Contribution of the CervicalSympathetic Ganglia to the

Innervation of the Pharyngeal ArchArteries and the Heartin the Chick Embryo

MARLIES E. VERBERNE, ADRIANA C. GITTENBERGER-DE GROOT,LIESBETH VAN IPEREN, AND ROBERT E. POELMANN*

Department of Anatomy and Embryology, Leiden University Medical Center,2300 RC Leiden, The Netherlands

ABSTRACTIn the chick heart, sympathetic innervation is derived from the sympa-

thetic neural crest (trunk neural crest arising from somite level 10–20).Since the trunk neural crest gives rise to sympathetic ganglia of theircorresponding level, it suggests that the sympathetic neural crest developsinto cervical ganglia 4–14. We therefore tested the hypothesis that, inaddition to the first thoracic ganglia, the cervical ganglia might contribute tocardiac innervation as well.

Putative sympathetic nerve connections between the cervical gangliaand the heart were demonstrated using the differentiation markers tyrosinehydroxylase and HNK-1. In addition, heterospecific transplantation (quailto chick) of the cardiac and trunk neural crest was used to study the relationbetween the sympathetic neural crest and the cervical ganglia. Quail cellswere visualized using the quail nuclear antibody QCPN.

The results by immunohistochemical study show that the superior andthe middle cervical ganglia and possibly the carotid paraganglia contributeto the carotid nerve. This nerve subsequently joins the nodose ganglion ofthe vagal nerve via which it contributes to nerve fibers in cardiac vagalbranches entering the arterial and venous pole of the heart. In addition, thecarotid nerve contributes to nerve fibers connected to putative baro- andchemoreceptors in and near the wall of pharyngeal arch arteries suggestinga role of the superior and middle cervical ganglia and the paraganglia of thecarotid plexus in sensory afferent innervation. The lower cervical ganglia 13and 14 contribute predominantly to nerve branches entering the venous polevia the anterior cardinal veins. We did not observe a thoracic contribution.Heterospecific transplantation shows that the cervical ganglia 4–14 as wellas the carotid paraganglia are derived from the sympathetic neural crest.

Abbreviations: AA, aortic arch artery; AD, adrenal; AM, atrialmyocardium; Ao, aorta; AoO, aortic orifice; AVS, atrioventricularsulcus; CA, coronary artery; CCA, common carotid artery; CG,cardiac ganglia; 8CSG, 8th cervical SG; CST, cervical sympathetictrunk; 13CSG, 13th cervical SG; CN, carotid nerve; DA, dorsalaorta; GN, glossopharyngeal nerve; H, heart; ICA, internal ca-rotid artery; IVS, interventricular septum; JV, jugular vein; L,lung; LA, left atrium; LCA, left coronary artery; LICA, leftinternal carotid artery; LV, left ventricle; NG, nodose ganglion;NT, neural tube; PO, pulmonary orifice; PG, carotid paraganglion;RA, right atrium; RACV, right anterior cardinal vein; RCB, rightcarotid body; RCCA, right common carotid artery; RCSG, right13th cervical SG; RDAB, right dorsal arterial vagal branch;RDAS, right ductus arteriosus; RICA, right internal carotidartery; RJV, right jugular vein; RL, right lung; RNG, right

nodose ganglion; RPA, right pulmonary artery; RSCA, rightsubclavian artery; RV, right ventricle; S, aorticopulmonary sep-tum; SCG, superior cervical SG; SEP, subepicardium; SN, spinalnerve; SPG, spinal ganglion; T, transplant; T1, first thoracic SG;T7, seventh thoracic SG; VN, vagal nerve.

Grant sponsor: Netherlands Institute of Scientific Research;Grant number: 902–16–162.

*Correspondence to: Robert E. Poelmann, Department of Anatomyand Embryology, Leiden University Medical Center, PO Box 9602,2300 RC Leiden, The Netherlands.E-mail: Poelmann@RULLF2.LeidenUniv.NL

Received 17 November 1998; Accepted 27 April 1999

THE ANATOMICAL RECORD 255:407–419 (1999)

r 1999 WILEY-LISS, INC.

The cardiac neural crest does not contribute to the neurons of the cervicalganglia. We conclude that the cervical ganglia contribute to cardiac innerva-tion which explains the contribution of the sympathetic neural crest to theinnervation of the chick heart. Anat Rec 255:407–419, 1999.r 1999 Wiley-Liss, Inc.

Key words: neural crest; quail-chick chimera; cervical sympathetic gan-glia; chick embryo; heart innervation; tyrosine hydroxylase

It is reported that in both embryonic (Kirby et al., 1980)and adult chickens (Baumel, 1975) the sympathetic car-diac nerves project from the first pair of thoracic sympa-thetic ganglia (SG) and reach the heart at day 10–11 ofincubation (Higgins and Pappano, 1979; Kirby et al.,1980). These thoracic SG are part of the paired sympa-thetic trunks. Each trunk consists on average of 37 SG: 14cervical, 7 thoracic, 13 synsacral, and 3 caudal (Baumel,1975). These SG originate from the trunk neural crest cells(Le Douarin and Teillet, 1974), migrating from the neuraltube caudal from somite level 5 (Kirby and Bockman,1984).

The trunk neural crest from somite level 10–20 (alsonamed the sympathetic neural crest) is shown to beimportant for the innervation of the heart. Ablation of thisarea results in a significant depletion of the norepineph-rine uptake in the atrium, the absence of sympathetictrunks in the corresponding region and the absence ofsympathetic nerves in the atria (Kirby and Stewart, 1984).

