- Email: [email protected]
Ascl1 mutants
Developmental Biology 295 (2006) 67 – 75 www.elsevier.com/locate/ydbio
Delays in neuronal differentiation in Mash1/Ascl1 mutants Alexandre Pattyn a,1 , François Guillemot b , Jean-François Brunet a,⁎ a
b
CNRS UMR 8542, Département de Biologie, Ecole normale supérieure, 75005 Paris, France Molecular Neurobiology, National Institute for Medical Research, Mill Hill, London NW71AA, UK Received for publication 16 December 2005; revised 7 March 2006; accepted 8 March 2006 Available online 4 May 2006
Abstract The inactivation of a developmental transcription factor may lead to the complete absence of a specific cell type. More commonly, though, it only partially impairs its generation. The modalities of this partial effect have rarely been documented in any detail. Here, we report a novel function for the bHLH transcription factor Ascl1/Mash1 in the generation of the nucleus of the solitary tract (nTS). In Mash1−/− late embryos, the nTS is markedly atrophic. Tracing back the origin of this atrophy, we show that nTS precursors appear in the mutants 1 day later than in the wild type and then accumulate at a slower pace. We also show that the previously reported atrophy of the sympathetic chain in Mash1 mutants is similarly preceded by a delay of 1 to 2 days in the appearance of differentiated ganglionic cells. Finally, we provide evidence that the acceleration imposed by Mash1, regardless of the production of post-mitotic cells, affects differentiation itself, both generic and type-specific. © 2006 Elsevier Inc. All rights reserved. Keywords: Mash1/Ascl1; Proneural gene; Neuronal differentiation; Transcription factor; Nucleus of the solitary tract; Sympathetic ganglia; Mouse; Knock-out
Introduction Proneural bHLH transcription factors are key players in the process of neurogenesis. In flies, where they were first discovered, their best characterized action, paradoxically, is in non-cell autonomous inhibition of neurogenesis, by the process of lateral inhibition (Gibert and Simpson, 2003). This process is also thought to occur in the neuroepithelium of vertebrates and maintain a pool of dividing progenitors throughout the protracted period of neurogenesis (Kageyama and Nakanishi, 1997; Ross et al., 2003). The cell autonomous promotion of neuronal differentiation by proneural factors is much less understood, be it in flies or vertebrates. It is commonly construed as the “selection” of “committed” neuronal precursors from dividing progenitors, the promotion of cell cycle exit (the clearest hallmark of “commitment”) and the upregulation of “generic” neuronal differentiation traits, which ensues. However, before “commitment”, proneural bHLH genes pattern the ⁎ Corresponding author. Fax: +33 1 44 32 23 23. E-mail address: [email protected] (J.-F. Brunet). 1 Present address: Institut des Neurosciences de Montpellier INSERM U583, 34091 Montpellier Cedex 5, France. 0012-1606/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2006.03.008
neuroepithelium itself (e.g., Filippi et al., 2005; Gowan et al., 2001); during differentiation, they control subtype specific traits along with generic or pan-neural ones (e.g., Fode et al., 2000; Lee and Pfaff, 2003; Lo et al., 2002; Parras et al., 2002); and they can influence such late events as the migration of neuronal precursors (Ohsawa et al., 2005; Tiveron et al., 2003). The emerging picture is that, collectively, the actions of proneural bHLH genes span the entire course of neuronal differentiation. There is evidence that several are expressed in succession in the same lineage (e.g., Cau et al., 1997; Fode et al., 1998), each gene possibly in charge of a different phase in a multistep differentiation process. The precise definition of these steps, however, is still elusive, and awaits the identification of direct target genes, few of which are known (e.g., Lee et al., 2004). Mash1/Ascl1 is the only murine member of the achaete/scute family of bHLH genes expressed in the nervous system. The list of neuronal types whose differentiation depend on Mash1 has steadily increased since the initial description of its null mutant phenotype (Guillemot et al., 1993). To date, it includes olfactory neurons (Cau et al., 1997; Guillemot et al., 1993) and interneurons (Parras et al., 2004), sympathetic, parasympathetic, and enteric neurons (Blaugrund et al., 1996; Guillemot et al., 1993), glomus cells of the carotid body (Kameda, 2005),
68
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
GABAergic neurons of the cortex (Casarosa et al., 1999), central noradrenergic (Hirsch et al., 1998), and serotonergic (Pattyn et al., 2004) neurons, retinal neurons (Akagi et al., 2004; Tomita et al., 1996), V2 and dl5 interneurons of the spinal cord (Helms et al., 2005; Li et al., 2005; Parras et al., 2002) and cranial motoneurons (Ohsawa et al., 2005). Here, we report that Mash1 is required for the proper formation of relay visceral sensory neurons of the hindbrain, i.e., of the nucleus of the solitary tract, as well as the associated primary chemoreceptors of the area postrema. We show that these two structures do form in the mutants but are atrophic, due in part to a delay in the onset of neuronal differentiation. We also show that the previously described atrophy of the sympathetic chain in Mash1 mutants results from a similar delay in the differentiation of sympathoblasts and discuss several mechanisms. Materials and methods
2001) (Fig. 1), making this region topologically equivalent to the progenitor domain for dl3 interneurons of the spinal cord (Helms and Johnson, 2003). The post-mitotic precursors of the nTS are first detectable at E10-E10.5 by their expression of Rnx/ Tlx3 (Qian et al., 2001), Lmx1b, and Phox2b (Dauger et al., 2003) (and Fig. 1A). The onset of Phox2b expression slightly precedes the downregulation of Mash1 (arrows in Figs. 1A, B). The nTS precursors migrate ventrally as soon as they are detectable and, eventually, aggregate just dorsal to the motor nucleus of the vagus nerve (Dauger et al., 2003; Qian et al., 2001) (and Figs. 1A, B). While the nTS of Lmx1b knock-outs has not been examined to date, inactivation of Tlx3 or Phox2b results in nTS agenesis, both genes being switched on independently of each other and Tlx3 being required for the maintenance of Phox2b (Dauger et al., 2003; Qian et al., 2001). It has been argued that Mash1, expressed in nTS progenitors, has a specification function in nTS formation, since it can ectopically activate Tlx3 in the chick
Genotyping and maintenance of mutant mice Mash1, Ngn2, and Mash1KINgn2 mutant mice were produced and genotyped as previously described (Guillemot et al., 1993; Parras et al., 2002).
In situ hybridization, immunohistochemistry, and immunofluorescence Riboprobes for DBH, dHand (gift of Y-S. Dai), Gata3 (gift of D. Engel), Mash1, Ngn2, Ret, Tlx3 (gift of Q. Ma), Olig3 (gift of C. Birchemeier) and Tubb3 (gift of C.W. Ragsdale) and antibodies against BrdU (Sigma), Lmx1b (gift of T. Jessell), Phox2a and Phox2b were used for in situ hybridization, immunohistochemistry, and in situ hybridization combined with immunohistochemistry as in Tiveron et al. (1996) and for immunofluorescence as in Pattyn et al. (2000). Double immunofluorescence experiments were analyzed on a TPS/ SP2 Leica confocal microscope and pictures were superimposed in Photoshop (Adobe).
BrdU analysis and cell counts BrdU was injected intraperitoneally into pregnant mice (6 mg/mouse) at E9.5, E10.5, E11.5, and E12.5, 24 or 48 h before dissecting the embryos. Embryos were then processed for double-immunofluorescence analysis using antibodies against BrdU and Phox2b, as described (Pattyn et al., 1997). For each experiment, the number of BrdU+/Phox2b+ cells was counted on transverse sections at the level of rhombomeres 7 and 8 (where nTS neurons can be identified unambiguously) of wild-type and Mash1 mutant embryos (for each genotype, n = 22 sections at E10.5, n = 14 sections at E11.5 and E12.5 and n = 15 sections at E13.5, from 2 embryos). Values are means ± SEM.
Results Normal development of the nTS The nTS is an elongated nucleus located in the dorsal part of the hindbrain formed by first order relay sensory neurons postsynaptic to baroreceptors, chemoreceptors and osmoreceptors of the distal VIIth, IXth and Xth cranial ganglia, whose central axons form the “solitary tract” (Blessing, 1997). Neurons of the nTS are born from E9.5 to E12.5 (Taber Pierce, 1973). They emerge at rhombomeric levels 4 to 7 (Pattyn et al., 1997) from the dorsal-most part of the Mash1-positive domain of the neuroepithelium (Hornbruch et al., 2005; Qian et al.,
Fig. 1. Hindbrain nTS precursors derive from a Mash1-positive domain of the neuroepithelium. Combined in situ hybridization with Mash1 (dark blue) and immunohistochemistry with a Phox2b antibody (orange) on transverse sections through r7 at E10.5 (A) and E11.5 (B). Only the right side of the sections is shown. Right panels are close-ups of the respective framed areas. The dorsalmost Mash1-positive progenitors give rise to nTS precursors, which transiently coexpress Mash1 and the early post-mitotic marker Phox2b (arrows). Soon after they exit the ventricular zone, nTS precursors migrate ventrally (blue arrowheads) while ventrally generated Phox2b-positive motoneurons (asterisk in panel A) migrate dorsally (red arrowheads).
