Feeding and breathing are essential motor functions and rely on the activity of hypoglossal and phrenic motor neurons that innervate the tongue and diaphragm, respectively. Little is known about the genetic programs that control the development of these neuronal subtypes. The transcription factor Tshz1 is strongly and persistently expressed in developing hypoglossal and phrenic motor neurons. We used conditional mutation of Tshz1 in the progenitor zone of motor neurons (Tshz1MNΔ) to show that Tshz1 is essential for survival and function of hypoglossal and phrenic motor neurons. Hypoglossal and phrenic motor neurons are born in correct numbers, but many die between embryonic day 13.5 and 14.5 in Tshz1MNΔ mutant mice. In addition, innervation and electrophysiological properties of phrenic and hypoglossal motor neurons are altered. Severe feeding and breathing problems accompany this developmental deficit. Although motor neuron survival can be rescued by elimination of the pro-apoptotic factor Bax, innervation, feeding and breathing defects persist in Bax−/−; Tshz1MNΔ mutants. We conclude that Tshz1 is an essential transcription factor for the development and physiological function of phrenic and hypoglossal motor neurons.

Breathing and feeding are motor behaviors that are essential for the survival of all terrestrial vertebrates. They rely on complex neuronal networks, including motor neurons that are the ‘executive’ components controlling the activity of specific muscle groups (Moore et al., 2014). The diaphragm muscle is essential for inspiration during breathing and is innervated by motor neurons of the phrenic motor column located in the mid-cervical spinal cord. The tongue has crucial functions in feeding and is innervated by the hypoglossal motor neurons in the lower brainstem. The activity of phrenic and hypoglossal motor neurons has to be tightly coordinated to avoid maladaptive outcomes such as the swallowing of air or the blockage of airways (Moore et al., 2014). Hence, hypoglossal and phrenic motor neurons and the circuitry that coordinates their activity are essential for animal survival. Nevertheless, relatively little is known about the formation of the motor neurons that relay breathing and feeding commands.

All motor neurons that innervate skeletal muscle, i.e. somatic motor neurons, derive from ventral progenitor cells expressing the transcription factor Olig2 and share a common transcriptional program (Isl1/2+, Hb9+; Hb9 is also known as Mnx1) during their early initial development that distinguishes them from other neuronal types (Jessell, 2000). However, at later developmental stages they acquire subtype-specific identities that correlate with the muscle groups they innervate (Dasen et al., 2008; De Marco Garcia and Jessell, 2008; Livet et al., 2002). According to their subtype identity, the ventrally derived motor neurons initiate expression of specific sets of transcription factors and either maintain or downregulate Isl1/2 and Hb9 (Arber et al., 1999; Briscoe et al., 1999; Pfaff et al., 1996). In addition, their cell bodies cluster, forming motor pools in specific positions along the rostro-caudal and medio-lateral axes of the hindbrain and spinal cord (Bonanomi and Pfaff, 2010; Dasen and Jessell, 2009). Compared with motor neuronal subtypes that innervate different limb muscles, little is known about the genetic programs important for the development of hypoglossal and/or phrenic motor pools.

Genes that specifically affect hypoglossal motor neuron development have not yet been identified. Hypoglossal motor neurons locate near the midline of rhombomeres 7/8 of the hindbrain and express Isl1/2 and Hb9. They are organized topographically according to which tongue muscle they innervate (Aldes, 1990, 1995; Krammer et al., 1979; Lewis et al., 1971; Odutola, 1976). Dorsal hypoglossal motor neurons innervate the retrusor muscles (i.e. extrinsic hyoglossus, extrinsic palatoglossus, intrinsic inferior longitudinalis and intrinsic superior longitudinalis) and express Foxp1 (Chen et al., 2016). In contrast, motor neurons located in the ventral half of the hypoglossal nucleus innervate the protrusor muscles (i.e. extrinsic genioglossus, intrinsic verticalis and intrinsic transversus) and express Pou3f1 (Chen et al., 2016). Finally, motor neurons innervating extrinsic and intrinsic muscles segregate in the lateral and medial hypoglossal nucleus, respectively (Aldes, 1995; Chen et al., 2016; Dobbins and Feldman, 1995; McClung and Goldberg, 2002).

More information is available for phrenic motor neurons that locate to the cervical spinal cord (C3-C5). Previous studies showed that mutations causing respiratory distress often affect the function and/or survival of phrenic motor neurons (Turgeon and Meloche, 2009). These mutations can either affect the survival of all motor neurons, for instance Hb9, ErbB3, CLAC-P (Col25a1) or Fzd3 (Hua et al., 2013; Riethmacher et al., 1997; Tanaka et al., 2014; Yang et al., 2001), or specifically phrenic motor neurons, e.g. Hoxa5/c5 genes and Pou3f1 (Bermingham et al., 1996; Machado et al., 2014; Philippidou et al., 2012). Indeed, Pou3f1 (also known as Scip or Oct6) is strongly expressed in phrenic neurons, and its ablation in mice leads to disorganization of the phrenic motor column and early postnatal lethality due to breathing failure within the first hours of life (Bermingham et al., 1996). The Hoxa5/c5 null mutation in motor neurons impairs the expression of phrenic specific markers and the correct localization and survival of phrenic motor neurons, resulting in a complete absence of diaphragm innervation (Philippidou et al., 2012). When phrenic motor neurons are generated from embryonic stem cells, the combined action of Pou3f1 and Hoxa5 is necessary for the acquisition of their identity (Machado et al., 2014).

The transcription factor teashirt zinc finger homeobox 1 (Tshz1) belongs to the evolutionarily conserved family of Teashirt genes that are characterized by the presence of three atypical zinc finger motifs, an N-terminal acidic domain and a binding motif for the co-repressor CtBP (C-terminal binding protein). In addition, mammalian Tshz genes also contain a homeobox domain (Caubit et al., 2000; Fasano et al., 1991; Manfroid et al., 2004; Santos et al., 2010). Tshz1 is strongly expressed in subtypes of developing and mature neurons of the mouse (Caubit et al., 2005; Kuerbitz et al., 2018; Ragancokova et al., 2014). Mutation of Tshz1 in mice (Tshz1NULL) leads to perinatal lethality and severe physiological dysfunctions, such as the inability to feed and respiratory distress, which have been attributed to morphological changes of the oral cavity (Coré et al., 2007). The function of Tshz1 in motor neurons has not been investigated.

Here, we define the role of Tshz1 during development of somatic motor neuronal subtypes. We show that Tshz1 is expressed persistently during development in hypoglossal and phrenic motor neurons, but transiently in other motor neuronal subtypes. We investigated the role of Tshz1 using a conditional Tshz1 loss-of-function mutation introduced with the Olig2Cre allele in ventral motor neuron precursors (Tshz1MNΔ). Tshz1MNΔ mice are born at a normal Mendelian ratio. However, the mutant pups do not feed but instead have air in their stomach and intestine, display respiratory distress and die within the first 10 h of life. Interestingly, two specific motor neuron populations are affected by Tshz1 loss of function: the hypoglossal nucleus and the phrenic motor column. These motor neurons were generated in appropriate numbers but many subsequently died by apoptosis, leading to a reduction in their numbers at birth and reduced muscle innervation. Elimination of the pro-apoptotic gene Bax in Tshz1MNΔ mice rescued motor neuron numbers. However, neither physiological nor innervation deficits were rescued in the Bax−/−; Tshz1MNΔ mice. Thus, Tshz1 loss of function impairs motor neuron function as well as survival. We conclude that Tshz1 is an essential transcription factor and coordinates the survival and function of phrenic and hypoglossal motor neurons.