The trunk neural crest gives rise to the SG of theircorresponding somite level (Kirby and Stewart, 1984; Yip,1983,1986) suggesting that the sympathetic neural crestdevelops into the cervical SG 4–14. We want to hypothesizethat, in addition to the first thoracic SG, the cervical SGmight contribute to the cardiac innervation as well. This,as well as reports of a cervical contribution to cardiacinnervation in several mammalian species (Randall et al.,1971; Armour and Randall, 1975; Janes et al., 1986), led usto investigate in detail the contribution of the cervical SGto cardiac innervation in the chick embryo. In addition, westudied the neural crest origin of the cervical SG to showtheir relationship with the sympathetic neural crest.

In the present study nerve connections between thecervical SG and the heart are studied by using an antibodyagainst tyrosine hydroxylase (TH), an enzyme involved incatecholamine synthesis and considered to be a marker forsympathetic nerve tissue. This is combined with heterospe-cific but homotopic transplantation (quail-chick chimera)of different levels of the neural crest, i.e. parts of either thecranial neural crest [this is the neural crest migratingcranial from somite level 6 as described by Kirby andBockman (1984)], or the trunk neural crest (caudal fromsomite 5). With these results we discuss the possiblefunction of TH-positive nerve fibers entering the heart andpharyngeal arch arteries (PAA), the specificity of TH-staining for sympathetic nerves and the homology ofcardiac innervation between different species.

MATERIALS AND METHODSChimera Technique

Fertilized eggs of the White Leghorn chick (Gallusdomesticus) and Japanese quail (Coturnix coturnix ja-

ponica) were incubated at 37°C for 41–48 hr and 37–46 hrrespectively, i.e. until the embryos reached stage HH10–13(Hamburger and Hamilton, 1951). The eggs were openedand the embryos stained lightly with nile blue sulphate.The dorsal part of the neural tube was removed at differentlevels using sharpened tungsten needles. The homotopicpart of a quail embryo at the comparable developmentalstage was excised and transplanted into the exposed site inthe chick embryo.

Transplantation of the cardiac neural crest, between thecentre of the otic vesicle (MO) and the third somite (S3),was performed at stage HH10. Likewise, different levels ofthe trunk neural crest were transplanted: between somitelevel 4–9 (S4–9) at stage HH11, S10–15 at stage HH12and S16–20 at stage HH13. Surviving embryos weresacrificed at stage HH281 (n 5 1), HH29 (n 5 1), HH291(n 5 1), HH35 (n 5 3), HH36 (n 5 4), HH37 (n 5 6),HH371 (n 5 1), HH38 (n 5 3) and HH39 (n 5 1). Theywere fixed in ethanol-acetic acid for 24–72 hr and embed-ded in paraffin.

Histology

Serial 5 µm sections of normal chick embryos (n 5 3)(fixed in ethanol-acetic acid for 24–72 hr and embedded inparaffin) and quail-chick chimeras (n 5 21) were alternat-ingly incubated with TH antibody (Boehringer Mannheim,Germany), a marker for sympathetic nerve tissue, andwith HNK-1 (American tissue type connection, see Aboand Balch, 1981), a marker for migrating neural crest andnerve tissue (Luider et al., 1993; Poelmann et al., 1994).TH (1:50) and HNK-1 (1:10) were diluted in phosphate-buffered saline (PBS) containing 0.05% Tween-20 (Merck,Schuchardt) and 1% Ovalbumine (Sigma). In addition,quail-chick chimera sections were incubated with QCPN(undiluted), a quail cell nuclear marker (Hybridoma Bank,Iowa City). The incubation with the first antibody lastedovernight. All sections were incubated for 1.5 hr in thesecond rabbit anti-mouse antibody conjugated to peroxi-dase and diluted in the same buffer (1:300). The sectionsincubated for TH were further incubated with a third goatanti-rabbit antibody (1:50; Nordic Immunology, Tilburg),followed by a fourth antibody, the rabbit peroxidase-antiperoxidase complex (1:500; Nordic Immunology, Til-burg) both for 1.5 hr in the same buffer as used for the firstand second antibodies. After treatment of all sections with0.04% diaminobenzidine tetrahydrochloride (DAB)/0.06‰H2O2 in 0.05 mol/L TRIS-maleic acid (pH 7.6) for 10 min atroom temperature, they were briefly counterstained withMayer’s haematoxylin, dehydrated via ethanol, trans-ferred to xylene and coverslipped using Entellan (Merck,

408 VERBERNE ET AL.

Darmstadt). In the quail-chick chimeras, neurons derivedfrom the transplanted neural crest can be identified whenthey are immunopositive both for the quail marker QCPNand the differentiation markers TH and HNK-1.

Normal embryos from stage HH37, HH40 and HH42were used as controls to check whether the HNK-1 andTH staining pattern found in the quail-chick chimerascan be considered as normal. No differences were obser-ved in HNK-1 and TH staining patterns between quail-chick chimeras and normal chick embryos of similarstages.

RESULTSTransverse sections of quail-chick chimeras (HH28139),

incubated with QCPN, were examined to study the originof the cervical SG. Table 1 summarizes the results of thequail-chick chimeras. The transplant often partially at-tached to the ablated region and/or showed an ingrowth inthe adjacent caudal non-ablated region (Table 1).

In general the highest number of QCPN-positive neuro-nal cells are situated in the SG at the corresponding levelsof the transplant. The number of quail-derived cells de-creases in SG at greater distance from the transplant (Fig.1). In some embryos the cranial spread could not bedetermined due to the removal of a larger part of the neckregion. On average, QCPN-positive neuronal cells areobserved in SG up to two somite levels cranial and threelevels caudal from the attached transplant. A much smallerlongitudinal spread of QCPN-positive neurons is presentin the spinal ganglia as compared to the SG (Table 1).

