cells in vasa vasorum(arrowhead).D: Nuclear staining with Hoechst 33258 of the same section.The arrows and arrowhead indicate the same localization shown in (C). E:A sagittal section of day 3 embryo immunolabeled with 1E12 MoAb andcounterstained with Hoechst 33258. F: Mouse embryo at 10.5 daypostcoitum was immunolabeled with SM a-actin MoAb and counter-stained with Hoechst 33258. D, Dorsal; L, lumen; V, ventral. Scale bar 550 µm.

after day 12 and later distributed in the cells of theinner layers of SMCs. This second gradient of theexpression of smooth muscle a-actin supports the ideathat initial population of the cells is replaced by cardiacneural crest cells, which differentiate into the tunicamedia (Le Lievre and Le Douarin, 1975). Therefore, thegradient of SMC differentiation in these vessels ap-pears to follow a different pattern. It would be ofinterest to determine the kinetics of proliferation ofneural crest and mesenchymal-derived SMC precur-sors within the media of vessels that have dual mesen-chymal contributions.It is assumed that blood vessels are formed gradually

by the recruitment and migration of mesenchymal cellsthat lay outside an endothelial layer. Therefore, it isreasonable to suggest that the endothelium might beused as a template for recruitment and migration ofpresumptive SMCs and perhaps might regulate thisprocess. Heparin-like molecules (Campbell and Camp-bell, 1986; Fritze et al., 1985; Castellot et al., 1982) andseveral growth factors such as PDGF, bFGF, and TGFbare secreted by the endothelium (Collins et al., 1987;Hannan et al., 1988; Sato et al., 1990). These and otherfactors may play a role in recruitment and migration of

undifferentiated SMC precursors in vivo. Our studiesshow that recruitment, proliferation, and differentia-tion of presumptive SMCs follow a ventral to dorsalgradient. If the endothelium plays a role in this process,interaction between aortic endothelium and SMCs maymaintain the balance of proliferation activity in theventral and dorsal surface of the vessel during earlyblood vessel formation. In fact, this has been shown tobe the case in interactions between endothelial cellsand pericytes (Saunders and D’Amore, 1992). Alterna-tively, one might speculate that the notochord mightinhibit recruitment or migration of mesenchymal cellsto the dorsal surface of the aorta. Proteins of thehedgehog family secreted by the notochord and neuraltube have been shown to regulate a variety of processesinvolved in embryonic patterning (Fan and Tessier-Lavigne, 1994; Leufer et al., 1994). If we consider SMa-actin and the 1E12 antigen as indicators of presump-tive SMC, these data suggest two possible scenarios forSMC recruitment to the condensing vessel wall; (1)factors responsible for SMC recruitment could be ex-pressed earlier or in a more localized fashion at theventral surface of the aorta, or (2) SMC might bepreferentially derived from a ventral pool of precursorcells.At this point, the mechanisms that regulate recruit-

ment of SMCs during vessel development are notknown. From our data on proliferation, it appears thatthe proliferation rate of mesenchymal cells recruited tothe vessel wall is not equivalent on the ventral anddorsal surfaces. The ventral to dorsal gradient of prolif-eration activity is transient; that is, once the vessel wallis built up by recruitment of mesenchymal cells, thematuration of vessel wall is maintained by equivalentreplication rates throughout aorta. Taken together theproliferation gradient and the differential expression ofSM a-actin and 1E12 suggest that recruitment, differen-tiation, and replication of SMCs to the vessel wallfollows a predetermined route. The findings open newquestions related to the mesodermic origin of SMC, i.e.,somato versus splanchnopleura, and the signals in-volved in this process. Further studies are required todetermine which signals regulate the ventral to dorsalgradient of recruitment and early differentiation ofSMC precursors to the aortic endothelium. A betterunderstanding of the pattern of this growth and differ-entiation might aid in identifying the molecular compo-nents involved in regulation ofmesenchymal cell recruit-ment during vascular development.