Tshz1 is expressed in specific motor neuron populations

We examined the expression of Tshz1 in motor neuronal subtypes during development by using heterozygous Tshz1GFP mice in which the second exon of Tshz1 was replaced by sequences encoding GFP (Ragancokova et al., 2014). Analysis of Tshz1-driven GFP expression in the spinal cord and brainstem showed the presence of GFP in several neuronal types, among them motor neurons. Detailed immunohistochemical analysis using anti-GFP and pan-motor neuron antibodies, such as anti-Isl1/2 or anti-choline acetyltransferase (ChAT), demonstrated the presence of GFP in hypoglossal motor neurons (nXII) from embryonic day (E) 12.5 to postnatal day (P) 0.5 (Fig. 1A). GFP was detected in 85% of the Isl1/2+ hypoglossal motor neurons at E12.5, but not in Olig2+ progenitors (Fig. 1A,B). GFP expression persisted and was detected in around two-thirds of hypoglossal motor neurons at E15.5 and P0.5. Subsets of GFP+ hypoglossal motor neurons co-expressed Foxp1 and Pou3f1 (Fig. S1A,B). The persistent expression of Tshz1 in hypoglossal motor neurons was confirmed by in situ hybridization at P0.5 (Fig. S1C). Tshz1 expression was spatially restricted to subdomains of the hypoglossal nucleus at P0.5, a finding that became apparent in a 3D reconstruction of the hypoglossal nucleus (Fig. 1C; see Fig. S5 for examples of sections used for the reconstruction). The position of hypoglossal motor neurons was digitally reconstructed in three dimensions by assigning Cartesian coordinates to each motor neuron identified in consecutive histological sections (see also Materials and Methods). The position of neurons of the adjacent dorsal nucleus of the vagus nerve (nX) was used as a landmark. In the rostral hypoglossal nucleus, GFP-expressing motor neurons were only detected ventro-laterally, whereas in the central part all motor neurons expressed GFP. Caudally, GFP was expressed differentially, with high expression in the dorso-medial portion, whereas other motor neurons displayed low expression (Fig. 1C,D). Subtypes of hypoglossal motor neurons are topographically organized in conserved and stereotyped positions according to the tongue muscles they innervate (Aldes, 1995; McClung and Goldberg, 2002). The distribution of Tshz1-expressing motor neurons indicates that they innervate the extrinsic genioglossus, the palatossus or the hyoglossus retrusor muscles, and the intrinsic inferior and/or superior longitudinalis muscles, but not the intrinsic verticalis and transversus protrusor muscles (Aldes, 1995; Dobbins and Feldman, 1995; McClung and Goldberg, 1999).

In the spinal cord, Tshz1-driven GFP was expressed in subpopulations of Isl1/2+, ChAT+ motor neurons. In particular, we detected GFP expression in the area where the phrenic motor column develops already at E10.5, and GFP expression persisted in the phrenic motor column throughout fetal development (Fig. 1E; Fig. S1B). Phrenic motor neurons innervate the diaphragm and express the transcription factor Pou3f1 in addition to the pan-motor neuron markers Isl1/2 and ChAT (Machado et al., 2014; Philippidou et al., 2012). GFP was expressed in 85% of phrenic motor neurons at E12.5 and P0.5 (Fig. 1E,F). In addition, Tshz1-driven GFP expression was transiently detected in motor neurons in the brachial and lumbar spinal cord (Fig. 1G; Fig. S2B). Low GFP levels were present in some differentiating motor neurons (Olig2+, Isl1/2+), whereas the majority of early postmitotic motor neurons (Olig2, Isl1/2+) expressed GFP (Fig. S2A,B). In the lumbar spinal cord, GFP persisted until E15.5 but was undetectable at P0.5 (Fig. 1G,H; data not shown). Next, we defined identities of brachial and lumbar spinal motor neuronal subtypes that expressed Tshz1 at E12.5 (Fig. S2D-O). GFP was detected in about two-thirds of Foxp1+ neurons in the lateral and medial parts of the lateral motor columns (LMCl and LMCm) at E12.5, but only in about 10% of neurons in the medial motor column (MMC; Isl1/2+, Lhx3+: Fig. S2D-G,L-O). In the thoracic spinal cord, GFP was expressed at a low level in about 5% of preganglionic motor neurons (PGCs; Foxp1+, Isl1/2+) and about 10% of hypaxial motor neurons (HMCs: Isl1/2+, Lhx3; Fig. S2H-K). In summary, persistent high levels of Tshz1-driven GFP were observed in hypoglossal and phrenic motor neurons, and only transient GFP expression in other subsets of motor neurons (Fig. 1I). We did not detect Tshz1-driven GFP in GATA3+ or Chx10 (Vsx2)+ V2 interneurons, or in Sim1+ V3 interneurons at E12.5 (Fig. S3A,B). The related protein Tshz3 is expressed in the nervous system (Caubit et al., 2005, 2010). Tshz3 protein was not present in hypoglossal (Caubit et al., 2010) or phrenic motor neurons at E12.5, but was expressed in cells of the lateral and medial motor columns in the lumbar spinal cord (Fig. S3C,D).

Tshz1 expression in motor neurons is necessary for the survival of newborn mice

This expression pattern suggested a possible role for Tshz1 during specification of distinct motor neuronal subtypes. We conditionally mutated Tshz1 in the ventral motor neuron progenitor domain using Olig2Cre (Olig2Cre; Tshz1Δ/Flox; hereafter called Tshz1MNΔ). It should be noted that Tshz1 is not expressed in V2 or V3 interneurons that are also generated from progenitors recombined by Olig2Cre (see Fig. S3A,B). Tshz1MNΔ mutant animals were born at a normal Mendelian ratio but did not survive beyond P0.5 (Fig. 2A). We noted that Tshz1MNΔ pups were unable to feed and had no milk in their stomach. Dissection of the abdominal cavity revealed that their intestines were filled with air (aerophagia; Fig. 2B). In addition, Tshz1MNΔ pups displayed signs of respiratory distress and cyanosis. These phenotypes are similar, if not identical, to those observed in Tshz1 null mutant mice (Tshz1Δ/Δ, referred as Tshz1NULL; Coré et al., 2007). Tshz1NULL mice display oral cavity malformations, i.e. absence of soft palate and flattened epiglottis. As expected for an Olig2Cre-specific mutation, we did not observe such changes in the soft palate or epiglottis of Tshz1MNΔ mice at P0.5 (Fig. 2C). We conclude that aerophagia and feeding deficits in Tshz1MNΔ pups are caused by a neuronal dysfunction.

Next, we quantified the breathing deficits in Tshz1MNΔ animals at P0.5 using plethysmographic recordings (Fig. 2D). Compared with controls, Tshz1MNΔ mutant mice displayed longer breath duration (Ttot; Fig. S4A) and smaller tidal volume (Vt; Fig. S4B). Therefore, ventilation (VE; Vt/Ttot), representing the amount of air that reaches the lungs per minute, was reduced by half (Fig. 2E). In addition, Tshz1MNΔ mutants spent more time in apnea, which was defined as a period longer than 2 s without a breath (Fig. 2F), although the length of individual apneic periods was similar in Tshz1MNΔ mutants and control mice (Fig. S4D). The coefficient of variation of the breathing frequency (calculated as s.d./mean) was higher in Tshz1MNΔ mutants than in controls, highlighting the irregular breathing of Tshz1MNΔ animals (Fig. 2G). We observed that the expiratory period (defined as the time between the end of inspiration and onset of the next inspiration cycle) was prolonged in Tshz1MNΔ animals (Fig. S4C). During normal breathing, expiration of newborn mice is passive and relies on the relaxation of the rib cage/lungs, indicating that the longer expiratory period might be due to less frequent inspiration events. To exclude expiratory deficits, we also analyzed expiration during vocalization, as vocal bursts depend on active expiration (Hernandez-Miranda et al., 2017). In Tshz1MNΔ mutants, active vocal expirations were normal in strength and frequency (Fig. 2H; Fig. S4E,F). Furthermore, we did not detect breaths that were abnormally short, apneustic-like prolonged inspirations, or the presence of forced expiratory efforts that would suggest impaired inspiratory/expiratory timing as a cause of apneas. Therefore, we conclude that Tshz1MNΔ mutants present with a reduced inspiratory motor drive, which results in shallow breathing, an increased incidence of apneas and rhythm irregularities.

Tshz1 is specifically required for hypoglossal and phrenic motor neuron survival

Previous studies have demonstrated that tongue innervation by hypoglossal nerves is required for suckling and feeding (Fujita et al., 2006; Fukushima et al., 2014). We hypothesized that impaired development of hypoglossal motor neurons might cause the feeding deficit observed in Tshz1MNΔ mice. To assess this, we counted hypoglossal motor neurons at different developmental stages (Fig. 3A-F). Motor neurons are generated in excess and up to 50% undergo apoptosis during a critical period from E13.5 to E14.5 (Dekkers et al., 2013; Yamamoto and Henderson, 1999). Similar numbers of hypoglossal motor neurons were present in Tshz1MNΔ and control animals at E12.5 (Fig. 3A,B). We observed a 20% loss of hypoglossal motor neurons at E14.5 (Fig. 3C), which became more pronounced by P0.5 when the loss amounted to 38% (Fig. 3E). The loss of motor neurons was due to increased apoptosis detected by staining for cleaved-caspase 3 (Fig. 3D). 3D reconstructions showed the hypoglossal nucleus was shortened along the antero-posterior axis in Tshz1GFP/Δ mutants (Fig. 3F). However, the overall nuclear organization was conserved and GFP-expressing motor neurons were located at the same relative positions (Fig. 3G-J; Fig. S5). Altogether, our data demonstrate that in Tshz1MNΔ mice, hypoglossal motor neurons are generated in normal numbers but their survival is impaired.