In addition, sections incubated with TH of both normalchick embryos (HH37–42) and the quail-chick chimeras(n 5 15, HH36–39) were examined for nerve connections

between the cervical SG and the heart. Figure 2 gives aschematic overview of the sympathetic trunk and its nerveconnections with the heart in a stage HH37 chick embryo.Three main levels in the cervical sympathetic trunk (supe-rior cervical, middle cervical and lower cervical SG) con-tain TH-positive nerve connections with the heart in thenormal embryos and quail-chick chimeras.

The Superior Cervical Sympathetic GangliaAt the level of the postotic hindbrain, the internal

carotid arteries are in close association with the superiorcervical SG (the most cranial part of the paired cervicalsympathetic trunks, Fig. 2), the cranial nerves IX and Xand the jugular vein (Fig. 3a–c). Tyrosine hydroxylaseantibody stains very specifically the SG and their nervebranches while other nerve branches, e.g. the vagal nerveand the glossopharyngeal nerve are TH-negative (Fig. 3b).In the trunk neural crest chimeras, caudal to S5, chimericSG contain both QCPN-positive neurons and satellite cells(Fig. 3d). In the cardiac neural crest chimeras (MO-S3),QCPN-staining is confined to the satellite and supportivecells of the superior cervical SG (Fig. 3e), while theneuronal somata are QCPN-negative. Each superiorcervical SG sends a branch, the carotid nerve, alongthe internal carotids in the neck, which is TH-positive(Fig. 2).

Cervical Sympathetic Ganglia 7–10At the level between the seventh and 10th cervical SG,

nerve connections are present between the cervical SG andthe carotid nerves (Fig. 2, 4a). These connections are lessprominent from stage HH40 onwards. The carotid nerves

TABLE 1. Neural crest origin of cervical sympathetic ganglia

Somite level(S) ablation Stage

Somite level(S) transplant

Somite level(S) QCPN

spinal ganglia

Somite level(S) QCPN

positive ganglia

Correspondingsympathetictrunk level

MO–S3 HH35 — — S6, 7 1MO–S3 HH35 — — S6, 7 1S4–9 HH281 S5–9 S6–8 S6–10 1–4Ca

S4–9 HH36 S6–13 S11 S6–14 1–8CS5–9 HH291 S6–8 S6–7 S6–8 1, 2CS7–9 HH35 S8–10 S8–10 S6–12 1–6CS10–12 HH37 S10–11 S9–11 S6–12 1–6CS10–12 HH36 S12–13 S11–14 S12–15 6–9CS10–14 HH371 S10–11 S10#–11 S10#b–15 4#–9CS10–14 HH38 S10–11 S13 S12#–19 6#–13CS10–14 HH37 S12–13 S9#–13 S9#–17 3#–12CS10–14 HH38 S12–14 S12#–13 S12#–15 6#–9CS10–14 HH38 S12–16 S11#–17 S11#–20 5#–14CS10–15 HH39 S13–14 — S9–18 3–12CS10–14 HH37 S15–16 S14 S14–17 8–11CS10–14 HH37 S16–17 S15–16 S12–19 6–13CS11–14 HH37 S15–17 — S6, 12, 13, 15–21 1, 6, 7, 9–14C, 1Tc

S12–13 HH36 S12–13 S11–14 S9–14 3–8CS12–13 HH36 S12–13 S13–15 S12–15 6–9CS13–14 HH37 S14–15 S13–16 S10–21 4–14C, 1TS16–20 HH29 S19–22 S19–21 S20–23 14C, 1–3T

aC, cervical sympathetic ganglion.b#, further cranial not determined.cT, thoracic sympathetic ganglion.

409CERVICAL SYMPATHETIC CARDIAC CONTRIBUTION

Fig. 1. a: Transverse section of an HH38 S12–14 quail-chick chimerastained with QCPN (340). Note that at the level of the transplant (T),located in the dorsal part of the neural tube (NT), the cervical SG (CSG) aswell as the spinal ganglia (SPG) are QCPN-positive. b–e: Transversesections of an HH38 S12-S16 quail-chick chimera stained with QCPN.b: Note that the 12th cervical SG (CSG) two levels caudal from the

transplant contains a large number of QCPN-positive cells (341).c: Magnification of the cervical SG in b (3205). d: At the level of the 14thcervical SG (CSG) four levels caudal from the transplant (341).e: Magnification of d (3210). Note that the 14th cervical SG (CSG)contains only a few QCPN-positive cells (arrows).

410 VERBERNE ET AL.

form a carotid plexus at these levels, which containsTH-positive paraganglia (Fig. 2, 4b). These paraganglia ofthe carotid plexus contain QCPN-positive cells in quail-chick chimeras (n 5 10) in which the transplant adheredsomewhere between the levels S10-S14 (Fig. 4c). Quail-

chick chimeras in which the transplant attached overlevels S15-S17 (n 5 3) show no QCPN-positive paraganglia(n 5 2) or these ganglia contain only a few QCPN-positivecells (n 5 1). The cervical SG 2–6 do not contain nerveconnections with the carotid nerves (Fig. 2).

Fig. 2. Schematic sagittal section of a chicken embryo of about HH37.It presents an overview of three levels of nerve connections (arrows)between the right cervical sympathetic trunk and the heart. The levels ofthe transverse sections of Figures 3, 4 and 5 are indicated. To the left thecorresponding somite levels (MO-S21) of origin are indicated. The mostcranial arrow points to the level where the superior cervical SG (SCG)

changes into the carotid nerve (CN). The middle arrow points at thenerve connections between the middle cervical SG (CSG) 7–10 and thecarotid nerve (CN). The most caudal arrow points at the connectionbetween the 13th cervical sympathetic ganglion (13CSG) and the jugularvein (JV).