EXPERIMENTAL PROCEDURESImmunohistochemistry

Embryoswere harvested, classified according toHam-burger and Hamilton (1951), fixed in either 4% parafor-maldehyde or methyl-Carnoy’s for 1 hr, embedded inparaffin, and then sectioned at 5 µm. Sections werecleared twice in 100% xylene and rehydrated through aseries of ethanols (100%, 95%, 70%, and 50%). Sectionswere then washed with 13 PBS (pH 7.8) and blocked

Fig. 7. Rate of proliferation in the ventral and dorsal half of thedeveloping vessel wall. Cross sections of dorsal aortae at day 4, 6, 8, and10 were stained with 0.5% toluidine blue and hybridized with an antisensehistone 3 riboprobe. The numbers of total and proliferating cells at ventral(solid bar) and dorsal surface (hatched bar) were counted and expressedas percentage of proliferating cells. Counts in each category and develop-mental stage were assessed in five animals (See Results). (Statisticalanalysis of the ventral and dorsal compartments differ significantly fromrandom in the 4, 6, and 8 days of embryonic development and havecontingenct X2 with Yates’ correction 5 28.85: P ! 0.001 [for day 4],12.85: P ! 0.02 [for day 6] and 5.3: P , 0.05 [for day 8].)

350 LEE ET AL.

with 1% goat serum in PBS for 1 hr at room tempera-ture. Tissue sections were subsequently incubated withsmooth muscle a-actin MoAb (Sigma, St. Louis, MO;1:200 dilution in PBS), anti-smooth muscle calponinMoAb (Sigma; 1:5,000 dilution in PBS), polyclonalantibody to bovine aorta smooth muscle myosin (1:200dilution, generous gift from Dr. Robert Adelstein, NIH),or 1E12 MoAb (hybridoma supernatant), for 1 hr atroom temperature. After several washes in PBS, sec-tions were incubated with isotype specific biotinylatedsecondary antibodies (1:200 dilution; Vector Laborato-ries, Burlingame, CA); the 1E12 MoAb and calponinMoAb are IgMs, while a-actin and myosin antibodiesare IgG. After three washes in PBS, avidin-FITC (1:200dilution, Vector) was applied for 30 min. Finally, allsections were incubated with Hoechst 33258 (10 µg/mlin PBS; Molecular Probes, Eugene, OR) for 5 min toprovide nuclear counterstaining and mounted in 90%glycerol in 13 PBS, pH 8.0. Sections were viewed andphotographed on a Zeiss Axiophot (Carl Zeiss, Thorn-wood, NY).

In Situ Hybridization

Digoxigenin (Dig)-labelled riboprobes were gener-ated by the method of Schaeren-Wiemers and Gerfin-Mosen (1993). A 1.5 kb fragment of histone 3 cDNAwaslinearized with SalI and transcribed with T7 RNApolymerase to generate anti-sense probe. The senseprobe was generated with SP6 polymerase after diges-tion of histone 3 cDNAwithEcoRI. Embryos at differentstages of development were fixed in 4% paraformalde-hyde for 1 hr, embedded in paraffin, and then sectionedat 5 µm. The sections were cleared and rehydrated asdescribed previously. All sections were treated withproteinase K (20 mg/ml, Sigma) for 8 min at roomtemperature and washed twice with PBS in DEPCtreated ddH2O. After postfixation in 4% paraformalde-hyde for 1 hr, the sections were rinsed with 0.1 Mtriethanolamine (pH 8.0) and acetylated with 0.25%acetic anhydride for 15 min at room temperature.Slides were rinsed with 23 SSC and hybridized withDig-labelled riboprobes in 50% formamide, 43 SSC, 13Denhardt’s, 103 dextran sulfate, and 0.5 mg/ml salmonsperm DNA overnight at 50°C. After posthybridizationwashes, sections were blocked with 2% sheep serum inbuffer 1 (100 mM Tris-HCl, pH 7.5, and 150 mM NaCl)for 1 hr at room temperature and incubated withanti-Dig-alkaline phosphatase antibody (1:500 dilutionin buffer 1, BMB, IN) for 1 hr. Several washes withbuffer 2 (100 mM Tris-HCl, pH 9.5, 100 mM NaCl, and50 mM MgCl2) followed. Development of color wasachieved by incubation with 4-nitro blue tetrazoliumchloride (BMB) and 5-bromo-4-chloro-3-indolyl-phos-phate (BMB) as previously described (Shepherd et al.,1996).

Quantification of Proliferation

Staged sections of dorsal aorta were stained withtoluidine blue or hybridized with histone 3 riboprobe.