Because Tshz1MNΔ animals displayed severe breathing deficits, we analyzed phrenic motor neurons, which innervate the diaphragm. Phrenic motor neurons can be distinguished from other motor neurons at the same axial level (C3-C5) by their location between the medial and hypaxial motor columns and by co-expression of Isl1/2 and Pou3f1. The number of phrenic motor neurons was unchanged at E12.5 but reduced at P0.5 in Tshz1MNΔ mice (Fig. 4A,B). Thus, phrenic motor neurons were generated in correct numbers but subsequently died. Genes specifically expressed in phrenic but not in other types of ventral motor neurons have been previously identified (Machado et al., 2014; Philippidou et al., 2012). We analyzed the expression of a subset of these phrenic-specific genes in Tshz1MNΔ mutants at E12.5, i.e. at a stage before motor neuron loss was observed. Expression of Alcam, Pou3f1, Hoxa5 and Hoxc5 was correctly initiated in phrenic motor neurons of Tshz1MNΔ mice, indicating that phrenic motor neurons retained their identity (Fig. S6). Tshz1 was also transiently expressed in other spinal motor neurons. However, in the lateral and medial motor columns, we detected no loss of ChAT+ motor neurons compared with controls in Tshz1MNΔ mice at P0.5 (Fig. 4C).

Tshz1 mutation impairs hypoglossal and phrenic nerve innervation

Next, we investigated the morphology of the hypoglossal and phrenic nerves at P0.5 (Fig. 5). Innervation of the tongue by the hypoglossal nerve was assessed using antibodies against heavy neurofilament protein (NF200) on whole-mount tongue preparations (Fig. 5A). The overall innervation pattern of the hypoglossal nerve was conserved (one ventral and one medial branch that separate inside the tongue), but the nerve was thinner and less branched in Tshz1MNΔ than in control mice. We analyzed the overall projections of medial and lateral branches of the hypoglossal nerve and observed a 25-30% reduction in the projection area of both branches in Tshz1MNΔ mice (Fig. 5B; Fig. S7). We also analyzed excitatory and inhibitory innervation of hypoglossal motor neurons using antibodies directed against Vglut2 (Slc17a6) and Vgat (Slc32a1) proteins, which mark presynaptic terminals (Fig. 5C). Quantification of the density of Vglut2 and Vgat punctae revealed an increase of both inhibitory and excitatory synapses onto hypoglossal motor neurons in the Tshz1MNΔ compared with control mice (Fig. 5D,E). This indicates that premotor neurons increase their connectivity to those motor neurons that are spared from cell death in the Tshz1MNΔ animals. We conclude that innervation deficits correlate with the impaired feeding and aerophagia of Tshz1MNΔ mice.

Phrenic nerve innervation of the diaphragm was visualized using either membrane-bound GFP expressed upon recombination with Olig2Cre (Fig. 5F; Tau-LSL-mGFP; Hippenmeyer et al., 2005) or a Hb9GFP transgene expressed in ventral motor neurons (Fig. 5G,H; Wichterle et al., 2002). Whole-mount preparations showed that the diaphragm of Tshz1MNΔ mice was innervated and that the overall innervation pattern appeared to be correct (Fig. 5F). However, phrenic nerves were thinner in Tshz1MNΔ mice (Fig. 5F). Neuromuscular junctions assessed by α-bungarotoxin labeling were correctly formed in Tshz1MNΔ mutants (Fig. 5G). Finally, innervation of thoracic muscle was unaffected (Fig. 5H). We propose that a deficit of the phrenic motor column accounts for the impaired breathing of Tshz1MNΔ mice.

Tshz1 does not affect respiratory rhythm generation but impairs the pattern of motor bursts

Breathing and swallowing are highly automated behaviors driven by central pattern generators located in the hindbrain (Del Negro et al., 2018). Phrenic and hypoglossal nerves burst synchronously but not simultaneously during the inspiratory phase of breathing, i.e. hypoglossal activity precedes phrenic activity. Hypoglossal activity results in tongue muscle contraction and expansion of the upper airways, which decrease airway resistance in anticipation of the inspiratory effort (Ghali, 2015). This synchronization is also necessary for feeding in order to avoid maladaptive outcomes such as the swallowing of air or airway blockage (Moore et al., 2014). We tested whether respiratory rhythm generation and coordinated phrenic/hypoglossal motor activity were affected in Tshz1MNΔ mutants. For this, we used E18.5 isolated brainstem-spinal cord preparations, which retain a fictive breathing behavior, to record spontaneous hypoglossal and phrenic nerve activity in vitro (Fig. 6A,B). In contrast to the reduction of the frequency of breaths in Tshz1MNΔ animals (in vivo plethysmographic recordings), burst frequencies of the phrenic nerve and regularity of the rhythm (assessed with the coefficient of variation) were similar in Tshz1MNΔ and control mice (Fig. 6C,D). Further, activities of the phrenic and hypoglossal nerves remained synchronous in preparations of Tshz1MNΔ animals (Fig. 6B). Careful examination of hypoglossal and phrenic bursts revealed that the total duration of hypoglossal and phrenic motor bursts was increased in Tshz1MNΔ animals (Fig. 6E), due to a lengthening of the burst decay (Fig. 6F). In addition, the temporal coordination between the onsets of hypoglossal and phrenic motor activity was altered in vitro, as hypoglossal nerves fired earlier in Tshz1MNΔ mutants than in control animals (Fig. 6G,H). Altogether, these data indicate that the irregular breathing of Tshz1MNΔ mutants in vivo is not explained by deficits in rhythm generation. However, changes in the coordination of hypoglossal and phrenic activity might underlie or contribute to the observed deficits in breathing and feeding in Tshz1MNΔ mutants.

Preventing apoptosis does not rescue the impaired survival of Tshz1MNΔ mutants

Mutation of the pro-apoptotic gene Bax is commonly used to rescue apoptosis in neurons, but the elimination of Bax cannot rescue additional functional deficits (Hua et al., 2013; Philippidou et al., 2012; Tanaka et al., 2014; White et al., 1998). We tested whether the mutation of Bax could rescue the cell death and physiological phenotypes observed in Tshz1MNΔ animals (Bax−/−; Tshz1MNΔ). The mutation of Bax was sufficient to rescue the loss of motor neurons, and no difference in the composition of the hypoglossal nucleus or phrenic motor column was observed in Bax−/−; Tshz1MNΔ and Bax−/− animals at P0.5 (Fig. 7A,B). Interestingly, preventing apoptosis did not rescue deficits in hypoglossal innervation or the impaired survival of Tshz1MNΔ pups (Fig. 7C-E). In particular, Bax−/−; Tshz1MNΔ pups were still unable to feed although they did not display air in their intestine (Fig. 7E). Plethysmographic analysis demonstrated that breathing parameters improved but were not fully restored in Bax−/−; Tshz1MNΔ animals (Fig. 7F-H). Specifically, the tidal volume was rescued, but breath duration did not recover, resulting in an incomplete rescue of ventilation. Thus, further motor neuron dysfunction(s) in addition to cell death cause the respiratory distress and inability to feed.

A common transcriptional program (Isl1/2+, Hb9+) defines the generic identity of ventral motor neurons. These neurons subsequently diversify and express subtype-specific genes that correlate with the muscle they innervate. We show here that Tshz1, a zinc-finger and homeodomain factor, is persistently expressed in developing hypoglossal and phrenic motor neurons and is essential for their survival and function. We investigated the role of Tshz1 in ventral motor neurons using conditional mutagenesis. In such mice, hypoglossal and phrenic motor neurons are born in correct numbers, but many die between E13.5 and E14.5. Severe feeding and breathing problems accompany this developmental deficit. The deficiency in neuronal survival can be rescued by the elimination of the pro-apoptotic factor Bax, but this does not suffice to fully rescue the dysfunction of phrenic and hypoglossal motor neurons. Thus, even when the motor neuron numbers are rescued, Bax−/−; Tshz1MNΔ mutants cannot feed, and display deficits in breathing and reduced innervation of hypoglossal nerve targets. Thus, Tshz1 coordinates the survival and function of phrenic and hypoglossal motor neurons.