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
neural tube (Hornbruch et al., 2005). However, the ectopic Tlx3+ neurons were not further characterized and could have been dl5 interneurons that are also Mash1-dependent (Helms et al., 2005 and this report). To pin down the role of Mash1 in nTS formation, we examined the nTS of Mash1 knock-out embryos (Guillemot et al., 1993). nTS formation depends on Mash1 At perinatal stages, the nTS and associated area postrema of Mash1 mutants were markedly atrophic (Fig. 2A). To trace the
69
origin of this defect, we monitored the formation of the nTS from its inception. In E10.5 Mash1 null mutants, no cell expressing Tlx3, Phox2b, or Lmx1b was detected in the dorsal part of the hindbrain in contrast to the wild type (Figs. 2B–D). The more ventral stripe of Lmx1b and Tlx3 expression was also abrogated, in agreement with the recent demonstration that dl5 interneurons depend on Mash1 (Helms et al., 2005). In contrast, Phox2b expression was intact in branchial and visceral motoneuronal (BM/VM) precursors whose initial differentiation is Mash1-independent (Pattyn et al., 2000, 2004). A subtle delay in their dorsal migration was detected (arrowhead in
Fig. 2. The development of the nTS and associated area postrema is affected in Mash1 mutant mice. Transverse sections through the caudal hindbrain of wild-type (left panels) and Mash1 mutant (right panels) mice at E18.5 (A), E10.5 (B–D), E11.5 (E–G), and E12.5 (H), stained as indicated. (A) View of the dorsal vagal complex at E18.5 stained with Phox2b antibody, showing that the nTS and the area postrema (ap) are atrophic in Mash1 mutants. (B–D) At E10.5, nTS precursors (nTSp) are not detected in Mash1 knock-out mice, as shown by the lack of Phox2b (B), Tlx3 (C) and Lmx1b (D) expression. Tlx3 and Lmx1b also reveal the absence of hindbrain dl5 neurons at this stage (C, D). In contrast, the ventrally born Phox2b+ visceral motoneurons (vMN) are present, although slightly delayed in their dorsal migration (arrowheads in panel B). (E–G) At E11.5 a few nTS precursors (nTSp) are detectable in Mash1−/− embryos, but no dl5 neuron (arrowhead in panels F and G). (H) At E12.5, an increasing amount of nTS neurons expressing Phox2b and Lmx1b (yellow cells) are present in Mash1 mutants (albeit in greatly reduced numbers compared to control animals) and undergo a normal ventral migration. Note that the Lmx1b+/Phox2b− cells born at that stage from the Mash1+ domain of the alar plate (red cells) are also produced in the mutants (their apparent increase relative to wild type is not significant) (white arrowheads).
70
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
Fig. 2B; see also Figs. 2E, G) in agreement with recent data showing the role of Math3 and Mash1 in the lateral migration of facial and trigeminal motoneurons (Ohsawa et al., 2005). One day later, at E11.5, when a large population of nTS precursors were already born in the wild type and had undergone part of their ventral migration, the first nTS precursors expressing Phox2b, Tlx3, and Lmx1b were detected in the Mash1 mutants (Figs. 2E–G). In contrast, dl5 interneurons were still completely absent in the mutants, while BM/VM precursors had reached their normal site of aggregation in the nucleus ambiguus and
dorsal motor nucleus of the vagal nerve (Fig. 2G). At E12.5, when the generation of nTS precursors is about to end, their number was conspicuously reduced in the mutants (Fig. 2H). In order to quantify the dynamics of nTS formation in wildtype and Mash1 mutants, we injected BrdU at E9.5, E10.5, E11.5, or E12.5 and counted newborn Phox2b-positive neurons 1 day after injection. This procedure confirmed a 24-h delay in the onset of nTS precursor appearance in the mutants compared to wild type (Figs. 3A, B). Afterwards, the accumulation of Phox2b-positive cells remained slower in the mutants than in
Fig. 3. Mash1 sets the timing of nTS precursor accumulation. Transverse sections through r7 at E10.5, E11.5, E12.5, and E13.5 of wild-type (A) and Mash1−/− (B) embryos, 24 h after injection of BrdU, immunostained for BrdU (in green) and Phox2b (in red). Pictures show the dorsal part of the neural tube where nTS neurons arise. Only the right side is shown, the ventricular zone is on the left, the mantle layer on the right. For each experiment, double-positive cells (in yellow) represent postmitotic nTS precursors born from a cell cycle whose S phase soon followed BrdU injection. For each stage, the number of BrdU+/Phox2b+ cells 24 h after injection (h.a.i) of BrdU is plotted on the graph in C (controls in black, Mash1 mutants in grey). In control embryos (A, C), the generation of nTS post-mitotic precursor starts at around E9.5, reaches a plateau between E10.5 and E12, and stops at around E13. In Mash1−/− embryos (B, C), the generation of nTS post-mitotic precursor is delayed by one day, and starts at around E10.5. Between E10.5 and E12.5 nTS neurons accumulate at a lower pace in the mutants than in control embryos.
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
the wild type and both abated synchronously at E12.5 (Figs. 3A, B). Thus, the generation of nTS precursors lasted 4 days and peaked between E11.5 and E12.5 in wild-type embryos, in agreement with earlier estimations (Taber Pierce, 1973), while in Mash1 mutants it was shortened by one day due to a delayed onset and never reached the same levels (Fig. 3C). The residual production of nTS precursors in Mash1 mutants suggests that another gene can partially compensate for Mash1. At the dorso-ventral level where nTS precursors arise, Ngn2 (but not Ngn1, not shown) is expressed in scattered neuroepithelial cells and in a dense population of cells in a zone intermediate between the neuroepithelium and the mantle layer (Fig. 4A). Some of the Ngn2+ cells also express Phox2b (arrows in Fig. 4A) and are therefore post-mitotic (Pattyn et al., 1997). This pattern of Ngn2 expression is similar to that described in the alar plate of the spinal cord (Helms et al., 2005). Examination of Phox2b+ nTS precursors in Ngn2 knock-out at E12 did not reveal any obvious phenotype, while the double Mash1/Ngn2 knock-out displayed a phenotype indistinguishable from that of the simple Mash1 knock-out (Fig. 4B). Hence,
71
Ngn2 is not in charge of the formation of the residual nTS neurons in Mash1 mutants. Nature of the delay in nTS formation The delay in the appearance of Phox2b+ precursors in the dorsal hindbrain of the mutants could be due to a delay in the production of post-mitotic cells, i.e., reflect the proneural function of Mash1. To explore this possibility, we supplied proneural activity in the nTS region of Mash1 null embryos by using Mash1KINgn2 animals in which the Ngn2 coding sequence replaces that of Mash1 in the Mash1 locus. This configuration is known to restore proneural activity at all sites examined so far (Parras et al., 2002; Pattyn et al., 2004). Mash1KINgn2 embryos displayed a delay in nTS formation indistinguishable from Mash1 null mutants (Fig. 4C). These data argue that a delayed production of post-mitotic cells cannot be the sole explanation for the defect in nTS formation in Mash1 mutants, and that the specification of nTS precursors must itself be delayed.
Fig. 4. Ngn2 cannot compensate for the absence of Mash1 during nTS formation. (A) Expression pattern of Ngn2. Combined in situ hybridization with Ngn2 (dark blue) and immunochemistry for Phox2b (orange) on transverse sections through r7 of a wild-type embryo at E10.5. Only the right side is shown. The right panel is a close up of the framed area. Ngn2 is widely expressed in the ventricular zone of the neural tube throughout the dorso-ventral axis. At the level where nTS precursors arise (framed area), Ngn2 is expressed in the ventricular zone (bracket) in fewer cells than Mash1 (compare with Fig. 1A). In the mantle layer, however, it is largely coexpressed with the post-mitotic marker Phox2b (arrows) and is switched off once the cells have started their ventral migration (arrowhead). (B) Ngn2 is not responsible for the late residual generation of nTS neurons. Transverse sections through r7 of E12 wild-type, Ngn2−/− and Mash1−/−; Ngn2−/− embryos as indicated, stained for Phox2b. No obvious phenotype is observed in Ngn2 mutants, and the phenotype of Mash1;Ngn2 double mutants is similar to that of Mash1 single mutants (compare with Fig. 2). (C) Ngn2 cannot replace Mash1 during the early generation of nTS neurons. Transverse sections through r7 of wild-type and Mash1KINgn2 at E10.5 and E12, as indicated, costained for Lmx1b (in red) and Phox2b (in green). Mash1KINgn2 animals exhibit the same delay of nTS neuron production as Mash1 mutants. Note that the production of dI5 neurons is not rescued either (arrowhead).
72
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
A possibility is that, in Mash1 mutants, each nTS precursor, although born on schedule, is slowed down in its post-mitotic differentiation. We injected BrdU at E9.5 (i.e., when the first nTS precursors are normally born) and assessed Phox2b/BrdU double labeling at E11.5 (i.e., when Phox2b first appears in the mutants, 1 day later than in wild types). No double labeling was detected in the mutants, in contrast to the wild type (Fig. 5). Thus, the first cells that end up expressing Phox2b are born later in the mutant than in wild type, but then switch on Phox2b on schedule, relative to their birth date. Instead, the delay in nTS specification could occur upstream of cell cycle exit. To test this idea, we monitored the bHLH gene Olig3, a marker coexpressed with Mash1 in the cycling progenitors of the nTS (not shown). There was no alteration in Olig3 expression from E9.5 to E10.5 in Mash1 nulls (not shown) (see Discussion). Delay in neuronal differentiation in autonomic ganglia of Mash1 null embryos Another case of an incomplete differentiation block in Mash1 mutants is represented by sympathetic ganglia. Previous reports mention an incompletely penetrant phenotype with agenesis of the superior cervical ganglion contrasting with mere atrophy in the caudal chain (Guillemot et al., 1993; Hirsch et al., 1998). We reexamined the timing of sympathoblast differentiation in Mash1 nulls in more detail. At E10.5, at thoracic levels, sympathoblasts aggregate around the dorsal aorta in both wildtype and mutant animals, albeit in a somewhat looser arrangement in the mutants (Fig. 6A). As previously shown, Phox2b expression was intact, and that of Ret reduced in the mutant ganglion anlagen (Guillemot et al., 1993; Hirsch et al.,
Fig. 5. Delayed birth of Mash1-null nTS precursors. Transverse sections through r7 of control and Mash1 mutants at E11.5 after injection of BrdU at E9.5, costained for BrdU (in green) and Phox2b (in red). In contrast to the control (yellow cells in the left panel), no double-positive cells are detected in Mash1 mutants (right panel). See text for details.
