Pioneer neurons in the mouse trigeminal sensory system

Transcription

1 Proc. Nadl. Acad. Sci. USA Vol. 87, pp , February 1990 Neurobiology Pioneer neurons in the mouse trigeminal sensory system (mesencephalic trigeminal nucleus/monoclonal antibody/specific neuronal antigen/axon outgrowth) DIDIER Y. R. STAINIER* AND WALTER GILBERTt Departments of tcellular and Developmental Biology and *Biochemistry and Molecular Biology, Harvard University, Cambridge, MA Contributed by Walter Gilbert, November 14, 1989 ABSTRACT Pioneer neurons establish preliminary nerve pathways that are followed by later-growing axons. The existence of pioneers and their importance is well documented in invertebrate systems. In mammals, early neuronal development has generally been difficult to study because of the size and complexity of the embryos, and the lack of adequate markers. Here we look at the time of earliest axonal outgrowth in the mouse embryo by using specific monoclonal antibodies to stain wholemount preparations. During the period of formation and closure of the neuropore beginning at embryonic day 8.5, we can follow the earliest trigeminal sensory neurons extending axons along stereotyped pathways. In the trigeminal ganglion, an early wave of neurogenesis gives rise to a small number of neurons whose axons pioneer the different trigeminal tracts in the periphery. After a brief pause (12 hr), these primary axons branch out to innervate individual targets. Emerging a day later, secondary fibers extend along the pioneers. By contrast, in the central nervous system, neurons of the mesencephalic trigeminal nucleus extend toward the rhombencephalon independently, ignoring preexisting fibers. These results show the existence of an early set of axonal tracts in the mouse peripheral nervous system that may be used for the guidance of later-differentiating neurons. Pioneer neurons establish nerve pathways when distances are short and tissues can be traversed (1-3). In insect embryos, they are essential for guiding later-differentiating neurons toward their targets (2, 3). Very recently, McConnell et al. (4) showed that the subplate neurons in the developing mammalian cortex invade the thalamus early in fetal life, thereby providing a possible pathway for later-born cortical neurons. A pioneer-like behavior has also been attributed to the earliest neurons of the chicken peripheral trigeminal system by the immunohistochemical studies of Moody et al. (5) and by the retrograde tracing studies of Covell and Noden (6). However, these latter observations (5, 6) were limited by the use of cryostat sections to analyze projection patterns. Non-serial thin-section microscopy does not reveal the threedimensional relationship of a putative pioneer with other axons. In this study, we describe the pioneering property of some of the earliest differentiating neurons in the mouse. We use specific monoclonal antibodies (mabs) to stain embryonic wholemounts and visualize the full extent of axonal outgrowth from a defined set of neurons. The wholemount preparations are optically dissected with the confocal microscope and the resulting images are then combined to generate a three-dimensional representation of the specimen. In the trigeminal ganglion, an early wave of neuronal differentiation gives rise to a small number of neurons, presumably derived from the epidermal placodes as has been shown in the chicken embryo (5, 6). These neurons pioneer the different trigeminal tracts in the periphery. (The term "pioneer" is The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C solely to indicate this fact. used merely to suggest that these are the first detected axons in these pathways and that secondary or trailing fibers appear to grow along them. No claim is made at this point about the requirement of these axons for the guidance of subsequent fibers.) Later-differentiating neurons, derived mostly from the neural crest (5, 6), seem to project along the preexisting tracts. By contrast, in the central nervous system (CNS), neurons of the, mesencephalic trigeminal nucleus (MesV) differentiate at a steady rate and project caudally in an independent manner; fasciculation does not take place. We show the developmental sequence of early axonal outgrowth in the mouse trigeminal system, focusing on the ophthalmic projection, and report that the early-born ganglionic neurons pioneer the peripheral tracts. MATERIALS AND METHODS mab Production and Characterization. mab E1.9 was isolated from a fusion in which the immunogen consisted of whole cells from embryonic day 12 (E12) rat mid- and hindbrain. It recognizes a cytoplasmic epitope in the primary sensory and motor neurons during axonal outgrowth, appearing at E8.5 and disappearing by E12 in the mouse. More specifically, it stains neural crest-derived sensory neurons (e.g., neurons of dorsal root ganglia) as well as placodederived sensory neurons (e.g., neurons of cranial nerves VII and VIII). Its full description will be published elsewhere. Inmmunohistochemistry. Pregnant