Migration of hypoglossal myoblast precursors

Migration of Hypoglossal Myoblast PrecursorsSARAH MACKENZIE, FRANK S. WALSH, AND ANTHONY GRAHAM*Department of Experimental Pathology, Kings College London, London, United Kingdom

ABSTRACT The intrinsic hypoglossal mus-culature develops from precursor myoblastswhich undergo long-range migration from theoccipital somites to the tongue. Little detail isknown about the precise spatiotemporal path-way taken by these cells or the factors controllingmigration. In this study, chick/quail chimeras inwhich the occipital paraxial mesoderm is quailderived, reveal that the pathway taken by thetongue muscle progenitors is both complex andhighly specific. Precursor myoblasts are Pax-3positive cells which descend from the somite andmigrate around the pharyngeal endoderm. Theythen course rostrally, following the base of thepharynx, remaining in a tight strand. We haveexamined a number of factors implicated in thecontrol of migration of the hypoglossal precur-sors. Replacement of the occipital somites withthose originating in the flank reveals that intrin-sic differences do not exist between these somiteswith respect to their capacity to respond to migra-tory cues. The lack of high level HGF/SF expres-sion along the pathway of the migrating hypoglos-sal precursors suggests that this factor is notinvolved in the actual process of migration of thehypoglossal precursors to the tongue. The path-way followed by the migrating precursors is iden-tical to that of both the developing hypoglossalnerve and the circumpharyngeal crest—a sub-population of the cranial neural crest, and impor-tantly these populations utilize this pathway be-fore the myoblast precursors. However, ablationneither of the hypoglossal nerve nor of the neuralcrest results in a perturbation in the ability ofthis Pax-3 positive population to migrate. Thisdemonstrates that migration of the precursors isindependent of both of these cell populations,and that it is controlled by the peripheral tissues.Dev. Dyn. 1998;213:349–358. r 1998 Wiley-Liss, Inc.

Key words: hypoglossal; precursor; migration;myoblast

INTRODUCTION

In vertebrates, all skeletal muscles derive from theparaxial mesoderm. Some muscle groups develop insitu, such as the epaxial muscles of the trunk, whichform the deep muscles of the back. In contrast, severalmuscles derive from cells that migrate over long dis-tances as undifferentiated precursor myoblasts beforeundergoing myogenesis and fusing. Hypaxial muscles

including those of the limb, the ventral body wall, thetongue and the diaphragm fall into this category. It iswith the migration of the hypoglossal musculature thatthis study is concerned.

Early studies showed that whilst the skeletal andconnective tissues of the limbs and the ventral wall ofthe body develop from the lateral plate mesoderm, themuscles in these regions form from cells which origi-nate in the adjacent somites. In the case of the limbmuscles, this was elegantly demonstrated by the analy-sis of chick/quail chimeras in which paraxial mesodermadjacent to either the fore or the hind limbs was of quailorigin (Chevallier et al., 1977; Christ et al., 1977). Inthese chimeras, the muscles of the limbs were derivedexclusively from quail tissue. Half somite transplantsfurther defined that the migrating cells derive from thelateral half of the somite (Ordahl and Le Douarin,1992). Similarly, chick/quail transplants were used toshow that the ventral abdominal muscles form from theventral bud of the dermamyotome which migrates intothe splanchnopleure of the lateral plate mesoderm(Christ et al., 1983).

An equivalent situation occurs in the formation of thehypoglossal musculature. Although the tongue devel-ops in the base of the pharynx from multiple origins(Kontges and Lumsden, 1996), the intrinsic hypoglossalmusculature originates from the occipital region(somites 2–5) (Hazelton, 1970; Noden, 1983) and mi-grates to the developing tongue. Furthermore, thisterritory, which lies immediately posterior to the oticvesicle, adjacent to rhombomeres 7 and 8, also contrib-utes neural crest cells, the circumpharyngeal crest, andthe motor innervation, the hypoglossal nerve, of thehypoglossal musculature (Shigetani et al., 1995; Kura-tani et al., 1988).

Although previous studies identified the sites oforigin of the hypoglossal musculature, they gave noinformation regarding the dynamics of myoblast migra-tion or the factors acting to control the migratorypathway. In this study we have used chick/quail chime-ras, which allow intact embryos to be examined in threedimensions, to establish a clear picture of the timing ofemigration of cells from the occipital somites as well as

Grant sponsor: Dunhill Medical Trust.Frank S. Walsh is presently at Department of Neuroscience Re-

search, SmithKline Beecham, Harlow, Essex, United Kingdom.*Correspondence to: Dr. Anthony Graham, Department of Experi-

mental Pathology, Kings College London, Guy’s Campus, LondonBridge, London SE1 9RT, U.K. E-mail: [email protected]

Received 8 July 1998; Accepted 12 August 1998

DEVELOPMENTAL DYNAMICS 213:349–358 (1998)

r 1998 WILEY-LISS, INC.

the precise pathway that they take. We have alsocarried out a series of embryological manipulations toassess the contribution of several different factors tothe control of the migration of the hypoglossal precur-sors.