1998) (Figs. 6A, D). None of the other differentiation markers normally switched on at E10.5 could be detected, be they subtype-specific (Phox2a, dHAND, Gata3, and Dbh) or panneural (βIII-tubulin, SCG10) (Figs. 6A, D). One day later, however, at E11.5, the expression of Ret, as well as that of dHAND was rescued, but the number of expressing cells was reduced. Phox2a and Gata3 were now weakly detected, while the pan-neural and subtype specific genes βIII-tubulin, SCG10, and Dbh were still silent (Figs. 6B, D). Finally, at E12.5, at levels where Phox2b+ ganglionic cells could still be detected (i.e., caudal to the cervical ganglion), the full sympathoadrenergic phenotype was expressed (Figs. 6C, D). A similar dynamic was observed at lumbar levels. Discussion While some transcription factors are absolutely required for the differentiation of neuronal types, others are only partially so, a phenomenon usually subsumed under the concept of “partial redundancy”. The same transcription factor can exemplify one or the other situation depending on the neuronal type examined. For example Mash1 is absolutely required for the differentiation of central serotonergic (Pattyn et al., 2004) or central noradrenergic (Hirsch et al., 1998) neurons but only partially so for that of ventral telencephalic neurons (Casarosa et al., 1999). Here, we demonstrate a complex pattern of requirement for Mash1 in the generation of the nTS. It can be deconstructed into two phases: an early phase lasting about 24 h during which Mash1 is absolutely required for the generation of nTS precursors and a late phase during which nTS precursors accumulate in the absence of Mash1, albeit at a reduced pace. We document a similar phenomenon in sympathetic ganglia, where differentiation of ganglionic cells fails in the mutants for a full day after its onset in wild-type embryos, then proceeds towards the full-blown sympathoadrenal phenotype, but at a slower pace. The latter data argue against the existence of Mash1-independent and Mash1dependent “subprograms” of sympathoadrenal differentiation (Sommer et al., 1995), in favor of a single differentiation program whose every step is Mash1-dependent at first and Mash1independent later. A similar phenomenon had been previously documented in retinal explants from Mash1 mutants where the appearance of rods, horizontal, and bipolar cells is not abrogated but delayed by several days (Tomita et al., 1996). Two mechanisms can be proposed to account for the early block of nTS and sympathetic neuron differentiation. According to a “modular” mechanism, the inactivation of Mash1 unveils a chronological heterogeneity in the genetic control of these structures: from E9.5 to E10.5, Mash1 would be the only gene capable of directing nTS and sympathetic neuron identity, respectively, while later on, another gene would be coexpressed, with the same capacity. In this second phase, Mash1 would only increase the efficiency of neuronal differentiation. Such a bimodal ontogeny is observed in dorsal root ganglia, involving first Ngn2 then Ngn1 (Ma et al., 1999). In the nTS, Ngn2, although coexpressed with Mash1 is not that putative second gene since its mutation has no phenotype, whether on a wild-
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
73
Fig. 6. Delayed neuronal differentiation in autonomic ganglia of Mash1−/− embryos. (A–D) Transverse sections through the thoracic sympathetic chain of wild-type (left panels) and mutant (right panels) embryos at E10.5 (A), E11.5 (B) and E12.5 (C), stained as indicated. Panel D shows a summary of the timing of expression of sympathetic neuron markers in wild-type and Mash1−/− embryos. The onset of all neuronal differentiation markers is delayed in the mutants-except for Phox2b, as previously reported (Hirsch et al., 1998). See text for details.
type or a Mash1 null background. It is conceivable that Olig3, being expressed in the hindbrain dl3 domain as it is in the spinal dl3 domain (Muller et al., 2005), would play such a role. In sympathetic ganglionic cells, a candidate is the bHLH factor dHand, implicated in sympathoadrenal differentiation by gain of function experiments both in vitro (Howard et al., 1999) and in vivo (Howard et al., 2000) and expressed after Mash1 (Howard et al., 2000) (but see below). The second mechanism posits that the initial block in differentiation is the result of a quantitative role of Mash1 combined with a threshold effect. For example, differentiation would occur at a precise level of accumulation of one or several transcription factors and that level would be shifted (or
the accumulation slowed down) in the Mash1 mutants. The proneural role of bHLH proteins is classically understood in such dose-dependent terms: cells are thought to commit to the neural differentiation pathway according to their level of proneural gene expression (Bertrand et al., 2002; Gibert and Simpson, 2003). Thus, if a second proneural gene was coexpressed with Mash1 from the start, concurring to the overall proneural activity, the attainment of a critical level in any cell would be delayed in the Mash1 mutants, resulting in the initial block. Again, the coexpressed bHLH factor Olig3 could underlie the residual proneural activity of Mash1 mutants. However, replacement of Mash1 by Ngn2, which presumably restores proneural function, rescues neither the
74
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
initial block nor the subsequent lowered rate of nTS precursor accumulation, suggesting that the production of post-mitotic precursors is not the only limiting step, if it is at all, in Mash1 knock-outs. Rather, our data argue that the neuronal differentiation aspect of Mash1 function is likely to be itself of a quantitative nature. This is best illustrated in sympathetic ganglia, where differentiation proceeds independently of cell cycle exit (DiCiccoBloom et al., 1990; Rohrer and Thoenen, 1987; Rothman et al., 1980): we document six genes (dHAND, Phox2a, Gata3, Tubb3, SCG10, and Dbh) whose onsets of expression, very precocious and roughly synchronous in the wild type, are not only delayed but staggered in the mutants. The onset of dHAND expression is delayed by one day and followed, a few hours later, by those of Phox2a and Gata3, themselves followed by those of effector genes such as Dbh, βIII-tubulin, or SCG10. The “modular” mechanism outlined earlier is improbable in this case, since it requires that each group of synchronously expressed genes would be controlled by yet another transcription factor with Mash1-like properties. More likely, the expression of every single marker is enhanced by Mash1 and the fast kinetics with which they normally reach the threshold for detectable expression—not resolved in this study, but see Ernsberger et al. (2000) and, for chick embryos Tsarovina et al. (2004)—are slowed down by Mash1 inactivation. This is further suggested by the progressive increase in the level of Ret expression and in the number of cells expressing Gata3 and Phox2a in the mutants. In the nTS, although the onset of expression of Phox2b is not delayed with respect to cell cycle exit in those cells that go on to express it, the competence of nTS progenitors to produce such cells is delayed. A molecular underpinning of this competence, expressed in progenitors, remains to be identified. Whatever the explanation for the initial block in neuronal differentiation and its subsequent relief, Mash1, rather than an indispensable part of the gene expression machinery, should be seen, at least in these structures, as a mere accelerator for the onset of gene expression. This role can explain, in the mutants, the atrophy of structures such as the nTS or the caudal sympathetic chain (Hirsch et al., 1998) (and this paper) or can be completely compensated for as in rods and horizontal cells of the retina (Tomita et al., 1996). However, it could also underlie some all-or-none effects, such as the lack of locus coeruleus (Hirsch et al., 1998) or of the superior cervical ganglion, would the delay in differentiation and the window of competence of a given cell population completely overlap. Acknowledgments We thank C. Birchmeier, Y-S. Dai, D. Engel, T.M. Jessel, Q. Ma and C.W. Ragsdale for antibodies and probes. This work was supported by the Centre National de la Recherche Scientifique and grants from the Association pour la Recherche sur le Cancer (ARC), the Fondation Franco-Norvégienne (FFN), the EC (QLG2-CT-2001-01467) and the French Ministry of Research (ACI 2002).