Genes that control hypoglossal and phrenic motor neuron formation

We show here that Tshz1 specifically functions during development of phrenic and hypoglossal motor neurons. However, Tshz1 is also expressed in other motor neuron populations, such as those contained within the lateral motor column, but these neurons express Tshz1 transiently and their generation is not affected by Tshz1 mutation. Thus, the action of Tshz1 is axially restricted to the caudal hindbrain and cervical spinal cord. The anterior-posterior identity of motor neurons depends on Hox genes expressed in a spatially restricted manner in the hindbrain and spinal cord. At a given axial level, Hox activity is refined in motor neurons by subtype-specific co-factors such as Foxp1 in the lateral motor columns (Dasen and Jessell, 2009). In Drosophila, the protein Tsh, homolog of Tshz1, has been shown to directly interact with the Hox protein Sex combs reduced (Scr) to promote trunk structure development and repress head formation (de Zulueta et al., 1994; Fasano et al., 1991; Röder et al., 1992). This raises the possibility that Hox genes cooperate with Tshz1 during the specification of hypoglossal and phrenic motor neurons.

In mammals, three Teashirt homologs (Tshz1-3) exist and any one of them can replace Tsh in Drosophila (Manfroid et al., 2004). Interestingly, Tshz1 and Tshz3 are co-expressed in the lateral motor column but not in hypoglossal or phrenic motor neurons (Fig. S3C; Caubit et al., 2010). Thus, the specificity of Tshz1 function in phrenic and hypoglossal motor neurons could reflect the absence of redundantly acting Teashirt factors in these motor neuronal subtypes. It is interesting to note that Tshz3 was previously shown to control survival of motor neurons in the nucleus ambiguus, i.e. a pool of brainstem neurons that innervate the upper airways (Caubit et al., 2008). Thus, both Tshz1 and Tshz3 can take over essential roles in development of motor neuronal subtypes.

Physiological consequences of Tshz1 mutations

Hypoglossal and phrenic motor pools control movement of the tongue and the diaphragm, respectively, and are specifically affected by Tshz1 mutations. In particular, Tshz1MNΔ mutants have feeding problems that we assign to a hypoglossal dysfunction. Suckling behavior and milk intake depend on the contraction of tongue muscles and surgical impairment of hypoglossal nerves prevents feeding in rats (Fujita et al., 2006; Fukushima et al., 2014; Lowe, 1980; McLoon and Andrade, 2013). The medial and lateral branches of the hypoglossal nerve innervate protrusive and retrusive tongue muscles, respectively, and the projection pattern of both branches was impaired in Tshz1MNΔ mutants. Tongue protruder muscles are required to keep the airways open (Edwards and White, 2011; Fregosi, 2011; McLoon and Andrade, 2013). In contrast, tongue retrusors are needed during swallowing and are inactive during normal breathing (Fujita et al., 2006; Fukushima et al., 2014; McLoon and Andrade, 2013). The observed changes in innervation of hypoglossal nerve targets might thus affect breathing in addition to feeding behavior.

Tshz1MNΔ mice also show long periods of apnea and breathe more shallowly. In addition, Tshz1MNΔ mice display an unusual phenotype, an accumulation of air in the intestine. Aerophagia in newborn mice has been previously observed in Sim2, Lgr5 and Tshz1NULL mutants (Coré et al., 2007; Morita et al., 2004; Shamblott et al., 2002). Aerophagia in Sim2 and Tshz1NULL mutant mice was assigned to cleft palate and other malformations of the oral cavity (Coré et al., 2007; Shamblott et al., 2002). In Lgr5 mutant mice, it was attributed to an abnormal fusion of the tongue to the floor of the oral cavity (Morita et al., 2004). Our analysis relied on a conditional mutation of Tshz1 in ventral motor neurons, and thus unambiguously shows that a neuronal dysfunction causes aerophagia.

Swallowing interrupts the respiratory cycle, resulting in an apneic period during which the larynx is closed, and the bolus moves through the pharynx into the esophagus (McFarland and Lund, 1993; Paydarfar et al., 1995). A disruption of this process might cause aerophagia. Aerophagia is also observed in humans where it is frequently assigned to abdomino-phrenic dyssynergia, a condition characterized by a deficit in abdominal and diaphragm muscle coordination (Dike et al., 2017; Ercoli et al., 2018; Villoria et al., 2011). We, therefore, propose that the phrenic dysfunction is responsible for breathing deficits and also contributes to the aerophagia.

The specificity of Tshz1 function in a small subset of motor neurons indicates that this factor determines a program unique for phrenic and hypoglossal neurons rather than a generic motor neuron transcriptional program. Whereas the innervation pattern of the tongue was disrupted, diaphragm muscle innervation was largely preserved in Tshz1MNΔ mice despite the severe consequences of the mutation on breathing behavior. Changes in premotor input that resulted in uncoordinated movement have been defined in locomotor circuits (Baek et al., 2017; Hinckley et al., 2015; Song et al., 2016). However, the neuronal circuitries that control complex orofacial behaviors such as swallowing and breathing are incompletely characterized and only phrenic premotor neurons but not hypoglossal ones have been defined (Wu et al., 2017). To test whether the overall synaptic input that shapes hypoglossal activity is altered in Tshz1MNΔ mutants, we counted the density of synapses onto hypoglossal motor neurons. We detected an increase in the number of excitatory and inhibitory synapses onto the remaining hypoglossal motor neurons in Tshz1MNΔ mutants. This indicates that the premotor neurons still recognize their correct targets and compensate for the neuronal loss in Tshz1MNΔ mutants by more densely innervating the remaining motor neurons. However, many neuronal cell types synapse onto motor neurons to regulate their activity, and we cannot exclude the possibility that deficits in innervation by specific subtypes contribute to the behavioral deficits in Tshz1MNΔ mutant mice.

Genetic mutation of the pro-apoptotic gene Bax is commonly used to rescue apoptosis in the central nervous system (Hua et al., 2013; Philippidou et al., 2012; Tanaka et al., 2014; White et al., 1998). However, the Bax mutation cannot rescue associated functional deficits. For instance, the Hoxa5/c5 mutation in mice causes a loss of phrenic motor neurons accompanied by a lack of diaphragm innervation, and an additional Bax mutation rescues neuronal numbers but not innervation (Philippidou et al., 2012). We show here that the additional mutation of Bax in a Tshz1MNΔ background rescued hypoglossal and phrenic motor neuron numbers but did not rescue hypoglossal innervation deficits or survival of the mutant animals. Thus, feeding and breathing behaviors are still impaired in Bax−/−; Tshz1MNΔ double mutant mice. We conclude that the deficits caused by the mutation of Tshz1 result not only in a loss of motor neurons, but also in functional deficits such as the changed coordination of phrenic and hypoglossal nerve activity. Thus, the mutation might impact coordinated movements of the tongue and of the diaphragm, thereby impairing the vital ability to coordinate orofacial muscles during respiration and swallowing.

Mouse strains

Mouse strains carrying Olig2Cre (Dessaud et al., 2007), Tshz1Flox, Tshz1Δ and Tshz1GFP (Ragancokova et al., 2014) alleles have been described. Bax mutant (Knudson et al., 1995) and Rosa26-Tomato Ai14Flox (Madisen et al., 2010) strains were obtained from The Jackson Laboratory. All experiments were performed in accordance with the guidelines and policies of the European Union, the Max Delbrück Center for Molecular Medicine and the Institut des Neurosciences Paris-Saclay.

Histology, immunofluorescence, in situ hybridization and quantification

Hematoxylin-Eosin staining was performed on cryosections (12-18 µm) using Weigert's iron Hematoxylin (Sigma-Aldrich) and Eosin Y. Immunofluorescence analyses were performed on cryosections or on whole-mount diaphragm (Lin et al., 2001) using the following primary antibodies: rat anti-GFP (1:1500, Nacalai Tesque, GF090R), mouse anti-Isl1/2 [1:400, Developmental Studies Hybridoma Bank (DSHB), 39.4D5 and 40.2D6], guinea pig anti-Isl1/2 (1:5000, a kind gift from T. Jessell and Susan Brenner-Morton, Columbia-University, New York, USA), rabbit anti-Foxp1 (1:500, Abcam, ab16645), mouse anti-Lhx3 (1:300, DSHB, 67.4E12), goat anti-ChAT (1:300, Millipore, AB144P), rabbit anti-Olig2 (1:500, Millipore, AB9610), rabbit anti-cleaved caspase3 (1:300, Cell Signaling, 5A1E), guinea pig anti-Tshz1 (1:4000; Ragancokova et al., 2014), rabbit anti-RFP (1:1000, Biotrend, 600-401-379), rabbit anti-Pou3f1/Oct6 (1:500, Abcam, ab126746), guinea pig anti-VGLUT2 (1:500, Synaptic Systems, 135404), rabbit anti-VGAT (1:500, Synaptic Systems, 131002), rabbit anti-GATA3 (1:500, Abcam, ab106625) and sheep anti-Chx10 (1:300, Abcam, ab16141) Secondary antibodies coupled either with Alexa Fluor 555, Alexa Fluor 488 or Alexa Fluor 639 were used at 1:500 (Molecular Probes). For whole-mount diaphragm staining, Alexa Fluor 488-conjugated α-bungarotoxin (1:1000, Molecular Probes) was used. Images were acquired on a Zeiss LSM 700 confocal microscope. Chromogenic and immunofluorescent in situ hybridization were performed as described (Wildner et al., 2013). All probes were synthesized using an RNA labeling kit (Roche) as specified by the manufacturer. Probe for Sim1 was a gift from Martyn Goulding (Salk Institute, San Diego, USA). Probe for Tshz1 was amplified from total brain cDNA.