Total cell number was obtained by counting toluidineblue stained nuclei within the entire vessel wall or inthe vascular compartments (tunica intima, media, andadventitia). Percentage of proliferating cells was deter-mined after counting the number of cells positive forhistone 3 within each compartment. Histone 3 mRNAexpression occurs during S phase and has a shorthalf-life making it a good identifier of cycling cells. Thismarker is a reliable indicator of cell proliferation ascompared to BrdU and cyclin D (Gown et al., 1996). Sixstages of aortic development were evaluated; days 4, 6,8, 12, 16, and adult. Each data point represents themean 6 SD of at least five animals. The average fromeach animal was generated after analysis of threesections from the same specimen. Sections were 30–50µm apart to ensure appropriate sampling of proliferat-ing cells. Double counting was avoided by proceedingsystematically from left to right through the thicknessof the vessel wall. Data points were obtained by count-ing cells (n . 500) within a microscopic field. Twoinvestigators (SHL and MLIA) counted independentlyto confirm the data obtained in each time point. Statis-tical analysis were performed in the raw numbers aswell as in the percentage analysis (see respective figurelegends).

ACKNOWLEDGMENTS

We thank Sujata Guin for her technical assistance,members of the laboratory for helpful discussions, andDr. T.F. Lane for his critical reading of this manuscript.We thank Dr. R.S. Adelstein in (NIH) for providingsmooth muscle myosin polyclonal antibody. This studywas supported by a Grant-in-Aid from the AmericanHeart Association to M.L.I.-A. (96-1218).

REFERENCESAikawa, M., Sivam, P.N., Kuro-o, M., Kimura, K., Nakahara, K.,Takewaki, S., Ueda, M., Yamaguchi, H., Yazaki, Y., Periasamy, M.,and Nagai, R. (1993) Human smooth myosin heavy chain isoformsas molecular markers for vascular development and atherosclerosis.Circ. Res. 73:1000–1012.

Applegate, D., Feng, W., Green, R.S., and Taubman, M.B. (1994)Cloning and expression of a novel acidic calponin isoform from rataortic vascular smooth muscle. J. Biol. Chem. 269:10683–10690.

Belknap, J.K., Grieshaber, N.A., Schwartz, P.E., Orton, C., Reidy,M.A., and Majack, R.A. (1996) Tropoelastin gene expression inindividual vascular smooth muscle cells: Relationship to DNAsynthesis during vascular development and after arterial injury.Circ. Res. 78:388–394.

Campbell, G.R. and Campbell, J.H. (1985) Smooth muscle phenotypicchanges in arterial wall hemostasis: implications for the pathogen-esis of atherosclerosis. Exp. Mol. Pathol. 42:139–162.

Campbell, J.H. and Campbell, G.R. (1986) Endothelial cell influenceson vascular smooth muscle phenotype. Rev. Physiol. 48:295–306.

Castellot, J.J., Favreau, L.V., Karnovsky, M.J., and Rosenberg, R.D.(1982) Inhibition of vascular smooth muscle cell growth by endothe-lial cell-derived heparin. Possible role of a platelet endoglycosidase.J. Biol. Chem. 257:11256–11260.

Chamley-Campbell, J., Campbell, G.R., and Ross, R. (1979) Thesmooth muscle cell in culture. Physiol. Rev. 59:1–61.

Coffin, J.D. and Poole, T.J. (1988) Embryonic vascular development:immunohistochemical identification of the origin and subsequentmorphogenesis of the major vessel primordia in quail embryos.Development 102:735–748.

351PROLIFERATION AND DIFFERENTIATION OF THE VASCULAR WALL

Collins, T., Pober, J.S., Gimbrone, Jr., M.A., Hammacher,A., Betsholtz,C., Westermark, B., and Heldin, C.H. (1987) Cultured humanendothelial cells express platelet-derived growth factorA chain. Am.J. Pathol. 126:7–12.

De Mey, J.G., Uitendaal, M.P., Boonen, H.C., Vrijdag, M.J., Daemen,M.J., and Struyker-Boudier, H.A. (1989)Acute and long-term effectsof tissue culture on contractile reactivity in renal arteries of the rat.Circ. Res. 65:1125–1135.

Drew, J.S., Moos, C., and Murphy, R.A. (1991) Localization of isoactinsin isolated smooth muscle thin filaments by double gold immunola-beling. Am. J. Physiol. 260 (Cell Physiol. 29):C1332–C1340.

Duband, J.L., Belkin, A.M., Syfrig, J., Thiery, J.P., and Koteliansky,V.E. (1993) Calponin and SM 22 as differentiation markers ofsmooth muscle: Spatiotemporal distribution during avian embryodevelopment. Differentiation 55:1–11.