References Akagi, T., Inoue, T., Miyoshi, G., Bessho, Y., Takahashi, M., Lee, J.E., Guillemot, F., Kageyama, R., 2004. Requirement of multiple basic helix– loop–helix genes for retinal neuronal subtype specification. J. Biol. Chem. 279, 28492–28498. Bertrand, N., Castro, D.S., Guillemot, F., 2002. Proneural genes and the specification of neural cell types. Nat. Rev., Neurosci. 3, 517–530. Blaugrund, E., Pham, T.D., Tennyson, V.M., Lo, L., Sommer, L., Anderson, D. J., Gershon, M.D., 1996. Distinct subpopulations of enteric neuronal progenitors defined by time of development, sympathoadrenal lineage markers and Mash-1-dependence. Development 122, 309–320. Blessing, W.W., 1997. Anatomy of the lower brainstem. In: The Lower Brainstem and Bodily Homeostasis. Oxford Univ. Press, New York London, pp. 29–99. Casarosa, S., Fode, C., Guillemot, F., 1999. Mash1 regulates neurogenesis in the ventral telencephalon. Development 126, 525–534. Cau, E., Gradwohl, G., Fode, C., Guillemot, F., 1997. Mash1 activates a cascade of bHLH regulators in olfactory neuron progenitors. Development 124, 1611–1621. Dauger, S., Pattyn, A., Lofaso, F., Gaultier, C., Goridis, C., Gallego, J., Brunet, J.-F., 2003. Phox2b controls the development of peripheral chemoreceptors and afferent visceral pathways. Development 130, 6635–6642. DiCicco-Bloom, E., Townes-Anderson, E., Black, I.B., 1990. Neuroblast mitosis in dissociated culture: regulation and relationship to differentiation. J. Cell Biol. 110, 2073–2086. Ernsberger, U., Reissmann, E., Mason, I., Rohrer, H., 2000. The expression of dopamine beta-hydroxylase, tyrosine hydroxylase, and Phox2 transcription factors in sympathetic neurons: evidence for common regulation during noradrenergic induction and diverging regulation later in development. Mech. Dev. 92, 169–177. Filippi, A., Tiso, N., Deflorian, G., Zecchin, E., Bortolussi, M., Argenton, F., 2005. The basic helix–loop–helix olig3 establishes the neural plate boundary of the trunk and is necessary for development of the dorsal spinal cord. Proc. Natl. Acad. Sci. U. S. A. 102, 4377–4382. Fode, C., Gradwohl, G., Morin, X., Dierich, A., LeMeur, M., Goridis, C., Guillemot, F., 1998. The bHLH protein NEUROGENIN 2 is a determination factor for epibranchial placode-derived sensory neurons. Neuron 20, 483–494. Fode, C., Ma, Q., Casarosa, S., Ang, S.L., Anderson, D.J., Guillemot, F., 2000. A role for neural determination genes in specifying the dorsoventral identity of telencephalic neurons. Genes Dev. 14, 67–80. Gibert, J.M., Simpson, P., 2003. Evolution of cis-regulation of the proneural genes. Int. J. Dev. Biol. 47, 643–651. Gowan, K., Helms, A.W., Hunsaker, T.L., Collisson, T., Ebert, P.J., Odom, R., Johnson, J.E., 2001. Crossinhibitory activities of Ngn1 and Math1 allow specification of distinct dorsal interneurons. Neuron 31, 219–232. Guillemot, F., Lo, L.-C., Johnson, J.E., Auerbach, A., Anderson, D.J., Joyner, A.L., 1993. Mammalian achaete-scute homolog 1 is required for the early development of olfactory and autonomic neurons. Cell 75, 463–476. Helms, A.W., Johnson, J.E., 2003. Specification of dorsal spinal cord interneurons. Curr. Opin. Neurobiol. 13, 42–49. Helms, A.W., Battiste, J., Henke, R.M., Nakada, Y., Simplicio, N., Guillemot, F., Johnson, J.E., 2005. Sequential roles for Mash1 and Ngn2 in the generation of dorsal spinal cord interneurons. Development 132, 2709–2719. Hirsch, M.R., Tiveron, M.-C., Guillemot, F., Brunet, J.-F., Goridis, C., 1998. Control of noradrenergic differentiation and Phox2a expression by MASH1 in the central and peripheral nervous system. Development 125, 599–608. Hornbruch, A., Ma, G., Ballermann, M.A., Tumova, K., Liu, D., Cairine Logan, C., 2005. A BMP-mediated transcriptional cascade involving Cash1 and Tlx-3 specifies first-order relay sensory neurons in the developing hindbrain. Mech. Dev. 122, 900–913. Howard, M., Foster, D.N., Cserjesi, P., 1999. Expression of HAND gene products may be sufficient for the differentiation of avian neural crestderived cells into catecholaminergic neurons in culture. Dev. Biol. 215, 62–77.
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75 Howard, M., Stanke, M., Schneider, C., Wu, X., Rohrer, H., 2000. The transcription factor dHAND is a downstream effector of BMPs in sympathetic neuron specification. Development 127, 4073–4081. Kageyama, R., Nakanishi, S., 1997. Helix–loop–helix factors in growth and differentiation of the vertebrate nervous system. Curr. Opin. Genet. Dev. 7, 659–665. Kameda, Y., 2005. Mash1 is required for glomus cell formation in the mouse carotid body. Dev. Biol. 283, 128–139. Lee, S.K., Pfaff, S.L., 2003. Synchronization of neurogenesis and motor neuron specification by direct coupling of bHLH and homeodomain transcription factors. Neuron 38, 731–745. Lee, S.K., Jurata, L.W., Funahashi, J., Ruiz, E.C., Pfaff, S.L., 2004. Analysis of embryonic motoneuron gene regulation: derepression of general activators function in concert with enhancer factors. Development 131, 3295–3306. Li, S., Misra, K., Matise, M.P., Xiang, M., 2005. Foxn4 acts synergistically with Mash1 to specify subtype identity of V2 interneurons in the spinal cord. Proc. Natl. Acad. Sci. U. S. A. 102, 10688–10693. Lo, L., Dormand, E., Greenwood, A., Anderson, D.J., 2002. Comparison of the generic neuronal differentiation and neuron subtype specification functions of mammalian achaete-scute and atonal homologs in cultured neural progenitor cells. Development 129, 1553–1567. Ma, Q., Fode, C., Guillemot, F., Anderson, D.J., 1999. Neurogenin1 and neurogenin2 control two distinct waves of neurogenesis in developing dorsal root ganglia. Genes Dev. 13, 1717–1728. Muller, T., Anlag, K., Wildner, H., Britsch, S., Treier, M., Birchmeier, C., 2005. The bHLH factor Olig3 coordinates the specification of dorsal neurons in the spinal cord. Genes Dev. 19, 733–743. Ohsawa, R., Ohtsuka, T., Kageyama, R., 2005. Mash1 and Math3 are required for development of branchiomotor neurons and maintenance of neural progenitors. J. Neurosci. 25, 5857–5865. Parras, C.M., Schuurmans, C., Scardigli, R., Kim, J., Anderson, D.J., Guillemot, F., 2002. Divergent functions of the proneural genes Mash1 and Ngn2 in the specification of neuronal subtype identity. Genes Dev. 16, 324–338. Parras, C.M., Galli, R., Britz, O., Soares, S., Galichet, C., Battiste, J., Johnson, J.E., Nakafuku, M., Vescovi, A., Guillemot, F., 2004. Mash1 specifies neurons and oligodendrocytes in the postnatal brain. EMBO J. 23, 4495–4505.
75
Pattyn, A., Morin, X., Cremer, H., Goridis, C., Brunet, J.-F., 1997. Expression and interactions of the two closely related homeobox genes Phox2a and Phox2b during neurogenesis. Development 124, 4065–4075. Pattyn, A., Hirsch, M.-R., Goridis, C., Brunet, J.-F., 2000. Control of hindbrain motor neuron differentiation by the homeobox gene Phox2b. Development 127, 1349–1358. Pattyn, A., Simplicio, N., van Doorninck, J.H., Goridis, C., Guillemot, F., Brunet, J.-F., 2004. Mash1/Ascl1 is required for the development of central serotonergic neurons. Nat. Neurosci. 7, 589–595. Qian, Y., Fritzsch, B., Shirasawa, S., Chen, C.L., Choi, Y., Ma, Q., 2001. Formation of brainstem (nor)adrenergic centers and first-order relay visceral sensory neurons is dependent on homeodomain protein Rnx/Tlx3. Genes Dev. 15, 2533–2545. Rohrer, H., Thoenen, H., 1987. Relationship between differentiation and terminal mitosis: sensory and ciliary neurons differentiate after terminal mitosis of precursor cells, whereas sympathetic neurons continue to divide after differentiation. J. Neurosci. 7, 3739–3748. Ross, S.E., Greenberg, M.E., Stiles, C.D., 2003. Basic helix–loop–helix factors in cortical development. Neuron 39, 13–25. Rothman, T.P., Specht, L.A., Gershon, M.D., Joh, T.H., Teitelman, G., Pickel, V.M., Reis, D.J., 1980. Catecholamine biosynthetic enzymes are expressed in replicating cells of the peripheral but not the central nervous system. Proc. Natl. Acad. Sci. U. S. A. 77, 6221–6225. Sommer, L., Shah, N., Mahendra, R., Anderson, D.J., 1995. The cellular function of MASH1 in autonomic neurogenesis. Neuron 15, 1245–1258. Taber Pierce, E., 1973. Time of origin of neurons in the brain stem of the mouse. Prog. Brain Res. 40, 53–65. Tiveron, M.-C., Hirsch, M.-R., Brunet, J.-F., 1996. The expression pattern of the transcription factor Phox2 delineates synaptic pathways of the autonomic nervous system. J. Neurosci. 16, 7649–7660. Tiveron, M.-C., Pattyn, A., Hirsch, M.-R., Brunet, J.F., 2003. Role of Phox2b and Mash1 in the generation of the vestibular efferent nucleus. Dev. Biol. 260, 46–57. Tomita, K., Nakanishi, S., Guillemot, F., Kageyama, R., 1996. Mash1 promotes neuronal differentiation in the retina. Genes Cells 1, 765–774. Tsarovina, K., Pattyn, A., Stubbusch, J., Muller, F., van der Wees, J., Schneider, C., Brunet, J.F., Rohrer, H., 2004. Essential role of Gata transcription factors in sympathetic neuron development. Development 131, 4775–4786.