We find that, even though occipital somites have beenshown to be intrinsically different from somites at otheraxial levels with regards their ability to establishsensory ganglia (Lim et al., 1987), there are no differ-ences with respect to the responsiveness of myoblasts tothe hypoglossal migratory cues. Recent mutationalanalysis of the c-met receptor tyrosine kinase and itsligand HGF/SF resulted in defects in the hypoglossalmusculature (Bladt et al., 1995), suggesting that thissignaling system may play a role in the development ofthese cells. However, we show here that HGF/SF is notexpressed along the migratory pathway during theperiod of myoblast migration, and as such it probablydoes not play a role in guiding these migratory cells. Wehave further analyzed the roles of the two other deriva-tives of the occipital region which play a role in tonguedevelopment—the hypoglossal nerve and the occipitalcrest. Both of these display developmental profileswhich would be consistent with their acting to controlmyoblast guidance, but we find that ablation of thesepopulations has no effect on myoblast migration. Rather,the myoblast precursors seem to use cues in the periph-ery to guide them along their pathway from the somitesto the hypopharyngeal region.

RESULTSMigratory Pathway of the HypoglossalPrecursors

Previous studies examining the development of thehypoglossal musculature pinpointed the occipitalsomites as the site of origin of these cells. However, dueto technical constraints, these works did not giveinformation regarding the timing of the onset of myo-blast migration or the precise pathway of these cells. Toaddress these points we have also used chick/quailchimeras but in this study the quail cells were detectedusing the QCPN antibody in wholemount immunohisto-chemistry. Occipital somites of quail origin were trans-planted into a chick host. This allowed us to establishthe precise timing and pathway of the migrating hypo-glossal precursors. The migrating quail cells follow anextremely narrow and circuitous pathway. Quail cellsleaving the ventrolateral edge of the somite can beidentified at stage 17 (Fig. 1A; n 5 3) with the cellsinitially migrating ventrally along one of two shorttracts (Fig. 1B). These join to form a common and quitenarrow stream of cells, sandwiched between the pharyn-geal endoderm, of the sixth (and last) pharyngeal arch,and the peripheral tissues of the trunk. By stage 20, themajority of cells have migrated to the ventral edge ofthe last pharyngeal arch where they curve in ananterior direction and move along the base of thepharyngeal arches (Fig. 1C, arrow; n 5 5). At this stage

cells can be seen along the pathway and can be identi-fied as forming a continuous stream from the somite(Fig. 1C, arrowhead). The pathway that the cells aretaking is well defined, and the migrating cells aretightly confined to a very precise and narrow tract (Fig.1D). In stage 26 chimeras, transverse sections at thelevel of the transplanted somite show that the cellsdescend from the somite and follow the endoderm of thepharynx (Fig. 1E, arrow; n 5 5). In more anteriorsections, cells are seen at the base of the pharyngealarches confined to a roughly circular tract (Fig. 1F),reflecting the restricted pathway seen in the whole-mount embryos.

Pax-3 Expression in the Occipital Region

It is only recently that genes expressed specifically inmigrating hypaxial muscle precursors have been identi-fied. The first of these was the paired-box/paired-homeobox containing transcription factor Pax-3 whichwas identified in the migrating myoblast precursors atthe limb level (Williams and Ordahl, 1994). However, inthe case of the migrating hypoglossal precursors, geneexpression has not been closely examined, so a study ofPax-3 expression in the occipital region was under-taken. The occipital somites show a similar Pax-3expression pattern to other somites. Expression isinitially throughout the somites and later becomesrestricted to the lateral and medial domains of thedermamyotome. Pax-3 positive cells can be seen leavingthe occipital somites until stage 17, which is consistentwith the timing of the onset of migration revealed bythe chick/quail chimeras (Fig. 2A). Furthermore, thePax-3 cells also form two short tracts which join anddescend ventrally (Fig. 2B). By stage 20, Pax-3 positivecells are found in a pattern identical to that identifiedin the chick/quail chimeras. They can be seen to curveanteriorly and migrate under the pharyngeal arches,and are still present in a trail connecting the somites tothe leading front where the majority of cells are locatedunder the pharyngeal arches (Fig. 2C, arrowhead).Furthermore, the Pax-3 expressing cells also show atightly defined pattern of expression with a clearlydemarcated dorsal boundary as the cells curve at thebase of the sixth arch.

In order to confirm that the expression of Pax-3colocalizes with the migratory cells, stage 20 chick/quail chimeric embryos were double-stained for Pax-3and the anti-quail antibody QCPN, and then sectioned.Transverse sections show that after homotopic graftingof occipital somites quail cells found at the base of thepharyngeal arches are also Pax-3 positive (Fig. 2D; n 53). The labeled cells are not spread evenly throughoutthe entire area of Pax-3 expression, most probablyreflecting the fact that only one quail somite wastransplanted whilst the contribution to the hypoglossalmusculature comes from four somites (somites 2–6).Similarly, the Pax-3 band of expression seen under the

350 MACKENZIE ET AL.

pharyngeal arches at stage 20 (Fig. 2C) is slightlybroader than the migrating trail seen with a singlesomite transplant alone (see Fig. 1D). However, the

patterns clearly overlap, demonstrating that Pax-3 is abone fide marker for migrating precursors of the tonguemuscles.