CD1 mice were obtained from Charles River Breeding Laboratories under a specific breeding schedule: mice were bred for 3 hr from noon to 3 p.m. on EO; our embryonic day ran from noon to noon. Animals were sacrificed by cervical dislocation or euthanized with halothane. Embryos staged according to Theiler (7) were slit longitudinally along the forming neuropore and processed as half-embryo wholemounts. Immunostaining was done as previously described (8). All the reactions were done at room temperature (with gentle shaking) in polystyrene culture dishes by sequential incubations in the following: 10% normal goat serum (NGS) in phosphate-buffered saline (PBS) for 1 hr; hybridoma culture supernatant overnight (12-18 hr); after three 10-min washes with 10% NGS/PBS, fluoresceinated goat anti-mouse IgM (Southern Biotechnology Associates, Birmingham, AL), diluted 1:100, for 2 hr. After three more 10-min washes in 10% NGS/PBS, the wholemounts were mounted between two coverslips separated by silicone grease, in fresh mounting medium. For E1.9 staining, the embryos were fixed with 1% paraformaldehyde for 10 min, rinsed several times with PBS, and incubated for 2 hr in 10% NGS/PBS. Saponin (0.04%) was included throughout the E1.9 immunostaining. We used a Lasersharp MRC-500 confocal microscope mounted with fluorescence optics to observe specific axonal pathways. Serial sections were collected and combined by projection. The resulting images Abbreviations: mab, monoclonal antibody; CNS, central nervous system; MesV, mesencephalic trigeminal nucleus; En, embryonic day n. 923

2 924 Neurobiology: Stainier and Gilbert were stored in a WORM optical disk (Maxtor, San Jose, CA) and printed using a video printer. The data in this paper come from analyzing about 120 embryos, some as early as the 5-somite stage (E8) and others as late as E12. RESULTS We used two specific mabs to describe the existence and behavior of pioneer sensory neurons in the mouse trigeminal system. mab E1.9 recognizes a cytoplasmic epitope expressed in growing primary sensory and motor axons from E8.5 to E12 (unpublished data). mab E1.9 immunoreactivity first appears in the central nervous system at E8.5 and in the trigeminal ganglion at E9; it disappears at E12. mab B30 recognizes a rare ganglioside expressed on trigeminal sensory neurons shortly after differentiation (8). Carefully staged embryos (7) were dissected, treated immunohistochemically as wholemounts, and examined with a Lasersharp MRC-500 confocal microscope equipped with fluorescence optics. During E9 (day of vaginal plug is E0), at the 14-somite stage, a few cells in the region of the prinntive trigeminal ganglion are E1.9-positive (Fig. la). The filamentous appearance of 1.9 immunoreactivity does not yet define neuronal cell bodies, but short axonal projections span the width of the b Proc. Natl. Acad. Sci. USA 87 (1990) ganglionic space and fibers are stretched out, orienting toward the periphery. The initially low neuronal density stays constant throughout E9, as suggested by birthdating studies in the rat (9) and as revealed by E1.9 immunoreactivity. A second wave of neurogenesis starts at E10, lasts for about 48 hr, and populates the trigeminal ganglion. These secondary neurons are E1.9+ as soon as they differentiate. Major axonal growth is visible by E9.5, at the 21-somite stage; Fig. lb shows that the ganglion's polarity is established as its dorsal tip has sent out a short (100,um) projection. A few axons form this budding ophthalmic projection. Although at this point the leading axons and their growth cones stand close together, observation at a slightly later stage shows that they are growing independently of each other. Fig. 2a shows that by E10, at the 30-somite stage, the pioneer growth cones of the ophthalmic projection have reached past the eyecup. At the base of these first axons, a second wave of axogenesis is appearing. The leading axons are few in number (three or four); the path of their outgrowth is specific, restricted to a narrow strip between the retina and the telencephalon, yet they seem to be growing individually. These axons grew approximately 450,um in 12 hr; this growth rate ofjust under 40,um/hr is about twice as fast as that calculated by Davies (10) for secondary fibers in the maxillary nerve. The trailing a (a) E9 (14-somite-stage) mouse embryo stained with mab E1.9. A few cells in the region of the primitive trigeminal ganglion are FIG. 1. E1.9-immunoreactive. Axonal profiles (arrowheads) span the width of the ganglionic space and orient toward the periphery. D, dorsal; A, anterior. (Bar = 50,um.) (b) Trigeminal ganglion of an E9.5 (21-somite-stage) mouse embryo stained with mab E1.9. A few axons are emerging at the dorsal tip of the ganglion and form the budding ophthalmic projection (o). The growth cones stand close together yet they grow independently. More medially in the ganglion, fibers are starting to exit the ganglionic space (arrows). These will pioneer the maxillary nerve. D, dorsal; A, anterior. (Bar = 50,um.)