411CERVICAL SYMPATHETIC CARDIAC CONTRIBUTION

Cervical Sympathetic Ganglia 13 and 14At the level of the 12th cervical sympathetic ganglia, the

carotid nerve joins the nodose ganglion of the vagal nerveas well as the recurrent nerve (Fig. 2, 5a). Near the nodoseganglia both the carotid bodies and the media of thecommon carotid arteries contain TH-positive nerve fibers

and cells (Fig. 5a–c). The 13th cervical SG (and also the12th, although less prominent) sends branches via theventral rami of the spinal nerve along the vertebral veintowards the jugular vein (Fig. 2, 5d). The 14th cervical SGsend nerve fiber branches via the ventral rami towards thesubclavian veins, joining the nerve branches along the

Fig. 3. Transverse alternate sections of the upper part of the neck ofan HH36 S6–13 quail-chick chimera stained with (a) HNK-1 (398), (b) TH(396) and (c) QCPN (3127). The superior cervical SG (SCG) in c isshown in higher magnification in d (3411). Note that the superior cervicalSG (SCG) is positive for TH (b) and QCPN (c,d) while the glossopharyn-geal (GN) and vagal (VN) nerves are negative for these markers (b,c) . All

nerves are positive for HNK-1 (a). Note that the superior cervical SG(SCG) contains QCPN-positive neurons (arrows) and satellite cells(arrowheads) (d). e: Transverse section of the superior cervical SG (SCG)(3485) in an HH39 MO-S3 (cardiac neural crest) quail-chick chimerastained with QCPN. Note that only the satellite cells (arrowheads) areQCPN-positive. The much larger neuronal somata are negative (arrows).

412 VERBERNE ET AL.

anterior cardinal veins (Fig. 2, 5e). The nerve branchesfrom the thoracic SG travel ventrolateral as well asventromedial from the lungs but do not reach the heart(Fig. 2). The ventromedial branches join the plexus aroundthe dorsal aorta and to some extent are in contact with thepulmonary plexus and the dorsal part of the plexus aroundthe oesophagus. The branches which move ventrolateralfrom the lungs bend into a lateral direction towards thebodywall (Fig. 5f).

Arterial and Venous PoleAt the level where the vagal nerve crosses the fourth

PAA, TH-positive nerve branches and cells are present inthe media of the dorsal part of the fourth PAA in eight of 12embryos (Fig. 6a,b). At the level of the sixth PAA, TH-positive nerve branches from the vagal and recurrentnerves join the arterial cardiac vagal branches whichtravel along the pulmonary arteries (sixth PAA) towardsthe arterial pole of the heart (Fig. 6c,d). At the arterial poleTH-positive nerve fibers and cells are located at the cranialpart of the aortico-pulmonary septum where the arterialvagal cardiac branches connect (Fig. 6e,f). In the thoraxventral to the lungs near the venous pole of the heart,nerve branches along the anterior cardinal veins are incontact with the TH-positive branches along the vagalnerves (Fig. 6g).

Intracardiac Nerves and Ganglia

At the arterial pole and along the ventral side of theheart between stage HH37–40, TH-positive nerve fibersare situated close to the aortic and pulmonary orifices. Atthe venous pole and dorsal side of the heart nerve fibersare present in the sinal cardiac vagal branch and branchesalong the anterior cardinal veins close to the orifices of thelarge veins near the sinus venosus. At stage HH42, at theventral side of the heart nerve fibers are located deeperinto the heart, i.e. in cardiac ganglia and nerve fiberbundles reaching the atrioventricular sulcus, especiallyaround coronary arteries (Fig. 7a). At the dorsal side of thestage HH42 heart, TH-positive nerve fibers are present innerve bundles and cardiac ganglia in the subepicardium ofthe atria unto the dorsal upper part of the ventricles andjust caudal from the sinus venosus (Fig. 7b). The ventricu-lar myocardium near the base of the heart including theupper part of the ventricular septum as well as the atrialmyocardium, including the ventral part of the interatrialseptum bordering the left ventricular outflow tract, con-tain a few nerve fibers (Fig. 7c,d). The neurons in thecardiac ganglia at both the dorsal and ventral side of theheart are TH-negative in all trunk neural crest chimerasas well as the normal embryos. The cardiac ganglia areQCPN-negative in the trunk neural crest chimeras (n 5

Fig. 4. a: Transverse section in the upper part of the neck of an HH37S15–17 quail-chick chimera (3132) showing a TH-positive connection(arrow) between the cervical sympathetic trunk (CST) just caudal from the8th cervical SG and the carotid nerves (CN) along the right (RICA) and left

(LICA) internal carotid arteries. b,c: Transverse alternate sections in theneck of an HH38 S12–16 quail-chick chimera (3220) showing a paragan-glion (PG) of the carotid plexus which is both positive for TH (b) andQCPN (c).

413CERVICAL SYMPATHETIC CARDIAC CONTRIBUTION

Fig. 5. a: Transverse section (3220) at the level of the right commoncarotid artery (RCCA) and the right nodose ganglion (RNG) in an HH37embryo stained with TH. Note the TH-positive branches (arrows) from thecarotid nerve and right carotid body (RCB) joining the nodose ganglion(RNG) of the vagal nerve. b,c: Transverse alternate sections of an HH38embryo at the level of the right common carotid artery (RCCA) stainedwith HNK-1 (b) and TH (c) (3205). Note that nerve fibers and eventuallycells (arrows) are positive for TH in the media of the common carotidartery (RCCA) as well as in the carotid body (RCB) (c). These fibers andcells are also positive for HNK-1 (arrows in b). d,e: Transverse sections ofan HH38 embryo stained with TH (344). Note a TH-positive nerve branch

(arrow in d) from the right 13th cervical SG (RCSG) running along thespinal nerve (SN) towards the right jugular vein (RJV) (d). e: Note aTH-positive branch (arrows) from the 14th cervical SG running towardsthe right anterior cardinal vein (RACV) at the level where the rightsubclavian artery (RSCA) joins the common carotid artery (RCCA).f: Transverse section of an HH37 embryo stained with HNK-1 (344). Notethe nerve branch from the first thoracic SG, traveling ventrolateral fromthe left lung (L) and bending in a lateral direction towards the bodywall(arrowheads). Arrows show the directions where the neural tube (NT) andheart (H) are located.