Motor neuronal subtypes were identified according to their location in the hindbrain and spinal cord, and by their expression of specific transcription factors. Hypoglossal motor neurons close to the midline in the caudal hindbrain (rhombomeres 7/8) were identified with Isl1/2 (E12.5) or ChAT (P0.5) antibodies. At E14.5 and P0.5, motor neurons in the entire nucleus were counted; at E12.5, only nine consecutive sections of the rostral half of the nucleus were analyzed. Phrenic motor neurons are located at cervical levels between the medial and lateral motor columns and specifically express Pou3f1. Motor pools were quantified on serial sections, using at least three animals per genotype. Motor neurons were counted unilaterally on every sixth section for characterization of Tshz1 expression, and bilaterally on every fourth section for Tshz1MNΔ mice analyses.

Images used for synapse quantification were acquired on a Leica SP8 microscope with a HC PL APO CS2 63.0×1.30 Glycerol objective. Stacks of nine or ten images were recorded at 330 nm increments. Laser power and settings were identical for all samples in an experiment. The recorded image stacks were deconvolved using Leica Lightning processing in LASX. Automated calculation of perisomatic bouton density was performed offline in MatLab (MathWorks) using the Image Processing Toolbox and custom-written routines. Briefly, image de-noising with edge preservation was performed using Guided Filtering. The GFP channel was then binarized using a locally adaptive threshold. Small processes were pruned by eroding the masked areas, removing small objects, and then dilating the remaining masked areas. A perisomatic sample region was defined within 6 pixels (∼400 nm) of the masked soma. Labeled boutons were binarized using Otsu's methods. Masked boutons that were found to reside within the sample region were counted and the synapse density was calculated as the number of counted boutons divided by the soma perimeter.

3D reconstruction

Cryosections of Tshz1GFP/+ or Tshz1GFP/Δ hindbrains at P0.5 were stained with Isl1/2 and GFP antibodies. Motor neurons of the vagus motor nucleus were used as a landmark. In the hypoglossal nucleus, motor neurons were either Isl1/2+, GFP or Isl1/2+, GFP+. To allocate Cartesian coordinates to the motor neurons, we acquired their positions using the spots function in Imaris software (bitplane). Their xy positions were recalculated using the midline as the y-axis, and the lowest point of the hypoglossal nucleus as the origin. The z-axis coordinate depended on the section: z=0 corresponded to the most rostral section with hypoglossal MNs. Subsequent sections were separated by 64 µm. Coordinates were exported to ‘R’ and the 3D reconstruction was made using a previously described script (Dewitz et al., 2018).

Whole-mount staining

Whole-mount staining of diaphragm was adapted from Lin et al. (2000). Briefly, trunks of P0.5 animals were fixed in 4% paraformaldehyde for 4 h at 4°C. After fixation, diaphragms were dissected, incubated in 0.1 M glycine in PBS, rinsed and permeabilized in 0.5% Triton X-100 in PBS. Diaphragms were blocked for at least 90 min in dilution buffer (500 mM NaCl, 3% bovine serum albumin, 5% goat serum in 0.5% Triton X-100 in PBS), and incubated with primary antibodies overnight at 4°C. After three washes in 0.5% Triton X-100 in PBS, preparations were incubated with secondary antibodies and Alexa Fluor 488-conjugated α-bungarotoxin (1:1000, Molecular Probes) for 2 h. After washes, diaphragms were flat-mounted on slides and pictured. Diaphragm innervation patterns are variable even between controls (Laskowski et al., 1991), which precluded a quantitative assessment.

Whole-mount staining of the tongue was carried out after P0.5 preparations of tongue and lower jaw were fixed for 4 h in 4% paraformaldehyde, bleached in Dent's Bleach (10% H2O2, 13.3% DMSO, 53.3% methanol) for 24 h and re-fixed in Dent's fixative (20% DMSO, 80% methanol) for 24 h. The preparations were stained for 5 days with chicken anti-NF200 antibody in blocking buffer [0.5% of normal goat serum (Invitrogen), 20% DMSO, 75% PBS, 0.025% sodium azide] and then incubated for 2 days with Alexa Fluor 555-coupled secondary antibody in blocking buffer. After washing, samples were dehydrated in methanol and cleared in BABB (1/3 benzyl alcohol. 2/3 benzyl benzoate). Preparations were placed in a homemade glass chamber filled with BABB. Images were acquired as z-stacks using a Zeiss LSM700 confocal microscope. For the analysis of the innervation of hypoglossal targets, we used a z-projection method with maximal intensity to define the area of the projections (ImageJ 1.42). The fraction of the innervation area covered by NF200 signals, i.e. the hypoglossal nerve and its medial and lateral branches, was quantified. A double-blind approach was applied to calculate the signal-occupied area.

Plethysmography

Breathing of pups was recorded 3-6 h after birth as described (Bouvier et al., 2010; Hernandez-Miranda et al., 2017). Briefly, freely moving P0 animals were placed in an airtight plethysmographic chamber of 20 ml volume connected to a differential pressure transducer (Validyne DP 103-14) and a reference chamber. Variation of pressures was sampled at 1 kHz and digitalized using a Labmaster interface. Breathing was monitored for 5 min. Injection of 2.5 µl of air into the chamber with a Hamilton syringe was used as a calibration and allowed the quantification of several breathing parameters. For quantification, apnea-free periods of more than 30 s were used. Breath duration (Ttot), tidal volume (Vt) and apnea duration, defined as ventilation arrest longer than 2 s, were obtained from analyses of plethysmographic traces with Elphy software (developed by Gerard Sadoc, UNIC, CNRS, France, https://www.unic.cnrs-gif.fr/software.html). Tidal volume (Vt) was normalized to the animal weight and used to calculate ventilation (VE=Vt/Ttot in µl/s/g).

Electrophysiology

Electrophysiological recordings of hypoglossal and phrenic nerves were performed as described (Ruffault et al., 2015). Briefly, E18.5 animals were delivered by caesarian section, anesthetized on ice and dissected on ice-cold oxygenated artificial cerebrospinal fluid (aCSF) containing 128 mM NaCl, 8 mM KCl, 1.5 mM CaCl2, 1 mM MgSO4, 24 mM NaHCO3, 0.5 mM Na2HPO4, 30 mM glucose, pH 7.4. Brainstem-spinal cord preparations were obtained after transverse sections through the ponto-medullary region and the C6 spinal cervical segment. Preparations were transferred to the recording chamber (4 ml) and perfused with oxygenated aCSF at 30°C. After 30 min of recovery, hypoglossal and phrenic nerves were recorded using extracellular bipolar suction electrodes (A-M Systems). During the recordings, nerves were perfused with normal aCSF. Raw signals were amplified (1000×) and filtered (0.3-30 kHz) through a high-gain AC amplifier (7P511, Grass Instrument). Signals were then integrated (time constant 50 ms, Neurolog System, Digitimer), digitalized at 10 kHz (Digidata A322A), recorded and analyzed using pClamp9 software (Molecular Devices). To determine the delay between hypoglossal and phrenic activities, we defined the starting point of the burst as 15% of the burst maximum amplitude. Thus, the initial ‘slow’ slope effect was not included in the quantification.

Statistical analyses

All statistical analyses were performed with GraphPad Prism 6 software. Normality of the data was tested with d'Agostino-Pearson omnibus normality test. For comparing two datasets, we used unpaired two-tailed t-test. For multiple datasets, we used ordinary one-way ANOVA with Tukey's multiple comparisons test. Burst durations, coefficient of variation (in vivo) and nerve delays did not follow a Gaussian distribution; therefore, we used Kruskal–Wallis test followed by Dunn's multiple comparisons test.