Delays in neuronal differentiation in Mash1/Ascl1 mutants Alexandre Pattyn a,1 , François Guillemot b , Jean-François Brunet a,⁎ a
b
CNRS UMR 8542, Département de Biologie, Ecole normale supérieure, 75005 Paris, France Molecular Neurobiology, National Institute for Medical Research, Mill Hill, London NW71AA, UK Received for publication 16 December 2005; revised 7 March 2006; accepted 8 March 2006 Available online 4 May 2006
Abstract The inactivation of a developmental transcription factor may lead to the complete absence of a specific cell type. More commonly, though, it only partially impairs its generation. The modalities of this partial effect have rarely been documented in any detail. Here, we report a novel function for the bHLH transcription factor Ascl1/Mash1 in the generation of the nucleus of the solitary tract (nTS). In Mash1−/− late embryos, the nTS is markedly atrophic. Tracing back the origin of this atrophy, we show that nTS precursors appear in the mutants 1 day later than in the wild type and then accumulate at a slower pace. We also show that the previously reported atrophy of the sympathetic chain in Mash1 mutants is similarly preceded by a delay of 1 to 2 days in the appearance of differentiated ganglionic cells. Finally, we provide evidence that the acceleration imposed by Mash1, regardless of the production of post-mitotic cells, affects differentiation itself, both generic and type-specific. © 2006 Elsevier Inc. All rights reserved. Keywords: Mash1/Ascl1; Proneural gene; Neuronal differentiation; Transcription factor; Nucleus of the solitary tract; Sympathetic ganglia; Mouse; Knock-out
Introduction Proneural bHLH transcription factors are key players in the process of neurogenesis. In flies, where they were first discovered, their best characterized action, paradoxically, is in non-cell autonomous inhibition of neurogenesis, by the process of lateral inhibition (Gibert and Simpson, 2003). This process is also thought to occur in the neuroepithelium of vertebrates and maintain a pool of dividing progenitors throughout the protracted period of neurogenesis (Kageyama and Nakanishi, 1997; Ross et al., 2003). The cell autonomous promotion of neuronal differentiation by proneural factors is much less understood, be it in flies or vertebrates. It is commonly construed as the “selection” of “committed” neuronal precursors from dividing progenitors, the promotion of cell cycle exit (the clearest hallmark of “commitment”) and the upregulation of “generic” neuronal differentiation traits, which ensues. However, before “commitment”, proneural bHLH genes pattern the ⁎ Corresponding author. Fax: +33 1 44 32 23 23. E-mail address: [email protected] (J.-F. Brunet). 1 Present address: Institut des Neurosciences de Montpellier INSERM U583, 34091 Montpellier Cedex 5, France. 0012-1606/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2006.03.008
neuroepithelium itself (e.g., Filippi et al., 2005; Gowan et al., 2001); during differentiation, they control subtype specific traits along with generic or pan-neural ones (e.g., Fode et al., 2000; Lee and Pfaff, 2003; Lo et al., 2002; Parras et al., 2002); and they can influence such late events as the migration of neuronal precursors (Ohsawa et al., 2005; Tiveron et al., 2003). The emerging picture is that, collectively, the actions of proneural bHLH genes span the entire course of neuronal differentiation. There is evidence that several are expressed in succession in the same lineage (e.g., Cau et al., 1997; Fode et al., 1998), each gene possibly in charge of a different phase in a multistep differentiation process. The precise definition of these steps, however, is still elusive, and awaits the identification of direct target genes, few of which are known (e.g., Lee et al., 2004). Mash1/Ascl1 is the only murine member of the achaete/scute family of bHLH genes expressed in the nervous system. The list of neuronal types whose differentiation depend on Mash1 has steadily increased since the initial description of its null mutant phenotype (Guillemot et al., 1993). To date, it includes olfactory neurons (Cau et al., 1997; Guillemot et al., 1993) and interneurons (Parras et al., 2004), sympathetic, parasympathetic, and enteric neurons (Blaugrund et al., 1996; Guillemot et al., 1993), glomus cells of the carotid body (Kameda, 2005),
68
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
GABAergic neurons of the cortex (Casarosa et al., 1999), central noradrenergic (Hirsch et al., 1998), and serotonergic (Pattyn et al., 2004) neurons, retinal neurons (Akagi et al., 2004; Tomita et al., 1996), V2 and dl5 interneurons of the spinal cord (Helms et al., 2005; Li et al., 2005; Parras et al., 2002) and cranial motoneurons (Ohsawa et al., 2005). Here, we report that Mash1 is required for the proper formation of relay visceral sensory neurons of the hindbrain, i.e., of the nucleus of the solitary tract, as well as the associated primary chemoreceptors of the area postrema. We show that these two structures do form in the mutants but are atrophic, due in part to a delay in the onset of neuronal differentiation. We also show that the previously described atrophy of the sympathetic chain in Mash1 mutants results from a similar delay in the differentiation of sympathoblasts and discuss several mechanisms. Materials and methods
2001) (Fig. 1), making this region topologically equivalent to the progenitor domain for dl3 interneurons of the spinal cord (Helms and Johnson, 2003). The post-mitotic precursors of the nTS are first detectable at E10-E10.5 by their expression of Rnx/ Tlx3 (Qian et al., 2001), Lmx1b, and Phox2b (Dauger et al., 2003) (and Fig. 1A). The onset of Phox2b expression slightly precedes the downregulation of Mash1 (arrows in Figs. 1A, B). The nTS precursors migrate ventrally as soon as they are detectable and, eventually, aggregate just dorsal to the motor nucleus of the vagus nerve (Dauger et al., 2003; Qian et al., 2001) (and Figs. 1A, B). While the nTS of Lmx1b knock-outs has not been examined to date, inactivation of Tlx3 or Phox2b results in nTS agenesis, both genes being switched on independently of each other and Tlx3 being required for the maintenance of Phox2b (Dauger et al., 2003; Qian et al., 2001). It has been argued that Mash1, expressed in nTS progenitors, has a specification function in nTS formation, since it can ectopically activate Tlx3 in the chick
Genotyping and maintenance of mutant mice Mash1, Ngn2, and Mash1KINgn2 mutant mice were produced and genotyped as previously described (Guillemot et al., 1993; Parras et al., 2002).
In situ hybridization, immunohistochemistry, and immunofluorescence Riboprobes for DBH, dHand (gift of Y-S. Dai), Gata3 (gift of D. Engel), Mash1, Ngn2, Ret, Tlx3 (gift of Q. Ma), Olig3 (gift of C. Birchemeier) and Tubb3 (gift of C.W. Ragsdale) and antibodies against BrdU (Sigma), Lmx1b (gift of T. Jessell), Phox2a and Phox2b were used for in situ hybridization, immunohistochemistry, and in situ hybridization combined with immunohistochemistry as in Tiveron et al. (1996) and for immunofluorescence as in Pattyn et al. (2000). Double immunofluorescence experiments were analyzed on a TPS/ SP2 Leica confocal microscope and pictures were superimposed in Photoshop (Adobe).
BrdU analysis and cell counts BrdU was injected intraperitoneally into pregnant mice (6 mg/mouse) at E9.5, E10.5, E11.5, and E12.5, 24 or 48 h before dissecting the embryos. Embryos were then processed for double-immunofluorescence analysis using antibodies against BrdU and Phox2b, as described (Pattyn et al., 1997). For each experiment, the number of BrdU+/Phox2b+ cells was counted on transverse sections at the level of rhombomeres 7 and 8 (where nTS neurons can be identified unambiguously) of wild-type and Mash1 mutant embryos (for each genotype, n = 22 sections at E10.5, n = 14 sections at E11.5 and E12.5 and n = 15 sections at E13.5, from 2 embryos). Values are means ± SEM.
Results Normal development of the nTS The nTS is an elongated nucleus located in the dorsal part of the hindbrain formed by first order relay sensory neurons postsynaptic to baroreceptors, chemoreceptors and osmoreceptors of the distal VIIth, IXth and Xth cranial ganglia, whose central axons form the “solitary tract” (Blessing, 1997). Neurons of the nTS are born from E9.5 to E12.5 (Taber Pierce, 1973). They emerge at rhombomeric levels 4 to 7 (Pattyn et al., 1997) from the dorsal-most part of the Mash1-positive domain of the neuroepithelium (Hornbruch et al., 2005; Qian et al.,
Fig. 1. Hindbrain nTS precursors derive from a Mash1-positive domain of the neuroepithelium. Combined in situ hybridization with Mash1 (dark blue) and immunohistochemistry with a Phox2b antibody (orange) on transverse sections through r7 at E10.5 (A) and E11.5 (B). Only the right side of the sections is shown. Right panels are close-ups of the respective framed areas. The dorsalmost Mash1-positive progenitors give rise to nTS precursors, which transiently coexpress Mash1 and the early post-mitotic marker Phox2b (arrows). Soon after they exit the ventricular zone, nTS precursors migrate ventrally (blue arrowheads) while ventrally generated Phox2b-positive motoneurons (asterisk in panel A) migrate dorsally (red arrowheads).