3 Neurobiology: Stainier and Gilbert Proc. Natl. Acad. Sci. USA 87 (1990) 925 D FIG. 2. (a) Trigeminal ganglion of an E10 (30-somite-stage) mouse embryo stained with mab E1.9. The pioneer axons of the ophthalmic projection (o) have reached past the eyecup (e). A few axons (three or four) pioneer this narrow pathway. At the base of this projection, more fibers are emerging and are starting to grow along the initial axons. Some of the pioneering maxillary (mx) fibers are seen extending. At this early E10 stage, the leading maxillary fibers are about 100,um away from their target area. The distal tip of the mandibular (md) projection has been severed accidentally. The first and second branchial bars are indicated by 1 and 2; dots delineate the first branchial bar pushed up against the maxillary process during mounting. (Bar = 50 Am.) (b) Trigeminal ophthalmic projection of an E10.5 (35-somite-stage) mouse embryo stained with mab E1.9. The pioneer axons of the ophthalmic (o) branch have not grown past the point reached at the 30-somite stage (a). The leading axons have now started branching and more fibers have grown along them. Branching ophthalmic fibers now surround the eyecup (e) dorsally as well as ventrally. D, dorsal; A, anterior. (Bar = 50 Am.) axons near the ganglion seem to be associating with the early fibers, and one can subsequently observe their growth along the pioneering axons. Fig. 2b shows that by E10.5, at the 35-somite stage, the ophthalmic projection has not grown much past the point reached at the 30-somite stage, in Fig. 2a, Df OA FIG. 3. Branching of pioneer fibers in the vicinity of the epithelium of the maxillary process, as seen in a 16-mm cryostat section of an E10.5 (37-somite-stage) mouse embryo stained with mab E1.9. Pioneer fibers first reach the proximity of the maxillary epithelium during E10 and start branching at E10.5. D, dorsal; A, anterior. (Bar = 50 Am.) Our embryonic day ran from noon to noon and we took CD1 embryos in the early afternoon and around midnight. Davies and Lumsden (11) obtained embryos from 8-hr overnight mating of CD1 (Charles River Breeding Laboratories) mice; their embryonic day ran from midnight to midnight. We believe that most of their work was done with embryos slightly older than ours. A - but that its leading axons have started branching, an early event in the innervation process. Neurogenesis and axonogenesis in the maxillary and mandibular branches proceed in a similar fashion, although with a slightly later onset than in the ophthalmic branch. A limited number of pioneer axons lead the way, directly growing to the general area of their respective target (see for example the maxillary fibers in Fig. 2a). After a period of little distal growth, these pioneer axons start branching in the vicinity of the target. Meanwhile, a second wave of axons projects and grows, apparently fasciculating on the leading fibers. Later, these secondary axons themselves branch to innervate the full extent of the target area. Fig. 3 shows a few pioneer fibers branching in the vicinity of the epithelium of the maxillary process at the 37-somite stage (E10.5). In the CNS, mesencephalic neural crest cells emerge and start migrating at the 4- to 7-somite stage (12). Some of them give rise to the primary sensory neurons that form the MesV (13). At E8.5, by the 10-somite stage, E1.9 immunoreactivity first outlines a few neurons that lie in the rostral part of the mesencephalon. This is the earliest neuronal outgrowth that we have detected in the CNS; previously, neurons were thought to first differentiate in the mouse CNS by E9-E10 (14). Fig. 4 shows two adjacent neurons that send short projections caudally towards the rhombencephalon. The location, time of birth, and direction of projection of the neurons in Fig. 4 are characteristic of MesV neurons. Indeed, mab B30, which we showed to be specific for MesV neurons in the mouse CNS (8), outlines such caudally projecting neurons at a slightly later stage in their naturation. As more mesencephalic neurons differentiate (MesV neurons differ-