Fig. 1. Migration pathway of the hypoglossal muscle precursors inchick/quail chimeras. Occipital somites 4 or 5 of stage 8–9 chick embryoswere replaced with the equivalent quail somite and migrating cells werevisualized with the anti-quail antibody QCPN. A: The bisected head of astage 17 chimeric embryo showing the labeled somite and the first cells(arrow) beginning to leave the somite. B: A higher magnification of thesame embryo showing two tracts of cells as they exit the somite (arrows)which subsequently join to form a single tract just ventral to the somite. C:Bisected head of a stage 20 embryo showing cells migrating along the

base of the pharyngeal arches in an anterior direction (arrow) as well atrail of cells linking them to the occipital somites (arrowhead). D: A highermagnification of the same embryo at the base of the sixth pharyngealarch. E: Transverse section through a stage 26 chimera showing cellsdescending from the somite and migrating along the pharyngeal endo-derm (arrow). F: A more anterior section showing that the migrating cellsare located in a roughly circular tract (arrows) at the base of the pharynx.Anterior is to the right in A–D. ba, pharyngeal arch; o, otic vesicle; p,pharynx; s, somite. Scale bars, 150 µm.

351HYPOGLOSSAL MYOBLAST PRECURSOR MIGRATION

Somitic Contribution to the HypoglossalMusculature

The descriptive analysis of the migrating cells, show-ing the complicated and yet extremely precise pathwaythat they follow, provoked an examination of potentialfactors that could control this process. One potentialfactor that could be involved is that the occipitalsomites are distinct from those found at other axiallevels. It has been shown previously the absence ofdorsal root ganglia at this axial level is due to anintrinsic property of occipital somites (Lim et al., 1987).To test whether the capacity to contribute to themigrating cell population is unique to the occipitalsomites, somites at this level from chick host embryoswere replaced with quail somites from the trunk region.Transplanted flank somites contributed to the migrat-ing population of hypoglossal precursors in a fashionalmost identical to that seen with control somites (n 5

4). Cells from transplanted somites migrate away fromthe somite following the same pathway as controlsomites, moving initially in a ventral direction thencurving in an anterior direction under the pharyngealarches (Fig. 3). Both the pathway and the timing ofmigration corresponded closely with the control grafts,although the contribution was not as consistently ‘‘ro-bust’’ as that seen with homotopic grafts. This demon-strates that the migratory cues are directed by theenvironment through which the cells traverse.

Scatter Factor/Hepatocyte Growth Factor

In light of the reported defects in the hypoglossalmusculature of both HGF/SF and c-met mutant mice(Bladt et al., 1995), the expression of HGF/SF duringthe early migration of hypoglossal precursors in theoccipital region was examined by in situ hybridization.Somewhat unexpectedly, the occipital region of the

Fig. 2. Pax-3 expression in the occipital region. Wholemount in situhybridization against Pax-3 was performed on chick embryos of differentstages. In some cases the in situ protocol was followed by wholemountanti-quail antibody (QCPN) treatment, and embryos were viewed aswholemounts or vibratome sectioned. A, B: Pax-3 expression in theoccipital region of a bisected stage 17 embryo, showing cells descendingventrally from the somite, initially forming two tracts (arrows) that then joinshortly after leaving the somite. C: Pax-3 expression in the occipital region

of a bisected stage 20 embryo showing the curved pathway of themigrating cells (arrow) underneath the pharyngeal arches as well as a trailof cells maintaining contact with the somite (arrowhead). D: Transversesection of a double-labeled stage 20 chick/quail chimera showing thecolocalization of Pax-3 positive cells with the migrating quail cells (arrow).Anterior is to the right in wholemounts. ba, pharyngeal arch; p, pharynx.Scale bars, 150 µm.


chick embryo remains conspicuously free of detectableHGF/SF, at the initiation of migration at stage 17/18(Fig. 4A) and remains absent during precursor myo-blast migration, as seen at stage 20 (Fig. 4B, arrow).This applies to both the region adjacent to the occipitalsomites as well as to the ventral regions of the pharynxthrough which the hypoglossal cord migrates.

Expression of HGF/SF is apparent at stage 18 in thehead region; however, it is not located along the migra-tory pathway, but rather in the first pharyngeal arch,the most distant arch from the occipital somites (Fig.4C). Transverse sections show that expression is con-fined to the mesodermal core of the arch which containsthe precursors for the muscles that will form variouscomponents of the facial musculature. Later, expres-sion is also seen in the second and third pharyngealarches (Fig. 4D), once again with expression restrictedto the mesenchymal core of each arch.

These results also suggested that one role for HGF/SFin the developing pharyngeal region is in muscle differ-entiation. Consequently, we conducted a study of MyoDexpression in this region to determine how closely theexpression of HGF/SF correlates with the induction ofmuscle formation. The MyoD pattern of expression isalmost identical to that of HGF/SF. However, there is aslight time delay, such that MyoD is first expressed inthe mesoderm core of the most rostral arch of stage 20embryos, two stages after the onset of HGF/SF expres-sion (Fig. 4E). Similarly, MyoD is only expressed in themesenchymal core of the remaining pharyngeal archesafter the initial expression of HGF/SF (Fig. 4F).