19), in contrast with the cardiac neural crest chimeras (n 52) which contain QCPN-positive cardiac ganglia.

DISCUSSIONTyrosine Hydroxylase Specificity

Earlier morphological studies used catecholamine histo-fluorescence to detect sympathetic cardiac nerves (Higginsand Pappano, 1979; Kirby et al., 1980), while TH activityhas only been measured quantitatively in the chick embry-

onic heart (Stewart and Kirby, 1985). From the presentstudy it appears that TH can be used to detect cardiacsympathetic nerves as well, since our results show thatTH-immunohistochemistry stains very specifically the sym-pathetic trunk and its branches. However, it is stilluncertain whether all TH-positive fibers near the embry-onic heart can be considered as sympathetic efferent nervefibers. Baumel (1975) has described in adult birds that thenerves around the internal carotids also contain branches

Fig. 6. a,b: Transverse alternate sections of the dorsal part of the 4thPAA of an HH37 embryo stained for HNK-1 (3106) (a) and TH (3121) (b).Note that some nerve fibers and/or cells (arrows) in the media of the 4thPAA (AA) are both positive for HNK-1 (a) and TH (b). c: Transversesection of an HH42 embryo stained with TH (341). Note TH-positivenerve fibers (arrows) in the right dorsal arterial vagal branch (RDAB)traveling along the right pulmonary artery (RPA) towards the arterial poleof the heart. d: Magnification of the area indicated in c (3103). Note that

only fibers are stained with TH. e,f: Transverse alternate sections of theaorticopulmonary septum (S) of an HH37 embryo stained with HNK-1(391) (e) and TH (383) (f). Note that a few HNK-1-positive fibers and/orcells (arrows in e) are also positive for TH (arrows in f). g: Transversesection at the level of the arterial pole of an HH42 embryo stained with TH(344). Note that nerve branches along the right anterior cardinal vein(RACV) are in contact with TH-positive nerve branches of the vagal nerve(VN) (see arrows).

415CERVICAL SYMPATHETIC CARDIAC CONTRIBUTION

from the glossopharyngeal nerve and fine variable ramifrom the vagus which contribute to the innervation ofthe parathryroid, thyroid, the ultimobranchial glandsand the carotid body. In the cat, sensory afferent branchesfrom the glossopharyngeal nerve which join the carotidnerve to the carotid body and the carotid sinus have beenshown to be TH-positive and are involved in relaying baro-and chemoreceptor information to the medulla (Massari etal., 1996). In this study TH-positive nerve fibers and cellsare present in the carotid body, in the media of the dorsalpart of the aortic arch artery, in the aorticopulmonaryseptum, and the media of the common carotid artery.These are probably putative chemo- and baroreceptors.The media of the mammalian aortic arch are known tocontain a similar structure as the carotid body (Ham andCormack, 1979), which might also apply to the chick. Thecatecholamine histofluorescence technique shows stainingof these receptor areas as Le Lievre and Le Douarin (1975)report the presence of fluorogenic monoamines-containingcells (using the formol-induced-fluorescence technique) inthe wall of the common carotid arteries. This shows thatcatecholamines and TH are located in the same areas.

Tyrosine hydroxylase staining in the upper part of theaorticopulmonary septum might also implicate chemorecep-tors as aortic bodies are described to be localized near theroots of the aorta and pulmonary trunk (Baumel, 1975).We conclude, therefore, that TH stains not only sympa-thetic postganglionic efferent nerve fibers but also chemo-and baroreceptor cells and their sensory afferent innerva-tion. Since the glossopharyngeal nerve and its petrosalganglion as well as the proximal part of the vagal nerve inthe neck did not show any TH-positivity, we assume thatTH-positive nerve fibers in the carotid nerve are derivedfrom the cervical SG and possibly the paraganglia of thecarotid plexus. This might implicate that in addition tosympathetic efferent neurons, the SG and paragangliacontain sensory afferent neurons as well. This means thatthe (para)sympathetic efferent and sensory afferent ner-vous system are not clearly separated in the chick. Untilnow, only the nodose ganglion of the vagal nerve is knownto contribute to sensory afferent cardiac innervation (Wak-ley and Bower, 1981). However, it can not be excluded thatsensory afferent neurons in the spinal ganglia or intrinsiccardiac sensory neurons on the heart itself might also have

Fig. 7. a,b,c,d: Transverse sections in the heart of an HH42 embryo.Note TH-positive nerve fibers (arrows) around the left coronary artery(LCA) in the atrioventricular sulcus (AVS) as depicted in a (3108). Panel b(3215) shows TH-positive nerve fibers in cardiac ganglia (CG) and nervebundles in the subepicardium (SEP) caudal from the sinus venosus. Notethat the neurons are TH-negative (arrows). Panel c (3197) shows

TH-positive nerve fibers (arrows) in the interventricular septum (IVS). d: Ahigher magnification of the indicated area in c (3440). Panel e (3207)shows TH-positive nerve fibers (arrows) ventrocaudal from the interatrialseptum near the atrioventricular node region. f: A higher magnification ofthe area indicated in e (3440).