We thank Carola Griffel and Bettina Brandt for excellent technical assistance, Petra Stallerow for her help with the animal husbandry, as well as Carola Dewitz and Dr Michael Strehle for their help with the 3D reconstructions and R troubleshooting, respectively. We acknowledge Thomas Müller for his advice and technical inputs and for critically reading the manuscript, and thank Elijah Lowenstein for proofreading the manuscript.

Author contributions

Conceptualization: C. Chaimowicz, A.N.G., C.B.; Methodology: C. Chéret, A.W.; Validation: C. Chaimowicz, P.-L.R.; Formal analysis: C. Chaimowicz, P.-L.R., C. Chéret, A.W.; Investigation: C. Chaimowicz, P.-L.R.; Resources: N.Z., G.F.; Writing - original draft: C. Chaimowicz, C.B.; Writing - review & editing: P.-L.R., C. Chéret, N.Z., G.F., A.N.G.; Visualization: C. Chaimowicz, C.B.; Supervision: A.N.G., C.B.; Project administration: A.N.G., C.B.; Funding acquisition: A.N.G., C.B., G.F.

Funding

This work was supported by the Helmholtz Association (C.B.), the Centre National de la Recherche Scientifique (G.F.) and the Agence Nationale de la Recherche (ANR-15-CE16-013-02 to G.F.). Deposited in PMC for immediate release.