4 926 Neurobiology: Stainier and Gilbert Proc. Natl. Acad. Sci. USA 87 (1990) FIG. 4. Earliest detectable neurons in the mouse CNS. An E8.5 (10-somite-stage) mouse embryo stained with mab E1.9 is shown at the level of the midbrain. Two neurons lie clustered in the mesencephalon (mes). Their location and time of birth are characteristic of MesV neurons. They also send short projections caudally towards the rhombencephalon. D, dorsal; A, anterior. (Bar = 50,um.) entiate at a slow but steady rate from E8.5 to E12), their axons also project caudally in an independent manner; fasciculation does not take place. This axonal separation remains constant and can be seen in postnatal animals (8). MesV neurons are clustered, but their axons remain distinct when projecting to the brainstem. When they exit the CNS at the level of the pons, these MesV axons pass through the trigeminal ganglion and grow along the pioneered mandibular tract to their peripheral targets. DISCUSSION In the 10-somite-stage mouse embryo, cells from the epidermal placodes start migrating into the region of the primitive trigeminal ganglion, differentiate, and give rise to the first trigeminal neurons shortly after E9 (15). Later-differentiating, neural crest-derived neurons populate most of the ganglion (5, 6, 9, 15). [3H]Thymidine birthdating studies in the rat (9) as well as electron microscope observations (11) confirm that trigeminal neurons differentiate shortly after E9 and that the first fibers leave the ganglion at its dorsal tip at E9.5. These developmental studies and the observation that mab E1.9 stains both neural crest-derived and placodederived sensory neurons (e.g., neurons of the dorsal root ganglia and neurons of the VIIth and VIIIth ganglia, respectively) indicate that the E1.9-immunoreactive axons that leave the ganglion at its dorsal tip by E9.5 (Fig. lb) are the first to grow out. Furthermore, E1.9 immunostaining of cryostat sections of E10.5-E12 trigeminal ganglion shows no E1.9-negative areas, indicating that mab E1.9 stains all trigeminal neurons. In the E10 embryo, electron microscopy shows small fascicles of nerve fibers leaving the ganglion (11). Our study complements this observation by showing that within these fascicles, a few fibers have reached the target area while the others have just started growing, apparently by fasciculating on the early fibers. This is best exemplified by the ophthalmic fibers in Fig. 2a: the pioneer axons have extended ,um and reached their target area while the secondary fibers have grown ,um out of the ganglion. Selective guidance cues allow axons to grow along very precise pathways to their targets (reviewed in ref. 16). In general, axons may follow pathways labeled by specific molecular cues (17), or they may respond to gradients of diffusible factors secreted by cells within the target (18-20). Two lines of in vitro work have provided evidence for the chemotropic guidance of axons: the observation that the floor plate orients the growth of rat spinal cord commissural axons (18) and a set of experiments done with the mouse trigeminal ganglion (19, 20). In this latter work, the observation of explants cocultured in collagen gels led Lumsden and Davies (19, 20) to conclude that trigeminal cutaneous target fields explanted at E10 and Eli release a diffusible factor which specifically directs neurite outgrowth from trigeminal ganglia of the same age (and that this factor is distinct