The Hypoglossal Nerve

The pathway taken by the myoblast precursors is thesame as that taken by the axons of the motor neurons ofthe hypolglossal, or XIIth cranial nerve, that willsubsequently innervate them (Fig 5A). Moreover, thehypoglossal nerve rootlets first begin to emerge, anduse this pathway, at stage 16 (Kuratani et al., 1988;data not shown), which is before the myoblast precur-sors have started to leave the lateral edge of thedermamyotome of the occipital somites. These factsobviously suggests that the XIIth nerve could be actingas a guidance cue for the precursor myoblasts. Todirectly test this role of the hypoglossal nerve, asymmet-ric ablations of the neural tube of stage 10–12 chickembryos were performed. Such a manipulation allowsone side of the neural tube and its associated nerves todevelop normally thus serving as a control, whilst thecontralateral side develops in the absence of the neuraltube and hence any nerves that would arise from it. Theablated embryos were incubated for a further 2 or 3days until they had reached approximately stage 21(n 5 5) or 26 (n 5 7). They were subsequently assessedsimultaneously for migration, by wholemount in situhybridization against Pax-3 expression, and for thesuccess of the nerve ablation, by use of an anti-neurofilament antibody. In all cases, Pax-3 positivemyoblast migration was apparent with no obviousdisruption to the extent, timing or pathway of migra-tion (compare Fig. 5A and B, arrows). Neurofilamentstaining verified that most operated embryos showed acomplete removal of the hypoglossal nerve (Fig. 5B,arrowhead). Sections through ablated embryos allow adirect comparison of Pax-3 expression in ablated andcontrol sides, highlighting that in the complete absenceof the nerve, no obvious effect on the presence or thelevel of Pax-3 expression is seen (Fig. 5C).

The Circumpharyngeal Neural Crest

While it is clear that the hypoglossal motor neuronsare not guiding the myoblast precursors, there is an-other embryonic cell population which has been previ-ously suggested as an attractive candidate for fulfillingthis role—the circumpharyngeal neural crest (Kura-tani et al., 1991). As with the motor neurons of the XIIthcranial nerve, this subpopulation of the migrating

Fig. 3. Contribution of flank somites to the tongue muscle precursors.Occipital somites 4 or 5 of stage 8–9 chicken embryos were replaced withepithelial flank somites from stage 13 quail donors and visualized usingthe QCPN antibody. A: Bisected head of a stage 20 control chimericembryo which had received a homotopic graft of a quail occipital somite,showing migrating cells (arrow) moving underneath the pharyngealarches. B: The bisected head of a stage 20 chimeric embryo in which anoccipital somite was replaced with a quail flank somite, showing cellsmigrating along an identical pathway. ba, pharyngeal arch; s, somite.Anterior is to the right. Scale bars, 150 µm.


cranial neural crest follow the same pathway, and againemerge into the periphery, at stage 11, prior to themyoblasts (Kuratani and Kirby, 1991; Lumsden et al.,1991). In order to determine whether the migration ofthe precursor myoblast population depends upon themigration of the circumpharyngeal crest we ablatedthis population. Even though it has been demonstratedthat crest from this axial level fails to regenerate afterneural tube ablation (Suzuki and Kirby, 1997), wefurther ensured the absence of crest at this axial levelby ablating whole neural tubes from rhombomere 4through to the seventh or eighth somite at stages prior

to neural crest emigration (Lumsden et al., 1991).Wholemount in situ hybridization with Pax-3 wasperformed on surviving embryos to identify the migra-tory population. In all cases (n 5 4), Pax-3 positive cellscan be seen streaming away from the lateral edge of thesomite (Fig. 5D, arrow) with no apparent effect on theextent, direction or timing of migration. As with theasymmetrical neural tube ablations, some regenerationof the neural tube was apparent. However, Pax-3 alsoacts as a marker for the efficiency of the neural tubeablation since it is expressed in the dorsal region of theneural tube. This confirmed that in each case the

Fig. 4. HGF/SF expression in the occipital region. Wholemount in situhybridization was performed on chick embryos at a variety of stages andembryos were viewed as bisected wholemounts. A: At stage 18 when thefirst precursors are starting to leave the somite, there is an absence ofHGF/SF expression in the region adjacent to the occipital somites (arrow).B: In stage 20 embryos, HGF/SF expression is still absent along themigratory pathway (arrow). C: The bisected head of a stage 18 embryo

showing HGF/SF expression (arrow) in the first pharyngeal arch. D: Astage 20 embryo showing that by now the embryo is expressing HGF/SFstrongly in the first three pharyngeal arches (arrows). E: MyoD expressionin the first pharyngeal arch of a stage 20 embryo (arrow). F: By stage 27strong expression of MyoD can be seen in the mesenchymal core of allpharyngeal arches (arrows) as well as in the myotomes of the somites. ba,pharyngeal arch. Anterior is to the right. Scale bars, 150 µm.


neural tube had failed to regenerate as far as theoccipital region (Fig. 5D, arrowheads). It is also interest-ing to note with these embryos, that removal of theentire neural tube at such an early stage in develop-ment has no apparent effect on pharyngeal arch mor-phology or somitic development. Pax-3 expression inthe somites adjacent to the ablation also appearsnormal. The important conclusion that can be drawnfrom these experiments is that the migratory pathwayof the hypoglossal myoblast precursors is controlled bythe peripheral tissues.