Aldes
,
L. D.
(
1990
).
Topographically organized projections from the nucleus subceruleus to the hypoglossal nucleus in the rat: a light and electron microscopic study with complementary axonal transport techniques
.
J. Comp. Neurol.
302
,
643
-
656
.
Aldes
,
L. D.
(
1995
).
Subcompartmental organization of the ventral (protrusor) compartment in the hypoglossal nucleus of the rat
.
J. Comp. Neurol.
353
,
89
-
108
.
Arber
,
S.
,
Han
,
B.
,
Mendelsohn
,
M.
,
Smith
,
M.
,
Jessell
,
T. M.
and
Sockanathan
,
S.
(
1999
).
Requirement for the homeobox gene Hb9 in the consolidation of motor neuron identity
.
Neuron
23
,
659
-
674
.
Baek
,
M.
,
Pivetta
,
C.
,
Liu
,
J.-P.
,
Arber
,
S.
and
Dasen
,
J. S.
(
2017
).
Columnar-intrinsic cues shape premotor input specificity in locomotor circuits
.
Cell Rep.
21
,
867
-
877
.
Bermingham
,
J. R.
,
Scherer
,
S. S.
,
O'Connell
,
S.
,
Arroyo
,
E.
,
Kalla
,
K. A.
,
Powell
,
F. L.
and
Rosenfeld
,
M. G.
(
1996
).
Tst-1/Oct-6/SCIP regulates a unique step in peripheral myelination and is required for normal respiration
.
Genes Dev.
10
,
1751
-
1762
.
Bonanomi
,
D.
and
Pfaff
,
S. L.
(
2010
).
Motor axon pathfinding
.
Cold Spring Harb. Perspect. Biol.
2
,
a001735
.
Bouvier
,
J.
,
Thoby-Brisson
,
M.
,
Renier
,
N.
,
Dubreuil
,
V.
,
Ericson
,
J.
,
Champagnat
,
J.
,
Pierani
,
A.
,
Chédotal
,
A.
and
Fortin
,
G.
(
2010
).
Hindbrain interneurons and axon guidance signaling critical for breathing
.
Nat. Neurosci.
13
,
1066
-
1074
.
Briscoe
,
J.
,
Sussel
,
L.
,
Serup
,
P.
,
Hartigan-O'Connor
,
D.
,
Jessell
,
T. M.
,
Rubenstein
,
J. L. R.
and
Ericson
,
J.
(
1999
).
Homeobox gene Nkx2.2 and specification of neuronal identity by graded Sonic hedgehog signalling
.
Nature
398
,
622
-
627
.
Caubit
,
X.
,
Coré
,
N.
,
Boned
,
A.
,
Kerridge
,
S.
,
Djabali
,
M.
and
Fasano
,
L.
(
2000
).
Vertebrate orthologues of the Drosophila region-specific patterning gene teashirt
.
Mech. Dev.
91
,
445
-
448
.
Caubit
,
X.
,
Tiveron
,
M.-C.
,
Cremer
,
H.
and
Fasano
,
L.
(
2005
).
Expression patterns of the threeTeashirt-related genes define specific boundaries in the developing and postnatal mouse forebrain
.
J. Comp. Neurol.
486
,
76
-
88
.
Caubit
,
X.
,
Lye
,
C. M.
,
Martin
,
E.
,
Coré
,
N.
,
Long
,
D. A.
,
Vola
,
C.
,
Jenkins
,
D.
,
Garratt
,
A. N.
,
Skaer
,
H.
,
Woolf
,
A. S.
, et al. 
(
2008
).
Teashirt 3 is necessary for ureteral smooth muscle differentiation downstream of SHH and BMP4
.
Development
135
,
3301
-
3310
.
Caubit
,
X.
,
Thoby-Brisson
,
M.
,
Voituron
,
N.
,
Filippi
,
P.
,
Bévengut
,
M.
,
Faralli
,
H.
,
Zanella
,
S.
,
Fortin
,
G.
,
Hilaire
,
G.
and
Fasano
,
L.
(
2010
).
Teashirt 3 regulates development of neurons involved in both respiratory rhythm and airflow control
.
J. Neurosci.
30
,
9465
-
9476
.
Chen
,
X.
,
Wang
,
J. W.
,
Salin-Cantegrel
,
A.
,
Dali
,
R.
and
Stifani
,
S.
(
2016
).
Transcriptional regulation of mouse hypoglossal motor neuron somatotopic map formation
.
Brain Struct. Funct.
221
,
4187
-
4202
.
Coré
,
N.
,
Caubit
,
X.
,
Metchat
,
A.
,
Boned
,
A.
,
Djabali
,
M.
and
Fasano
,
L.
(
2007
).
Tshz1 is required for axial skeleton, soft palate and middle ear development in mice
.
Dev. Biol.
308
,
407
-
420
.
Dasen
,
J. S.
and
Jessell
,
T. M.
(
2009
).
Hox networks and the origins of motor neuron diversity
.
Curr. Top. Dev. Biol.
88
,
169
-
200
.
Dasen
,
J. S.
,
De Camilli
,
A.
,
Wang
,
B.
,
Tucker
,
P. W.
and
Jessell
,
T. M.
(
2008
).
Hox repertoires for motor neuron diversity and connectivity gated by a single accessory factor, FoxP1
.
Cell
134
,
304
-
316
.
De Marco Garcia
,
N. V.
and
Jessell
,
T. M.
(
2008
).
Early motor neuron pool identity and muscle nerve trajectory defined by postmitotic restrictions in Nkx6.1 activity
.
Neuron
57
,
217
-
231
.
de Zulueta
,
P.
,
Alexandre
,
E.
,
Jacq
,
B.
and
Kerridge
,
S.
(
1994
).
Homeotic complex and teashirt genes co-operate to establish trunk segmental identities in Drosophila
.
Development
120
,
2287
-
2296
.
Dekkers
,
M. P. J.
,
Nikoletopoulou
,
V.
and
Barde
,
Y.-A.
(
2013
).
Death of developing neurons: new insights and implications for connectivity
.
J. Cell Biol.
203
,
385
-
393
.
Del Negro
,
C. A.
,
Funk
,
G. D.
and
Feldman
,
J. L.
(
2018
).
Breathing matters
.
Nat. Rev. Neurosci.
19
,
351
-
367
.
Dessaud
,
E.
,
Yang
,
L. L.
,
Hill
,
K.
,
Cox
,
B.
,
Ulloa
,
F.
,
Ribeiro
,
A.
,
Mynett
,
A.
,
Novitch
,
B. G.
and
Briscoe
,
J.
(
2007
).
Interpretation of the sonic hedgehog morphogen gradient by a temporal adaptation mechanism
.
Nature
450
,
717
-
720
.
Dewitz
,
C.
,
Pimpinella
,
S.
,
Hackel
,
P.
,
Akalin
,
A.
,
Jessell
,
T. M.
and
Zampieri
,
N.
(
2018
).
Nuclear organization in the spinal cord depends on motor neuron lamination orchestrated by catenin and Afadin function
.
Cell Rep.
22
,
1681
-
1694
.
Dike
,
C. R.
,
Bishop
,
W.
and
Pitcher
,
G.
(
2017
).
Aerophagia and abdomino-phrenic Dyssynergia in a 3-year-old
.
J. Pediatr. Gastroenterol. Nutr.
65
,
e76
-
e77
.
Dobbins
,
E. G.
and
Feldman
,
J. L.
(
1995
).
Differential innervation of protruder and retractor muscles of the tongue in rat
.
J. Comp. Neurol.
357
,
376
-
394
.
Edwards
,
B. A.
and
White
,
D. P.
(
2011
).
Control of the pharyngeal musculature during wakefulness and sleep: implications in normal controls and sleep apnea
.
Head Neck
33
Suppl. 1
,
S37
-
S45
.
Ercoli
,
P.
,
García
,
B.
,
del Campo
,
E.
,
Pinillos
,
S.
and
Javier Ercoli
,
P.
(
2018
).
CASE REPORTS CASE REPORT Aerophagia due to abdomino-phrenic dyssynergia in a 2-year-old child
.
Rev. Esp. Enferm. Dig.
110
,
332
-
333
.
Fasano
,
L.
,
Röder
,
L.
,
Coré
,
N.
,
Alexandre
,
E.
,
Vola
,
C.
,
Jacq
,
B.
and
Kerridge
,
S.
(
1991
).
The gene teashirt is required for the development of Drosophila embryonic trunk segments and encodes a protein with widely spaced zinc finger motifs
.
Cell
64
,
63
-
79
.
Fregosi
,
R. F.
(
2011
).
Respiratory related control of hypoglossal motoneurons--knowing what we do not know
.
Respir. Physiol. Neurobiol.
179
,
43
-
47
.
Fujita
,
K.
,
Yokouchi
,
K.
,
Fukuyama
,
T.
,
Fukushima
,
N.
,
Kawagishi
,
K.
and
Moriizumi
,
T.
(
2006
).
Effects of hypoglossal and facial nerve injuries on milk-suckling
.
Int. J. Dev. Neurosci.
24
,
29
-
34
.
Fukushima
,
N.
,
Yokouchi
,
K.
,
Kawagishi
,
K.
,
Kakegawa
,
A.
,
Sumitomo
,
N.
,
Karasawa
,
M.
and
Moriizumi
,
T.
(
2014
).
Quantitative analysis of survival of hypoglossal neurons in neonatally nerve-injured rats: Correlation with milk intake
.
Arch. Oral Biol.
59
,
616
-
620
.
Ghali
,
M. G. Z.
(
2015
).
Vagal modulation of pre-inspiratory activity in hypoglossal discharge in the decerebrate rat
.
Respir. Physiol. Neurobiol.
215
,
47
-
50
.
Hernandez-Miranda
,
L. R.
,
Ruffault
,
P.-L.
,
Bouvier
,
J. C.
,
Murray
,
A. J.
,
Morin-Surun
,
M.-P.
,
Zampieri
,
N.
,
Cholewa-Waclaw
,
J. B.
,
Ey
,
E.
,
Brunet
,
J.-F.
,
Champagnat
,
J.
, et al. 
(
2017
).
Genetic identification of a hindbrain nucleus essential for innate vocalization
.
Proc. Natl. Acad. Sci. USA
114
,
8095
-
8100
.
Hinckley
,
C. A.
,
Alaynick
,
W. A.
,
Gallarda
,
B. W.
,
Hayashi
,
M.
,
Hilde
,
K. L.
,
Driscoll
,
S. P.
,
Dekker
,
J. D.
,
Tucker
,
H. O.
,
Sharpee
,
T. O.
and
Pfaff
,
S. L.
(
2015
).
Spinal locomotor circuits develop using hierarchical rules based on motorneuron position and identity
.
Neuron
87
,
1008
-
1021
.
Hippenmeyer
,
S.
,
Vrieseling
,
E.
,
Sigrist
,
M.
,
Portmann
,
T.
,
Laengle
,
C.
,
Ladle
,
D. R.
and
Arber
,
S.
(
2005
).
A developmental switch in the response of DRG neurons to ETS transcription factor signaling
.
PLoS Biol.
3
,
e159
.
Hua
,
Z. L.
,
Smallwood
,
P. M.
and
Nathans
,
J.
(
2013
).
Frizzled3 controls axonal development in distinct populations of cranial and spinal motor neurons
.
eLife
2
,
e01482
.
Jessell
,
T. M.
(
2000
).
Neuronal specification in the spinal cord: inductive signals and transcriptional codes
.
Nat. Rev. Genet.
1
,
20
-
29
.
Knudson
,
C. M.
,
Tung
,
K. S. K.
,
Tourtellotte
,
W. G.
,
Brown
,
G. A. J.
and
Korsmeyer
,
S. J.
(
1995
).
Bax-deficient mice with lymphoid hyperplasia and male germ cell death
.
Science
270
,
96
-
99
.
Krammer
,
E. B.
,
Rath
,
T.
and
Lischka
,
M. F.
(
1979
).
Somatotopic organization of the hypoglossal nucleus: a HRP study in the rat
.
Brain Res.
170
,
533
-
537
.