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
neural tube (Hornbruch et al., 2005). However, the ectopic Tlx3+ neurons were not further characterized and could have been dl5 interneurons that are also Mash1-dependent (Helms et al., 2005 and this report). To pin down the role of Mash1 in nTS formation, we examined the nTS of Mash1 knock-out embryos (Guillemot et al., 1993). nTS formation depends on Mash1 At perinatal stages, the nTS and associated area postrema of Mash1 mutants were markedly atrophic (Fig. 2A). To trace the
69
origin of this defect, we monitored the formation of the nTS from its inception. In E10.5 Mash1 null mutants, no cell expressing Tlx3, Phox2b, or Lmx1b was detected in the dorsal part of the hindbrain in contrast to the wild type (Figs. 2B–D). The more ventral stripe of Lmx1b and Tlx3 expression was also abrogated, in agreement with the recent demonstration that dl5 interneurons depend on Mash1 (Helms et al., 2005). In contrast, Phox2b expression was intact in branchial and visceral motoneuronal (BM/VM) precursors whose initial differentiation is Mash1-independent (Pattyn et al., 2000, 2004). A subtle delay in their dorsal migration was detected (arrowhead in
Fig. 2. The development of the nTS and associated area postrema is affected in Mash1 mutant mice. Transverse sections through the caudal hindbrain of wild-type (left panels) and Mash1 mutant (right panels) mice at E18.5 (A), E10.5 (B–D), E11.5 (E–G), and E12.5 (H), stained as indicated. (A) View of the dorsal vagal complex at E18.5 stained with Phox2b antibody, showing that the nTS and the area postrema (ap) are atrophic in Mash1 mutants. (B–D) At E10.5, nTS precursors (nTSp) are not detected in Mash1 knock-out mice, as shown by the lack of Phox2b (B), Tlx3 (C) and Lmx1b (D) expression. Tlx3 and Lmx1b also reveal the absence of hindbrain dl5 neurons at this stage (C, D). In contrast, the ventrally born Phox2b+ visceral motoneurons (vMN) are present, although slightly delayed in their dorsal migration (arrowheads in panel B). (E–G) At E11.5 a few nTS precursors (nTSp) are detectable in Mash1−/− embryos, but no dl5 neuron (arrowhead in panels F and G). (H) At E12.5, an increasing amount of nTS neurons expressing Phox2b and Lmx1b (yellow cells) are present in Mash1 mutants (albeit in greatly reduced numbers compared to control animals) and undergo a normal ventral migration. Note that the Lmx1b+/Phox2b− cells born at that stage from the Mash1+ domain of the alar plate (red cells) are also produced in the mutants (their apparent increase relative to wild type is not significant) (white arrowheads).
70
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
Fig. 2B; see also Figs. 2E, G) in agreement with recent data showing the role of Math3 and Mash1 in the lateral migration of facial and trigeminal motoneurons (Ohsawa et al., 2005). One day later, at E11.5, when a large population of nTS precursors were already born in the wild type and had undergone part of their ventral migration, the first nTS precursors expressing Phox2b, Tlx3, and Lmx1b were detected in the Mash1 mutants (Figs. 2E–G). In contrast, dl5 interneurons were still completely absent in the mutants, while BM/VM precursors had reached their normal site of aggregation in the nucleus ambiguus and
dorsal motor nucleus of the vagal nerve (Fig. 2G). At E12.5, when the generation of nTS precursors is about to end, their number was conspicuously reduced in the mutants (Fig. 2H). In order to quantify the dynamics of nTS formation in wildtype and Mash1 mutants, we injected BrdU at E9.5, E10.5, E11.5, or E12.5 and counted newborn Phox2b-positive neurons 1 day after injection. This procedure confirmed a 24-h delay in the onset of nTS precursor appearance in the mutants compared to wild type (Figs. 3A, B). Afterwards, the accumulation of Phox2b-positive cells remained slower in the mutants than in
Fig. 3. Mash1 sets the timing of nTS precursor accumulation. Transverse sections through r7 at E10.5, E11.5, E12.5, and E13.5 of wild-type (A) and Mash1−/− (B) embryos, 24 h after injection of BrdU, immunostained for BrdU (in green) and Phox2b (in red). Pictures show the dorsal part of the neural tube where nTS neurons arise. Only the right side is shown, the ventricular zone is on the left, the mantle layer on the right. For each experiment, double-positive cells (in yellow) represent postmitotic nTS precursors born from a cell cycle whose S phase soon followed BrdU injection. For each stage, the number of BrdU+/Phox2b+ cells 24 h after injection (h.a.i) of BrdU is plotted on the graph in C (controls in black, Mash1 mutants in grey). In control embryos (A, C), the generation of nTS post-mitotic precursor starts at around E9.5, reaches a plateau between E10.5 and E12, and stops at around E13. In Mash1−/− embryos (B, C), the generation of nTS post-mitotic precursor is delayed by one day, and starts at around E10.5. Between E10.5 and E12.5 nTS neurons accumulate at a lower pace in the mutants than in control embryos.
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
the wild type and both abated synchronously at E12.5 (Figs. 3A, B). Thus, the generation of nTS precursors lasted 4 days and peaked between E11.5 and E12.5 in wild-type embryos, in agreement with earlier estimations (Taber Pierce, 1973), while in Mash1 mutants it was shortened by one day due to a delayed onset and never reached the same levels (Fig. 3C). The residual production of nTS precursors in Mash1 mutants suggests that another gene can partially compensate for Mash1. At the dorso-ventral level where nTS precursors arise, Ngn2 (but not Ngn1, not shown) is expressed in scattered neuroepithelial cells and in a dense population of cells in a zone intermediate between the neuroepithelium and the mantle layer (Fig. 4A). Some of the Ngn2+ cells also express Phox2b (arrows in Fig. 4A) and are therefore post-mitotic (Pattyn et al., 1997). This pattern of Ngn2 expression is similar to that described in the alar plate of the spinal cord (Helms et al., 2005). Examination of Phox2b+ nTS precursors in Ngn2 knock-out at E12 did not reveal any obvious phenotype, while the double Mash1/Ngn2 knock-out displayed a phenotype indistinguishable from that of the simple Mash1 knock-out (Fig. 4B). Hence,
71
Ngn2 is not in charge of the formation of the residual nTS neurons in Mash1 mutants. Nature of the delay in nTS formation The delay in the appearance of Phox2b+ precursors in the dorsal hindbrain of the mutants could be due to a delay in the production of post-mitotic cells, i.e., reflect the proneural function of Mash1. To explore this possibility, we supplied proneural activity in the nTS region of Mash1 null embryos by using Mash1KINgn2 animals in which the Ngn2 coding sequence replaces that of Mash1 in the Mash1 locus. This configuration is known to restore proneural activity at all sites examined so far (Parras et al., 2002; Pattyn et al., 2004). Mash1KINgn2 embryos displayed a delay in nTS formation indistinguishable from Mash1 null mutants (Fig. 4C). These data argue that a delayed production of post-mitotic cells cannot be the sole explanation for the defect in nTS formation in Mash1 mutants, and that the specification of nTS precursors must itself be delayed.
Fig. 4. Ngn2 cannot compensate for the absence of Mash1 during nTS formation. (A) Expression pattern of Ngn2. Combined in situ hybridization with Ngn2 (dark blue) and immunochemistry for Phox2b (orange) on transverse sections through r7 of a wild-type embryo at E10.5. Only the right side is shown. The right panel is a close up of the framed area. Ngn2 is widely expressed in the ventricular zone of the neural tube throughout the dorso-ventral axis. At the level where nTS precursors arise (framed area), Ngn2 is expressed in the ventricular zone (bracket) in fewer cells than Mash1 (compare with Fig. 1A). In the mantle layer, however, it is largely coexpressed with the post-mitotic marker Phox2b (arrows) and is switched off once the cells have started their ventral migration (arrowhead). (B) Ngn2 is not responsible for the late residual generation of nTS neurons. Transverse sections through r7 of E12 wild-type, Ngn2−/− and Mash1−/−; Ngn2−/− embryos as indicated, stained for Phox2b. No obvious phenotype is observed in Ngn2 mutants, and the phenotype of Mash1;Ngn2 double mutants is similar to that of Mash1 single mutants (compare with Fig. 2). (C) Ngn2 cannot replace Mash1 during the early generation of nTS neurons. Transverse sections through r7 of wild-type and Mash1KINgn2 at E10.5 and E12, as indicated, costained for Lmx1b (in red) and Phox2b (in green). Mash1KINgn2 animals exhibit the same delay of nTS neuron production as Mash1 mutants. Note that the production of dI5 neurons is not rescued either (arrowhead).
72
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
A possibility is that, in Mash1 mutants, each nTS precursor, although born on schedule, is slowed down in its post-mitotic differentiation. We injected BrdU at E9.5 (i.e., when the first nTS precursors are normally born) and assessed Phox2b/BrdU double labeling at E11.5 (i.e., when Phox2b first appears in the mutants, 1 day later than in wild types). No double labeling was detected in the mutants, in contrast to the wild type (Fig. 5). Thus, the first cells that end up expressing Phox2b are born later in the mutant than in wild type, but then switch on Phox2b on schedule, relative to their birth date. Instead, the delay in nTS specification could occur upstream of cell cycle exit. To test this idea, we monitored the bHLH gene Olig3, a marker coexpressed with Mash1 in the cycling progenitors of the nTS (not shown). There was no alteration in Olig3 expression from E9.5 to E10.5 in Mash1 nulls (not shown) (see Discussion). Delay in neuronal differentiation in autonomic ganglia of Mash1 null embryos Another case of an incomplete differentiation block in Mash1 mutants is represented by sympathetic ganglia. Previous reports mention an incompletely penetrant phenotype with agenesis of the superior cervical ganglion contrasting with mere atrophy in the caudal chain (Guillemot et al., 1993; Hirsch et al., 1998). We reexamined the timing of sympathoblast differentiation in Mash1 nulls in more detail. At E10.5, at thoracic levels, sympathoblasts aggregate around the dorsal aorta in both wildtype and mutant animals, albeit in a somewhat looser arrangement in the mutants (Fig. 6A). As previously shown, Phox2b expression was intact, and that of Ret reduced in the mutant ganglion anlagen (Guillemot et al., 1993; Hirsch et al.,
Fig. 5. Delayed birth of Mash1-null nTS precursors. Transverse sections through r7 of control and Mash1 mutants at E11.5 after injection of BrdU at E9.5, costained for BrdU (in green) and Phox2b (in red). In contrast to the control (yellow cells in the left panel), no double-positive cells are detected in Mash1 mutants (right panel). See text for details.