DISCUSSION

In this study, we have defined the onset and pathwayof migration of the hypoglossal myoblast precursors.We have further analyzed a number of factors thatcould control myoblast migration. We find that somitesfrom axial levels, other than the occipital, can producemigrating myoblast precursors and that these will usethe same pathway as the endogenous cells. We also findthat the expression pattern of HGF/SF in the pharyn-geal region is not consistent with it playing a role incontrolling myoblast migration but rather that this

molecule is more likely to be involved in muscle differen-tiation. We have further demonstrated that even thoughboth the hypoglossal nerve and the neural crest use thismigratory pathway before the myoblast precursors,ablation of either of these populations does not affectmyoblast migration. These results suggest that themigration of the hypoglossal myoblast precursors iscontrolled by peripheral cues.

Myoblast Migration

Replacement of the occipital somites with their quailequivalents facilitated an in depth analysis of theprocess of cell migration from the occipital somites tothe tongue. This analysis was complemented by that ofPax-3 expression, which we have shown to be anexcellent marker of the migrating myoblasts. Thesecells begin migrating from the lateral edge of theoccipital somites at stage 17, but unlike the precursorsat the limb level, they do not disperse throughout theadjacent lateral plate mesoderm. Instead they appearto join one of two tracts which coalesce almost immedi-ately as they leave the lateral edge of the dermamyo-tome, then travel as a coherent stream moving ven-

Fig. 5. Ablation of the hypoglossal nerve and the neural crest.Hypoglossal nerves were ablated by the asymmetric removal of neuraltubes from stage 10–12 embryos which were then double-stained bywholemount in situ hybridization against Pax-3 (purple) followed by ananti-neurofilament antibody (RMO; brown) and were either vibratomesectioned or viewed as bisected wholemounts (A–C). Neural crest wasablated in 8–10 somite embryos by removal of the entire neural tube (D).A: Control side of the bisected head of a stage 20 operated embryoshowing the colocalization of the migrating Pax-3 positive cells (arrow)and the hypoglossal nerve (arrowhead). B: Contralateral operated side ofthe same embryo showing the absence of the hypoglossal nerve (arrow-

head) although the vagus nerve (X) has still formed normally. Pax-3positive cells are still seen migrating in the absence of the hypoglossalnerve (arrow). C: Transverse section through an ablated embryo showingthe absence of the nerve on the operated side (compare arrowheads)whilst Pax-3 expression is present on both sides (arrows). D: Thebisected head of a stage 20 embryo in which the neural tube was ablatedat the 8 somite stage. The arrow shows the migrating stream of Pax-3positive cells moving under the pharyngeal arches, arrowheads show theextent of the neural tube ablation as determined by Pax-3 expression. Inwholemount embryos, anterior is to the right. Scale bars, 150 µm, exceptin D, where it is 300 µm.


trally, following the line of the circumpharyngeal ridge,curving around the posterior limit of the sixth pharyn-geal arch. On reaching the base of this arch, they makean arc-shaped turn in an anterior direction, fanning outslightly and begin to travel along the base of the archesat the ventral aspect of the pharynx. Cells migrate asfar as the base of the second arch, still as a tight strandof cells and along a very well-defined and apparentlyquite complex pathway.

The Production of Hypoglossal MyoblastPrecursors Is Not Intrinsic to Occipital Somites

It is possible that the production of migratory myo-blasts destined for the hypoglossal is intrinsic to somitesat the occipital level. Indeed, several differences areknown to exist between occipital somites and thosefound in other regions. First, it has been suggested thata property specific to the sclerotome of the occipitalsomites is responsible for the absence of normal dorsalroot ganglia in the occipital region (Lim et al., 1987).Another potential difference between these somites isrevealed by the pattern of migration of the neural crest.Occipital crest displays a pattern of migration distinctfrom that seen in the trunk region. Early migratingneural crest cells in the trunk leave the neural tube andfollow a ventral pathway through the anterior half ofthe somite. However, upon encountering the occipitalsomites, early migrating crest cells do not initiallymove through the somites. Instead they remain at thedorsal edge of the somites until the first somite disap-pears, and then migrate around the anterior edge of thesecond somite before descending ventrally along thecircumpharyngeal ridge to form the vagus nerve (Kura-tani and Kirby, 1991). In addition, the pattern of geneexpression in these somites often differs from theremainder of the somites along the anteroposterioraxis. This is seen in the expression of the paired boxtranscription factor Pax-1, which is only down-regu-lated in the occipital somites (Wilting et al., 1995). Yetanother distinction can be made between the occipitaland trunk somites in that the fate of the sclerotomalcells deriving from the occipital somites is distinct fromthat of the trunk somites. The occipital sclerotomecontributes to several bones at the base of the skull(Couly et al., 1993), whilst in the trunk, sclerotomegives rise to the vertebrae. However, despite the sugges-tion that differences exist between the occipital somitesand those of the flank, replacement of the occipitalsomites with those of the flank revealed that any suchdifferences do not extend to the capacity for thesesomites to contribute cells to the epaxial musculature.This clearly demonstrates that precursor hypoglossalmyoblast migration is controlled by cues within theenvironment.