Kuerbitz
,
J.
,
Arnett
,
M.
,
Ehrman
,
S.
,
Williams
,
M. T.
,
Vorhees
,
C. V.
,
Fisher
,
S. E.
,
Garratt
,
A. N.
,
Muglia
,
L. J.
,
Waclaw
,
R. R.
and
Campbell
,
K.
(
2018
).
Loss of intercalated cells (ITCs) in the mouse amygdala of Tshz1 mutants correlates with fear, depression, and social interaction phenotypes
.
J. Neurosci.
38
,
1160
-
1177
.
Laskowski
,
M. B.
,
Norton
,
A. S.
and
Berger
,
P. K.
(
1991
).
Branching patterns of the rat phrenic nerve during development and reinnervation
.
Exp. Neurol.
113
,
212
-
220
.
Lewis
,
P. R.
,
Flumerfelt
,
B. A.
and
Shute
,
C. C.
(
1971
).
The use of cholinesterase techniques to study topographical localization in the hypoglossal nucleus of the rat
.
J. Anat.
110
,
203
-
213
.
Lin
,
W.
,
Sanchez
,
H. B.
,
Deerinck
,
T.
,
Morris
,
J. K.
,
Ellisman
,
M.
and
Lee
,
K.-F.
(
2000
).
Aberrant development of motor axons and neuromuscular synapses in erbB2-deficient mice
.
Proc. Natl. Acad. Sci. USA
97
,
1299
-
1304
.
Lin
,
W.
,
Burgess
,
R. W.
,
Dominguez
,
B.
,
Pfaff
,
S. L.
,
Sanes
,
J. R.
and
Lee
,
K.-F.
(
2001
).
Distinct roles of nerve and muscle in postsynaptic differentiation of the neuromuscular synapse
.
Nature
410
,
1057
-
1064
.
Livet
,
J.
,
Sigrist
,
M.
,
Stroebel
,
S.
,
De Paola
,
V.
,
Price
,
S. R.
,
Henderson
,
C. E.
,
Jessell
,
T. M.
and
Arber
,
S.
(
2002
).
ETS gene Pea3 controls the central position and terminal arborization of specific motor neuron pools
.
Neuron
35
,
877
-
892
.
Lowe
,
A.
(
1980
).
The neural regulation of tongue movements
.
Prog. Neurobiol.
15
,
295
-
344
.
Machado
,
C. B.
,
Kanning
,
K. C.
,
Kreis
,
P.
,
Stevenson
,
D.
,
Crossley
,
M.
,
Nowak
,
M.
,
Iacovino
,
M.
,
Kyba
,
M.
,
Chambers
,
D.
,
Blanc
,
E.
, et al. 
(
2014
).
Reconstruction of phrenic neuron identity in embryonic stem cell-derived motor neurons
.
Development
141
,
784
-
794
.
Madisen
,
L.
,
Zwingman
,
T. A.
,
Sunkin
,
S. M.
,
Oh
,
S. W.
,
Zariwala
,
H. A.
,
Gu
,
H.
,
Ng
,
L. L.
,
Palmiter
,
R. D.
,
Hawrylycz
,
M. J.
,
Jones
,
A. R.
, et al. 
(
2010
).
A robust and high-throughput Cre reporting and characterization system for the whole mouse brain
.
Nat. Neurosci.
13
,
133
-
140
.
Manfroid
,
I.
,
Caubit
,
X.
,
Kerridge
,
S.
and
Fasano
,
L.
(
2004
).
Three putative murine Teashirt orthologues specify trunk structures in Drosophila in the same way as the Drosophila teashirt gene
.
Development
131
,
1065
-
1073
.
McClung
,
J. R.
and
Goldberg
,
S. J.
(
1999
).
Organization of motoneurons in the dorsal hypoglossal nucleus that innervate the retrusor muscles of the tongue in the rat
.
Anat. Rec.
254
,
222
-
230
.
McClung
,
J. R.
and
Goldberg
,
S. J.
(
2002
).
Organization of the hypoglossal motoneurons that innervate the horizontal and oblique components of the genioglossus muscle in the rat
.
Brain Res.
950
,
321
-
324
.
McFarland
,
D. H.
and
Lund
,
J. P.
(
1993
).
An investigation of the coupling between respiration, mastication, and swallowing in the awake rabbit
.
J. Neurophysiol.
69
,
95
-
108
.
McLoon
,
L. K.
and
Andrade
,
F.
(eds). (
2013
).
Craniofacial Muscles; Chapter 12
.
New York, NY
:
Springer New York
.
Moore
,
J. D.
,
Kleinfeld
,
D.
and
Wang
,
F.
(
2014
).
How the brainstem controls orofacial behaviors comprised of rhythmic actions
.
Trends Neurosci.
37
,
370
-
380
.
Morita
,
H.
,
Mazerbourg
,
S.
,
Bouley
,
D. M.
,
Luo
,
C.-W.
,
Kawamura
,
K.
,
Kuwabara
,
Y.
,
Baribault
,
H.
,
Tian
,
H.
and
Hsueh
,
A. J. W.
(
2004
).
Neonatal lethality of LGR5 null mice is associated with ankyloglossia and gastrointestinal distension
.
Mol. Cell. Biol.
24
,
9736
-
9743
.
Odutola
,
A. B.
(
1976
).
Cell grouping and Golgi architecture of the hypoglossal nucleus of the rat
.
Exp. Neurol.
52
,
356
-
371
.
Paydarfar
,
D.
,
Gilbert
,
R. J.
,
Poppel
,
C. S.
and
Nassab
,
P. F.
(
1995
).
Respiratory phase resetting and airflow changes induced by swallowing in humans
.
J. Physiol.
483
,
273
-
288
.
Pfaff
,
S. L.
,
Mendelsohn
,
M.
,
Stewart
,
C. L.
,
Edlund
,
T.
and
Jessell
,
T. M.
(
1996
).
Requirement for LIM homeobox gene Isl1 in motor neuron generation reveals a motor neuron-dependent step in interneuron differentiation
.
Cell
84
,
309
-
320
.
Philippidou
,
P.
,
Walsh
,
C. M.
,
Aubin
,
J.
,
Jeannotte
,
L.
and
Dasen
,
J. S.
(
2012
).
Sustained Hox5 gene activity is required for respiratory motor neuron development
.
Nat. Neurosci.
15
,
1636
-
1644
.
Ragancokova
,
D.
,
Rocca
,
E.
,
Oonk
,
A. M. M.
,
Schulz
,
H.
,
Rohde
,
E.
,
Bednarsch
,
J.
,
Feenstra
,
I.
,
Pennings
,
R. J. E.
,
Wende
,
H.
and
Garratt
,
A. N.
(
2014
).
TSHZ1-dependent gene regulation is essential for olfactory bulb development and olfaction
.
J. Clin. Invest.
124
,
1214
-
1227
.
Riethmacher
,
D.
,
Sonnenberg-Riethmacher
,
E.
,
Brinkmann
,
V.
,
Yamaai
,
T.
,
Lewin
,
G. R.
and
Birchmeier
,
C.
(
1997
).
Severe neuropathies in mice with targeted mutations in the ErbB3 receptor
.
Nature
389
,
725
-
730
.
Röder
,
L.
,
Vola
,
C.
and
Kerridge
,
S.
(
1992
).
The role of the teashirt gene in trunk segmental identity in Drosophila
.
Development
115
,
1017
-
1033
.
Ruffault
,
P.-L.
,
D'Autréaux
,
F.
,
Hayes
,
J. A.
,
Nomaksteinsky
,
M.
,
Autran
,
S.
,
Fujiyama
,
T.
,
Hoshino
,
M.
,
Hägglund
,
M.
,
Kiehn
,
O.
,
Brunet
,
J.-F.
, et al. 
(
2015
).
The retrotrapezoid nucleus neurons expressing Atoh1 and Phox2b are essential for the respiratory response to CO2
.
eLife
4
,
e07051
.
Santos
,
J. S.
,
Fonseca
,
N. A.
,
Vieira
,
C. P.
,
Vieira
,
J.
and
Casares
,
F.
(
2010
).
Phylogeny of the teashirt-related zinc finger (tshz) gene family and analysis of the developmental expression of tshz2 and tshz3b in the zebrafish
.
Dev. Dyn.
239
,
1010
-
1018
.
Shamblott
,
M. J.
,
Bugg
,
E. M.
,
Lawler
,
A. M.
and
Gearhart
,
J. D.
(
2002
).
Craniofacial abnormalities resulting from targeted disruption of the murineSim2 gene
.
Dev. Dyn.
224
,
373
-
380
.
Song
,
J.
,
Ampatzis
,
K.
,
Björnfors
,
E. R.
and
El Manira
,
A.
(
2016
).
Motor neurons control locomotor circuit function retrogradely via gap junctions
.
Nature
529
,
399
-
402
.
Tanaka
,
T.
,
Wakabayashi
,
T.
,
Oizumi
,
H.
,
Nishio
,
S.
,
Sato
,
T.
,
Harada
,
A.
,
Fujii
,
D.
,
Matsuo
,
Y.
,
Hashimoto
,
T.
and
Iwatsubo
,
T.
(
2014
).
CLAC-P/collagen type XXV is required for the intramuscular innervation of motoneurons during neuromuscular development
.
J. Neurosci.
34
,
1370
-
1379
.
Turgeon
,
B.
and
Meloche
,
S.
(
2009
).
Interpreting Neonatal Lethal Phenotypes in Mouse Mutants: Insights Into Gene Function and Human Diseases
.
Physiol. Rev.
89
,
1
-
26
.
Villoria
,
A.
,
Azpiroz
,
F.
,
Burri
,
E.
,
Cisternas
,
D.
,
Soldevilla
,
A.
and
Malagelada
,
J.-R.
(
2011
).
Abdomino-phrenic Dyssynergia in patients with abdominal bloating and distension
.
Am. J. Gastroenterol.
106
,
815
-
819
.
White
,
F. A.
,
Keller-Peck
,
C. R.
,
Knudson
,
C. M.
,
Korsmeyer
,
S. J.
and
Snider
,
W. D.
(
1998
).
Widespread elimination of naturally occurring neuronal death in Bax-deficient mice
.
J. Neurosci.
18
,
1428
-
1439
.
Wichterle
,
H.
,
Lieberam
,
I.
,
Porter
,
J. A.
and
Jessell
,
T. M.
(
2002
).
Directed differentiation of embryonic stem cells into motor neurons
.
Cell
110
,
385
-
397
.
Wildner
,
H.
,
Das Gupta
,
R.
,
Brohl
,
D.
,
Heppenstall
,
P. A.
,
Zeilhofer
,
H. U.
and
Birchmeier
,
C.
(
2013
).
Genome-wide expression analysis of Ptf1a- and Ascl1-deficient mice reveals new markers for distinct dorsal horn interneuron populations contributing to nociceptive reflex plasticity
.
J. Neurosci.
33
,
7299
-
7307
.
Wu
,
J.
,
Capelli
,
P.
,
Bouvier
,
J.
,
Goulding
,
M.
,
Arber
,
S.
and
Fortin
,
G.
(
2017
).
A V0 core neuronal circuit for inspiration
.
Nat. Commun.
8
,
544
.
Yamamoto
,
Y.
and
Henderson
,
C. E.
(
1999
).
Patterns of programmed cell death in populations of developing spinal motoneurons in chicken, mouse, and rat
.
Dev. Biol.
214
,
60
-
71
.
Yang
,
X.
,
Arber
,
S.
,
William
,
C.
,
Li
,
L.
,
Tanabe
,
Y.
,
Jessell
,
T. M.
,
Birchmeier
,
C.
and
Burden
,
S. J.
(
2001
).
Patterning of muscle acetylcholine receptor gene expression in the absence of motor innervation
.
Neuron
30
,
399
-
410
.

Competing interests

The authors declare no competing or financial interests.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.

Supplementary information