1998) (Figs. 6A, D). None of the other differentiation markers normally switched on at E10.5 could be detected, be they subtype-specific (Phox2a, dHAND, Gata3, and Dbh) or panneural (βIII-tubulin, SCG10) (Figs. 6A, D). One day later, however, at E11.5, the expression of Ret, as well as that of dHAND was rescued, but the number of expressing cells was reduced. Phox2a and Gata3 were now weakly detected, while the pan-neural and subtype specific genes βIII-tubulin, SCG10, and Dbh were still silent (Figs. 6B, D). Finally, at E12.5, at levels where Phox2b+ ganglionic cells could still be detected (i.e., caudal to the cervical ganglion), the full sympathoadrenergic phenotype was expressed (Figs. 6C, D). A similar dynamic was observed at lumbar levels. Discussion While some transcription factors are absolutely required for the differentiation of neuronal types, others are only partially so, a phenomenon usually subsumed under the concept of “partial redundancy”. The same transcription factor can exemplify one or the other situation depending on the neuronal type examined. For example Mash1 is absolutely required for the differentiation of central serotonergic (Pattyn et al., 2004) or central noradrenergic (Hirsch et al., 1998) neurons but only partially so for that of ventral telencephalic neurons (Casarosa et al., 1999). Here, we demonstrate a complex pattern of requirement for Mash1 in the generation of the nTS. It can be deconstructed into two phases: an early phase lasting about 24 h during which Mash1 is absolutely required for the generation of nTS precursors and a late phase during which nTS precursors accumulate in the absence of Mash1, albeit at a reduced pace. We document a similar phenomenon in sympathetic ganglia, where differentiation of ganglionic cells fails in the mutants for a full day after its onset in wild-type embryos, then proceeds towards the full-blown sympathoadrenal phenotype, but at a slower pace. The latter data argue against the existence of Mash1-independent and Mash1dependent “subprograms” of sympathoadrenal differentiation (Sommer et al., 1995), in favor of a single differentiation program whose every step is Mash1-dependent at first and Mash1independent later. A similar phenomenon had been previously documented in retinal explants from Mash1 mutants where the appearance of rods, horizontal, and bipolar cells is not abrogated but delayed by several days (Tomita et al., 1996). Two mechanisms can be proposed to account for the early block of nTS and sympathetic neuron differentiation. According to a “modular” mechanism, the inactivation of Mash1 unveils a chronological heterogeneity in the genetic control of these structures: from E9.5 to E10.5, Mash1 would be the only gene capable of directing nTS and sympathetic neuron identity, respectively, while later on, another gene would be coexpressed, with the same capacity. In this second phase, Mash1 would only increase the efficiency of neuronal differentiation. Such a bimodal ontogeny is observed in dorsal root ganglia, involving first Ngn2 then Ngn1 (Ma et al., 1999). In the nTS, Ngn2, although coexpressed with Mash1 is not that putative second gene since its mutation has no phenotype, whether on a wild-
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
73
Fig. 6. Delayed neuronal differentiation in autonomic ganglia of Mash1−/− embryos. (A–D) Transverse sections through the thoracic sympathetic chain of wild-type (left panels) and mutant (right panels) embryos at E10.5 (A), E11.5 (B) and E12.5 (C), stained as indicated. Panel D shows a summary of the timing of expression of sympathetic neuron markers in wild-type and Mash1−/− embryos. The onset of all neuronal differentiation markers is delayed in the mutants-except for Phox2b, as previously reported (Hirsch et al., 1998). See text for details.
type or a Mash1 null background. It is conceivable that Olig3, being expressed in the hindbrain dl3 domain as it is in the spinal dl3 domain (Muller et al., 2005), would play such a role. In sympathetic ganglionic cells, a candidate is the bHLH factor dHand, implicated in sympathoadrenal differentiation by gain of function experiments both in vitro (Howard et al., 1999) and in vivo (Howard et al., 2000) and expressed after Mash1 (Howard et al., 2000) (but see below). The second mechanism posits that the initial block in differentiation is the result of a quantitative role of Mash1 combined with a threshold effect. For example, differentiation would occur at a precise level of accumulation of one or several transcription factors and that level would be shifted (or
the accumulation slowed down) in the Mash1 mutants. The proneural role of bHLH proteins is classically understood in such dose-dependent terms: cells are thought to commit to the neural differentiation pathway according to their level of proneural gene expression (Bertrand et al., 2002; Gibert and Simpson, 2003). Thus, if a second proneural gene was coexpressed with Mash1 from the start, concurring to the overall proneural activity, the attainment of a critical level in any cell would be delayed in the Mash1 mutants, resulting in the initial block. Again, the coexpressed bHLH factor Olig3 could underlie the residual proneural activity of Mash1 mutants. However, replacement of Mash1 by Ngn2, which presumably restores proneural function, rescues neither the
74
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75
initial block nor the subsequent lowered rate of nTS precursor accumulation, suggesting that the production of post-mitotic precursors is not the only limiting step, if it is at all, in Mash1 knock-outs. Rather, our data argue that the neuronal differentiation aspect of Mash1 function is likely to be itself of a quantitative nature. This is best illustrated in sympathetic ganglia, where differentiation proceeds independently of cell cycle exit (DiCiccoBloom et al., 1990; Rohrer and Thoenen, 1987; Rothman et al., 1980): we document six genes (dHAND, Phox2a, Gata3, Tubb3, SCG10, and Dbh) whose onsets of expression, very precocious and roughly synchronous in the wild type, are not only delayed but staggered in the mutants. The onset of dHAND expression is delayed by one day and followed, a few hours later, by those of Phox2a and Gata3, themselves followed by those of effector genes such as Dbh, βIII-tubulin, or SCG10. The “modular” mechanism outlined earlier is improbable in this case, since it requires that each group of synchronously expressed genes would be controlled by yet another transcription factor with Mash1-like properties. More likely, the expression of every single marker is enhanced by Mash1 and the fast kinetics with which they normally reach the threshold for detectable expression—not resolved in this study, but see Ernsberger et al. (2000) and, for chick embryos Tsarovina et al. (2004)—are slowed down by Mash1 inactivation. This is further suggested by the progressive increase in the level of Ret expression and in the number of cells expressing Gata3 and Phox2a in the mutants. In the nTS, although the onset of expression of Phox2b is not delayed with respect to cell cycle exit in those cells that go on to express it, the competence of nTS progenitors to produce such cells is delayed. A molecular underpinning of this competence, expressed in progenitors, remains to be identified. Whatever the explanation for the initial block in neuronal differentiation and its subsequent relief, Mash1, rather than an indispensable part of the gene expression machinery, should be seen, at least in these structures, as a mere accelerator for the onset of gene expression. This role can explain, in the mutants, the atrophy of structures such as the nTS or the caudal sympathetic chain (Hirsch et al., 1998) (and this paper) or can be completely compensated for as in rods and horizontal cells of the retina (Tomita et al., 1996). However, it could also underlie some all-or-none effects, such as the lack of locus coeruleus (Hirsch et al., 1998) or of the superior cervical ganglion, would the delay in differentiation and the window of competence of a given cell population completely overlap. Acknowledgments We thank C. Birchmeier, Y-S. Dai, D. Engel, T.M. Jessel, Q. Ma and C.W. Ragsdale for antibodies and probes. This work was supported by the Centre National de la Recherche Scientifique and grants from the Association pour la Recherche sur le Cancer (ARC), the Fondation Franco-Norvégienne (FFN), the EC (QLG2-CT-2001-01467) and the French Ministry of Research (ACI 2002).