The Role of HGF/SF

The phenotypic abnormalities presented in c-met andHGF/SF mutant mice (Bladt et al., 1995) suggestedthat this signaling system could be generally involved

in regulating the migration of myogenic precursorscells. While this may apply to the developing limb, anumber of points argue strongly against HGF/SF’splaying an active role in the process hypoglossal myo-blast precursor migration. While HGF/SF is expressedat the start of myoblast migration, expression is re-stricted to the first pharyngeal arch, which is distantfrom the site of origin of the myoblast precursors.Indeed, the absence of HGF/SF throughout the path-way during the early period of migration of the hypoglos-sal precursors of the chick suggests that HGF/SF is notinvolved in controlling migration along this pathwayper se. Support for this conclusion also comes from theanalysis of mutant mice. While the c-met and theHGF/SF knockout animals do display defects in thehypoglossal musculature, these only relate to numberof muscle cells in the developing tongue, not to theirabsence (Bladt et al., 1995). The fact that there areintrinsic hypoglossal myotubes in the knockout ani-mals demonstrates that HGF/SF is not required toguide the migration of these myoblast precursors. How-ever, in these animals it was noticed that there was areduction in the number of myotubes in the hypolglos-sal musculature, and taken with our results, it is likelythat HGF/SF is directly involved in muscle differentia-tion in the developing tongue. Further evidence support-ing such a role comes from work in which it was shownthat forced expression of this molecule can cause theformation of ectopic muscle in the central nervoussystem (Takayama et al., 1996).

Roles of the Hypoglossal Nerveand the Circumpharyngeal Crest

The very well-defined route that the precursors fol-low suggests that this migratory process is tightlyregulated. Two other derivatives of the occipital regionwhich also contribute to the developing tongue werealso assessed as to their role in guiding myoblastmigration, namely the hypoglossal nerve (Kuratani etal., 1988) and the circumpharyngeal crest (Kurataniand Kirby, 1992). Both the hypoglossal nerve and thecircumpharyngeal crest take the same pathway as themigrating myoblasts and importantly both embarkupon it before the emergence of the myoblasts from theoccipital somites. However, ablation of the hypoglossalnerve failed to make any identifiable impact on theextent, timing or pathway of the migrating cells, sug-gesting that precursor migration is independent ofmigration of this nerve. Similarly, ablation of themigrating neural crest population, prior to precursoremigration, produced no alterations to the migration ofthe Pax-3 positive cell population. These results showthat the myoblast precursors, and probably also hypo-glossal motor axons and circumpharyngeal crest, areguided by peripheral cues.

The results in the current study show that migrationof the hypoglossal precursors is independent of theaxial tissue, a fact which has also been shown to applyto the migration of the limb muscle precursors (Rong et


al., 1992). An examination of migrating hypoglossalmyoblast precursors shows that after leaving the occipi-tal somites these cells appear to change their directionon encountering the pharyngeal endoderm. They thentrack along the circumpharyngeal ridge, which liesbetween the lateral mesoderm of the trunk and thecaudal end of the pharynx. On reaching the base of thearches these cells then migrate rostrally. Given thecomplexity of the migratory pathway of these myoblastprecursors, the peripheral cues that are involved arelikely to be multiple, and more complex than thosefound at limb levels.

The occipital region is a particularly complex area ofthe body with both somites and pharyngeal archesco-existing at this axial level. However, these twosegmental systems arise independently of each otherduring development, and tend not to intermingle. Thisfact is reflected in the migration pattern taken by thehypoglossal myoblasts. In all vertebrates analyzed,including humans chickens, mice, fish and lampreys(Kuratani, 1997; Kuratani et al., 1997; O’Rahilly andMuller, 1984; Romer, 1972), it has been shown that thehypoglossal/hypobranchial nerve fibers, and the musclecells where studied, do not take the most direct routetowards their final destination. Rather, they seem toavoid the pharyngeal arches, skirt round their caudaledge, and only project rostrally on reaching the base ofthe pharynx. This migration pattern would be consis-tent with the guidance of the precursors to their finallocation as a result of inhibitory cues from this region,possibly emanating from the pharyngeal endoderm,and interestingly it has been suggested that somiticderivatives cannot pass through a pharyngeal arch(Kuratani, 1997). These migratory events provide alink between somitic and pharyngeal regions, and arein keeping with studies which have shown these twoseparate systems of segmentation both predate theexistence of the hypoglossal/hypobranchial apparatus,and suggest that they may be linked with the evolutionof this structure.