416 VERBERNE ET AL.

a share in cardiac afferent innervation as has been shownin the cat (Oldfield and McLachlan, 1978) and has beensuggested for the guinea-pig heart (Steele et al., 1996)respectively.

Connections Between the Cervical SympatheticTrunk and the Heart

Our results show a contribution of the superior cervicalSG and the middle cervical SG 7–10 to the carotid nervewhich joins the nodose ganglion of the vagal nerve. Morecaudally, TH-positive branches of the vagal nerve enter thearterial and venous pole of the embryonic chick heart.Furthermore, the lower cervical SG contribute predomi-nantly to atrial innervation via the anterior cardinal veins,although some of these nerve fibers might also reach thearterial pole due to exchange with the vagal nerve. Untilnow only one study mentions the contribution of thesuperior cervical SG to heart innervation in the adultchick, without using any specific markers (His, 1893).Recent studies do not report a cervical sympathetic contri-bution. Although Baumel (1975) mentions in adult chicks‘‘collaterals’’ connecting the paraganglia interposed alongthe retrocarotid trunk (carotid nerve) with the paraverte-bral trunk (cervical SG), a connection between the carotidnerve and the heart is not described. The connectionsbetween the middle cervical SG and the carotid plexusmight serve as transient migratory pathways for neuro-blasts populating the carotid paraganglia, since they be-come less prominent from stage HH40 onwards. Baumel(1975) also described that these connections are reducedand difficult to display in adult chickens. This phenom-enon might be due to degeneration of some of theseTH-positive nerve fibers in combination with a furtherapparent dilution caused by an increasing number ofTH-negative sensory collaterals joining the remainingTH-positive nerve connections during later stages. Tran-sient differentiation has at least been described for thecervical visceral preganglionic nerve fibers at day 3 and 4which completely degenerate at day 5 (Levi-Montalcini,1950). Consistent with our findings is that Baumel (1975)describes bilateral sympathetic cardiac nerves travelingventrally at the cranial edge of the lungs joining thejugular and subclavian veins to continue their coursealong the cranial cardinal veins. However, both Baumel(1975) and Kirby et al. (1980) report that these sympa-thetic cardiac nerves arise from the first thoracic gangliawhile in our study these sympathetic cardiac nerves derivefrom the 14th cervical SG. In the present study, nervebranches, arising from the thoracic SG which run lateralfrom the lungs, change into a lateral direction towards thebody wall without joining the anterior cardinal veins asKirby et al. (1980) reported.

Studies in mammals, including humans, also mention acontribution of the middle and lower cervical SG to sympa-thetic innervation of the heart (Randall et al., 1971;Armour and Randall, 1975; Janes et al., 1986). However,these studies do not describe a connection between thesuperior cervical SG and the heart. In our study and thoseof Kirby et al. (1980) and Baumel (1975), the majorsympathetic cardiac nerves arising from the lower cervicalSG and/or first thoracic SG are homologous with thesympathetic cardiac nerves arising from the stellate SG inmammals (Randall et al., 1971; Armour and Randall,1975; Niehoff and Sullivan, 1980; Janes et al., 1986; Mo et

al., 1994; Wallis et al., 1996), the latter are shown toinfluence the heart when stimulated (Randall et al., 1971;Armour and Randall, 1975; Op’t Hof et al., 1991; Yano etal., 1992; Yamazaki et al., 1997). The stellate ganglion is afusion product of the inferior cervical ganglion and the first(and sometimes even the second) thoracic sympatheticganglion (Kanagasuntheram and Dharshini, 1994). Inchicks these ganglia do not fuse. In dogs stimulation of theright stellate ganglion predominantly influences the myo-cardium of the right atrium including the sinuatrial nodearea while stimulation of the left stellate ganglion affectsonly left atrial sites and AV junctional rhythm (Kralios andMillar, 1981). Since we describe that in the chick the rightand left TH-positive nerves arising from the lower cervicalSG also appear to enter predominantly the right and leftatrium via the anterior cardinal veins, the correlationbetween anatomical location and area of function mightalso hold for the chick.

Considering functionality, the appearance of the firstsympathetic axons in the myocardium at stage HH42 inour study is consistent with the time, day 16, at which thefirst release of endogenous catecholamines is measuredafter electrical stimulation (Higgins and Pappano, 1981).

Neural Crest Origin of the CervicalSympathetic Trunk

Our results show that SG receive their largest amount ofcells from the neural crest of the corresponding level.However, there appears to be a superimposed longitudinalmigration as SG cranial and/or caudal from the trans-planted level also contain QCPN-positive neurons, thenumber of which decreases at increasing distance from thetransplant level. Yip (1983, 1986) finds similar resultsafter transplantation of trunk neural crest from a singlesegmental level, but did not study the contribution of theneural crest originating cranial from the level of somite 12.In addition, in our study there is no significant longitudi-nal spread of quail sensory neurons in the spinal gangliawhich is consistent with the findings of Yip (1986) whodescribed that quail cells where only present in spinalganglia at the level of the transplant. Kirby and Stewart(1984) show that ablation of the sympathetic crest oversomite level 10–20 causes a significant decrease in norepi-nephrine uptake and absence of sympathetic nerves in theatria of the chick embryo. This suggests that the largestcontribution to sympathetic atrial nerves is derived fromthis area. According to our results and those of Yip (1983,1986) the sympathetic neural crest gives rise to thecervical sympathetic ganglia 4–14 which means that neu-rons located in this range contribute to atrial innervation.This is consistent with the findings that we observe amajor contribution from the lower cervical SG to nervesentering the venous pole via the anterior cardinal veins.While ablation of the neural crest between somite levels1–10 does not significantly influence the norepinephrineuptake in the atria (Kirby and Stewart, 1984), it suggeststhat the sympathetic contribution of the superior cervicalSG, derived from the neural crest at somite levels 6 and 7,to the atria is of minor importance. However, the norepi-nephrine uptake of the ventricles, on which these abla-tions might have influence, was not examined. As alreadydiscussed above, it is possible that the superior cervical SGare only involved in sensory afferent innervation of chemo-and baroreceptors situated in and near the PAA at the