Fan, C.M. and Tessier-Lavigne, M. (1994) Patterning of mammaliansomites by surface ectoderm and notochord: Evidence for sclerotomeinduction by hedgehog homolog. Cell 79:1175–1186.

Fritze, L.M., Reilly, C.F., and Rosenberg, R.D. (1985) An antiprolifera-tive heparan sulfate species produced by postconfluent smoothmuscle cells. J. Cell Biol. 100:1041–1049.

Glukhova, M.A., Kabakov, A.E., Frid, M.G., Ornatsky, O.I., Belkin,A.M., Mukhin, D.N., Orekhov, A.N., Koteliansky, V.E., and Smirnov,V.N. (1988) Modulation of human aorta smooth muscle cell pheno-type: a study of muscle-specific variants of vinculin, caldesmon, andactin expression. Proc. Natl. Acad. Sci. USA85:9542–9546.

Gonzalez-Crussi, F. (1971) Vasculogenesis in the chick embryo. Anultrastructural study. Am. J. Anat. 130:441–460.

Gown, A.M., Jiang, J.J., Matles, H., Skelly, M., Goodpaster, T., Caso,L., Reshatof, M., Spaulding, D., and Coltrera, M.D. (1996) Valida-tion of the S-phase specificity of histone (H3) in situ hybridization innormal and malignant cells. J. Histochem. Cytochem. 44:221–226.

Hamburger, V. and Hamilton, H.L. (1951) A series of normal stages inthe development of the chick embryo. J. Morph. 88:49–92.

Hannan, R.L., Kourembanas, L.S., Flanders, K.C., Rogelj, S.J., Rob-erts, A.B., Faller, D.V., and Klagsbrun, M. (1988) Endothelial cellssynthesize basic fibroblast growth factor and transforming growthfactor beta. Growth Factors 1:7–17.

Hirakow, R. and Hiruma, T. (1981) Scanning electron microscopicstudy on the development of primitive blood vessels in chickembryos at the early somite stage. Anat. Embryol. 163:299–306.

Hood, L.C. and Rosenquist, T.H. (1992) Coronary artery developmentin the chick: Origin and deployment of smooth muscle cells, and theeffects of neural crest ablation. Anat. Res. 30:291–300.

Hungerford, J.E., Owens, G., Argraves, W.S., and Little, C.D. (1996)Development of aortic vessel wall as defined by vascular smoothmuscle and extracellular matrix markers. Dev. Biol. 178:375–392.

Hungerford, J.E., Hoeffler, J.P., Bowers, C.W., Dahm, L.M., Falchetto,R., Shabanowitz, J., Hunt, D.F., and Little, C.D. (1997) Identifica-tion of a novel marker for primordial smooth muscle and itsdifferential expression pattern in contractile versus noncontractilecells. J. Cell Biol. 137:925–937.

Kirby, M.L. and Waldo, K.L. (1990) Role of neural crest in congenitalheart disease. Circulation 82:332–340.

Kirby, M.L. and Waldo, K.L. (1995) Neural crest and cardiovascularpatterning. Circ. Res. 77:211–215.

Kuro-o, M., Nagai, R., Nakahara, K., Katoh, H., Tsai, R.C., Tsuchimo-chi, H., Yazaki, Y., Ohkubo, A., and Takaku, F. (1991) cDNA cloningof a myosin heavy chain isoform in embryonic smooth muscle and itsexpression during vascular development and in arteriosclerosis. J.Biol. Chem. 266:3768–3773.

Le Lievre, C.S. and Le Douarin, N.M. (1975) Mesenchymal derivativesof the neural crest: Analysis of chimeric quail and chick embryos. J.Embryol. Exp. Morphol. 34:125–154.

Leufer, E., Nelson, C.E., Johnson, R.L., Morgan, B.A., and Tabin, C.(1994) Sonic hedgehog and Fgf-4 act through a signalling cascadeand feedback loop to integrate growth and patterning of thedeveloping limb bud. Cell 79:993–1003.

Miano, J.M., Cserjesi, P., Ligon, K.L., Periasamy, M., and Olson, E.N.(1995) Smooth muscle myosin heavy chain exclusively marks the

smooth muscle lineage during mouse embryogenesis. Circ. Res.75:803–812.

Miano, J.M. and Olson, E.N. (1996) Expression of the smooth musclecell calponin gene marks the early cardiac and smooth muscle celllineages during mouse embryogenesis. J. Biol. Chem. 271:7095–7103.