from nerve growth factor). Lacking specific markers for trigeminal neurons, Lumsden and Davies designed their experiments on temporal assumptions drawn from observations of silver-stained wholemounts and thin sections (11). By E9.5, a small number of fibers have emerged from the dorsal side of the ganglion to pioneer the ophthalmic projection. But, whereas Lumsden and Davies reported that fibers first reach the maxillary target epithelium early in Eli (they also described a few fibers contacting the epithelium of the mandibular process at E10.5), we observe that pioneer axons first reach the proximity of the maxillary target epithelium during E10 and start branching out at E10.5 (Fig. 3). During Eli, the secondary fibers growing along these initial axons first approach their target. Thus, in culturing E10 and Eli explants, Lumsden and Davies were not looking at the earliest sensory outgrowth but at secondary outgrowth approaching tissue previously explored by the pioneer axons. Tropism may be involved in conjunction with fasciculation in guiding the secondary fibers. Alternatively, such a target-derived tropic effect in vitro may only be a by-product of the differentiation of the target cells after an

5 Neurobiology: Stainier and Gilbert initial axonal approach (one expects such factors to be secreted from the differentiating target cell to adjust the extent of its innervation, primarily through controlling local sprouting). In fact, there was a small but consistent increase in target-directed outgrowth in their collagen gel experiments on moving from E10 to Eli explants, which may reflect the ongoing differentiation of the target tissue in the presence of the pioneer axons [as has been suggested for embryonic chicken lumbosacral motoneurons (21)]. Tropic guidance of the earliest trigeminal sensory neurons remains an attractive hypothesis which should be tested in cocultures of E9-E9.5 explants. Our results show that in the periphery, the initial trigeminal axons may constitute a "labeled" pathway available for the guidance of later fibers; thus, different mechanisms may be involved in guiding the pioneers versus the later axons. By contrast, in the CNS, MesV axons do not fasciculate; the initial axons of the MesV cannot directly serve as guides for subsequent MesV fibers. We thank Dr. A. Ghysen for interesting discussions and Dr. C. Fulwiler and M. Grether for critical reading of the manuscript. 1. Bate, C. M. (1976) Nature (London) 260, Raper, J. A., Bastiani, M. J. & Goodman, C. S. (1984) J. Neurosci. 4, Klose, M. & Bentley, D. (1989) Science 245, Proc. Natl. Acad. Sci. USA 87 (1990) McConnell, S. K., Ghosh, A. & Shatz, C. J. (1989) Science 245, Moody, S. A., Quigg, M. S. & Frankfurter, A. (1989) J. Comp. Neurol. 279, Covell, D. A. & Noden, D. M. (1989) J. Comp. Neurol. 286, Theiler, K. (1989) The House Mouse (Springer, Berlin), 2nd Ed. 8. Stainier, D. Y. & Gilbert, W. (1989) J. Neurosci. 9, Altman, J. & Bayer, S. A. (1982) Adv. Anat. Embryol. Cell Biol. 74, Davies, A. M. (1987) Development 100, Davies, A. M. & Lumsden, A. G. S. (1984) J. Comp. Neurol. 223, Chan, W. Y. & Tam, P. P. L. (1988) Development 102, Narayanan, C. H. & Narayanan, Y. (1978) J. Embryol. Exp. Morphol. 43, Jacobson, M. (1978) Developmental Neurobiology (Plenum, New York), pp Verwoerd, C. D. A. & van Oostrom (1979) Adv. Anat. Embryol. Cell Biol. 58, Dodd, J. & Jessell, T. M. (1988) Science 242, Ghysen, A. (1978) Nature (London) 274, Tessier-Lavigne, M., Placzek, M., Lumsden, A. G. S., Dodd, J. & Jessell, T. (1988) Nature (London) 336, Lumsden, A. G. S. & Davies, A. M. (1983) Nature (London) 306, Lumsden, A. G. S. & Davies, A. M. (1986) Nature (London) 323, Lance-Jones, C. & Landmesser, L. (1981) Proc. R. Soc. London Ser. B. 214, 1-18.