417CERVICAL SYMPATHETIC CARDIAC CONTRIBUTION

arterial pole of the heart and not in sympathetic efferentinnervation in the heart itself. The presence of QCPN-positive cardiac ganglia in the quail-chick chimeras inwhich the cardiac neural crest was heterospecifically re-placed, is consistent with previous reports from our labora-tory (Poelmann et al., 1998; Verberne et al., 1998) andothers (Kirby and Stewart, 1983) showing that thesecardiac neurons are derived from the cardiac neural crest.Moreover, this study shows that the cardiac neural crestdoes not contribute to the neurons of the cervical SG exceptfor the satellite cells in the superior cervical SG. Further-more, the absence of TH-staining in cardiac ganglionicneurons in this study as well as the positivity for acetylcho-linesterase staining as shown by others (Rickenbacher andMuller, 1979) support the idea that these cardiac gangli-onic neurons are parasympathetic in nature.

In conclusion, we show, that the cervical SG significantlycontribute to cardiac innervation. We did not observe athoracic sympathetic contribution. The superior cervicalSG, the middle cervical SG 7–10 and the carotid paragan-glia contribute to TH-positive branches entering the ve-nous and the arterial pole of the heart via the carotidnerves and arterial and venous cardiac vagal branches. Inaddition, the results suggest that these ganglia are in-volved in sensory afferent innervation in and near thePAA. The lower cervical SG 13 and 14 mainly contribute tonerve branches along the jugular and anterior cardinalveins entering the venous pole. These cervical ganglionicnerve connections with the heart as well as the findingthat the carotid paraganglia and the cervical SG 4–14 arederived from the sympathetic neural crest explains thecontribution of the latter to atrial innervation. The nerveconnections between the cervical SG and the heart showgreat homology between avian and mammalian speciesexcept for the superior cervical ganglia.

The homology of cardiac innervation between chicks andvarious mammals will be useful when the cardiac nervoussystem of these animals is used as a model for studyingaltered heart function due to disturbed cardiac innerva-tion. These models might be used to explain cardiacarrhythmias during embryonic development or poor car-diac nerve regeneration after heart surgery and infarction.Disturbed development of the cardiac nervous system canbe artificially induced by treating chick embryos withretinoic acid (Broekhuizen et al., 1998) or by ablation of thecardiac or sympathetic neural crest (Kirby and Stewart,1983, 1984; Kuratani et al., 1991). Studies on nerveregeneration after cardiac surgery in humans reveal thatthe amount of regeneration is dependent on the age of thepatient at the time of operation (Kondo et al., 1998) andthe extent of cutting nerves (Bernardi et al., 1998). Properunderstanding of the quantitative distribution of parasym-pathetic and sympathetic nerve fibers in different parts ofthe heart are needed to explain the differences in regenera-tion between these two kinds of autonomic nerves causingautonomic imbalance.

ACKNOWLEDGMENTSWe want to thank Jan Lens and Bas Blankevoort for

photographic assistance.

LITERATURE CITEDAbo T, Balch CM. 1981. A differentiation antigen of human NK and K

cells identified by a monoclonal antibody HNK-1. J Immunol127:1024–1029.

Armour JA, Randall WC. 1975. Functional anatomy of canine cardiacnerves. Acta Anat 91:510–528

Baumel JJ. 1975. Aves nervous system. In Getty R, editor. Sisson andGrossman’s The anatomy of the domestic animals. Philadelphia:WB Saunders.

Bernardi L, Valenti C, Wdowczyck-Szulc J, Frey AW, Rinaldi M,Spadacini G, Passino C, Martinelli L, Vigano M, Finardi G. 1998.Influence of type of surgery on the occurrence of parasympatheticreinnervation after cardiac transplantation. Circulation 97:1368–1374.

Broekhuizen MLA, Gittenberger-de Groot AC, Baasten MJ, Wladimi-roff JW, Poelmann RE. 1998. Disturbed vagal nerve distribution inembryonic chick hearts after treatment with all-trans retinoic acid.Anat Embryol 197:391–397.

Ham AW, Cormack DH. 1979. Histology. Philadelphia: JB Lippincott; p609–610.

Hamburger V, Hamilton HL. 1951. A series of normal stages in thedevelopment of the chick embryo. J Morphol 88:49–92.

Higgins D, Pappano AJ. 1979. A histochemical study of the ontogeny ofcatecholamine-containing axons in the chick embryo heart. J MolCell Cardiol 11:661–668.

Higgins D, Pappano AJ. 1981. Development of transmitter secretorymechanisms by adrenergic neurons in the embryonic chick heartventricle. Dev Biol 87:148–162.

His W Jr. 1893. Die Entwickelung des Herznervensystems bei Wirbel-tieren. Abhandl Math-Phys Kl sachs Akad Wiss 18:1–64.

Janes RD, Brandys JC, Hopkins DA, Johnstone DE, Murphy DA,Armour JA. 1986. Anatomy of human extrinsic cardiac nerves andganglia. Am J Cardiol 57:299–309.

Kanagasuntheram R, Dharshini P. 1994. The cervical sympathetictrunk: a new hypothesis. Ann Acad Med Singapore 23:923–928.