Noden, D.M. (1993) Embryonic origins and assembly of blood vessels.Am. Rev. Respir. Dis. 140:1097–1103.

Owens, G.K., Loeb, A., Gordon, D., and Thompson, M.M. (1986)Expression of smooth muscle-specific alpha-isoactin in culturedvascular smooth muscle cells: relationship between growth andcytodifferentiation. J. Cell Biol. 102:343–352.

Owens, G.K. (1995) Regulation of differentiation of vascular smoothmuscle cells. Physiol. Rev. 75:487–517.

Poole, T.J. and Coffin, J.D. (1989) Vasculogesis and angiogenesis: Twodistinct morphogenetic mechanisms establish embryonic vascularpattern. J. Exp. Zool. 251:224–231.

Rosenquist, T.H. and Beall, A.C. (1990) Elastogenic cells in thedeveloping cardiovascular system: smooth muscle, nonmuscle, andcardiac neural crest. Ann. N.Y. Acad. Sci. 588:106–119.

Ross, R. (1993) The pathogenesis of atherosclerosis: A perspective for1990s. Nature 362:801–809.

Ruzicka, D.L. and Schwartz, R.J. (1988) Sequential activation ofa-actin genes during avian cardiogenesis: vascular smooth musclea-actin gene transcriptsmark the onset of cardiomyocyte differentia-tion. J. Cell Biol. 107:2575–2586.

Samaha, F.F., Ip, H.S., Morrisey, E.E., Seltzer, J., Tang, Z., Solway, J.,and Parmacek, M.S. (1996) Developmental pattern of expressionand genomic organization of the calponin-h1 gene: A contractilesmooth muscle cell marker. J. Biol. Chem. 271:395–403.

Sato, Y., Tsubol, R., Lyons, R., Moses, H., and Rifkin, D.B. (1990)Characterization of the activation of latent TGF-beta by co-culturesof endothelial cells and pericytes or smooth muscle cells: a self-regulating system. J. Cell Biol. 111:757–763.

Saunders, K.B. and D’Amore, P.A. (1992) An in vitro model for cell-cellinteractions. In Vitro Cellular Dev. Biol. 28A:521–528.

Sawtell, N.M. and Lessard, J.L. (1989) Cellular distribution of smoothmuscle actins during mammalian embryogenesis: expression of thea-vascular but not the a-enteric isoform in differentiating striatedmyocytes. J. Cell Biol. 109:2929–2937.

Schaeren-Wiemers, N. and Gerfin-Mosen,A. (1993)Asingle protocol todetect transcripts of various types and expression levels in neuraltissue and cultured cells: In situ hybridization using digoxigenin-labeled cRNAprobes. Histochemistry 100:431–440.

Schwartz, S.M., Campbell, G.R., and Campbell, J.H. (1986) Replica-tion of smooth muscle cells in vascular disease. Circ. Res. 58:427–444.

Schwartz, S.M. and Reidy, M.A. (1987) Common mechanisms ofproliferation of smooth muscle in atherosclerosis and hypertension.Hum. Pathol. 18:240–247.

Shepherd, I., Luo, Y., Raper, J.A., and Chang, S. (1996) The distribu-tion of collapsin-1 mRNA in the developing chick nervous system.Dev. Biol. 173:185–199.

Skalli, O., Ropraz, P., Trzeciak, A., Benzonana, G., Gillessen, D., andGabbiani, G. (1986)Amonoclonal antibody against a-smoothmuscleactin: a new probe for smooth muscle differentiation. J. Cell Biol.103:2787–2796.

Sugi, Y. and Lough, J. (1992) Onset of expression and regionaldeposition of a-smooth and sarcomeric actin during avian heartdevelopment. Dev. Dyn. 193:116–124.

Takahashi, K. and Nadal-Ginard, B. (1991) Molecular cloning andsequence analysis of smooth muscle calponin. J. Biol. Chem. 266:13284–13288.

Topouzis, S. and Majesky, M.W. (1996) Smooth muscle lineage diver-sity in the chick embryo: Two types of aortic smooth muscle celldiffer in growth and receptor-mediated transcriptional responses totransforming growth factor-b. Dev. Biol. 178:430–445.

Thyberg, J., Hedin, U., Sjolund, M., Palmberg, L., and Bottger, B.(1990) Regulation of differentiated properties and proliferation ofarterial smooth muscle cells. Arteriosclerosis 10:966–990.

352 LEE ET AL.