EXPERIMENTAL PROCEDURESSomite Transplantation

Fertile Rhode Island Red hens’ eggs (Needle Farm,Enfield) and quail eggs (Rosedean Farm, Cam-bridgeshire, U.K.) were incubated at 38°C in a humidi-fied atmosphere. Hamburger-Hamilton (HH) stage 8–9or stage 13 donor quail embryos were treated withdispase I (1 mg/ml, Boehringer) for 2–10 min dependingon the stage of the embryo and the epithelial occipitalsomites, 1 through 4, were separated from the adjacenttissues using flame-sharpened tungsten needles andcollected in L15 medium containing 1% fetal calf serum(Gibco, Paisley, UK). Individual, recently formed occipi-tal somites, up to somite 4, were removed from theright-hand side of HH stage 8–9 host chicken embryoswith a tungsten needle and replaced with donor quailsomites. Eggs were re-incubated for 1–3 days.

Neural Tube Ablations

Asymmetric ablation of the hypoglossal nerve wasperformed using stage 10–12 chick embryos. Neuraltube from the level of the otic vesicle to the eighthsomite was removed on only one side of the embryousing a tungsten needle. For neural crest ablation, theentire neural tube was removed from the level of theotic vesicle to approximately half way down the unseg-mented paraxial mesoderm of HH stage 8–10 embryos.Embryos were then incubated for a further 2–3 daysand collected for analysis.

Wholemount In Situ Hybridization

In situ hybridization was performed using a modifica-tion of the protocol described by Henrique et al. (1995).The probes used in this study were as previouslydescribed—the HGF probe is that of Thery et al. (1995)and the Pax-3 probe is that of Goulding et al. (1993).Embryos were washed twice in PBT (PBS/0.1% Tween),dehydrated in 50% then 100% methanol and rehy-drated through a methanol series (100%, 75%, 50% and25%), washed twice in PBT and transferred to PBT:hybridization buffer (1:1), then transferred to freshhybridization buffer containing 1 µg/ml DIG-labeledriboprobe. Hybridization was performed overnight at70°C, after which embryos were washed twice at 70°Cin hybridization buffer, then once at 70°C in hybridiza-tion buffer:MABT (1:1). This was followed by twowashes in MABT and then embryos were blocked inMABT containing 2% Boehringer Blocking Reagentand 20% goat serum for 2 hr before being incubatedovernight in alkaline phosphatase conjugated anti-DIGantibody (Boehringer Mannheim). Embryos werewashed extensively in MABT then the DIG-labeledprobes were visualized by reacting the embryos with4-nitroblue-tetrazolium chloride and 5-bromo-4-chloro-3-indolyl-phosphate (Boehringer) in NTMT at roomtemperature for 1 hr to overnight, until the desiredlevel of color was achieved. Embryos were post-fixed in4% paraformaldehyde and stored in 0.4% paraformalde-hyde. Whole embryos were dissected and mountedunder coverslips in 90% glycerol/PBS and photo-graphed using 64T film (Kodak) on a Zeiss Axiophotmicroscope. Embryos for sectioning were embedded inalbumin-gelatine (0.45% gelatine, 25% albumin, 20%sucrose), fixed with 2.5% glutaraldehyde and vibratomesectioned at 40–50 µm.

Wholemount Immunohistochemistry

Embryos that had been treated for wholemount insitu hybridization were fixed in fresh 4% paraformalde-hyde for 1 hr at room temperature. Endogenous peroxi-dases were inactivated with 0.05% hydrogen peroxide.Quail nuclei were detected using the quail-specificantibody QCPN (1:100; Developmental Studies Hybrid-oma Bank, University of Iowa) for 4 days at 4°C, andneurofilament was detected using an anti-NF-M anti-body (1:10,000, clone RMO-270, Zymed) for 3 days at


4°C. Both primary antibodies were detected with anHRP-conjugated goat anti-mouse secondary (1:100;Dako, P447) after overnight incubation at 4°C. Primaryantibodies were visualized using diaminobenzidine (0.5g/L in 0.1 M Tris-HCl, pH 7.4). Whole embryos weredissected and mounted under coverslips in 90% glycerol/PBS and photographed using 64T film (Kodak) on aZeiss Axiophot microscope. Embryos for sectioning wereembedded in gelatine-albumin (0.45% gelatine, 25%albumin, 20% sucrose), fixed with 2.5% glutaraldehydeand vibratome sectioned at 40–50 µm.

ACKNOWLEDGMENTS

We thank Dr. M. Goulding (Salk Institute, USA) forthe gift of the Pax-3 probe, and Prof. Claudio Stern forthe HGF probe. The QCPN antibody developed byB. & J. Carlson was obtained from the developmentalStudies hybridoma bank maintained by the Universityof Iowa, Department of Biological Sciences, Iowa City,IA 52242, under contract NO1-HD-7–3263 from theNICHD. We would also like to thank members of thelab, Ian Mckay and Tom Shilling for valuable commentson this manuscript.

REFERENCESBladt F, Riethmacher D, Isenmann S, Aguzzi A, Birchmeier C.

Essential role for the c-met receptor in the migration of myogenicprecursor cells into the limb bud. Nature 1995;376:768–771.

Chevallier A, Kieny M, Mauger A. Limb-somite relationship: Origin ofthe limb musculature. J. Embryol. Exp. Morphol. 1977;41:245–258.