References Akagi, T., Inoue, T., Miyoshi, G., Bessho, Y., Takahashi, M., Lee, J.E., Guillemot, F., Kageyama, R., 2004. Requirement of multiple basic helix– loop–helix genes for retinal neuronal subtype specification. J. Biol. Chem. 279, 28492–28498. Bertrand, N., Castro, D.S., Guillemot, F., 2002. Proneural genes and the specification of neural cell types. Nat. Rev., Neurosci. 3, 517–530. Blaugrund, E., Pham, T.D., Tennyson, V.M., Lo, L., Sommer, L., Anderson, D. J., Gershon, M.D., 1996. Distinct subpopulations of enteric neuronal progenitors defined by time of development, sympathoadrenal lineage markers and Mash-1-dependence. Development 122, 309–320. Blessing, W.W., 1997. Anatomy of the lower brainstem. In: The Lower Brainstem and Bodily Homeostasis. Oxford Univ. Press, New York London, pp. 29–99. Casarosa, S., Fode, C., Guillemot, F., 1999. Mash1 regulates neurogenesis in the ventral telencephalon. Development 126, 525–534. Cau, E., Gradwohl, G., Fode, C., Guillemot, F., 1997. Mash1 activates a cascade of bHLH regulators in olfactory neuron progenitors. Development 124, 1611–1621. Dauger, S., Pattyn, A., Lofaso, F., Gaultier, C., Goridis, C., Gallego, J., Brunet, J.-F., 2003. Phox2b controls the development of peripheral chemoreceptors and afferent visceral pathways. Development 130, 6635–6642. DiCicco-Bloom, E., Townes-Anderson, E., Black, I.B., 1990. Neuroblast mitosis in dissociated culture: regulation and relationship to differentiation. J. Cell Biol. 110, 2073–2086. Ernsberger, U., Reissmann, E., Mason, I., Rohrer, H., 2000. The expression of dopamine beta-hydroxylase, tyrosine hydroxylase, and Phox2 transcription factors in sympathetic neurons: evidence for common regulation during noradrenergic induction and diverging regulation later in development. Mech. Dev. 92, 169–177. Filippi, A., Tiso, N., Deflorian, G., Zecchin, E., Bortolussi, M., Argenton, F., 2005. The basic helix–loop–helix olig3 establishes the neural plate boundary of the trunk and is necessary for development of the dorsal spinal cord. Proc. Natl. Acad. Sci. U. S. A. 102, 4377–4382. Fode, C., Gradwohl, G., Morin, X., Dierich, A., LeMeur, M., Goridis, C., Guillemot, F., 1998. The bHLH protein NEUROGENIN 2 is a determination factor for epibranchial placode-derived sensory neurons. Neuron 20, 483–494. Fode, C., Ma, Q., Casarosa, S., Ang, S.L., Anderson, D.J., Guillemot, F., 2000. A role for neural determination genes in specifying the dorsoventral identity of telencephalic neurons. Genes Dev. 14, 67–80. Gibert, J.M., Simpson, P., 2003. Evolution of cis-regulation of the proneural genes. Int. J. Dev. Biol. 47, 643–651. Gowan, K., Helms, A.W., Hunsaker, T.L., Collisson, T., Ebert, P.J., Odom, R., Johnson, J.E., 2001. Crossinhibitory activities of Ngn1 and Math1 allow specification of distinct dorsal interneurons. Neuron 31, 219–232. Guillemot, F., Lo, L.-C., Johnson, J.E., Auerbach, A., Anderson, D.J., Joyner, A.L., 1993. Mammalian achaete-scute homolog 1 is required for the early development of olfactory and autonomic neurons. Cell 75, 463–476. Helms, A.W., Johnson, J.E., 2003. Specification of dorsal spinal cord interneurons. Curr. Opin. Neurobiol. 13, 42–49. Helms, A.W., Battiste, J., Henke, R.M., Nakada, Y., Simplicio, N., Guillemot, F., Johnson, J.E., 2005. Sequential roles for Mash1 and Ngn2 in the generation of dorsal spinal cord interneurons. Development 132, 2709–2719. Hirsch, M.R., Tiveron, M.-C., Guillemot, F., Brunet, J.-F., Goridis, C., 1998. Control of noradrenergic differentiation and Phox2a expression by MASH1 in the central and peripheral nervous system. Development 125, 599–608. Hornbruch, A., Ma, G., Ballermann, M.A., Tumova, K., Liu, D., Cairine Logan, C., 2005. A BMP-mediated transcriptional cascade involving Cash1 and Tlx-3 specifies first-order relay sensory neurons in the developing hindbrain. Mech. Dev. 122, 900–913. Howard, M., Foster, D.N., Cserjesi, P., 1999. Expression of HAND gene products may be sufficient for the differentiation of avian neural crestderived cells into catecholaminergic neurons in culture. Dev. Biol. 215, 62–77.
A. Pattyn et al. / Developmental Biology 295 (2006) 67–75 Howard, M., Stanke, M., Schneider, C., Wu, X., Rohrer, H., 2000. The transcription factor dHAND is a downstream effector of BMPs in sympathetic neuron specification. Development 127, 4073–4081. Kageyama, R., Nakanishi, S., 1997. Helix–loop–helix factors in growth and differentiation of the vertebrate nervous system. Curr. Opin. Genet. Dev. 7, 659–665. Kameda, Y., 2005. Mash1 is required for glomus cell formation in the mouse carotid body. Dev. Biol. 283, 128–139. Lee, S.K., Pfaff, S.L., 2003. Synchronization of neurogenesis and motor neuron specification by direct coupling of bHLH and homeodomain transcription factors. Neuron 38, 731–745. Lee, S.K., Jurata, L.W., Funahashi, J., Ruiz, E.C., Pfaff, S.L., 2004. Analysis of embryonic motoneuron gene regulation: derepression of general activators function in concert with enhancer factors. Development 131, 3295–3306. Li, S., Misra, K., Matise, M.P., Xiang, M., 2005. Foxn4 acts synergistically with Mash1 to specify subtype identity of V2 interneurons in the spinal cord. Proc. Natl. Acad. Sci. U. S. A. 102, 10688–10693. Lo, L., Dormand, E., Greenwood, A., Anderson, D.J., 2002. Comparison of the generic neuronal differentiation and neuron subtype specification functions of mammalian achaete-scute and atonal homologs in cultured neural progenitor cells. Development 129, 1553–1567. Ma, Q., Fode, C., Guillemot, F., Anderson, D.J., 1999. Neurogenin1 and neurogenin2 control two distinct waves of neurogenesis in developing dorsal root ganglia. Genes Dev. 13, 1717–1728. Muller, T., Anlag, K., Wildner, H., Britsch, S., Treier, M., Birchmeier, C., 2005. The bHLH factor Olig3 coordinates the specification of dorsal neurons in the spinal cord. Genes Dev. 19, 733–743. Ohsawa, R., Ohtsuka, T., Kageyama, R., 2005. Mash1 and Math3 are required for development of branchiomotor neurons and maintenance of neural progenitors. J. Neurosci. 25, 5857–5865. Parras, C.M., Schuurmans, C., Scardigli, R., Kim, J., Anderson, D.J., Guillemot, F., 2002. Divergent functions of the proneural genes Mash1 and Ngn2 in the specification of neuronal subtype identity. Genes Dev. 16, 324–338. Parras, C.M., Galli, R., Britz, O., Soares, S., Galichet, C., Battiste, J., Johnson, J.E., Nakafuku, M., Vescovi, A., Guillemot, F., 2004. Mash1 specifies neurons and oligodendrocytes in the postnatal brain. EMBO J. 23, 4495–4505.
75
Pattyn, A., Morin, X., Cremer, H., Goridis, C., Brunet, J.-F., 1997. Expression and interactions of the two closely related homeobox genes Phox2a and Phox2b during neurogenesis. Development 124, 4065–4075. Pattyn, A., Hirsch, M.-R., Goridis, C., Brunet, J.-F., 2000. Control of hindbrain motor neuron differentiation by the homeobox gene Phox2b. Development 127, 1349–1358. Pattyn, A., Simplicio, N., van Doorninck, J.H., Goridis, C., Guillemot, F., Brunet, J.-F., 2004. Mash1/Ascl1 is required for the development of central serotonergic neurons. Nat. Neurosci. 7, 589–595. Qian, Y., Fritzsch, B., Shirasawa, S., Chen, C.L., Choi, Y., Ma, Q., 2001. Formation of brainstem (nor)adrenergic centers and first-order relay visceral sensory neurons is dependent on homeodomain protein Rnx/Tlx3. Genes Dev. 15, 2533–2545. Rohrer, H., Thoenen, H., 1987. Relationship between differentiation and terminal mitosis: sensory and ciliary neurons differentiate after terminal mitosis of precursor cells, whereas sympathetic neurons continue to divide after differentiation. J. Neurosci. 7, 3739–3748. Ross, S.E., Greenberg, M.E., Stiles, C.D., 2003. Basic helix–loop–helix factors in cortical development. Neuron 39, 13–25. Rothman, T.P., Specht, L.A., Gershon, M.D., Joh, T.H., Teitelman, G., Pickel, V.M., Reis, D.J., 1980. Catecholamine biosynthetic enzymes are expressed in replicating cells of the peripheral but not the central nervous system. Proc. Natl. Acad. Sci. U. S. A. 77, 6221–6225. Sommer, L., Shah, N., Mahendra, R., Anderson, D.J., 1995. The cellular function of MASH1 in autonomic neurogenesis. Neuron 15, 1245–1258. Taber Pierce, E., 1973. Time of origin of neurons in the brain stem of the mouse. Prog. Brain Res. 40, 53–65. Tiveron, M.-C., Hirsch, M.-R., Brunet, J.-F., 1996. The expression pattern of the transcription factor Phox2 delineates synaptic pathways of the autonomic nervous system. J. Neurosci. 16, 7649–7660. Tiveron, M.-C., Pattyn, A., Hirsch, M.-R., Brunet, J.F., 2003. Role of Phox2b and Mash1 in the generation of the vestibular efferent nucleus. Dev. Biol. 260, 46–57. Tomita, K., Nakanishi, S., Guillemot, F., Kageyama, R., 1996. Mash1 promotes neuronal differentiation in the retina. Genes Cells 1, 765–774. Tsarovina, K., Pattyn, A., Stubbusch, J., Muller, F., van der Wees, J., Schneider, C., Brunet, J.F., Rohrer, H., 2004. Essential role of Gata transcription factors in sympathetic neuron development. Development 131, 4775–4786.