Kirby ML, Bockman DE. 1984. Neural crest and normal development:a new perspective. Anat Rec 209:1–6.

Kirby ML, Stewart DE. 1983. Neural crest origin of cardiac ganglioncells in the chick embryo: identification and extirpation. Dev Biol97:433–443.

Kirby M, Stewart D. 1984. Adrenergic innervation of the developingchick heart: neural crest ablations to produce sympatheticallyaneural hearts. Am J Anat 171:295–305.

Kirby ML, McKenzie JW, Weidman TA. 1980. Developing innervationof the chick heart: a histofluorescence and light microscopic study ofsympathetic innervation. Anat Rec 196:333–340.

Kondo C, Nakazawa M, Momma K, Kusakabe K. 1998. Sympatheticdenervation and reinnervation after arterial switch operation forcomplete transposition. Circulation 97:2414–2419.

Kralios FA, Miller CK. 1981. Sympathetic neural effects on regionalatrial recovery properties and cardiac rhythm. Am J Physiol 240:H590–H596.

Kuratani SC, Miyagawa-Tomita S, Kirby ML. 1991. Development ofcranial nerves in the chick embryo with special reference to thealterations of cardiac branches after ablation of the cardiac neuralcrest. Anat Embryol 183:501–514.

Le Douarin NM, Teillet M-A M. 1974. Experimental analysis of themigration and differentiation of neuroblasts of the autonomic ner-vous system and of neurectodermal mesenchymal derivatives, usinga biological cell marking technique. Dev Biol 41:162–184.

Le Lievre CS, Le Douarin NM. 1975. Mesenchymal derivatives of theneural crest: analysis of chimaeric quail and chick embryos. JEmbryol Exp Morphol 34:125–154.

Levi-Montalcini R. 1950. The origin and development of the visceralsystem in the spinal cord of the chick embryo. J Morphol 86:253–282

Luider TM, Bravenboer N, Meijers C, Kamp AWM van der, Tibboel D,Poelmann RE. 1993. The distribution and characterization of HNK-1antigens in the developing avian heart. Anat Embryol 188:307–316.

Massari VJ, Shirahata M, Johnson TA, Gatti PJ. 1996. Carotid sinusnerve terminals which are tyrosine hydroxylase immunoreactiveare found in the commissural nucleus of the tractus solitarius. JNeurocytol 25:197–208.

Mo N, Wallis DI, Watson A. 1994. Properties of putative cardiac andnon-cardiac neurones in the rat stellate ganglion. J Auton Nerv Syst47:7–22.

418 VERBERNE ET AL.

Niehoff ML, Sullivan JM. 1980. Autoradiographic study of the rightstellate ganglionic fibers to the cat heart. J Auton Nerv Syst2:105–115.

Oldfield BJ, McLachlan EM. 1978. Localization of sensory neuronstraversing the stellate ganglion of the cat. J Comp Neurol 182:915–922.

Op’t Hof T, Misier ARR, Coronel R, Vermeulen JT, Verberne HJ, FrankRGJ, Moulijn AC, Capelle FJL van, Janse MJ. 1991. Dispersion ofrefractoriness in canine ventricular myocardium: effects of sympa-thetic stimulation. Circ Res 68:1204–1215.

Poelmann RE, Mentink MMT, Gittenberger-de Groot AC. 1994. Rostro-caudal polarity in the avian somite related to paraxial segmenta-tion: a study on HNK-1, tenascin and neurofilament expression.Anat Embryol 190:101–111.

Poelmann RE, Mikawa T, Gittenberger-de Groot AC. 1998. Neuralcrest cells in outflow tract septation of the embryonic chicken heart:differentiation and apoptosis. Dev Dyn 212:373–384.

Randall WC, Armour JA, Randall DC, Smith OA. 1971. Functionalanatomy of the cardiac nerves in the baboon. Anat Rec 170:183–198.

Rickenbacher J, Muller E. 1979. The development of cholinergicganglia in the chick embryo heart. Anat Embryol 155:253–258.

Steele PA, Gibbins IL, Morris JL. 1996. Projections of intrinsic cardiacneurons to different targets in the guinea-pig heart. J Auton NervSyst 56:191–200.

Stewart DE, Kirby ML. 1985. Endogenous tyrosine hydroxylaseactivity in the developing chick heart: a possible source of extraneu-ronal catecholamines. J Mol Cell Cardiol 17:389–398.

Verberne ME, Gittenberger-de Groot AC, Poelmann RE. 1998. Lineageand development of the parasympathetic nervous system of theembryonic chick heart. Anat Embryol 198:171–184.

Wakley GK, Bower AJ. 1981. The distal vagal ganglion of the hen(Gallus domesticus): a histological and physiological study. J Anat132:95–105.

Wallis D, Watson AHD, Mo N. 1996. Cardiac neurones of autonomicganglia. Microsc Res Tech 35:69–79.

Yamazaki T, Akiyama T, Kitagawa H, Takauchi Y, Kawada T, Su-nagawa K. 1997. A new, concise dialysis approach to assessment ofcardiac sympathetic nerve terminal abnormalities. Am J Physiol272:H1182–H1187.

Yano K, Hirata M, Mitsuoka T, Matsumoto Y, Hirata T, Hashiba K.1992. Interactive effects of verapamil and sympathetic nerve stimu-lation on the sinus and AV nodes in the canine hearts. Jpn Heart J33:83–93.

Yip JW. 1983. Formation of paravertebral sympathetic chain gangliain chick embryos. Neurosci Abstr 9:937.

Yip JW. 1986. Migratory pathways of sympathetic ganglioblasts andother neural crest derivatives in chick embryos. J Neurosci 6:3465–3473.

419CERVICAL SYMPATHETIC CARDIAC CONTRIBUTION