Christ B, Jacob HJ, Jacob M. Experimental analysis of the origin of thewing musculature in avian embryos. Anat. Embryol. 1977;150:171–186.

Christ B, Jacob M, Jacob HJ. On the origin and development of theventrolateral abdominal muscles in the avian embryo.An experimen-tal and ultrastructural study. Anat. Embryol. 1983;166:87–101.

Couly GF, Coltey PM, Le Douarin NM. The triple origin of skull inhigher vertebrates: A study in quail-chick chimeras. Development1993;117:409–429.

Goulding MD, Lumsden A, Gruss P. Signals from the notochord andfloorplate regulate the region-specific expression of two Pax genes inthe developing spinal cord. Development 1993;117:1001–1016.

Hazelton RD. A radioautographic analysis of the migration and fate ofcells derived from the occipital somites in the chick embryo withspecific reference to the development of the hypoglossal muscula-ture. J. Embryol. Exp. Morphol. 1970;24:455–466.

Henrique D, Adam J, Myat A, Chitnis A, Lewis J, Ish-Horowicz D.Expression of a Delta homologue in the chick. Nature 1995;375:787–790.

Kontges G, Lumsden A. Rhombencephalic neural crest segmentationis preserved throughout craniofacial ontogeny. Development 1996;122:3229–3242.

Kuratani S. Spatial distribution of postotic crest cells defines thehead/trunk interface of the vertebrate body: Embryological interpre-

tation of peripheral nerve morphology and evolution of the verte-brate head. Anat. Embryol. 1997;195:1–13.

Kuratani SC, Kirby ML. Initial migration and distribution of thecardiac neural crest in the avian embryo: An introduction to theconcept of the circumpharyngeal crest. Am. J. Anat. 1991;191:215–227.

Kuratani SC, Kirby ML. Migration and distribution of circumpharyn-geal crest cells in the chick embryo. Formation of the circumpharyn-geal ridge and E/C81 crest cells in the vertebrate head region. Anat.Rec. 1992;234:263–280.

Kuratani S, Tanaka S, Ishikawa Y, Zukeran C. Early development ofthe hypoglossal nerve in the chick embryo as observed by thewhole-mount nerve staining method. Am. J. Anat. 1988;182:155–168.

Kuratani SC, Miyagawa-Tomita S, Kirby ML. Development of cranialnerves in the chick embryo with special reference to the alterationsof cardiac branches after ablation of the cardiac neural crest. Anat.Embryol. 1991;183:501–514.

Kuratani S, Ueki T, Aizawa S, Hirano S. Peripheral development ofcranial nerves in a cyclostome, Lampetra japonica: Morphologicaldistribution of nerve branches and the vertebrate body plan. J.Comp. Neurol. 1997;384:483–500.

Lim TM, Lunn ER, Keynes RJ, Stern CD. The differing effects ofoccipital and trunk somites on neural development in the chickembryo. Development 1987;100:525–533.

Lumden A, Sprawson N, Graham A. Segmental origin and migration ofneural crest cells in the hindbrain region of the chick embryo.Development 1991;113:1281–1291.

Noden DM. The embryonic origins of avian cephalic and cervicalmuscles and associated connective tissues. Am. J. Anat. 1983;168:257–276.

O’Rahilly R, Muller F. The early development of the hypoglossal nerveand occipital somites in staged human embryos. Am. J. Anat.1984;169:237–257.

Ordahl CP, Le Douarin N. Two myogenic lineages within the develop-ing somite. Development 1992;114:339–353.

Romer AS. The vertebrate as a dual animal—somatic and visceral.Evol. Biol. 1972;6:121–156.

Rong PM, Teillet MA, Ziller C, Le Douarin NM. The neural tube/notochord complex is necessary for vertebral but not limb and bodywall striated muscle differentiation. Development 1992;115:657–672.

Shigetani Y, Aizawa S, Kuratani S. Overlapping origin of pharyngealarch crest cells on the postotic hind-brain. Dev. Growth Differ.1995;37:733–746.

Suzuki HR, Kirby ML. Absence of neural crest cell regeneration fromthe postotic neural tube. Dev. Biol. 1997;184:222–233.

Takayama H, LaRochelle WJ, Anver M, Bockman DE, Merlino G.Scatter factor/hepatocyte growth factor as a regulator of skeletalmuscle and neural crest development. Proc. Natl. Acad. Sci. USA1996; 93:5866–5871.

Thery C, Sharpe MJ, Batley SJ, Stern C, Gheradi E. Expression ofHGF/SF, HGF1/MSP, and c-met suggest new functions during earlychick development. Dev. Genet. 1995;17:90–101

Williams BA, Ordahl CP. Pax-3 expression in segmental mesodermmarks early stages in myogenic cell specification. Development1994;120:785–796.

Wilting J, Ebensperger C, Muller TS, Koseki H, Wallin J, Christ B.Pax-1 in the development of the cervico-occipital transitional zone.Anat. Embryol. 1995;192:221–227.


Migration of hypoglossal myoblast precursors

Documents

Transcript of Migration of hypoglossal myoblast precursors